The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis 9781407314310, 9781407344119

Table of contents :
Front Cover
Title Page
Copyright
Table of Contents
Abstract
List of Figures
List of Maps
List of Tables
Introduction
Chapter 1: The Roman Settlement Analysis
Chapter 2: The Northwest Integration within the Roman World
Chapter 3: An Overview on the Entre-Douro-E-Minho Roman Settlement Patterns
Chapter 4: The Shaping of Bracara Augusta Peripheries
Chapter 5: The Shaping of Mountain, Valley and Coastline Settlement
Conclusion
Catalogue
Appendix I- Maps
References

Citation preview

________ Helena Paula Abreu de Carvalho is a Professor of Classical History and Archaeology at the Institute of Social Sciences of the University of Minho, Portugal and a research member of Lab2Pt – Landscape, Heritage and Territory Laboratory. She has been a visiting professor at several universities in Europe (University Blaise Pascal-Clermont Ferrand, University of Bologna, University of La Coruña, University of Vigo, Institut Català d’Arqueologia Clàssica) and also in Brazil (University of Vale de São Francisco, University of S. Paulo, Federal University of Espírito Santo). Her research interests include Roman and landscape archaeology.

BAR S2789 2016  CARVALHO  ROMAN SETTLEMENT PATTERNS IN THE CONVENTUS BRACARENSIS

The main aim of this study was to analyse the Roman settlement patterns of the western part of the Conventus Bracarensis, located within a vast province of the Iberian Peninsula: the Hispania Citerior or Tarraconensis. In order to fully understand the Roman settlement patterns of the area under study, it was necessary to verify the political and administrative procedures carried out at the Tarraconensis province together with their direct impact upon the landscape, alongside the varied occupation techniques implemented after the conquest, assessing the different patterns impacting upon the indigenous space. The nature of the available data was analysed through G.I.S. and four operative concepts: entities, relationships between entities, scales and limits to those relationships. Using these elements, the interpretation and description of the studied centres aimed to shed some light on the basic evolutionary features of settlement patterns. Through their complex interactions, the analysed areas provided the minimum conditions with which to achieve the proposed objectives.

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis Helena Paula Abreu de Carvalho

BAR International Series 2789 9 781407 314310

B A R

2016

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis Helena Paula Abreu de Carvalho

BAR International Series 2789 2016

First Published in 2016 by British Archaeological Reports Ltd United Kingdom BAR International Series 2789 The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Copyright © Helena Paula Abreu de Carvalho 2016 The Author’s moral rights under the 1988 UK Copyright, Designs and Patents Act, are hereby expressly asserted All rights reserved. No part of this work may be copied, reproduced, stored, sold, distributed, scanned, saved in any form of digital format or transmitted in any form digitally, without the written permission of the Publisher.

ISBN 9781407314310 paperback ISBN 9781407344119 e-format DOI https://doi.org/10.30861/9781407314310 A catalogue record for this book is available from the British Library

All BAR titles are available from: British Archaeological Reports Ltd Oxford United Kingdom Phone +44 (0)1865 310431 Fax +44 (0)1865 316916 Email: [email protected] www.barpublishing.com

Contents Abstract ..............................................................................................................................................vii List of Figures .................................................................................................................................. viii List of Maps ........................................................................................................................................ix List of Tables ....................................................................................................................................... x Introduction .......................................................................................................................................... 1 Chapter 1 .............................................................................................................................................. 3 The Roman Settlement Analysis .......................................................................................................... 3 Introduction .......................................................................................................................................... 3 1 Analysis axes..................................................................................................................................... 3 1.1 Dominant records and challenges .................................................................................................. 3 1.2 Reference matrix, transection and frameworks.............................................................................. 7 2 Data asymmetries and options .......................................................................................................... 8 2.1 Available data ................................................................................................................................ 8 2.2 Data variability............................................................................................................................... 9 2.3 Conducted and systematic field walking surveys ........................................................................ 10 2.4 Adopted nomenclatures................................................................................................................ 12 3. The methodological infrastructure: the GISARM ......................................................................... 15 3.1 The alphanumeric and cartographic Databases ............................................................................ 15 3.2 The Geographic Information System: opportunities and constraints........................................... 16 4. Operational concepts ...................................................................................................................... 16 Chapter 2 ............................................................................................................................................ 19 The Northwest Integration within the Roman World ........................................................................ 19 Introduction ........................................................................................................................................ 19 1. Pre-Roman settlement .................................................................................................................... 19 1.1 Bronze Age and the transition to the Iron Age ............................................................................ 19 1.2. Communities diversity throughout the first millennium BC ...................................................... 21 1.3 The Northwest in the late first millennium BC ............................................................................ 23 1.4. Roman expeditions in northern Douro ........................................................................................ 24 2. Roman conquest and pacification ................................................................................................. 25 2.1. The campaigns of Augusto in the Northwest .............................................................................. 26 2.2. Organization of the territory ....................................................................................................... 26 2.2.1 A new social geography: populi and civitates .......................................................................... 26 2.2.1.1 Textual sources ...................................................................................................................... 26 2.2.1.2 Epigraphic data ...................................................................................................................... 27 2.2.1.3 Populi and civitates ................................................................................................................ 28 iii

2.2.2 Power centres: the urban politics of Augusto ........................................................................... 30 2.2.3 Communications network .......................................................................................................... 34 2.2.3.1 Land itineraries ...................................................................................................................... 34 2.2.3.2 River and sea itineraries ......................................................................................................... 36 3. The consolidation of the Roman power ......................................................................................... 37 3.1 The development and hierarchy of the urban centres .................................................................. 37 3.2 Reinforcement of the road network system ................................................................................. 44 3.3 Economic intensification.............................................................................................................. 46 3.3.1 Exploitation of resources .......................................................................................................... 46 3.3.2 Market economy and currency circulation ............................................................................... 49 Chapter 3 ............................................................................................................................................ 52 An Overview on the Entre-Douro-E-Minho Roman Settlement Patterns.......................................... 52 Introduction ........................................................................................................................................ 52 1. The fortified settlement .................................................................................................................. 52 1.1 Analysis categories and spatial distribution ................................................................................. 52 1.2 Chronological approaches ............................................................................................................ 59 1.3 Distribution areas ......................................................................................................................... 62 1.3.1 The area between the Lima and Ave rivers ............................................................................... 63 1.3.2 The mountainous area ............................................................................................................... 63 1.3.3 The area between the Sousa and Tâmega rivers....................................................................... 64 2. The road network: trajectories and road stations ........................................................................... 65 3. Central sites and new settlements .................................................................................................. 69 4. Villas: identification challenges ..................................................................................................... 78 Chapter 4 ............................................................................................................................................ 83 The Shaping of Bracara Augusta....................................................................................................... 83 Peripheries .......................................................................................................................................... 83 Introduction ........................................................................................................................................ 83 1 Analysis frameworks....................................................................................................................... 83 2. Entities and relationships ............................................................................................................... 84 2.1 The road network system and the necropolis ............................................................................... 84 2.1.1 The main itineraries .................................................................................................................. 84 2.1.2 The city necropolis and associated epigraphy .......................................................................... 86 2.2 The ancient circuits ...................................................................................................................... 88 2.2.1 The hill-forts hierarchy ............................................................................................................. 88 2.2.2 Maintenance, retraction and abandonment .............................................................................. 91 2.3. Roman settlements ...................................................................................................................... 92 2.3.1 Vici and road stations ............................................................................................................... 92 2.3.2 Villas and associated epigraphy ............................................................................................... 94 iv

2.3.3 Necropolis and associated epigraphy ....................................................................................... 98 2.3.4 Agricultural and artisanal exploitation units............................................................................ 99 3 Peripheral limits and scales ........................................................................................................... 100 3.1 Settlement concentration and dispersion areas .......................................................................... 101 3.2 Rural land division ..................................................................................................................... 103 3.2.1 Analysis methodologies ........................................................................................................... 104 3.2.2 The area around the XIX itinerary, between Braga and the Cávado River............................ 105 3.2.3 Termini and limiting devices ................................................................................................... 109 3.2.4 The characterization of the cadastre ...................................................................................... 113 Chapter 5 .......................................................................................................................................... 116 The Shaping of Mountain, Valley and Coastline Settlement ........................................................... 116 Introduction ...................................................................................................................................... 116 1. The spine settlement throughout the Via Nova trajectory ........................................................... 116 1.1 Entities and relationships ........................................................................................................... 116 1.1.1 The XVIII itinerary or Via Nova: trajectory and associated epigraphy ................................. 116 1.1.2 The fortified settlement............................................................................................................ 119 1.1.3 The ‘spine’ settlement: vici, mansiones and mutationes ......................................................... 122 1.2 Limits and scales ........................................................................................................................ 125 1.2.1 The road as an organizational element of a ‘spine’ settlement network ................................ 125 1.2.2 The Romanization of a mountainous territory ........................................................................ 126 2. Valley and coastline settlement.................................................................................................... 128 2.1 Analysis frameworks.................................................................................................................. 129 2.2 Entities and relationships ........................................................................................................... 130 2.2.1 The fortified settlement............................................................................................................ 130 2.2.2 Transport network and associated settlements ....................................................................... 136 2.2.3 The exploitation of resources .................................................................................................. 138 2.2.3.1 Villas and associated units ................................................................................................... 138 2.2.3.2 Small and medium size Farms and artisanl units ................................................................. 141 2.3 Limits and scales ........................................................................................................................ 142 2.3.1 Diffusion areas and influencing boundaries ........................................................................... 142 2.3.2 The Romanization of the valley landscape and coastline façade............................................ 144 Conclusion ....................................................................................................................................... 147 Catalogue ......................................................................................................................................... 152 Appendix I- Maps ............................................................................................................................ 210 References ........................................................................................................................................ 225

v

‘’Inexact words, gaps, lost words, recovered, impossible words, careful words, ‘dictionary’ words, place us on guard; there are realms in which the read words haven’t been written. The silence is also a word’’. Giorgio Manganelli

vi

Abstract The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis This work aimed to characterise the settlement patterns in the Entre-Douro-and-Minho region, beginning with a large and synthetic analysis of the changes launched by the Roman conquest in the Northwest Iberia. The foundation of Bracara Augusta and the development of the main road network led to a redefinition of the settlement structuring axes, among which are the creation of secondary settlements within the area. The archaeological data at our disposal allowed us to locate some small towns, which appear closely connected with the main road network. Among the very considerable archaeological data at our disposal indicating Roman sites we would like to emphasise the presence of Roman villas. Finally, the continuity of some Iron Age settlements in the Roman period seemed to be linked with the control of the landscape, the road network and the valleys. The archaeological remains of the region, comprising more than one thousand and two hundred sites together with the diversity of settlement and the variety of studies produced and published, have required different scales of analysis. We decided to start by establishing a transection cutting the region. This transverse section obeyed to the following three criteria: geomorphic diversity, archaeological data and, of course, the presence of Roman city of Bracara Augusta. In this transect we analysed three areas with greater detail: the region around Bracara Augusta; the mountainous surrounding region boosting settlement patterns organised by the Flavian Roman road system and, finally, the valley and coastline areas. In the mountainous region the key features were the gradual rarefaction of settlements as we move towards higher altitudes and the implementation of sites directly connected with the roman roads. There were also distinctive features between the settlement patterns in the valley and near the coast. Density and dispersal have been the features of settlement in the valley; the coastal area was scarcely occupied with some settlements from the Iron Age and some villas, apparently associated with salt-making or fish products. Finally, we can identify a density of roman occupation patterns around Bracara Augusta. While analysing this region, we studied the organization of the rural landscape in order to verify a ‘centuriated’ area. We have confirmed this idea. In an area of 10km around Bracara Augusta, an area framed by the highest hills in the region, we found traces of a cadastre that organised itself in modules of 20 by 20 actus.

vii

List of Figures Figure 1. Map of Hispania representing the division of existing provinces and conventus (provided by M. Martins).... 19 Figure 2. Boundaries of the three northwest conventus and their respective capitals (provided by M. Martins)............. 23 Figure 3. Simplified Roman road network system in Hispania, with particular emphasis towards the connection with Bracara Augusta (Morais 2005: 66). ...................................................................................................................... 35 Figure 4. Location of the Entre-Douro-e-Minho area in Roman Hispania (elaborated from http: //commons.wikimédia,org/wiki/category: Maps-of-Roman-Hispania) ................................................................... 52 Figure 5. Bouça da Mó mutatio plan (Lemos and Baptista 1995-1996). ......................................................................... 69 Figure 6. Bouça da Mó interpreted plan, according to M. Martins .................................................................................. 69 Figure 7. S. Vicente do Pinheiro thermal plan according to Jorge Alarcão and based upon works from Teresa Soeiro (Alarcão 1988 1/473, fig. 103; Soeiro 1984). .......................................................................................................... 76 Figure 8. The Via Cova plan, according to M. Martins ................................................................................................... 81 Figure 9. Bracara Augusta: plan of the Roman town, according to M. Martins and M.C. Ribeiro ................................. 84 Figure 10. Votive altar found in the outskirts of Bracara Augusta (Carvalho et alii 2006) .......................................... 102 Figure 11. a, b, c Perspective of cippus 1, in 2004 ........................................................................................................ 110 Figure 12. Perspective of cippus 1, in 2006 ................................................................................................................... 110 Figure 13. a, b Survey perspective of cippus 1, in 2006 ................................................................................................ 111 Figure 14. Survey plans of cippus 1, in 2006 ................................................................................................................. 111 Figure 15. Perspective of cippus 2, in 2004 ................................................................................................................... 111 Figure 16. Location of the Cippi (1,2) and of the archa petrinea (3) ............................................................................ 112 Figure 17. Votive altar dedicated to Mars...................................................................................................................... 112 Figure 18. Roman Cadaster............................................................................................................................................ 115 Figure 19. Genius Viriocelensis pedestal (Carvalho 1998) ............................................................................................ 123 Figure 20. Via XVIII with some milestones in situ ....................................................................................................... 128

viii

List of Maps Map 1. Entre-Douro-e-Minho region and delimitation of selected areas ....................................................................... 210 Map 2. Hillforts in Entre-Douro-e-Minho region .......................................................................................................... 211 Map 3. Hillforts of A category ....................................................................................................................................... 212 Map 4. Hillforts of B category ....................................................................................................................................... 213 Map 5. Hillforts of C category ....................................................................................................................................... 214 Map 6. Iron Age hillforts (categories A, B and C) ......................................................................................................... 215 Map 7. Road network system, vici and road stations ..................................................................................................... 216 Map 8. Road network system and vici ........................................................................................................................... 217 Map 9. Villae.................................................................................................................................................................. 218 Map 10. Bracara Augusta peripheries: archeological sites and roman roads ................................................................ 219 Map 11. Main influence area around Bracara Augusta ................................................................................................. 220 Map 12. Roman cadastre, roman roads and archeological sites ..................................................................................... 221 Map 13. Roman cadastre with “Couto de Tibães” delimitation ..................................................................................... 222 Map 14. Framework of the mountain region .................................................................................................................. 223 Map 15. Valley and coastline settlement ....................................................................................................................... 224

ix

List of Tables Table 1. Hill-forts of A category ...................................................................................................................................... 55 Table 2. Hill-forts of B category ...................................................................................................................................... 57 Table 3. Hill- forts of C category ..................................................................................................................................... 58 Table 4. Road stations ...................................................................................................................................................... 68 Table 5. Entre-Douro-e-Minho vici.................................................................................................................................. 73 Table 6. Entre-Douro-e-Minho Villas .............................................................................................................................. 80 Table 7. Deities associated with the road network system around Bracara Augusta. ...................................................... 87 Table 8. Distances between Villas and Bracara Augusta and the nearby roads (in meters) ............................................ 97 Table 9. Agricultural exploitation units ......................................................................................................................... 100 Table 10. Occupation, abandonment and reoccupation of ............................................................................................. 135

x

the establishment and development of new organizational patterns by the Romans.

Introduction

The difference and reciprocity of both processes – namely in terms of when and where the indigenous ended and the Roman started – generated a network of constraints and challenges. On the one hand, constraints arouse from our attempt to control vast and different archaeological data that in turn had multiple credibility levels. Furthermore, challenges emerged due to the fact that we tried to provide our research with a strong recovery investment, as well as, to homogenize our archaeological data so that it prompted a territorial sustained cartographic analysis. This aim was only achieved through a continuous georeference of archaeological areas. Thus, two guidelines also influenced our research. First, the expectations generated by the regional study of the Roman settlement, initiated in Portugal in the 1980s, had critical on-going results and secondly, the challenges included in methodological analysis introduced by Landscape Archaeology. The relationship between the landscape and settlement patterns is inherent to the emergence of gradual questions and to the establishment of different analysis exercises that will be developed throughout this work. Such exercises started with the establishment of a Geographic Information System, conceived to process, map and interpret available data related with the Entre-Douro-e-Minho region. By available information we understand not only the Roman archaeological remains but also all the Iron Age settlements regardless of the fact whether they have been “Romanized” or not.

The main aim of this study is to analyse the Roman settlement patterns of the western part of the Conventus Bracarensis, located within a vast province of the Iberia Peninsula: the Hispania Citerior or Tarraconensis. The choice over both the studied region and the research procedures to apply has been the end result of a varied group of aims, data, challenges and analysis that will be presented in this introductory section. To contextualise the problems within our research, we necessarily had to start by considering the wider Romanised framework of the peninsular Northwest. This was believed to be a valid approach in understanding the diversity of the pre-Roman world and the different conquest strategies led by Rome over the indigenous population. Adding to this, and in order to fully understand the Roman settlement patterns of the area under study, we also had to verify the political and administrative procedures carried out at the Tarraconensis province together with their direct impact upon the landscape, alongside the varied occupation techniques implemented after the conquest. Furthermore, it became impossible to detect the organizational space models without fully understanding the historical scale of the settlement challenges. Scales in this context must be understood as a back-and-forth of spatial and temporal plans, often discontinuous, sometimes of wide range, either further confined, where sites are fully circumscribed. We understand that without these scales the archaeological and historical data would greatly lose credibility. Adding to this, the imperial policies underlying the Roman conquest of Hispania were developed after assessing the different patterns impacting upon the indigenous space. Specifically, the regional territorial organization and the development practices were analysed together with the strands sustaining both the urban and economic development. Hence, we aimed at shedding some light on the territory geographical features as these are believed to have generated different entities and relationships, which in turn have conditioned the traits of the human occupation and what has been powered as a global dynamic result of the conquest constraints and the new structuring policies for the territory. Under these circumstances, we tried to analyse a specific region landscape evolution process, although a region wide enough to allow the research of communities and landscape reorganizational models.

It is important to emphasise that the physical boundaries of the Entre-Douro-e-Minho region do not overlap with any previous administrative frontiers of Hispania. Under these circumstances, the studied physical limits were not determined by the Roman political geography or by other specific feature of the region. The chosen area was a pragmatic choice related with the coherence and handling of available data regardless of the fact that the selected region included features allowing its study as a geographic and geomorphologic unit within a wider scale. After determining the area to be studied and analysing all the available data regarding the Romanization of the region, being the latter a task which was crucial in the homogenization of available sources and in the establishment of a hierarchy within the population, we produced substantial, vast and heterogeneous data. The sites catalogue and the synthesis cartography represent the achievement of these tasks, while, simultaneously, providing an advanced base of interactive support of future analysis.

We progressively adjusted the correlation between a dilated space and a reference array, namely between the peninsular Northwest and the western façade of the Conventus Bracarensis, located within the present day Portuguese region of Entre-Douro-e-Minho. We attempted to value a group of core problems inherent to the study of the Roman conquest of new territories. The preliminary enquiry aimed to explore two main themes: the evolution of the indigenous population under Roman domain and

The results led us into other spatial boundaries and made us consider a strategic analysis area that prompted a thinner layered scale analysis and in turn produced further challenges and observational horizons. This selection considered issues of geomorphological nature and, at the same time, socio-political and economic issues, which impacted upon the development of both the landscape and the population.

1

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

The strategic area under analysis corresponded to a transversal part of the region, shaped by a transection to which Bracara Augusta has been the core centre. This selection was based upon three criteria: the stability and quantity of available data, the region boundaries shaped in as a natural amphitheatre running from the mountains to the coastline and the impact of Bracara Augusta upon the surrounding region. Adding to this, we defined three distinct morphological criteria: the region where the Roman city was established, the mountainous area and the area integrating both the medium and low hydrographical watersheds of the Cávado and Ave rivers, corresponding to the coastline and valley areas.

searched data, we believed it was important to present a synthetic overview of the main coordinates shaping the region in both its landscape archaeology and Roman occupation. We established, within the possibilities, the occupational area and the complex transition from preRoman to Roman settlements. And from this analysis exercise should emerge realistic frameworks to approach the aims mentioned above. In chapter 4 the study moves towards the outskirts of Bracara Augusta. We emphasise that both the direct and indirect impact of the city in its surrounding area has always been the core aim of this work, but this intention has successively been broadened into other approaches. This has been related with the fact that Bracara Agusta detained a significant physical and urban dimension; it was a conventus capital and generated a gradual interest amongst researchers.

We are aware that choosing one or more sampling areas when assessing settlements boundaries may either risk cancelling or aggravate the impact of their specificity in terms of wider contexts. Defining one or more study areas may also encompass the illusion of establishing ‘natural’ observational units. Nevertheless, we cannot escape these challenges and we need to further emphasise the importance of observational scales.

What motivated us forward was the possibility to understand how the surrounding area of the city responded to its foundation and how the road network system impacted upon the change or implementation of the outskirts. The possibility of a potential cadastral land organization within the surrounding areas constituted a main interest point within this study. Following on the previous chapters it is not too risky to state that perhaps the establishment of connections between the several peripheral areas could perform the end to this particular chapter.

The nature of the available data was analysed through Geographical Information Systems and followed four operative concepts: entities, relationships between entities, scales and limits to those relationships. When organised in accordance with these four previous elements, the interpretation and description of the studied centres aimed to shed some light on the basic evolutionary features of settlement patterns. Thereby we did not attempt to interpret our findings in a predetermined and rigid manner. Consequently, we tried to map and correlate the entities – sites and remains – that could only be achieved with the understanding of their contexts together with their acting scales and territorial boundaries. The observation technique catered for a better rigor mainly when filtering the material registrations of fragmented and scattered data.

In the last chapter (5) we assessed settlement models for both the mountain areas crossed by the Via XVIII itinerary and the valley and coastline areas. These performed regions where Bracara Augusta had a significant impact; however we do not wanted to divide the territory when it came to the impact of the city within the experimenting and shaping models of organizing the human landscape. We merely attempted to understand the different features within the area.

In chapter 1 we will focuse on analysing the dominant interpretation of archaeological and Ancient History record procedures related with the region under study, in a wider framework of presenting available data of the system built by us and of the concepts mentioned above. We also will attempt to include some approaches on typologies and nomenclatures of the archaeological sites.

In the conclusion part, we interrogated ourselves about the perspectives, the limits and the potential that a work focused on a cataloguing framework and georeferenced norms of archaeological data may offer the development of a research agenda that is collectively more efficient. All the mentioned aims probably depended on a critical tension inherent to the work. The will to synthetize the archaeological data within a more global and integrated realm and, at the same time, analyse, on a different level, the specific events in which man structured the territory. The analysed areas, due to their complex interaction provided us with the minimum conditions to achieve our proposed objectives.

In chapter 2 a synthesis of the main results with regards to the Romanization of the peninsular Northwest will be developed. The analysis included the Bronze Age in an attempt to differentiate the varied rhythms and patterns related with the development of the communities and habitats. We have provided special attention to the Romanizing programs for the Northwest, specially the ones related with the establishment of urban areas and the development of their surrounding regions.

The catalogue and cartography, perform an organic complement to this study.

With these two frameworks in mind, one related with the methodological analysis of the Roman settlements (chapter 1) and the other focused on the integration of the Northwest within the Roman world (chapter 2) we focused our study on the Entre-Douro-e-Minho region (chapter 3). Despite the challenges of producing a generalist level and the possibility to correlate previously re-

2

Chapter 1 1 Analysis axes

The Roman Settlement Analysis

1.1 Dominant records and challenges Introduction Defining the challenges underlying this work invited us to track down archaeological research rhythms and paradigms conducted in the region under study. The heterogeneous nature of archaeographical data gathered on the site catalogue on which this work was based clearly reflects the existence of different criteria, exploitation methods and data interpretation.

We start by approaching the main challenges faced by researchers with regards to the study of Roman settlements in the peninsular northwest. At the same time, we will attempt to highlight the paradigmatic records that shaped and contributed towards the recollection, prospection, excavation and treatment of all archaeological data used.

The framing of these ‘’methodological’’ trends may be broken down into two key challenges mobilizing and structuring the peninsular northwest research tradition: the ’castro culture’ and the Romanization process.

We were mainly interested in framing historiographical prejudices underlying the peripheral and shallow Romanised nature of the region when compared to other southern regions in the peninsula shaped by civilizational inflows from the Mediterranean. Adding to this theoretical record, almost ideological, exists a geographical record that conceived the peninsular northwest as an homogeneous entity. Both the theoretical and geographical records profiled a research tradition that led towards a systematic approach of settlements as analysis entities consequently originating the Roman remains to be side lined. In this context, the available timeline record needed to establish a valid chronology for the Romanization process, as it was rather fragmented and widespread.

The inherent challenges present in the ‘’castro culture’’, in the sense of defining its concept and its integration within wider historic-archaeological studies, was the main analysis aim of several authors (Martins, 1990; González Ruibal, 2006-2007), who developed a framework of their conducted research and critically and clearly included it within the wider archaeological research of the peninsular northwest. Both the scholar and archaeological studies carried out between the end of the nineteenth century and the first half of the twentieth century in northern Portugal were central in identifying and overestimating a cultural indigenous universe, found in the remains and in several settlements excavations. The results were synthetized in the so-called ‘’castros culture’’, first put forward by P. Bosch-Gimpera (1921), who included it within the second peninsular Iron Age and framed it in terms of uncovered settlements and houses layout.

Point 1 will focus on the geographic options of our research, from the reference array to the definition of a transection crossing the region that will then provide us access to different data exploitation areas. Under these circumstances, we hope to overcome both the geographical assumption of a homogeneous Northwest and the solely concentration in a circumscribed area, also allegedly homogeneous.

Martins Sarmento brilliantly shaped the beginning of this type of research. Other than his numerous site inventories and material collections, the researcher started the excavation work both in Citânia de Briteiros and Castro Sabroso at the end of the nineteenth century. His findings, mainly the ones related with the Citânia of Briteiros, are of paramount importance both at a regional and national archaeological level (Lemos 1985; Martins 1995; Pimenta 2007).

Point 2 is entirely focused on the managed archaeological data. We must, however, highlight a structuring point: the interpretation of new habitats featuring the Roman occupation is difficult to acknowledge when based upon the fragmented nature of the available data. Therefore, and despite the high number of analysed sites, few have been excavated not allowing us to rigorously assess them. Even when they have been subjected to excavations they rarely provided the necessary elements to perform accurate assessments. We end this point with an approach of the conducted field surveys and a discussion of some concepts and of Roman archaeological sites nomenclatures.

The death of Martins Sarmento performed a clear fracture within northern Portuguese Hill-forts research, and as stated by A. González Ruibal, 2006-2007 ’since Sarmento’s death and up to the 1930s when Mario Cardozo comes into the scene, the archaeological research in northern Portugal could be compared to the Iron Age’’ (2006-2007:33). The paradigmatic case of the Citânia of Briteiros illustrated these events. Here, the excavations only continued nearly three decades after the death of researcher.

In point 3 we describe the Geographical Information System infrastructure adopted by the present work. The last point (4) will attempt to define the use and definition of four operative concepts: entities, relationships, boundaries and scales.

The same trends can be observed in Galicia at the beginning of the twentieth century, where the growth of archaeological research and inventory focused on Hill-forts took place with regards to inventorying and excavating 3

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis sites. To consolidate this research strategy, the ‘Grup Nós’ and Galegos Study Seminar were created in the 1920s and shaped by figures as Cabré Aguilló, Frederico Macinera, Bouza Brey and later on Garcia and Bellido (Martins 1990).

the ’time’ was hidden in the absence of accurate chronological data. A direct consequence of such trends ended up being the integration within the ‘castros culture’ of artefactual indicators not related with the proposed chronology, both by P. Bosh-Gimpera and Santa Ollala, who clearly included them within the II Iron Age. Thus, a generalist trend of considering ‘castrejo’ everything found in settlements emerged and disregarded a needed chronological assessment.

However, the peninsular northwest archaeology focused on the study of Hill-forts has mainly been focused on the works of F. Lopez Cuevillas, who boosted the research agenda between the 1920s and 1960s by producing extensive monographic and synthesis works over the ‘castro culture’. It was this particular researcher who gradually synthetized and mapped the settlements geographical dispersion (Lopez Cuevillas 1933), the architecture and habitats (Lopez Cuevillas 1946-1947) and several other aspects of the existing material culture, such as the weapons (Lopez Cuevillas 1946-1947), the ceramics (Lopez Cuevillas 1953; 1958) or the sculpture (Lopez Cuevillas 1951) amongst others.

The anachronism of the ‘castreja culture’ is then related with the lack of stratigraphic excavations and knowledge over the Bronze Age archaeological artefacts, but also of Roman and Medieval origin. These trends shaped the past of the peninsular northwest as purely ‘castrejo’, following on the path of Martins Sarmento, Garcia Bellido and F. Lopez Cuevillas. These events were also due to the great relevance given to the recognition and acknowledgement of new Hill-forts up to the 1970s.

One may consider that Portugal entered a new settlement research stage in the decades following the 1930s. This research stage was mainly shaped by the work of Mário Cardozo, who roughly systematized both the material and ‘artefactual’ culture (Cardozo 1929; 1932; 1934; 1937; 1938; 1942; 1944; 1946a; 1946b; 1952, 1959; 1962; 1965; 1966). Less generalist were the works of Félix Alves Pereira (1933; 1938-1941) who disclosed most of Briteiros artefacts, of Abel Viana (1926; 1930) who focused on inventorying and cataloguing sites, an approach also adopted by Carlos Teixeira (1963) in his written works focused on the Braga region, Irisalva Moita (1962; 1971) in her northern Portugal research, and Fransisco Manuel Alves (1934 – 38) for the Trás-os-Montes region.

The concept of ‘’castreja culture’’ adopted by researchers started to guide the identification and study of several settlements scattered around a wide conceptualized region that was also culturally convergent. From here onwards, Hill-forts started to be excavated with the aim of establishing a homogenous framework of that particular cultural area and determining its chronology (Martins 1990). Questions related with Hill-forts origin and terminus performed key challenges within the conducted studies, while at the same time, researchers tried to identify elements related with the type of habitat, ceramics, jewellery and statues, assumed as Hill-forts fossils.

The notorious accumulation of data emerging from inventories and excavations was conducted without any stratigraphic concern between the 1920s and the 1950s. However, it helped to specify Hill-forts features from a domestic and military architectural point of view, as well as, the metallurgical, pottery and jewellery production, enabling a geographical framework of this type of habitat to be uncovered.

The identification of new settlements and elements of material culture, together with their interpretation focused on comparison studies determined the research rhythms between the end of the 19th century and the 1970s. This research stage focused almost entirely in excavating several settlements, recollecting artefacts, creating museums and publishing results in conferences and specialized journals, both in Portugal and Galicia.

Nevertheless, and with rare exception, 1 the studies were segmented in terms of subjects and value comparisons of data coming from different sites, which led to a systematic anachronism. The lack of knowledge over the artefacts stratigraphic contexts prompted a gradual devaluation of the habitats chronological evolution generating a challenged inheritance.

Northern Portugal and Galicia have persistently been recognised as the land of ’castros’ and a region where the Romanization process seemed to be systematically absent. Absent from the research objectives that were ideologically sealed by regionalisms and respective identity processes. Absent also from remains not identified or devalued by researchers, who cater for an indigenous discourse that resisted the Roman invader, although a few attempts had been made in properly interpreting the territorial evolution between the Late Iron Age and the medieval times by Alberto Sampaio (1899-1903), especially for the Portuguese territory.2

The main concerns were related with proving the material culture homogeneity and set the geographical boundaries of the ‘castro’ phenomenon, which in the region under study is clearly different from other peninsular areas. This was then translated into a rather simplistic comparison of these settlements features and severely minimized their differences. The ’space’ was valued as a context and

The systematic devaluation of the Northwest Romanization process, with particular emphasis towards northern 2 We owe to this researcher the first and clearer attempt to systematize the long-term historical reality of settlements. This reality was only archaeologically demonstrated a few decades later (Martins 1995).

1 Where one may have a critical perspective similarly to Rui de Serpa Pinto with regards to the Terroso excavations results in terms of the group of Hill-forts located in Portugal (Pinto 1932).

4

The Roman Settlement Analysis

Portugal, compromised the research of Roman remains and restricted findings and sites to occasional news, almost always found at the time of their destruction and primarily focused on the study of the road network system, the mining processes and the acculturation of Hill-forts.

In practical terms, the presence of Roman control seemed only to be sustained by the presence of road networks and by economical exploits of the land, although severely reduced to gold-mining activities. The knowledge over Augustan cities located on the peninsular northwest was scarce up to the 1970s and very little was known with regards to their urban features very much due to the lack of excavation works and research (Fabre 1970; Pastor Muñoz 1976; Gorges 1979). They only started to be included within the wider development of urban archaeological projects in Braga in 1976, and in Lugo and Astorga in the 1980s.

Literature produced up the 1970s revealed the fragmented and thematic character of conducted researches regardless of the fact they explored different sources. The conquest and organization of the peninsular northwest mainly utilized literary sources and epigraphic data (Shulten 1943; Torres Rodriguez 1948; 1951-52; 1976; Syme 1970; Diego Santos 1975; Roldán Hervás 1974; 1976; Alarcão 1973; Jonas 1972; Wiegals 1978; Rodríguez Colmenero 1979) and valued the social transformation of the indigenous world through epigraphy (Vigil 1963; Taboada Chivite 1968; Caro Baroja 1970; Albertos Firmat 1975; 1977; Santos Yanguas 1977).

The consequence of such trends can be fully understood in the following statement: ‘the lack of knowledge over both the urban and rural world has led to the conclusion that both the peninsular northwest and northern Portugal endured a small degree of Romanization due to its suburban location.

The Roman road network system benefited from the identification of several Milestones suggesting the layout of key itineraries crossing the Hispanic Northwest (Capela 1895; Taboada Chivite 1946; Barradas 1956; Estefânia Alvarez 1960; Rodriguez 1970; Martinez Tamuje 1975; Caamaño Gesto 1977-1978; Almeida 1979; Santos 1979).

The lack of traditional artistic artefacts, the absence of Roman buildings and monumental cities seemed to have performed the key reasons for those who stated that History in northern Douro did not include the Romans. The ‘granite civilization’ was not embellished with marble statues and was more likely associated with ‘castro culture’ or with the castles and churches that seemed to had been the standard Portuguese construction. Nothing resembled the delicacy and elegance of the cultural prototypes of Roman origin present in Baetica and Lusitânia (Alarcão 1973; Blázquez Martinez 1977).

The mining processes were studied by researchers mainly from the 1970s onwards (Domergue 1970a; 1970b; Allan 1970; Jones 19702; Domergue and Sillières 1977) and the monetary exchanges performed a key interest point from that decade onwards both in Portugal and Galicia (Villaronga 1970; Acuña Castroviejo and Cavada Nieto 1971; Cavada Nieto 1972; Centeno 1976-1977; 1977; 1978; 1987; Pereira et lii 1974).

The great visibility of the settlements against the invisibility of the Roman presence added to preconceptions in research interpreting the Northwest that addressed it as a peripheral land, marginal, a real ‘finisterra’, later pacified and hardly cultured.

Several studies regarding the Hispanic region focused on the Northwest religion and valued the voting epigraphy and literary sources (Étienne 1958; Blázquez Martinez 1957; 1962; 1970; Encarnação 1970; 1975; Albertos Firmat 1974; Pastor Muñoz 1974; 1976; Toboada Chivite 1976).

Gradually these prejudices shifted, mainly due to the adoption of new research approaches in the late 1970s focused on studying the Roman cities, developing sites inventories and excavating settlements. They introduced new methodologies in stratigraphic analysis enabling the generation of chronologies and the acknowledgement of the occupation timeframe of those settlements. They also contributed towards improvements in spatial analysis and the understanding of their variability and distribution.

The thematic character of Roman occupation research left out a needed appreciation of Roman sites, together with their chronology and spatial distribution. If it hadn’t been the case, this could have led to an understanding of the Roman settlements network when comparing it with clearly indigenous Hill-forts. However, such appreciation demanded knowledge over sites and findings that were only briefly referenced in literature. Not being possible to accurately date them, one understands the strength given to generalist interpretations with regards to the incipient Romanization of the territory, which in turn was originated from the ways the indigenous populations organized themselves, clearly attested by the survival of the ‘castrejo’’ habitat in Roman times (Cardozo 1966; Vigil 1973; Blázquez Martinez 1977).

Simultaneously, the first synthesis works focused on the ‘castro culture’ were brought to light both in terms of their broader context (Silva 1986; Höck 1986; Calo Lourido 1993) and in terms of establishing their geomorphological areas (Soeiro 1984; Carballo Arceo 1986; Martins 1990; Almeida 1990; 1996; Dinis 1993; Maciel 1997). The redefinition of research strategies supported by the new work methodologies and the first radiometric dating carried out in settlements changed the knowledge and understanding of this region in the 1980s. Several factors were decisive within the shift in research perspectives.

In this context, what are the indicators behind the low levels of Romanization? The problems are linked with: the lack of urban buildings, the reduced use of slaves’ labour, the lack of industrial production or commercial economic activities sustained by monetary trade only present within urban areas.

A rather peculiar phenomenon was the shift in research focused in the ‘castros world’. This is corroborated by González Ruibal, who states that from the 1970s the Ro-

5

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

mans are omnipresent in all existing archaeological settlements literature both in Galicia and Portugal. Quoting Ruibals’ words ’this Roman-centred vision of the ‘castros’ culture’ placed the focus on Roman influences: the chronologies were adapted to these events together with the main items of the ‘castros originality’ seemed to be dated and interpreted within this new hegemonic timework’ (González Ruibal 2006-2007).

the presence of similar urban infrastructures as to what had been found in other Roman cities. We highlight the presence of forum complexes known in Lugo and Astorga, public baths existing in all three cities, entertainment infrastructures, with Bracara Augusta presenting a theatre (Martins, Ribeiro and Magalhães 2006) and an amphitheatre (Morais 2001). Gradually, Urban Archaeology impacted upon traditional conceptions linked with the existence of an incipient urbanism, which would relate to a shallow Romanization process.

If this shift brought with it a better understanding of the Roman presence in the Northwest it did not substantially changed the knowledge with regards to the Roman presence in the region. This first stage mainly re-assessed the materials thought to be of Roman origin. The improvement in settlements chronology ended up revealing that several of the settlements believed to be of Roman influence were in fact of indigenous origin. Amongst others, we may refer the proto-urbanization of Hill-forts or to the presence of exotic items ranging from amphorae to ornaments uncovered in contexts that were clearly pre-Roman (Silva 1986; Martins 1990; 1991).

Several cities, including secondary ones, confirmed their orthogonal plans together with the presence of quality infrastructures, as was the case of Gijón (Fernández Ochoa et alii 1996; Fernández Ochoa 2003), Aquae Flaviae (Rodríguez Colmenero and Alcorta Irastoza 1998), Tongobriga (Dias 1997), León (Vidal Encinas 1986), Tude (Pérez Losada and Acuña Piñero 1998) and Iria Flavia (Perez Losada 1998; 2002: 87-109). Urban studies led to research focused on the relationship between the rural and urban settings.

In the 1980s the study of the Romanization process changed the traditional interpretations under analysis and these studies were based upon two distinct research approaches, both initiated in the previous decade.

The adoption of surveys and excavation work improved the knowledge about RRoman sites and triggered a reassessment of available data (Tranoy 1981; 1982; 1983; Fernández Ochoa 1982; 1983; 1983-1984; Martins 1995; 1996; Amaral 1999; Dias 1997).

One was based upon a re-assessment of epigraphy linked with the Hill-forts and with the indigenous world cultured processes that ended up being the base for Alan Tranoy and Patrick Le Roux’s works (Le Roux and Tranoy 1973; 1974; 1975; 1982; 1984; Le Roux 1974; 1975; Tranoy 1977, 1980; 1981). It is worth highlighting the remarkable work of Alan Tranoy on the peninsular northwest and published in 1981. The author used a vast array of data for the three convents and elaborated an integrated reading of the regional Roman occupation.

Within this context we highlight the work of Jorge de Alarcão (1988), who systematized data present in available literature, on Portuguese Roman sites and greatly contributed towards a shift in research methodologies through the compilation of scattered data while, at the same time, shed some light on the need for wider and more thorough regional studies.

The results obtained by both French authors set the basis for the reflection on the Roman occupation of the Hispanic Northwest presenting an overview detached from the settlements world. We also highlight the works of Iglesias Gil (1976), Albertos Firmat (1977), Árias Villas et alii (1979) and G. Pereira Menault (1982; 1983) who greatly contributed towards the understanding of the indigenous communities transformation processes.

In summary, the accumulation of available data about Roman sites on the peninsular northwest allowed a better convergence of research results. However, this convergence was more a virtual concept than a reality based upon collective and comparative research programs. Thus, it may add to wider complexities, from a landscape and settlement archaeological point of view, when adjusting intersection lines and synthesis of data with a diverse methodological support.

The other research approach significantly changes the traditional interpretation of the Roman occupation of the Hispanic Northwest was linked with urban studies. We are referring to the systematic excavations conducted in cities and part of the advances in Urban Archaeology studies.

Under these circumstances, the role of field survey was enhanced from the 1970/80s although it was conducted in non-systematic manners. Nevertheless, the territory started being archaeologically known. Three directions have been explored. First, a growing number of Roman foundational sites were uncovered, namely Villas, in most cases have been uncovered at the time of their destruction having, therefore, been subjected to punctual archaeological interventions. Secondly, the knowledge over Hill-forts was improved due to research challenges taking place in well-defined areas, as for example the case of water basins where excavations improved the overall understanding of their chronology, placing into perspective the generalist idea that they existed in Roman times. Lastly, settlement studies correlating

Archaeological works conducted in Braga since 1976 revealed an urban orthogonal plan present since the city foundation (Martins and Delgado 1989-90a; 1999), and a similar plan was recognised in Lugo (Gonzalez Fernadez and Carreño Gascón 1998; Rodriguez Colmenero et alli 1995; Rodriguez Colmenero and Carreño Gascon 1999) and in Astorga (Garcia Marcos and Vidal Encinas 1990; 1995; 1996). The recurrent excavations conducted within the three Roman Conventus capitals in the past decades determined

6

The Roman Settlement Analysis

Hill-forts with other habitats of clear Roman origin improving the indicators of the Roman presence in the region while, at the same time, redefined the research agenda to include the analysis on the landscape changing processes in the long-term, placing the ‘castro habitat’ in its rightful place.

1997; Fernández Ochoa and Morillo Cerdan 1999; Pérez Losada 2002), excavations were carried out in villas (Pérez Losada 1995; Pérez Losada 2000; Fernández Ochoa and Gil Sendino 2004), mining areas were acknowledged together with their associated infrastructures (Lemos 1993; 2005; Lemos and Morais 2004; Sánchez Palencia et alii 1990; Sanchéz Palencia and Fernández Posse 1993; Sánchez Palencia 2000) and the first works focused on the landscape morphology were developed (Orejas Saco del Valle 1996; 2001; Ruiz del Árbol 2001).

The abandonment of large scale approaches and the choice over studies of regional character conceived under a perspective of valuing both the proto-historic and Roman settlements evolution, delimited by coherent geographic boundaries represented the most important research focused on understanding the unknown Roman universe.

Although there have been significant improvements over the knowledge of the Romanization process within the peninsular northwest there is still a lot to be researched and the prejudice over this region marginality, when compared to other southern peninsular and Mediterranean Romanized areas, still needs to be overcome.

The field surveys conducted in the Sousa and Tâmega region (Soeiro 1984), those aiming to assess the medium Cávado river (Martins 1990; 1995), the study of the Lima river (Almeida 1990), the Ave river (Dinis 1993) and the Coura river valleys (Silva 1994), together with the Minho coastline (Almeida 1996), the Neiva valley (Maciel 1997), the western region of Trás-os-Montes (Lemos 1993) and the Chaves region (Teixeira 1996) were of paramount importance.

We also believe that the diversity of this overall process present in the vast region of the Hispanic Northwest is yet to be properly understood. In reality, the significance of the available data, yet rather fragmented and uncovered only in properly researched areas, revealed considerable discrepancies within the changes and socio-cultural evolution of this vast region. These events cannot be ignored and they make it difficult to generalize results as a whole.

Progress in research methodologies allowed the Roman occupation of the peninsular northwest to be as valued as the study of the previous occupation and started to include a diachronic and spatial overview within their study. Furthermore, the cultural changing processes hidden behind the concept of Romanization could, therefore, be analysed (Fernández Ochoa and Morillo Cerdan 1999; Orejas Saco del Valle 1996; Ruiz del Arbol 2001; Martins et alii 2005; Aris Vilas and Villa Valdês 2005).

Moreover, it seemed obvious that we need to ‘revisit’ the very own Romanization concept itself, which cannot be understood as a synonymous with replacing of social, economical and cultural indigenous structures by Roman ones (Reece 1997; Barret 1997; Mattingly 1997). The Romanization process was not a static nor an unilateral event but a dynamic, multidirectional and interactive one, mainly when it came to the adoption of Roman organizational models together with the ways indigenous communities apprehended and dopted them and answered to those models, all in a time that witnessed different rhythms very much dependent on the nature of the agents involved.

We will now review the main vectors of the overall process. In terms of the Hispanic Northwest region, research carried out throughout the last three decades determined the pre-Roman settlement patterns that conditioned the diversity and different chronology identified within the Romanization process of the three Conventus (Fernández Ochoa and Morillo Cerdam, 1999; Sastre Prats 2001; Martins et alii 2005; Árias Vilas and Villa Valdês 2004: Orejas Saco del Valle 2004).

The nature of available data related with the Romanization of Northwest has been rather fragmented, mainly descriptive and excluding wide areas of the region under study, which in turn made it difficult to overcome the traditional indigenous prejudices and the ability to homogenize a region that was truly diverse in its morphology, pre-Roman occupation models and clearly in its adaptation to a Roman order imposed by Augusto.

Urban Archaeology has shown that Augustan cities, regardless of their different origin, all displayed an orthogonal core plan and were developed alongside the staple features known in all Roman cities. They all presented public buildings, bathing infrastructures, entertainment structures, remains of temples, residential areas, a hierarchized street network, sanitary and water supply networks on top of other structures linked with production and trading processes (Martins 2004; 2005; 2006; Garcia Marcos and Vidal Encinas 1998; SeVillano Fuertes and Vidal Encinas 2002; Rodríguez Colmenero and Carreño Gacón 1999).

1.2 Reference matrix, transection and frameworks

The approach of Roman settlements, the core aim of this work, was conducted in two geographic realms that were distinct in terms of their analysis scales. The wider realm refers to the territory that corresponds to the present Portuguese region of Entre-Douro-e-Minho, where we attempted to assess the impact of the area integration within the Roman empire together with the processes related with the transformations in both the settlements and landscape, the establishment of the political

From a territorial archaeology perspective, the Roman road network projects were intensified (Lemos 2000; 2002c; Caamaño Gesto 1995-96; Rodríguez Colmenero et alii 2004), secondary urban settlements were identified, vici, mansiones and mutationes were discovered (Dias

7

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

capital of Bracara Augusta, the construction of the network of main itineraries key in controlling the territory and in its interaction with other regions and provinces part of the Roman world (Map 1).

The second chosen subunit referres to a mountainous region, crossed by the Via XVII itinerary also known as Via Nova and included the Homen river valley up to its junction with the Cávado River.

The research enabled us to equally assess the development process of new Roman habitats, such as the vici, the villas but also the mansiones and mutationes associated with the transport network.

The third subunit is represented by a landscape moderately rugged, with a wide coastline fringe and wide valleys crossing the region between the low Cávado and Ave rivers and corresponds, administratively, to Barcelos and Esposende municipality and part of the Póvoa de Varzim. With the exception of smaller areas reaching 500 metres high, this subunit is shaped by a 200 metres high-level curve.

In this wider realm, we tried to understand the abandonment and the persistence of pre-Roman communities, the inclusion of the surviving communities within new clusters and Roman infrastructures established in the region. We attempted to acknowledge the phenomenon of communities’ transference within landscape and settlement changing.

2 Data asymmetries and options

The challenges in research within this wide work scale did not enable us to efficiently formulate all these proposed aims with regards to the features of the Roman settlement, reason why we think to be important conducting a further detailed analysis within smaller regional areas. Only the cross-referencing and diversity in scales ensures with a minimum of efficiency the adjustment and finding of several analysis and synthesis levels over the territory.

Initially, we attempted to value a wide group of available data for the Entre-Douro-e-Minho region, which demanded the manipulation of a vast array of data, mainly bibliographical and included in catalogues and archaeological maps developed by several authors as part of their regional studies. The support needed to interpret the amount of data was provided by geomorphologic and topographic background, which enabled us to create an interpreted cartography, an analysis that will be conducted on chapter 3.

We have selected areas that prompted more thorough analysis scales that in turn led us towards other challenges and observational horizons. The selection process was based upon the following criteria: the stability and quantity of available data; the landscape features, the amphitheatre layout of the region, from the mountains to the coastline and finally the different impacts Bracara Augusta had over the surrounding territory. These guidelines also took into consideration the socio-political and economical reasons behind the transformation and shaping of the landscape under Roman domain.

This research scale tried to assess the potentialities granted by several, yet fragmented, available data for the Entre-Douro-e-Minho region, enabling us to acknowledge a group interpretation. The three analysis transept subunits created and approached on chapter 4 and 5 focused on different types of data. It included inventory results conducted by other researchers, works that have already been published and prospection works conducted by us together with a permanent manipulation of cartography and available aerial photography.

Three subregional units were defined, as they seemed suited for a detailed observation and the ability to express different organizational strategies: the Entre-Douro-eMinho reference matrix is here moved in terms of an analysis transection.

Point 2 will focus on summary features of the available data sources (mainly inventories and maps), on a critical overview of collected data, on the typology and conducted prospection orientation and also on the discussion over typological criteria of Roman archaeological sites.

The first subunit is Bracara Augusta and its surrounding area. We are dealing with a region with around 200 metres altitude located in the wide Cávado river valley on the northern side and by some significant landscape features outlining it in the south, east and west. With around 19.5 km N/S, by 23km E/W, the area includes the present municipality of Braga and part of Vila Verde, Amares, Póvoa de Lanhoso, Guimarães, Vila Nova de Famalicão and Barcelos.3 The western part presented valleys up to the coastline and mountains and narrow valleys mark the eastern area. The territorial features of this subunit link it with other selected subunits.

Some technical and conceptual aspects that will be handled on section 3 and 4. It is also important to refer that, by option, we summarized the cartographic data used and only actually included it when this data was in fact, seized and studied.

2.1 Available data

The catalogues and inventories conducted within the archaeological inventories are numerous and produced mainly from the 1980s onwards for the Entre-Douro-eMinho region.

3 This surface corresponds to the following sheets from the Portuguese CMP with a 1:25.000 scale: s. 56, 70 and part of the s. 55, 69, 83, 84, 85, 71 and 57.

8

The Roman Settlement Analysis

The catalogues are normally found in works focused on settlements analysis of regions that are more or less homogeneous and organized around hydrographical basins that are considered valid geomorphological units.

2.2 Data variability

The surveying of available data required the consultation of archaeological bibliography of Roman classified sites and remains indicating the Roman presence within the region, namely inscriptions, coins, varied archaeological elements and Milestones. We soon acknowledged the inevitability of conducting inventories of all pre-Roman settlements, as a thorough analysis of the Roman occupation history, which would require a solid assessment of the first Iron Age settlements framework.

The vast majority of these catalogues include prospection and excavations results conducted by different researchers and focused on the pre-Roman and Roman settlement and, in some cases, in an attempt to assess the long-term evolution of both settlement patterns. In general, these works offer great amount of valid data, many of which is new but is often heterogeneous and different due to distinctions in criteria adopted by the different authors. Differences range from remains descriptions and sites features.

All the references gathered performed a wide and chaotic group of data that can be seen or have been seen. Above and foremost, available literature provides a rather wide, heterogeneous and rarely interpreted amount of data. This is mainly due to the different backgrounds and interests of the authors but also related with the distinct criteria authors used to produce data and classify findings. While some findings are only focused on a group of tegulae as a match to a specific site, others are focused on the type of ceramics found or other remains enabling a characterization, although if brief, of listed sites.

From the group of works studying settlements and focused on well-defined areas, we highlight the works of Teresa Soeiro (1984) for the area between the Sousa and Tâmega rivers, Manuela Martins (1990) for the Cávado region, Carlos Alberto Brochado de Almeida (1990) for the Lima valley and coastline region, António Pereira Dinis (1993) for the Ave valley, Maria de Fátima Martins Silva (1994) for the Coura river valley, Tarcísio Maciel (1997) for the Neiva valley and António Lopes (2003) for the Minho river mouth.

Adding to this, monographic or synthesis works we needed to consult used different systems to register coordinates or simply not included that type of information.

There were also other catalogues included in works focused on the cities hinterlands, similarly to what happens in Tongobriga (Dias 1997) or in municipalities as for example the work by Maia Marques (1984) for Monção.

In fact, these sources were inevitably unbalanced when it came to the description and location of the findings. Furthermore, the accumulative effect of this type of data generated artificial scenarios of evidence and absence that were rapidly transformed in occupational trends.

From the works focused on the studies of wider regions, also including catalogues and most often identified sites through prospection surveys, we refer the ones conducted by Carlos Alberto Brochado de Almeida and focused on the Roman settlement of the coastline between the Cávado and Minho regions (1996). Following a similar approach were the works by Armando Coelho Ferreira da Silva (1986) that included a wide Hill-forts catalogue for the Entre-Douro-e-Minho region and also the catalogue produced by Jorge Alarcão (1988) and included in his work entitled Roman Portugal focused on the Roman remains referenced for the same region.

If there were areas subjected to a more or less intense prospection and excavation work, other sites seemed to lack information. These trends occurred when lesser research was conducted on site and literature related with the research was produced. Standard site cartography, based upon this type of sources, revealed quite clearly densities and gaps, which were unrelated with its occupation. Nevertheless, they could be useful for the researchers who consult them looking for evidence of settlement patterns for a certain region.

The present work also required the analysis of inventories of archaeological sites from archaeological maps with an emphasis towards the ones conducted for the Braga municipality by Luís Fontes (1993) and for Viera do Minho (Fontes and Roriz 2007). Less extensive inventories, although equally key in interpreting settlements were the ones produced for Vila Verde (Regalo 1986) and Póvoa de Lanhoso (Regalo and Brito 1991).

Thus, we ignored ‘remains’ or ‘neglected’ references of certain sites. The organization of the collected data from literature followed two basic processes. The first process aimed at normalizing data in terms of descriptions and locations whereas the second classified data by adopting standardized site categories related with research themes and adapted data taking the regional features into consideration while providing descriptions based upon literature produced by different authors.

Other than these catalogues and inventories, we also analysed archaeological surveys present in the Minho University Archaeological Unit archives, namely the ones produced by the institution for the different municipalities.

Such categories performed the database entry index that was developed to manage information needed to the present work (cf point 3). It is described on the Site Catalogue introductory section of this thesis. No data standardization and classification tasks are simple. Both are challenged by successive readings together with 9

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

remains and sites directories. If often enough the identification is clear, sometimes the terminology is confused and data reliability is small or even inexistent.

The case study of the Via Cova Villa, located in the municipality of Póvoa de Lanhoso, is a valid paradigmatic example. The Villa was uncovered by excavations and performed the pars rustica of a Villa. Prospections conducted within a wide area around the site were not sufficient to identify any associated remains. Another pertinent case study encompassed works that demanded soil displacement. At the time, it was detected in a joining area a demolition level associated with a building believed to have been a Villa pars urbana of around 2,5 meters deep.

Moreover, the manipulation of data coming from excavations represents data producing more conclusive findings in terms of chronology and functionality. On sites that underwent little prospection works, data may often produce unbalanced findings. The adopted criteria for the normalization and classification of data generated a series of permanent options and constant adjustments while producing a critical review over the operability of adopted methodology and confronting problems emerging from different stages of the work organization.

Other than these geomorphologic regional challenges, it is critical to understand the problems related with the dominant land system of the studied region, which is characterized by small size Farms together with the features and type of soil usage.

It seemed pertinent to adapt the standardization of description criteria and the site classification to the geographical, geomorphological and archaeological reality under study. Thus, we avoided developing constant new terminologies or site categories that would complicate archaeological records readings. In the same context, it seemed efficient to assess terminologies used in similar works and take into consideration the necessary geographical adaptions and the applied archaeological data (cf point 2.4).

The dominant Minho landscape consists of land divided in different and several fields dominated by polyculture. Each field performs a small area bounded by vines, walls, roads and trees. This landscape is not exclusive of valley areas and it extends itself throughout the slopes of several hills that shape the region and are dominated by small parcels sustained by land or rock ramparts. On the other hand, higher areas served other aims. Several sites are presently abandoned, are intensely wooded and more conducive to fires than to archaeological prospections. Even in sites where archaeological remains have been found, we can only estimate a settlement of a considerable size, due to the quantity of material and the presence of matching items. And we do so with a considerable degree of creativeness.

A balance in available data must focus on the considerable quantity of remains existing in the Entre-Douro-eMinho region and this allows the assumption that it is impossible to assign a plausible classification. How can a tegulae group, found in a churchyard, be categorized? Do we provide it with a status in a map or ignore it? The several options in cases like this may cause inevitable risks. However, we tried to minimize them by explaining the choices made when it came to organising catalogues (cf site catalogue).

The conduction of systematic field walking surveys is only possible within small areas, in particular regions with systematic work clearances. In the majority of cases, this task is extremely difficult as different fields are subjected to different and diverse cropping processes, being rather rare the finding of two fields tilled together.

Another challenge was linked with the criteria normally applied to the valorisation of pottery remains found on the surface and based upon the analysed dispersion area (Gorges 1979; Millet 1979; 1985; 2000). For the Entre-Douro-e-Minho region it was impossible to adopt, with rare exceptions, the criteria generating hypothetic site classification. This circumstance was related with the geomorphologic features of the territory, but also, with the soil use and land division. Both these events put pressure on the conduction of systematic prospections within wide areas and, as a consequence, the adequate adoption of these criteria. The geomorphology helped explaining the reason why we sometimes find Roman infrastructural remains at two or three metres deep. These circumstances mainly happened in valley areas, precisely the ones where these establishments are thought to have existed. The layered sedimentation present in valley areas has been the result of soils displacements on the slopes of nearby hills. It normally occurred through colluvial processes, and an intense yet persistent shift of the lands that ended up generating historic soils in the region (Bouhier 1979). Holders of great agricultural potential, these soils offer strong resistance to the detection of archaeological remains.

In order to adopt site classification criteria based upon the dispersion of remains and taking the examples provided into consideration, we have to acknowledge the fact that we are dealing with different specific geomorphologic conditions with particular features of rural space division and usage. Thus, it is paramount to adapt the hierarchy of sites to the reality of the studied region. Adding to this, it is important to emphasise the contribution of certain works carried out in mountainous regions, due to their specificity, the questions they raise (Ruiz del Árbor 2001) together with the propositions and their end results. We highlight the works of Luís Fontes in Lindoso, Serra Amarela and the ones conducted in Vieira do Minho (Fontes and Roriz 2007).

2.3 Conducted and systematic field walking surveys

Territorial and landscape studies brought to light a series of new questions linked with prospection methodologies and inevitably with the operational definition of the con-

10

The Roman Settlement Analysis

cept itself, its adequacy to the analysis of territorial features and period study.

to the region size and density levels of referenced remains. Under these circumstances, conducted prospections had specific aims and were meant to address particular issues emerging from previous data analysis together with challenges we set ourselves to investigate as part of the present work.

New Archaeology field walking surveys improved in the 1970s became one of the most relevant methodologies of the currently known Landscape Archaeology. It rapidly evolved in terms of techniques and general use boosting in turn wider theoretical debates about its benefits and the validity of results (Potter 1979; Barker and Hodges 1981; Dyson 1983; Alcock 1994; Millet 2000).

Thus, several of the field walking surveys can be considered ‘selective’ or ‘conducted’ as they were meant to resolve specific issues related with site catalogues and inventories found in the literature. They were carried out to improve records, geo-references and confirmations. We also conducted prospections in wider areas where previous works focused on territorial recognitions or in areas never subjected to research.

The methodologies and challenges related with archaeological prospection are represented by a set of techniques related with territorial studies presenting similar features, as for example, in the lands that is normally rugged or presenting soils with small degrees of archaeological value. These events greatly shifted the archaeological prospection exercise to become as important and demanding as excavation works.

We focused on the particular regions in greater detail, highlighting the transection crossing the Minho region between Terras de Bouro and Amares municipalities, going through Braga and Barcelos up to the coastline, in Esposende.

Field walking survey became a paramount work tool in research projects taking place in valley areas. However the debate related with adopted methodologies focused on the analysis of the geomorphological features of distinct areas, mainly mountain regions, remained.

We are then referring to conducted and not systematic prospections, which means they have been either based upon literary references or chosen due to their potential in terms of housing settlements or other forms of habitats.

In order for prospection results to be more than just cartographical data on a map, they need to be interpreted in ways to clarify questions, such as the relationships between different sites or their impact on the differential use of soils or even their repercussions within the evolution of ancient landscapes.

This type of assessment to uncover the ‘archaeological potential’ of sites adopted traditional support sources. First and foremost, we used 1:25000 cartography namely the oldest one produced in the 1930s and 40s. Although out-of-date in terms of transport networks and recent populated areas, this cartographic support performed an excellent work tool due to its toponym value revealing old roads. Geological and agricultural maps were a key data source in sites characterization.

Other than the location and quantification of surface remains, it is important to reflect upon prospection methodologies adapted to the type of region aiming to study and the challenges inherent to the field work recognition. More than just collecting scattered materials on the soil, it is important to understand their origin so that their archaeological record can be formulated. This task demands that the elements provided by the current landscape are analysed.

The photo interpretation conducted on stills from different surveys and with different scales contributed to a better understanding over the territory, the identification of archaeo-sites and the detection of underground structures suggestive of sites, however not visible on the surface.

Basic and common problems faced by field walking survey analysis of a given region are related with sedimentation, deposition, transport and revolving processes throughout times. Areas that have been intensely occupied are, by definition, intensely handled, revolved and recycled. This trend demands the consideration of several factors and a thorough and interdisciplinary work, as all these may affect the nature of the detected archaeological records.

We find it necessary and possible to conduct systematic field walking surveys in part of the territory around the city of Braga and in a small area of the Cabeceiras de Basto municipality. Selected areas are located between Braga and the Cávado river valley a region considered of top priority in prospection terms due to the following reasons: a)

It has been a densely occupied region throughout the ages, and in Roman times it boosted a rather particular type of settlements that transformed itself in the Suevi and Visigoth periods; b) This performs a high archaeological risk area presenting a landscape in rapid, radical and constant transformation due to intense public regeneration works and the growth of populated area in the outskirts of Braga; c) It is located in the surroundings of Bracara Augusta offering the potential to understand ancient rural land division other than the fact that it was crossed by Via XVIII and Via XIX iteneraries,

Taking into consideration our work aims and the nature of available literature, in its vast majority generated by prospection works conducted in non-systematic manners, it became paramount to conduct prospection in archaeological sites referenced in literature with the intent to complete and standardize data. Other prospections took place in areas with substantial information gaps, or to areas never prospected in archaeological terms. Moreover, it was not always possible to visit or verify available data for all the chosen sites within the EntreDouro-e-Minho region. This task seemed impossible due

11

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

the later with a challenging layout that is important to analyse in detail.

makes it difficult to produce comparative exercises existing in different backgrounds. The explanation over the adopted criteria to identify sites favours the equilibrium between the need to adapt classifications to the regions in which the sites are located and, at the same time, produce a comparative platform.

The selected areas for systematic field walking surveys around Braga are presented in Chapter 4 together with the adopted methodology in terms of prospectors’ numbers, record sheets, base cartography and collected materials.

The difference between vici, mansiones and mutationes is one of the cases in which it is necessary to find a parallel between the observed reality and what is referenced in literature. Not wanting to multiply nomenclatures, we must always apply the Latin terminology to different sites as it may correspond to the terms used by the Roman authors. However, we understand the risk of applying terminology to a wide group of sites that do not present the features used by Roman writers.

The base cartography used and produced by local authorities for the prospected areas had a 1:5000 scale. The areas were bounded in accordance with our aims and available means. Prospections had a team of six people. All the found material was collected, divided in chronological terms and managed, listed and subjected to statistic analysis. We also carried out systematic prospective works at Cabeceiras de Basto municipality. The researched area is located around the Santa Comba Roman settlement. It is an area presenting several literary references and Roman remains including honorary inscriptions. It is believed this site to have been a civitas capital (Alarcão 1990a).

We will list a few terms used and quoted in order to determine some approached angles. The Latin term vicus is rather ambiguous as it may refer to different realities. In fact, the haziness of the term can already be found in the Roman authors themselves, as it presents several meanings. Isidoro de Sevilha defined the term vicus as a group of houses, with streets but without walls, stating that one could also name vici to the buildings within a city, hence a potential neighbourhood (Tarpin 2002). The meaning of this concept has been widely discussed by several authors (Wightman 1986; Curchin 1986; Alarcão 1995; Tarpin 2002; Pérez Losada 2002).

2.4 Adopted nomenclatures

For sites classification, we adopted suggested nomenclatures that have been systematized by several authors studying the Proto-historic and Roman settlements for the region under study and part of their dissertations research (Soeiro 1984; Marques 1984; Martins 1990; Lemos 1993; Dinis 1993; Almeida 1996; Dias 1997; Maciel 2003; Teixeira 1996).

Legally the vici were organic sites and they can be characterized as open settlements without walls. They registered a spontaneous formation process, organic growth and lacked any type of foundation ritual (Broise 1976; Tarpin 2002).

We also analysed syntheses developed by Jorge Alarcão (1995-1996; 1996; 1998b; 1999a; 1999b; 2000), together with recent works focused on the same chronology, namely ones produced by João Bernardes for Collippo civitas (2007), Conceição Lopes for the territorium of Pax Iulia (2003) and Pedro Carvalho for the Beira Interior region (2006). All geographic differences have been considered.

Significant features have also been related with their diversity, which can be translated into different local economic realities. Moreover, the term can be applied to places boosting important urban buildings, as for example temples and public baths, a wide variety of administrative, economic, religious, thermal or port infrastructures or even small conglomerates that do not gather these features but are open populated clusters (Provost 1993).

Due to the imminently comparative nature of sites classification we also took into consideration studies conducted for the peninsular northwest (Carballo Arceo 1986; Pérez Losada 2002; González Ruibal 2006/07), for the Gallia (Provost 1993; Drinkwater 1983; Favory and Van der Leuw 19998; Van der Leeuw et alii 2003; Leveau 2002; Dousteyssier and Trément 2006-2007) and also, for Britannia (Millett 1982; 1990; Hingley 1989).

Thus, the term vicus seemed to have been associated with several types of rural concentration sites and it should be used to designate sites that have been less important than civitates capitals. On the other hand, F. Pérez Losada referred that vici performed individual groups of houses and buildings with a small degree of urbanization (Morel 1996:156) and their existence was dependent on nearby transport networks. Both their trade and administrative functions could lead them to become political centres in rural areas.

Although the comparative realm is relevant it did not seem pertinent to unify the descriptive criteria of sites when located in different regions, be it geomorphologic or in terms of human occupation due to the fact that identifiable items of an archaeological site are largely dependent on its regional and surrounding contexts.

They would normally be located in western provinces, although they can be found throughout the Empire. They present varied shapes and functions.

Equally, it seemed important not to generate new concepts or site categories. If the observed reality is indeed heterogeneous, the multiplying effect of nomenclatures

The fact that the Latin term encompassed a wide dimension led to the development of other descriptive nomenclatures less guided by Latin terms in order to improve the overall definition of these concentrated habitats. ‘Small 12

The Roman Settlement Analysis towns’ and ‘Villages’ would be examples in English language whereas ‘bourg / bougarde’ and ‘hameau’ for the French.

Private road infrastructures included big scale hostels: deversorium, hospitium or caupona and tabernae that were also hostels (Mantas 1996).

Although the creation of new terminologies aimed at designating certain realities that may not be easily adjusted with descriptions used by Latin authors, some contemporary authors (Tarpin 2002) believed that the proliferation of new meanings to reference similar populated sites brought new inconveniences. The use of classic concepts to identify and interpret sites seemed however the best way to proceed. F. Perez Losada (2002) followed this idea when using Latin terms to value the diversity and specificity of sites classified as vici.

The difficulty in distinguishing these terminologies has been challenged in the tabernae case as they could be located nearby cities, villas, vici, on the outskirts of mansiones and mutationes or simply present along roads, wherever and whenever human traffic would demand it. This difficulty has also been aggravated by the simple nature of these types of constructions, which in archaeological terms means these infrastructures can be interpreted as rural exploitation units or simply an undetermined site. Adding to this, recognising tabernae or the archaeological difference between mutationes and mansiones, which performed the more common structures in the territory under study, is not an easy task.

Latin terminologies for road stations were equally problematic as they included a wide group of terms that were not always clear. Adding to this, their identification performed one of the major challenges in understanding the overall Roman road network system (Mantas 1996).

The basic criteria to follow when aiming to differentiate these sites is related with distances calculations and attempts to adjust results with the geomorphologic reality of each studied site and with detected remains uncovered by the systematic works undertook in some of the main itineraries within this region (Mantas 1996; Lemos e Baptista 1995-96; Rodriguez Colmenero 1995-96).

One of the first challenges was related with the difference between official and private stations. The former obeyed to norms imposed by cursus publicus4 and the latter were dependent on several factors linked with secondary roads supply needs or with a source of income, as they tend to have been located nearby existing populated sites.

An even more complex issue is related with the Villa terminology present in the rural Roman world (Pérez Losada 1987).

Within the official road stations categories, the classic sources referred to mansiones, praetoria, stationes and mutationes.

For Tito Lívio, a Villa referred to any type of rural residence, but Varrão provided us with more thorough definitions for this type of infrastructure. In his opinion, a Villa entailed a cultivated fundus of significant size. On the other hand, Culumela defined a Villa in tree distinct categories: the urban, the rural and the trading structure, none of which corresponded to Varrao’s definitions referred to the former two as Villa urbana and Villa rustica, respectively.

The mansiones referred to key road stations within the cursus publicus and they provided a daily stop equipped to house overnight services destined for high state officials, including the Emperor himself. Furthermore, these infrastructures presented other types of services required by official couriers. Thus, a mansio required a wide variety of a complex set of infrastructures including mews, workshops, porches and accommodation services prepared to house large groups of people.

The application of the term Villa to rural Roman sites encompasses a great degree of subjectivity, even though most authors who refer to the term agree it relates to agricultural-pastoral exploitations boosting a set of residential infrastructures, other structures needed to the type of activities developed on site and a fundus that could have been continuous or discontinuous. A unanimous belief between researchers is that a Villa was subjected to a continuous evolution process that could have initially been a simple one but complicated itself in time and was very much dependent on the landowner’s economic situation.

The term praetoria was applied to important buildings destined to welcome high state officials and governors and they could include basilicae aimed at public function services and baths infrastructures or other key structures providing a pleasant stay for state Roman officials. The term mutatio referred to small-scale road stations and they had a more frequent presence alongside roads. They provided simpler services and accommodation facilities although they could include baths infrastructures. The stationes were mainly surveillance checkpoints and they played a key role in securing the safety of the roads. They could be fortified and have military deployment structures.

Conceição Lopes referred to the term Villa in the territorium of Pax Iulia similarly and considered this infrastructure as ‘an agricultural-pastoral exploitation unit encompassing two inseparable elements: a group of residential buildings, storehouses, mews structures and a continuous or discontinuous property (Lopes 2003:236). The author also highlighted the varied organizational possibilities around this type of structure (presenting distinct parts or functioning as a whole) together with the changes endured by the structure over time namely in terms of size. However, the author only integrated in this category sites pre-

4 The Emperor Augusto was behind the creation of the cursus publicus, which had the overall aim of organizing the official mailing services. This organization demanded the installation of structures with programmed distances throughout the viae militares. We recommend the work of Vasco Gil Mantas (1996) for the study of the road network system, its inherent typologies, existing road stations distances and evidently, the layout of the Via XVI itinerary, which will be referred in the upcoming chapters.

13

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

senting remains that have been interpreted and corroborate ‘the existence of certain types of buildings and wider infrastructures as for example, painted plasters, marble, tiles or imported materials in a significant number’ highlighting also that ‘epigraphy’s mentioning foreign settlers or the identification of a tria nomina should perform complementary indicators sustaining the presence of villae’ (Lopes 2003: 236-237). Last but not least, Conceição Lopes did not classify as Villa ‘any site registering a Roman occupation prior to the last quarter of the first century BC’ (Lopes 2003:237).

there does not seem to exist a clear and unanimous definition of the archaeological elements serving identifiable criteria. J. Gorges (1979: 14-17) referred that a topographic implementation or the distribution area of remains may be sufficient to identify a Villa. In his opinion, Villas would have privileged locations within the landscape and an extensive construction area. The author also suggested sizes for the dispersion areas. J. Alarcão (1995: 390) on the other hand, considered that the presence of tiles, bathing structures or funerary inscriptions fully corroborated the existence of a Villa. In order to distinguish Villas from other agricultural infrastructures such as Farms, the author combined criteria including size areas for remains dispersion and the presence or absence of certain and specific materials6.

Evidently this was a purely functional Villa classification, but also one that catered for the presence of prestigious structures that normally would be present in later and wealthier Villas. Adding to this, the application of the term Villas to feature wider establishments focused on rural exploitation processes in the southern Iberian Peninsula, or ones exploiting large scale fundi in the Meseta area, is clear.

Considering the suggestions of J. Alarcão and J. Gorges, for João Bernardes (2007: 20-21) a Villa may be classified as any site presenting architectonic and artistic elements, such as tiles, bathing structures and sculptures or sites in which these elements are present within a dispersion area of around one to two hectares. The remains would need to include fine and imported wares, worked stones, coins and inscriptions.

Problems started when trying to apply this terminology in areas with smaller fundi or in areas where the capacity to invest by their owners was significant less, as referred by M. Martins for the region of Braga (Martins 1995). For the author, the type of regional raw materials, the underqualified labour force and the socio-economic conditions of the region, largely shaped the type of existing structures. These tended to be small to medium size exploitation units fully adapted to regional raw materials, to the non-specialised local labour force and to the socioeconomic conditions of their owners. All these events shaped the small and medium size exploration units with ‘reduced construction areas’.

Considering the geomorphologic features of the EntreDouro-e-Minho region, M. Martins defined the criteria to identify Villas in sites including the presence of Roman pottery (tegulae and/or others), architectonic elements (column bases, shafts and capitals) or the existence of imported wares and epigraphic monuments (Martins 1990: 224; 2005).

Furthermore, in his work regarding Collippo civitas, João Bernardes (2007: 20-21) defined these structures as agricultural exploitation units existing in the rural world but subjected to city control, although the author acknowledged this concept as being rather confusing. Nevertheless, the author believed it applied to buildings representing a certain degree of wealth, displaying elements inherent to well being and the material wealth of their landowners.

The work conducted by Fransisco Sande Lemos in western Trás-os-Montes, defined that the identification of Villas would not demand any of the above criteria and the author utilized more standardized approaches (Lemos 1993). The challenges in classifying sites as Villas is a rather curious phenomena as it requires a permanent revision of the produced historiography and a constant thoroughness in interpreting different criteria to apply to this Roman sites typology.

There is no material or archaeological definition for this term.5

Agricultural exploitations also included another type of structures that normally would present more modest architectural elements when compared to Villas. Some authors defined these structures as Villas rusticae (Mantas 1987), small Farms (Mantas 1987; Alarcão 1990b) or Farms also know as ‘fermes’ in French literature.

In the absence of excavated buildings where one can fully acknowledge the traditional elements used to feature a Villa both functionally and materially, the challenges in the identification process of these sites will persist. The acknowledgement of this site category based only upon remains found in prospection work is complex as

The concept of ‘small Farm’ is not utilized by classic authors and only appeared in medieval times when representing agricultural exploitation units.

5 This author states the following with regards to the use of the term villa: ‘The archaeological problem is that we cannot say which site a “Roman” would have considered a villa since there is no unambiguous archaeological correlation with the Roman ideals. Furthermore, and although a villa recognized by a Roman generally included adjoining lands, here is no way of identifying the lands attached to any particular site. Thus, the widely used term villa-estate has no archaeologically useful function’ (Millet 1990: 92). The author addressed this ambiguity by providing the Latin term with an English terminology, suggesting the expression ‘villa’.

In Portugal, J Alarcão (1976: 11) introduced the term when designating small mixed exploitation units, with an agricultural or fishing character and fully dependent of 6 This group of criteria has been defined for the study of agricultural exploitation units in the Entre-Douro-e-Minho region (Alarcão 1995:387-401).

14

The Roman Settlement Analysis

larger Villas. However, the author also admitted the existence of another category for agricultural establishments, which he named as Farms and consisting of larger units administered by wealthier landowners with higher sociocultural levels (Alarcão 1998b).

formed the underlying aim leading to the application of Geographic Information Systems (GIS) in archaeology.7 The core aim when producing and applying a Geographic Information System to pre-Roman and Roman occupational remains in the Entre-Douro-e-Minho region (GISARM) was focused on three basic principles: the need to standardize historic-archaeological available data for the studied area, to develop settlement maps based upon archaeological sites inventories and to attempt to hierarchize the settlement.

Small Farms would then be rural establishments with a familiar character and they would not integrate the classic urban Villas. They would have small construction areas and a lesser quantity and quality of materials both in terms of constructions and needs. On another level, one can admit small Farms have performed other roles, namely workmanship and forest exploitation units (Lopes 2003:241).

The third stage of data processing for this work focused on varied analysis processes aiming to develop settlement models by just-positioning different data layers to generate a critical reading and a distribution of sites in space and in time.

It is extremely difficult to distinguish between both categories when we are not dealing with proper buildings but rather with simple remains and wide dispersion areas. Hence, it is challenging to distinguish what would have been a small Villa or a Farm when all we can assess are surface materials.

Data related with the nature of the remains and their respective spatial dispersion arising from both literature and conducted prospection works enabled the formulation and revision of classifications and criteria previously defined for each site category.

According to J. Bernardes (2007:19) the archaeological distinction between different rural exploitation units referenced in literature should only be considered as operative differences.

The ability to control groups of archaeological data and the possibility to provide meaning to cartographic distribution, greatly contributed towards a more rigorous and efficient analysis and interpretation of mechanisms and strategies for the spatial occupation, as well as to guide the overall process.

The fact that Farms may have had a plural functionality or even developed specialized productive activities not necessarily linked with agricultural endeavours makes the classification of sites even more challenging as they can easily be misinterpreted with handcrafts establishments (Lopes et alii 1998: 139-140).

The GISARM (Carvalho and Bernardes 2000) was based upon an archaeological site database developed in the software tool 4Th Dimension for Macintosh and linked with the cartographic application Map Grafix also for Macintosh. In 2007 the process of transferring GISARM to the ArcInfo version 9.2 started.

The existing difficulties in identifying rural establishments only presenting tiles or stone elements scattered around circumscribed areas, does not allow the identification of differences between Farm types, especially in areas as the one under study, given the impossibility to assess the size of their remains dispersion areas.

3.1 The alphanumeric and cartographic databases

In the present work we will apply the concept of ‘small Farm’ when referring to sites believed to have been agricultural-fishing exploitation units presenting remains including ceramics, millstones, worked stones and walls and also associated with slags highlighting in this case the craftsmanship character of the site.

The standardization process inherent to existing historic and archaeological data of the studied area was accomplished with the development of a database, where all the necessary data was stored under alphanumeric coded schemes. This process encompassed a thorough process of literature research in an attempt to unify all available data.

Lastly, we highlight the fact that the nomenclatures to adopt in the identification of Roman sites is still open to debate when researching rural occupations throughout the Empire. This is sustained by specialised literature, although attempts are being made towards standardizing the identification criteria and terminologies (Dousteyssier and Trément 2006-2007; Trément (Dir.) 2013).

The features of the available data made us decide over a database of relational files and the advantages of this type of application are to allow the files hierarchy. Secondary files are linked to a central file, meaning that a record of a certain file can match several records of another dependent file. Due to the non-quantitative nature of data inserted in most fields, we had the need to define in precisely the exact attributes for each case. Lists of attributes were created for each field and this presented clear advantages in terms

3. The methodological infrastructure: the GISARM

The need to list, standardize, analyse and represent archaeological data in a respective time and space per-

7 Available bibliography on the topic is vast. We will synthetize the works that have influenced the initial stage of this work: Allen et alii 1990; Arroyo-Bishop 1995; Baena et alii 1997; Morgan and Brusegard 1989; Reillly and Rahtz 1992; Richards and Ryan 1985.

15

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

of speed both when introducing data and in terms of data output. Each field can be questioned and the obtained results can be projected as a graphic or be cartographically observed. The connection between files and from each file to the cartographic database is their ID number and is based upon the National Geographic Code that follows a serial number. This number is the identifiable variable prompting a fast cross-reference of several attributes independently of their nature or the file they are in.

analysis models produced by geographers and applied to research on the past GIS become a key tool within the overall archaeological work. A priority and challenging issue is linked with the concept of typologies. The already mentioned standardization of data is an underlying condition in assessing data and performs the base for sites representation becoming inevitable to adopt in any archaeological work. The fixation of items related with standardized data follows classic data hierarchies, typologies and presents the distinct advantage of enabling the application of high volumes of data, but also the risk of analysis categories, perpetuating typologies and guiding models.

The cartographic base encompasses 3D digital information with regards to the hydrography and altimetry of the study area. The basic information is related with the altimetry and hydrography of the Entre-Douro-e-Minho region, in a scale of 1:250000, having been used for some specific areas a digital and cartographic scale of 1:25000.

The advantages of data handling and standardization may however, fossilize options and turn certain challenges recurrent. There is the risk of certain options to become definite in certain contexts and they can be valued to such an extend they become a problem themselves.

For the present work, we have selected a UTM geographic projection, although the system allows the conversion of all cartographic data into other types of projections. This Database also entails a wide variety of information, as for example a soil map and all the georeferenced archaeological sites. The latter are signalled by symbols representing their typologies that are automatically generated in the cartographic base resulting in an alphanumeric database.

Curiously, the possibility of producing tables and maps in a short period of time should prompt an alternative situation, as it enables the visualization of several possibilities arising from data interaction and largely dependent on the utilization of several surveys. This was perhaps the most interesting issue generated by this work as there is no doubt that the non-critical and committed use of GIS may prompt theoretical and methodological challenges that should be thought, revisited, discussed and questioned.

Each site was exported on a text file with the following information: ID number, X and Y coordinates, the symbol signalling the type of site and the site toponym.

As in any work model or method, there is the permanent need to face the nature of data and the development of support systems with a critical attitude.

3.2 The Geographic Information System: opportunities and constraints The development of a GIS 8 to manage available data required the learning of a language inherent to the development of a relational database, the managing of a cartographic base, as well as the use of the necessary schemes needed to interlink both databases.

4. Operational concepts

One of the main challenges to standardize and classify data starts here and one example would be the relevancy of the objects under study. This problem, present in most scientific works is enhanced by the ‘ability’ of generate General System Databases. These events are linked with the need to thoroughly handle data and the demands obtained results have in terms of stability, permanency and ’omnipresence’.

By conceptual decisions we mean the following: the pragmatic expression of looking for a minimum efficient structure that articulates and interprets the landscape in terms of archaeological data without interfering in data reinterpretation and benefits coming from enhancement and development of new research agendas. Throughout our work, four operative concepts became extremely relevant: ‘entities’, ‘relationships’, ‘limits’ and ‘scales’.

The idea behind this type of approach is related with the evolution of technologies and cartographic analysis softwares that in turn generate a growing possibility to handle increasingly amounts of data together with the speed and precision in analysis demanding spatial elements.

We may define as the operative challenge of the present work the identification and interpretation of relationships between entities composing RRoman settlements located in the western part of the Conventus bracarensis and referring to the present Entre-Douro-e-Minho region. The main challenge was not the archaeological site itself or the remains identifying it, but the possibilities of articulating them with other types of data, in an attempt to understand dynamics between settlements and the territorial organization present in a certain chronological and cultural period. The accomplishment of this aim demanded its articulation with the pre-Roman settlement but also its

It is relevant to explain certain conceptual decisions linked with the interpretation of findings, and, as a consequence, the progress towards settlement analysis stages.

Geographic data used in archaeology as a tool to manage Geographic Information Systems performs a methodological natural process. Similarly to several other spatial 8 The adopted methodology was presented in detail by J. P. Bernandes and H. Carvalho (1996), H. Carvalho (2000) and H. Carvalho and J. P. Bernardes (2000).

16

The Roman Settlement Analysis

cross-reference with the geographical features of the landscape in question, as these variables were generators of scales and limits that contributed to this region occupation specificities.

The research of these correlations and distinctions allowed the reference of entities as poles, or spots distributed across the landscape. They may be interconnected and spread in settlement networks that were wide and very much dependent on the chosen analysis scale.

We redirected previously referred observations, in order to determine two referential ‘angles’ shaping the problem.

The identification of entities catered for an understanding of the successive landscapes. It was focused on questioning territorial polarities and the need to identify natural or anthropic primary entities providing the landscape with permanent features.

‘Angle’ 1 In order to identify the Roman settlement matrix prompted by the adaption of political and administrative models imposed by Rome for the region, we needed to understand the role of Bracara Augusta and the road network system developed to link the city with the wider Northwest and other Hispanic urban centres, together with the displacement process of the pre-Roman social organization. Moreover, we hoped to understand the way indigenous communities, mainly the ones located in Hill-forts, adapted themselves to the new territorial resources exploitation, the new urban supply and to the Roman market economy. What survived, how and why of the previous indigenous settlement network, how the introduction of new types of Roman sites was processed, how secondary settlements such as Vici or Villas articulated with indigenous ones and how the latter persisted after Augusto, performed some of the challenges guiding the programmatic decisions of the present work.

The preference of the term ‘entity’ over ‘type’ was mainly linked with the excessive standardization of the latter concept. We cannot state that a Roman road is a ‘type’ site. It seemed more accurate to refer to it as an ‘entity’ structuring and shaping the landscape, while at the same time, conditioning the development of other entities. Furthermore, it was critical to clarify how the concept ‘entity’ could be applied to archaeological research. One of the central concepts in programming methodologies and database structures has precisely been this concept. We designate by ‘entity’ something that exists or existed and may be represented in a computer system. Entities include objects and things that really existed (like a certain person or a certain building) but also encompass abstract constructions as for example a ‘family’, ‘relationships’ or ‘events’ (Carvalho 1997: 313).9

‘Angle’ 2

Thus, we speak of ‘entities’ as something that existed and that in archaeological language prompted the development of site categories, constructions or findings, be they a Roman Villa, a coin, a group of buildings or a necropolis. However, the concept of ‘entity’ may also be used to designate abstract constructions, for example relationships between sites or anything we understand as paramount within landscape interpretation. In the latter sense, the concept may add to the understanding of ‘entities’ as poles or spots organised according to function and enabling several other entities to shape relationships and establish boundaries.

Previously, we summarised how data presented in this work was handled and combined in ways to include all the dimensions inherent to archaeological work together with how a computer system was generated. The managing and manipulation of a vast array of complex data has a multiplying effects in terms of reformulating work fronts, be they the development of software tools allowing the handling of data relation, the control of data through prospection works, the assessment of results, the critical revision of subjects directly linked with challenges present in landscape studies, the development of site typologies, the definition of potential areas and the establishment of chronologies.

The identification of ‘entities’ shaping the territory and functioning as poles allowed the creation of spots that in turn generated other observational pathways when analysing settlement. When understood as a spot, these ‘entities’ may be connected and become settlement networks with a considerable size and dependent on the adopted analysis scale. We hope to highlight their capacity to be faced as ‘operative entities’ within the overall historic construction of the territory throughout Roman times.

We proceed by presenting some perspectives related with the four operative concepts. Ph. Leveau highlighted the challenge of the ‘tense’ relationship between site analysis and its inclusion in a particular space. ‘Pour définir le statut d'un site dans un territoire donné et à une époque précise, les archéologues recourent habituelment à l' approche typologique qui l'isole de l'espace. Ils procèdent selon une méthode analytique et une logique déductive à partir de critères définis a priori et hiérarchisés. Mais il est possible de caractériser un site en le plaçant dans un ensemble, soit en donnant la priorité au système relationnel qui permet d'en préciser la fonction’. (Leveau 1997: 7).

Both definite and abstract entities have been defined by their attributes. 9 These concepts are key in the Database General Systems and belong to a model entitled ‘Model of Entity-Relationship’. For further and clearer explanations of these concepts or other data modeling and representation of historical sources, please consult Joaquim Ramos de Carvalho (1997). Database General Systems use concepts that we also find in Semiotics and in Language Philosophy. What is pertinent in this particular concept concerns wide areas of theoretical and methodological thought and regards the polysemy ability that enable us to acknowledge terms and concepts that are difficult to use and associate with archaeological practice.

17

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis The attributes of an ‘entity’ refer the ones we classify as for example a Roman city, certain geographic coordinates, a group of buildings, their remains or the length of time they were occupied. On the other hand, ‘abstract entities’ referred to attributes of that same entity, for example its historic role within the politics of a given period, certain events or their centrality and settlements. For each ‘entity’ there will be one or more attributes that distinguish themselves from others and allow us to use it as an interpretative pole of several archaeological records scattered around the landscape. Their use also simplified the interpretation of data mainly when we were dealing with ‘entities’ that we may call primary or ‘entities’ functioning as settlement networks catalysts. ‘Main entities’, which in a structural manner seemed to polarize the landscape are several: the city, the road network, the vici, the RomanisedHill-forts, the nonRomanised Hill-forts and the Villas. These are poles with sizes, relevance and duration that are difficult to hierarchize due to errors generated when assessing their typology. These challenges lead us towards a second concept: one related with the relationships between several entities.

the concept led us to establish boundaries between wider entities, as the ones within a city or within a territory. Still on this scale, this concept prompted a discussion related with the evolution of landforms and their constant configuration and reconfiguration, thus allowing particular territorial dynamics to be approached. ‘Entities’, ‘relationships’ and ‘limits’ demanded their articulation with several scales of analyse in which data may be observed both in chronological and spatial terms. These scales were closely related with an attempt to investigate further pertinent criteria and the real meaning of data. How to control the specificity or the similarity of these data? How to remake this path? How to improve the rigour of hypothesis and observations? It is important to draw attention towards certain significant effects of these scale games. In fact, it helped us to place the research of settlement models on a comparative level, while assessing the modalities enabling the landscape to evolve, produce cuts and successive reuses within configurations that were often displaced from the created typologies needed to analyse them. From the definition of pertinent scales, only unfolding analysis areas will ensure the efficiency and the comparative worth of settlement analysis.

The concept of ‘relationship’ allowed us to compensate the unbalances present in frequent analysis, consisting in overlapping almost automatically certain primary ‘entities’ with the surrounding settings. Similarly to ‘entities’, the ‘relationships’ may be described as attributes. A ‘relationship’ between an ‘entity’ defined as a landscape pole and other ‘entities’ may be the proximity, the chronology or a certain degree of landscape hierarchy. This concept allowed us to work on several platforms of analysis: to detect the relationship between sites, to assess the relationship between continuity and discontinuity over territorial occupation, prompt relationships in ways to add logic when identifying sites or entities, meaning that the establishment of relationships may help improving the overall classification and interpretation of archaeological sites. The concept of ‘limits’ may help us to work on a series of problems faced by archaeological research, although it is paramount to define what we understand by the concept of ‘limit’. On a first level, the concept was linked with the spatial demarcation of entities, or with the definition of the influential sphere of considered primary entities. On a second level, naturally articulated with the first, the concept was related with the maintenance, reuse and recycle of other settlement limits. The use of this concept necessarily implied the need to set up precise scales of analyse as the concept encompassed different opportunities depending on whether we work with larger or smaller scales. On a wider scale the concept of ‘limit’ allowed us to closely handle spatial regularities and irregularities, mainly with the maintenance and/or retraction of frontiers, or with the persistence, reuse and retreat of potential exploitation territories. On a smaller scale, the operationally of

18

1. Pre-Roman settlement

Chapter 2 The Northwest Integration within the Roman World

New directions in research and methodologies focused on the study of Iron Age settlement located in the peninsular northwest in the past two decades enabled us to identify important variations in terms of research patterns and interpretations of the complex and diverse pre-Roman regional occupation. The registered changes perform a framework of knowledge surpassing the apparent homogeneity of the so called ‘castreja culture’, allowing us to differentiate regions, assign chronologies and have a much clearer understanding of the different events the Roman power had to deal with at the end of the first millennium BC. On the other hand, research advances related with the Iron Age period took place alongside research developments over the Bronze Age era. These events significantly improved the type of available information and knowledge related with these periods.

Introduction

The main aim of the initial part of the present chapter is to contextualize the diversity of the pre-Roman peninsular northwest and understand its different conquest tactics and the regional domain on behalf of Rome. This overall aim can be explained through several questions that will be systematized in different points. The question approached in point 1 will seek to synthetize knowledge over the Iron Age in the Northwest region while taking into consideration the regional variations in terms of settlement patterns, Hill-forts sizes and chronologies, the use of economical resources, the profit and the overall social organizational patterns.

Under these circumstances and considering the time of conquest and integration of the settlements located in the region under study, we have been able to understand the long-term effect of this cultural process. Furthermore, it became important to initiate an approach that would also consider the pre-Roman world, mainly from the Bronze Age period onwards.

The conquest and integration of the northwest region within the Roman world and the impact these events had upon the indigenous social organization will perform the overall aim of point 2. Equally, we will attempt to verify the utilized procedures present in the political domain and in the administrative managing of this Hispanic region, namely between the Augusto and Flavian dynasty.

1.1 Bronze Age and the transition to the Iron Age

Point 3 will focus on the organizational patterns and rhythms in terms of territorial exploitation and in the consolidation of the Roman domain, both in Flavian and Antonine periods.

When González Ruibal contextualized the Bronze Age problematics within the peninsular northwest, the author also referred a new phenomenon: the emergence of the rural landscape - a parallel phenomenon that took place alongside the establishment of the first communities at the beginning of the first millennium BC (González Ruibal 2006-2007). The phenomenon naturally encompassed new concepts and a wider social space understanding and was responsible for the ‘development’ of a new type of landscapes. This era registered a more active intervention of communities with their surrounding landscape and its transformation into a cultural delimited space. This process has been explained in the expression ‘the landscape transforms itself into territory’ developed by F. Criado Boado (1993). Furthermore, the process of concentrating population emerged alongside the development of communal spaces organization. Both these processes profoundly impacted upon the overall landscape evolution. The strengthening of communities’ identity eventually emerged at this time, and it seemed to equally perform one of the most significant changes occurred, although this process was only fully consolidated throughout the Iron Age.

Figure 1. Map of Hispania representing the division of existing provinces and conventus (provided by M. Martins)

Nevertheless, when we look at the peninsular northwest we find that these processes revealed significant differences and challenges already referred by several research-

19

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

ers (Carballo Arceo 1986; 1990; Fernandez-Posse 1998; Villa Valdês 2002).

For the studied region in particular, the explanation related with the emergence of a type of habitat that would appropriate space and intensified agricultural exploits, events that have been acknowledge by researchers as having had its origin in the Chalcolithic (Bettencourt 2003b), was also linked with the development of metallurgical and long distance trading activities. The latter was a phenomenon that occurred in the Late Bronze Age (Peña Santos 1996; Bettencourt 1999.

There were particularly dynamic areas and others registering a slower evolution. The transition between the first to the second millennium BC revealed two basic traditions in this region. On the one hand, the foundation of settlements in high altitude locations can be found in the south and western areas, with the Sil and Lérez rivers performing the territorial northern boundaries. For the remaining region, we mainly witnessed the maintenance and development of open landscape processes that were more common throughout the Bronze Age. Within this context, the area between Minho and Lérez performed an intermediate sector for the two mentioned regions (Carballo Arceo and González Ruibal 2006-2007).

Throughout this period, mainly between 1000 and 700 BC, there seemed to have been developed a settlement pattern emerging in geomorphological and different topographical contexts (Silva 1986; Martins 1990, Bettencourt 2005). This phenomenon may be explained by the need to diversify the exploitation of economical resources that likely led to a social hierarchy and the affirmation of leadership positions (Dinis 1993; Martins 1996; Bettencourt).

When referring to northern Portugal in general and according to González Ruibal research works developed in the last decades have been revealing a series of new and important findings over the human occupational processes and landscape organization related with the transition of the Bronze to the Iron Age. It is possible to state that the Late Bronze witnessed a great diversification in habitats and space occupation practices as observed in several sites (Jorge 1988; 1998; Bettencourt 1999, 2000c; 2005).

The available environmental data for this period indicated an intense economic activity linked with a systematic reuse of available resources (Figueiral 1990; Bettencourt et alii 2003; Figueiral and Bettencourt 2004). The pollen spectrums registered a continuous presence of cereals, non-tree pollen and species signalling a strong anthropic action within areas close to the communities.

The existence of open landscapes in mountainous areas and other geomorphological and topographical contexts has been acknowledged, as for example in riverside areas, in open uplands and in small hills next to valleys or in slopes.

On top of cereal and vegetable remains, detected by both Palynology and by species macro-remains identified in conducted excavations (Figueiral 1990; Bettencourt 1999; 2005), chestnut and oak trees seemed to perform the main forest species ensuring the tree coverage of the region (Figueiral 2000). The scots pine also performed key resources.

These types of settlements were known to exist in both northern Portugal and southern Galicia (Jorge 1987; 1988; Suárez Otero 1998). They normally had small sizes and included constructions and storage units made of perishable materials suggesting a systematic exploitation of the land and the use of agro-forestry resources10.

Overall, the data uncovered by Palynology (Aira Rodriguez et alii 1989), Antracology (Carballo Arceo 1998; Figueiral 1990; 2000) and Edaphology (Díaz-Fierros Viqueira 1996) revealed an intensive use of available resources in the recollection and food production, in hunting and grazing activities. The antracological data revealed human activities including deforestation to create Farming or construction areas (Figueiral and Carcaillet 2005)

Other than typical Bronze Age open settlements another habitat emerged in high areas between the end of the second and beginning of the first millennium BC. It occupied prominent locations within the landscape, encompassed defensive structures or included delimitation elements in their physical space (Martins 1996; Bettencourt 2000b; 2005).

From the end of the second millennium BC onwards, intensification in agricultural exploitation was registered together with a more proactive relationship with the surrounding landscape on behalf of communities. These changes were likely accompanied by social changes in the sense of providing communities with a greater complexity and social hierarchy. These changes can also be seen in the material marks present in the settlements (Martins et alii 2005; González Ruibal 2006-2007)

The presence and growth of these types of habitats throughout the Late Bronze Age was not exclusive of the peninsular northwest region, as it actually took place throughout Europe. These events have generated varied explanations.

Another important issue in this context is the importance of long distance exchanges. It is possible to state, based upon available archaeological data, that the northwest region was part of an integrated and complex trading network structured before the Late Bronze Age and linked with northern Europe (Jorge 1998). From the Late Bronze Age onwards communications were extended into the Mediterranean, making available to the northwest region

10 Although Bouça do Frade may have had a Midlle Bronze Age occupation, the majority of these types of habitats were established in the Late Bronze Age period (1250/700-600BC) (Jorge 1987). Similar sites are known to have existed in the Baião area, namely the Lavra and Tapado da Caldeira (Jorge 1980; 1987). In the Cávado valley area there were some habitats registering trenches as it was the case of Alto de S. Bento and Alto da Cividade, in Braga (Bettencourt 2000c). They were equally present in the Lima valley, as for example the Regueira site (Almeida et alii 1994) and in the Minho valley, the Fraião castle case (Almeida et alii 1995).

20

The Northwest Integration within the Roman World

productions and innovations influencing socialeconomical changes that can be observed throughout this period (González Ruibal 2006-2007).

will perform the dominant habitat throughout the II Iron Age. The defensive structures were now made of stone and they increased in number and visibility. Another key factor was linked with the type of houses existing in these Hill-forts. If, in some cases, there seemed to have been a continuation of the Late Bronze Age traditions, with communities living in houses presenting round structures and made of perishable materials, in others cases already occurring in the sixth century BC, the presence of houses made of stone was registered in Galicia and the Minho valley (Silva 1986; González Ruibal 2006-2007).

During the first millennium BC a progressive intensification of agricultural activities was registered and another set of transformations becoming more visible in the last two centuries BC. The economic development emerging from more intense and progressive exploitations of agricultural resources represented a long term process originated at the Bronze Age. It endured retraction moments mainly associated with the First Iron Age, in a time registering an absence of valley settlements (Martins 1990). Between the ninth and eight centuries BC the archaeological data indicated the abandonment of open settlements both in valleys and mountainous areas. Similar events took place with high altitude based communities. Nevertheless, several settlements emerging in valley borders were also abandoned while others were further developed (González Ruibal 2006-2007; Martins 1996). This changing framework has led researchers to locate at this period the transition into the First Iron Age (González Ruibal 2006-2007).

1.2 Communities diversity throughout the first millennium BC

The study of first millennium BC settlement demanded, in terms of present knowledge, the consideration of all the variations that this phenomenon has known within the geographical context of the peninsular northwest. The observed heterogeneity constituted a substantial interpretation challenge. Currently, the existence of more dynamic regions seemed to be certain and one of these was located in the Galician zone including both the Portuguese Entre-Douro-and-Minho and southeast Galicia regions (González Ruibal 2006-2007).

One of the possible reasons behind these trends was likely linked with climatic changes resulting from the passage of the Sub-boreal to the Sub-Atlantic, which was characterized by an increase in rain and humidity, turning valleys and low lands into swamp areas (Ramil Rego 1993). On the other hand, the degradation of the fertile mountainous areas and the fact that exploitation activities of the previous millennium might have impacted upon the continuous production, could explain the abandonment of mountain settlements (Martins 1996). Both the climatic changes and the mountain soil degradation could help explain the need for a human landscape re-organization together with the need to diversify food resources. Thus, these events likely led communities towards an agro-forestry and pastoral economies, common practices within Iron Age communities (Martins 1990).

The pre-Roman world in the peninsular northwest is better known in its southwest area, when comparing to the remaining northwest regions of the peninsula. Thus, the assessing of results may highlight structural differences. By excellence this area was the region of the ‘castreja culture’. It registered a high density of settlements and we have recognised the existence of wide settlements, also known as ‘citânias’. The excavations of ‘citânias’ were behind the creation of the idea of a ‘castro culture’, with features that remain to this day the ‘brand’ of the northwest history prior to the Roman conquest (Martins 1990). This image demanded, however, a more thorough study.

Although less abundant, the available archaeological data for the beginning of the Iron Age revealed a clear decrease of goods linked with long distance exchanges together with smaller metallurgical productions (Martins 1996; Bettencourt 1999; González Ruibal 2006-2007). The ritual deposits featuring the Bronze Age diminished and the relationships with the Atlantic and Europe seemed to have lost their previous dimension (Ruiz-Gálvez Priego 1998; Kristianses 1998).

Between the sixth and fourth centuries BC the retraction that some authors associated with the transition stage between the Bronze Age and the first Iron Age gave place to a significant impulse development in the Entre-Douroand-Minho region. The monographic (Martins 1990; 1991; Almeida 1996; Dinis 2001) or the synthesis works (Silva 1986; Hock 1986; Queiroga 1992) developed in the last decades provided a settlement framework concentrated in sites with a significant topography within the landscape and where communities’ strategic objectives in controlling the surrounding landscape seemed to be critical.

The abandonment of settlements located in lower valleys and the concentration of the population in settlements located near the borders of important water basins prompted structural changes, both within the communities and in the relationships they kept between themselves.

As much as it is possible to verify the fortification of settlements became a standard procedure and represented a consolidation of communities with their surroundings, mainly in terms of controlling the valleys that crossed the Entre-Douro-e-Minho amphitheatre topography. This was a territorial consolidation process on behalf of the communities alongside their effective sedentarization (Silva

The exploitation of agricultural resources took place within existing settlements bounded by wall lines and likely created pressure over the control of the conquered land, as suggested by several authors (Martins 1990; Dinis 1993; Almeida 2000; González Ruibal 2006-2007). Hill-forts

21

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

1986; Martins 1990; Alarcão 1992; Martins 1996; Almeida 1996).

erside forest areas also suffered a decrease due to the rise of settlements in low altitude lands (Martins 1988).

The beginning of the fourth century BC witnessed a substantial growth in settlements. In some cases, these habitats existed in sites occupied from the Late Bronze Age onwards, while others were created as part of a trend of intense occupation of the coastline and valley areas. The general framework revealed a significant increase in occupied areas in this region.

In Galicia the emergence of settlements took place at a later stage when compared to northern Portugal, and no remains have been found prior to the ninth century BC, as opposed to what researchers found in Entre-Douro-andMinho where the process was initiated in the transition between the 2nd to the 1st millennium BC (Carballo Arceo 1989; González Ruibal 2003). Nevertheless, the historical evolution of Iron Age societies throughout the first millennium BC seemed to follow the same patterns.

Alongside settlements located in strategically important sites when it came to control and visibility, we registered a growing need to establish new Hill-forts in order to fixate populations that were by then more interested in exploiting the valley resources. These trends prompted an intensification of land exploitation and significantly increased the worked lands.

In Lugo, the variability of settlement patterns continued to be a prominent feature, and one inevitably linked with the geomorphological differences between the western granitic region and the eastern areas registering the presence of schist (Árias Vilas and Villa Valdês 2005). Settlements throughout the first millennium BC, for this particular region, may be grouped around three main categories.

The structuring of a new scenario may be synthetized in three main processes – clustering, fortification and territorialisation. They were directly linked with the need to demark economical exploiting areas within communities, ensuring land ownership and its effective appropriation (Martins 1996). The diversity of the Portuguese northwest Hill-forts was, without any doubt, the most prominent feature of this overall process. Alongside wide settlements or oppida housing considerable population numbers there were medium and small Hill-forts bounded by either a wall line or a trench and located in valley areas. They varied in size, structure and cultural materials. The latter also varied in quantity and variety in terms of ceramics and metallic artefacts. Diverse were also the areas used for exploiting resources. The agricultural economy practiced in the valleys coexisted with the exploitation of bush areas located on settlements surrounded by strands and oak or cork forests but also by riverside forests located nearby water lines or around wider valley areas. This vast and diverse set of resources served an agro-forest and pastoral economy that allowed the production of cereals, the recollection of fruits for both men and animals, fed the cattle and enabled the storage of energetic and construction resources that were important to the survival of communities (Martins 1996).

The first was represented by coastline settlements, located near the coastline in hill tops or in small hills near the sea. They presented simple defensive systems, normally composed by one wall line followed by a trench. This first settlement group was common in all of Lugo and Asturias region and was extended across the Galician western façade. The second category, defined by F. Árias Vilas and A. Villa Valdês as interior Hill-forts is located in higher areas of medium altitude and were reasonably prominent in the surrounding landscape. This type of settlement is more significant and also better known within communities mainly occupying the borders of the Galician valleys. Finally, the third category existed in the eastern mountainous region next to metal bearing exploits. These were highland settlements boosting rather complex defensive systems encompassing more than one wall line and trench. Their development already took place in Roman times, due to their role in exploiting the metal bearing areas (Sanchez-Palencia Ramos and Fernández-Posse 1993; 1998; Sanchez-Palencia Ramos et alii; Oregas saco del Valle and Sanchez-Palencia Ramos 2002). Three chronological stages enabled us to contextualize the evolution of settlements throughout the 1st millennium BC in the Lugo region.

The technological advances have been documented by the presence of agricultural iron instruments which marked the development of the II Iron Age metallurgy (Silva 1986; Martins 1990).

On a first stage, between the ninth and fifth centuries BC Hill-forts became the only dominant habitat. On a second stage, between late fifth and the beginning of the fourth up to the second century BC, the initial habitat model consolidated itself and a more active exploitation of the surrounding area took place. Finally, the third stage, between late second and throughout the first century BC up to late second century AD there was an internal and external change in settlement types very much influenced by the Roman presence. The Roman domain in this region seemed not to have prompted the abandonment of the Hill-forts, but rather prompted an interaction between new habitat types. These were clearly linked with the conquest and integration of the region within the Empire, a phenomenon very much present in the habitats located in

The economic intensification likely led to a significant disintegration of areas previously occupied by forest, and registered a strong anthropic impact within the areas surrounding the Hill-forts. Communities’ need for wood prompted the opening of clearings with agricultural ends and these events might explain the paleo-environmental data revealing a progressive decrease of forest areas on behalf of communities throughout the first millennium BC (Aira Rodrigues and Ramil Rego 1992; Figueiral 1990; 1993; Díaz Fierros Viqueira et alii 1994). The riv-

22

The Northwest Integration within the Roman World

mountainous and mining regions (Sanchéz Palencia Ramos and Fernandez Posse 1986; Árias Vilas and Villa Valdês 2005; Orejas Saco del Valle 2005).

Lucus Augusti and Asturica Augusta, by the Emperor Augusto.

In the eastern Lugo, Asturias and Cantabrian regions the archaeological data revealed a rather diverse and less developed settlement framework (Villa Valdês 2002). The northern Cantabrian mountainous range only housed Hill-forts in late third and second centuries BC. They were located in medium altitude regions in small hills or coastline areas and they were normally bounded by one wall line and surrounding slopes (Camino Mayor 1995). The only known exception to such a late chronology referred to the Campa Torres settlement, with an occupation chronology ranging between the six and fifth centuries BC (Maya Gonzaléz and Cuesta Toribio 2001). The defence mechanism of the hill-forts in this area was rather simple and encompassed slopes and stony walls without any reinforcement. The archaeological data demonstrated these settlements had a survival economy that was predominately pastoring and included embryonic trading practices (Carrocera Fernandez 1995).

Figure 2. Boundaries of the three northwest conventus and their respective capitals (provided by M. Martins)

In the southern Cantabrian Mountains the Iron Age period is not well studied. Nevertheless, three distinct areas can be defined. In the Leon region, habitats had small and medium sizes, stony walls and, according to E. Carrocera Fernandez (1995) they had a clear agricultural vocation. The Hill-forts in the Zamora region were mainly mountain habitats located in relatively high altitudes, isolated and displayed defensive mechanisms composed by walls and reinforced with trenches and sharp stones. The main economic activity of these communities was pastoralism. The mountainous region housed settlements that had isolation as their most prominent feature, and communities clearly controlled the surrounding region. In the area that is currently Portuguese territory, the settlement types seemed to include pastoral resources more than agricultural ones (Lemos 1993; Lemos 1995).

An increase in agricultural production, the development of trading, a shift in settlement patterns and the ‘internal’ reorganization of certain Hill-forts performed the changes occurring at this time. In this period, the changing processes seemed to have already been consolidated and are well documented. This is the first prominent note. Several Hill-forts had at the time impressive defensive systems with many wall lines and numerous occupied platforms. In some cases, they were highly populated and encompassed a proto-urban organization with well defined residential units and other attached equipment infrastructures bounded by walls and circulation axes. The consolidation of a habitat model accompanied by the presence of large oppida in the late first millennium BC is typical of the Entre-Douro-e-Minho and the Portuguese Minho region. This type of settlement prolonged itself further to the North towards the coastline and southern Galicia, mainly in the area around Rias Bajas (Martins 1990).

Lastly, there was a part of this region where one can find wide Hill-forts, mainly nearby flattened areas and these often boosted complex defensive systems. E. Carrocera Fernandez (1995) defined them as the region ‘el borde suroriental’. This area corresponded to the future Asturicensis Cenventus and mainly included circular houses with the exception of settlements located in the Lubian region (Zamora) and Corona de Corporales (Leon) that presented squared housing plants (Carrocera Fernandez 1995).

The refrerred changes inevitably brought social transformations to communities, mainly in terms of disseminating hierarchized societies (Silva 1986; Martins 1990) in some areas, or the idea of ‘potential social hierarchy’ to other areas (Orejas Saco del Valle 1996).

1.3 The Northwest in the late first millennium BC

In most part of the southern northwest, the social hierarchy process seemed to have been more present throughout the first century BC and was visible in the dependence relationship existing between Hill-forts (Martins 1990). The new type of settlements can be considered as true oppida. Some examples were the ‘citânia’ of Sanfins, Briteiros, Stª Luzia, Cividade of Ancora and Romariz amongst others. They presented a series of clear cultural indicators that were specific to this region, as for example, a proto-urban structure of space organization, thermal

As previously referred the last couple of centuries BC witnessed a set of changes profoundly shaping the peninsular northwest Iron Age societies. This period was normally understood to have taken place between the D. Julius Brutus expedition in 138/136 BC and the foundation of the three main northwest cities: Bracara Augusta,

23

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis infrastructures and the presence of warriors’ sculptures (Silva 1986).

roborated by the presence of varied imported materials with late-republican chronologies, by the adoption of new technologies, by the presence of coins and by the profound renovations of particular infrastructures. All these changes have been intensified throughout the first century BC (Silva 1986; Martins 1990; Almeida 1992).

In another northwest region and in part of the Meseta area integrated in the Asturias territory a social hierarchy process seemed to have taken place, extended itself towards the Douro basin. Here, a ‘celtiberization’ process occurred from the third/second centuries BC onwards and the appearance of important oppida was also registered. Some are known in Roman times due to their reference in Latin sources, as it is the case of Brigaecium, Baedunia and Lancia (Orejas Saco del Valle 1996; Esparza Arroyo 1986).

From late second centuries BC onwards the southern Galicia region alongside northern Portugal registered the emergence of very particular group of habitats, the wide oppida that shared amongst them monumental defensive systems, great population concentration, proto-urban organization, economical diversification and control over wider areas, in which other dependent Hill-forts existed. According to some authors, from the second century BC onwards, the premature contact of the referred region with Roman power might have accelerated these change processes (Silva 1986; Martins 1990; Martins 2005). This fact performed a distinguishing factor namely concerning the research based on cultural differences dependent on the Roman contacts and on Romanising agents.

The key changes should be framed upon non-exclusive patterns and should not be generalized and synchronized, automatically or diffusively towards nearby regions. In fact, in the last century BC wide northwest regions did not participated in this complex and social hierarchy processes featuring more exposed regions to the Roman world. The Trás-os-Montes and western Zamora regions represented an enclave location between the two previously defined areas. This region performed a continuum of the current archaeological area of ‘Las Medulas’. Hill-forts in these regions seemed to have been self-contained and self-sufficient in terms of survival goods. Each habitat encompassed their own independently controlled territory and there did not seem to have been any territorial expansion, population concentration or hierarchy amongst them (Lemos 1993). Inside the settlements there was an isolation of family groups and there did not seem to have exist any differentiation between them or any social hierarchy (Esaparza Arroyo 1986; Orejas Saco del Valle 1996).

The interpretation of available archaeological data is consistant with social complexity process of indigenous communities, as it has been stated in the study of other European regions (Drinkwater 1983; Millett 1990), where the contact with Roman power seemed to have played a key role in accelerating the development and change of the indigenous world. Nevertheless, these processes cannot be seen as simple imitation phenomena of Mediterranean models. D. Iunius Brutus’s expedition to the Entre-Douro-eMinho region in 138/136 BC should be considered a clear marker in the analysis of the evolution of indigenous communities located in the peninsular northwest in the period before our era, just as the critical importance of Caesar's expedition to Britannia in understanding the change processes taking place in the region until the conquest Claudius (Millett 1990).

Communities in other areas, as for example the ones located in the later entitled Asturia Transmontana presented similar development patterns. It is possible that these trends could have been applied to all western Leon area, as well, as in the current Viana do Bolo region. Moreover, we should highlight the fact that regional diversity seemed to dominate the Northwest in the late first millennium BC. The contact with the Roman power reached this region at different times and in different ways inevitably prompting different integration rhythms and indigenous community’s development patterns before the pacification process of Hispania carried out by Augusto.

1.4 Roman expeditions in northern Douro

The Roman landing in Emporion in 218 BC marked the beginning of the Roman interest in Hispania although this exact date does not imply a clear and planned stage conquest on behalf of Rome. The Emporion landing represented instead a war episode and signalled the first Roman incursion outside the existing Roman frontier: the Italic peninsula and Sicily. Under these circumstances, 218 BC performed a key date not only for the Hispania conquest but also for the Roman Empire overall expansion and history.

In this process, the southern Galicia region detained clear markers distinguishing it from the remaining other Northwest areas. However, it is important to highlight the greater diversity present in cultural areas later integrated in the Asturicensis and Lucensis Conventus. A key factor in this distinction, and one that allowed us to reference southern Galicia in opposition with the remaining Northwest, was the fact that there does not seem to have occurred a great resistance on behalf of these communities to the Roman presence. On the contrary, and after D. Iunius Brutus expedition the archaeological data has been revealing an acceptance from the communities in the area of the already pacified southern regions, a fact that prompted a series of changes. These have been cor-

In 218 BC Hispania was not a well-known land. A substantial part of its southern and eastern coastline might have only been known for existing Mediterranean populations, but all the remaining territory was only accountable in vague reports mentioning unimaginable richness alongside crude and barbaric communities.

24

The Northwest Integration within the Roman World

There are not many reports but we know the existence of premature Roman commercial exchanges with the eastern peninsular region, although we consider that some classic authors might have magnified the territory mining richness (Tranoy 1981)11.

Before these excursions throughout the peninsular northwest the territory was already part of proto-historic commercial routes. The northwest region was linked to long distance exchange networks with northern and eastern Europe, before the Late Bronze Age. These trading networks were extended to the Mediterranean, although the more active regions were located in southern Douro.

Up to the Carthaginian defeat a true organised Roman domain over the territory did not existed. Only the control over a small eastern peninsular area and only in 197 BC two praetors were nominated for the Citerior and Ulterior Hispania provinces. After dividing the peninsula in two parts the limits of both provinces were extended from east towards the west across two centuries. These advances came with the challenges inherent to land conquest and the contingencies of an internal political framework marked by internal crisis. The territorial domain took place without a truly organised administrative structured background. Hispania was only a part of the territory linked to the capital by the military forces due to its wealth, which was later exploited. It was a place of cultural and ethnical disparities and a privileged arena for civil Roman wars.

The Phoenician enclave of Santa Olaia, nearby the Mondego river valley underlined the importance of this route, which is believed to have been extended to the Northwest. Although the exchanges with the Mediterranean might have been sporadic, there were materials in Sardinia and Sicily that likely came from the Northwest (Gonzalez Ruibal 2006-2007). On the hand, archaeology has been exposing the presence of Phoenician and Greek artefacts in the northern Douro and Minho coastlines, especially in the Galician Rias Baixas area (Gonzalez Ruibal 2006-2007). From the second century BC onwards, but mainly in the first century BC the presence of Roman pottery in settlements located in the northern peninsula were predominant. This corroborated the idea that after D. Iunius Brutus‘s campaign and Publius Crassus’s excursion this region kept commercial exchanges with the southern peninsula region. These events also underlined the idea that this region was pacified and under Roman control at this time (Silva 1986; Martins et alii 2005; Morais 2005).

The Numancia victory and the Baleares conquest in 123 BC marked the beginning of a general space planning and rose the need for territorial control, which led in the first century BC to the foundation of urban centres, the opening of communication networks and towards a progressive interest in knowing the land, the communities and their wealth (Montenegro Duque et alii 1986; Fabião 1992; Richardson 1996).

2. Roman conquest and pacification

Between the Emporion landing and the arrival of Augusto in Tarraco the written sources informed us of four main expeditions to the northern Douro region.

Between the Emporion landing in 218 BC and the conquest and pacification of the peninsular northwest at the end of the first century BC two centuries unfolded and they were crucial to the history of Rome. Rome goes from being a civitas to be the capital of an empire. Such transformations encompassed varied changing processes and the use of certain regions as ‘laboratorial’ privileged spaces. This was the case with Hispania, without any doubt.

The Ulterior Hispania province governor, D. Iunius Brutus, conducted the first, between 138 and 136 BC. Other than the reference to other military events that took place in southern Douro, we are aware about the combat between Roman troops and the ‘Bracari’. This expedition had punitive aims alongside the intention of knowing the region. The Roman troops did not cross the Minho river or extended their visit to the current Trás-os-Montes area.

The landing of the Roman Empire in the Iberian Peninsula together with the effective conquest of the region shaped the concepts of ‘province’ and ‘empire’ that Rome applied to a vast conquered land.

Between 96 and 94 BC another Ulterior Hispania governor, Publius Crassus, performed a second excursion to the northern Douro region but this time with clear commercial interests. This excursion marked the opening of a commercial route between the south and the gallician region.

In the Hispania case, the conquest process and space appropriation took place alongside the evolution of these concepts. The evolution of semantics in the concept of ‘province’ allowed us to set the path that followed the process of space appropriation, on behalf of Rome. Initially, the concept only entitled the length of a magistrate’s responsibility. It evolved and started to include the overall administration of a conquered region and later acquired a circumscribed territorial dimension. This evolution directly accompanied the administrative organization of conquered areas, leading, in early first century AC to a progressive ‘crystallization’ of the term in its ‘territorial’ meaning (Nicolet 1984).

In 74 BC Perpena travelled to Cale and had the same punitive intentions. The acknowledgement of the territory and the opening of new commercial routes led Iulius Caesar to visit the gallician coast up to Brigantium (La Coruna) by boat between 61 and 60 BC (Tranoy 1981; Alarcão 1990b; Fabião 1992; Silva 2000).

11 The authors approaching the Hispanic mining wealth in general and the Northwest in particular were Estrabão when quoting Posidonius, Silius Italicus and Florus.

25

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

With the establishment of the Principate, Augusto had the power over a wide province encompassing a great part of the conquered Hispania territories. The definite conquest, in which Augusto was directly involved performed a triumph of his ‘empire’ and became the administrative organizational model for the territory. Initially, this model included the subdivision of the Empire into provinces and demanded other hierarchic organization levels giving birth to conventus and to civitates. To serve this successively fragmented structure of the territory in increasingly smaller units, there was the also need to fragment the capital power.

aspects that were mandatory of a revision of the Plinian list and the chronology of the foundation of the juridical conventus (Dopico Cainzos 1986; 1988). From the year eighth BC was an inscription (Rodriguez Colmenero et alii 1995) that dealt with a hospitality pact between the Lougeiorum Civitas and Caio Asinio Gallo. It referred the Arae Augustae conventus: ‘’... ex gente Asturum conventus Arae/Augustae/civitas Lougeiorum hospitium fecit ...’’. The reference about this new Ara Augusta conventus raised several interpretations. Dopico Cainzos (1986) suggested the conventus might have preceded the Asturicensis and that its capital might have been located on the site of the future Asturica Augusta. G. Pereira Menault (1983) corroborated the above statements.

The Roman foundation of coloniae and municipia and the urbanization of the conquered regions performed key structuring strategies of the new political and administrative order. Other key tool also served this new order: the development of a communication network linking the urban centres and ensuring the effective functioning of the overall system.

Rodriguez Colmenero presented a rather different interpretation. The author believed that in a first moment the Gallaecia and Asturia regions were divided in the Gallaecia Bracarense with the capital in Bracara Augusta, in the Gallaecia Lucense with the capital in Lucus Augusti, in the Asturia Augustana with the capital in Asturica Augusta and in the Asturia Transmontana or maritime with the capital in Ara Augusta. The author considered the hypothesis that Ara Augusta may be the current Campa Torres, nearby Gijon.

2.1. The campaigns of Augusto in the Northwest

Augusto settled in Tarraco in 27BC. In the following year he prepared the war against Cantabria and Asturias, focusing the army in a combat line centred in Segisama (Sasamón), located in the eastern part of the current Burgos province. Publius Carisius, legacy of Lusitania, settled nearby the future Asturica Augusta, a site where part of the Citerior army camps was also positioned.

In the first century AD the area of that initial primitive convent may have been integrated in the Asturicensis Conventus. This may explain its absence from the Plinian list that could also be related with the short duration of that convent effective existence (Rodriguez Colmenero et alii, s.d.; Rodriguez Colmenero 1996a; Rodriguez Colmenero 1996b).

The first phase of the conquest took place in the year 25 BC and included several combat fronts. Caio Antistio Veto won the war against the Cantabria and established a camp in Lugo from where future attacks were conducted.

C. Fernandez Ochoa and A. Morilo Cerdan had a different interpretation about the geography of the Ara Augusta conventus. For these authors this conventus corresponded to a first organizational stage, which included the peninsular northwest region under military control and the area of the future Asturica Augusta. This first administrative outline performed the basis from which the future Asturicensis, Lucensis and Bracarensis conventus would emerge. The division probably took place in the late Augustan period and the beginning of the Tiberius’ reign (Fernandez Ochoa and Morillo Cerdan 1999).

There were successive risings in 24, 22 and 19 BC, the latter being controlled by Marco Vipsanio Agrippa. This victory marked the political-military control of Hispania by Rome, although there was a rebellion around 16 BC marking the end of hostilities. Agrippa was responsible for the first attempt in organizing the new territories within the administrative peninsular map. The Lusitania frontier was extended to the Cantabria region that started to include the Asturia and Gallaecia regions. Nevertheless, the development of this first political and administrative map naturally suffered later adjustments in its boundaries.

2.2. Organization of the territory

The Lusitania frontier progressively retreated towards the south, and was then located in the Douro region determining the integration of the Citerior or Tarraconensis northwest province. This province ended up being divided into juridical conventus as referred by Plinio in 70 AD.

2.2.1 A new social geography: populi and civitates

2.2.1.1 Textual sources

Although traditional historiography considered the conventus to have been a Flavian creation (Tranoy 1981; Alarcão 1990b), we believe that they were Augustan.

The main sources behind the study of the population living in the peninsular northwest were the works of Estrabão, Ptolomeu, Plínio, Pomponio Mela, Floro, Dion Cassio and Orosio (cf. Sources Hispaniae Antiquae).

The discovery of the Tabula Lougeiorum uncovered in the Caurel region and currently located in the San Antón Museum in La Coruña shed some light on very important 26

The Northwest Integration within the Roman World

Regardless of the existence of several authors, the provided information was scarce and often contradictory. Works were conducted at a certain time and for varied ends and they should be read taking into account their ‘production aim’ that was often propagandistic. In this line of thought and taking into consideration the Cantabrian wars, A. Morillo Cerdan (1995) stated that the reports ‘’actuaron como una caja de resonancia de las virtudes y excelências del princeps’’. To the author the conquest challenges and the virtues of Augusto were the relevant subject matters.

the social organization of pre-Roman communities. Often, they confirmed certain references only known through the works of Latin authors. In other cases, they added data or suggested communities’ territorial boundaries. Other than the ‘’Padrão dos Povos’’, found in Chaves, well known and analysed (Tranoy 1981) there were other key epigraphic documents that are key to reference. An important source towards the study of the familiar units has been the entitled ‘’Zoelas Pact’’, an inscription found in Astorga in which a hospitality pact was renewed. The pact is dated from 27 AD and referes to several gentilitates belonging to the Zoelas gens, framed within the populus of Astures. In 152 AC the pact was enlarged to include not only other gentilitates of the Zoelas but also other gentes of Orniacos (Morillo Cerdán 1995).

The pax Augusta reached the ends of Hispania and both the literature and iconography served the imposition practices of a new regime and the necessary glorification of their author. The main source contributing to the knowledge of the indigenous populations was the Geographia III by Estrabão, a document written between 29 and 7 BC and reviewed until 18 AD.

The entitled ‘’Tabula O Caurel’’ found in Torre de Cabreira, O Caurel, dated from 28BC (Colmenero 1996a), mentioned a hospitality pact celebrated between a private and the castellani Tolentenses. Adding to this reference, the origin of Tilego was included: “...Tillegus Amabati f(ilius) Susarrus/ ) Aiobaigiaeco hospitium fecit cum Lougeis Castellanis/Toletensibus(...) clientelamque sua/m suorumque in perpetuo cas/tellani Tolentenses receperunt (...)”. Not less interesting was the reference to the judges, of indigenous origin, present – Latino (son of) Ari and Aio (son of) Temari – and to the practicing Roman consulates – Appio lunio Silano and P. Silio Nerva.

Contemporary of Augusto, the work of Estrabão was very informative about the way we see and understand the Roman world as it was structured around the opposition of the Greek/Barbarians or Civilization/Barbarism. This opposition was rather cultural and not racist, in the sense of how we understand it today (Nicolet 1984). It was decisive in its political inventorying of the known world, contributing towards the knowledge of the spatial distribution of communities and their wealth against the impact of the Roman expansion and their modelling effects.

The entitled ‘Édito del Bierzo’ was found in 1999 and was often referenced due to its relevance towards the mentioning of the ‘transduriana’ province (Transdurianae provinciae), unknown before this bronze tabula was found. Nevertheless, part of the text mentioned the castellani Paemeiobrigenses (castellanos Paemeiobrigenses), its gens Susarri (gens Susarrorvm), and contained explicit references with regards to the question of land ownership and its boundaries.

The expression ‘space obsession’ suggested by Claude Nicolet (1984) and related with Augustan politics materialized with the public stage of a map produced by Agripa and painted in a portico, on a location close to where the Res Gestae tabulae was to be placed, right next to the Mausoleum of Augusto.12 Although its nature and aims written sources provided a particular type of knowledge about the indigenous reality the Romans were confronted with. They referred the list of the populi living in the Northwest, provided us with an idea about the authors’ vision regarding important aspects of these communities’ way of life, namely their political and social organization.

There were other epigraphic sources that enable us to understand territory boundaries. We are naturally referring to the termini. The Augustan terminus of Santa Comba de la Vega established the territory of Cohors IV Gallorum in relation with the city of Luggones. Other inscription concerning Astures and Luggones has been interpreted as a landmark boundary between the asture tribe of Luggones and the Cantabrian (Morillo Cerdán 1995).

2.2.1.2 Epigraphic data

Within the region of the conventus bracarensis some inscriptions can also be found and they might have been termini of community settlements.13Admitting this interpretation, we should take into consideration the problem of the territory controlled by a settlement, as well as the chronology to which this delimitation referred. The men-

Epigraphic sources are of extreme relevance with regards the knowledge and analysis of several issues related with 12 As referred by C. Nicolet (1988:20-21) “(...) l’oeuvre posthume de l’empereur Auguste (…) ses Res Gestae, peut être considéré, dans sa seconde moitié au moins, comme une véritable somme géographique (...) le texte appararaît presque comme le commentaire d’une carte, on croirait qu’il demande lui-même à être suivi sur le dessin. Or ce dessin, en effet (la carte d’Agrippa), n’était pas três loin: le portique qui le contenait se trouvait à quelques centaines de mètres du Mausolée”.

13 To have an idea about the quantity of inscriptions interpreted as termini and specially, in terms of the difficulties in reading and interpreting, them we should reference the corpus presented by Rodriguez Colmenero (1995) at the Pre-Historic Epigraphy Symposium in Santiago de Compostela in 1992, together with the challenges faced by pre-historic epigraphy in Portugal presented by J. Encarnação (1995) at the same scientific Symposium.

27

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

tioned termini seemed to perform, in several cases, the population of a Hill-fort, as the references appeared in nominative and not in genitive, as it would be normal.

might have occurred before the Roman domain and it could have been ratified through inscriptions within the new administrative Roman framework or be related with the territorial reorganization process established with Augusto.

The lack of certainties with regards to the interpretation of these inscriptions are related with the difficulty in knowing if we are facing a ‘toponymic’ or an ‘antroponimic’ reference (Rodriguez Colmenero 1993; Gonzalez Ruibal 2006-2007).

2.2.1.3 Populi and civitates

One of the most important inscriptions of this kind was found in a rock located in Valdecastiñeiro, Ourense, which was believed to delimit the castellum INT … within the Interamniense territory, in the Gallaecia Bracarense. The inscription – STP/) INT COM CI I was rather controversial. G. Pereira Menault (1983) suggested the following interpretation: S(altus) T(erritorium) P(ublicum)/ ) (castelli) INT(er) COM(prehensus) CI(vitate) I(nteramniense?).

One of the most referenced excerpts of the book II of Naturalis Historia by Plinio has been the one referencing some of the existing populi within the peninsular northwest, providing the number of the free population existing at the time: ‘Nam in Cantabricis VII populis Iuliobriga sola memoratur (…)" (NH III 27). “Iunguntur iis Asturum XXII populi divisi in Augustanos et Transmontanos Asturica urbe magnifica: in iis sunt Gigurri, Paesici, Lancienses, Zoelae. Numerus omnis multitudinis ad CCXL liberum capitum. Lucensis conventus populorum est sedecim, praeter Celticos et Lamavos ignobilium ac barbarae appellationis, sed liberum capitum ferme CLXVI. Simili modo Bracarum XXIIII civitates CCLXXXV capitum, ex quibus praeter ipsos Bracaros Bibali, Coelerni, Callaeci, Equaesi, Limici, Querquerni citra fastidium nominentur’ (NH III 4, 28).

A. Rodriguez Colmenero (1977, 1987-1988, 1988, 1995) presented several other interpreting possibilities. In the author’s last work the proposal was the following: S(tatuerunt) T(erminum) P(onere)/ CROBRI (?) (et) TRI (?) ) (cognationes)/ INT(er) COMO. I(nter) EBO (ou Ebno?) (Rodriguez Colmenero 1995).14 Another monument found in Alvarelhos and related with the known inscription MADE/QVIS(enses) STATV/ERVNT/ LADRO/NO CA/MALI/ F(itio) ANTONIO/ ALMO has been prompting different interpretations. For A. Tranoy this monument can be interpreted as the territorial boundary of the Madequisenses, established by the representatives of both this group and of another one, not mentioned (Tranoy 1981). A. F. Silva (1986?) suggested the Madequisenses might have dedicated the monument to Ladronus, son of Camalus. González Ruibal (2006-2007) believed the A. Tranoy version was more plausible as the verb statuere was not normally used in monuments dedicated to a person.

In this referred excerpt, Plinio spoke of the seven populi in Cantabria and twenty-two populi in Astures, who were divided in ‘transmontanos’ and ‘augustanos’, the 16 populi in conventus lucensis and twenty-four civitates in the bracaraugustanus one. The use of the terms populi and civitates within the same excerpt, was rather curious, with the author apparently mentioning the same reality. However, in some passages the work of the naturalist mentioned both terms in different ways. For this reason, this question seemed to be related with the diversity of the used formulae on the plinian text, even if only due to the use of different sources from different times.

Another known inscription is referred to a Milestone located in the Chaves Museum, which delimited the boundaries between two castella territories. On one side it contained the word PRAEN and on the opposite side the inscription COROQ… (CIL II 2489; Tranoy 1981; Rodríguez Colmenero 1996; Rodríguez Colmenero s.d).

The formulae varied in accordance with the used sources being possible to sate a rough chronology: the description of conventus in Lusitania would be from Augustan times; between Augusto and Claudio would be the descriptive formulae for Baetica, whereas the Citerior would have had as main sources and formulae the ones from the Claudio’s era (Rodríguez Colmenero 1996).

Taking into account the interpretation suggested by A. Tranoy with regards to the Alvarelhos inscription, the castella should have had legal and political powers to delimit their territories and to establish pacts within certain limits under Roman law (Gonzalez Ruibal 20062007). However, and considering the termini are dated from Roman times, the problem with regards to the delimitation of the castella territories’ chronology persisted. It

This different ‘times’ in used sources by Plinio, contained in his works as different descriptive formulae, enabled us to explain the apparent inconsistencies of the author with regards to the author referencing the concepts populus / civitas or populus / oppidum.

14 According to the author, this reinterpretation was developed after the item was studied. The author read the words CROBRI and TRI. Based upon this reading, Rodriguez Colmenero translated the inscription as the following: "(…) Determinaron poner este mojón terminal las parentelas de Crobros y Tri…(?) entre Como e Ebo?’’ (Rodríguez Colmenero 1995: 128).

28

The Northwest Integration within the Roman World

lands15. Sources referred nine communities: Orgenomescos, Salaenos, Vadinienses, Avariginos, Concanos, Tamaricos or Camaricos, Coniacos or Coniscos. Estrabão mentioned the Plentuisos, whereas some other authors suggested the Coniacos or Coniscos and the Concanos were the same gens. Also curious is the dubious mentioning of the Blendios (Morillo Cerdán 1995).

However, the relevant question were the different postures related with the problem of the pre-Roman and Roman organization of the region, in which, the interpretation of Plinio’s concepts or formulae may serve as a reference. The key issue was related with understanding whether the terms populus and civitas, in the sense of space organization, referred as the base for pre-Roman groups or were instead administrative realities imposed by the Roman power. The opinions on the matter differ considerably. In fact, several authors believe that despite the different terminology utilized, the concepts referred to identical administrative realities (Rodríguez Colmenero 1996a) fully related with previous pre-Roman communities. Other researchers, based on the same premise, stated that the concepts were applied to realities unrelated with the indigenous communities and were administrative and political Roman realities imposed by the imperial power, in order to organize the management and exploitation of the territory (Fernández-Posse and Sánchez Palencia 1996; Sastre Prats 1998).

With regards to the Astures, Plinio suggested twenty-two populi16. Within the Astures Transmontanos, the Pesicos and the Lungones or Lugones were registered. For the Astures Augustanos, there was the reference to Gigurros, Amacos, Luggones, Zoelas, Superatios, Brigaecinos, Bedunienses, Orniacos, Lancienses and Saelinos (Morillo Cardá 1995). The Orniacos featured in the Zoelas Pact and Ptolomeu only mentioned the Saelinos. The Zoelas were a neighbouring people of the ‘galaics’ according to a reference by Plinio (NH XIX 10). Epigraphy placed them within the bordering area between Portugal and Zamora (Morillo Cerdán 1995). Within the Astures Augustanos area some probable populi capitals were identified: Petavonium of the Superatios; Fuentes de Ropel (Zamora) of the Brigaecinos; Bedunia has been identified as San Martín de Torres of the Bedunienses and Lancia of the Lancienses.

In sum, one might admit that such differences might have been intentional. The ‘interpolate’ mention to ‘’Bracarum XXIIII civitates’’ for the Bracarugustanus territory could have been the reflex of a different social reality within the Northwest, a reality which archaeological data sustained and Plinio might have wanted to underline. This interpretation has been suggested by both Santos Yanguas (1985) and González Ruibal (2006-2007).

In the Roman area of the Conventus Lucensis, the naturalist author mentioned sixteen populi (N.H. III 4, 28). From all these, it was only possible to identify the following: Neri, Supertamarici, Praestamarci, Cileni, Arrotrebae, Copori, Arroni, Addovi, Varri Namarini, Cibarci, Egi, Albiones, Seurri and Lemavi, other than the Baedi, whose identification still rises doubts (Rodríguez Colmenero s.d).

Based on Plinio’s work, and adding on to the references by Ptolomeu and Estrabão and to the epigraphic references several authors attempted to define the geography of these populi within the referred region.

With regards to the Conventus Bracarensis region, the same author referred to twenty-four civitates, from which only a few numbers were referenced.

The Cantabros and Astures were located within the farther northern side of the Iberia Peninsula, in both sides of the Cantabria mountain range. Cantabria was the name given by Roman geographers to the ocean sector, between the high course of the Minho River and the Mare Cantabrum, and between Bidasoa and the Atlantic Ocean. The Astures were linked with the Cantabros and located between the Bierzo and the Galician-Leonese mountains. The region, to become the Conventus Asturum, was located between the Cantabria and the Douro region, referring in terms of present geography to the larger region of Zamora. It included small sectors of both the Trás-os-Montes province and the Ourense region (Morillo Cerdán 1995). This wide territory was normally divided, with regards to the populi and also in terms of ethnographic possibilities, in three areas: the Cantabros, the Astures Transmontanos and the Astures Augustanos.

The Plinio’s work together with the remaining written sources (Estrabão and Ptolomeu) and the epigraphic data allowed us to recognize the name of twenty-one of the twenty-four mentioned civitates: Amphilochi, Aulobrigenses, Bibali, Bracari, Callaeci, Coelerni, Equaesi, Grovii, Helenni, Interamici, Leuni, Limici, Luanci, Lubaeni,

15 A. Morillo Cerdán (1995:47) stated the following with regards to Cantabrian literary references: "en este espacio geográfico, que coincide a grandes rasgos con la actual Cantábria y las tierras montañosas limítrofes, habitaban diversas etnias. Las Fuentes mencionan la existencia de siete gentes (NH II 27), aunque sólo conocemos el nombre de algunas (…). Muchas de estas tribus parecen corresponderse directamente con las civitates citadas por Plinio (NH IV 110) y Ptolomeo (Geog. II, 6, 50)”. The term gentes may lead us towards the wrong direction, as in the same excerpt we also read gentes for the Cantabrian, populi for the Astures and Lucensis and civitates for the Bracari. We rather use the Plinio reference alongside another group of references from known sources and thus, accounting 9 instead of 7 references. 16 Plinio referred 22 populi, however only mentioned the Gigurros,

Estrabão located the Astures between the Douro and Cantabria areas. The author listed them after the Galician and before the Cantabros (Geographia III, 3.8), events confirmed by Pomponio Mela (Chorographia 13-14). The geographic space occupied by the Cantabros matched the present Cantabria and its surrounding mountainous

Péscicos, Lancienses and Zoelas (N. H. III 28) adding in another work the location of the Pesicos in a peninsula (N. H. IV 111). In the second century, Ptolemy listed the civitates of the Begaecinos, Beduneoenses, Orniacos, Lungones, Saelinos, Superatios, Amacos, Tiburros and Figurros (Geog, II, 6, 28-37) alongside the Pesicos in the coastline (Geog, II, 6, 6).

29

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Naebisoci, Narbasi, Nemetati, Quarquerni, Tamagani, Turodi (or Aquiflavienses).

Seurbi,

sustaining a hierarchical social organization, was fully structured right before the end of the millennium BC. This idea is behind the understanding of the civitates model that comprised well defined and fully consolidated ethnic territories. The Roman administration could have taken advantage of the pre-existing indigenous structures in order to define administrative entities, linking this region with the general model applied by the Roman Empire to other western provinces.

Some authors attempted to establish the mapping of these civitates, although a few discrepancies have been registered in the presented proposals (Tranoy 1981; Silva 1986; Alarcão 1988; 1992; 1995-96; Rodríguez Colmenero 1996a). Other than interpretation issues already approached with regards to the Plinio’s text, the question persisted with the remaining sources. The information was contained in a group of documents, both texts and inscriptions of Roman origin and dating back to mid first century AD.

However, it might be legitimate to consider that the Roman administration might have imposed delimitation in some civitates that may not match to previous ethnic and political units, namely in areas of the Northwest revealing different socio-cultural and economic indicators.

Although it might not be possible to fully understand the overall territorial organization, the Roman reports containing the known geography previous to the occupation was rather obvious and consensual. The debate was then placed on another level and linked with the need to know if that organization reflected, with more or less changes, a pre-Roman structure and if, by the contrary, was not already a product of the Roman administrative management over the occupied territories. I. Ines Sastre Prats (1998, 2001) defended the most radical ideas over the topic based upon the archeological works carried out in Asturias (Fernández-Posse and Sánchez Palencia Ramos 1996). For the author, both terms – populi and civitates – were applied to the territory delimited by the Romans with purely administrative and tax revenue aims and were not correspondent to a previous ethnic and political formation. Following this, the civitates were in place within this region simply as territorial boundaries needed to reorganize the indigenous communities by the Romans and were meant to facilitate the control and management of the three conventus capitals.

This idea seems particular applicable to the Asturias region, when considering the organization that has been defended in its key studies (Sastre Prats 1998, 2001; Orejas Saco del Valle Sánchez-Palencia 2002). The strong self-sufficiency of the local communities, namely sustained by a domestic economic model, together with the absence of social stratification and of any other settlement hierarchy, performed a development framework entirely different from what is registered on the western and southern northwest, throughout the same period. Moreover, it seems reasonable to suggest that the management and organization of the territorial units within the Asturias region developed by the Romans, do not actually correspond to any pre-Roman reality, having instead been imposed on communities and aimed at exploiting resources and taxes, as it has been defended by Sastre Prats (1998; 2001). We could be facing different realities that have to be regionally valued and separately researched.

2.2.2 Power centres: the urban politics of Augusto

Nevertheless, other authors considered that the civitates overlapped the ethnic pre-Roman territorial structures, even though it was difficult to conceptualize their exact meaning (Tranoy 1981; Martins et alii 2005). In sum, this was the general idea behind what might have equally happened within the Gallia and the Britannia regions (Drinkwater 1983; Millett 1990).

Having drawn the administrative structure it was important to establish power centres linked by a transport network system, so that an effective control could be achieved within the conquered territory, in order to outline its economic exploitation efficiency.

The Roman power would have, then, used the peninsular northwest region alongside other Empire areas, as preexisting indigenous structures to support and enable the establishment their own administrative structure.

The choice over the places that would become the centres of Roman power was crucial within a territory without cities. For this reason, Augusto established three urban centres within the peninsular northwest that started functioning as administrative capitals. These events were paramount towards the pacification, control and organization of this region.

Considering that the social pre-Roman reality was different within the territories matching the three northwest conventus, each reaching different levels of development as revealed by archaeology, we can suggest distinct actions to have taken place within the territory organization on behalf of the Romans, although it seems paramount to highlight the need to carry out further analysis focused on this challenges.

Even though the establishment of these urban centres was the probable result of a political decision that could have occurred, more or less simultaneously, around the year 16 BC, the history, features and development of each of these capitals occurred differently due to the different cultural pre-Roman communities of the region in which they were located (Martins et alii 2005).

Taking into consideration the available data it seemed plausible to suggest that the western area of the bracarense conventus, clearly acknowledge by several researchers as the more ‘developed and dynamic’ area and 30

The Northwest Integration within the Roman World

Before proceeding with the study of each of these cities from an urban morphology point of view, we will establish a few convergences and differences that have been more of less evident.

Archaeological works carried out in Braga within the last few decades have enabled us to understand the importance and dimension of an ever-growing city. Bracara Augusta presented an orthogonal layout according to the main orientation axes, NNW/SSE and SWW/NEE, with square modules of 150 feet defining quarters, with the latter possessing a construction area of around 1 square actus (Martins 1999; Martins 2004). This urban plan was recovered from walls and from the study of network infrastructures like streets and porticos together with the understanding of the cloaca orientation, data that enable to put forward a theoretical proposal of the city plan in 1990 (Martins and Delgado 1989-90a). Archaeological works carried out ever since has confirmed this proposal.

What were the main converging points? Bracara Augusta, Asturica Augusta and Lucus Augusti shared amongst themselves, other than a chronological similar foundational date, a similar function within the Tarraconensis province administrative framework. They were all conventus capitals of probably juridical origin under Roman law (le Roux 1994). They were planned urban speces with centralized functions as road network centres within the territory. What were the key differences between the three capitals? From start, the social fabric composition, but also the different role each had within the general imperial political framework with regards to the Hispania northwest region.

The dominant orientation axes recovered from a series of excavations after 1976 have been fully confirmed with an insulae that was located in the joining terrains of the ‘Antigo Albergue Distrital’. A cloaca was also found on site and these events have enabled us to state that the city had sanitary infrastructures from the Julian-claudian period.

Asturica Augusta was a city strongly shaped by the presence of administrative and military agents linked with gold exploitations of the heart of the northwest mining region. This fact was well documented by epigraphic sources, which revealed a large predominance of Roman citizens within the city social composition (Tranoy 1986; Iglesias and Ruiz 1998).

It is believed the western decumanus maximus would follow on the trace of the present S. Sebastião Street that culminates on the eastern side of the Paulo Orósio square. The eastern part of the decumanus maximus seemed to match the current design of the Alcaide Street, passing through the Santiago square and approximately following the current Falcões street (Martins 2005). This outline is today fully confirmed with the orientation of the Via XVII orientation found in the excavations of CTT quarter in Braga, which prolonged the overall decumanus maximus axis (Martins et alii 2010; Fontes et alii 2010a; 2010b;).

Bracara Augusta on the other hand, performed a city with an indigenous elite promoted from existing pre-Roman communities and revealed a weak military presence (Tranoy and Le Roux 1989-90). The city witnessed a florescent economic life and had a strong interaction with its surrounding territory, partially ensured by its strategic location as a crossroad of terrestrial, fluvial and maritime communication circuits, which strengthened its preferential link with Baetica and Lusitania (Morais 2005).

On the northern part of the city, the cardus maximus seems to be represented by the remains of a large street found at the excavation of the Albergue Distrital area, underneath which an important cloaca was discovered (Lemos and Leite 2000; Martins et alii 2012; Martins and Ribeiro 2013).

Lastly, Lucus Augusti seemed to have fundamentally had an integration and consolidation role of the Roman power within the region, which, generically allowed the definition of the city role between the one carried out by Asturica in administrative terms (Tranoy 1986; Rabanal 1988) and Bracara in commercial terms (Martins 2005; Morais 2005).

In the southern city area the only axis with an orientation of NNW/SSE matched the current Bombeiros Voluntários Street that follows up to the crossroad with the Damião de Góis Street. This allowed the assumption that the cardus maximus should follow the same orientation (Martins 2004).

Moreover, we hoped to understand the structure of each of these cities, taking into consideration the new knowledge framework. The overall aim was to highlight the main value of the Augustan intervention programme when establishing urban centres in the peninsular northwest

The limits of the city were always developed with a wall in mind. This wall has been referred in documents from the 12th and 13th centuries and described by scholars between the seventeenth and eighteenth centuries (Cunha 1634-1635; Argote 1732-1734).

Bracara Augusta was, as much as the archaeological data enabled us to reveal, an ex novo foundation. It was erected between the Cávado and Este rivers’ fluvial area. The choice over the city location seemed to have been determined by the central position of its site in relation with the indigenous communities of the region, but also linked with the possibilities the region offered as a junction area of several pre-Roman communication axes that articulated the Atlantic coast with both the remaining Northwest territory and the already pacified world in the southern peninsula (Martins 1990; Lemos 2002; Morais 2005).

In 1910, J. Teixeira developed the first urban map of Braga where the city was bounded by a wall (Oliveira 1978). The trace has generically been confirmed by excavation works conducted throughout the past few decades (Lemos et alii 2007). The archaeological remains suggest the Roman wall was erected in the late third century and early fourth. It defined an area comprising 48 hectares and leaving out some buildings of the previous city. Some

31

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

light still needs to be shed on the possibility of Bracara Augusta having had a foundational wall a possibility already suggested but never confirmed (Martins and Delgado 1989-90a). The cartography of the axes corresponding to the city orthogonal plan suggested that the planned area did not encompass the totality of the urban area that was defended by the Low Empire wall (Martins 2004). A few constructions were detected with different alignments from the urban plan axes, some of which were associated with crafts areas that might have developed in the surroundings, so expanding the city area that reached its maximum extension during the 2nd century (Martins 2004). The density and the degree of development of the preRoman communities existing within the region where the city was established may explain the predominance of indigenous people within the city social composition, the small representation of Roman and military citizens and the dominant presence of pilgrimages, free citizens and slaves (Tranoy 1981; Tranoy and Le Roux 1989-1990). Bracara Augusta witnessed, from an early stage a considerable commercial dynamism, corroborated by epigraphy, by the type of importing products and by its own local production that would certainly developed the region (Morais 1998, 2001; Cruz 2001).

where the Legio X Gemina camped. The passage from military camp to a uraban centre was clearly framed within the project carried out by Augusto aimed at establishing three cities within the Northwest following the Cantabria wars (Sevillano Fuertes and Vidal Encinas 2002). Urban archaeology in Astorga was developed in the 1970s and ensured the conditions needed to both exhume several infrastructures of the Roman city and uncover the respective orthogonal plan. The end results of around 30 years of consecutive excavations impacted upon the acknowledgement of the urban area and the city architecture, placing Asturica Augusta, similarly to Bracara and Lucus within the group of planned cities possessing a high level of Romanization (Garcia Marcos and Vidal Encinas 1998). The existence of a permanent military camp within the first occupational stage of the future city of Asturica is an archaeological fact corroborated by the existence of two trenches facing NW/SE and associated with a stone structure (González Fernández 1999), which is interpreted as part of a primitive fortification 17. Other remains of this camp were documented on a series of orthogonal structures associated with trenches and terrain level zones found on the first occupational level of the entitled ‘’Casa do Pavimento de Opus Signium’’ (Opus Signium Pavement House). These structures preceded a domus erected in Julian-Favian times and could, according to M. Burón Álvarez (1998) have been supplemented by walls of stucco, adobe and wood and belong to a primitive military camp.

A key element linked with the economic development of the city was the role that the city must have had as a controlling site over the mining exploitation in the conventus bracarensis territory (Lemos and Morais 2004; Lemos 2004d; Martins 2005). Contrary to what happened in the remaining Augustan sites in the peninsular northwest, Bracara Augusta developed an important leading role in the late third century, following on the reforms carried out by Diocletian, when promoting the city as the capital of the new Galician province (Tranoy 1981).

Between the late first century BC and the beginning of the first AC, Asturica went from being a military camp to becoming a city. It went through an orthogonal planning adapted to the local topographic features. Adding to this, the western city sector was dominated by axes running NW/SE, whereas the eastern sector was organized according to axes running N/S and fully adapted to the area features (Garcia Marcos and Vidal Encinas 1998).

This promotion, archaeologically documented by an outbreak of urban renewals occurring in both public and private buildings, was related with an overall improvement of the administrative role of the city. These events likely justified the continuity and fortification of a significant urban area that showed no signs of population retraction in opposition of what occurred in several other Hispania cities over the same period (Arce 1982).

The construction of a series of urban infrastructures was initiated namely with the development of the road network, public buildings and wealthy private residences construction. The city sanitary infrastructure was also established and encompassed a wide cloaca network, a construction dating back to the beginning of the 1 st century (Sevillano Fuertes and Vidal Encinas 2002). A prime example of the city’s richness happened at the Julius-Claudius era with the establishment of the ‘’Opus Signinum Pavement House’’ (Burón Álvarez 1997). From the domus built at that time it was possible to identify

The gradual use of necropolis and the maintenance of steady rhythms in products importing from the Mediterranean area between the fourth to sixth centuries revealed a continuity of the city administrative role and the permanence of a religious elite linked with the functions the city had as a province and episcopal capital (Martins 2004).

17 The uncovered excavated materials were from the end of the first century BC / beginning of the first century AD and included lamps, thin wall ceramics, italic sigillata, coins, glasses and an important set of wine amphorae coming from the eastern Mediterranean. The presence of the latter was rather scarce in the Italic peninsula and some examples known to be located in Baetica. Other remains found include indigenous ceramics and Roman common productions (González Fernandez 1999). The primitive wall seemed to have existed only for a short period of time as it has been corroborated by the edification of a domus in the late first century. This structure demanded that part of the primitive defensive wall was destroyed (Garcia Marcos and Vidal Encinas 1998).

As much as archaeology can attest, Bracara Augusta maintained and reinforced its importance as an urban centre, even after the Suevi location in the region, being able to keep its cultural leading role in Late Antiquity (Martins 2004; Fontes et alii 2010;). Referred by Plinius, the Elder as urbs magnifica, Asturica Augusta had a military origin, attested by a military camp

32

The Northwest Integration within the Roman World

twelve areas organised around a central space - a potential atrium. The most important room, also providing its name to the house, is a triclinium with a pavement in opus signinum, holding a group of black and white tesselae forming ornamental patterns18. This pavement performed the western example, within the Iberia peninsula of this type of ground, and the part with the later chronology refers to changes conducted in the Tiberius / Claudius times (Burón Alvarez 1998).

nications network also led towards the city being surrounded by an impressive wall with semi-circular towers with a similar chronology to the other urban centres in the Northwest (Sevillano Fuertes 1999; 2007). In 465, Teodorico plundered the city starting the beginning of its decline and depopulation (Garcia Marcos and Vidal Fuertes 1998). The city of Lucus Augusti was erected in a platform registering a slight declination towards the southwest and was surrounded by wide horizons reaching the coastline regions. This was a strategically relevant road network location (Rodríguez Colmenero 1995).

The forum was located at one of the city’s highest areas and is articulated both the western and eastern urban city sectors. According to the latest know interpretations this space was organized around a wide area housing an ambulacrum with a two-stored arcade, initially believed to have been part of a basilica (García Marcos and Vidal Encinas 1995). On the western axis a rectangular building was found ending on a semi-circular apse, interpreted as being an aedes Augusti. The western area of the forum was organised around a monumental structure known as ‘’Ergástula’’, interpreted by Garcia Marcos and Vidal Encinas (1998) as ‘’la substructure’’ of a majestic building located on the eastern part of the square.

The hypothesis that the city evolved from a pre-Roman setting has not been well accepted between researchers working on the Roman city, despite the discovery of three huts in R. Mourelo Avenue and the indigenous pottery found throughout the city. 19 The stated evidences have been interpreted as the result of an indigenous occupation on site at the first organisational stage of the city foundation and not necessarily a clear indicator of the existence of a previous settlement. The site lacks indigenous construction evidence and the topographic features of the location do not corroborate this conclusion. (Rodríguez Colmenero 1995).

The city performed a strategic location to control de heart of the Northwest gold-mine region and was a key road platform to export products. Asturica Augusta reached its peak during the intense exploitation of the mining resources existing in the region throughout the first two centuries AD.

The traditional hypothesis conjured by A. Schulten (1943) with regards to the military origin of the northwest cities seemed to be accepted in the case of Lugo, due to the presence of a military camp established before the city foundation. This was corroborated by the finding of over 70 ‘caetra’ coins of crafted on location and traditionally attributed to the Publio Carisio camp (Centeno 1986).

Throughout the High Empire the city was a key administrative urban centre where personalities with important senior positions interacted, namely procuratores augusti, legati juridici and positions linked with the military, finance, recruiting, census and taxation (Tranoy 1981).

Furthermore, the L. VI inscription also supported the existence of a camping site in Lugo, inscription which was interpreted as L(egio) VI, one of the legions responsible for the northwest conquest. Other factors supporting this idea can be found in older archaeological remains, mainly a group of trenches associated with the camping site structure (Rodríguez Colmenero et alii 1995; Rodríguez Colmenero and Covadonga Carreño 1999). The military origin of Lugo would, then, had been the result of the installation on site of the Caio Antistio Veto camping site, at around 25 BC, after the Roman victory over the Cantabrian. The original military function of the site was lost with the end of the wars and with the overall administrative reorganization of the territory carried out by Augusto, from 16 BC onwards (Tranoy 1981).

The city social fabric revealed by its inscriptions, registered a high number of immigrants coming from all Hispania, as well as from other areas of the Empire, namely from the west (Tranoy 1981; 1986). Due to the socio-economic level of the living population, the city met several urban renewal programmes when a great number of wealthy residencies were built (Burón Alvarez 1997). The downfall of the gold-mining exploitation throughout the 3rd century AD led towards a visible city decline and prompted Asturica to lose its key role as a strategic centre within the Northwest in favour of Bracara Augusta, although the city resisted these trends throughout the fourth and fifth centuries.

Thus, Lugo became a city and its patron was Paulus Fabius Maximus the person to whom the necessary religious power to head rites and perform foundational ceremonies is believed to have been delegated. Similar events must have happened in Bracara, where a pedestal statue dedicated to Augusto by the bracaugustanus on the anniver-

The city kept some degree of administrative functions, mainly as a centre of road network leading to the Cantabria and Atlantic ports and the outflow of the annona militaris. The relevance of the city geo-strategic commu18 This type of pavement emerged in the republican times in Italic peninsula (second century BC) and was used up to the beginning of the first century AD. Its diffusion in the Iberia peninsula took place in Augustan times and its important is related its association with "a la presencia de pobladores itálicos, o bien miembros de una clase social elevada con alto grado de Romanización" (Burón Alvarez 1998: 1045-1046).

19 Urban archaeology in Lugo was born in the 1970s and was a consequence of the regeneration of the city historical center that was defined by a Roman wall. The uncovered remains confirmed the orthogonal planning and enabled the acknowledgment about the architectural features, the urban infrastructures and the degree of Romanization of this urbe.

33

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

sary date of Paulus Fabius Maximus likely existed (Martins 2004).

We will now enlist the main overland transport circuits that refer to the main itineraries.

Lucus Augusti had, therefore, a first occupational stage as a military camp preceding the urban setting and had been occupied by indigenous communities, who were clearly involved in all tasks related with the foundation of a city.

2.2.3.1 Land itineraries

Throughout the Julius-Claudius times the city endured a first stage of establishing urban infrastructures and ‘’desarollo del embrión monumental’’ delivered with the construction of public buildings within and outside of the city, "(…) sacrificadas al urbanismo flavio y posterior" (González Fernández and Carreño Gascón 1998: 1176).

Bracara Augusta was connected with the other cities founded by Augusto: Lucus Augusti and Asturica Augusta. A series of secondary itineraries connected Asturica with Gallia and from there with the empire capital: Rome. The importance of the road network system between Bracara and Lusitania led us to include in this first level the main known itineraries between this city and the southern peninsula.

The first urban phase of Lucus Augusti took place between the end of the first century BC and mid first century AD and, similarly to Bracara Augusta was the less know stage of its development.

The main itineraries have been acknowledged through three main sources: the Antonine Itinerary,20 the Ravena Anonymous and the Astorga tabulae. The Milestones following on the pathways allowed us to adjust circuits and have an idea over their construction chronology and the conducted maintenance works, due to the emperors’ name references on the inscriptions.

Some buildings and construction sites from the beginning of the first century AD were documented and there exists evidence related with several types of pottery from this time, such as italic sigillata, thin walls, indigenous ceramic and betic winery amphorae, specially the Haltern 70 type, found in all areas of the Northwest (Morais 2005).

Most of these itineraries were built in the Emperor Augusto reign but maintenance works were carried out in the reigns to come.

The first Augustan-Tiberius period witnessed a progressive urban structural implementation (González Fernández and Carreño Gascón 1998), which will be reinforced throughout Claudius times with the consolidation of the transport network system and the establishment of further infrastructures. This was a period when the city was structured around a ‘hypodamic’ urban layout running NW/SE that seemed to follow on the foundational plan. It could also have corresponded to the pre-existing camping site design (González Fernández and Carreño Gascón 1998).

The Via XVI itinerary (ab Olisipone Bracarum Augustam), referenced in Antonine Itinerary and connected with Lusitania, was a circuit that went through the western part of what is today the Portuguese territory and through several important urban centres. The Bracara – Cale section performed part of the itinerary between Olisipo and Bracara Augusta and can be considered a main road, with a continuous millage count and with its caput viae being the city of Bracara Augusta (Mantas 1996; 2000).

This was also the time when a monumental thermal building was constructed. It was located near the Minho River and had an initial construction phase taking place on the first half of the first century (Arias Vilas and De Vega Rodríguez 1996; Arias Vilas 1998). Within city walls remains from another thermal building have been uncovered, and they were believed to have been a public bath infrastructure dating back to the Claudius / Nero times.

The connection with Lusitania was also made through the Bracara – Tongobrigam-Emeritam circuit. From Bracara Augusta it would follow to Tongobriga (Freixo, Marco de Canaveses) and go through the Douro. It would then follow south where another road would cross leading up to Viseu (Dinis 1997). From there, the road led to Egitania (Idanha-a-Velha) and to Emerita Augusta, crossing the Alcantara Bridge. With several Milestones, this trajectory was not mentioned in the Antonine itinerary.

The available epigraphic data related with the city corroborated the presence of the military and religious elites linked with the imperial worship, alongside the presence of a high level of indigenous populations (Tranoy 1981).

2.2.3 Communications network

The implementation of a transport network system represented a key strategy within the overall conquest process and land control system. Alongside the establishment of urban centres, Augusto initiated the structuring of the Northwest main itineraries ensuring the connection of the three main Augustan urban centres and their articulation with the province capitals and other key sites within the Roman Hispania.

20 The Antonine Itinerary reached our days through successive printed copies matching the first editions from the sixteenth century. This key source listed the main road itineraries of the Empire and mentioned 372 land trajectories, several mansiones and the distances between sites. The itineraries are displayed in simple, independent lines. With regards to the area crossing the current region of EntreDouro-e-Minho, this source referenced five trajectories: the road from Cale (Oporto) to Bracara Augusta, the Geira Romana (Via Nova), the road linking Bracara Augusta to Asturica via Chaves and the circuit between Braga and Lugo (Lucus Augusti) via Tui.

34

The Northwest Integration within the Roman World The Via I itinerary (De Italia in Hispanias…ad Legio VII Geminam) connected Rome with Iberian Peninsula via southern Gallia. It went through Tarraco and Caesaraugusta and from there it led towards Legio (Fernández Ochoa 1995b; Menéndez de Laurca and Navia Osorio 2000). The Via XXXII itinerary (item ab Asturica Tarracone) was described as the connection between Asturica and Tarraco and linked with the I itinerary on the Ebro valley. It ensured the link between northwest Hispania and Rome (Fernández Ochoa 1995b, Menéndez de Luarca and Navia Osorio 2000). The Via XXXIV itinerary (de Hispania in Aquitania Ab Asturica Burdigala) followed the outline of the XXXII itinerary inside the Asturias territory going up north until Pompaelo (Pamplona), then Burdigala (Bordeaux) through the Perineum mountain range. Figure 3. Simplified Roman road network system in Hispania, with particular emphasis towards the connection with Bracara Augusta (Morais 2005: 66).

The Via XXVI itinerary (item ab Asturica Caesaraugustana) ensured the connection with Caesaraugusta (Zaragoza) crossing in Brigaecium (Fuentes de Ropel, Zaragoza) the Silver Route. From Oceloduri (Zamora) this trajectory would follow up to Caesaraugusta.

The Via XVII itinerary (item a Bracara Augusta) was documented both in the Antonie itinerary and in the Astorga tabulae. It connected Bracara Augusta with Asturica Augusta in a trajectory that crossed northern Portugal up to Aquae Flaviae (Chaves), and then the Northwest lands until Venetia Petavonium, Argentiolum and Asturica (Lemos 1993, 2000, 2002x; Lemos and Morais 2004; Fernández Ochoa 1995b).

The Via XXVII itinerary (item ab Asturica per Cantabria Caesaraugusta) allowed an alternative connection between Asturica and Caesaraugusta. It had a similar outline to the XXVI itinerary until Brigaecium (Fuentes de Ropel, Zaragoza) and from there it would venture across Vaceus lands until it reached its destination (Fernández Ochoa 1995b; Menéndez de Luarca and Navia Osoio 2000). The ‘Rota da Prata’, with an outline that certainly goes back to pre-historic times, was strategic in connecting the Northwest with the southern peninsula. It practically linked northern and southern Hispania. The establishment of Asturica Augusta on the far north and of Emerita Augusta in the south turned the Silver Route into one of the most important peninsular transport trajectories.

The Via XIX itinerary (item a Bracara Augusta) linked Bracara Augusta to Asturica via Lucus Augusti. Leaving Bracara, this trajectory went through Limia (Ponte de Lima) following up north lands until Tude (Tuy) after crossing the Minho River. From here, the circuit would go to Iria Flavia and then east up to Lucus Augusti. From this city there was a trajectory leading up to Asturica (Almeida 1968; Almeida 1979; Pérez Losada 2002; Fernández Ochoa 1995b).

Other than these main itineraries, the Roman road network integrated a series of secondary intersections and vias vincinales and privatae performing a complex hierarchy of pathways. The description of this network, due to its density and difficult spatial conditions will not be analysed on the present work.

The Via XX trajectory (item per loca maritime a Bracara) linked Bracara and Asturica in a trajectory following the coastline, also connecting Brigantium (La Coruña) with Asturica Augusta. The chronology of this circuit is still unknown due to the absence of Milestones and its trajectory is controversial. In fact, most authors agreed that the layout of this itinerary on Portuguese territory, between Bracara and Tude (Tuy), would coincide with the outline of the XIX itinerary, following on a coastal pathway shaped by the Gallician coastline until Brigantium. Another hypothesis has been suggested for this trajectory. They will be analysed in the following chapter, when we approach the Entre-o-Douro-e-Minho region.

Secondary roads were by nature extremely difficult to identify. They did not have Milestones along their trajectories and they attracted smaller investment funds with regards to their construction, events that led towards their faster disappearance. However, some restricted works of territorial and geographical analysis have enabled us to acknowledge their existence. With regards to the bracarensis conventus, we can identify a few secondary trajectories in the Tongobriga (Dias 1997) and the Aquae Flaviae (Teixeira 1996; Amaral 1999) territory. Research carried out by F. Pérez Losada (2000) at the Ourense region shed some light on a dense network of secondary roads, allowing us to further analyse the scope of the territorial occupation.

Asturica Augusta was connected with some of the most important itineraries in the southern and the eastern peninsula and with Gallia and Italic Peninsula (Sevillano Fuertes and Vidal Encines 2002).

35

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Chapter 4 will analyse Bracara Augusta suburbs and include some of the viae vicinales and privatae uncovered in the area. This data enabled us to exemplify and widen the knowledge about the features of this type of tertiary road network. They are, without any doubt, the most difficult ones to identify (Lemos 2003).

glasses, documented an interchange in which the metals would be traded by Phoenician and Cádiz products (Gonzalez Ruibal 2006-2007). There are doubts with regards to the fact whether the Cantabria area had been or had not been attended on that period (Camino Mayor 1995). Even though it was likely that for a few centuries travels of exploratory nature, or ‘’périplos noratlanticos’’ as interpreted by Morillo Cerdán (2003), would have taken place. These events likely prolonged the military and economic Roman interest in the northwest region.

2.2.3.2 River and sea itineraries

How do we understand the articulation of river and maritime routes within the scope we have been exploring for the peninsular northwest? We will therefore, provide some coordinates in relation to the matter.

Caesar’s conquests in Gallia, together with his expedition to Britannia in mid first century BC, polarized a first advance attempt towards the construction of an Atlantic sea route controlled by Rome. This route would go through the Straits of Gibraltar up to the Mancha Channel following on the Hispania western coastline. It presented favourable conditions to port as wide estuaries and old lagoon areas punctuated this circuit. Currently, these natural features still favour short and long-course circulation (Fabião 2006: 130-131).

As synthetized by Ángel Morillo Cerdán (2003) “el control de las costas setentrionales de Hispania por parte de Roma abría la puerta a una articulación conjunta de todo el Atlántico, y a la creación de una ruta que enlazaba los nuevos territórios de la Gália con el Mediterrâneo a través del estrecho de Gibraltar”. However, the Roman territorialisation of the Atlantic sea is older. In fact, the incorporation of Cádiz (Gadir) within the Roman influential sphere, namely through the signing of a foedus with Rome (Morais 2005) had already opened the possibility for the ocean to perform “una auténtica frontera política del Estado Romano” (Morillo Cerdán 2003).

Nevertheless, Morillo Cerdán (2003) underlined that it was the Augustan Germanic campaigns that decisively shaped the Atlantic into a grand scale sea route, linking the Mediterranean with northern Europe. Even though the chronology of this consolidation is still under debate it does not seem possible to place a regular Atlantic circulation in Augustan times. This would likely only take place afterwards, during Claudio’s kingdom.

There were, at least, two issues worth considering when attempting to understand the literature and archaeological data regarding the use of maritime and river routes within the peninsular northwest.

The third century witnessed a consolidated and growing use of the Atlantic circulation however this time also testified a generalized economic crisis felt throughout the Empire. This trend would then be inverted in the beginning of the fourth century when a more circumscribed traffic developed regional trading. The latter even resisted the fall of the Western Roman Empire (Morillo Cerdán 2003).

The first issue was related with the changing rhythms imposed by the Roman occupation with regards to the sea circulation. These events underlined the consideration of scales surrounding the peninsular Atlantic coast starting in the Mediterranean. The second issue, interlinked with the first, concerned the acknowledgment of river iteneraries, not only when linked with maritime circulation, but also from the perspective of their internal role and impact upon the land communications network.

The value of the Atlantic traffic as the key transport communication network required an understanding of its articulation with local river routes, as well as its apparent overlap with the main continental road networks linking the Mediterranean with the northern Empire frontiers. In truth, the Atlantic sea route should offer itself as a circulation itinerary boosting multiple advantages in relation with to terrestrial trajectories in terms of time, money and safety regardless of the inherent dangers of maritime travel (Morais 2005)

If it is not possible to understand the maritime trajectories without the support of river ones, the same cannot be stated with regards to river routes themselves, which were closely linked with settlements networks and with the land transport system, which in turn were not necessarily subordinated to the development of sea routes. Text references and archaeological remains contributed towards the acceptance over the existence of an Atlantic circuit in Antiquity, or from another perspective, of a complex maritime Atlantic circulation. We know that the Phoenician and Punic circulation was made on western peninsular waters since, at least, the sixth century BC. This adds to the conclusion that the mineral resources performed a prime stimulus for this circulation, according to Estrabão (Fabião 2005: 83). Nevertheless, the presence of Mediterranean origin materials in several coastline locations throughout the Atlantic coast of Hispania, such as amphorae of Iberia-Punic typology, Attic pottery or

The infrastructures developed to support maritime traffic together with its adequate articulation with land and river trajectories ensured an internal distribution of products traded within the Atlantic routes. They presented themselves as a key in the developing and consolidating this important communication network within the western part of the Empire and to ensure the link of this region in particular with the rest of the Roman provinces. It is also important to underline that the structuring and articulation of this circulation with the terrestrial transport circuits was developed considering Roman frontiers.

36

The Northwest Integration within the Roman World

Within the peninsular northwest, the Atlantic route detained a leading role, having in the port of Gades (Cádiz) a key platform linking the Iberian Peninsula with the Mediterranean (Morais 2005).

processes both in the Flavian and Antonine periods. These were achieved through the overall improvement of the main urban areas. We will now synthetize the main issues regarding northwest cities.

On the other hand, the morphology, the stream features and the dense river route network of the northwest region allowed the installation of costal stopovers, some of which were simple anchorage sites.

In Bracara Augusta the main known buildings date back to the Flavian or Falvian-Antonine periods, events we believed were linked with the profound regeneration process that took place between the last quarter of the first century and mid second century. Within this context, it is important to highlight the dynamic that probably dismantled or camouflaged several architectonic remains from the previous period, event that, if not well understood, can significantly distort the assessment of the overall urban evolution process.

The rivers mouths of Minho and Douro performed, with all certainty, strategic supporting ports for small and larger coasting. The estuaries of the main navigable rivers in the Entre-e-Douro-e-Minho regions would have equally fulfilled an important function linked with small coasting navigation.

In the beginning of the second century the public thermal building of Alto da Cividade was built re-using materials and shapes from a pre-existing building referred to as prethermal structure with a foundational chronology or perhaps from the Julian-Claudius period. It was partially destroyed in order to accommodate new buildings (Silva 2000; Martins 1999; 2005).

In turn, the shape of the Galician coastline enabled the installation of important port establishments, certainly including the one in Vigo, the one in Iria Flavia and the key Brigantium port controlling Atlantic navigation (Naveiro López 1991; Morais 2005). In the Cantabria coast, the ports that have been identified are the following: Gijón, Portus Victoriae (Santander), Portus Amanum-Flaviobriga (Castro Urdiales) and Oiasso (Irun) (Morillo Cerdán 2003).

Nearby the thermal area, a theatre was constructed in the north side of the referred thermal building. The baths structure and the theatre were perfectly articulated and had a prime location on the highest platform of the Roman city, very close to the place where the forum is thought to have been located (Martins et alii 2006; Martins et alii 2011; Martins and Ribeiro 2012; Martins et alii 2012; Martins and Ribeiro 2013; Martins et alii in print)

3. The consolidation of the Roman power

A generalized process of strengthening the Roman power within the Iberia peninsula characterized both the Flavian and Antonine periods.

Accepting the amphitheatre location suggestion put forward by the aerial photographic analysis and by descriptions made by antiquarians and scholars referring a building seen in S. Pedro de Maximinos (Morais 2001), the city likely had a second leisure building located on a platform of lower altitude in relation with the one at Alto da Cividade.

This process demanded the renovation and the urban network of the existing cities alongside the consolidation of the urbanization within the Northwest through either the foundation of new centres or with the promotion of previous communities. These new secondary urban centres were articulated by the road network system and significantly grew at this stage. They detained key roles within the new space organization from an economic and political point of view, as well as within the Romanization of the whole territory. These events accelerated the communities’ integration process and the spread of the Roman way of life. These are the points to be approached in the analyses of the consolidation of the Roman power within the peninsular northwest.

The alignment of the leisure building, located within the forum axis, provided Bracara Augusta with a scenography adjusted to the prototypes of monumental parameters featuring other Roman cities. This also suggested these infrastructures could have been built on the same period (Martins 2005). Nearby the Cathedral another public building was erected dating back to the Flavian period and was interpreted as being a macellum, presenting a plant that still offers some doubts in its interpretation (Fontes, Lemos and Cruz 1997-1998). However, the existence of a market on that location has equally been suggested by the discovery of an inscription dedicated to Genius Macelli by Flavius Urbicio.

3.1 The development and hierarchy of the urban centres

Cities of Augustan foundation met key urban development stages between the Flavian times and the 2nd century, in particular during Trajan and Adriano reigns. This process was well documented in the three main conventus capitals of the Northwest that acquired at this stage new buildings of prime quality. This urban renewal programme was not exclusive of the northwest region. In fact, all cities in Hispania have met significant regeneration 37

With regards to private architecture, the excavation of the Carvalheiras domus uncovered a structure that represents a paramount site in the overall understanding the city urbanism and domestic architecture features. This site enabled us to estimate the urban modulation due to the fact the excavation brought to light an entire residential quarter, delimited by streets and walls. Furthermore, a residence with atrium and a perystyle contemplating the

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

south and western areas was brought to light, alongside a group of tabernae that opened up to the porticos flanking the streets. The residential area was built in the last quarter of the first century and witnessed several renovations thereafter (Delgado and Lemos 1985; Delgado and Lemos 1986; Martins 1997-1998; 1999; Silva 2000).

full quarter (Sevillano Fuertes and Vidal Encinas 2002): 49-50). The construction and reshape of public buildings helped to fully establish the city layout, according to the words by V. García Marcos and J. M. Vidal Encinas (1998). From a public buildings perspective, other than the forum, two baths buildings are known to have existed. The ‘’Termas Maiores’’ located on the west side of the city and built on the second half of the first century (Sevillano Fuertes and Garcia Marcos 2000). The same chronology is given to the ‘’Termans Menores’’ (Burón Álvarez and García Marcos 2000).

From the Flavian period was the construction of another domus identified within the area adjacent to the Albergue Distrital. The conducted excavations revealed a street of around 25 feet (7.50 meters) separating two insulae and street crossing section boosting the 150 Roman feet module also identified at the Carvalheiras site (Martins 1997/1998; Martins et alii 2012). Although the group of buildings that existed in this insula are difficult to interpret, due to the small portion of excavated area, the conducted interventions allowed the identification of a Roman cloaca (Lemos and Freitas 2000) and confirmed the idea of a city with streets flanked by porticos, adjacent to tabernae constructed on ground level of houses (Martins 1999; Ribeiro 2008; Martins 2009; Martins et alii 2012).

The construction and remodelling programme of both public and private buildings greatly helped to consolidate the image of magnificent urbs of this particular city, a place where the wealth from the gold exploitations of the Northwest would interact with several products and people coming from various Hispania and Empire locations, including prominent administrative employees that were needed in the overall managing of the gold mining exploits of the region (Tranoy 1986; Rabanal 1988).

Based upon available data, we can consider that Bracara Augusta reached its maximum extension on the Antonine period. This seemed to be accompanied by a significant process of overall infrastructures improvement of the city and was characterized by the construction of two leisure buildings (Martins et alii 2011; Martins et alii in print), unprecedented within the other conventus capitals of the Northwest. Although no available data can corroborate these ideas, this regeneration process would also affect the forum area, similarly to what happened in other cities in Hispania (Martins 2005).

When assessing the remains of building constructions located outside the walls, we understand that the built city, between the first and third centuries, was wider than the core area enclosed by the late wall (Garcia Marcos et alii 1997; Sevillano Fuertes and Vidal Encinas 2002). However, it seemed evident that the urban development was interrupted in the third century, at a time when a sharp decline of the city took place. This was evident in the deactivation of the public bathing infrastructure and in the gradual abandonment of most private residential properties.

Having evolved from a military camp, Asturica Augusta displayed a known ortogonal layout, with squared and rectangular quarters and streets flanked by porticos (SeVillano Fuertes and Vidal Encinas 2002), similar to Bracara and Lucus urbanism.

The urban development and growth reach its peak in the Flavian period, when a wide construction programme further developed the street network system modifying existing streets and building new circulation areas. New porticos were built and changed the urban layout axes into a N/S orientation over the NW/SE foundational one (González Fernández and Carreño 1998).

From the Flavian times onwards, Asturica witnessed a growing urban development resulting from its accumulated wealth originated from the generated income from regional mining exploitations. There was a change within the urban network structure. The streets were elevated and porticos were constructed. Health infrastructures were developed in works that began at the beginning of the first century, also including the construction of new cloaca.

From the Flavian dynasty onwards, the city revealed a more consolidated urban plan with quarters presenting different modules: 60 x 70 meters and 70 x 100 meters. The forum was then relocated to the present Plaza Mayor area (González Fernández and Carreño Gascón 1998).

Private constructions from the same period were the reflection of a wealthy and profoundly romanized society. At the Flavian-Antonine period, luxurious domus were built. Good examples were the ‘’Casa de la Muralla’’ (House of the Wall), located on the far west side of the city, the ‘’Casa de las Columnas Pintadas’’ and the ‘’Domus del Mosaico del Oso y los Pájaros’’. All of these were built at the end of the first and beginning of the second centuries AD (García Marcos and Vidal Encinas 1995; 1996; 1998).

Similarly to Bracara Augusta, Lucus Augusti revealed the existence of porticos flanked by streets with sizes varying between four to eight meters width (González Fernández 1995a: 53-54). The numerous excavations conducted in the city detected several elements linked with both the sewage system and with water caption techniques (Álvarez Asorey et alii 2003). With regards to public buildings, other than the public baths built nearby the Minho river (González Fernández 1995b: 67-69) there was a tabularium that has been documented by epigraphy, and there is the possibility about the existence of a theatre or amphitheatre structure that is

In several cases, new buildings overlapped previous constructions, as it was the case of the ‘’Domus del gran Peristilo’’, construction which reshaped a previous structure dated back from mid/late first century occupying the

38

The Northwest Integration within the Roman World

currently under study (Gonzalez Fernandez and Carreõ Gascón 1998).

ies focused on the establishment and specific developments of each of the urban centres. Let us ponder now about the scale issues.

From the Flavian and Antonine periods was the construction of most private residences known within the city, with the main feature being the presence of a bathing infrastructures and the regular use of tiles (Rodríguez Colmenero 1995).

For a start, the ‘invisibility’ of the urban programmes with regards to the three Augustan cities located on the northwest peninsular has not been specific for these three centres. It should instead be thought as a consequence of the occurred transformations within the cities from the Flavian period onwards. In fact, the heavy regeneration programmes overlapped previous spaces or recycled existing constructions, operations and dynamics, thus camouflaging the remains of the pre-existing city spaces, turning them opaque and difficult to acknowledge, from an archaeological perspective.

At the time the city was a wide area and some previously built constructions were demolished with the construction of the Low Empire wall, similarly to what happened in Bracara and Asturica. The mentioned wall was built at the end of the third beginning of the fourth centuries and was rather similarly in terms of construction techniques to the other defensive walls present in the other northwest conventus capitals. The wall was subjected to several regeneration processes throughout the centuries to follow, which did not changed its original layout (Rodriguez Colmenero 2007; Alcorta Irastorzsa 2007).

Furthermore, we believed that the problem with the Augustan and Julius-Claudius programmes for northwest cities should be assessed not only in global terms, and rather contextualized within the evolution of the cities in Hispania, followed by studies of each city in particular while considering the results emerging from archaeology. Moreover, there is an undeniable need for further data and for a detailed assessment and re-evaluation of the available archaeological information.

The magnitude of the urban programmes developed in the Flavian period may easily lead us to the conclusion that before the 70s of the first century AD there were not ‘real cities’ in the Northwest. The conviction, underppined over decades, regarding a local promotion of these cities throughout the Flavian period and the absence of epigraphic data has vitiated all the interpretations about this subject, mainly regarding the apparent inexistence of outstanding urban infrastructures between foundation times and the Flavian period, in all the three cities of the Northwest. Adding to this, all the building works at the Flavian period have been interpreted as a natural consequence of the promotion of the cities into municipalities or conventus capitals (Tranoy 1981).

The conducted works at the Flavian and Antonine periods took place throughout Hispania and were the result of the growth and consolidation of the Western Empire. Sites suffering from these impacts were evidentially different but strategically key. In fact, the regeneration process took place both within cities established during the Late Republic and in the ones founded by Augusto. This process was also associated with the promotion of secondary urban centres, and their growth underpinned the Roman power in space where the marginal territories becoming gradually smaller.

Although the hypothesis about the legal status of Bracara Augusta in terms of Latin law goes back to 1994 (Le Roux 1994), and despite the fact that this suggestion did not raise any debate at the time, the acknowledgement about the existence of a tabula that corroborates the organization of these conventus capitals to an Augustan era and finally the fact that archaeological data revealed that some urban structures were prior to the Flavian period in all the northwestern cities, there is still the question regarding the apparent ‘invisibility ‘of the ‘Augustan cities’.

This interpretation was, without doubt, the more suited when it came to understanding the issue. We can, therefore, add: the regeneration policies of cities were articulated with their spatial expansion, underlying the consolidation of the circulation network system and the investment in the outskirt areas, especially when it came to the first delimited urban plans drew in Augustan times. This articulation revealed a global political reach demonstrating and reinforcing the image and power of Rome over the provinces.

We chose to leave aside the question regarding the legal status of the analysed cities considering the lack of irrefutable testimonies to discuss it. On the other hand, it is questionable to continue defending the local promotion of these cities in Flavian times after considering all the debate that has been emerging in the last few years with regards to the early organization of the peninsular northwest conventus, as it is accepted by the majority of researchers (Rodriguez Colmenero 1996a; 1996b; Fernandez Ochoa and Morillo Cerdán 1999).

Furthermore, bibliography referring to the majority of the Roman Hispania – Tarraco, Clunia, Emerita Augusta and even Conimbriga, corroborate a common process. All these cities have met, in the Flavian and Antonine times, wide urban renovation programmes regardless of their legal status or whether their foundation was of Republican, Caesarean or Augustan origin. The Tarraco case study was paradigmatic: the Flavian programme provided the city with a powerful scenography; widely visible to those arrived by sea or by land. These trends may be understood as a clear display of the power of Rome outside the empire capital (Mar 1993; Mar and Pensabene 2001).

Regardless of the legal status of the three Augustan foundations on the peninsular northwest, the general framework of their urban development has to be further analysed within a wider context including all of Hispania territory. This does not bypasses the confined and ‘individual’ understanding that should be included in the stud-

What we hope to assess afterwards concerned the three cities of Augustan foundation in the peninsular northwest.

39

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

They had development stages and evolution rhythms rather different between themselves. They shared, however, a common framework with regards to the strengthening process of the Roman power within their territories. This was linked with a growing dynamic and economic increase as the urban centres performed powerful arenas to attract and disseminate population.

In Julius-Claudius times, the city of Bracara Augusta also displayed the ability of the population to pay tribute to the imperial family or to his legates through buildings that should have been located around the forum. We know that the city housed traders that were in fact Roman citizens, epigraphically confirmed in the Claudius period. The civic community of the early city was able to construct buildings of excellent quality, open up roads and keep quarries functioning and even organize and structure the associated population by dividing the land of the surrounding region. It was certainly in the city started to build in the Augustan time that lived, traded and enriched - that is, have been structured – the elites who will later become decision-makers with regards to the regeneration programmes occurring in Flavian times.

This common and wide framework of urban expansion reverberated itself within the elites. With varied agents and chronologies, the Augustan cities displayed, between their foundational period and the Flavian and Antonine periods, the necessary time and conditions to enrich and establish their elites. This supported the urban renovation programmes that were gradually more sophisticated. The perception of that period, necessary to the population growth and the development of trading and handmade activities seemed to be paramount in the assessment of the overall urban evolution programmes of northwest cities.

Alongside the consolidation in power and the establishment of urban centres in the peninsular northwest, there was a clear expansion of new secondary urban centres throughout the region, namely between Augustan times ad the Flavian and Antonine periods.

In fact, the urban programmes starting in the Flavian period matched the needs arising from the internal growth and development of the cities and associated territories. This led towards the reshape of buildings, to the relocation of spaces dedicated to handcraft activities towards suburban areas and was linked with the need to reinforce the private construction programmes. Furthermore, it also led towards the deactivation and the reshape or recycling of other urban spaces in a natural process that suited new needs and altered the facies of cities.

This has been a well-documented process associated with a second moment of territorial reorganization that consolidated several urban sites, some of which stood out by their local promotion (Fernandez Ochoa and Morillo Cerdan 1999; Perez Losada 2002; Martins, Lemos and Perez Losada 2005: 279-296). Furthermore and adding to this reorganization is the expansion of a rural establishments network, with the development of Villas and the abandonment of Hill-forts.

This panorama seemed to be present in all three northwest cities and it should be studied and assessed as an urban development dynamic process, due to the fact these cities were spaces enduring a continuous transformation.

It seemed pertinent to accept that some of these secondary conglomerates ended up becoming regional capitals due to their gradual relevance in terms of concentrating population and boosting economic impact in certain territories. Furthermore, these trends seemed to have been documented in the areas located within the bracarensis conventus, a process studied by F. Pérez Losada (1998; 2002). These regional capitals could be considered urban centres despite their role as caput civitates. They were added with functions related with control over other centres of lesser importance (Martins, Lemos and Pérez Losada 2005).

The understanding of the city space before Flavian times can only be achieved with further archaeological research, which should focus the problem within the evolution process dynamic of cities in political, social and economic terms. This can be exemplified in the synopsis encompassing the foundation of Bracara Augusta and the time of its main buildings construction, the ones already excavated and better known that had been erected mainly during the Flavian and Antonine times. In this city, as in other cities, the assessment of archaeological data can never disregard the 70 to 80 years between the foundation and the construction of the Flavian buildings as well as, the agents that were, according to epigraphy, a population mainly composed of indigenous origin coming from the Hill-forts scattered around the city. The generations that cover the time period in question were sufficient to construct a building preceding the baths of Alto da Cividade, to acknowledge handcrafts infrastuctures that were later demolished to give place to private residences, as it was the case of the Carvalheiras domus, to develop commercial relationships with the exterior and to accumulate wealth that ended up being key to the process of regenerating the city. Moreover, the social body developed with the arrival of immigrants from other regions.

40

Some factors prompted the emerging of a new regional capitals network located between the conventus and the civitates capitals. One of those factors was the importance invested in those urban centres. Another one was linked with the economical development of certain regions as a consequence of the mining exploitations, the intensification of Farming activities but also the importance of trading activities ensured by both maritime and land comunications networks, all of these ensuring the conditions for the emergence of key centres. Although it was difficult to rigorously document these capitals on a northwest scale, due to the absence of archaeological data, it was still possible to identify some of the urban centers emerging as regional capitals. Within the bacarensis conventus territory, it is legitimate to suggest that were three centres: Aquae Flaviae, Tude and Tongobriga. The Iria Flavia could be part of the

The Northwest Integration within the Roman World

conventus lucensis territory, and finally, the city of Gijón can be placed within the asturicensis conventus region (Pérez Losada 1998).

to a cloaca and pipelines structures. Well studied was the city water supply system that encompassed a dam and an aqueduct (Alfenim 1995; Rodríguez Colmenero s.d; Rodríguez Colmenero and Alcorta Irastorza 1998).

We will attempt to characterize the urban centres that might have performed a key role in the development of different northwest regions in the sense of consolidating the Roman power, in improving the efficiency of local resources exploitation, in integrating the population and in crystallizing the assimilation of the economical, social and cultural Roman models.

Research works conducted on site produced several hypothesis related with the development of this urban center, both in the High and Low Empire times. The studies put forward the potential location of the forum and the theatre (Rodríguez Colmenero s.d; Rodríguez Colmenero and Alcorta Irastorza 1998).

The city of Aquae Flaviae was originated from a settlement located in a rocky formation with 375 metres altitude, nearby the Tâmega River dominating the wide valley area of Chaves that spread towards the east. This site matched the Turodi civitas capital (Pérez Losada 1998) mentioned by Ptolomeu and believed to have been Aquae Laiae. To most researchers there must have been an error within the geographer work and it is believed that it should be read instead as Aquae Flaviae (Tranoy 1981).

Moreover, urban excavation conducted in the city in the last decade provided numerous unprecedented data believed to improve the knowledge about the development of this city. Recent known archaeological data exposed a thermal building, very well preserved but partially destructed by an earthquake. These events corroborate the high level of Romanization and regeneration process of this key urban center. Another bathing complex located outside the wall also existed at the city, although it has been rather difficult to fully acknowledge its trace due to the intense and successive regeneration programs endured by the infrastructure throughout time.

In Flavian times, the importance of the pre-existing communities, who flourished throughout the first half of the first century, would likely justified its local promotion. It was at this time that the centre received the name of Aquae Flaviae, the headquarters of the Aquifavienses, a term mentioned in the ‘Padrão dos Povos’ dating back to the year 79 AD (CIL II 2477; Tranoy 1981) and also referred in epigraphic sources (Rodríguez Colmenero 1997).

Considering the city relevance within the conventus bracarensis context and the overall peninsular northwest territorial organization, the civitas capital of Aquae Flaviae possibly has functioned as regional capital, as defended by some authors (Pérez Losada 1998; Martins, Lemos and Perez Lousada 2005). Favoring this hypothesis, we may consider this urban center to have functioned as caput viarum.

The centre was promoted in Flavian times likely due to its strategic relevance in terms of location. It was at the very centre of the Via XVII itinerary linking Bracara Augusta to Asturica Augusta. This itinerary also went through key mining regions.

The first populated center of city of Tuy must have had a defensive system. The city is known in classic references as Tyde or castellum Tyde (NH IV 112). This site was likely the headquarters of the civitas tudensis and probably belonged to the Grovii territory (Pérez Losada 2002).

The Via XVII itinerary and its Aquae Flaviae territorial branches turned this civitas into an important circulation platform of products, equipment and people, which develops due to mining exploration witnessed in its surrounding region. A gold mining complex was established around Aquae Flaviae, known as the Outeiro Seco, or the Outeiro Machado and the entitled ‘Poço de Freitas’, together with the key territorium metallorum of Jades – Três Minas (Three Mines). The exploitation of local resources started in the Tiberius reign (Lemos and Morais 2004).

The Roman city was established in a low altitude area, around 1/2km from the castellum Tyde and it probably dated back to the Flavian period when it was known as Tude (Pérez Losada 2002). The finding of architectonic and epigraphic elements, together with references to private residences, enabled us to consider Tude as a key civic and political Roman center, one that consolidated the urbanization process of the Northwest and prompted the integration of local communities within the Roman world.

The strategic circulation and economic relevance of the civitas, promoted to municipium in the Flavian period, benefited from an urban reorganization and regeneration programmes that implanted an orthogonal layout of Flavian origin (Rodríguez Colmenero and Alcorta Irastorza 1998) also documented in the surrounding area (Rodríguez Colmenero 1997). These changes included the construction of a bridge over the Tâmega River in Trajan times. The bridge was connected with the Via XVII itinerary and included Aquae Flaviae in the official main itineraries network of the peninsular northwest.

The analysis of aerial photography allowed F. Pérez Losada (2002) to reconstruct a potential urban layout for this city. It had a NNE/SSW orientation that the author defined as a ‘pseudo-orthogonal street complex’. Considering the available archaeological data and taking into consideration the existence of a pseudo-orthogonal layout, the existing public buildings and the city central role both in the transport and trading networks allowed us to consider Tude as a regional capital. This interpretation has also corroborated by the city role as caput viarum of

Despite the abundant epigraphic data and archaeological excavation works conducted in Chaves, the Roman urban layout is still not well known. The same can be said about the sewer network, regardless of the numerous references

41

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

the Via XIX itinerary, linking Bracara Augusta to Lucus Augusti.

nine period reflected its urban centrality and relevance (Dias 1997; 1999).

Tongobriga has been referenced in literary and epigraphic sources. Ptolomeus called it Tuntobriga and included the city within the Callaeci Bracari.

The development of other northwest seconderay urban sites throughout the decades before the Flavian period might justify the changes within individual legal status, as it happened to Aquae Flaviae. Moreover, and although the status of Tude or Tongobriga have still not been confirmed, both centres are thought to have been regional capitals.

From 1980 onwards, the site has been thoroughly excavated enabling us to acknowledge some of the features of this settelement. The city has been interpreted ever since as a civitas capital (Alarcão 1988; 1995-96; Dias 1997).

There was another promotion within the bracarensis conventus territory. This was the case of the Limicorum forum site (the current Xinzo de Limia), a location that from late first and early second centuries AD had a municipium status (Pérez de Losada 2002). This site displayed a proto-urban layout, similar to ‘povoado-rua nuclear’, structured upon the construction of the road linking Chaves to Ourense and by some transversal road axes, witnessing the remains of important edifications (Xusto Rodríguez 1998). The epigraphy revealed the existence of several immigrants and Roamn citizens within the region and we know about the promotion of indigenous enrolled in the Quirina tribe.

Conducted excavations revealed several occupational phases. Between the end of the first century BC and the beginning of the first AD and up to the Flavian period, this urban center was likely occupied by pre-Roman communities (Dias 1998: 767), who built several construction infrastructures of circular shape and a preRoman bathing complex. Between the Flavian and Antonine periods, namely between Vespasian and Trajan, the site was structured and displayed typical Roman urban infrastructures with rectangular layout residences. The public baths infrastructure is from the same period (thermal complex I) and was located nearby the pre-Roman thermal complex (Dias 1997). In mid second century, specifically in both Trajan and Adrian reigns, a wide forum was built (139m x 68,5m). It housed a basilica and a temple and fully articulated with the pre-existing thermal building (Dias 1999).

Iria Flavia performed an important urban centre on the western area of Galicia. Its role as a mansio of the Via XIX itinerary was referenced in several sources, as it was the case of the Antonino Itinerary, the Ravena Anonimous and the Astorga tabulae.

The first half of the second century registered key developments within residential areas together with the regeneration of the bathing complex (thermal complex II) that included from then onwards a natatio (Dias 1997).

The geographer Ptolomeus referred to Iria Flavia as one of the two ‘’poles existing in the Coporos territory’’ alongside Lucus Augusti (Pérez Losada 2002:88). However, F. Pérez Losada (2002) believed Iria Flavia was in fact not a secondary vicus when compared to Lucus Augusti due to the importance this urban centre displayed in both epigraphic and archaeological sources. Furthermore, the author suggested the possibility of a Copori territory dismembering in the Flavian period, fact that likely reinforced the importance of Iria Flavia as a civitas capital. Under these circumstances, Lucus Augusti kept its strategic role as conventus and civitas capital of the eastern Copori region, leaving Iria Flavia the capital of the Copori communities living on the west.

Throughout the fourth century and early fifth new developments took place in the thermal infrastructure, in the forum and in residential areas. In the seventh century the urban centre witnessed a decline period even though it became parish headquarters of the Porto diocese as referred in texts of the ‘Paroquial Suévico’ (Dias 1997). In the first half of the twelfth century the site was known as Santa Maria do Freixo. Other than the referred public infrastructures, the city was believed to have housed a theatre complex not yet excavated and benefited from the natural slopes existing nearby the axes identified as the city kardus (Dias 1997). Analysis of aerial photography seemed to indicate the existence of a circus located outside the city and nearby the road linking Tongobriga to the area crossing the Tâmega River in Canaveses (Dias 1997; 1999).

Two inscriptions revealed the original name of this urban center in Roman times and, simultaneously, enabled us to understand its origin as a forum, a place specifically built by the Roman power as a space for trading transactions (Pérez Losada 2002). In both inscriptions the site was named as Forum Iriesis. One of the pieces was a votive monument dedicated to Neptune. The following can be read: NEPTVNO FOR(o) I(ir)E(n)S(is) D(e) S(uo) P(osuerunt). The other inscription was engraved in a funerary stele presenting the following reading: CAMBAVIVS CORALI F(ilius) SENA FOR (o) IRIE(n)S(is) ANNORVM L H(ic) S(itus) E(st) S(it) T(ibi) T(erra) L(evis).

Despite the identified alignments of the residential area, the city seemed not to display an orthogonal layout, an absence that can be justified by the geo-morphological features of the site or by the existence of a pre-Roman settlement before the Flavian period (Dias 1997). The importance of Tongobriga as a civic and road centre, located in the surroundings of the Bracara – Emerita road (Alarcão 1995-96) might explain the city dynamism as a regional capital and its centrality in the provision of products and services. Furthermore, the public work programme that this centre endured in the Flavian and Anto-

According to the available archaeological data, Iria Flavia performed an ex novo foundation emerging in the transition years between the first century BC and the first AD. It probably encompassed a forum (Forum Iriensis), a

42

The Northwest Integration within the Roman World

civic center established to fulfill an imminent commercial role, namely in a territory dominated by disperse indigenous communities (Pérez Losada 2002). The growing importance of the site, mainly as a maritime trading center that was strategically located, alongside its status as civitas capital led to its urban development throughout the Flavian period. Its importance remained throughout the second century. The name Iria Flavia revealed the importance the city achieved on the second half of the first century, although F. Pérez (2002) referred that this events do not prevent the possibility of the city having a municipal status.

current Coruña region but its origin and foundation are still questionable. The author Miguel Santa Cruz (2003:121) considered the city foundation to have been an initiative of Rome in the second century AD and an establishment that overlapped a previous commercial center. The city was provided with a port infrastructure that would have been located under the present paseo maritimo (Bello Diéguez 1994). Other authors referred the Roman developments in Brigantium started at the end of the first century and were mainly located in the upper area of the old town. Regardless, the city most significant development stage took place on the second half of the first century, throughout the Antonine period (Fernández Ochoa and Morillo Cerdán 1994:62). This was a time when the city was further expanded.

Conducted excavations in Iria Flavia revealed the existence of an urban planning displaying an orthogonal layout running N/S and alongside the XIX itinerary. The latter itinerary crossed the city and was its kardo maximus. Infrastructures excavated did not reveal public buildings with the majority of structures uncovered referring to private residences. Nevertheless, some residences displayed a notable wealth whereas others presented rather poor features and are believed to have been either domestic residence of crafting and trading establishments.

At the Brigantium port it has been constructed a pharos known as the ‘Torre de Hercules’ (Romero Masia 1992). This performed the only surviving Roman infrastructure likely to have been financed by the Roman state (Santa Cruz 2003:131). This monument, studied by S. Hutter (1973) and T. Hauschild (1976) displayed in its surrounding area an inscription allusive of C. Sevius Lupus, a citizen from Aeminium and most likely the constructor.

The materials found in the excavations are abundant, varied, mainly including common ceramics, ‘thin walled’ ceramics, lamps, and glasses, alongside a wide number of coins, sigillata and amphorae.

The Roman state use of this urban center would be strategic not only in terms of transporting goods but also troops (Naveiro López 1991: 154). Under these circumstances, Brigantium performed a paramount site in supporting the Atlantic route in the transport of military troops and goods between the Mediterranean, Britannia, Gallia and the Germanic limes, also functioning as a scaling port (Santa Cruz 2003). Its privilege location turned the center into a good receiver key site (Naveiro López 1991; 1996) while exporting regional mining products from the Asturias region (Pérez Losada 2002:136).

Within the imported pottery uncovered assets, it was surprising the quantity of found sigillatas and the variety of represented forms in Hispanic, Gallic, late Hispanic, Africa productions studied by López Pérez (2002). This has led to the identification of Iria Flavia as a key ceramic products redistributor. The studies conducted on the immediate surrounding area revealed scattered settlement representing a ‘’continuum da cidade Romana (…) disposto nas inmediacións da via XIX entre Iria e xacemento de Porto en Cesures’’ (Pérez Losada 2002:99). Under these circumstances, the urban center seemed to have performed a key role as a road node of the Via XIX itinerary and a guaranteed commercial importance, both for the passage this road, and also for its proximity to the scale points essential in the Atlantic route, as for example the Arousa case and the water course of the Ulla River.

Little is know about the occupation of the city, although the results of some excavations enabled us to acknowledge its regular urbanism displaying an orthogonal layout (Naveiro López 1991). The city of Gijón is located in a peninsula and boosts a key geo-strategic potential. It performed a paramount site for the Cantabria navigation between Antiquity and our days (Fernández Ochoa and Morillo Cerdán 1994). Established in the Flavian period, the city is believed to have been the heiress of the Campa Torres settlement, a site abandoned in the late High Empire (Fernández Ochoa et alii 2003: 104-105)21.

Iria Flavia and its harbor would function as a key commercial site and redistribution center, namely a supplier of Lucus Augusti. The urban center reached its peak between late first and the second centuries, witnessing since then a progressive decline (López Pérez et alii 1999: 252, apud Pérez Losada 2002). The retraction of the maritime trade since late second century, the consolidation of land transport networks and the proximity to Brigantium might explain the city loss appeal from the third century onwards. It only recovered some dynamism in late fourth century and throughout the fifth when it became episcopal headquarters.

21 The Campa Torres settlement is normally identified as oppidum Noega as referenced in the classic sources. It was a wide habitat where the oldest Roman materials from the region can be found, including italic sigillata, Tiberius coins issued in the Ebro valley and one coin issued in Graccurris. An important inscription was also uncovered in Campa Torres. It was dedicated to Augusto and dated back to the years 9-10AD. It has been identified by several authors as one of the three Arae Sestianae, referenced in the classic sources and constructed in Augustan times in the Gallian coast by L. Sextus Quirinatis, a governor of the Citerior province between 22 and 19BC. These arae marked the arrival of Rome to unknown territories (Maya Gonzaléz and Cuesta Toribio 1995).

Within the lucencis conventus area the importance of Brigantium is well attested. The city was located in the

43

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

The foundation of Gijón can be included within the urban development processes occurring in the Northwest throughout the Flavian dynasty. These processes also encompassed the establishment of several secondary centers, enabling the development of a network of smaller cities between the conventus capitals and the wide rural surrounding areas (Fernández Ochoa 1995a).

in scale of the territorial control under the impact of the urban expansion.

3.2 Reinforcement of the road network system

The reinforcement of the road network system conducted on the Northwest witnessed a new construction, maintenance and repair stage occurring in the main Augstan itineraries. On the other hand, the communications network was then consolidated with the construction of new bridges and the improvement of the cursus publicus alongside the construction of new itineraries (Tranoy 1981; 1995-1996; Lemos 1993; 2000; 2004).

Gijón rapidly reached an important status within the Conventus Asturicensis context. The city became a maritime port, turning this urban center into a key site supporting the circulation of products in the Cantabria route, fully articulated with the Atlantic route (Fernández Ochoa et alii 2003). Moreover, several maritime and land routes crossed in this city linking the site with the rest of the Peninsula, with the Gallia and Britannia ports supplying goods and transporting the troops located nearby the Adrian wall. From a regional and peninsular perspective, its role as the core center of the ‘Via de la Plata’, boosting communications of Gijón with the central and western region of Asturias, with the Meseta and with the southern peninsula, was also behind the relevance of this urban center.

The repair works of the main itineraries are well documented in the Milestone and in other road inscriptions, mainly the ones referencing emperors after Claudius due to the fact that the first phase of development of new itineraries developing a perfect triangulation between the three northwest conventus capitals only ended at the end of the Julius-Claudius dynasty.

The first excavations only took place in 1982 and they uncovered a series of important infrastructures, such as the Campo Valdés thermal building built in the late Flavian period and the beginning of the second century (Fernández Ohoa et alii 2000; Fernández Ochoa 2003) and the late Roman wall constructed in late third and fourth centuries and reaching an area of up to 16 hectares (Fernández Ochoa 1997; Fernández Ochoa and Gil Sendino 2007). This wall, with semi-circular towers of around four to five meters diameter matches with the walls found in Lucus, Legio, Asturica and Bracara and integrated the regional program of defensive constructions that aimed to ensure the safeguard of the Empire western part (Fernández Ochoa 1997; Fernández Ochoa and Gil Sendino 2007). Arcaheological remains related with the salting industry were also discovered. It seems to have had a range of export limited to the most nearby and it operates between the third and fourth centuries AD (Fernández Ochoa et alli 1996; Fernández Ochoa 2003). The Gijón bay was a refuge and trading site, reinforcing the belief the city played a strategic role as a Cantabria port (Fernández Ochoa and Morillo Cerdán 1994), with its maritime infrastructures likely located in the entitled ‘BajoVilla medieval’’ (Fernández Ochoa et alii 2003:115). All these urban centers attested the gradual densification of the peninsular northwest urbanization after the establishment of the three main conventus capitals by Augusto. The economic development of the rural areas, the functioning of land, river ad maritime routes together with the needs prompted by the new political, economical ad social order gave origin to newurban centres, some of which are only consolidated in the Flavian and Antonine periods. They displayed an orthogonal urban structure, housed infrastructures of clear Roman typology and some even became regional capitals. This was a time of readjustment

44

Some examples of the road network system repair works took place at the XVI, XVII and XIX itineraries, all registering an absence of Flavian Milestones. However, there were some Milestones from the Antonine period, mainly from the reigns of Trajano, Adriano and Marco Aurélio. These events suggested improvement works in the layout of these itineraries and also an overall the improvement in utilized materials. The Via XVI itinerary has Milestones referencing the Antonine dynasty in site where this itinerary crossed the territory integrated in the conventus bracaransis, between Bracara and Cale. Furthermore, there is another Milestone from Adriano at Quinta do Paiço in Santo Tirso, in Portela de Baixo and S. Cosme do Vale in Vila Nova de Famalicão. It is also referenced Milestones from Marco Aurelio in S. Cristovão do Mouro (Santo Tirso) and in Quinta de Santa Catarina (Vila Nova de Famalicão) (Mantas 1996). Milestones from Adriano times have also been uncovered in the Via XVII itinerary, namely in Ponto do Arco (Montalegre), Venda dos Padrões (Chaves), in Babe (Bragança) and San Vitero (Zamora). On the other hand, from the group of Milestones located in the Via XIX itinerary, we empahsise the ones from Adriano located in S. Paio de Merelim (Braga), Santa Marinha de Arcozelo (Ponte de Lima), in Pedrón, Saxamonte, Redondela, Vilaboa and Ponte do Burgo (Pontevedra) (Rodríguez Colmenero et alii 2004). The maintenance works that took place at the main land itineraries consolidated the overall management of the territory and imprinted the imperial signature in distant regions. Moreover, the innumerous Milestones existing on the main road circuits of the Northwest also revealed the political propaganda and the intention in keeping operational the main circuits, in order to ensure the connection between the cities, the homogenization of the territory and the circulation of people, products and information.

The Northwest Integration within the Roman World

Some of the most important known bridges located on main circuits were established in the Flavian and Antonine periods, as is the case of the Aquae Flaviae Bridge crossing the Tamega River. Dating back to Trajano times, this bridge displayed twelve arches and was linked with the Via XVII itinerary (Rodriguez Colmenero 1997). Its construction was probably related with the existence of a wide group of civitates quoted in the famous ‘Padrão dos Povos’ (Alarcão 1988). Another work of Roman engineering was the bridge over the Lima River on the Via XIX itinerary, although its precise chronology is still difficult to assess.

The layout of the Via XVIII itinerary is well known due to the fact that it has been the object of several studies (Caamaño Gesto 1995-1996; Lemos e Baptista 19951996; Rodríguez Colmenero et alii 2004). Exiting Bracara Augusta, the trajectory crossed the Cávado River and went through the current municipality of Amares and Terras de Bouro,22 heading towards Portela do Homem, where it would enter the current Ourense province and crossed Rio Caldo and Baños de Bande. It would then follow up to Baños de Molgas, Alto de Rodicio and Serra do Burgo towards Trives and Valdeorras. There, it crossed the Encina da Lastra Mountain before reaching Astorga (Caamaño Gesto 1995-1996).

There were other bridges with unclear chronologies. They were located on the circuits of several main itineraries. Even though not all of these were built on the period when the transport system was consolidated, it is possible that the economical and urban development of the overall region taking place between Augusto and the Flavian and Antonine periods and the promotion and the further development of secondary urban centers contributed towards the edification of these infrastructures. These works were likely at the expense of local populations, as it was the Chaves bridge case.

The Via XVIII itinerary had several bridges. In the Portuguese territory, near Portela do Homem, there were three known bridges. The trajectory between miles XXXII and XXXIII in the Albergaria area has the remains of two bridges crossing the Maceira and Forno river streams, tributaries of the Homem River, both with one arch. A wider structure would have been S. Miguel Bridge with at least three perfectly round arches, however this structure was destructed in 1642 (Lemos and Baptista 19951996:118-121).

Little is know regarding the bridge crossing the Lima River in Ponte de Lima. However the bridge is thought to have had a Roman foundation. The key relevance of the Lima settlement, referenced in the Antonine Itinerary, was located on a strategic site of the itinerary and likely prompted the construction of such infrastructure.

Galiza had several Roman bridges. The Pedriña Bridge crossing the Lima River, the Navea Bridge and the Bibei Bridge in Trives are good examples. The latter performed one of the best-preserved examples in Hispania, dating back to Trajan times (Casal Garcia and Blanco Sanmartín 2005:116-117).

Fransisco Sande Lemos referenced some bridges located in the conventus bracarensis, although the author does not indicate any precise chronology for these structures. From the list of bridges referenced by the author we reference the Pedrinha Bridge over the Tuela River in Torre de Dona Chama located on the Via XVII itinerary, the São João Bridge and the entitled Campelo Bridge in Guimarães built over the Ave River. Furthermore, we also provide a particular relevance to the Arco and S. Martinho do Campo bridges built over the Vizela River in the route linking Bracara to Emerita Augusta (Lemos 2005: 171172). The author also mentioned the Arquinho Bridge in Possacos, in the Valpaços municipality serving the circuit between Bracara and Asturica via Chaves and the Arco Bridge in Alfarela de Jales signaling a secondary trajectory crossing the mining district of Serra da Padrela (Lemos 2005:173).

Due to the abundance of Milestones, remains of roads and bridges signaling with unusual precision its layout, this trajectory represents a paradigmatic example of the consolidation of Roman power in the peninsular northwest. Furthermore, this new itinerary should be understood within a context of regional economical growth and articulated with the intensification of the gold-mining exploitation, paramount to the Empire overall economy (Tranoy 1981; 1995-1996; Lemos 2004). Other than the maintenance works taking place in roads opened in the Augustan and posterior periods, the region was gradually covered by a complex network of secondary trajectories ensuring the Roman domain and the progressive penetration of communities in the territory. These pathways linked the secondary urban centers with the economic exploitation areas but also allowed the access of exogenous consumption goods in rural areas, events progressively altering the traditions of indigenous communities.

From this period is also the consolidation of the cursus publicus served by a wide and efficient network of mansiones and mutationes. Their names are an important legacy of the famous Antonine Itinerary, even though several of the sites supporting the road network are still unknown to archaeology (Tranoy 1995-96). Related with the support of the main road network was the opening of the Via XVIII itinerary. This trajectory is also know as Via Nova and was referenced in the Antonino Itinerary (item alio itinere a Bracara Asturica). It was built in the Flavian period, more precisely in 80AD, opening a new trajectory between Bracara and Asturica Augusta through the heart of the Northwest mining region.

45

22 The outline of this itinerary on the current Portuguese territory measures around 34 miles and will be described in detailed in the following chapter.

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

3.3 Economic intensification

There are, however, some emerging works in surveying lesser known mining areas, namely in the conventus bracarensis territory. Within the conducted works, we highlight the one recently carried out by F. Sande Lemos and P. Morais (Lemos and Morais 2004) focused in inventorying the main gold mines located on the outskirts of the Via XVII itinerary, alongside studies undertaken by C. Martins (2005).

3.3.1 Exploitation of resources

Setting aside the idea about a synchronous and homogenous development in the peninsular Northwest Territories we must understand the intense development taking place in urban centers and the roads crossing the territory establishing the Roman domain. We must now question about the exploitation of basic resources and our attention must focus on the gold mining, Farming and fishing activities.

The process of metal exploitation seemed to have started in the Julius-Claudius period23 and it registered intensification in the late first century AD. There are, however, significant differences in the way the available resources were explored, depending on the deposits potential and of the type of exploitation. A distinction must be made between the territoria mettalorum, that were wide mining districts, represented by Las Medulas region in Asturias and Três Minas-Jales mining site located on the conventus bracarensis, in Portuguese territory, and the mining complexes of medium and small dimensions that were subjected to different exploitation techniques.

The Roman mining exploitation of the peninsular northwest started with the Augustan Pax and was fully organized in the Julius-Claudius dynasty. The development of this exploitation was intimately linked with the establishment and consolidation of a system focused on production outflow. The cast of the main mining northwest regions has gradually been broaded since the 1970s, with Claude Domergues playing a key role in cataloguing the repository and studying the overall gold exploitation system (Domergue 1970a; 1970b; 1986; 1987; 1990; Domergue and Herail 1978; Domergue and Martins 1977; Domergue and Sillières 1977). The works of C. Domergue have been followed by studies conducted since the 1980s by the CSIC team led by Sánchez-Palencia over the gold mines located in the Asturicensis conventus (Sánchez-Palencia Ramos 1979; 1983; 2000). The conducted research over the Asturias gold mining exploitation led to a series of works detailing the processes of gold exploitation and analyzing the settlements and the associated social organization (Sánchez-Palencia Ramos e Fernández Posse 1993; 1998; Sánchez-Palencia Ramos et alii 1990; 1999).

The lack of systematic studies prevented us from acknowledging the fact whether known mining areas in northern Portugal would have had or did not have the territoria mettalorum status, as recently suggested by F. Sande Lemos (2004:237). Some areas deserving further research are Fojo das Pombas in Valongo, the FrançaBragança mining in Serra de Montesinho and the Macedinho area in Vila Flor. The relevance about the exploitation of the iron deposits in Torre de Moncorvo at the Reboredo Mountains is equally suggestive about this region having been a mining site. This was also suggested by Jorge de Alacão (1988) and confirmed by F. Sande Lemos (1993). Both northern Portugal and Galicia registered numerous testimonies regarding other gold exploitation areas. One example refers to the area around the Valongo, Gondomar and Paredes municipalities (Alarcão 1988). The relevance of this mining region suggest probably was a mining site explored by the state, similarly to what occurred with Três Minas-Jales (Lemos 2005).

Regardless of the extensive available bibliography on the topic, the mining activity still needs further research in terms of its dimension and meaning, alongside the economic and social organization of these activities. The chronological timeframe of these types of exploiation is not yet fully known. We are only aware about the location of the main mining sites of the Northwest where the gold exploits prevailed and were fully dependent of the state administration and control. We are referencing mainly the Asturias and the mining region of Jales-Três Minas studied by J. Wahl (1988; 1988-1989; 1998). Still missing are the systematic inventory works of the Roman mining exploitations and there is a considerable absence in information regarding the Galician and Portuguese coastlines (Lemos 2004). The same can be said about the Cantabria coastline where there are some indicators related with gold-mining and iron, led and zinc exploits (Fernández Ochoa and Morillo Cerdán 1994:173). There are also few monographic and synthesis works enabling the real assessment of the impact of the Roman mining exploits in the peninsular northwest. This type of research performs a paramount source of knowledge about the presence of the Romans within the region (Ferrer Sierra 1995; Lemos and Morais 2004).

46

There were also other gold mining complexes with medium sizes in the conventus bracarensis region and they greatly impacted upon the strategic economic relevance of the area. The main one, due to its dimension, was the entitled Poço das Freitas in the Aquae Flaviae territory (Lemos 2004:237). Prospecting works carried out by F. Sande Lemos (1993) in western and eastern Trás-os-Montes and more recently in Entre-Douro-e-Minho allowed the identification of a wide group of small mining complexes, certainly privately exploited although some may have had considerable sizes. From this group, we enlist Monte Furado in the Coura River, the one existing at the Nó mountainous range (Viana do Castelo), the mines referenced in 23 This took place in the Trás.os-Montes region where both the mining exploitation and the establishment of the road network system go back to the Julius-Claudius dynasty (Lemos 1993).

The Northwest Integration within the Roman World

Monção and the ones in Santa Justa and the Banjas Mountains. Small structures of auriferous and argentifera mining exploitation have also been identified in the upper basin of the Neiva River (Maciel 2003) in both Marrancos and Oleiros areas, in Vila Verde (Lemos and Morais 2004).

Northwest Territories. There are, however, areas in all the three conventus that have been better researched but still presenting significant research loopholes. Research challenges are mainly related with the overall approach of this subject by researchers and this is the end result of several circumstances. In some cases problems arise due to the difficulty in identifying Villas, as these structures are often located deeply underground, as it happens in some Entre-Douro e Minho and Galicia regions where only a small number of Villas have been excavated. Nevertheless, the presence of architectonic materials, epigraphy and ceramics around wide surfaces might indicate their presence. In other cases, identified Villas have only been partially excavated with limited identification of their rightful chronology, evolution and features. Moreover, the fact that we are dealing with structures that endured profound refurbishments throughout time also impacts upon the study of these sites.

If the exploitation of the mining resources of the Northwest took different ways, it is equally certain that they were associated with different settlement patterns. The gold mining exploitation conducted in the conventus asturicensis and in part of the lucencis conventus seemed to be linked with hii-forts founded in Roman times (Sánchez Palencia et alii 2000). However, in other cases, the mining exploitation implied the consolidation of preRoman Hill-forts, providing them with the role of secondary centers. Furthermore, there is evidence about the existence of open settlements of Roman foundation, namely of vici typology, that ensured the necessary exploitation mechanisms and outflow of production. These were naturally located nearby main road circuits (Martins et alii 2000; Lemos 2004).

When tracking down available bibliography related with the study of villas, we verify that most suggested chronologies present in catalogues and inventories has been based upon the study of mosaics, items that had only been present in villas in the later stages of occupation.

Available data related with the Roman exploitation of agricultural resources is yet rather fragmented due to the small number of excavations carried out on sites where establishments of the Villas is identified. However, we are aware about the changes these infrastructures brought to Farming and fishing activities and the consequent exploitation of these resources. Still, it has been difficult to identify the buildings associated with the new ways of exploring the land, as it is rather difficult to rigorously date the beginning and the consequent evolution of the process associated with the privatization of the soil demanded by the new Roman model of land exploitation. Thus, we are yet to fully understand the rhythm of Romanization in the rural areas of the peninsular northwest.

Moreover, foundation chronologies for excavated villas are older than chronologies suggested for the dissemination of this type of establishments. Examples are the Veranes and Murias de Bolono Villas, in Gijón, presenting High Empire foundations and the Boides Villa in Asturias (Garcia Entero 2001:50; Fernández Ochoa 1982:273, 284; Fernández Ochoa et alii 1997). In the Leon region we can find a similar situation with some excavated Villas of older chronology - El Soldón in Santa Colomba de Somoza, located in the surrounding area of Asturica Augusta and dated back to the Tiberius period, the Pago de Piégalo (Cimanes de Veja) Villa dated from the first century, the Monasteruelo (Velilla de los Oteros) Villa with a foundational date ranging between late first and early second centuries, the San Arian (Villoria de Órbigo) Villa, and Navatejera (Villaquilambre) Villa, located nearby León and dating from the end of the first and beginning of the second centuries (Garcia Entero 2001: 138; 141; 148; 150; 156).

As Villas defined a new model of ownership and exploiting properties, the establishment of these infrastructures must have had a profound impact in the overall organization of the territory and in the transformation of the rural landscape. Villas demanded the constitution of fundi that were in turn originated from the fragmentation of indigenous communitary properties of previous pre-Roman Hillforts (Pérez Losada 1987).

Based upon available data, any aproach concerning the development of the Villas in the peninsular northwest deserves deserves to be viewed with caution, due to research handicaps research, despite the crucial importance of the subject for the overall understanding of the reorganizational process of the territory and of the Farming, fishing and maritime resources exploitation, in particular.

The difficulties in understanding the phenomenon in a northwest regional scale has been aggravated by the regional diversity that the process might have encountered. This is an idea also supported by available bibliography on the subject. If it has been certain that differential pre-Roman economical development level took place in all three conventus areas, it must have influenced an early or later dissemination of Villas. We believe that the chronological differences witnessed by the dissemination of these types of infrastructures encompassed research challenges that are important to reference.

It seemed undeniable that the western Bracara Augusta area witnessed an early dissemination of Villas, namely from the first century AD onwards. Most of them have been located around the city and alongside the main itineraries (Martins 1990). The Villas location also followed the region surviving Hill-forts (Martins 1995) as these structures demanded an appropriation of lands previously controlled by indigenous communities.

We are not aware about the density of this type of exploitation, as there is no registration of systematic studies on the subject that can validate a rightful assessment for the

47

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Nevertheless, the generalization of this private model of agricultural exploitation is widely acknowledged as having taken place at a later stage in other regions, probably in Flavian times or from the second and even third centuries onwards. This has been the chronology given to Villas located in southern Galicia (Pérez Losada 1995). This has also been the case for most of the rural and coastline Villas located in the conventus lucensis. They have been dated with Low Empire chronologies with some few exceptions (Árias Villa and Villa Valdés 2005). Some of the best examples of these agrarian units, exploited bay landholders, are present in the Villas of Noville, Mugardos (Pérez Losada 1995), Toralla, in Vigo or Cambre, in Corunha (Casal Garcia and Blanco Sanmartín 2005:113).

and even when they were, most often they have not been subjected to extensive archaeological works needed to uncover the architectonic features of the sites. Nevertheless, available fragmented data suggest a clear dominance of the rectangular shaped plant Villas, presenting a corridor or portico flanked by lateral bodies or towers delimiting an open patium. The late Roman Noville Villa, located in Mugardos, is a perfect example of this architectural planning, alongside the Boelhe Villa, in Penafiel (Pérez Losada 1990-91; Pérez Losada et alii 1992; Pérez Losada 1995:175-179). The conventus bracarensis area is the one producing higher amounts of available data, although there is a significant information asymmetry for the western and eastern part of the referred territory (Martins 1990; Lemos 1993). The western part witnessed a high number and an early dissemination of Villas, events that seemed clearly associated with the foundation of Bracara Augusta. It is also possible to observe a different distribution of Villa between the High and Low Empire times, being these trends also the result of a gradual densification of these type of establishments (Martins 1995). This process, intensified from late first and throughout the second centuries AD, seemed to be associated with the consolidation of road network and with the development of secondary urban centers, , as it was the case of Aquae Flaviae (Amaral and Teixeira 1999) Tongobriga (Dias 1997) and Tude regions (Pérez Losada 2000).

The general idea of a late chronology for Villas located in the conventus lucensis, has also been suggested to the conventus asturicensis, although there has been some Villas existing in the León and Gijón region presenting clear High Empire chronologies. This type of sites is almost absent in the Cantabria coastline region and only a few examples, studied by F. Pérez Losada (1991:434), are known to have existed. They had excellent locations facing the coastline, likely practiced a mixed economy of exploiting agricultural and maritime resources, dating back to the Low Empire. The Noville Villa performed one of the best examples of these structures at the Cantabria coast, although only the pars urbana is known. The chronological diversity of what seemed to feature the privatizing process of land associated with the foundation of Villas is interlinked with the geomorphological diversity featuring the three conventus regions.

The expansion of Roman sites (vici, mansiones, mutationes) represent the thorough ‘colonization’ of the territory and also contributed towards a wider efficiency in exploiting agricultural and fishing resources, impacted in the growing of production and surpluses that prompted the consolidation of a market economy (Almeida 1996; Lemos 2004). This has been a key indicator of the Romanization process within the rural landscape and of the social and cultural transformation of its inhabitants.

While in the conventus lucensis the majority of Villas emerged on virgin lands, in the conventus bracarensis, namely within the Cávado, Lima and Minho river valleys, emerged on sites where pre-Roman settlements had been located. This suggests that the land exploitation in the bracarensis region could have been organized alongside potential aedificia housing pre-Roman communities located on the main river valleys.

Nevertheless, the economic development of the Roman northwest did not limit itself to the exploitation of minerals and to the intensification of food production or salting establishments, taking place in Villas also linked with the exploitation of maritime resources.

The differences observed in the three conventus have equally been observed in the choice over the location of Villas.

The new economic model was introduced and followed the integration of the region within the Roman world. These events greatly impacted upon the gradual exploitation of different types of available resources that allowed the increase of the production of goods needed by the new life habits generated in the context of a society, which has adopted constuction and consumption patterns clearly distinct from the ones existing in the previous period.

On the western part of the peninsular northwest, these infrastructures were mainly located in valleys or on smooth slopes with a good visual control over the cultivated lands. A concentration of such sites has been identified in the coast waterline, believed to reconcile Farming activities with the exploitation of maritime resources, including salt and fish preparing industries.

New activities arising from the new needs were associated with the emergence of new living spaces - cities, secondary towns and Villas. These sites demanded new construction techniques and developments requiring the exploitation of wood, stone and clay needed to produce ceramic construction materials. Unfortunately, not much is known about the social organization of these type of exploits, although they were likely conducted under private initia-

Villas located in interior regions seemed to have been dispersed around lands presenting high agricultural potential, thus occupying valleys and upper lands, although they seemed to be absent from mountain areas (Alarcão 1998; Pérez Losada 1995, 1996). Not much is known about the architecture of northwest Villas. It is true that only a few ones have been excavated,

48

The Northwest Integration within the Roman World

tive, with the exception perhaps of developments in cities, that were likely promote and supported by the municipal management of these types of production.

Augusta. Several kilns of elliptic and rectangular shape are known to have produced great quantities of common pottery alongside with construction ceramics.

The exploitation of clay was greatly intensified in this period, although the indigenous communities already practiced it in their wares production. In fact, this raw material was indispensable in the construction of buildings of Roman typology, as they demanded greater quantities of clay materials used in pavements, roofing and in the thermal buildings. This use rapidly spread around the region as public buildings and private housing infrastructures integrated domus spaces. On the other hand, clay and kaolin were needed to produce ceramic, a product indispensable in daily lives, especially in their mundane culinary practices.

3.3.2 Market economy and currency circulation

From the group of resources existing in the Hispanic northwest, only gold had a strategic importance to Rome. However, the Roman economy was based upon taxed savings and fully depended on charged taxes, both to province capitals and to other regional centers. The exploitation of resources and the increase in production performed key strategies to generate wealth, raise exchanges and develop a market economy associated with the currency circulation.

Clay was a widely available resource found in the valleys of the main rivers crossing the Northwest Territory, as well as, at the coastline. Under these circumstances, the necessary conditions to the development of producing centers of construction ceramic were in place. These centers were located outside the cities. Bracara Augusta was supplied by a center located in the Prado/Ucha area, at around 14 km distance from the city. However, it is likely that pottery centers producing wares used in table or in kitchen would have been also located in the city. We know that the potters located in Bracara Augusta produced common ceramic, lamps, amphorae and probably sought to produce some products that would imitate the most appreciated shapes of imported fine wares. One of those productions was entitled ‘’bracarense ceramics’’. It was produced in the High Empire period and imitated the shapes of imported sigillata and lamps. With a considerable regional diffusion, the great quality of this production manufactured with clay of the Minho coastline displayed an excellent knowledge about the available regional resources alongside their systematic and selective exploitation (Morais 2005). Nevertheless, we are aware about other fine wares production, namely thin-walled potery. The assessment of available remains led to the conclusion that the ceramics industry was one of the most significant activities functioning at the city of Bracara Augusta. There are referenced to the city kilns, tanks and pottery accessories, as for example molds (Morais 2005).

Although our knowledge regarding these specific aspects of the Romanization process are still scarce due to the lack of research, it seems reasonable to state that the Northwest Territory endured significant economic changes linked with the diversification and the privatization of different resources available within the region. These trends prompted a significant growth in circulating goods, fed regional markets and originated mechanisms to distribute them at an interregional level. The rise in exchanges substantially increased the densification of the road network system, with the creation of both urban and rural markets and with the widespread of currency. Wider magnitude markets were naturally located in the conventus capitals boosting an economic relevance that has been well archaeologically documented through the numerous importing products uncovered. Bracara Augusta had a rather diverse economic life helped by its proximity with the coastline that was in turn linked to river and land routes ensuring the safe arrival of products circulating in the Atlantic to the city. According to R. Morais (2005:134) the city disputed with Lucus and Asturica the role of redistribution center although its action did not surpassed the south margin of the Douro River. Research focused on the study of amphorae located in the city shed some light on the privileged relationships taking place with other production centers in Baetica throughout the High Empire. Wine amphorae performed the majority (around 80%) of uncovered remains and were followed by fish sauce (around 10%) and olive (around 1,5%) amphorae (Morais 1998; 2005). The city also registered the presence of imported thin-walledpottery, mainly of Italian production performing the majority existing and in relation with the Gallic production of the pre-Flavian period (Morais 1997-98:70-71). Both types of production were, however, scarce when compared with the high volume of Hispanic production coming from Trititium Magallum (La Rioja) (Morais 2005a).

Other than the ateliers associated with the production of pottery (Morais 2005), there were others areas dedicated to the production of glass objects, namely recipients for domestic use, window glass and jewelry, as for example beads, ring stones and bracelets (Cruz 2002). Similarly to Bracara Augusta, we are aware about the existence of ceramic production centers located in other northwest cities and regions, namely in Asturica and in Lucus Augusti where the development of the pottery production seemed to have had an explosive moment from the Flavian period onwards. The city potters were able to sustain a florescent activity throughout two decades, producing several types of ceramic from which we highlight the manufacture of red ‘engobe’, also present in Bracara

Throughout the Low Empire, there was a predominance of late Hispanic production, when compared to other imported fine wares productions. Equally significant was the presence of late African sigillata throughout the fourth 49

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

century, revealing a strong dependence on the Atlantic route, where products supplying other markets of the Cantabrian region circulated. The predominance of amphorae coming from eastern markets in relation with African containers also revealed the type of power the city elites had throughout the fifth century. Under these circumstances, the exchanges between Bracara Augusta and the eastern Mediterranean markets should be highlighted. These events have been fully attested by the presence of late ‘focence’ and Cypriot productions (Morais 2005:134).

From the late Nero and Flavian period, the majority of sigillata products coming from Hispanic production centers would have their origin in Tritium Magallum. With regards to amphorae containers, these revealed different origins and a wide variety of shapes. Other type of products also arrived in Astorga, as for example glass items and adornment objects come from several Empire regions. A complex road network converging to the city ensured the overall city supply. Paramount trajectories were the circuits linking Asturica to Tarraco and the ‘Via de la Plata’ itinerary. Asturica Augusta performed the starting and end point of all routes going through the Cantabrian ports, alongside land routes connecting the peninsular northwest with the Meseta region.

Bracara Augusta, with an economic relevance well documented by epigraphy and archaeology (Morais 2005), had an important commercial activity that would distributed itself throughout the city mainly through the numerous establishments sheltered by the porticos flanking the streets. Some of this activity could also have taken place inside buildings established and focused on trading activities. The macellum located within the Cathedral area and with an epigraphic reference is a prime case.

With their starting point in Asturica Augusta, these products were distributed across the Asturias territory, both in the north and the south of the Cantabrian mountain range. In fact, sigillata had been found in several rural establishments and semi-urban centers of the region (Morillo Cerdán et alii 2004).

Similarly to Bracara Augusta, Lucus Augusti benefited from the Atlantic traffic and was supplied by the Brigantium port from the Flavian period onwards, although the city might has been served by the Iria Flavia port before this time. This performed the trajectory responsible for the arrival of thin-walled pottery from Italy and sigillata and lamps from southern Hispania and Africa. The Gallic sigillata came from production centers located in southern Gallia (La Graufesenque and Montans). It was replaced at the end of the Julius-Claudius dynasty by Hispanic productions coming mainly from the Ebro valley (Tritium Magallum). Throughout the fourth century, Hispanic sigillata products dominated the northwest markets and the African production items became rather scarce.

Secondary cities would equally perform important trading nodes and were supplied by the conventus capitals that would function as redistribution centers. These secondary central places would then supply the local communities alongside communities living in the surrounding rural areas. The supply of rural communities could also be achieved through the vici, if we take on the example of the Alvarelhos vicus, in Trofa. This site was located on the Via XVI itinerary and boosted a public area uncovered through excavations. This site is believed to have been an example of a market site, with a square surrounde by a porticus from where tabernae could be accessed. The fora also referenced in the Northwest might have functioned as important supplying markets of inner regions.

Amphorae containers were well represented, with a predominance of amphorae Haltern 70 and a scarcity of oil containers, in fact similarly to what happens in Bracara Augusta. The importing process seemed to include products from wider origins including the Egeu, Palestine, Africa and Rhone. Italian products were well represented and the ones of Hispanic origin come mainly from Baetica and the Tarraconensis region, overpowering Portuguese productions. In later stages, there was a clear preference for eastern wines of Jonian and Palestinian origin.

The economic currency of the peninsular northwest was not a single consequence of the growth in production or with the significant increase in circulation of local and imported goods. These two indicators, already visible on the first century AC, should not be considered as the mechanisms behind the development of a Roman market economy (Amaral 2007). In fact, it is important to consider other factors related with the integration of this region within the Empire, if we are to explain the new exchange patterns, from which the emergence of currency circulation is the prime example. Therefore, only the mechanisms associated with the Roman state power may explain the behavioral changes within the regional economy that from this moment onwards started to be included within the overall imperial economy.

Furthermore, Bracara Augusta registered the presence of glass items coming from the Rhine and eastern European regions alongside adorning objects coming from varied locations of the Empire. The uncovered material remains in Astorga confirmed that this city was certainly an important economic center since its foundation up to the third century. There were great quantities of imported thin-walled pottery mainly represented by Italic sigillata, lamps and Gallic productions coming from the La Graufesenque region, as well as lamps derived from Dressel 3 and produced in the southern region of Hispania and circulating alongside the ‘Via de la Plata’ route.

It was only after the integration of the commercial process between several northwest regions, encompassing a significant reshape of the landscape and impacting upon the relationship between production levels and consumption needs, that there was a structuring in the overall development of a type of economy based upon market exchanges alongside the adoption of currency as payment. Conducted research has suggested that throughout the first century

50

The Northwest Integration within the Roman World

BC Coins were not common, if not accidental, forming prestigious items more likely to be hoarded than to be manipulated as currency objects (Centeno 1987; Zabaleta Estevez 2000).

inance of coins emitted in Rome. The volume of these coins significantly grew both in the Flavian and Antonine dynasties, however, endured a circulation breakdown at the end of that dynasty until mid third century.

One of the main impulses behind the beginning of currency in the Northwest was related with the need to pay the military troops and with the development of trading activities in the southern peninsula. These events took place in the late republican period (Centeno 1987). However, it was the Cantabrian Wars that demanded a growing in the currency level, an events corroborated by the finding of the caetra coins both in Bracara and Lucus. These coins were stamped in order to pay soldiers (Martins 1999).

The increase in currency circulation at urban centers taking place in the third century seemed to be associated with an overall system reform, when the Antonianus was created and substituted the Denarius. The most abundant coins of this time, well documented in Bracara Augusta, related with the Galieno and Claudius II cases. With the Diocletian reform in the late third century, new bronze currency was emitted and an overall financial recovery was registered (Zabaleta Estévez 1999).

Furthermore, it would be after the campaigns of Augusto that a regular currency circulation was adopted for the peninsular northwest. Thus, it is important to underline that the currency circulation, together with the catalyst role that coins had within the regional economic life, strongly impacted upon the motivation and needs of the regional economic agents, with the Roman state being directly involved in the regional economy through the mining exploitation (Amaral 2007).

Both the issuing and the currency circulation at the fourth century seemed to initiate a new economical cycle encompassing three main stages. The first took place between 300 and 330 and the third happening after 360, featured smaller emissions and a rather scarce coin circulation, while the second stage, between 330 and 360, was associated with a significant increase in currency circulation, with propagandistic nature. The coins found in Bracara Augusta during the later period are Honorius emissions dating back to 393-395. From then onwards, the currency flows decreased although coins from the fourth century were still in used, at least, until mid fifth century, events attested by their systematic presence in treasures hidden in the fifth century throughout the northwest region.

There is another element to be considered: the relationship between the spread of currency circulation and the economy and urban development rhythms. These rhythms were partially lined with urban investments, public buildings regeneration programs and were in some cases inspired by the Emperors themselves. It is also important to notice the regular and growing presence of imported goods alongside the increase in currency circulation. These trends also revealed the way articulation between different levels of production and market were processed.

Currency circulation occurring in Bracara Augusta displayed a similar pattern to circulation flows registered in the remaining cities of the Hispanic northwest, and these trends can certainly be generalized across the rural territories.

The development of the peninsular northwest urban centers economy was accompanied by the spread of currency within the wider region. Hence, the assessment of currency circulation rhythms demanded a thorough study of coins present in the cities, since it is there that coins occur in greater quantities, but also because its long occupation allows evaluating the evolution of currency flows in larger cycle times.

The absence of studies related with currency circulation patterns in secondary urban centers prevented us from fully understanding the way the spread of market economy was processed outside the main peninsular northwest urban sites. These sites would have functioned as secondary and tertiary markets supplying the surrounding regions.

The available data emerging from conducted studies and regarding the cities of Bracara Augusta (Centeno 1987; Zabaleta Estévez 1999; Amaral 2007) and Lucus Augusti, mainly from Augusto and Tiberius periods, revealed that both cities currency flow followed on similar patterns, with currency becoming a paramount element in the overall urban economy (Centeno 1987). Initiated in the Augusto period, the northwest currency circulation took place with caetra coins, likely issued between 27-23 BC and linked with military campaigns. The closing of Hispanic emitting centers during Caligula’s reign, whose currencies have master the circulation in the time of Augustus and Tiberius, prompted a shortage in currency circulation up to Claudius period. The coins from this Emperor, issued in Hispania after the year 41 AD, were well represented in Bracara Augusta and in the northwest in general. From then onwards, currency flows started to register a predom-

51

What has been stated is not compatible with the exercise we set ourselves to develop on the present chapter. We must look upon the group of available archaeological data and verify its spatial distribution. The goal is to identify, although temporarily, layers of information that can be valued on other work scales. As referred in the introductory part of the present work, the increase in intelligibility over the space occupation is only meaningful alongside a permanent ‘coming-and-going’ effect of different analysis scales.

Chapter 3 An Overview on the Entre-Douro-EMinho Roman Settlement Patterns Introduction

Research focused on the Roman settlement patterns of the conventus bracarensis western façade started with a wide set of tasks developed in the Entre-Douro-e-Minho region. In the present work, this particular region will be entitled as ‘reference matrix’.

On chapter 2, we acknowledge some of the settlement features from the Bronze Age onwards, and these traits have distinguished this region from the remaining peninsular northwest. When synthetizing different behaviors, we do not hope to turn them into something rigid; thus we understand the three northwest areas defined by the three conventus, as pragmatic entities that were able to fixate some significant trends in the ways they were occupied and evolved. We are dealing with urban centres established by the Romans and with a progressive ‘filling’ of space by secondary urban sites and with the implantation of new rural establishments. Taking these trends into consideration and bearing in mind the future analysis to carry out, the organization of the present chapter started with an approach of fortified habitats aiming to obtain a basic cartography of its distribution (point 1). The cross-reference of data will enable us to debate, with some difficulty, the chronological challenges inherent to these settlements. The Roman road network system will be summarized in point 2. Then, we will assess the continuity and the transformation of the casual centrality of some Hill-forts, in articulation with the geographic implantation of both the castella and the vici (point 3). However, we will not summarize of the main regional urban centers, namely Bracara Augusta and Tongobriga, as their integration was provided on the previous chapter and included a general approach over the Roman urban programs established on the Northwest. Point 4 will include a list of Villas remains and some considerations over their features and distribution.

Figure 4. Location of the Entre-Douro-e-Minho area in Roman Hispania (elaborated from http: //commons.wikimédia,org/wiki/category: Maps-ofRoman-Hispania)

With a well-defined geomorphological unit, the recovered archaeological data provided us with a general perspective on the region; however we must clarify how we utilized this information. What have been relevant to us throughout this chapter are not the development mechanisms that help us to identify the regional settlement patterns, but instead the acknowledgment over the presence of certain occupational features in the territory. This exercise is only relevant if we consider it as an attempt to highlight work hypothesis and key distribution points related with their own development patterns, thus enabling a readjustment of the archaeological perspective over the settlements patterns of the region.

1. The fortified settlement

1.1 Analysis categories and spatial distribution

The study of settlement throughout the Roman times requires an understanding about the pre-Roman reality. If this premise must be present in any work on the subject, it is mandatory in the region we are dealing with. Only an understanding about the settlement patterns prior to the dissemination of the Roman power may explain the overall implantation and development frameworks.

We will not develop on the present chapter what we will approach in further detail on chapters 4 and 5. We believe that the operative definition of settlement patterns, valued on those chapters, demanded a stronger link of the archaeological remains existing in a circumscribed geographical context. Taking into consideration our present knowledge, it was not possible coordinate that particular goal for the Entre-Douro-e-Minho territory as the reliability, detail and scope of data were insufficient for us to conduct such procedure.

The synthesis of the main approaches defining the indigenous reality within the peninsular northwest context has been elaborated. What matters at this moment is to change

52

An Overview on the Entre-Douro-E-Minho Roman Settlement Patterns

the scale of analysis, produce inventories on available data and analyze the features of pre-Roman settlements located in the Entre-Douro-e-Minho territory.

tlements located in small hills of low altitude and located around the valley areas (C).24

This region has 365 sites interpreted as Hill-forts. However, a substantial part of these references cannot be subjected to detailed analysis, as they do not hold sufficient and valid data. These circumstances also apply to the Roman settlements. We only have considered sites that provided us with credible data: site description and respective topographic implementation, occupational area, number of walls and, in the absence of excavations, a presentation of the material remains visible on the surface, even if only as a summary. After analyzing the sites from a total of 365 surveyed Hill-forts, we have selected 278, which meant a loss of 24% of our total sample. We then tried to apply criteria that enabled us to detect the differences existing in this still wide group of sites. The organization of this data must be done according to a group of variables of diverse nature, namely chronology (implementation, retraction, abandonment and reoccupation processes) and the smaller or largest degree of Romanization present in each example.

It is important to refer that this topographic criteria does not imply, on the present case, any attempt to apply to the Entre-Douro-e-Minho region the conclusions drawn by Manuela Martins in terms of hierarchy and spatial distribution of the Iron Age settlements located in the medium Cávado River basin (Martins 1990: 206-211). In fact, this task is only possible to conduct in circumscribed areas. We only considered the distinction between fortified sites in relation to their location in the landscape, as this approach seemed to be more suited with regards to the group of data under analysis. Moreover, the topographic criteria raised some problems resulting from difficulties with proceeding with a required altitude calibration altitude, considering the altimetric spectrum of the various geomorphological units under observation. On the other hand, the small and vague descriptions of some Hill-forts prevented a rightful assessment of their topographic implementation. Despite these challenges, the topographic criteria enabled us to organize data according to a structuring element: the space. This basic arrandgement included a series of other type of data, such as size or the occupation chronology that was, whenever possible, assessed together.

After testing the potential variables, we understand our lack of knowledge in terms of occupational chronology, Hill-forts sizes or variables able to assess the Romanization level. Most of these issues will continue outstanding, above all because they imply projects of intensive field surveys and excavations, preferably occurring in small areas where the geomorphological conditions could be conveniently assessed and fully integrated within the wider settlement analyses. It is also quite necessary to conduct a critical reflection and develop a consequent redefinition over the assumptions that perform indicators of the Romanization level on each site and over the chronology of its different degrees.

We will now classify the selected sites in terms of their defined topographic categories, while commenting on some evident correlations. As the tables include the sites ID, also present on the maps, we will not include them on the text.

Hill-forts of A category

We included in this category a total of 69 Hill-forts. They share a privileged position within the surrounding landscape and register a significant development in the last two centuries BC and in the first half of the first century AD. From this group, sites registering no sign of Romanization are extremely rare. Although, it is not always possible to acknowledge the number of existing walls it is within this category we can find the best defensive mechanisms (sites presenting between 3 and 4 wall lines). It is also this category to include the widest sites in the region that may be classified as oppida (Map 2 and 3)

Thus, the contribution emerging from our study of indigenous Hill-forts should only be understood as an attempt to organize the amount of available data in relation with their interpretation. However, our work aims to be more than just an inventory of sites and different from the catalogues and inventories we normally have access to. After considering available data constraints, we produced a first list based upon topography that performed this selection differentiator criterion. This first list was also based upon the three categories used by Manuela Martins (1990: 206-211) for the medium-course Cávado valley region. The first category grouped settlements located on rock knobs or rock knobs ends formations, hence detaining a privileged position within the landscape (A). The second included habitats located on hills registering medium altitudes, thus detached from the surrounding landscape, however lacking the visibility held by first category sites (B). Finally, the last category encompassed set-

24 We didn’t include in the A, B and C categories some sites tha we were not sure about, although they are present on the final catalogue (vol. III). This was mainly related with the fact that they were from the Bronze Age or the Middle Ages periods.

53

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

ID 3011602 3020501 3020601 3027402 3021202 3022201 3023202 3023701 3025401 3028101 3031302 3034401 3031801 3032002 3040601 3041301 3040001 3061501 3084101 3091402 3091901 3092305 3112001 3110101 3111303 3121702 3121801 3123201 3130601 3131101 3133803 3134501 13070102 13072202 13073102 13091501 13110101 13110501 13111301 13112602 13112902 13131201 13140301 13141301 13160501 16013101 16020101 16021901 16040101 16040401 16041101 16042404 16042802 16050401 16051901 16071821 16073901 16075101 16090301 16090503 16090705 16090901 16091001 16092201

Site Castelo de Espinho Outeiro do Crasto Crasto Monte de Lousado Carmona Monte da Sala Monte de S. Mamede Castro de Faria Citânia de Roriz Picoto dos Mouros Monte da Falperra Eiras Velhas Monte das Caldas S. Mamede-Guizande Cividade de Chacim Alto dos Moiros Monte do Castelo Castro de S. Lourenço Citânia de Briteiros Santo Tirso Castro de Lanhoso Monte Castro Castro de Vieira Castro Anissó Castro de Rossas/Mte. Castelo Castro de Peniche Castro das Ermidas Castro das Eiras Castro do Barbudo S. Julião Penedos da Portela Castro de Escariz Monte de Santiago S. Tiago (Bogalhos) Castro de Quires Citânia de Sanfins Penha Grande Cividade de Eja Castro de Abujefa Outeiro do Dino Quintela Cividade de Terroso Castro de Alvarelhos Castro do Monte Padrão Cividade de Bagunte Castro de Reboreda Cividade de Âncora Coto da Pena Penha da Rainha Castro da Assunção Castro de São Caetano Cividade Riba de Mouro Castro da Sª da Vista Forte da Cidade Alto da Cividade Stº Estevão da Facha Castelo da Ermida Alto de Valadas Alto da Coroa Monte Crasto S. Silvestre Alto dos Mouros Castelo de Neiva Santinho/Roques

54

Municipality Amares Barcelos Barcelos Barcelos Barcelos Barcelos Barcelos Barcelos Barcelos Barcelos Braga Braga Braga Braga C. de Basto C. de Basto C. de Basto Esposende Guimarães P. Lanhoso P. Lanhoso P. Lanhoso V. Minho V. Minho V. Minho V. N. Famalicão V. N. Famalicão V. N. Famalicão V. Verde V. Verde V. Verde V. Verde / P. Lima M. Canavezes M. Canavezes M. Canavezes P. Ferreira Penafiel Penafiel Penafiel Penafiel Penafiel P. Varzim Trofa Santo Tirso Vila do Conde A. Valdevez Caminha Caminha Monção Monção Monção Monção Monção P. Coura P. Coura P. Lima P. Lima P. Lima V. Castelo V. Castelo V. Castelo V. Castelo V. Castelo V. Castelo

Altitude (m) 480 222 260 309 300 303 409 298 324 226 563 480 304 427 640 320 550 200 336 461 385 202 550 733 730 99 209 384 331 296 453 235 480 500 512 570 248 207 330 274 119 153 222 413 206 300 187 79 472 265 342 259 244 376 280 142 443 347 212 160 281 385 183 269

An Overview on the Entre-Douro-E-Minho Roman Settlement Patterns

16092402 16092902 16090001 16092602 16093801

Castelo de S. Martinho Peso Citânia de Santa Luzia Vieito Cotorinho

V. Castelo V. Castelo V. Castelo V. Castelo V. Castelo

470 183 226 122 226

Table 1. Hill-forts of A category

The concentration of these Hill-forts within the most significant reliefs of the hydrographic Minho, Lima, Neiva, Cávado and Ave river basins is visible. Furthermore, this concentration mainly occurred in areas corresponding to medium and low river basins, normally occupying rock knobs or rock knobs ends formations and located in areas delimited by contour lines of levels between the 200 and the 450 meters of altitude.

river basins, we find the S. Caetano Hill-fort located in the Minho River. If we head south, we have the Escariz and Castro Seco Hill-forts, as well as the ones of S. Julião, Citânia de Briteiros, Eiras and Citânia de Sanfins.

Hill-forts of B category

We have included in this category a total of 98 Hill-forts that are located on predominant hills within the landscape. They registered altitudes of around 100 and 350 meters, however there are a small number of sites with altitudes higher than the average.

It is obvious that the Hill-forts altimetry offered variations adjusted to their position in relation with the different hydrographic basin areas. Hence, on the coastline, where the relief is less stressed, there were Hill-forts of this category located between 100 and 200 meters of altitude, while the ones located around mountains foothills registered altitudes of around 500 meters of altitude.

One of the challenges already signaled for the medium Cávado region was linked with the difficulty in detecting in this category of Hill-forts the defensive systems, remains of residences or even surface remains, which is due, above all, to their topographical features (Martins 1990). The hills where the sites are located are constantly wooded, with thick layers of furze, which makes it difficult to acknowledge the existence of structures and material remains. On the other hand, these habitats proximity to present day Villages make them the constant objects of stone looting activities, subjected to urbanization processes and included in the overall agricultural exploitation of the land.

The observation of cartography catered for other considerations. On the coastline, there was a group of large Hillforts, which seemed to define a group of settlements intimately related with the control over the coastline and river mouths, alongside the exploitation of the resources. These performed the a settlement pattern that benefited from both the maritime and fluvial trade, likely functioning as potential redistribution centers of products circulating in the Atlantic proto-historic route. Some examples of these Hill-forts are Cividade de Âncora, Coto da Pena, Citânia de Santa Luzia, Neiva Castle, S. Lourenço, Cividade de Terroso and, probably Cividade de Bagunte and Alvarelhos.

Regardless of these challenges and disparities, it was still possible to observe some patterns. The defensive apparatus of the Hill-forts included in the B category were clearly less elaborated than the one found in A category of Hill-forts. Most sites only displayed 1 or 2 wall lines.

This group of coastline Hill-forts extended towards the Galician coastline and included the well-known Santa Tegra, Vigo and Monte do Facho settlements (VVAA 2006).

In terms size, the number of large Hill-forts is residual, and the majority referred to sites identified as ‘’small’’ settlements.

This is a category of large Hill-forts located between the coastline and the interior valley areas. From the coast and moving for areas of the valley, we have identified the settlements of S. Silvestre, Santinho-Roques, Carmona, Citânia de Roriz, Faria, Monte da Saia, Ermidas and Penices.

Still in this category, we find a high number of sites lacking Romanization remains (35 cases) or registering insignificant surface data to enable us to assess their Romanization level (Map 4).

In a region normally known as a transition area, located between open valleys and areas corresponding to medium

55

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

ID 3010602 3011601 3012101 3020102 3021201 3022301 3022401 3022502 3022903 3022902 3022901 3026201 3032901 3033703 3031402 3041402 3040701 3060102 3070701 3070901 3072501 3072801 3082704 3082901 3084902 3087301 3085402 3087202 3110601 3111501 3111801 3120802 3120801 3123901 3123301 3123601 3124002 3124101 3124701 3130602 3132001 3131301 3131701 3132104 3135402 13061101 13071102 13072203 13110102 13111604 13112202 13130801 13160101 13162601 16010101 16010402 16020102 16021101 16021402 16021401 16021801 16031102 16040201 16040301

Site Monte de Caldelas Chã do Castro Monte de Sequeiros Monte do Facho Srª da Aparecida Crasto Castro de S. Simão Monte do Crasto Alto da Giesta Castelos Altar do Senhor do Lírio Santa Marinha Castro da Consolação Castro Máximo Pau de Bandeira Santa Comba Alto dos Mouros Cividade de Belinho Alto da Retortinha Santo Ovídio Castro da Portela Cividade de Paredes Castro de Sabroso Picoto de Santo Amaro Monte de S. Bartolomeu Monte da Forca Monte da Santa S- Miguel-o-Anjo (V/O) Outeiro do Crasto Outeiro da Coroa Castro da Atafona S. Miguel-o-Anjo (Cale) Castro do Facho Castro de Santa Tecla Castro de Santa Cristina S. Miguel-o-Anjo /R/D) Castro da Bóca Castro do Cruito Castro de Vermoin Castelhão Moega Dossãos Castro de Santa Engrácia Castro Mau Monte de Santa Cruz Castelo da Maia Castilho Soalhão Castro de Abragão Castelões Cristelo Monte de S. Félix Castro de Argifonso Castro de Boi Monte do Castro Monte de S. Miguel-o-Anjo Cútero Monte Góis Monte de Stº Amaro Picoto dos Mouros Nª Srª do Crasto Cividade de Paderme Cividade de Anhões Castro da Senhora da Graça

Municipality Amares Amares Amares Barcelos Barcelos Barcelos Barcelos Barcelos Barcelos Barcelos Barcelos Barcelos Braga Braga Braga C. de Basto C. de Basto Esposende Fafe Fafe Fafe Fafe Guimarães Guimarães Guimarães Guimarães Guimarães Guimarães Vieira do Minho Vieira do Minho Vieira do Minho V. N. Famalicão V. N. Famalicão V. N. Famalicão V. N. Famalicão V. N. Famalicão V. N. Famalicão V. N. Famalicão V. N. Famalicão Vila Verde Vila Verde Vila Verde Vila Verde Vila Verde Vila Verde Maia M. Canavezes M. Canavezes Penafiel Penafiel Penafiel Póvoa do Varzim Vila do Conde Vila do Conde Arcos de Valdevez Arcos de Valdevez Caminha Caminha Caminha Caminha Caminha Melgaço Monção Monção

56

Altitude (m) 220 310 228 170 250 186 236 202 298 160 140 240 340 170 570 310 283 137 250 332 500 580 278 309 421 266 249 385 337 450 521 194 268 328 336 291 289 210 356 200 300 250 217 217 170 108 416 442 200 275 209 208 157 123 285 180 270 344 100 302 130 159 453

An Overview on the Entre-Douro-E-Minho Roman Settlement Patterns

16040701 16041801 16042001 16050602 16050504 16060501 16061202 16070101 16070302 16070406 16071303 16071302 16071601 16071602 16072801 16071826 16072907 16073701 16075103 16075102 16100402 16080301 16080503 16080701 16090101 16090801 16091103 16091302 16092301 16092604 16092603 16092701 16092801 16093601

Castro de Cambeses Castro do Outeiro da Torre Monte dos Penedos Monte de S. Sebastião Portela da Bustarenga Penedo de S. Martinho Cidadelhe Monte Crasto Castelo da Formiga Castro de Stº Ovídio Castelo Pequeno Castelo de Genço Nª Senhora da Conceição S. João Castro de Vilar Castro de Trás-Cidade Portela Grande Castro de S. Simão S. Simão Cresto Torre Monte Crasto Alto do Corgo Crestos Castro de Agrichousa Crastos Velhos Monte Arculo Nª Srª do Crasto Nª Srª do Crasto Molhadoiros do Galo Calvário Cresto Monte dos Castelos Sabariz

Monção Monção Monção Paredes de Coura Paredes de Coura Ponte da Barca Ponte da Barca Ponte de Lima Ponte de Lima Ponte de Lima Ponte de Lima Ponte de Lima Ponte de Lima Ponte de Lima Ponte de Lima Ponte de Lima Ponte de Lima Ponte de Lima Ponte de Lima Ponte de Lima V. N. Cerveira Valença Valença Valença Viana do Castelo Viana do Castelo Viana do Castelo Viana do Castelo Viana do Castelo Viana do Castelo Viana do Castelo Viana do Castelo Viana do Castelo Viana do Castelo

315 247 131 198 493 221 180 210 365 388 248 274 297 54 250 210 450 442 180 176 220 269 120 178 140 112 169 248 106 312 53 200 177 110

Table 2. Hill-forts of B category

It is in this category that the called ‘castros agrícolas’ are included, if we take into consideration topographic criteria. This concept was used, for the first time, by Carlos Alberto Brochado de Almeida (1996). According to the author, the designation of ‘castros agrícolas’ aimed to individualize a group of sites detected on valley areas, particularly located in the Lima and Neiva hydrographic river basins. Moreover, the concept is destined to categorize small Hill-forts presenting one wall earth line followed by a trench. To these features we must add the fact that ‘castros agrícolas’ present remains on their surfaces indicating a foundation period between the late first century BC and early first century AD (Almeida 1996).

The distribution area of Hill-forts included in B category corresponded to the region registering the highest presence of sites included in the A category. The region between the Sousa and Tâmega rivers, namely the area located in the final course intersecting the Douro River, was where the presence of settlements was more residual. This particularity alongside with the distribution of the remaining Hill-forts seems to suggest a rather specific situation occurring in these areas, one that is mandatory of further detailed analysis in order to be fully understood.

The chronology of low altitude Hill-forts, in particular, the ones known as ‘castros agrícolas’ is very difficult to assess or even detail. This difficulty is visible in the successive dating suggestions of some authors (Almeida 1996; 2003b). Valley Hill-forts may encompass shapes that are identical in terms of topography, but with different chronologies that can go back to proto-historic times, to early Roman times or even date back to late Roman and High Middle Ages periods. The chronological differences suggested for many of these Hill-forts required a more systematic overview to be conducted on these types of sites. However, this particularity is not exclusive of the known ‘castros agrícolas’ (Fabião 2001).

Hill-forts of C category

We have listed a group of 40 hiil-forts in the C category that we can denominate as valley or low altitude settlements, as they have been addressed by Manuela Martins (1990). The majority of these sites are located in altitudes lower than 150 meters and they present only one defense wall that was normally followed by a trench. In some cases, there was the presence of one or two wall lines.

57

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

In this context, the case of the Lago Hill-fort, located in Amares (ID 3011503) performed a prime example due to the vulnerability of the chronological assessments based upon topographic and morphological criteria. Excavations supervised by Manuela Martins (1990) enabled us to conclude that this Hill-fort had an Iron Age chronology and registered no sign of Roman occupation, although initially we believed we were facing a Romanized site or, as was also suggested, a Roman military camp. The excavations conducted at the Sola settlement located in Palmeira (Bettencourt 1999: IIa 260-261) revealed a Bronze Age occupation, even though initially the habitat was believed to have had a similar chronology to the Lago Hill-fort.

reflect all the group of available Hill-forts that could have been included. We only attempted to list the valley Hillforts registering an Iron Age chronology or with a chronology from the transition period towards the Roman era. Furthermore, Table 3 does not include Hill-forts classified in 1996 by C. Alberto Brochado de Almeida as ‘castros agrícolas’ or the ones reclassified in 2003 by the same author as late Roman Hill-forts (Almeida 1996; 2003a; 2003b). In fact, low altitude Hill-forts with a late Roman chronology were excluded from our chronological approach and from the challenges of the present work. We did not attempt to establish the chronological distinction of listed Hill-forts in our cartography, although we have included the suggestions, whenever possible, of the authors with regards to the description of sites included in the catalogue.

These challenges greatly impacted upon the construction, although only temporarily, of a table including all low altitude Hill-forts. So, Table 3 and the corresponding Map (Map 5) should be interpreted with caution, as they do not ID 3011902 3010704 3011503 3021801 3024601 3060701 3092802 3130401 3130802 3135701 13111502 13113201 13160102 13160502 13160801 13161301 13162101 13162802 16020601 16020901 16021201 16021501 16042801 16043001 16043002 16070607 16070703 16071607 16072906 16074201 16072206 16075001 16100901 16101101 16080303 16080302 16081301 16091201 16091507 16093201

Site Outeiro Castelo do Catro Castro de Lago Castro de Carapeços Egreja Outeiro dos Picotos Monte de Santa Iria S. Sebastião Lugar do Gaião Outeirinho Castro de Irivo Suvidade de Recezinhos Castro de Casais Castro de Santagões Stª Marinha de Ferreiró Castro de S. Paio Castro da Retorta Castro de S. João Barrocas Boucinha do castro Barroca Crasto Crastêlo Monte Redondo Cristêlo Crasto Castro de Vinhó Castro do Eirado Castro de Baixo Stª Comba Quinta do Crasto Cividade das Donas Castro de Mentrestido Castro Paços Bacelar Outeiro da Madorna Cividade de Deião Cividade de Lanheses Igreja

Municipality Amares Amares Amares Barcelos Barcelos Esponsende Póvoa de Lanhoso Vila Verde Vila Verde Vila Verde Penafiel Penafiel Vila do Conde Vila do Conde Vila do Conde Vila do Conde Vila do Conde Vila do Conde Caminha Caminha Caminha Caminha Monção Monção Monção Ponte de Lima Ponte de Lima Ponte de Lima Ponte de Lima Ponte de Lima Ponte de Lima Ponte de Lima V. N. Cerveira V. N. Cerveira Valença Valença Valença Viana do Castelo Viana do Castelo Viana do Castelo

Table 3. Hill- forts of C category

58

Altitude (m) 175 96 65 180 61 58 110 100 50 74 157 285 85 63 42 14 40 30 60 279 90 35 127 73 78 120 220 30 96 13 163 30 253 44 50 140 55 21 45 30

An Overview on the Entre-Douro-E-Minho Roman Settlement Patterns

If we observe the distribution of Hill-forts included in the C category (Map 5) we verify that they were mainly concentrated in the Lima and Cávado rivers basins, with the exception of some sites located around the Minho River, particularly in an area close to the river mouth and to the intersection of the Ave and Este rivers in the coastline. They also corresponded to a distribution area including the main Hill-forts of A and B categories.

Although registering late chronologies, agricultural settlements may also represent a product of the reorganization of the sites and associated with the economic development featuring the region under study and occurring within the last two centuries BC and early first century AC. These events can also be materialized by the growing needs inherent to the land organization and to the exploitation of diversified food resources (Martins 1990; Alarcão 1999; Silva et alii 2005).

Although the chronological challenges are complex, it has been possible to state that the foundation of valley Hillforts is not a specific phenomenon of a transition era, similarly to what occurred in other peninsular northwest regions (López González and Alvarez González 1997; Martins 1988; González Ruibal 2006-07). The foundational periods of this type of Hill-forts throughout the Iron Age seemed to be associated with a clear proximity of the communities with the valley resources. The exploitation of agricultural and pastoral resources was significantly intensified in the last two centuries of our era (Martins 1990; Silva 1986).

Nevertheless, it is obvious the difficulty in advancing more than a list, forcibily "unstable", of the so called valley Hill-forts, given the wide chronological range attributed to habitats holding similar foundation and morphology. Finally, we must say that despite the concentrated settlement caractheristic of the Hill-forts, analized in the A, B and C categories, we will incorporate another type of habitat within the occupational shapes of the Iron Age period. They were located on mountainous bases, close to Hill-forts and don’t match the typical settlement standards attributed to this period.

The challenges inherent to the occupational chronology are therefore crucial. In the case of ‘castros agrícolas’, it directly interfered with an explanation regarding a phenomenon, which apparently was of short term. How to explain the foundation of these types of sites when the occupational periods, at times, does not even last a century?

They are referring to isolated habitats, with foundational periods prior to the Roman times and focused on the ‘’exploitation of the Hill-forts surrounding area (…) in centers, similar to Farms, and perhaps registering a seasonal occupation depending on the agricultural cycles’’ (Silva et alii 2005: 31). This type of pre-Roman occupation has been detected in field work surveys conducted in the Ave valley, where sites following these features have been identified and located around common Hill-forts (Millet et alii 2000; Silva et alii 2005:31).

If we acknowledge the chronology suggested by Carlos Alberto Brochado de Almeida (1996; 2003; 2003b), we are then dealing with settlements established at the same time as the foundation of Bracara Augusta and of the main road itineraries. This chronology also includes changes occurred at several pre-Roman Hill-forts from the second century BC onwards. Hence, the attributed chronology for ‘castros agrícolas’ must be articulated with the foundation of some Villas and with retraction or abandonment processes taking place in some Hill-forts. Moreover, it should also be related with an increase in commercial activities and with the effective start regarding the integration process of this part of the conventus bracarensis within the Roman Empire.

Quinta do Paço (ID 16071820) can be included in this typology of sites. It is located around the Monte da Facha Hill-fort, where a circular residence has been excavated and is believed to represent a model for this type of isolated aedificia of clear indigenous origin (Almeida et alii 1981; Almeida 1983; Silva at alii 2005).

1.2 Chronological approaches

It is clear that low altitude Hill-forts articulated themselves with the exploitation of valley resources and held distinctive foundation chronologies. However, it does not seem valid to associate the spread of ‘castros agrícolas’ with wider Roman strategies or programs linked with resettlement processes occurring in the region. Furthermore, some valley settlements, as for example is the Lago Hillfort case, in Amares, emerged in a time prior to the definite conquest of the Northwest and the foundation of Bracara Augusta. These events represented phenomena that, alongside the development of large Hill-forts, may be understood as a changing sign within the indigenous world and taking place in the last two centuries BC. This change was articulated with hierarchy and reorganization processes of proto-historic habitats.

The group of presented data included significant variations in terms of the topographic location of settlements, despite of occasionally providing chronological data. Although available data rarely enabled the assessment about the chronology of most of the selected sites, we attempted to define some chronological parameters that may be used to understand the Hill-forts occupation, if we base ourselves upon the referenced material suggested by some authors. We are mainly referring to surface materials. In the impossibility to produce a timeframe table, we tried to conduct a chronological approach accounting the different scenarios occurring at the Hill-forts located in the region under study. We defined five major chronological groups in which most of the selected Hill-forts were included. 59

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

The first group included a significant set of sites registering an abandonment process still occurring in the Iron Age, probably between the late second and first centuries BC. This chronology has been justified by the absence of remains with a Roman origin. Included in this group were also the majority of Hill-forts from the B category, meaning the ones located on hills and presenting medium and small sizes when compared to the large Hill-forts and the ‘citânias’ existing in the region. This group also encompassed some valley fortified settlements included in the C Category and the best studied example referred to the Lago Hill-fort located in Amares (Martins 1988). We can include in the second group Hill-forts including Roman remains, although this presence may be justified by trading processes. The remains account for imported wares, such as Late Republican or Augustan amphorae and italic sigillata, as well as exogenous pieces that were normally adorning objects. The chronology of such materials enabled us to conclude that this type of Hill-forts have been abandoned between the late first century BC and early first century AD.

settlements of the region and played roles in the strong economic and urban dynamism that occurred in the last century BC and in the first half of the first century AD. These trends were accompanied by a careful internal reorganization, with streets and quarters that could be considered from an oppida. Some exemples are Cividade de Âncora, Citânia de Santa Luzia, S. Caetano Hill-fort, Citânia de S. Julião, Citânia de Briteiros, Citânia de Sanfins, Cividade de Terroso and Cividade de Bagunte. In this context, we understand that the Romanization took on aspects not only related with the importing exogenous products from which wine amphorae are good examples, but also related with the acquisition and assimilation of consumption habits that were visible in the use of common Roman ceramics and other goods. Simultaneously, these hil-forts adopted new religious practices documented by existing altars and they started to integrate rectangular houses within their residential areas, covered by tegulae and imbrex. These changes have been corroborated by the significant presence of Roman remains seeming to indicate we were facing a process that was clearly Roman.

A perfect example of this situation is Santo Ovídio Hillfort (ID 3070901) located in the Fafe municipality. The site displayed several artificial platforms where some houses can still be detected alongside the remains of one wall line. This Hill-fort has been referenced since the late nineteenth century, due to the finding of the Gallician warrior statue. The excavation was supervised by Manuela Martins and took place between 1980 and 1984. It led to the conclusion that this Hill-fort revealed no remains of Roman origin due to the lack of Roman construction materials (tegulae and imbrex), although rare italic sigillata and amphorae Haltern 70 fragments have been uncovered together with several acus crinalis. These findings corroborate the commercial relationship conducted between this site inhabitants and Roman traders (Martins 1991).

The region wide settlements, included in the A category and in this particular group, were abandoned or, at least, endured a gradual retraction process within their residential areas in the late first century. However, it is possible that they met occasional occupation practices of wider chronologies in areas closer to the valleys. Finally, the fifth group corresponded to a group of Hillforts registering an occupation clearly subsequent to the Flavian period and surviving until the end of the first century or after. These are the cases of Citânia de Roriz and Carmona, S. Lourenço, Eiras and Vieira Hill-forts. In an attempt to systematize these trends we must add that Hill-forts chronological challenges must be, therefore, articulated with the abandonment, continuity or reoccupation processes of habitats. So, how is it possible to assess these phenomena?

The third group includes a wide group of Hill-forts where the only remains of Roman origin were reduced to fragments of tegulae, imbrex and other ceramic remains believed to be Roman. Due to the nature of available data, this is an extremely vague group in terms of several sites being described with the words: ‘Roman ceramic’ or ‘indigenous ceramic and scarce fragments of tegulae’ or even ‘rather incipient Romanization’. These were some of the references emerging from conducted work on sites, which made it extremely difficult to define chronologies and the Romanization degree, in particular in the vast majority of the selected sites.

First and foremost, we have to rethink and attribute greater value to the ceramic materials that can potentially be used to assess either abandonment or continuity stages. Several of the Hill-forts included in the valley site group and registering no sign of Roman occupation, may effectively have endured continuity in their occupation, surpassing the late first century. Hence, the exclusive existence of ceramics of indigenous manufacture may only mean that the Hill-forts performed habitats outside the circulation of imported products routes however they were occupied up to the end of the first century BC.

This data requires further research conducted in circumscribed and smaller areas and an extremely detailed excavation with special attention given towards the study of ceramics that are both Roman and of ‘indigenous manufacture’.

The presence of Villas within an occupation span including the existence of indigenous ceramics justified the need for an exact chronology to be attributed to material remains that were generically understood as of indigenous manufacture.

A fourth group included Hill-forts offering more definite evidence of an occupation dated from the mid and late first century BC. This group comprises most of the wide

Furthermore, the difference between indigenous ceramic dating back to the late first century BC and mid first century AD and previous productions from the Iron Age has 60

An Overview on the Entre-Douro-E-Minho Roman Settlement Patterns

been highlighted by Manuela Martins for the medium Cávado region (1990:160-163). It was actually possible to individualize both ceramics categories in technological and morphological terms, with the chronology covering most of the remains uncovered from the surface interpreted as ‘ceramics of indigenous manufacture’. The issue has been linked with a need to conduct further studies on similar ceramic artifacts found in other regions and to gradually adopt precise terminologies when describing surface findings.

cess is normally considered as a Hill-fort occupation continuity, as in reality what really exists is a different functional and spatial occupation of the same area. A later abandonment stage may encompass a wide variety of community settling choices, be it in the Hill-forts slopes, in Villas, in open sites of the vici type or in road stationes. The latter performed a typology of sites that in the meanwhile had been included in the rural landscape of the region. What has been interpreted as occupational continuity present in some Hill-fort surroundings may only actually mean the survival of a small number of inhabitants that still used part of the previous spaces and infrastructures, however pursuing different economic logics and ways of life, when compared to the ones existing at the previous habitats. Thus, in this period the Hill-forts would only perform a living alternative to communities that were already participating in an economic and social Roman context. These small surviving centers could be justified by their strategic role in terms of topography, which provided them with control duties over Roman road trajectories. This type of occupational continuity of the same spaces may be, in several cases, only related with the retraction of the population from these occupied areas.

Naturally, there is a difference between abandoned Hillforts in the third and second centuries and the ones that have been provided with abandonment chronologies between the late first century BC and early first century AD, due to terminology imprecisions. This group of sites should include a more distinctive category encompassing the group of habitats registering some degree of contact with the Roman world, mainly with regards to imported ceramics and with an abandonment process taking place in the transition period. If the presence of amphorae remains of the Haltern 70 production or of other imported ceramic artifacts can be related with commercial relationships, the use of tiles of clear Roman origin as it is the example of tegulae implied a significant shift in the habitats construction techniques.

In other cases, the continuity in occupation may have entailed the development of a settlement boosting different features from the previous Hill-forts, although it could also just had been an extension of the initial center and have occupied slopes and valleys surrounding areas. One example is the case of Alvarelhos Hill-fort (ID 13140301), which displayed a remarkable strategic location, a considerable size and a defense comprising three wall lines. The identified residential remains were located mainly at the intermediate platform excavated in the 80s and 90s of the twentieth century. They performed a site where circular houses are overlapped by structures of clear Roman typology, which indicated the presence of a vicus on the site and suggested a mansio function for the Via XVI itinerary (VVAA 2006: 82-83; Mantas 1996).

Although the presence of Roman objects was frequent within Entre-Douro-e-Minho sites from mid second century BC, the generalization of products supporting ‘Romanization’, which also meant the adoption of certain ways of life (food habits, construction techniques, dressing codes, amongst others) was not prior to the first century ADC and reached the indigenous habitats in different ways and with different intensities. In some cases, there was intensification in the use of construction materials, as for example the use of tegulae and imbrex destined to cover houses, and the use of Roman common pottery. The adoption of this type of artifacts that meant more than a mere substitution of containers, may be understood as a significant change in food and eating habits on behalf of indigenous population. Thus, they performed a key indicator of the cultural changes taking place within the region at this time, and related with the Roman impact upon existing communities. Another issue was related with the communities’ mobility, and it can only be fully understood in circumscribed timeframes and using different spatial scales. The abandonment of Hill-forts necessarily entailed the spread of their population. To fully understand this scope, we must take into consideration different scenarios for the location of their inhabitants and carefully consider the chronology for this process.

A third problem has been related with continuity and reoccupation processes, arising from the absence of excavations, of ceramic materials classification and from a bad interpretation of those concepts and artifacts. Some Hill-forts registering a similar medieval occupation have been included in this group. In several of these cases, it was not easy to assess whether there was in fact been a continuum use of the space up to the Middle Ages, or if we are facing a reoccupation of a previous Hill-fort that served different aims in its pre-Roman occupation and throughout the following up periods. In some Hill-forts, there seemed to have been an effective discontinuity between the Iron Age and the Middle Ages. In others, the presence of Roman materials enabled the interpretation of a continuous occupation span although, without further excavations it has been rather difficult to assess the timeframe and the spatial context of this same continuity.

Moreover, the abandonment of a Hill-fort in the late first century BC and early first century AD may have implied a transference of a large part of the community towards another site or to a surrounding city, or even the relocation of population in lower mountainous slopes, that in time became new residential areas, structured under new occupational logics, although they could maintain traditional construction techniques. Curiously, the latter pro61

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Alto dos Mouros Hill-fort (ID 3028101), also known as Picoto dos Mouros, served as a prime example. The site chronology has been established by an excavation supervised by T. Maciel (2003). Surface field work uncovered construction materials, indigenous ceramics, common Roman pottery and some amphorae and tegulae fragments (Almeida 1996). However, the research works also revealed an abandonment process occurring in early Roman times and a later reoccupation on the fourth century AD (Maciel 2003).

structural remains. This site is believed to have been a Romanized site, reoccupied in the country ‘reconquest’ and entitled ‘’Alpe Tamiel’’ in eleventh century documentation (Almeida 1996). The medieval occupation has been corroborated both by the findings of ceramics and the reusing of a wall surrounding the acropolis that seemed to have been rebuilt on that same period. Nevertheless, the vast majority of examples mentioned in literature as medieval Hill-forts reoccupation left us with serious doubts with regards to their real features and to the adaptability of the concept ‘reoccupation’.

A similar process may have occurred at Castro do Barbudo, located in Vila Verde (ID 3130601), where the excavations revealed an abandonment process occurring in Roman times and a clear medieval reoccupation located on the highest part of the hill. At this time, a castle was also established on a site with no previous occupational remains (Martins 1989). Equally, Monte do Castelo, or Castro de Lanhoso (ID 3091901) registered an occupation that goes back to the Chalcolithic period and subsisted throughout the Iron Age. However, the settlement seemed to have endured a later occupational stage throughout the Iron Age and a Roman incipient occupation. In the Middle Ages a castle was built at the top of the hill, in a place that didn’t seemed to have known previous human occupation (Martins 1990; Almeida 1978). The medieval occupations of both Castro do Barbudo and Castro de Lanhoso raised a pertinent question related with the meaning of the term ‘reoccupation’. It doesn’t seem to exist a clear overlapping of the previous spaces, but only of the hills, something that is natural considering the importance assumed by the defense and the strategic control needs by both indigenous and medieval communities.

The reoccupation of Hill-forts, therefore, implies a rigorous assessment of the whole interpretation process. The inclusion of medieval infrastructures on sites where old indigenous settlements existed may not imply the later reoccupation. Instead, the topographic features of the location act as catalyzers and actually determine the establishment of their defensive infrastructures. The abandonment, continuity, retraction and reoccupation processes demand a valorization of the mobility behind communities’ movements, be it in a large, medium or small scale. The mentioned processes hide the subtle movements of population (re)composition, and they demand an adequate research to be conducted and focused on the ‘thin’ mobility study. This particular study is critical in correctly assessing the transformation and appropriation mechanisms of the space and of the construction and development of territories.

1.3 Distribution areas

In our research focused on the study of the Hill-forts located in the Entre-Douro-e-Minho region, it was only possible to identify general trends related with location, distribution and chronology of the sites performing the settlement network. Such trends hide key regional and micro-regional variations. In fact, between the western part connecting with the Atlantic route and linking it with the Mediterranean and the eastern part of the conventus bracarensis reaching the Trás-os-Montes region and prolonging itself towards the Spanish territory, different settlement patterns have been found, both in pre-Roman and Roman times (Lemos 1993; Amaral 1993; 1999). Hence, there were significant variations to take into consideration.

Defensive organizational techniques from indigenous and medieval communities differed, as were different their concepts of living and protected spaces. This can be fully understood with the study of the Vieira Hill-fort/castle, located in Viera do Miho (ID 3112001). Recent research (Fontes and Roriz 2007:56) confirmed both a protohistoric and a Roman occupations, as it considered the remains of circular and rectangular constructions alongside the wide quantity of artifacts. At a later stage, a castle was built in the same area and it became the ‘Terra de Vieira’ headquarters, located in a higher area and presenting a plant that was rather easy to rigorously define. With some rare exceptions, the vast majority of descriptions available in literature are manifestly insufficient in adding up to the overall understanding about the true nature of these Hill-forts reoccupation practices. Thus, we must question if this particular terminology should only be applied to cases in which there really was a physical use of the previously occupied site, in terms of re-using residential infrastructures or defensive mechanisms.

The observation of the map representing the settlement distribution pattern of the A, B and C Hill-forts categories and the acknowledgement regarding the existing information gaps, enabled us to clearly distinguished three main dispersion patterns that seemed to display different features. Beyond the region demarcated by rivers and Lima and Ave, where lies the main area of mapped Hill-forts, it must be highlighted another area, which includes the group of sites implanted in the periphery of the mountainous interior region, particularly those located between the Cávado and Tâmega rivers. A third pattern encompassed

Crasto Aldreu, in Barcelos (ID 3020601) is an example. The Hill-fort registered the presence of surface materials including Iron Age pottery fragments and ceramic artifacts of Roman and medieval origin, alongside residential 62

An Overview on the Entre-Douro-E-Minho Roman Settlement Patterns

a group of Hill-forts located between the Sousa and Tamega rivers and representing a distinctive group from the one observed for the Lima and Ave region.

somehow associated with the transport network system. This is the case of Monte de S. Mamede - Guizande (ID 3032002), located in Braga and near the Via XVI itinerary, Eiras Velhas - S. Mamede de Este, located in Braga (ID 3034401) near the Via XVII itinerary and the Castro de Caires, located in Amares (ID 3010501) around the Via XVIII itinerary.

Creating a settlement hierarchy was not the idea behind our study, nor it was to analyze the political dependencies existing between sites. However, this doesn’t mean that they don’t exist or that they can’t be detected. In fact, settlement hierarchy models were identified in certain regions as it was demonstrated for the Cávado region (1990) by Manuela Martins or by Tarcísio Maciel for other areas of the Neiva River (2003).

It is possible that, in other circumstances, the Hill-forts may have functioned as mansiones. This could, for example, be the case of the S. Lourenço habitat (ID 3061501), located in Esposende, nearby the sea and on the Cávado River outfall. This case is particularly pertinent due to its proximity to Roman road remains that existed on the site foothill. This circumstance could have ensured the site continuity and relevance in Roman times (Almeida 1996). A deeper analysis of the group of Roman remains scattered around the valley area and around this Hill-fort should be conducted in the future, trying to understand some important issues. Did the S. Lourenço site meet the needs and necessary conditions to support the circulation on the road? Was this circumstance, yet to be determined, incompatible with the existence of a simple mutatio located on the valley?

1.3.1 The area between the Lima and Ave rivers

One of the most significant areas concentrating the majority of available records is located between the Lima and Ave rivers. It is delimited on its eastern part by the high basins of the Lima and Ave rivers, where the foothills of the region inner mountain ranges are also located. We are dealing with a region revealing an intensive protohistoric occupation. This occupation is well represented by the existing Hill-forts that are located in different topographic contexts and present different chronologies and sizes. Moreover, this is a region that has been subjected to an intensive contact with the Mediterranean, mainly through the Atlantic route. These circumstances are also well attested by the presence of Punic and Atic pottery fragments. Furthermore the initial contact of the region held with the Romanized world justified the presence of ‘campaniense’ pottery and some republican currency items. This is the region where large oppida emerged, resulting from an evolucionary process, with circular houses distributed without apparent planning to proper rectilinear organization. Habitats revealed a progressive delimitation of residential neighborhoods, streets and opened areas emerging sometimes inside the acropolis, as for example occurred with the regional wide citânias.

The case study of the S. Mamede-Guizande Hill-fort located between the Este and Ave river basins, nearby the Via XVI itinerary and Eiras Velhas, overlooking the Via XVII itinerary in Serra do Carvalho mountain range, precisely on the transition area between the Braga region and Póvoa de Lanhoso, are suggestive of the way certain Hill-forts may have enhanced their centrality when compared to others. This would have ensured their apparent occupational continuity and would have been dictated by their location in proeminent topographic sites or in areas crossed by Roman road circuits. Nevertheless, there are cases in which the road control functions, primarily assumed by Hill-forts, ended up being exercised by settlements of Roman origin, located in valleys areas. This was the case of the Caires site, a site believed to have endured a retraction process or even abandonment when another settlement - the Cividade de Biscaia (ID 3010503) - presumably a vicus or a mansio was established around the Via XVIII itinerary.

The survival of Hill-forts throughout the Roman period and their articulation with the network of the main itineraries revealed the role some of these sites detained at the time in terms of supporting circulation. This importance has also been associated with different chronological moments.

1.3.2 The mountainous area

If we consider the main road trajectories, excluding the Via XVIII itinerary and the Bracara - Tongobriga road, projected and initiated in the Augustan period and completed during Claudius reign, we realize that the Romanized Hill-forts located around the road circuits may have performed a reference point or acted as a supporting structure in the overall development and establishment of the road network.

Mountainous area settlements, precisely the ones located on transition area between Entre-Douro-e-Minho and Trás-os-Montes, witnessed a rather different reality from the one we have been analyzing.25

25 To this scale of analysis one must add the expression ‘mountainous areas settlements’ and this was mainly used to highlight basic differentiation axis within a wide Entre-Douro-e-Minho region. With other scales, the phenomenon changed. Thus, we just needed to compare, even if superficially, the analyzed area of the Cabreira Mountain, in detail (approached here on the part integrating the Vieira do Minho municipality) with the one observed in the Gerês Mountain range and crossed by the Via XVIII itinerary, or with the

It is clear that the systematic work of articulating the road itineraries with the Romanized settlements, assessing the presence of Roman materials, performed a key task to conduct. In the meanwhile we may suggest some examples of Hill-forts with a life span that seemed to have been 63

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

The first impression regards the visible landscape rarefaction, a long-term phenomenon demonstrated by the synthesis maps elaborated for the Vieira do Minho region (Fontes and Roriz 2007: 10-20). On the other hand and as previously observed, Hill-forts of B category were there either absent or simply residual. With regards to the valley Hill-forts included in the C category, they just did not exist in this area.

been associated with the S. Pedro necropolis (ID 3111301).

1.3.3 The area between the Sousa and Tâmega rivers

If there is a distribution model more or less homogenous for the proto-historic settlements in the region under study, the model should also be circumscribed to the Lima and Ave region and not be applicable to the wide mountainous territories of the inner Minho territory. In order to characterize the landscape in the region, we have analyzed a study developed by Luís Fontes and Ana Roriz (2007) that exemplified the key ideas we aim to highlight. This will be approached in detail in chapter 5.

Keeping our observations at a macro scale level, we must highlight the area located between the right margin of the Douro River and the Ave river basin. This area registered different settlement patterns from the ones observed in the Lima and Ave area, and also from the ones that seemed to exist in the inner Minho region.

The settlement patterns existing in the area, located between the Sousa and Tâmega River, refer to the downstream territories, nearby the rivers intersection with Douro. This is a region that has been subjected to several excavations and archaeological field surveys, both in the territorium of Tongobriga (ID 13070901) (Dias 1997) and in the Sousa valley region, where Monte Mózinho (ID 13112101) is located (Soeiro 1984).

For the Vieira do Minho region, six fortified sites have been identified (Fontes and Roriz 11.12) and they exhibited a distribution pattern including two different situations, in terms of topographic location and size. The left margin of the Cávado River has been the location of Outeiro do Crasto (ID 3110601) and Outeiro da Coroa (ID 3111501). They existed on top of rocky formations and headlands at the center of the Cabreira mountain range slopes.

Research work supervised by Tavares Dias in Tongobriga shed some light on a particular detail. Other than acknowledging the fact the city performed an important urban center, the works also revealed the city territorial hypothetic boundaries. These limits were bounded by vici, namely Várzea do Douro (ID 13072801), Quinta de Guimarães (ID 13021502), Meinedo (ID 13051301), Gatão (ID 13011601) and Lomba (ID 13011902). All these habitats were located in strategic sites in terms of road trajectories and may have been, in some cases, from indigenous settlements.

Connected with the eastern area of the Ave river there was a group of four settlements located on ‘’prominent reliefs dominating the numerous inner valleys of the river head’’ (Fontes and Roriz 2007:12). This group comprised Atafona (ID 3111801), Anissó (ID 3110101) and Monte do Castelo also known as Castro de Rossas (ID 3111303) and a wide habitat known as Castro or Castelo de Vieira (ID 3112001). The latter included an area of around 10 hectares and an impressive defensive system.

Some important sites should be highlighted in this region, namely Arados, Quires, Soalhão, Barreiro or Meinedo. Some registered occupation continuity in Roman times that can be explained by their location in strategically key areas in terms of accessing road itineraries. However, it seemed clear that in this region significant population movement took place in the lower mountain slopes, mostly in vici settlements, with some of them becoming important Suevi parish headquarters (Dias 1997; Fernandes 1968).

With the exception of Atafona, all the remaining sites included Romanization signs, and the Vieira and Rossas Hill-forts have actually been romanized. The proeminence and strategic location of Castro de Vieira made it a central site within the pre-Roman regional settlement context (Fontes and Roriz 2007:11) and a site that kept an important occupation throughout both the Roman times and the Suevi-Visigothic times. The location recouped its strategic relevance at a later stage, during the Middle Ages, when a castle was built and the site functioned as the Terra de Vieira headquarters.

The Santiago-Arados settlement (ID 13070102) located at 480 meters of altitude is nearby the road linking Tongobriga to Várzea do Douro. This may have justified its survival throughout Roman times (Dias 1997:303-304).

This region did not register evident community dispersion, mainly in terms of inhabitants moving from large Hillforts towards the valley in Roman times. The only exception that can be admitted occurred in the Rossas Hill-forts, or in the Monte do Castelo, as the Pombeiro habitat later emerged in the mountain slope (ID 3111301) and has

region belonging to the Amarela Mountain and with the Marão in order to acknowledge the differences. These differences can only be properly assessed within a regional work scope.

64

The Quires Hill-fort (ID 13073102), located on a hill with 512 meters of altitude, had four defensive lines and an occupation span up to the fourth century. The habitats occupied a privilege location and had access to fertile soils, which likely justified the continuation of its occupation and the potential articulation of the role of controlling the Roman road crossing the Croca valley and going towards the Canaveses bridge. The progressive installation of population in the hill slopes originated a settelement where Almeida Fernandes located the Baubaste-Bidebaste

An Overview on the Entre-Douro-E-Minho Roman Settlement Patterns

parish - civitas Bauvaste (Fernandes 1968:77; Dias 1997:315). The Soalhão Hill-fort (ID 13072203) located on top of a mountain included three defensive lines and presented round shaped houses. Apparently, this site was not romanized, although there was a Roman circuit located on the lower mountainous area nearby an incineration necropolis. These events seemed to indicate a potential community movement towards sites located on lower lands with altitudes between 280 and 300 meters and on sites where field crops known as ‘vale de Trajano’ existed, or even perhaps lower parts of the mountain range (Dias 1997:309). The Barreiro settlement in Stª Marinha do Zêzere (ID 13021503) and located at 430 meters of altitude seemed to represent the case of an abandoned Hill-fort perhaps in Roman times. This may be behind the displacement of communities towards lower mountainous sites, specifically towards Quinta de Guimarães, where a vicus that later became the Melga parish headquarters emerged (Fernandez 1968: 80; Dias 1997: 300-301).

It is possible that the lack of remains may be linked with research challenges. Only two settlements have been referenced to have existed in the Valongo municipality: Couce and Staª Justa also known as Cavada dos Crastos (Silva 1986: 85). The latter housed an important mining exploitation unit located in the Fojo das Pombas area (ID 13150502). This area has always been linked with gold mining activities. This is the region where the key mining unit of Serra das Banjas (ID 13102002) was also located. This relationship was likely behind the fact that the territory knew a specialized type of fortified sites and that they were developed in under a certain logic, that would include for example the awareness about the presence of shale in the soils and the fact this was not an ideal scenario for the exploitation of agricultural resources. Furthermore, we should also mention the supporting role played by fluvial routes and we highlight the outflows of Douro and Leça rivers and the exploitation of maritime resources. In this context, the Monte de Castelo-Guifões located in Matosinhos (Silva 1986:84), the Lavra villa (ID 13080302) and the Ivanta site in Valongo (Morais 2007) perform prime examples of where and how maritime and mining resources were exploited.

The Meinedo settlement (ID 13051301) performed another example of a pre-Roman site likely abandoned before the Flavian period. Its population may have been transferred to another site identified under the same name and classified as a vicus. This center, occupied in medieval times, was a Suevi parish headquarters and was known by the name of Magnetum (Fernandes 1968:77; Dias 1997).

2. The road network: trajectories and road stations

The territorial organization model under study has been structured around the importance of Tongobriga as civitas capital and the existence of key entities within the settlement planning, as for example has been the case of road itineraries. Another relevant factor is the important fluvial route that the Douro River was, which ensured the connection between the coastline and the inner Trás-osMontes territory.

When determining the main Romanizing element of the peninsular Northwest in the previous chapter, we have summarily detailed the connecting elements of this space. Focusing in the Entre-Douro-e-Minho territory, we will summarize the main features of the main transport itineraries crossing the region (Map 7).26 The main itineraries in the region corresponded to XVI, XVII, XVIII and XIX Roman roads and they all have been referred in the Antonine Itinerary. The region also included the Via XX itinerary – also known as via per loca maritime - and a circuit connecting Bracara Augusta and Emerita Augusta via Tongobriga and the inner Lusitania land.

This area was crossed by a dense terrestrial transport network that made Tongobriga a key strategic point for the communication of Bracara Augusta with Lusitania, also connecting the conventus capital with the mining exploitation region of the Marão mountainous range and with the western region of the northern margin of the Douro.

Some of the trajectories and secondary roads will be further detailed in the next two chapters. The itineraries exiting Bracara Augusta will be further described in the initial sections on chapter 4. On that same chapter, we will also analyze the secondary road of the Via XIX itinerary between the Roman city and the Cávado River. On the first half of chapter 5, the outline of the Via XVIII

The relevance of Tongobriga and Monte Mózinho, a fortified site established in Roman times, alongside several other human occupational remains known in the region both of pre-Roman and Roman origin, suggests a different settlement pattern from the one found in the area located between the Lima and Ave river basins. The absence of sites on the northern margin of Douro, mainly in the area between the Sousa river and the Atlantic sea, where important mining areas were also located, could be linked with the absence of research reflected in the lack of cartography. Some recent works account for projects in progress in this region will surely allow settlement patterns and exploitation of resources practices analysis to be further known (Morais 2007).

26 The outline of the transport system, with the exception of the XVII, XVIII and part of the XIX and XX itineraries, was based upon works conducted by Carlos Alberto Ferreira de Almeida (1968), Carlos Alberto Brochado de Almeida (1979; 1980), Jorge de Alarcão (1988) and on the synthesis maps integrating the work by José Ramón Menéndez de Luarca and Navita Osório (2000). With regards to the outline of the referred trajectories, these were based upon field works conducted by Fransisco Sande lemos, Luís Fontes, Ana Roriz, Rui Morais and works conducted by myself and referenced when necessary.

65

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

itinerary will be analyzed on its trajectory up to Portela do Homem. On the second half of that same chapter, we will put forward with a proposal for the Via XX itinerary, mainly between Braga and the Cávado river mouth. Specific data, as well as the interpretation of the road will be provided and follow a more specific approach regarding the selected three sub-regional units in order to develop a more detailed analysis. We understand that these three units entailed different organizational and territorial models and this in-depth analysis aims to assess the outline of itineraries and understand the relationship these had with the surrounding living habitats.

Starting in Bracara Augusta, the itinerary crossed S. Martinho de Sande, where a Milestone (CIL II 6214) from the Trajan emperor was found. It would then inflect towards Caldas das Taipas, Caldas de Vizela and Meinedo (Dias 1997). Here, it has been admitted the itinerary encountered a bifurcation: one road would lead to Monte Mózinho via Tâmega and Várzea do Douro crossing the Douro River, and the other would lead towards Tongobriga (Alarcão 1988; Dias 1997). It has been admitted that the connection to Tongobriga would be also made through more than one road. Lino Tavares Dias suggested two itineraries linking Bracara Augusta to Tongobriga that the author identified as 1 and 2. Both trajectories went via Caldas de Vizela and from then trajectory 1 would follow across the Odres river valleys, Póvoa, Constance and Canavezes, where the Tâmega River would be crossed through a road leading to Tuias, where it would finally reached Tongobriga (Dias 1997:319-320).

In summary, the following performs a general overview of our conducted approach. The Via XVI itinerary ensured the connection with Lusitania. It runned across the western part of the current Portuguese territory and link several important urban centers. The road linking Bracara-Cale performed a part of the itinerary also connecting Olisipo to Bracara Augusta. It had a permanent mile count and had as its caput viae the city of Bracara Augusta (Mantas 1996). The 35 miles mentioned in the Antonine Itinerary for this particular trajectory going up to Cale matched the existing field registration and the Adrian Milestone indication (CIL 4748) found in Braga (Mantas 1996: 840-851).

From this location the circuit headed south towards Aregos and Porto Manso. The crossing of the Douro River would take place at Porto Manso and then follow up to Castro D’Aire and Viseu (Dias 1997). In summary, the itinerary between Bracara Augusta and Tongobriga had a total of 5 Milestones. The first was found in S. Martinho de Sande and dated back to Trajan times (CIL II 6214). The Tuias Milestone, from Valentinian I and Valens periods (364-375), was found 1472,5 meters from Tongobriga ‘in the road existing nearby the city circus’ (Dias 1997: 320). It is believed this item performed the last mile before Tongobriga, which made Bracara Augusta the starting point.

Leaving Bracara Augusta, the road crossed the Veiga do Penso through Esporões and Bocas and exhibited an outline corresponding to the current national road 110. It then followed Portela, Muro (Quintiães) and Pousada. In the Pelhe river valley, the road matched the current national road 309 crossing S. Martinho do Vale, Santiago de Antas and Cabeçudos. The crossing of Ave River was likely done in a boat nearby the Santiago de Bougado site or via the Langocinha Bridge, where a Roman bridge is believed to have existed preceding the current medieval structure. From here, the trajectory would run all the way up to Trofa Velha, Sedães, Peça Má and Muro. According to Vasco Mantas, the latter site would link to a secondary road connecting the area with Guilhabreu via the coastline.

The itinerary between Tongobriga and the Douro River housed the remaining three Milestones present in this itinerary. The first one was found at the city exit, in the road linking with the Canaveses Bridge from the third/fourth centuries. Another Milestone was found in Soalhães and this item is of paramount importance as, apparently, it signaled the mile count starting in Tongobriga. It is a Constans monument (337-340) indicating mile 8 corresponding to the distance between the site and the city (Dias 1997). Finally, the Milestone found in Carreirinhas dating back to Gallienus (253-268) did not indicate the caput viae or the miles count.

From Muro, the itinerary would go all the way up to Ferronho, S. Pedro de Avioso, the Maia Castle and Santana do Bailio. The crossing of the Leça River would lead to S. Mamede Infesta, in a route with small deviations matching the current national road 14. The road would then follow the present Amial and Almada streets before reaching Cale.

The existence of an itinerary connecting with Tongobriga definitely dated back to the first century, although Trajan’s Milestone found in S. Martinho de Sande and the potential construction of the Canaveses Bridge are still believed to be the oldest transport elements of this itinerary.

The itinerary linking Bracara to Emerita Augusta allowed the connection with the Lusitania capital via Tongobriga and Egitania.

Regardless of the suggested Trajan chronology, it is possible that a prior trajectory connecting to Tongobriga existed before the Antonine period, and this belief is linked with the relevance of this urban center from the Flavian period onwards.

The circuit between Bracara Augusta and the Douro River included a dense network of secondary connections that have already been identified (Alarcão 1988; Dias 1997) and prompted the regular contact of Bracara with the important urban sites of Tongobriga and Monte Mózinho.

An identical process could have occurred in Acquae Flaviae. We acknowledged the existence of a Roman bridge on location from the Trajan period, similar to the Canav66

An Overview on the Entre-Douro-E-Minho Roman Settlement Patterns

ezes Bridge, and several of the found Milestones around the city are of later chronology. However, the Via XVII itinerary was probably constructed in the Julian-Claudian period as suggested by the uncovered Milestones suggest, although Aquae Flavia only became a municipality during the Flavian period.

Flaviae (CIL II 4783) (Lemos 2002; Fontes and Roriz 2007).

We believe in the potential existence of a road connection between Bracara Augusta and Tongobriga before the Trajan period. This road could have been subjected to further consolidation works in the Antonine period, and in the third and fourth centuries it started to indicate the mile count to whoever travelled from southeast (AregosDouro river) and Northwest (Bracara Augusta-Canaveses) (Dias 1997:321). It is also important to highlight the relevance of the road network consolidation works. They started in the Flavian dynasty and were carried out through the Antonine period, when the repair of the older itineraries is known to have taken place and when new roads were constructed, namely the Via Nova and the edification of several bridges. The Antonine period transport imperial program was mainly developed during the Trajan and Adriano reigns and it focused on improving the access to strategically important sites, in order to reinforce the integration of the northwestern peripheral regions. The potential repair works conducted on the Bracara-Tongobriga road and the construction of the Canaveses Bridge are acceptable to have been conducted as part of the overall Antonine emperors transport policy. The must also refer the identification of further four itineraries in Tongobriga region considered to have ‘’regional or local interest’ (Dias 1997: 322). One of them followed the left margin of the Tamega River and headed towards Várzea do Douro, an important vicus that marked the crossing of the Douro River towards the Paiva river valley. Another road would link the city to Bairral and Estrada and would cross, in Bairral, the previously mentioned trajectory linking with Várzea do Douro. A third circuit ensured the connection between Tongobriga and the Marão mountainous range via the Ovelha River and crossing the Vila Real area. Finally, the last itinerary would accompany the Marão mountain slopes connecting with Gondar, Gorbachã and the Teixe mines (Dias 1997:322-323).

The Via XIX itinerary connected Bracara and Asturica via Lucus Augusti. The trajectory between Braga and the Cávado River will be the studied and detailed on chapter 4 of the present work. From Prado, we respected the outline presented by Carlos Alberto Brochado de Almeida (1979) that suggested the passage via Moure and Marrancos up to Anais, Queijada, Fornelos, Feitosa, Arca and Ponte de Lima (Almeida 1979:107). From Limia, a mentioned mansio in the Antonine Itinerary, the circuit would run near Labruja River via Cancelinhas and the Arcozelo church up to Geira Bridge, also known as Geia (Almeida 1979:114). From here, it would cross the Arco Bridge located in Arcozelo and lead to Salgueiro, Devesa, Câmbua, Romarigães, the Codeceira Bridge, the Gândara mountainou range, Ponte Velha de Buibães and Portela de S. Bento. Finally, the itinerary would reach Valença and Arinhos, where the Minho River would be crossed before reaching Tuy (Tude). From Tude, the trajectory would connect to Lucus Augusti through Iria Flavia (Almeida 1968; Almeida 1979; Pérez Losada 2002; Fernández Ochoa 1995b). The Via XX itinerary linked Bracara Augusta to Asturica in a pathway following the coastline. It connected Brigantium (Lá Coruña) with both Asturica Augusta and Bracara Augusta. The exact chronology of this particular itinerary is unknown due to the absence of Milestones on its itinerary, which in turn makes it somehow controversial. The majority of authors considered that the circuit went through the current Portuguese territory up to Tuy. From Tuy onwards, it would follow a coastline trajectory accompanying the sea up to Brigantium. Other hypotheses have been suggested for its outline. One considered that the road followed on the XIX itinerary along a shorter trajectory via the Portuguese coastline up to the Minho River and from there it would match the already described circuit leading to Brigantium (Almeida 2003). Another thesis considered that the Via XX itinerary had an autonomous path when exiting Braga and it would accompany the Cávado River up to the coastline. In general, this is the hypothesis, which has behind the need to link Bracara Augusta to a watercourse port located on the same river and to the coastline. This connection is believed to have been critical in ensuring the supply and the outflow of products coming from the Atlantic route to the Roman city (Morais 2005).

The Via XVII itinerary ensured the connection between Bracara and Asturica Augusta via Aquae Flaviae. Exiting Bracara Augusta on the east, the trajectory would follow the Este river valley up to Carvalho mountains, where it would pass nearby the Eiras Velhas Hill-fort. From there, it would led towards a valley overlooked by the Póvoa do Lanhoso Hill-fort and via the northern slopes of both Cantelães and Cabreira mountainous ranges, already located in Vieira do Minho municipality. It would then leave Tabuaças and Campos, in a trajectory that may correspond to the national road 103 up to Outeiro dos Púcaros and Ruivães. In Boticas, Ruivães two Milestones have been uncovered but their whereabouts are currently unknown. One of them lacked any inscription and the other would indicate mile XLIII counting from Aquae

The identification of the road stationes has been, in the majority of cases, a complicated task to perform, in particular when it came to integrating settlements of wider dimensions, as it is the example of vici. The problem augmented when it came to the study of circuits not subjected to the official cursus publicus program, as it was the case of secondary roads study (Mantas 1996). The establishments needed to support stopovers in secondary circuits are believed to have emerged in rather spontaneous manners and dependent on the specific needs of the existing traffic. 67

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Nevertheless, a considerable number of road stations has been identified at the Entre-Douro-e-Minho region (table 4, Map 7). A large number corresponded to transport vici and performed sites that accumulated more than one func-

ID 13140301 3010701 3010503 3011501 3100605 3100602 3100309 3111403 16073501 3134202 3060701 3080801 3085201 13051301 13072301 13020104 13072801 16091102 3060107 16100801 16061204

Site Alvarelhos Lugar da Igreja, Carrazedo Cidade de Biscaia Vilela / Mojeje S. Sebastião da Geira Chãs de Vilar Bouça da Mó, Terras de Bouro São Cristovão Ponte de Lima Igreja Nova, Prado Outeiro dos Picotos Caldas das Taipas Caldas de Vizela Meinedo Caldas de Canavezes Porto Manso / Quinta de Mosteiro Várzea do Douro Capela de S. Lourenço Alto da Ponte Forte de Lobelhe Cabeço do Leijó

tion. Around the Minho and Douro rivers, some of these vici are believed to have detained the key role of fluvial ports.

Municipality Santo Tirso Amares Amares Amares Terras de Bouro Terras de Bouro Terras de Bouro Vieira do Minho Viana do Castelo Vila Verde Esposende Guimarães Guimarães Lousada Marco de Canavezes Baião Marco de Canavezes Viana do Castelo Esposende Vila Nova de Cerveira Ponte da Barca Table 4. Road stations

From the identified mutatio, we will reference Bouça da Mó located around the XXX mile of the Via Nova. This performed a rare example of a road station that has been fully excavated. This was an isolated building measuring 21x14 meters made of granite with probable wood-made internal divisions. On the southeast side, one can find remains of a small tavern that likely functioned on the outside area (Lemos and Baptista 1995-1996; Martins 2005:14-15).

68

Itinerary XVI XVIII XVIII XVIII XVIII XVIII XVIII XVII XIX XIX XX Via Bracara - Emerita Via Bracara – Emerita Via Bracara – Emerita Via Bracara – Emerita Via Bracara – Emerita Via Bracara – Emerita Secundary Via Secondary Via Secondary Via

Site Type vicus mutatio / vicus vicus mansio / vicus mutatio vicus mutatio mansio vicus? mansio mansio vicus vicus vicus? vicus mansio vicus mutatio mutatio vicus vicus?

An Overview on the Entre-Douro-E-Minho Roman Settlement Patterns

Several authors tried to define the network of central sites from the Iron Age based upon their relative topographic location, in order to understand the pre-Roman settlement organization and the potential existing hierarchy of the several Hill-forts existing in the region. Some suggestions have been made by researchers who studied the preRoman settlement network in general, and in restricted areas normally located around the Minho hydrographic basins, in particular. From the list of autors developing suggestions, we highlight Alain Tranoy (1981:72), Armando Coelho F. Silva (1986:43; 1999:44), Manuela Martins (1990), Jorge Alarcão (1992:62-66), Carlos Alberto Brochado de Almeida (1995; 2003b) and Tracísio Maciel (1997).27 Despite the extensive number of potential central sites identified by different authors dating back to the Iron Age for the Entre-Douro-e-Minho region, there are still several Hill-forts that have not been excavated and that occupy strategic key positions within the regions they are located in. The available inventories are not sufficient and further research must be conducted around smaller areas, in order to produce a more complete analysis of the castella detaining a key geopolitical role within the pre-Roman settlement network. Nevertheless, a significative number of potential central sites from the Iron Age is already available and sufficient enough to enable us to clearly understand the inability of establishing a clear continuity lined in terms of the geopolitical relevance these sites had within the established territorial organization network by the Roman administration focused on this region.

Figure 5. Bouça da Mó mutatio plan (Lemos and Baptista 1995-1996).

Regardless of the survival of several of the mentioned large castella, performing real pre-Roman oppida - in Roman times, they seem to had lost some degree of the power they detained prior to the administrative reorganization implemented by Augusto for the peninsular northwest. Thus, the majority of ‘citânias’ was either abandoned in the mid/late first century AD or endured a significant retraction process in their occupational areas. Part or the majority of their population moved towards mountainous slopes or instead, to new Roman centers established in the region. These events may be behind the origin of two important factors that articulated between themselves and were the consequence of the integration of the Northwest within Figure 6. Bouça da Mó interpreted plan, according to M. Martins

27 Potential core sites from the Iron Age were the Citânia de Sanfins, the Cividade de Âncora, the Cividade de Terroso, Cale, the Citânia de Stª Luzia, the Citânia de Briteiros, Castro de S. Lourenço, Castro de Faria, Castro de Roriz, Castro das Eiras, Monte Mozinho, Castro de Alvarelhos, Cividade de Bagunte, Castro de Vandoma, Castro S. Silvestre de Cardielos, the Citânia de S. Julião, Castro de Moldes/Castelo de Neiva, Castro de Roques, the Cividade de Carmona, Castro do Cossourado, or the Castro de Stº Ovídio (Ponte Lima). To the list above C. Alberto Brochado de Almeida added Stº Estevão da Facha, Santa Maria da Madalena (P. Lima), S. Lourenço da Armada settlements, the Cividade de Romarigães, Cristelo, Coto da Pena and the Forte Lobelhe (Almeida 2003b: 162-165). On the other hand, Luís Fontes suggested the existence of a central site at the Castro de Vieira, Cantelães, in Vieira do Minho (Fontes and Roriz 2007).

3. Central sites and new settlements

The analysis of the civitates capitals located in the EntreDouro-e-Minho region required an assessment of the castella. These were sites that earned a central position in Roman times. 69

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

the Roman world, mainly after Hispania was pacified. This integration inevitably demanded a new political and administrative framework, while at the same time catered for the development of conditions needed to establish a new social and economical order.

man phenomenon, which leads to the conclusion that the populi represented, in some way, the ethnic-political preRoman structure that was likely to have been respected when civitates were defined.

One of the factors we believe is paramount in explaining the process related with the loss of relevance of some of previous oppida is necessarily linked with the movement of the pre-Roman elites to Bracara Augusta, where they occupied administrative and religious positions that the city, as a civitas capital and a juridical conventus, required. The second factor has been linked with the emerging of new centralities established and structured, in the majority of cases, around the existing road itineraries. These would be built from Augustan times onwards and crossed the territory allowing the dissemination of new models of space occupation under the Roman influence.

What we can state with some degree of certainty is that these territorial units allowing Rome to integrate the indigenous communities were of a considerable size, had an administrative capital, a self-government structure and that the indigenous elites were in charge. This organizational model was extremely efficient and applied to several regions within the Western Roman Empire (Drinkwater 1983; Millett 1990; Provost 1993) However it raises some chronological problems and questions related with the archaeological acknowledgement of not only their boundaries but also of the location of their headquarters. With regards to the chronology, the problem is centered around the possibilities of the peninsular northwest civitates system was developed by Augusto, if we consider that they were referred in the ‘Edicto del Bierzo’ of 15 BC (Grau Lobo and Hoyas 2011; Sastre Prats 2001; 2001:24), or were in fact a Flavian creation (Alarcão 1995-96), if we consider the capita viarum character of certain civitates mentioned in Milestones created after the Flavian dynasty. Although it is not consensual, we believe the reference to civitates occurring in the Augustan period clearly suggests that the implantation of this model was part of a package detaining political directives applied by the emperor within the Hispanic northwest.

It is widely believed that only rarely and for a limited amount of time, some castella were able to keep their functions of central sites after Augustan tiames, if we consider the new territorial planning structure developed by Rome and fully consolidated by in the Flavian period. We believe it was possible that these trends initially occurred in the surrounding areas or in farther areas from the new political center represented by the city of Bracara Augusta. It also seems possible that some castella might have performed a key role within the new economical order, since its location was favorable to the Roman road control or to exploitation of specific resources, namely in terms of mining activities.

The exercise conducted by some researchers aiming to define the referred populi in both written and epigraphic sources and to identify key castella functioning as capitals inevitable led to the valuing of the Iron Age central sites. However, the challenge remains as the geography of both the populi and civitates is far from being consensual between the different authors approaching the subject (Tranoy 1981; Silva 1986; Alarcão 1992; 1997-98; Rodriguez Colmenero 1996). We must add the fact that very rarely this exercise took into consideration the settlement occupation chronology and mainly valued the strategic position held by sites in a pre-Roman context. If it is certain that the establishment of the abandonment chronology for the large ‘citânias’ presented several problems due to the absence of excavations, it seems inevitable to think that the fixation of a civitas capital in an indigenous settlement network required the existing sites long persistence throughout time, an intense Romanization process on location and the existence of compatible equipment able to perform and function accordingly.

The issues related with the loss of relevance of central sites from the Iron Age are directly related with the establishment of civitates. These were territorial and administrative units within the conventus territory. Furthermore, the challenges related with the establishment and the identification of civitates within the Peninsular Northwest framework has been approached on chapter 2 of the present work, and they are not very different from the challenges concerning the more restrict conventus bracarensis. Hence, and despite the numerous available literatures on the subject, a paramount problem still needs to be resolved and it concerns the knowledge over the civitates organization and if it did reflect in any way the preRoman settlement structure (Tranoy 1981; Silva 1986; Alarcão 1992). Or, on the contrary it only represents the result of a political Roman action that profoundly impacted upon the territory and its new social order and, in fact, wasn’t related with the previously existing ethnic groups (Sastre Prats 1998; 2001).

If the political geography of the civitates list inventoried by Pliny the Elder, concluded by Ptolemy and epigraphically referenced in the ‘Padrão dos Povos’ of Chaves is difficult to accept, the subsequent task of identifying the location of the respective capitals also is.

This actually performs a crucial question that still needs to be resolved, when we consider the challenges inherent to the literary and epigraphic sources that perform the fundamental database to frame the problem. Moreover, such sources are also of Roman origin, so they are confused and lack precision in their references regardless of them being from the populi or from the civitates. Some authors considered that both realities were clearly a Ro-

Jorge Alarcão tested several hypothesis related with understanding the geographical implementation of civitates. One of the attempts originated from the belief that these corresponded to pre-Roman populi and that Augusto elevated these central sites to headquarters in a transition period, when the emperor provided the elites with power 70

An Overview on the Entre-Douro-E-Minho Roman Settlement Patterns

(Alarcão 1992). This belief led the author to list Gallician warriors’ statues in an attempt to represent the leaders while attempting to ackowledge important local hierarchies, from which the Roman power would have sustained a network of civitates. Inherent to this argument are also a set of problems related with the number of known statues (Silva 1986:305-309) that doesn’t seem to match the number of known civitates and is not geographically adjusted with the the dispersion of these entities, as it has been also referred by the referenced author.

tained and reinforced its centrality throughout Roman times. If we list the remaining populi/civitates existing in the region, we verify that the problem regarding the territorial fixation of certain populi mentioned in the classic sources persisted, as well as the difficulty in identifying the eventual capitals (Alarcão 1995_96; 1998; 2000a; 2004). The cross-reference of several sources and methodologies still is the more detailed mechanism to advance research in a field that, due to several challenges, is rather complex.

The regressive method focused in utilizing Suevi parishes as platforms from which we could identify the existing Roman settlements functioning as civitates capitals provided casual however suggestive results for the region under study (Alarcão 2000a).28

Under these circumstances, and in order to move our research forward we must consider the probability of potential civitates capitals existing without attempting to confirm the political geography of their populi. We believe that it is possible the Vila Mou vicus (ID 16093705), located on the northern margin of Lima river, performed that role. This was a proto-historic settlement of low altitude that became an important Roman center and a Suevi-Visigothic central site, as suggested by Jorge Alarcão and opposing the hypotheses, by Almeida Fernandes, that Ovinia was in fact located in the Citânia de Santa Luzia (Fernandes 1997; Alarcão 2000a).

Moving from the Suevi parishes to the setlements that originated them, we can consider as civitates capitals Caldas de Vizela, a location that firstly was a vicus and then a parish headquarters under the name of Oculis Calidarum (Costa 1959; Alarcão 1999a) and Tongobriga, a widely acknowledge urban center equally referred in the ‘Paroquial Suévico’ as Tongobria (Dias 1997; Alarcão 1999a; 2000a). In some cases this exercise didn’t entitle the identification of parishes functioning as civitates, and instead suggested the existence of Roman sites. This would be the case of Meinedo, in Lousada referred to as a diocese headquarters under the Magnetum toponym and performing a site admitted to have been a vicus (Dias 1997:302; Alarcão 2000a:42).

This theory also suggested that several other potential civitates headquarters corresponded to sites of Roman origin, which at a later stage regained a certain degree of importance and became Suevi parish’s headquarters. It is interesting to consider the importance that the new Roman settlements had as potential civitas headquarters, at least on the western side of the conventus bracarensis and in the region more intensely populated and hierarchized in the Iron Age period.

Although casual, the observation allowing the understanding that the important Suevi parish headquarters may provide, at times, the identification of important Roman habitats functioning as civitates capitals should be considered. The identification of populi/civitates seemed to be consensual in the identification of Callaeci in the Porto region. Cale, their capital, was a site identified as o the location of the city cathedral and a place that would have been a castellum. However, conducted excavations at several city sites suggested a significant Roman occupation in the surroundings of the mentioned castellum, alongside the existence of port infrastructures nearby the Douro river, which suggested that Cale might have been a rather important urban center (Silva 2000:100-101). The role of Bracara Augusta as the Bracari capital has been widely accepted by several authors. However this is a case when we are presented with an ex novo foundation ensuring the capital function of a populus or of a group of populi 29 and not a proto-historic settlement that main-

New data emerging from excavation works conducted within the last decade in Roman sites believed to have been vici have been shifting the knowledge related with these sites. They are more abundant than previously believed. This new framework of knowledge may contribute towards the relocation of some classic challenges related with the overall Romanization process, from which the issue related with the populi/civitates and their headquarters is also framed. Hence, we believe that for the region under analysis, the civitates status was provided to new Roman setllements established within the region, which performed a more suitable administrative framework to have in place by the Romans. The possibility for vici to perform as administrative capitals has been, for a long time, a controversial subject and implied the discussion regarding the civitates capitals’ status. For some authors, the organic and rural growth inherent to the vici was inexorably incompatible with the political and ideological planning of the foundation and growth of civitates in the western part of the Empire (Jones 1987:54; Whittaker 1990). ‘Secondary towns’ would have functioned, throughout Roman times, as a type of settlement that simultaneously encompassed features from the main Roman cities and from the rural indigenous habitats. They

28 The great effort behind the identification of the referred parishes and pagi in the Parochiale Suévico (work from the sixth century – circa 572 and 582) documenting the dioceses of the Suevi kingdom and inside these the parishes and the pagi was produced by Almeida Fernandes in several articles that were gathered in 1997. 29 With regards to the problematic of the Bracari and Bracaraugustanus, we will not forward the problem to the extensive literature produced on the subject. We will only reference Alarcão 2003.

71

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

would distinguish themselves from civitates capitals by their production and market functions.

The case of the Santa Comba Hill-fort (ID 3041402) in Terras de Basto, exemplified this second possibility.

Thus, the economic functions would separate them from civitates capital cities, where the political function was predominant (Mangin 1983; Jones 1987). Whittaker synthetized rather well this belief:

Starting from a hypothesis put forward by Jorge Alarcão regarding the possibility that around this Hill-fort might have existed a civitas capital, we researched the area (cfr.; supra, chapter 1, paragrah 2.3) and came up with the conclusion that it was from this center that a vicus was later established and that it possibly included civitas capital functions.

‘Villas were a phenomena of a central civitas town, not of the so-called small towns. That (...) must have been the decisive factor dictating the difference between them, despite physical urban similarities between civitas centres and small towns. The civitas town was the social and political center of the state. The vici were firmlly rural and similar to Villas’ (1990: 116).

In other cases identified in the region, from which Alvarelhos, Cividade de Biscaia or Vila Mou vici are examples, the new Roman habitats were established from proto-historic settlements that extended themselves towards the slopes and surrounding areas. However, in Santa Comba, we can be facing a different circumstance in relation with the Tongobriga example, or the ones in Caldas de Vizela and in Vila Mou. We can admit however, that all these four cases might have functioned as civitates capitals.

Hence, the problem is firstly placed in the epigraphic references to civitates capitals with a vicus status. The most popular case referred to Sens, the main urban center of the Senones, where an inscription corroborating its vicus status has been found. Other than the referred case, there were several examples of vici, civitates capitals that would be placed within the ‘’vici qui proprios fines havent’’, as opposed to vici that were included in ‘fines’ of a ‘maior civitas’ and described in the ‘Código Justiniano’ (Tarpin 2002: 262).

The Caldas de Vizela site was a Roman open settlement probably founded due to its thermal and potential transport supporting functions. Tongobriga corresponded to an open urban site where the remains from the Iron Age don’t support the assumption of it having been a preRoman Hill-fort. Vila Mou was a low altitude site without any centrality when compared with the wide ‘citânias’ around, namely the Citânia de Santa Luzia. The size of the Santa Comba settlement together with the finding of a Gallician warrior statue led us to consider that the civitas status on this particular location resulted of a previous existing Hill-fort.

The systematic study of this phenomenon should encompass a multipolar overview, namely of a vici network without having them performing an administrative role. If we should admit that vici may have encompassed, in the region under analysis, an administrative capital status, we should also consider the possibility of vici existing without being capitals. It seemed important to reference the following statement from M. Tarpin:

We will now enlist the sites that may be classified under the vici typology (Table 5), and briefly describe them afterwards. In some cases, we will provide references to chapter 4 and 5, where we will analyze them in further detail (Map 8).

‘Il faut encore préciser qu’il peut y avoir au sein de la même cite un vicus chef-lieu et un vicus, voire de vici, sans statut particulier..(…) Ceci implies that vicus ne définit pas le chef-lieu et que, réciproquement, le fait d’être chef-lieu ne donne pas automatiquement un statut plus gratifiant que celui de vicus (2002: 263).

What we referred should be carefully considered and include the articulation of the indigenous power under Roman domain in the region under analyses and understanding that it could have originated, at least, two potential case scenarios presenting different geographic realities. In the western part of the region some of the vici may have had administrative capital power. In the inner and mountainous regions, away from Bracara Augusta, in areas displaying a scarce indigenous settlement pattern, some Hill-forts may have had a predominant position and functioned as civitates capitals.

72

An Overview on the Entre-Douro-E-Minho Roman Settlement Patterns

ID

Site

Municipality

Site Type

3010503 3010601 3011501

Cidade de Biscaia Caldelas Vilela / Mojeje

Amares Amares Amares

Road vicus, mansio Thermal vicus Road vicus / mutatio?

3085201

Caldas de Vizela

Guimarães

Thermal and road vicus

3080801

Caldas das Taipas

Guimarães

Thermal and road vicus

3134202

Prado

Vila Verde

Road vicus

3123602 3041402 3100602

Perrelos Santa Comba Chãs de Vilar

Vila Nova de Famalicão Cabeceiras de Basto Terras de Bouro

vicus? vicus Road vicus

3111403

São Cristovão

Vieira do Minho

vicus?

Amarante

Road vicus?

13011601 Gatão

13011902 Lomba Amarante 13021502 Quinta de Guimarães Baião

vicus? vicus?

13051301 Meinedo Lousada 13072301 Caldas de Canavezes Marco de Canavezes

vicus? Thermal and road vicus

13072801 Várzea do Douro

Port vicus

Marco de Canaveses

13112601 S. Vicente do Pinhei- Penafiel ro 13140301 Alvarelhos Santo Tirso 16073501 Ponte de Lima Viana do Castelo

Thermal vicus

16100801 Forte de Lobelhe 16093705 Cividade de Vila Mou 16061204 Cabeço do Leijó

Vila Nova de Cerveira Viana do Castelo

Port vicus Mining vicus

Ponte da Barca

Road vicus?

Road vicus Road vicus

Observations Mile X of the Via XVIII itinerary Inscriptions dedicates to the Nymphs Inscriptions dedicated to Genio Viriocelense; Via XVIII itinerary Inscriptions dedicated to Jupiter and Bormanico, mentioning of the legacy from the emperor; ‘sede’ of the Suevi parish Oculis Calidarum Honorary inscription dedicated to Trajan from the year 103 AD; located nearby the itinerary between Bracara and Emerita Augusta Via XIX itinerary. Important functions linked with clay exploitation Metallurgical activities Civitas capital? Milestone signalling mile XVII from the Via XVIII itinerary. Should correspond to Salaniana Wide habitat that may have functioned as a mansio from the Via XVII itinerary Currently missing inscription mentioning Vicus Atucausis Perhaps the parish headquarters of ‘Melga’? Suevi parish headquarters: Magnetum Located nearby the itinerary Bracara – Tongobriga Votive and funerary inscriptions. Bracara – Emerita itinerary

Via XVI itinerary Should correspond to Limia, mansio from the Via XIX itinerary located at XVIII miles from Bracara Potential fluvial port of the Minho river Inscriptions and a wide mining exploitation area Wide habitat associated with a Hercules inscription

Table 5. Entre-Douro-e-Minho vici30

30 We will not include in this list the Croca site, in Penafiel (ID 13110801), a site associated with a necropolis dominating the Croca river valley and crossed by a Roman road coming from Meinedo. Lino Tavares Dias was hesitant in classifying the site as a vicus (Dias 1997: 315-316). For different reasons linked with the nature of available data, we will not include the site on this group of potential vici located on the valley and coastline areas. This challenge will be further analyzed on chapter 5.

73

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis Bracara – Tongobriga – Emerita Augusta enabled the assumption that a settlement with thermal functions alongside duties including road support would have existed on site. The importance of the site is also corroborated by significant epigraphic documentation including a dedication to Iupiter Optimus Maximus and two other offered to a indigenous deity named Bormanicus. The presence of a wide public building structure is documented by a lintel of almost 3 meters length, where the following inscription can be read: Dedicavit Titus Flavivs Claudianus Archelaus, Leg. Aug. Other than these findings, we also highlight the presence of tiles, varied ceramic materials and scattered remains from bathing infrastructures. The settlement persisted throughout the Middle Ages as a Suevi parish headquarters unit under the name of Oculis, which seemed to have evolved into Oculis Calidarum as referred in a document dating back to the year 1014.

Cividade de Biscaia, Castro de Caires / Freixeiro, Amares The Cividade de Biscaia vicus occupies a wide area including the lower slopes of Castro de Caires or Gróvios and the Freixeiro site (cfr. Cap. 5).

Caldelas, Amares A potential thermal vicus located on the currently existing thermal site. Several Roman remains have been found on this location, from which two votive arae dedicated to Ninphas should be referenced. A necropolis has also been identified nearby (cfr. Cap. 5).

Mojeje / Vilela (Viriocelum), Amares A potential vicus located in the surrounding of the Via XVIII itinerary. The habitat could have been extended towards the Mojeje area located around a Roman road. (cfr. Cap. 5).

Caldas das Taipas, Guimarães The site is located near the road linking Bracara Augusta – Emerita Augusta via Tongobriga. An important honorary inscription dedicated to to Trajan emperor (CIL II 4796) and engraved on a rock dating back to the year 103 has been recovered from this site. The existence of salutiferous waters may be behind the development of the thermal infrastructure, from which few material remains still exist. It is beliebed that this building was overlapped by recent baths building structures. Other than the site being located on the Bracara – Tongobriga – Emerita road, it was also close to the Ave River that would be navigable to the vicinity of this site.

Santa Comba, Cabeceiras de Basto The settlement of Santa Comba may have originated a Roman site occupying an indigenous Hill-fort and located on the western part of the hill slope. It has been precisely on this slope that the ‘Casa de Santa Comba’ was founded and its landlords gathered assets uncovered on location including a Gallician warrior statue that became famous due to its excellent conservation state and by an inscription reading ‘’Artifices Calubrigenses E(x)s Albinis F(aciendum) C(uraverunt)’’ (Silva 1986). At the ‘Casa de Santa Comba’ there was also an honorary inscription dedicated to Gordian emperor found at the house foundations (CIL II 2382) and another honorary monument dedicated to Antoninus Pius (CIL II 2381) found at the Santa Senhorinha de Basto site (Alarcão 1988: 1/272:15). From the same location is also a group of Milestones and Roman architectonic elements that are believed to have been part of a Roman building entablature. Research conducted on site led us to believe that this performed the most important Roman settlement in the region and it could have been extended towards the slope lands. Furthermore, the inscription dedicated to Constantius Chlorus and later deposited in Abadim (Alarcão 1988:1/268:15) corresponded to a Milestone and not an to a honorary epigraphic element.

Prado, Vila Verde The existence of a vicus in the Prado area is suggested by the findings of tombstones and by the presence of abundant ceramic materials around several areas of the current site and in the church area. These remains should be articulated with the importance of the site, not only as a road station of the Via XIX itinerary, but also due to its location on an important ceramics production center. Other than functioning as an extracting clay site destined for the production of pottery in Bracara Augusta, it may also have functioned as a tiles production center (cfr. Cap. 4).

Caldas de Vizela (Oculis), Guimarães

Chãs de Vilar (Salaniana), Terras de Bouro

In the current Vizela area, the existence of a wide and important group of Roman remains suggested the presence of a vicus. The existence of thermal waters and the presence of an itinerary linking

This performed a wide Roman settlement that seemed to have occupied several artificial platforms. It is possible that it corresponded to the Salaniana mansio refered to in the Antonine Itinerary (cfr. Cap. 5). 74

An Overview on the Entre-Douro-E-Minho Roman Settlement Patterns

Meinedo (Magnetum), Lousada S. Cristovão, Vieira do Minho

The site is located near the road connecting Bracara – Emerita Augusta via Tongobriga. We highlight the presence of abundant Roman ceramic material on Quinta dos Padrões and on the football field currently existing in the surrounding area. To this remains we add the fact that the site may have been a Suevi parish under the name of Magnetum.

The site occupies a flattened area near Ruivães right next to the XVII itinerary. It is still possible to observe today, across a wide area, abundant material remains, as construction stones, and a wide quantity of Roman construction ceramics alongside with common pottery fragments. The location of the site near the road together with its features allowed the assumption that it was a mansio. The site also witnessed a medieval occupation.

Caldas de Canaveses, Marco de Canaveses Near by Canaveses bridge, on the road started in Bracara Augusta and going through Tongobriga up to Emerita Augusta, a thermal building was found exhibiting walls covered in Roman mosaics and with an abundant set of assets, encompassing coins, pipelines, construction ceramics and an inscription destroyed by building works carried out in 1743. In the surrounding area a necropolis was found. This was certainly a vicus housing functions that were simultaneously thermal and transport related.

Perrelos, Vila Nova de Famalicão The site is located in the lower area of the S. Miguel-o-Anjo site, in the Delães municipality. Excavations conducted on site and still in progress identified houses and a thermal building dating back to the third/fourth centuries. This particular building was subjected to a restructuring phase in the late fifth and early sixth centuries in order to turn it into a small artisanal structure. Although the excavations are in progress and the extension of the settlement is not yet fully known alongside its exact foundational chronology, there is a strong possibility that the site may have functioned as a vicus (Silva et alii 2005).

Várzea do Douro, Marco de Canaveses The material remains present on this site are scattered around a wide area including several platforms of a hill located on a tigh curve of the Douro River margins. On the highest platform, registering a maximum level of 114 meters of altitude, houses of circular plant can be found together with Roman coins. One coin dates back to the 2 BC / 14 AD and four others from 117 / 192 AD. Lino Tavares Dias (1995:267-270) conducted several archaeological interventions on the site and found Roman walls, column bases and capitols alongside abundant ceramic materials, from which sigillata Clara C should be highlighted. On another intervention, the researcher exhumed a circular house associated with Roman ceramics from which he identified sigillata Clara C. There were eight known funerary and votive inscriptions on site (Dias 1995; 1997). Várzea do Douro ended the road coming from Tongobriga and Monte Mózinho. It is widely believed that it was very likely a vicus functioning as a fluvial port allowing the continuation of traffic across the Paiva valley after the river crossing.

Gatão / Quinta dos Pascoais (Atucausis), Amarante The area where ‘Quinta dos Pascoais’ is located would have included the Atucausis vicus, if we consider an inscription found at the site and dedicated to Jupiter by the vicani Atucausenses (CIL II 6287). On site, plain ceramic fragments and another inscription mentioned as coming from S. João de Gatão, currently missing, have been recovered (CIL II 2383).

Quinta de Guimarães (Melga), Sta. Marinha do Zêzere, Baião The site is located on the southeast region of Tongobriga and its interpretation as a vicus is related with the presence of an necropolis at the Quinta de Guimarães and of material remains scattered across the slope area of the same site. Furthermore, Almeida Fernandes (1968:80) located the headquarters of the Melga parish in Sta. Marinha do Zêzere. The fact that the Melga toponym still exists in this area together with the presence of remains, led Lino Tavares Dias (1997) to suggest the possibility over the existence of a vicus, to which a Suevi parish headquarters succeeded.

S. Vicente do Pinheiro, Penafiel On this site a thermal building has been referenced to have existed and exploited the medicinal waters. Other than this building structure, a kiln to bake ceramic and a necropolis were also found and located at around 100 meters away from the bathing structures. The plan interpreted by J. Alarcão (Figura 16) had the following interpretation: ‘’the entrance was done on A and from there one would go to room B. One of those could have been an ‘apo75

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis dyterium’. Both opened to C, ‘frigidarium’ and there were two bathtubs in D and E. The heated rooms by ‘hipocaustum’ were H and I. Perhaps this room performed a ‘guarda-vento’ for both heated rooms. Both J and K areas were atriums for thermal services’’ (Alarcão 1988:28).

F(ilio)/ANTONIO/ ALMO. 31 Excavations conducted throughout the 1980s and 1990s identified a commercial area comprising an open square surrounded by a potential portico opening up to a tabernae and a thermal structure (Moreira 1992; VVAA 2006: 82-83). As it probably happened with the Castro de Caires in Amares case, the previous setllement might have been extended towards the slope and valley areas, in sites more suited to road supporting functions.

Ponte de Lima (Limia), Ponte de Lima Limia is located at 18 miles from Bracara Augusta on the axis of Via XIX itinerary. It was mentioned in the Antonine Itinerary and its location corresponds to the current Ponte de Lima area. There is the presence of Roman ceramics at several sites of the present village, namely in Sangarinhal, Sobral, Quinta da Graciosa and Avenida dos Plátanos. There is a wide density of remains near the areas where the road crossed, south after the present Village. The fact that they are nearby sites not continuos, both in Ponte de Lima and in the surrouding region led Carlos Alberto Brochado de Almeida (1996) to be prudent and refer to them one by one on the works he developed for the region. However, we believe it is possible to locate Limia here, and that the site would have functioned as a mansio of the Via XIX itinerary. The Bertiandos Milestone from Maximianus and Maximus indicated the XVIII mile starting in Bracara Augusta.

Figure 7. S. Vicente do Pinheiro thermal plan according to Jorge Alarcão and based upon works from Teresa Soeiro (Alarcão 1988 1/473, fig. 103; Soeiro 1984).

Alvarelhos, Trofa The Alvarelhos vicus originated from a hill-fort located on a slope with the same name and nearby the Via XVI itinerary. It is possible the site was a mansio located 23 miles away from Bracara Augusta and 12 miles from Cale. Its relevance as a road station supporting unit is linked with the fact that this was the area from where a secondary road leading towards to coastline would also have started (Mantas 1996:687). The vicus could have included the still valley area near the settlement location, namely the Quinta do Paiço where a silver ‘patera’ was found and boosting a dedication to Mars. The commemorator of this particular item was a free man identified as tria monina. From the Alvarelhos settlement are believed to also be three coin treasures, from which the most recent coin dates to the year 27 BC. Nine balls of silver engraved with the word CAESAR were also found together with a rectangular stone, where on one side the Roman numeral XII was inscribed and on the other the words CAESAR. To be highlighted is still the finding of a capitol and a fragment of column shaft. Nearby the settlement, a votive altar to Genius dedicated by Saturninus, son of Caturão, was also recovered alongside an inscription where one can read: MADE/QVIS(enses) STATV/ERVNT/ LADRO/NO CA/MALI/

Forte Lobelhe, Vila Nova de Cerveira The site was established on a fluvial terrace of the Minho river, particularly on a small hill with 46 meters of altitude. The first occupational stage took place on a proto-historic Hill-fort of low altitude surrounded by a slope and included houses of circular plan implanted on landings that seemed to have extended themselves towards the nearby river. According to the uncovered ceramic materials, this exact occupational moment took place in the mid first century BC. Throughout this century, buildings of rectangular plan substituted the old houses and from the same period should be a kiln that was constructed to produce common pottery (Almeida 1996:81). The excavations conducted meanwhile uncovered several material remains from which we highlight common Roman pottery, southern Gallic, Hispanic and Clara D sigillata, glasses, bronzes, coins and several amphorae fragments predominantly Haltern 70 and Dressel 7-11. One of the excavated structures is believed to have been an amphorae storage unit still presenting excavated holes destined for stowage. According to the interpreta31 The challenge associated with the interpretation of this inscription has been approached on chapter 1.

76

An Overview on the Entre-Douro-E-Minho Roman Settlement Patterns

tion of uncovered ceramic remains, the site was functional from the early first century onwards (Almeida 1996:117-118). Other than the referred materials, column bases and shafts and a fragment from a mosaic were also found alongside with varied pottery fragments, namely of dolia and common pottery as well as glass matirials. According to the data avaliable, the old Iron Age Hill-fort became a vicus settlement presenting a plan that cannot be rigorously defined and dated. The proximity of the site with the Minho River that was navigable up to the Tuy region and the importance of this fluvial route are of extreme importance in the assessment of this vicus role. The site occupation continued throughout the Suevi-Visigothic and medieval eras. Finally, the regional military and defensive requirements led to the construction of a fortress on the second half of the seventeenth century.

Cabeço de Leijó, Ponte da Barca The site is located on the northern slope of the Amarela mountainous range facing the Lima River. It was a settlement of wide dimensions occupied since the Roman times up to the high Middle Ages/‘Reconquest’ period. The excavations conducted on site32 brought to light wide houses of quadrangular plan. Artificial platforms, wall alignments and houses can still be seen. At the Lindoso castle (ID 16061201), a dedication to Hercules was uncovered and, in our opinion, it can be a statue pedestal. Other than this item, a stele representing a togged figure was also uncovered. It is possible that both monuments came from Cabeço de Leijó.

The group of sites identified as vici allowed us to assess their importance in the settlement network of the region as entities that generate new centralities and enabled us to understand more clearly the new landscape organization of the wide region under analysis.

Vila Mou, Ponte de Lima

The great majority of identified sites are directly articulated with the road itineraries, although that particular functionality didn’t perform the exclusive function of the inventoried vici.

The site occupies an elevated area and is located on a fluvial terrace, precisely on the right margin of the Lima River. The occupied area distributed itself across three platforms and presented remains of indigenous produced ceramics, tegulae, dolia, and amphorae fragments, manual grading wheels, imbrex, scoria and animal bones. The abundance of ceramic remains between the Cividade site and the valley area nearby the Lima River is the main reason behind the importance of this site. A necropolis has been found at Quinta da Torre (ID 16093704), located on the southern slope of Cividade de Vila Mou. In 1892, the demolition of the Vila Mou parish church (ID 16093703) uncovered an altar dedicated to Jupiter where the following can be read: RVFI. GRO/VIVS. VOTV/M. IOVI. OP/TIMO. M/AXIMO. With the altar, some capitols from the Visigothic times were also discovered together with an inscription to Victoriae (currently missing) and worked stones. Nearby the referreded church, Roman construction ceramics and granite column shafts have also been found. The mining exploitation has been testified to have taken place in Rasas (ID 16093701), a tin exploitation area located a few hundred miles from Monte da Cividade, where one lamp was also found (Rasas de Cima) and Mata de Vila Mou (ID 160937), a site where tin was also exploited. Located on the crossroad of some secondary Roman trajectories and around 9 km distance from the Via XIX iinerary, this habitat performed a important site within the Roman mining exploitation industry, both exploiting tin and gold (on the Seixo and Tinto rivers). Furthermore, it also held a privileged position on a key transport intersection, events that have led us to consider it a potential vicus.

77

The divison of vici in four categories (port, thermal, road stations and plurifunctional vici, where arts and crafts and trade industries were prevalent), as used for the Loire region by Michel Provost (1993:138) may help the overall understanding a first reading of these settlements. The vast majority however had, most definitely, more than one function when it comes to the region under analysis. The term ‘plurifunctional’ created by M. Provost to distinguish sites also focused on both trade and arts and crafts functions can easily be applied to the majority of vici in this region. To this plurifunctional character should be added a significant variation of vici sizes and morphology. Advances in archaeology studying this particular type of habitat, in particular in Gallia and Britannia, enabled the identification of a high morphological variety of this type of settlements (Pérez Losada 2001),33 a circumstance that very likely also occurred in region under study. With regards to rural vici,34 the information compiled by M. Tarpin (2002) revealed that this type of settlements 32 Excavations conducted by Luís Fontes to whom we thank the provided information. See also L. Fontes (2012). 33 F. Pérez Losada presented a morphology synthesis work of ‘aglomerados secundários’ based upon the works by Burnham and Wacher (1990) for Britannia; Brulet (1994) for Valonia-Luxemburg and Tassaux (1994) for Aquitania. The works from these authors shed some light on the considerable variety in domestic architecture found at these habitats (Pérez Losada 2001: 41, 43-48). 34 The concept of ‘rural Vicus’ enabled us to distinguish it from ‘urban vici’, a common practice in Rome and relate with city neighborhoods. The concept may also relate to ‘bâtiment urbain’, meaning a group of apartments that were different from insulae, referring instead to quarters. Furthermore, pagus, associated with vici, may be

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

was very common in Roman times, particularly in areas where the Roman state needed to closely control and monitor the territory in general and the road network system, in particular. In Hispania, the majority of epigraphic references to vici are precisely located in the western region, which according to M. Tarpin is related with the control issues of the mining exploitation regions. Moreover, the archaeological reality under analysis allowed us to add new data to the ‘’epigraphic’’ perspective. The variety of this type of settlements included vici with functions related with mining exploitation, vici linked with managing the salutiferous features of the thermal water sources and others vici with port functions. The control over the road network system seemed to be paramount in the implantation of these habitats. However, they seemed to have had a key role in the territorial planning and in the overall Romanization of this land. In fact, several of these vici may have had administrative functions, which ensured a rather peculiar process of establishment and territorial cohesion.

‘Est-ce (...) qu’il n’y a pas de trace des cultures pré-romaines dans les uici et les pagi? Certes non; bien des pagi (...) correspondent au territoire occupé par une communauté avant l’organisation romaine, et les uici, bien souvent, ont une population d’origine indigène. Mais ni les uici ni les pagi n’ont pour function d’assurer une continuité culturelle au nom d’un quelconque «droit à la différence», parfaitement anachronique. Leur rôle, bien au contraire, est de formaliser la possession du sol et l’intégration des individus dans un ensemble administratif et culturel fondé sur la suprématie de Rome’ (Tarpin 2002: 245).

4. Villas: identification challenges

Despite of more than 60 references to potential rural infrastructures that can be integrated in the villa category, the vast majority of remains reports back to surface findings and are very rare the cases of villas that actually have been excavated in the region under study.

Lastly, it is important to highlight that the majority of these setllements had a ex novo foundation, which in several cases was the end result of a displacement of preexisting communities towards slopes and valley areas, in a process that is rather similar to what was observed by F. Pérez Losada for the Galicia region (2001). One must confront this statements with the controversy related with the indigenous origin of these sites theory.

There is a knowledge gap with regards to a key part of the Roman occupation in the region and articulated with the exploitation of the rural landscape. Moreover, the key issue refers to knowing the conditions we presently have to produce a believable synthesis related with the patterns and with the evolution of Villas, while developing on the same scale of work a wider idea about the rural Roman settlement network and its geographical distribution.

The indigenous character of vici, particularly the ones located in the Gallia region, corresponded to the idea that these sites may represent the survival of the pre-Roman settlements. The cultural continuity was, for example, suggested by other advocates of the indigenous theory, in the mentioning of Cesar when describing the Gallia habitat making references to oppida, vici and aedificia units. Articulated in the geographic distribution of the vici mainly present in Gallia and Germania, the vici phenomenon has been seen as an adaptation of the Celtic roots of these provinces (Picard 1986; Buchsenschutz 1994; Metzier 1994).

Our study needed to take into consideration available data. We will now enlist all Villas combining the elements defined on chapter 1, sites that include the findings of Roman ceramics - tegulae, common pottery and/or other and are associated with elements such as topographic position, toponym or type and significant quantities of ceramic materials that will be assessed on chapters 4 and 5.

The question of the indigenous character seemed to be surpassed by more recent works on the subject. We were helped by the words of F. Pérez Losada, when stating that: “este “indixenismo” característico dos vici en absolute implica unha obrigatoria orixe prerRomana do hábitat dado que, de feito, a meirande parte deles son implantacións ex novo xurdidas em plena época Romana” (Pérez Losada 2001: 32). It is not necessary to justify a relatively obvious phenomenon – the fact that the vici inhabitants were in their vast majority of indigenous origin – based upon a cultural continuity. This idea was efficiently synthetized and widened in an incisive way by Michel Tarpin: defined as a territorial circumscription, topographically defined and holder of an important census status. On the subject, one must consult the work by M. Tarpin (2002).

78

The present inventory diminished considerably the total of sites identified in this category. A rapid analysis of the table and of the produced cartogram revealed some indications (Map 9).

An Overview on the Entre-Douro-E-Minho Roman Settlement Patterns

ID 3020104

Paço

Site

Municipality Barcelos

3026501 3028301 3028602

Assento Assento Paço-Vila Cova

Barcelos Barcelos Barcelos

3030103

Adaúfe-Igreja

Braga

3031001 3031201 3031301 3031901 3032301 3032904 3033101

Dume-Igreja Espinho-Igreja Esporões-Igreja Gualtar Lomar Pinhel-Pateira Outeiro de S. Sebastião

Braga Braga Braga Braga Braga Braga Braga

3034603 3035301 3060201

Coturela-Parretas Santarão Vila Mendo

Braga Braga P. Varzim

3061301 3086402

Linhariça Abação

Esposende Guimarães

3091501 3110501 3121001 13032401 13071202

Via Cova Igreja Velha, Cantelães Castelões Villa de Sendim Urrô, Casa da Babilónia

P. Lanhoso Vieira do Minho V. N. Famalicão Felgueiras M. Canavezes

13072702 13101701

Casa do Outeiro Parada Todeia

M. Canavezes Paredes

13111701 13131001 13140303 13161102

Outeiro Martim Vaz Sobre Sá (Alvarelhos) Paiço

Penafiel P. Varzim Santo Tirso Vila do Conde

13162801 13160103 16100404

Vila do Conde Caxinas Stª Luzia, Covas

Vila do Conde Vila do Conde V. N. Cerveira

16070701 16071605

Quinta do Paço, Beiral do Ponte de Lima Lima Paço/Travasselas Ponte de Lima

16071818

Prazil, Facha

Ponte de Lima

13080302

Lavra

Matosinhos

16050901

Nª Srª do Livramento

P. de Coura

16050801

Moimenta, Formariz

P. de Coura

16071804 16073204

Quinta do Paço Velho Boudilhão

Ponte de Lima Ponte de Lima

Features Funerary inscription; walls Architectonic elements Architectonic elements Excavated Villa; necropolis Architectonic elements; votive an funerary inscriptions Excavated Villa Funerary inscription Votive and property inscriptions Architectonic elements; 2 funerary inscriptions 2 funerary inscriptions Architectonic elements Architectonic elements; coin treasure Architectonic elements; walls Architectonic elements; walls; coins Architectonic elements; walls; necropolis Architectonic elements; walls Architectonic elements; funerary inscription Necropolis Excavated villa- pars rustica Excavated villa Columns; architectonic elements Excavated villa Common ceramics; architectonic elements grading wheels; tegulae Structures; architectonic elements Structures; tiles necropolis Architectonic elements Excavated Villa Inscription; architectonic elements; figurative bronzes Structures; columbarium; tiles coin treasure Tiles; necropolis Excavated structures; Villa associated with a necropolis Common ceramics; indigenous ceramics; dolia; amphorae; ‘cushioned’ stones Hispanic sigillata, tegulae Building of unknown size and plant; sigillata; common potery; incineration necropolis Tuscan capitol; construction ceramics worked stones Roman common pottery; construction ceramics; indigenous ceramics; worked stones; loom weight Construction and domestic ceramics; marble sculpture fragments; column bases; tiles Roman common ceramics; indigenous ceramics; Columns; granding wheels; walls: sigillata; tegulae Roman construction ceramics; coins; necropolis Excavated Villa Circular houses; architectonic elements; ceramic kiln

79

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

16071605 16090104 16093002

Paços/Travasselas Baganheiras Igreja (Geraz do Lima)

Ponte de Lima V. Castelo V. Castelo

Architectonic elements; walls Structures; architectonic elements; coin treasure Architectonic elements; Roman necropolis; circular houses

Table 6. Entre-Douro-e-Minho Villas

Obs: the excavated villas are highlighted

The distribution of findings signaled two preferential locations of villas, without however telling us if it corresponded to a pattern: a group of villas located near the coastline and another group concentrated on the open valley areas. The region corresponding to the mountainous range separating Entre-Douro-e-Minho and Trás-osMontes wasn’t part of the distribution pattern. Moreover, the remains are scarce and prevent us from establishing from start a hierarchy pattern focused on size and chronologies.

ical Unit of the Minho University in 1980 allowed the identification of buildings remains associated with a necropolis. The construction materials emerging from what it is believed to be the main building are very poor: walls, basic foundation remains and ground pavements. The oldest materials referred to indigenous ceramics and Haltern 70 amphorae fragments, which indicated an occupational stage occurring between the late first century BC and early first century AD. The chronology of the findings discovered in a well located next to the building and in the adjacent necropolis allowed to estimate an occupation between late first century BC and the fourth century.35

Nevertheless, we can state that coastline villas were focused on the exploitation of maritime resources that were present on the tools linked with preparing fish products and seemed to spatially distribute themselves in more or less regular intervals. Villas located in open valley areas were normally located on gentle slopes, detained a positive sun exposure, dominated fertile agricultural lands and were established nearby watercourses. Their concentration between the Lima and Ave rivers is part of the highly populated general pattern this particular area had since Proto-History. In Roman times, the city of Bracara Augusta exerted a significant attraction role, which likely continued this area development.

Via Cova, Póvoa de Lanhoso36

The following villas have been excavated.

Dume, Braga A wide Roman villa was identified on this location, and some walls from the residential areas and a thermal building are well known. In Luís Fontes’ opinion, the uncovered funerary epigraphs should be associated with the Carquemije necropolis, believed to have been the villa necropolis (Fontes 1987; Fontes 1991-1992) (cfr. Cap. 4).

Alto de Martim Vaz, Póvoa de Varzim A villa with a plant that cannot be fully recovered revealed infrastructures linked with the production of fish. It is believed the site met two occupational stages, one taken place in late first century AD and the other one at a later stage (cfr. Cap. 5).

The Via Cova Roman villa is located in Cimo da Vila on a small valley nearby the Pregal watercourse. Its natural boundaries are easily identifiable: on the western part by the Carvalho mountainous foothills, on the eastern and northern by the elevations corresponding to the S. Pedro and Castro de Lanhoso mountains and on the southern by the Pregal watercourse. Excavations conducted in 1990 accompanied works destined to construct a house. The excavated area allowed the identification of part of a building with 6 meters wide and surrounded by two walls. The part of the exhumed building was compartmented and included six divisions alongside an open and paved area corresponding to a patio. Part of the identified construction boosted an NW/SE orientation and was extended towards NE on one of the walls founded in granitic alteration. These circumstances favored the preservation of the walls that at some point reached 1,30 meters high. The excavations reports stated that due to the nature of these walls and other found materials, this structure was remodeled from a different and previous construction presenting a morphology that is impossible to reconstruct. This report concluded the features of the compartments and the findings, and enabled us to acknowledge a construction with a multifunctional character meant to store food and other agricultural instruments. Amongst other remains, the bases of 35 Excavations conducted by F. Sande Lemos and M. Delgado, to whom we thank the access to the reports and to the uncovered ceramic material deposited at the D. Diogo de Sousa Museum. 36 Excavations conducted by Manuela Martins, to whom we thank

Caxinas, Vila do Conde The group of structures identified in Caxinas during the excavations developed by The Archaeolog-

the access to the produced documentation.

80

An Overview on the Entre-Douro-E-Minho Roman Settlement Patterns

the wide dolia and a stone basin were found, other than a grading wheel and an excellent group of iron instruments of clear agricultural nature. We concluded that the excavated area matched the pars rustica of a villa dated back to the first century, which suffered a reconstruction process throughout the first half of the fourth century AD.37 There seemed to be a permanent occupation of this site up to the sixth century, if we consider the grey Paleo-Christian ceramic fragments chronology. In 1999, approximately 100 meters from the site were identified alongside walls and architectonic elements, namely a Corinthian capital taken from levels of great depth, which allowed us to place on this exact location the villa pars urbana, reinforcing the interpretation of the site discovered in 1990 as being the rustic part of the villa. Confirming the assumptions made regarding the second place, we will be facing a typical Roman villa with pars rustica and pars urbana, a site benefiting from an excellent location both in terms of environmental resources and in its proximity with the Via XVII itinerary linking to Bracara Augusta.

cated a first occupation stage occurring between late first and early first century AD, have taken place in the building or in the surrounding areas (Almeida 1996). The discovered ceramic material is varied and in great quantity: common Roman ceramic, dolia, amphorae, bracarenses, sigillata imitations, loom weights, Paleo-Christian pottery and a belt plaque with an inscription dating back to the fifth century. Throughout its wide chronological period, the villa was reconstructed, at least in part of its infrastructure. A necropolis located in Bouça da Gateira has been associated with this Villa, where a few dozens of graves have been discovered and present varied typologies. They range from tombstones with triangular sections and made of tegulae and imbrex to coffins either made of stone, wood or even tombs excavated on rock lacking any type of artifacts.

Igreja Velha-Cantelães, Vieira do Minho The so-called “old church of Cantelães is an archaeological site subjected to three archaeological campaigns, from 2010 to 2012. These interventions have revealed an impressive roman building from the 1st century AD, in some way displaced in this remote region of the Conventus Bracarensis. A well-documented fire has entirely destroyed the building by the end of the 1st century AD, which was permanently abandoned and deeply dismantled afterwards.

Paço-Vila Cova, Barcelos The site presented a great quantity of indigenous ceramic and structures dated from late first century BC and early first century AD. The first building endured recontruction stages at the High Empire period and witnessed considerable transformations at the fourth century, at a time when a thermal building was also constructed. The site included also a Visigothic building with much destroyed walls and a medieval occupation. Linked with the Roman villa is a necropolis located in the nearby terrains (cfr. Chap. 5).

Figure 8. The Via Cova plan, according to M. Martins

Quinta do Paço Velho, Ponte de Lima This villa is located on the southern hillside of the Santo Estevão da Facha settlement, on a small gentle slope. The conducted excavations revealed a group of compartments with walls presenting 50 cm width that could have been organized around a ‘peristyle’ from which the impluvium existed and has been excavated. From this location, a Tuscan capital proceeded and was found in the excavations. The majority of compartments were rectangular and paved with tiles. The presence of indigenous ceramic, although scarcely represented, indi-

Vila Mendo, Póvoa de Varzim On the Vila Mendo site, a rectangular building was found and it is believed to have been part of a Villa, if we consider the ceramic construction and the domestic materials found scattered around a wide area (cfr. Chap. 5).

From the potential villas inventory and the summary description from the excavated sites, we can withdraw some assumptions.

37 The precise chronology of the villa foundation is still unknown due to the fact that the excavation had to be interrupted for external circumstances.

81

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

First and foremost, the few conducted excavations do not provide sufficient data with regards to these sites infrastructures, which prevented us from fully acknowledging its integrating elements.

back to early first century AD, which is clearly corroborated by the existence of indigenous ceramics in the majority of the excavated sites. The villa of Paço, Vila Cova, Barcelos and Quinta do Paço Velho, Ponte de Lima witnessed a first occupational stage presenting circular houses that were very similar to those existing in preRomanHill-forts38 and that could have existed prior to the future villas construction.

Moreover, we are not even in a position to provide the ‘villa status’ to several of the mentioned sites. This is a common reality happening to most of the locations that are only known due to their surface remains, but occurring also with excavated sites, as it has been the case of Vila Mendo and even Caxinas cases.

The relationship between hill-forts and villas or any other Roman agricultural units seemed to have been inevitable and occurring at an early stage of the Roman occupation. Throughout a certain period, namely the first half of the first century AD, the indigenous occupation and the new sites of Roman typology may have coexistid in a process of interaction and mobility of the indigenous population. On the other hand, the spread of villas implied a change in the territory occupation and in the land privatization, processes that encompassed rhythms and characteristics that were certainly varied.

Under these circumstances, we still need to balance two prejudices: to established the classification of a site as a villa due to inconclusive elements or to reject the fact that we might be facing a potential villa in the absence of attributes considered as significant for their classification, a problem already approached on chapter 1. These events also cater for a more focused future and systematic research over the territory, without losing sight from the fact villas were not the only regulating entity within the rural landscape in Roman times.

The chronology of the abandonment of villas is believed to have occurred in late fifth and sixth centuries, although this is not in any way a common feature with these habitats.

This topic also borders the inescapable challenge regarding the continuity of villas in the Entre-Douro-e-Minho territory.

In the case of the Dume villa, the continuity of space occupation may be explained by its proximity with Braga. Its special status may also be explained by its role within the space reshaping in a Late Antiquity and medieval times, namely including the role of S. Martinho in the mid sixth century. By the constrary the Suevi-Visigothic and medieval occupation of the Paço-Vila Cova villa in Barcelos, required us to be caution with regards to the provided abandonment chronology of these site.

The synthesis provided by Alberto Sampaio (1979) over this subject raised more doubts than certainties; however the author’s intuition continued to be stimulating when considering Roman villas as paramount entities of the northern Portuguese landscape. Jorge Alarcão, in a global critique of Alberto Sampaio’s thesis, questioned this continuity and started by suggesting that villas have been ‘a common property feature of the Romanized northwest’ Alarcão 1980: 174). Based upon data emerging from researched Villas in prospections conducted in the late 1970s and using assessment criteria understood to be unsuitable to the region reality, this author raised questions that are still unanswered to this moment.

This subject also requires the cross reference of several data sources, namely historic references. With regards to the medieval settlement perspective on the region, Luís Carlos Amaral drew attention to the continuity of using of the term villa throughout the Middle Ages, while observing alongside other authors that the term has been the object of successive terminological simplifications that hide distinct functional realities (2007:163-164). It is peculiar to understand the term villa in the eleventh and twelfth centuries. It corresponded to what the author addressed as a ‘growing individualization process of space’ with the most visible effect being the gradual property division.

Focusing on this issue, Manuela Martins attempted to study villas with regards to the archaeological reality of the studied area, as has already been approached on chapter 1. The author suggested that the term ‘villa’ should be ued only to ‘reference sites located in rural areas, established under a Roman architectonic model and focused on agricultural and animal breeding activities’ (Martins 1995: 97), while also considering that some villas within the region could also have had a small construction area and fundi that were of smaller sizes, when compared to structures located in other regions.

Being impossible to elaborate a synthesis work over this topic or answer several of the questions with regards to villas size and architecture, one of the possibilities presented has been to articulate these remains with other space occupation entities, a task only possible to undertake on a different scale, a task that we will address in the following chapters.

The number of excavated villas and the uncovered remains do not add much to this equation. It is not possible yet to draw a synthesis over their architecture or over average sizes of these structures, an issue also applied to the overall size of properties. Moreover, with regards to the foundation chronology of these sites we accept the idea exceeding the classic belief regarding their late establishment. In fact, most recovered data seemed to suggest that most of these buildings dated

38 Again, this issue raised questions related with the materials and indigenous constructions chronologies. It is necessary to investigate the precise date of the living spaces and wider uncovered remains.

82

context, we have not included detailed descriptions, forwarding readers instead to specialized bibliography and to conducted works by other authors and in detailed descriptions related to geology, tectonic and climatic features of the region (Martins 1990; Pereira 1992; Ferreira et alii 2000).

Chapter 4 The Shaping of Bracara Augusta Peripheries

We then analysed the main road network system and its articulation with the city foundation, in order to understand the distinction of the main routes linking it with other Hispanic cities.

‘The city (…) shapes itself as a simultaneous living system of all its precedencies. This stratified simultaneity is systematically sustained. Without a life of its own and an organized starting point, it could not maintain, contain, transmit and continually increase all its inherent remains, together and for eternity’. A. Gnisci, ‘Rome as a ruins system’, in Leonardo Express, Coimbra, 2004, pp 61-83.

Equipped with this data, we focused on a synthetized observation of the ancient settlement patterns, mapping the distribution of indigenous Hill-forts and assessing the registered changes within influential territories. We were equally interested in the assessment of the continuity, retraction and abandonment processes of pre-Roman Hillforts resulting from the impact of Romanization.

Introduction The foundation of Bracara Augusta was part of an overall process that aimed to control the Hispanic Northwest territory and established a new political-administrative framework within the region, in which the urbs performed a key role.

After defining the previous settlements framework, our attention shifted towards the identification and description of several Roman entities including vici, Villas and other sites and remains registering a Roman chronology. Articulating the mentioned observational frameworks, we analysed the city surroundings as a whole attempting to obtain a clearer idea about the different occupational areas around Bracara Augusta. One of the aims of this chapter performed the analyses of the rural land division, which enabled us to put forward a proposal regarding a hypothetic reconstruction of the Roman cadaster.

In a landscape where the urban centres were still unknown, the outline of a city territorial integration, together with its evolution naturally presented them as a rather specific research problem prompting a targeted survey: Which was the city role in defining the structural direction of landscape organization and what were the impacts upon the surrounding territories? In which way the matrices and the evolution of the rural world interfered with the development and growth of cities? How did both worlds merged and how can the processes emerging from this interaction be explained?

1 Analysis frameworks

The site where the city of Braga was established matched a platform registering a maximum altitude of 199 metres and is located on the joining area of the Cávado and Este rivers valleys, the latter being a tributary of the Ave River.

Research tradition naturally created several 'surveys' where cities emerged almost invariably as overlapping entities within their territory. Thus, the surveying of rural landscape imprints and their rigorous mapping are not easy tasks. On the other hand, and providing a special emphasis towards the attention given to conducted research works, the results are not always reliable, both in terms of site analyses and their chronological assessment.

A series of significant reliefs draws a semi-circular outline, located in the southern part of the Cávado River and framed on the mentioned platform with maximum altitude contours ranging between 500 to 600 meters. The altitudes and the lesser elevated areas, registering heights between 20 to 60 meters, shaped a contrasting region (Teixeira and Medeiros 1973) in the center of wich featured the platform occupied by the city and the surrounding plains.

The task of correlating a city with the surrounding rural area may generate a map with few entities that are ‘typologically stable’ providing, at the same time, the urban centre with a key role with regards to the remaining territory. The mentioned unbalance, a product of research differences, seemed to be evident. To achieve a balanced analysis platform ranging between the necessary introduction of known remains and the ability to overcome data differences in both the rural and urban world performed the paramount aim of this chapter.

The rivers in the region follow structural fractures that also refer to their dominant courses . The Cávado river valley and their tributaries present a shift between the erosion alveolus, watercourse valleys and stressed relieves, as do the Ave River and its main tributary, the Este river, which perfors a hydrographical reference for the region. The tectonic origin of these valleys justified the appearance of mineral waters, a factor to be taken into consideration when analysing this area due to its importance in Roman times in terms of the use of thermal sources in bathing complexes.

The studied area was included around the city and referred to approximately 19,5 Km N/S by 23 Km E/W (cfr. Chapter 1, subparagraphs 1.2 and Map 10). The chapter structure was based upon a summarized geomorphological framing of the researched area. With regards to the guidelines of Bracara Augusta surrounding 83

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

A key element within the geomorphological structure of the region is the Cávado River that registers in its upstream another important water course: the Homem River. Up to Barcelos, the river follows a NE/SW course and then a E/W direction ending its course with a N/W orientation. Respecting the landscape features, this important watercourse follows an either stretched or narrow and even wide course with low margins. Near Braga, it presents an open valley and performs one of the natural entities structuring the landscape. These events have helped us explaining the human territorial occupation. Alongside the main hydrographical network, an intense network of water lines comprising tributaries and sub-tributaries follow transversal orientations and explain both the fertile land and the dominant orientation we find in the rural occupation of the region.

2. Entities and relationships

2.1 The road network system and the necropolis

2.1.1 The main itineraries

The structuring of an articulated road network performed a decisive process within the overall organization of the landscape. Functioning as a ‘skeleton’, itineraries ensured in Roman times both the defence and affirmation of the power of Rome over the conquered land and established varied connections between the main populated settlements (Chevallier 1997). They articulated urban centres and regions, but also promoted the link between cities and their surrounding areas. Thus, itineraries performed one of the main regulating entities of landscapes and city/field relationships (Map 10).

From a geological perspective, granitic and metassedimentary formations dominated the territory. The different lithology and their correlation with external geodynamic agents match some morphological variations (Ferreira et alii 2000). Relief variation directly links with tectonic structures and the NNE/NW/SE hircine fracturing is due to a systematic installation of a wide variety of granitites matching over 80% of the area rock support. The region also presents old volcanic rocks of varied compositions. The quartz veins seams to be relatively scarce in the area in some cases with associated mineralizations, namely tin, which was certainly exploited in pre-Roman and Roman times. The dominant granitic rock subtracts provided construction materials of top quality to the communities of Braga and has been intensely explored throughout times. We should also highlight the presence of clay deposits areas (Ferreira et alii 2000) being the object of systematic exploitations. In Roman times, it was certain the use of clays coming from the medium Cávado area that were exploited around Prado, Cabanelas and Ucha. Also certain were the exploitation of clays present in the coastline area and used for the production of fine wares (Morais 2005).

Figure 9. Bracara Augusta: plan of the Roman town, according to M. Martins and M.C. Ribeiro

The region where Bracara Augusta was founded presented features clearly understandable as a junction zone of proto-historic circulation routes and it is this assumption that values the suggestion put forward by Alain Tranoy (1981) regarding the choice over the foundation site of the future city as a ‘meeting point’ of the region indigenous communities.

The region is intensely populated, and in the last few decades has been subjected to dramatic landscape changes both within its public infrastructures and wider several urban ventures. Aerial photograms attest the exponential growth of the city and the alteration of its surrounding areas. In fact, it has been rather difficult to establish comparisons between the 1950 photograms and the current satellite images. In the 1950 photograms, or to be more specific, in the 1948 ones, the rural world still seemed to overlap the city and the clear definition of both ends was not evident. The difficulty persisted in current images, however for different reasons. The city seemed to ‘consume’ rather quickly the rural surroundings, events that made it very difficult to assess the features of the territorial organization throughout time and up to the past few decades. Today, it is only perceptible in old maps or in aerial photography.

Conducted excavations at Alto da Cividade allowing the identification of some Republican coins (Zabaleta Estevez 2002) and some late Republican ceramics (Morais 2005) reinforce that suggestion and enable the assumption that the site of Braga was used as a ‘market place’ (Elena Garrido et alii 2007). Bracara Augusta was established on a privileged site from a circulation point of view, which led Francisco Sande Lemos to consider it as a ‘transport platform’ for the wider peninsular northwest. The relief and hydrographical networks outline, mainly running NE/SW, can explain the main Roman road network for the Braga region. To these structural natural axes are added a group of others itineraries running NW/SE very much linked with the organization of the tributary 84

The Shaping of Bracara Augusta Peripheries

rivers following the relief facture lines and matching the valleys and mountain passages.

sible that the road could have contemplated both exits (western and south) and that these were eventually linked.

In the lower regions, topography enabled an easy circulation. These areas are concentrated on the margins of the Cávado and East rivers and around the courses of their tributaries. In the southeast area of Braga, the areas registering lower altitudes are included within the Ave river basin and communicate with the previously referred area, including in turn rather winding circuits of mountainous locations.

An Adriano milestone (Rodriguez Colmenero et alii 2004) found in Braga indicated mile XXXV to Cale, and may signal mile 0 and not mile 1 as it has been suggested. This Milestone was found in the southern gate of the medieval wall and it is possible that it was moved from the city Roman door to the site where it was later found (Lemos 2002: 104). The Via XVII itinerary exited the city following the decumanus maximus western axis that ended in the Falcões Street near the current Carlos Amarante square. The presence of a necropolis precisely on this area seemed to confirm this possibility (Martins and Delgado 1989-90). The recent discovery of part of this trajectory running inside the old CTT building (Martins et alii 2010; Fontes et alii 2010a; 2010b) seemed to fully confirm this possibility. From there, circulation was extended to the Senhora-a-Branca square via the current S. Victor and D. Pedro V streets alongside Padre Júlio Fragata avenue (also known as Rodovia), S. Victor-o-Velho, Pulo and Nova de Santa Cruz streets. From the Nova de Santa Cruz Street, the road would then run across the ‘Peões Quinta’, a parallel area of the current 103 national road. From there it would follow towards Areal Abaixo and from S. Simão chapel towards the Este river source (Martins 1990; Lemos 2002).

One of the more significant interventions of Augusto in the Northwest was articulated with the development of land itineraries linking the triangle comprising Lucus, Asturica and Bracara alongside circuits ensuring the connection with the remaining Hispanic regions. Thus, the main itinerary networks enabled an effective conquest of the region but also its administrative organization, economic exploitation and the outflow of its agricultural and mining resources. This transport structure was fully consolidated with the Flavian emperors in the second half of the first century AD with the construction of the Via Nova or the Via XVIII itinerary (Tranoy 1995-96). Six main itineraries converged in Braga, five of which were mentioned in the Antonine Itinarary: the Via XVI itinerary – Bracara – Cale – Olisipo; the Via XVII itinerary linking to Astorga via Chaves; the Via XIX itinerary connecting Braga to Lugo via Prado, Ponte de Lima, Valença and Tui; the Via XX itinerary also known as via per loca maritima with an outline still under debate; the itinerary not mentioned in the Antonine Itinerary linking Emerita Augusta via Tongobriga and Egitania, and finally the Via XVIII itinerary constructed in the Flavian dynasty and ensuring the connection to Astorga and crossing the northwest mining region (Tranoy 1981).39

This outline has been partially confirmed by the existence of two Milestones, one located in the current Orfeão square and attached to the Pedro V street final part and another one found a long time ago in Areal de Baixo. The first was dedicated to Tiberius and dates back to the year 32. It has been uncovered in pristine conditions and may indicate mile 0 as it does not indicate any distance within the city (Martins 1990: 89; Rodriguez Colmenero et alii 2004). The second Milestone (ID 3035101) dedicated to the emperor Constant was truncated and does not specify any mile, dating back to the fourth century (333350) (Martins 1990:88-89; Rodriguez Colmenero et alii 2004).

The study of the road network system has been given a special attention within the last few years and intensive field surveys have been carried out in the Via XVII and Via XVIII itineraries alongside part of the Via XIX trajectory. The Via XVI itinerary followed a trajectory between Bracara Augusta and Olisipo and Vasco Mantas studied its outline (1996). The itinerary between Bracara and Cale headed south and would cross Mouta, Capela and Esporões. On these sites there would have been Milestones indicating mile I, II and III (Mantas 1996:850). The same author put forward a proposal referencing a western exit on that road, however several other researchers have questioned this assumption and a southern city exit has also been suggested, on a site near the current ‘Rua dos Pelames’ or ‘Rua de S. Geraldo’. Nevertheless, it is pos-

The Via Nova, mentioned in the Antonine Itinerary with the number XVIII, was built on the Flavian dynasty in the year 80 AD if we consider the oldest Milestone chronology found on this road and dedicated to Tito and Domitian. The trajectory performed an alternative circuit to the Via XVII itinerary between Bracara Augusta and Asturica Augusta and was built in a time witnessing intensive mining activities within the Northwest and a significant economic development of the overall region. More direct than the old trajectory between Bracara Augusta and Asturica, via Chaves with 365,5 km against the 318 km of the Via Nova, this new itinerary crossed regions boosting severe reliefs and low population densities.

39 We forward to the extensive published literature the list and description of the found milestones, both in the city and in the referred road trajectories. The recently published work focused on the milestones and Roman road circuits within the bracarensis, lucensis and asturicensis conventus summed the existing knowledge about this region. It is a valuable and pioneer work tool (Rodriguez Colmenero et alii 2004).

The road exited Braga on the northwest and followed the current outline of the Janes Street, São Francisco square, Chão and São Vicente streets and Infias square, following up to the Areal area. From the Sete Fontes area the trajectory opened up to the Cávado valley and went down to 85

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Adaúfe. A possibility that seemed pertinent entailed a trajectory bypassing Montariol and avoiding the level difference that the trajectory would found if going through the Sete Fontes area. Taking this possibility into consideration, the itinerary would then follow to Adaúfe without major altitude variations. Considering the possibility of made Sete Fontes as the origin of a Roman aqueduct (Martins and Ribeiro 2012; Martins et alii Ribeiro and Meireles 2012), the road would likely keep a certain distance from this infrastructure as a requirement for the necessary preservation of the water supply system. From Adaúfe the road would then run upwards to Ancêde, a site where the Cávado River could only be crossed on boats.40

1983:199). The other Milestone was found in the Frei Caetano Brandão Street, on the basement of ‘Paços do Concelho’ (City Hall) (Fontes 1993:76). Other than the two referred Milestones, there was another monument found on a wall of a shop on the western side of the Frei Caetano Brandão Street, recently refurbished. The Milestone was located north of a door of the Roman wall identified in excavations carried out in 2007. 41 This latter milestone was found in pristine condition and may mark this road mile 0. The outline of Via XX itinerary linking Bracara to Lucus Augusti per loca maritima and mentioned in the Antonine Itinerary has been considered matching the outline of the Via XIX itinerary, at least in its trajectory across the Portuguese territory. However, this is not our opinion and we believe in an alternative circuit between Bracara and the coastline. We suggest the trajectory exited the city via the western door and followed the Naia sidewalk bypassing S. Gregório hill and the south part of ‘castro das Caldas’, in Gondizalvez. Its outline towards the coastline will be further analysed when we approach the settlement in chapter 5.

The suggested trajectory related with the city exit, mainly with regards to its crossing of the S. Francisco square towards the Chão Street seems to have been confirmed with the finding, on the mentioned square, of an altar dedicated to Lares Viales (Cunha et alii 2005) together with the identification of graves associated with a necropolis located nearby and unknown until 1995, when excavations were conducted in order to build a car park in Avenida Central. The trajectory of the Via XIX itinerary between Bracara and Tude was studied by Carlos Alberto Brochado de Almeida (1979) and referred to a circuit following a general outline. However, the prospections conducted to assess the road between Braga and the Cávado River allowed a more plausible consideration that the road exit would have been done via the northern door of the Roman city and followed the cardus maximus orientation. These events seemed to match the presence of tombstones around the Pópulo Convent and the finding of two Milestone fragments nearby (Martins 1990; Fontes 1993).

The suggested proposal for the outline of the Via XX itinerary does not invalidate the suggestion of a secondary road crossing the northern Cávado region that we believe was necessary and very likely confirmed in future archaeological research. It seems, however, that the priveleged road between Bracara and the coastline, important enough to be referred in the Antonine Itinenary may have crossed the southern margin of the Cávado River, as already suggested by Rui Morais (2005). A road not mentioned in the Antonine Itinerary linked Bracara Augusta to Emerita Augusta, the capital of Lusitania via Tongobriga and Egitania. It probably exited the city through a door located on the southeast side, the site that is currently referred as Fujacal where a necropolis was also found. Its outline followed south nearby the Falperra hill via an area accessing the Ave valley and crossing S. Martinho de Sande and Caldas das Taipas (Lemos 2002).

The road would follow a trajectory located on the western part of the referred convent and went around the current street accessing Dume and the new stadium turning left near the latter and continuing north crossing the current S. Lourenço chapel. From there, it would follow a trajectory crossing the Gafos stream, the latter site presenting a name certainly taken from the old ‘gafaria’ that existed on site nearby the referred chapel built near the road. On this site, there was also a crossroad named ‘Quatro caminhos’ that was paved in the past. From here the road would follow slightly to the east up to Carcavelos on the Dume municipality. From the Carcavelos square it would follow to Felgueiras up to Castro Mau and continuing towards Póvoa and Formoso up to the Cávado River that was crossed either by boat or on a bridge that eventually existed on site. The justification of this possibility will be dealt with analysing the rural land division of the Braga region.

As it has been widely accept, the road network system cannot not be fully understood without the analysis of a wide range of associated infrastructures, be they bridges but also supplying structures or settlements functioning as mutationes and mansiones and ensuring the cursus publicus. They have been mentioned in the classic sources namely the Antonine Itinerary (Tranoy 1981; Alarcão 1988).

One of the milestones found near the Pópulo convent was found in one of the building walls of the ‘Patronato de Nossa Senhora da Torre’ (Araújo 1982:142; Santos et alii

2.1.2 The city necropolis and associated epigraphy

The epigraphic information associated with the itineraries was particularly abundant in the region under study. We 40 The Porto Bridge still exists today. It is a medieval construction and it is likely that it substituted a previous construction. However, it is also possible that the river crossing would not have been done via a bridge but in boats a bit, further downstream, precisely at the Ancêde site.

41 Excavations conducted by the Braga Local Authority Archaeological Department, led by Dr. Armandino Cunha, to whom we thank the provided information.

86

The Shaping of Bracara Augusta Peripheries

aim to highlight the group of mentioned Milestones that presented itself as exceptional relevant data when compared to the group of items present in the Hispanic road networks, in general, and even within the Empire. However, there were other epigraphic sources equally representative of the importance of the ‘transport skeleton’ as a structuring entity of the territory and a tool to control the regions. We are referencing the inscriptions dedicated to Lares Vialibus and to deities normally associated with traveling and with the protection of travellers, as for example Mercury (Table 7).

Table 7. Deities associated with the road network system around Bracara Augusta. Deity Lares Vialibus

Orderer Itinerary Chronology Observations Maternus, Second (1) son of Rufus century

Lares Vialibus

Calus Iullus Saturninus

Lares Falvius Vialibus Sabinus Mercury

XVIII

First half of the second century

The identified funerary epigraphic monuments referred in their vast majority to steles, almost always adorned, although there are also some votive altars, plaques and the use of a lintel fragment has also been registered (Tranoy and Le Roux 1989-1990). Data emerging from excavations and Funerary epigraphy allowed the assumption about the significant differences registered in the Via XVII itinerary associated with a necropolis when compared with the remaining ones. While the Rodovia and Campo da Vinha necropolis had inscriptions exclusively revealing the presence of indigenous population, the necropolis from Via XVII had twelve citizens, nine pilgrims and included monuments dedicated to free slaves (Tranoy and Le Roux 1989-1990).

(2)

(3) XVIII

The funerary epigraphic corpus that has been, as much as possible, associated to each of these burial sites, has also been uncovered. This corpus revealed an onomastic panorama registering a strong presence of indigenous population that represented over half of the referred entities present in the inscriptions (51%). The presence of Roman citizens corresponded to only seventeen epigraphic references (21%), fourteen entries accounted free citizens (19%) from where five may have been citizens (7%) and five have been identified as slaves (7%).

(4)

The changes brought by Christianity were naturally felt on new burial sites nearby basilicas from, although not systematically, the fourth and fifth centuries onwards.

Table 7. Deities associated with the road network system around Bracara Augusta.

(1) Altar; unknown precedence (CIL II 2417) (2) Altar; Largo de S. Francisco (Cunha et alii 2005) (3) Altar; unknown precedence and location (Sousa 1973:21) (4) Inscription fragment; Santiago Seminar (Santos et alii 1983)

One of those new sites was the one associated with S. Victor church presenting an inscription dating back to 618 and registering a series of burials in the sixth and seventh centuries (Costa 1979:11-14). The area around the church, more precisely the space between S. Vitor and Martins Sarmento streets and opposite the southeast corner of the church, was subjected to excavations in 2000 and the results revealed a opus signinum pavement (3,3, meters N/S and 2,60 meters E/W). The pavement was probably covered in polychromatic mosaics from which a small fragment was found. There are no exact chronological data, although the graves dating may indicate fourth and fifth centuries burials (Lemos 2001). These findings led Francisco Sande Lemos (2001) to suggest the existence of a villa on site, although we believe the pavement may equally cater for another type of infrastructure.

The exit of roman roads has been closely associated with the location of necropolis due to the religious and legal requirements where the dead were mandatorily located outside the spaces used by the living. This element allowed the use of necropolis as a bordering factor of the urban space and, in some cases their location marked the beginning of the road itineraries. In Bracara Augusta there are five known wide necropolis that can be associated with the exit of previously referred itineraries (Martins and Delgado 1989-1990; Martins 2000): the entitled necropolis of the Via XVII itinerary, the Maximinos necropolis associated with the Via XX itinerary and also linked with the Via XVI itinerary; the Rodovia necropolis marking the southern and southeast exits (exit of the circuit linking Bracara to Emerita and also one of the potential exits of the Via XVI itinerary); the Campo da Vinha necropolis that we associate with the Via XIX itinerary and located on the northern part of the city and, finally, the necropolis found on the western side of Avenida Central and associated with the Via XVIII itinerary located on the northwest side. The known necropolis revealed varied chronologies that went back to the early first century (Martins and Delgado 1989-90b; Braga 2010; Fontes et alii 2010a; 2010b; Martins et alii 2010) and spanned to the sixth and seventh centuries.

The emergence of new burial sites was articulated with the development of new settlement areas as it naturally suggested a movement of the population towards the suburbs and organized around the ancient temples. These events will be paramount within the reconstruction of the urban and suburban spaces in Braga throughout Late Antiquity (Braga 2010).

87

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

2.2 The ancient circuits

Monte das Cadas, Braga (ID 3031801) Located on the slope of a rock knob end formation, it boosts an altitude of 304 meters. It encompassed three wall lines, both circular and rectangular houses and ceramic artefacts from both the Iron Age and Roman times. The site is not well preserved and the lack of archaeological excavations conducted with rigorous scientific methodologies prevents us from acknowledging the exact size of this settlement and the degree of its Roman occupation.

2.2.1 The hill-forts hierarchy

What were the settlement shapes and structures existing in Braga before the foundation of Bracara Augusta? First and foremost, we will analyse the foundational space of Bracara Augusta by deleting all the Roman settlement elements beginning with the city itself. We believe this to be a key exercise in assessing the indigenous ommunities settlement that surely influenced the first moments of Roman history in Braga and the organization of the surrounding rural territory.42

Monte de S. Mamede and/or Monte Redondo, Braga (ID 3031002) Located on a hill reaching 427 meters of maximum altitude, the site had three wall lines and two trenches. Excavations conducted on site in the late nineteenth century uncovered circular and rectangular houses, streets, pipelines and a wide range of artefacts from which we highlight a votive altar and a granitic bust. The presence of streets suggests a wide oppida.

Our studied area included on its western part a valley and the coastline region and on the east side a transition mountainous region with very specific features with regards to its human occupation. We will now proceed with the presentation of Hill-forts (Map 10) that are integrated in A, B and C categories and already mentioned on the previous chapter (Maps 3,4, 5 and 6).

Castro de Eiras Velhas, Braga (ID 3034401) Occupying the highest rock knob end formation with 480 metres of altitude, this is a wide Hill-fort presenting four wall lines and infrastructural remains that are visible on the surface alongside ceramic artefacts from the Iron Age and Roman times. The lack of excavations prevents us from acknowledging its abandonment chronology, as well as, the degree of its Romanization.

Hill-forts of A category

Monte da Falperra, Braga (ID 3031302) Located on the end of a rock knob end formation with a maximum altitude of 563 meters, the site registered an occupation that spanned back to the Bronze Age and was prolonged to the Iron Age according to found artefacts on site. However, no structures from the referred periods are known, although it is likely that they are scattered around different strongly forested platforms. Nevertheless, the site entails two wall lines surrounding the mountain range that can be attributed to a preRoman occupation. The Roman occupation has been difficult to assess with precision, and it is only possible to state that the site registered a late Roman occupation attested by the presence of ceramic materials, mainly sigillata and amphorae and by coins. The upper platform registered a Suevi and Visigothic occupation. Moreover, we can witness some building constructions, one of which can be identified as a palatium associated with a paleochristian basilica.

Citânia de Briteiros, Guimarães (ID 3084101) Occupying the end of a rock knob end formation with 336 meters of maximum altitude, the site had three wall lines that were reinforced with a fourth wall line on the northern side. It was organized around two main streets, with a pavement dividing the occupation area into quarters. The artefacts are rich and mainly encompass indigenous production and several Roman remains, which corroborated a continuity line in occupation of at least, up to the first century. Recent conducted excavations on site led by the University of Minho Archaeology Unit have been attempting to document the degree of Romanization on this relevant oppida, and have determined the pre-Roman character of the protourban organization of this settlement (Lemos and Cruz 2007a; 2007b).

42 Although numerous settlements have been known to exist for some time, we still have some information gaps with regards to the knowledge about these sites. Some settlements located in the Guimarães municipality have data that goes back to Martins Sarmento’s fieldwork in the late nineteenth century and not much data from contemporary research is available. On the other hand, data emerging from excavations is rather scarce despite it having enabled us to address puctually some doubts.

88

The Shaping of Bracara Augusta Peripheries

Castro das Ermidas, Vila Nova de Famalicão (ID 312801).

abandonment chronology. In fact, all the known structures only corroborate an indigenous occupation.

Occupying a rock knob end formation with 209 meters of maximum altitude, there are still remains of a wall reinforced with two trenches and a rampart. Excavations conducted on site revealed a human occupation between the 4th century BC and the 1st century AD. The remains mainly account for indigenous ceramics, Greek and Campanians pottery, glasses and metallic objects. The Hill-fort seems to have witnessed great development processes between the late first century BC and early first century AD.

Castro da Consolação, Braga (ID 30322901) Located on a small hill with 340 meters, this Hillfort likely had a wall and a small trench being also perceptible some ancient platforms. Scattered across the surface are fragments of indigenous ceramics and tegulae. On site are equally referenced capitals and granding wheels findings. The lack of excavations does not allow the assumption over the site Romanization degree.

Citânia de Roriz, Barcelos (ID 3025401) Pau de Bandeira / Monte Castro, Braga (ID 3031402)

Occupying a hill registering an altitude of 324 meters, this wide settlement had, at least, three wall lines and it was possible that there was another one located on the western slope and registering a gentle slope. On the highest part of the site, there are remains of streets, circular and rectangular houses and ceramic materials dating back to the first Roman occupational period of the region. The artefacts register a wide chronological range attesting the idea that this site endured a continuous occupation since the Bronze Age up to the Low Empire. Among the found remains there was a ‘seated’ statue (Bettencourt 2003a).

Occupying a high hill with three wall lines, this Hill-fort registers remains of circular houses and indigenous ceramics dating back to the Bronze and Iron Ages.

Castro de Sabroso / Monte de Sabroso, Guimarães (ID 3082704) This Hill-fort occupies a small hill with 278 meters altitude and presents one defensive wall line. Martins Sarmento has excavated a total of thirty-five houses of round shape and further three rectangular houses. The artefacts are predominantly of indigenous origin and the rare Roman artefacts found in the excavations suggested late first century BC abandonment for this site.

Outeiro do Castro, Barcelos (ID 3020501). Located on a rock knob end formation dominating the intersection of the Covo and Cávado rivers reaching 222 meters of altitude, this Hill-fort was large and surrounded by a wall line that followed its slope and likely boosted other defensive lines and a southern trench. On the surface, there are abundant tegulae fragments, some amphorae remains, imbrex and ‘micacea’ indigenous pottery. Between the first and the second wall lines a fragmented altar and a decorated srone were found.

Monte da Forca / Moinho Guimarães (ID 3087301)

de Ventom,

This is a small settlement located on a hill registering 266 meters high and presenting two wall lines. The rare artefacts found on the surface are from the Bronze Age, Iron Age and Roman times. However, the lack of excavations does not contribute towards a rigorous assessment in terms of the Romanization level on the site and its organizational features.

Hill-forts of B category

Castro Máximo, Braga (ID 3030703) Located on a hill with 170 meters of maximum altitude, Monte Castro or Castro Máximo is today integrated within the urban area of Braga and was located on the suburbs of Bracara Augusta. Circular houses, remains of a wall and a wide range of artefacts from the Iron Age have been found on site, together with occasional Roman remains, such as amphorae and coins. It is believed that it had an occupation up to the first century although the excavation works conducted on site could not determine the degree of Romanization it endured and its

S. Miguel-o-Anjo, Guimarães (ID 3087202). Located on a hill with 385 meters of altitude, this settlement boosts one wall line. Both Roman and indigenous artefacts can be collected from the surface. Its abandonment chronology and its Romanization degree are unknown.

Castro de Boca, Vila Nova de Famalicão (ID 3124002) 89

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

This is a small settlement with 289 meters of altitude located at the very end of a rock knob formation. It has two wall lines and one defensive trench. The site Romanization degree is unknown.

Furthermore, several other authors have also mentioned other sites located within the area and included in our catalogue. Thus, it became difficult to confirm their defensive systems and prevented their inclusion within the following list of sites.43

Low altitude Hill-forts, included on the C category, are only two, and they are both located on the right margin of the Cávado river. We are referring to the Lago Hill-fort excavated by Manuela Martins (1988a) and Lugar de Gaião with a chronology and interpretation still under debate. These challenges will be approached on this chapter.

Monte de S. 3084902)

Bartolomeu,

Guimarães (ID

This Hill-fort is located on a hill with 421 meters of altitude. A trench and a rampart are visible on the northern side. The collected artefacts included an iron scoria and indigenous ceramics together with remains of tegulae.

Hill-forts of C category Monte de 3081502)

Castro de Lago, Amares (ID 3011503)

Santa

Eulália,

Guimarães

(ID

This is a settlement boosting an excellent natural defensive system due to its location on a hill with 285 meters high. In 1933, Martins Sarmento did not register any remains of wall lines but determined the existence of a platform located on the east side, where he found several fragments of indigenous and Roman construction ceramics.

Located on top of a hill with 65 meters of altitude, this site is located near the Cávado River. It presents an irregular polygonal platform surrounded by a wall overlapping another wall line made of ground soil materials. The defensive system included one trench and conducted excavations on site allowed the identification of ceramics from the II Iron Age, having also been found bronxe and iron artefacts. The settlement had a high number of pits, one of them funerary. The conducted excavations did not identify stone houses. However some ground holes suggested the existence of perishable structures. According to the available data Manuela Martins (1988:73, 1990:68-69) considered that this site was not Romanised.

Monte Castro, Guimarães (ID 3084801) Located at 288 meters high, this site is mentioned by Martins Sarmento as having excellent natural defenses mainly on the west side and abundant tegulae, however the author did not mentioned any wall remains (Sarmento 1933:266-268).

Lugar de Gaião, Vila Verde (ID 3130802) This is a small settlement occupying a top of a hill with 60 meters high overlooking the Cávado River. It has an occupational platform surrounded by a slope with a trench on the west side. The ceramics are both indigenous and Roman.

The group of settlements that generate doubts was left out, although the available archaeological data for the majority of the inventoried sites is not homogenous. The cases mentioned below raise challenges in terms of their identification and inclusion of a category. For example, both the Monte de S. Bartolomeu and Santa Eulália performed settlements with an abandonment chronology that can be explained by changes that took place in the late Iron Age, hence they could be included in the B category that also included the Consolação and Sabroso Hill-forts. However, the particular nature of available references demanded extra caution when suggesting their features and chronology.

90

43 The following settlements have been excluded from our database for reasons resumed here: Monte de Vasconcelos located at Adaúfe municipality and inventoried by Manuela Martins and Stªa Lucrécia de Algeriz located at Stªa Lucrécia de Algeriz with survey works both conducted by A. Bettencourt in 1994. Both sites presented an occupation in the upper platforms at the transition age between the Final Bronze and the Initial Iron Age (Bettencourt 1999: vol. IIa 251); Pedroso located at Adaúfe municipality – A. Betterncourt disagreed with L. Fontes (Fontes 1990: 131; Fontes 1993: 36-37) with regards to the site defensive system and included it, chronologically in the transition stage between the Bronze and the Iron Age (Bettencourt 1999, vol: IIa 251); Cabeço da Ribeira located at Adaúfe municipality – the site was inventoried by M. Martins and included within the C category typology (Martins 1990: 209), later, A. Betterncourt classified it as a slope site and included it in the Bronze Age chronology (Bettencourt 1999: vol. IIa 249 – 250), while L. Fontes claimed not having found any remains from an older chronology on site (Fontes 1993: 35); Pedrais located in Gumarães was referenced by M. Sarmento due to the great quantity of uncovered tegulae (Sarmento 1901_ 123-4) although recent works did not provided the site with a clear typology (Dinis 1993:86); Castro da Sola, in Palmeira, Braga – A. Bettencourt conducted excavation on site between 1991 and 1994 and identified three occupational stages: the first going back to the Chalcolithic era, the second was included in the first half of the second millennium BC and the third took place at the High Middle Ages (Martins 1990: 83; Fontes 1993: 62; Bettencourt 1999: vol. IIa 260-261).

The Shaping of Bracara Augusta Peripheries

The observation of this group of hill-forts distribution map (Map 10) in relation with the site descriptions provided a general framework that may help explain some changes between the Iron Age and the late first millennium BC. Moreover, it shed some light in terms of the territorial appropriation rhythms and the population integration in a time under Roman influence.

Considering these base factors, several other events are to be taken into consideration. The region houses hill-forts with no Roman remains, which adds to the conclusion that they were abandoned at the end of the Iron Age, an event recurrent in smaller sized sites located in valley areas and included in the B category.

In general, we may consider that the Hill-forts are located in the most significant reliefs of the region. Their distribution matches the predominant morphology outlines and the hydrographical network of the territory under study.

To understand the exact moment when the pre-Roman Hill-forts were abandoned was not an easy task due to several chronological inconsistencies. In some cases, the retraction of the occupied space or its abandonment time was linked with changing social patterns already taking place during the Iron Age, although the archaeological corroboration of this possibility demands further excavations.

The group of Hill-forts included in the A category registered altitudes ranging between 200 and 500 meters and comprised wide Hill-forts with considerable defensive systems and an occupation which, in some cases, goes back to the Bronze Age and spans up to the first decades of the Roman occupation. In the majority of cases, this occupation revealed a clear dynamism between the late first millennium BC and the first century AD. This group included sites located on the Cávado valley and studied by Manuela Martins. They define the ‘'structural network of settlements in the region’’ durind the Iron Age (Martins 1990:216).

One example is Pau de Bandeira Hill-fort presenting only indigenous ceramics dating back to the Iron Age. In other cases, the presence of some Roman ceramics may suggest an abandonment period between late first century BC and early first century AD. This could, for example, be the case of both Consolação and Sabroso Hill-forts, being the latter the object of on-going excavation works. The settlements located on rock knobs ends formations registered prolonged occupation periods throughout the Roman times and are included on the A category. Two events took place on this category of Hill-forts: an extraordinary development in late first century BC and throughout the first century AD and a gradual retraction from this moment onwards. In other cases, an occupation span throughout the Roman and even medieval times took place although, the details are not fully known.

The group of hill-forts included in the B category is predominantly located in hills with altitudes ranging between 170 and 340 meters and have smaller sizes. There is a lack of Roman remains on site and their abandonment chronologies date back to the millennium transition. The lower altitude Hill-forts included on the C category referred to two cases in the region under study. The Castro Lago was not romanised and was likely abandoned at the end of the first century BC, probably in Augustan times. With regards to the site known as Lugar do Gaião, it is possible that it had a key role in terms of supporting the crossing of the Cávado River since Roman times up to the Modern Age. If we consider these events, it was probably romanised and this possibility emerged with the analysis of its strategic location but this can only be confirmed with future excavations.

These different abandonment stages of indigenous Hillfort seem to have emerged from different phenomena related with the dynamic of settlements development, their strategic importance within the region and with the foundation of Bracara Augusta. Moreover, the effective Roman control over the territory conducted through the structuring of a road network and the establishment of new centres following Roman typologies could also be behind the processes conducting to Hill-forts abandonment.

2.2.2 Maintenance, retraction and abandonment

Changes took place within indigenous communities on the second half of the second century BC. These were associated with the contact from this moment onwards with the Mediterranean and with other Romanised areas in southern Hispania. These relationships seemed to perform a practical example of the Roman military excursions within the Northwest started with D. Iunius Brutus expedition. They enabled new commercial routes to be opened in other areas of the Iberian Peninsula. Respecting this data and the development of A category sites, most of them becoming oppida, the contacts with the south regions should have essentially been maritime and conducted from the coastline. In turn, they demanded the regeneration of proto-historic routes namely the ones located on the Cávado and Ave river valleys.

The region pre-Roman sites endured changes that may be explained by several factors. The transformations emerging from the organization of the indigenous communities alongside their hierarchy processes were consolidated before the foundation of Bracara Augusta taking place before our era. This was a time that also witnessed the establishment of wide oppida within the region (Martins 1996; Martins, Lemos and Pérez Lousada 2005). Another changing factor was associated with the contacts emerging from the Roman incursions in northern Douro between mid second century and the transition era (Martins 1990). Finally, the transformations within the sites structure performed a natural consequence of the definite conquest of the northwest region, the foundation of the city and the development of the Roman road network.

This interaction produced changes that may be observed in several sites of the peninsular northwest. They were 91

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

present in the constructions and/or the defensive systems, in the reorganization of the living spaces and in the protourbanization phenomenon, of which the Citânia de Briteiros is a prime example. On the other hand, the commercial relationship of the northwest with the southern peninsula and the Mediterranean, a predominately coastline phenomenon, impacted upon some regional inner sites as for example Castro das Ermidas, where Greek and Campanian ceramic has been found (Queiroga 1992; Dinis 1993).

The Eiras Velhas Hill-fort registered a certain degree of Romanization justified by its strategic location, both as a control and maintenance site of the Via XVII itinerary. The Roman occupation of Monte da Falperra is not easy to explain, but its location dominating a wide area crossed by two Roman itineraries (the Via XVI and the one connecting to Emerita Augusta via Tongobriga and Egitania) suggested that it may have had a key role in controlling the southern exit of the city, which may have justified the maintenance, at least residual, of the Hill-fort occupation.

Wide settlements as Citânia de Briteiros and Castro de Ermidas witnessed great dynamism and development processes in the late first century BC and throughout the first century AB. Both sites were abandoned shortly after this latter period.

The destruction endured by the Caldas-Gondizalves Hillfort prevented us from having a clear perception over its degree of occupation in Roman times and demanded, therefore, special attention. Similarly to what took place in other regional Hill-forts this site may have been progressively abandoned alongside the first century AD. The fact that it was located nearby a secondary trajectory linking the city with the coastline may have provided this site with some degree of relevance that would in turn, be decisive in terms of a prolonged occupation that we are not able, however, to corroborate.

Different examples refer to sites still occupied after the first century AD. It is important to understand the retraction stage of some Hill-forts and the continous development of others. Furthermore, it is reasonable to state that some Hill-forts were widely active from mid second century and first century BC up to late first century AD due to external phenomena not directly related with the conquest or with the Cantabria wars. In fact, these factors were linked with their strategic location within the préRoman settlement framework, and naturally with the contacts with the Roman power starting with the expedition of D. Iunius Brutus in 138 BC.

Finally, the Castro Maximo or Monte Castro also demanded a special attention due to its proximity to the Bracara Augusta foundation area. Available data enabled the assumption that this Hill-fort was occupied up to mid first century AD coexisting with the first stages of Bracara Agusta occupation, a process that may be explained by the need of indigenous labor for the city construction. If this workforce could be established within city it could also be settled in the outskirts of city, in the Hill-forts that would function as a supporting structure for the city development.

The development of the referred Hill-forts took place alongside the foundation of Bracara Augusta and the city first century of existence. Such circumstances may have occurred from their strategic location within the preRoman settlement, even keeping some strategic importance after installing of new types of habitat that will set new settlement networks.These sites were also privileged in the context of the new circulation routes and the new economic centralities.

As suggested by Manuela Martins (1990), Castro Máximo may in fact be the castellum Agripia / Agripa mentioned in two funerary inscriptions found in the city. The monument was named after Agripa, a tenant of Augusto and one of the most influential characters related with the Emperor and apparently with the foundation of Bracara Augusta.

However, the new routes and centralities defining a new social order required time to be fully established and disseminated. Thus, there seemed to have been a coexisting period with old and new settlements existing alongside. This period was inevitably followed by retraction phenomena in certain sites and continuity in development in others.

2.3 Roman settlements

The maintenance in occupation in some Hill-forts may be explained with the establishment of a new settlement network and with the development of economic routes. In some cases, the survival or continuity processes took place on sites located nearby road trajectories, as for example, Monte Redondo-Guizande, Monte da Falperra and Eiras Velhas and Caldas-Gondizalves. They were certainly key in controlling the regional transport structure.

The geographic and hydrographical features of the Bracara Augusta, together with the establishment of the Roman road network prompted a considerable human occupation that was witnessed by the remains mapped. Among the entities that we can consider of Roman origin it is necessary to examine their separate categories and the relationships that seem to establish with each other (Map 10).

The Monte Redondo-Guizande Hill-fort dominated the Via XVI itinerary between the fertile plain of Penso and the Pelhe river valley, on the south side. This occupational continuity throughout Roman times enabled the assumption over the significant relevance of this settlement as a control and support site of this itinerary.

2.3.1 Vici and road stations In the analysis of ‘aglomerados secundários’, the road network is evidently a regulating entity. It included itineraries that were articulated with a wide range of secondary 92

The Shaping of Bracara Augusta Peripheries

and private circuits structuring the landscape and ensuring the circulation of people and goods.

for this site, identified as a low altitude Hill-fort, to have functioned as a supporting station to the Cávado river passage. A similar situation took place with Outeiros dos Picotos Hill-fort (ID 3060701), a site offering similar features and located a few kilometres distance from the Cávado river mouth (Morais 2005: 62,67).

When analysing the main itineraries, one must understand that this complex road ‘skeleton’ also included a series of road stations and supporting infrastructures structuring the cursus publicus.

The permanence and relevance of this passage was well attested in the Modern Ages, through the well-known ‘’Estrada da Graça’’, an important circuit supervised by the Tibães monastery community, who also controlled the river crossing in ‘’Barco da Graça’’, as previously referred by Aurélio de Oliveira:

As previously observed, it was possible that some Hillforts were invested with this type of role, at least, throughout the first stage of the Roman occupation within the territory. However, other sites may have performed the role of road stations. Jorge Alarcão (1996) suggested predetermined distances, both for short stops - mutationes - and for longer stops demanding overnight stays - mansiones. In some cases each of these sites may have undertook wider roles, becoming greater sites, namely vicus.

‘The important crossing of Cávado, in Graça, was of exclusive jurisdiction of the Abbey (…). This passage not only served people, but also horses, cars, troughs and litters. The renting costs of Barco da Graça were also an indicator of the relevance of this passage (…). This pathway would cross the Cávado river via S. Eulália de Cabanelas and Cervães. On site, we found a reference, near Prazos da Mitra de Braga and dating back to the 17th and 18th centuries, of a road leading up to Viana do Castelo. This road probably was the extension of this important passage and of the Couto circuit, which must have lost its relevance with the improvement and enhancement works of the Prado bridge (…)’ (Oliveira 1979:69-70).

We accept that in Prado (ID 3134202), located around 7,2 km distance from Bracara Augusta, there was a settlement that initially may have had a role of supporting the Via XIX itinerary but later evolved into a vicus. It had, first and foremost, a privileged location on the northern margin of the Cávado River linking not only to the river and to the Via XIX itinerary but also to other secondary circuits on the western and eastern sides. Other than the mutatio or mansio roles existing on the Via XIX itinerary, this location was at the core centre of a key region in terms of economic relevance, which may have resulted in a significant destruction of the site due to intensive clay exploitation works carried out throughout time. (Martins 1990; Morais 2005). We are aware the region between Prado and Ucha was a clay exploitation area in Roman times. It is possible that it also functioned as a pottery production area in progress throughout the Middle Ages up to today (Morais 2005:355).

The author also mentioned reconstruction works conducted on this circuit in mid eighteenth century, stating: ‘’All the paths leading to Graça were widened and a new pavement was implanted alongside with a maritime terminal providing an easy and safe docking (…)'' (Oliveira 1979:72).

The existence of a Roman production centre between Prado and Cabanelas is almost certain. This idea is linked with the great variety of ceramic remains found in the region. It may have functioned as a pottery centre supplying the city and the area is of paramount importance in understanding the economic role of this particular site within the overall territory. The identification of furnace remains demanded an intense field survey study in order to confirm the exact location of the Prado vicus, its size and the relevance of the area as a supplier and production centre for the region. These factors must be taken into consideration within the overall perception of the road network crossing the area.

The relevance of this passage in the Middle and Modern Ages was related with the status the Tibães monastery detained in the region and with the need the monastery had to maintain and manage the itinerary crossing the ‘Couto de Tibães’ area. This road circuit, of strategic relevance to the monastery, was the end result of the appropriation of an old road, probably of Roman origin and presenting an outline ensuring the connection between the exploitation and the production units based in the Prado/Cabanelas area.

Accepting that the Via XIX itinerary crossed the river somewhere around the current Prado Bridge does not prevent the existence of other passages crossing the Cávado River.

Nearby the road exiting Braga on the western side, located on the southern part of the Cávado River and around 7 km distance from the city, exists a site entitled Martim (ID 3024701), that has always been identified as a Roman Villa (Almeida 1972; Martins 1990).

In the case of Prado / Cabanelas area, it is likely that another trajectory crossing the Cávado existed near the Gaião site (ID 3130802) on the northern margin, allowing a secondary circuit to link with the Via XIX itinerary. It was also possible that an alternative route existed linking this area with Braga. The fact that this itinerary was crossed around ‘Lugar do Gaião’ site makes it possible

Excavations carried out by Carlos Alberto Ferreira de Almeida on site allowed the identification of remains from a building, one that seemed to have had at least, two occupational stages: one in Roman times and another in the Suevi and Visigothic times. On site, some Roman padded stones were uncovered together with a door 93

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

threshold, opus signium, part of an ionic capital, the base of a column, part of a temple podium and some column shafts (Almeida 1972). If we consider the assumption that this was in fact the main road circuit distributing all the goods arriving to Braga coming from the coastline, particularly from a fluvial port near Areias de Vilar, it was possible that there was a supporting station in Martin, near the circuit of the Via XX itinerary. The location and the type of remains found in Martim, nearby the stated road itinerary, suggested the existence of an open settlement connected with the road instead of an agricultural site of villa typology. We must, however, understand what type of infrastructure we are dealing with. If it was a mutatio, it would have been a small building supporting short stops. Nevertheless, the presence of a temple implied a wider social and religious dimension present on site, justifying it has persisted up to Suevi and Visigothic times. Although excavations have been conducted on site, it was not possible to fully understand the type of infrastructure we were dealing with. We could only question the fact that the site was not only a Villa, but instead a populated settelement housing functions that likely surpassed the ones required by a simple mutatio. In southern Braga, nearby an itinerary linking Bracara Augusta to Emerita Augsta via Tongobriga, a location called Caldas das Taipas existed (ID 3080801) at around 9 km distance from the city. It is believed that it was a vicus encompassing several transport and thermal functions (Martins 1995). An important honorific inscription dedicated to the emperor Trajan engraved on a rock was found on site (CIL II, 4796=5560). It dates back to the year 103.

All the villas listed on the present work have in common a privileged location, nearby lands with great agricultural potential, good sun exposure, sheltered from winds and with access to a dense hydrographical network. Their location also followed other common parameters. They would either be located on slopes or on small hills with altitudes ranging between 50 and 100 meters. This association of features, also present in other peninsular northwest sites (Pérez Losada 1996), provided these infrastructures with a polycultural economy based upon communities' self consumption but also on the production of needed surpluses destined to supply Bracara Augusta (Martins 1995b). The attraction of road circuits exerted upon population centres was recurrent, although the establishment of these roads respected detailed criteria. Thus, they were structured at a certain distance from main itineraries in order to provide them with a safety measures so they could properly function and grow. For the region under study, this distance has been accepted to be between 500 and 1000 meters. Under these circumstances, the transport skeleton of the region not only encompassed the main itineraries but also an intense network of ramifications, namely private and narrow trajectories. All these factors prompted a close relationship between the city and the rural world, with the city exerting a significant attraction within rural world. The list of identified villas located in the outskirts of Bracara Augusta will follow (Map 10). Couturela, Braga (ID 3034603) The Couturela villa is located on the immediate suburbs of the Roman city, occupying the upper platform of a small hill. The contemporary urbanization on site, conducted without any archaeological supervision, uncovered building remains, architectonic elements, several Roman construction and domestic ceramics. Luís Fontes considered this site to be the ‘Villa Pascasi’, a Villa referenced in 1911 and believed to have existed around the Dume municipality (Costa 1965: 38-40; Fontes 1993).

Although, the remains of a Roman thermal infrastructure seemed to have long gone, perhaps due to the construction of another modern thermal structure, the nature of the mentioned inscription attested the importance of the site, at least, during Trajan times. Thus, the Caldas das Taipas site may have functioned as a thermal vicus including road functions. Encompassing multiple roles may have been common with theses type of sites. Furthermore, the proximity of the site with the Citânia de Briteiros (ID 3084101) may hypothetically illustrate the changes operated within the region on the first century BC.

Infias / Pachancho, Braga (ID 3030352) The site is located on the outskirts of Bracara Augusta and is mentioned as being named Villa Infidias. It is also located in Dume (Costa 1965:3840). Architectonic elements have been uncovered at the old Pachancho factory, namely Roman columns, in excavations conducted in 1917 (Oliveira 1979: 174-175 and 182). The fact this site is relatively close to the ‘Casa Infias’ enabled the interpretation of it being an archaeological site of the Villa typology.44

The Citânia de Briteiros site safely maintained an occupation throughout the first century of our era, but likely endured a progressive population dispersion of its inhabitants towards other Ave valley areas that, at this time, registered intense rural exploitation works and towards sites with growing dynamism, such as Bracara Augusta. Other settlements emerged nearby the roads, as for example Caldas das Taipas or Caldas de Vizela (ID 3085201), the latter an important Roman vicus.

44 The findings were uncovered in the western part of the Nova do Areal Street when construction works took place in 1917 to buil a block of a new hospital, building that was not finished. Although the current Infias Palace is located around 400 meters from the site

2.3.2 Villas and associated epigraphy 94

The Shaping of Bracara Augusta Peripheries

Near the parish church of Espinho, a funerary monument has been uncovered (Santos et alii 1983, 198). On the surrounding area, it was possible to collect fragments of tiles and Roman pottery. The presence of an inscription and other Roman remains on the surface together with the topographic features of the site suggest we are dealing with a villa.

Santarão, Semelhe, Braga (ID 3035301) Since the nineteenth century, coins have been found at this site, namely from the Tiberius, Nero, Agrippina and Antoninus reigns. Roman bricks, pipelines, building remains and architectonic elements, from which two capitals in granite, have also been found. The site, of 75 meters high, was referenced in medieval documentation by Parata Samueli (Costa 1965:38-40), a name that Luis Fontes believed originated the Parada (de Tibães) and Semelhe municipalities (Fontes 1993:77). The finding of a statue pedestal dedicated to the emperor Augusto, dating back to the year three BC (EE, VIII, 280=ILER, 1028) was found on the Senhor dos Lírios chapel, also located in Semelhe, adding importance to the site. It is widely believed that this statue may have been moved from the city to a Farm on the outskirts. Chronological findings attest that this Villa is from the first century AD.

Lomar-Igreja Velha, Braga (ID 3032301) In the surrounding area of the Lomar monastery we can find Roman construction materials. A Crispus Milestone associated with the Via XVI itinerary and two funerary inscriptions are also believed to be from this site. One of the inscriptions is particularly relevant due to the fact it refers a citizen inscribed in Quirina tribe (CIL II, 2438), which under the name of Titus Caelius Flacus, seemed to have had an indigenous origin but likely acquired Roman citizenship in the Flavian times (Martins 1995).

Dume, Braga (ID 3031001) Conducted excavation led by Luis Fontes from 1987 onwards revealed the existence of an important Roman villa located on the surroundings of the current church. The identified structure seemed to correspond to the villa pars urbana. It had a fully preserved balneum with a layout not matching the villa structure. Instead, it had one that adapted to the terrain morphology. Moreover, a Suevi basilica has also been identified and was associated with the Dume diocese. It has been regenerated in the High Middle Ages. This is an important site registering continuous occupational stages up to our days. It started as a Roman villa dating back to the first century AD and was later transformed into a monastery attached to a basilica (Fontes 1992; 1993: 39-43).

Esporões-Igreja, Braga (ID 3031301). At the Esporões parish churchyard, a votive altar dedicated to Jupiter alongside a property stone were recovered (Santos et alii 1983:186-187; 201). In the nearby area, it is still possible to recollect tegulae, tiles and common Roman ceramic. The topographic features of the site, reaching 220 meters high, together with the mentioned Roman remains suggest the existence of a villa in the surrounding area of the church.

Pateira-S. Pedro de Merelim, Braga (ID 3032904) Field surveys conducted in the area nearby the Via XIX itinerary detected the existence of column shafts and a stone winepress in a property with boundaries coinciding to the outline the Via XIX itinerary and matching on western side with the current outline of the circuit linking Prado to Braga. Construction and domestic ceramics dating back to the Roman times can still be found in the nearby area. We believe this exploitation unit was linked with an inscription uncovered in S. Pedro de Merelim mentioning a Roman citizen from the Quirina tribe (ID 30032903).

S. Miguel de Gualtar, Braga (ID 3031901) In the current church surrounding area, it is possible to observe remains of tegulae and tiles. Equally known are two unparalleled funerary inscriptions. Excavations conducted in the 1990s found padded stones that were reused in the medieval construction of the primitive church. This attested the existence of a Roman site on this location assumed to be a Villa due to its topographic features (Martins 1990:84, Fontes 1993:55-56).

Santo Estevão Velho or das Pardelhas, Braga (ID 3033102)

Espinho-Igreja, Braga (ID 3031201)

In Santo Estevão Velho also known as Pardelhas, located on the southern area of a distillery building, great quantity of construction and common Roman ceramics alongside traces of walls have been found. The topographic features of the site, as well as the

where the mentioned columns were found, Luís Fontes believed there was a coorelation between the structures and the Infidias villa (Fontes 1993). This theory seems more probably rather than the suggestion that the items were milestones from the Via XVIII itinerary.

95

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

mentioned findings, suggest the existence of a Roman villa.

Some sites included on this category left us with interpretation doubts, reason why we listed them separately. They are:

Outeiro de S. Sebastião, Braga (ID 3033101) The site occupies a flattened area where great quantities of Roman ceramics can be found. The area is located between the Outeiro de S. Sebastião and the airfield area of Palmeira. At the Outeiro de S. Sebastião nearby the parish church and located on a small hill, a treasure with around 4000 coins was found, presently disappeared. A column base and two capitals coming from Palmeira are currently part of the Pio XII Museum assets (Martins 1990:84).

S. Frutuoso de Montélios, Braga (ID 3033701) A coin treasure and fragments of Roman common ceramic, of both construction and domestic typologies, were uncovered in a wide area that is today included in the entitled ‘Horta Pedagógica’ nearby S. Frutuoso. Although this site presented us with doubts we believe it is possible that the site of the ancient church of S. Frutuoso may have been a Roman villa.

Adaúfe Church, Braga (ID 3030103)

Assento/Manhente, Barcelos (ID 3024501)

In the surrounding area of the Adaúfe parish church, column shafts were found alongside Roman construction and domestic ceramic remains. A votive inscription devoted to ‘’Larisefius or Lari Sefio’ was found on the same area although its exact location is currently unknown (Martins 1990: 80-81). A funerary inscription found on a wall of the Santo André church in the same municipality may be associated with a potential Villa on this location to which all the mentioned finding may also be associated. This site has been identified as a mutatio by F. Sande Lemos.45

In a relatively wide area including a churchyard and the adjacent garden, tegulae and common pottery have been uncovered. This site may be linked with the ceramic kilns present in Telheiras. According to Carlos Alberto Brochado de Almeida (1996), the fact that within the proximities there was a secondary road crossing the Cávado River and going towards Ponte de Lima, should be taken into consideration when assessing the site typology. We are not aware, however, if we are dealing with a villa or a farm infrastructure.

Santa Luzia, P. Lanhoso (ID 3091701)

Lajes, Braga (ID 3033501)

In a relatively wide field area located between Lamas and Castro streams and corresponding to river terraces with a maximum altitude of 100 meters, tegulae fragments have been found. At the Souto site, one column used as an infrastructure to support a rural housing porch has also been uncovered (Martins 1990:91).

This site is located on an agricultural terrain overlooking the Cávado River. It occupies a large Farm between Lages and Moutorre. There is a wide area where it is possible to find great quantity of Roman construction and domestic ceramics alongside some indigenous ceramic fragments. The findings seem to be concentrated around the main house indicating the presence of a potential villa, also suggested by M. Martins (1990: 88) due to the great quantity and dispersion of the materials.

The villas proximity in relation with Bracara Augusta performed an assessment exercise that prompted several interpretation lines ranging between the ones directly articulated with the influence the urban centre exerted in terms of delimiting the rural area and the ones where the political role assumed by the terratenants within the Roman city administrative organs, living outside the city but in the surrounding area, so they could fulfil their duties and implement their influence.

Paço, Barcelos (ID 3025201) This site is located between the parish church and the Monte da SaiaHill-fort. On site there are tegulae, 'imbrices', common Roman ceramic and remains of a wall. If we related these findings with a funerary inscription found in ‘Quinta de Fralães’ one may suggest the existence of a Villa. The area registers topographic features in line with this type of infrastructure.

The attraction Bracara Augusta had over villas dissimination is noted in the wide presence of remains attributed to this type sites existing within an area of 5 km around the city (Map 11).

45 This interpretation has not been published, however we thank the researchers the suggestions put forward.

Here we find a variety of sites listed as Villas together with a reasonable density of other remains that may be 96

The Shaping of Bracara Augusta Peripheries

connected with this type of establishments. On the other hand, it is important to highlight their proximity with regards to main itineraries linking the city with other urban centres and the fact they could have been located nearby private or vicinal routes. The presence of these infrastructures may have conditioned the habitat organization although for this region it is difficult to distinguish Sites

Bracara

Infias Coturela Lomar S. Frutuoso Dume Santarão Pateira Gualtar Esporões Outeiro S. Seb Adaúfe Stº Estevão Espinho Quinta de Stº André Lajes, Braga Assento/Manhente (1) Stª Luzia, P. Lanhoso Paços, Barcelos

Via XVI

950 1000 2000 2000 2400 2500 3900 4000 4100 4700 4900 5100 5300 5900 9900 12200 13000 15700

where the city influence ended and where the road one started attracting this type of Roman rural sites.

Via XVII

Via XVIII 200

Via XIX

Via Bracara /Emerita

Via XX

700 300/400

10

400 1000

350 508 1500 80

1900

510 984

500 1200 950

2200 200

2700 600

3400

1590 3500

Table 8. Distances between Villas and Bracara Augusta and the nearby roads (in meters)

1) 585 meters from the Cávado River Observation: Highlighting the Villas located on the outskirts of Bracara Augusta

A group of other findings may become relevant in the assessment of the present Villa distribution framework located in the rural surroundings of Braga.

Thus, the Flavian chronology suggested by Alain Tranoy for this dedication (1981:329-339) must be, in our understanding, reassessed. Tranoy's suggestion was linked with the belief that the conventus organization took place in the Flavian dynasty. We are currently aware that the conventus may be earlier and very likely took place in either Augustan or Tiberius times. Furthermore, we should highlight the onomastic and the filiation scheme from the commemorator, who was clearly of indigenous origin.

The Dume Villa presented a considerable and important group of funerary inscriptions, both votive and honorary, fact the corroborated the relevance of the site since the beginning of the Roman presence within the region, although it should be taken into consideration the possibility of some of these pieces may not come from this site.

The name Camalus is attested in several inscriptions at the Citânia de Briteiros and it is equally present within the epigraphic file of Bracara Augusta. Accepting the Augustan/Tiberius chronology for the conventus foundation, we may admit that the honoured Camalus within the Dume inscription may have been the first conventus priest. Moreover, it is also key to highlight the fact that this priest added to his filiation – (CA)MALO MELG/(AECI FILIO – the fact that he is B)RACAR/(V)GVSTANO.

From this group we highlight an honorific voting altar dedicated to Camalo, son of Melgeco, Bracaraugustano, priest of Rome, of Augusto (and) of the Caesars, from the bracaraugustanus conventus (Tranoy 1981:329-330). It is therefore, an inscription dedicated to a priest of the imperial cult, whose name and affiliation documented an indigenous origin. Admitting that this inscription is from Dume and was not moved from the city, a place more suited to its location, we may consider Camalo the probable founder of the Dume Villa, or a descendant of the founder. This interpretation demands a reassessment of the inscription chronology, while taking into consideration the date suggested to the creation of the peninsular northwestern conventus (Dopico Cainzos 1988; Rodriguez Colmenero 1996a; 1996b; Fernandez Ochoa and Morillo Cerdan 1999).

Another inscription coming from Dume was the vow dedicated to Jupiter Depulsor or Repulsor – IOVI (D VEL R)EPVLSORI – by Durmia Pursina (CIL 2414=ILER, 102). Two other inscriptions related with Votive epigraphy presented similar texts: one was found in Esporões, Braga and the other in the S. Mamede Hill-fort in Guizande. The 97

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

commemorator was, in both cases, a free man under the name of Hermes.

There are three known necropolis located in the area around the city, excluding the ones directly related with the urban centre.

The votive altar found in Esporões (ID 3031301), was dedicated to Jupiter Optimo Maximo by Hermes for the health of Tiberius Claudius (?) Niger (or Nigrinus), probably the latest master: PRO S(alute) / TI(beri) C(laudi?) NI(S)G(ri vel rini) I(ovi) O(ptimo) M(aximo) / HERMES/ V(otum) L(ibens) S(olvit).

Carquemije necropolis, Braga (ID 3031002) This necropolis is most certainly related with the Dume villa. Located around 1km away from the urban part of the Villa where three graves were found at the time when building works were conducted to open a modern road (Martins 1990:83). The inscriptions were, in their vast majority, found outside their original context and they had been taken to the Dume church or found in other areas of the municipality. The known epigraphy caters for a social framework of great relevance in terms of understanding the role this Villa had throughout Roman times (ID 3031001). One of the inscriptions regards the castellum Agripa and is included in an inscription dedicated to Apilo, son of Arquio (CIL II, 2433). This dedication goes back to the beginning of the region Romanization process, and referred an indigenous man coming from a castellum (perhaps the Castro Máximo Hill-fort), who is then located and buried here. In another funerary votive inscription dating back to the first century, a reference is made to a citizen, Lucretius Saturninus inscribed in the Quirina tribe (CIL II, 2444). Furthermore, and on another funerary inscription of extreme relevance a reference was made the Citerior Hispania Province flaminia, pro Nigrina, in a monument enshrined by the heirs (Tranoy 1981:330). From this necropolis was also a fragmented funerary inscription, where only the last two lines can be read. They indicate the age of the deceased (CIL II, 2456). The only other inscription found on the necropolis site (ID 3031002) referenced Abronia (or Apronia) Quinta, daughter of Abronio (or Apronio) Reburro (Tranoy and Le Roux 1982, 32-33).

At the S. Mamede Hill-fort, in Guizande (ID 3032002), the dedication is made by Hermes to ANTISC/REO, an indigenous deity: PRO S(alute) / TI(beri) C(aesaris) SE(…) / HERMES / V(otum) S(olvit) L(ibens) M(erito). The reason behind these homages was likely the acquisition of freedom on behalf of Hermes and, assuming we were dealing with the same person, it is curious the need to perform two votes asking for health for the same person: to Jupiter in a site where we assumed have existed a Villa, and to an indigenous deity in a settlement located nearby the Esporões Villa. Finally, on another votive inscription, one eventually linked to a Villa, there was a reference to Lares Sefius. Lares may assume a protecting function to both the commemorator and his family (Comes pro salute sua et suorum). Some Villas referenced for the region around Braga are associated with necropolis or funerary inscriptions that apparently have proximity relationships with the described entities. The inscriptions performed a precious data source that enabled us to understand the onomastic behind the owners of the mentioned Villas and the proximity with the social outskirts region of Bracara Augusta.

2.3.3 Necropolis and associated epigraphy

Necropolis, alongside funerary epigraphy, performed an important settlement sign. If the location of necropolis does not provide us with the precise location of a habitat, it indirectly performed, due to proximity relationships, a coherent network of ‘’signs’’. Funerary monuments have a different value, as they may be found outside their original location. Inscriptions have often been found in churchyards or have been used as construction materials. This was typical in both rural and urban sites. Nevertheless, and although many funerary epigraphic cases do not provide georeferenced data, they performed a substantial added value as they provided written data that generated an image, even though an incomplete one, of the people who have been buried. Furthermore, if we analyse data provided by funerary inscriptions present in the region under study with the inscriptions of votive monuments, we are able to generate a social framework similar to the one found in the city with regards to Romanised indigenous presences, to citizens and in some rare cases to free man. 98

Cabanelas necropolis, Vila Verde (ID 3130801) Identified between Veiga and Cachada, this necropolis was entirely destructed when it was uncovered. The few findings regard tombs constructed with tegulae associated with ceramic vases. From Cabanelas and without any precise information, was a votive inscription currently located at the Pio XII Museum, in Braga. We are certain that this necropolis is associated with the settlement that may have existed in the Prado exploring the clays from the region of Prado and Cabanelas. Nevertheless, it is not possible to put forward any further interpretation lines without further information.

The Shaping of Bracara Augusta Peripheries

Stª Maria Madalena necropolis, Areias de Vilar (ID 3021103)

manised indigenous. It was a tribute to Mario, son of Adrono, who had at the time six years old.

Albano Belino (1909:25) published the finding of a necropolis nearby the referred chapel. In the immediate surroundings there were also Roman tiles (Almeida 1996:218).

2.3.4 Agricultural and artisanal exploitation units The most significant ‘’industrial’’ area acknowledged to have existed in the region under study is without any doubt the one located between Prado/Cabanelas and Ucha. It functioned as a supplying centre of clays to the pottery units located in Bracara Augusta. It is also likely that the 'industrial' status of the site is behind its transformation into a vicus (Martins 1990).

Other than the funerary inscriptions mentioned and related with the Dume necropolis and Villa, there are other five monuments of this typology located in the region under analysis. One of the known inscriptions was found in a wall of the S. Pedro de Merelim parish church (ID 3032903) wall that has then been reused as construction material. The inscription referred a Roman citizen, Lucio Valerio Rufino, who was registered in the Quirina tribe (CIL II, 2450=ILER, 5112). The homage was paid, by heritage disposition, by his son and heir, seemingly indicating a family necropolis of citizens living in the Villa.

The Cabanelas area was also believed to have been a pottery centre (Morais 2005). It produced fine ware and construction ceramicl used in the city buildings and infrastructures. In the Prado/Cabanelas area, a necropolis (ID 3130801) was found alongside an overwhelming amount of ceramic remains scattered around a wide area. Further west, a low altitude hill-fort was found in Ucha, Barcelos and is today known as Crasto (ID 3028201). It may have been linked with the exploitation of clays, although the scarce collection of surface materials only enabled us to suggest this, and not confirm.

Field surveys conducted nearby the Via XIX itinerary suggested that this necropolis was linked with the remains found in Pateira, a Villa located around 4 km from Bracara Augusta. The reuse of epigraphic monument on the wall of a church is not a surprising fact and it also suggests that the necropolis was in fact located nearby. The western boundary of the Pateira Villa was likely located in terrains nearby the present road linking Braga and Prado, the same site where the S. Pedro de Merelim church is currently located.

Nevertheless, the presence of a centre that was simultaneously a supplier of clays and a producer of ceramics located around 7/8 km from Braga was of particular relevance within the economic context of the city with its surrounding region. However, the size and real status of the Prado site as a production centre may only be fully understood with further deeper research and further field work to shed some light on the potential relationship between the site and the Roman city.

In a wall from the Santo André chapel, another epigraphic monument was found and its reading enabled the recognition of its funerary typology. The inscription is very likely related with the potential occupation of ‘Quinta de Santo André’ (ID 3030101), a site located around 6 km from Braga, nearby the Via XVIII itinerary and where abundant Roman ceramic fragments have been uncovered.

In terms of what we can assume as ‘’industrial’’ infrastructures, there is a fullonica associated with a family probably coming from Lugo (Tranoy 1981) that should have had its workshop in S. Vicente, Montariol where a votive inscription paying homage to Marte Tarbucelis (ID 3031003) was identified.

Also related with a Villa located in Espinho (ID 3031201) and around 5 km from Bracara Augusta was a votive inscription presenting a text that corroborates the funerary typology of the item and was clearly of Roman origin: [... / … ]TOV […/…] AN XXI [.. /.] V […] MAT[(er?)] / FILIO CAR[I/S]SIMO FECI[T]... From Lomar (ID 3031301) were two funerary inscriptions. One reads [A]RRVNTI VEGETI [...] and the other referenced a Roman citizen registered in the Quirina tribe: T CAELIO T F / QVIR / FLACCO, an apparently promoted indigenous, already mentioned when referring the Villa. A funerary inscription of the legionary Aurélio, son of Patrício, was found at the ‘Quinta de Fralães’ (ID 3025101) and is very likely related with the Paço Villa (ID 3025102) located on the foothill of the Hill-fort entitled Monte da Saia. Finally, on the Ribeiro site located in Lago, Amares (ID 3011301) a funerary inscription was uncovered and due to its onomastics, it may be associated with a family of Ro99

In the western region of Braga, currently Barcelos municipality, there were remains of ceramic kilns, which in some cases were associated with agricultural establishments as it was the case of Stª Maria de Galegos (ID 3026801). A series of sites that we assume had smaller sizes may also be linked with agricultural exploitation units. The observation of their distribution identified a great number of these types of establishments in the western region of Braga. This will be the object of a detailed study on chapter 5. The observation of the intense dispersion of these small sites seemed to respect a different type of organization and exploitation of the overall rural space. This fact will also be analysed on chapter 5 and we will now focus in the area closer to the city.

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

ID Number Designation 3030104 Fontela – Adaúfe, Braga 3031003 Casas Novas 3031004 Remelhe 3032902 Castro Mau 3033103 Stº Estevão 3134204 3091801

Barreiro, Vila Verde Quinta de Arrabalde, Lanhoso

Findings features Ceramics

stops throughout the roads can helped us estimate the possibilities we were dealing with.

Ceramics Ceramics; pipelines Ceramics Ceramics; anepigraph votive altar Ceramics Ceramics

Taking this criteria into consideration and after assessing the available data, we may admit that Bouça Alta, in Braga (ID 3032701) could have been a mutatio from the Via XVIII itinerary and Castro Mau (ID 3032902) or Santo Estevão (ID 3033103) other supporting infrastructures of the Via XIX itinerary.

Table 9. Agricultural exploitation units

The general trend observed with regards to the preferential location of small agriculturalsites, suggested by areas with ceramic remains reveals that they exist in areas of fertile soils and nearby Roman roads, which is in line with the distribution observed for Villas. However, the identification of some of these rural exploitation areas is impossible to confirm regarding their size and chronology. The proximity relationship of some Villa with the location of sites with sactered ceramic materiasl suggesting the existence of agricultural associated establishments confirmed the idea the latter structures may had been associated or subsidiary infrastructures from Villas. This could be the case of the Remelhe site (ID 3031004) located nearby the Dume Villa, the Santo Estevão site likely related with the Santo Estevão Velho Villa and also a site listed as Quinta de Arrabalde (ID 3091801) nearby the Lajes Villa, in Póvoa de Lanhoso (ID 3033501). Other than small exploitation units corresponding to integrated villae structures, there were certainly familiar agricultural exploitation units of small sizes that could be listed under the Farm category. They would explore the agricultural resources and were located in small or medium size properties. We may include in this category the Fontanela site (ID 3030104) or the Barreiro site (ID 3134204), in Vila Verde. The remains found in Monte do Ouro, although very close to a Villa located in ‘Quinta de Stº Andre’ (ID 3030101), may have corresponded to a later occupation. It dates back to medieval times, making it rather difficult to assess if this infrastructure was originated from an older one of Roman chronology. The natural attraction exerted by the road network system in structuring agricultural exploitation units is well known. They could have been Villas, aedificiae, or small Farms and follows the general trend observed for the concentrated settlement existingin in the area, a fact that is not surprising given the intense road network across regions with more fertile soils.The main challenge here is linked with the identification of these infrastructures, The question that arises in these cases, is whether we are facing a agricultural holding or if the areas of found remains may correspond to any other type of equipment whose function was articulated with the circulation of goods and people. In this case the calculation of distances needed to provide 100

We may also admit an infrastructure located nearby a road may have had agricultural functions or added that function to their role of circulation support, which in turn would have demanded the existence of smaller service areas, such as a tabernae or caupona. This could, perhaps have been the case of Casas Novas (ID 3034604), a site located nearby the Via XIX itinerary and relatively close to Bracara Augusta.

3 Peripheral limits and scales

The conducted assessment of entities and of some relationships shaping the settlement patterns in the surrounding region of Bracara Augusta led us to the conclusion that the city did not have a homogenous impact upon the rural hinterland. With a particular dynamic and far away from mere consumption, the relationships between the city and the rural surroundings seemed to encompass a strong interaction 46 . However, the understanding of the overall phenomenon performed a difficult task. 46 The debate around the relationship between the city and the rural world goes back to Max Weber (1891) and M. Finley (1977). In 1990, C. R. Whittaker conducted a critical revision of Finley’s thesis and its main opponents with regards to the role Roman cities sustained with their surrounding rural space. The author included in this debate the importance of vici. M. Finley’s (1977) thesis, following on other authors footsteps, defended the idea of a consumer/parasite city and highlighted some key points: I) the ancient city, an opposite of the medieval city, did not sustain separate functions between the city and the surrounding area; II) the ancient city was depend on the countryside; III) the income from the city consumers came from the countryside and not from commercial activities. Some authors thoroughly disagreed. G. Pucci (1983) and J. P. Morel (1981) defended the idea that the main changes in the thin ceramic production industry matched stages of great development in cities, suggesting the existence of ‘urban industries’ serving and supporting the rural world. In K. Hopkins’s (1980) opinion, cities had vital roles in collecting taxes; they paid the rural producers in real money so they could pay the taxes in return. The money from the city inhabitants come from trading and exporting manufactured goods, which means that both the trading industry and artisanal production paid for the city consumer needs. Ph. Leveau (1983; 1984; Leveau and Goudineau 1984) and J. Wacher (1974) amongst other researchers developed the thesis defending that the city must be understood as the rural space ‘organizer’ due to the need to organize and redistribute agricultural products. Ph. Leveau did not agree on the ‘parasite’ status of cities as the concept implied decadence. In C. R. Whittaker’s opinion, Ph Leveau was not accurate in associating Finley’s city consumption needs with the idea of decadence. The idea is that the organizing city in economical terms is behind the idea that the city would pay for its own consumption needs with trading surpluses. For this author, the most important thing in Leveau’s theory was the intimate relationship between the city and the countryside villas. For Whittaker, this idea does not contradict Finley’s thesis with regards to the importance of rural rents in organizing the city. The rural landlords were considered

The Shaping of Bracara Augusta Peripheries

What we can suggest is to focus on the cartographic analysis of the concentration and dispersion levels of archaeological remains. Moreover, we will attempt to understand the existence of different articulation areas within the suburban region. Furthermore, we will try to study land division features of the Roman rural world and present the first research results over the centurion of the cadastre.

3.1 Settlement concentration and dispersion areas

When analysing the city surroundings, Penelope Goodman (2006) determined that the concept suburbanum used in Roman times defined the surroundings reality of the city, encompassing, at the same time, a variety of events developing in space in a heterogeneous manner, in a degree that can only be fully understood case by case. The concept included the difference between the city, the suburbium and the rural space, and as the author concluded it was diffuse in a gradual scale of the components determining the human landscape. The cartography of archaeological remains around the city, located within a perimeter of 19,5 km N/S by 23 km E/W enabled the understanding over the structuring and articulation of populated areas (Map 11). First and foremost, there was a clear concentration of remains occupying a fertile and flattened area covering a region of 5 km around the city. Within this area, the northern part of the city presented a wider occupational diversity matching the mountainous area of Montariol on the east and Gondizalves on the west. This wider concentration of remains was also extended, although in a less intense manner, towards the south and across the area entitled Veiga do Penso up to Lomar and Esporões region. It is evident that the Cávado River crossing this region does not represent any boundary in terms of significant changes within the populated area. This is in line with the assumption that many Farms currently located on the river left margin naturally extend themselves towards the north. The northern area of the Cávado River housed a group of remains that can be explained by the proximity in relation to the roads itineraries, both main and secondary ones, and also in terms of the agricultural lands existing nearby the river or in the areas between Prado and Cabanelas. an urban elite, who supported several of the urbs needs. In sum, these are the key ideas behind the theoretical approach to the relationship between cities with the rural world. What is key to highlight with regards to the relationship of Bracara Augusta with its rural surroundings is that one must place the debate within the archaeological reality in question. Under these circumstances, if the idea of the consumption/parasite city is suited to some Roman cities, this does not seem to be the case with the Roman city of Braga. Moreover, Ph. Leveau’s thesis revealed an intimate relationship between cities and villas ad this seemed to be certain to have occurred in the Caesarea territory, in Mauritania. If the organizing criteria was linked with the concentration of villas around the city, thus establishing the influential boundaries and rural space division at the very beginning of the ‘empty surroundings’, the similarities between the relationships of the Caesarea in Mauritania with its countryside and the city of Bracara Augusta and its rural surroundings seem to be wider than their differences.

The northern region of the Cávado River, performing a small percentage of area under study, was crossed by the Via XVIII and the Via XIX itineraries alongside other secondary roads that were clearly important within the overall articulation of the region. This is followed by a gradual decrease of populated areas located between 5 and 8 km from the city, in areas bounded by significant reliefs but with some Hill-forts from the Iron Age. Some of these sites are located nearby road circuits linking the region with external areas in several directions. It is believed that they may have functioned as strategic points for support and surveillance. With the exception of the area bounding the Cávado River on the north side, this area is similar to all other areas and bounded the region presenting habitats not fully dependent on the city. We highlight three main areas: on the west, on the northwest and on the southeast. The rural area west of the city will be analysed in detail in chapter 5. We will draw attention to the intense setllement that is visible on this part of the countryside. The passage of a secondary road linking Bracara Augusta to the coastline may help to explain the location of some of the habitats, however the most impressive image refers to the its dispersion and wide number. The explanation behind these events should be sought within the regional opportunities existing in a region offering an open valley and wide areas presenting agricultural and artisanal resources. The northeast area of the city presents a settlement pattern that is visible on the north side of the Eiras Velhas Hillfort and of the outline of the Via XVII itinerary. Despite its proximity to the city, this area was not part of what we believe was Bracara Augusta surrounding territory. Instead, we are dealing with a encased valley area presenting a sparsely settlement housing some villas and registering the presence of the Hill-fort of Póvoa de Lanhoso. On the east and southeast areas, there was a great density of settlements occupying a relief area providing passage to the Cávado and Ave rivers valleys. The significant differences in terms of size and chronology of this wide group of Hill-forts including the Citânia de Briteiros are relevant. The road linking Bracara to Emerita Augusta via Tongobriga crossed a valley housing several resources namely thermal sources. The explanation for this settlement cluster may also be justified by the probable dispersion towards the valley area of population coming from the regional Hill-forts. According to available data, it seems reasonable to focus again on the region under study, however we should now only pay attention to a more restrict part, thus enabling the assessment of other issues related with the rural settlement around Bracara Augusta. The city surrounding space corresponded to an area of around 3 km that up to recent years seemed to be vacant in terms of Roman occupation. However, this perception has systematically been questioned due to the intense construction works that have been carried out in the area throughout the past decades and were demanding of archaeological surveys. The idea of an ‘empty’ occupation

101

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

circle around the peryurban area of the Roman city is shifting towards the conviction that it registered significant occupation levels, attested by the findings of tuguria, viae vicinales and viae privatae (Lemos 2002: 109-122), alongside with infrastructures believed to had been Villas.

altar associated with pavement remains indicating the presence of a fanum in the Via XIX itinerary.

Among the Villas believed to have existed in the outskirts of Bracara Augusta, we refer the Coturela (ID 3034603) and the Infias (ID 3035201) cases, both located around 1 km from the city. An intricate network of communication circuits enabled the link of these sites among themselves and their articulation with others located both in the rural area and within the city itself. The archaeological identification of some patsways and other small rural sites, hardly detected in surface surveys became of paramount importance. They have been detected as a result of excavations conducted in 2001 by the Archaeology Unit of Minho University, when building works have been carried out to open a secondary road of EN14 national road. These works led to the identification of three viae in the southwest of the city’s train station and of the ‘Cerca do Mosteiro da Visitação’. These viae were built in the 1st century and were likely operating up to the 5 th/6th centuries. A roadway was found in Capitão Alberto de Matos Street leading to, according to Francisco Sande Lemos, other sites settled in the Monte de Cones, Semelhe and Senhor dos Lírios (Lemos 2002). Between Estação Street and António Macedo Avenue in an area corresponding to the oldMosteiro da Visitação two other viae should also be highlighted, presenting a NW/SE orientation with around 4/4,5 meters width alongside another one with a perpendicular orientation. The via running NW/SE may have led, according to the author, to the Coturela Villa and to S. Frutuoso de Montélios (Lemos 2002: 118-122). Conducted excavations identified a vast chronological framework for this viae, and the material remains uncovered and dating back to the beginning of the first century is rather relevant. From the archeological materials found from this period, there were indigenous ceramics, italic sigilatta fragments (late first century BC and early first century AD), sud-Gallic sigillata fragments from the first century, several Haltern 70 amphorae and fragments of amphorae dating back to the High Empire period (Lemos 2002: 109-110; 112-115). The archaeological works also shed some light on a group of artefacts interpreted by the same author as tuguria remains. We are dealing with pits that may correspond to huts remains made of perishable materials and structures located alongside of the viae. Remains uncovered inside the pits generally date back to the first century (Lemos 2002). This led to the conclusion that these types of establishments had early chronologies. The image of the outskirts of the city has gradually been improved and further completed by monuments findings such as sanctuaries or rural temples, namely the sanctuary/fountain of Fonte do Ídolo (Elena Garrido et alii 2007) located nearby the Via XVII itinerary and an votive 102

Figure 10. Votive altar found in the outskirts of Bracara Augusta (Carvalho et alii 2006) Within the immediate outskirts, an extension corresponding to an area located around 1,5 km to 3 km from Bracara Augusta, we can find Dume (associated with the Carquemije necropolis), S. Frutuoso de Montélios, Semelhe, Lomar and Monte de Cones sites. As previously observed, the concentration of remains around Bracara Augusta took place in an area comprising 5 to 7 km. The geomorphological features of this area helped to explain this cluster of settlements. This is a rather slightly hilly area, with rich agricultural soils. Here we may identify the archaeological sites of Pateira – S. Pedro de Merelim, Outeiro of São Sebastião, S. Miguel de Gualtar, Stº Estevão, Quinta de Stº André and Espinho church. Within the 5 km area starting in the city, one can also find some Hill-forts. Some of them did not registered Roman occupation remains and their territories seemed to have been integrated within the ager occupied by new types of infrastructures, as it was the case of the Castro da Consolação Hill-fort, likely abandoned in the late first century BC. There are three further hill-forts presenting Romanization traces and they represent a ‘counterfort’ of Hill-forts with occupation up to Roman times. We refer to Monte da Falperra, Monte das Caldas in Gondizalves and Monte Redondo in Guizande occupying the higher hills around the city. On the other hand, we verified the abandonment of some Hill-forts located on the north side of Bracara Augusta. This was the case of Monte Castro and Pedroso located in the lower regions. When assessing the abandonment, retraction or maintenance of the regional Hill-forts, we must take into consideration several variables. If we apply to these sites the analysis model of potential exploitation areas – known as Site Catchment Analysis – we verify the maintenance of Hill-forts with territories of exploitation of half an hour.47 47 The analysis of territories or habiats exploitation areas applied to archaeology in the 1970s by Vita Finzi and Higgs (Vita Finzi and Higgs 1970: 1-37, 1972: 27-36, Vita Finzi 1975: 225-231; 1978) was

The Shaping of Bracara Augusta Peripheries

It is precisely in the surrounding of these Hill-forts that new Roman sites were established, as it was the case with the main Villas. This can be justified by the clear preference over valley lands to settle this type of rural infrastructures. Thus, the inherent logic in pre-Roman land exploitation was, of course, reassembled or remade alongside the Roman presence. In this particular case, it was shifted due to the presence of a Roman city in the surroundings and due to the need to regain lands in order to established new agricultural exploitation practices.

3.2 Rural land division

The identification of a series of alignments that seemed to extend the Roman city axes in the outskirts of the city has led to the belief over the existence of a possible Roman rural cadastre (Martins 1995). This suggestion demanded prospections to be conducted in order to verify the traces of potential modulating landscape factors in the city countryside.48 Inside walls, the city of Bracara Augusta reached 48 hectares in the Low Empire period. The confirmation of the cadastral organization of its rural territory helped explain the development mechanisms of the urbs itself, as they were likely connected to the formation of terratenants elite exploiting the agricultural resources in the region. It is acknowledged that Roman operations for measuring and calculating surfaces and properties took place in several sites of the Empire. They had legal aims and supported ‘un contrôle objectif des surfaces’ (Moatti 1993: 94). However, we still need to understand if there was surfaces control or control over land boundaries and their consequent land distribution. This control may have occurred under several circumstances, and we highlight two ways they could have taken place: through the simple measurement and delimitation of property perimeters with taxation aims in a procedure entitled ager per extremitatem mensura comprehensus (Ariño Gil et alii 2004:178181; Ariño Gil 2005); and depending on the imposition of cadastral land division logics within the landscape, thus organizing the rural area and the control of surfaces,

based upon a simple assumption and was further developed into other scientific works by Von Thunen and Chistaller: the idea that communities minimized costs and maximized profits abling the generation of models of ‘minimum costs’. When applied to archaeology, this idea encompassed an acceptance of the theory that communities detained resources within provisioning areas located at a certain distance from their living habitats. Thus, we may define potential exploitation areas and influential territories. This same model was applied by Manuela Martins (1990: 211-216) to sites located in the medium Cávado river basin. Alessandro Launaro conducted a reassessment and updated the knowledge on the subject (2004). 48 Rui Morais started this work. We thank the author the first publication on the theme.

103

alongside the delimitation and distribution of lands, structuring therefore, a cadastre. The need to fully understand these criteria led us towards diagnostic research approaches in an area that was, at the beginning, difficult to study in archaeo-morphological terms, as it is the region of Braga. Archaeological research focused in the relationships between cadastral shapes, the territory and the settlement has been exponentially growing in the last few decades, as a consequence of the acceleration of rhythms and intervention mechanisms within contemporary landscapes (Favory et alii 1982; Chouquer and Favory 1991; Clavel Levêque et alii 1994; Clavel Levêque and Vignot 1994; Chouquer 1996, 1997, 2000). Nevertheless, research focused on landscape historical morphology has a complex nature and interferes with the landscape ability to be fully understood due to its long sedimentation processes. It is the end result of complex human occupational stages and the overall dynamics of natural factors, but also of the processes erasing the marks of that sedimentation throughout time. The analyses of landscape evolution following a diachronic perspective performed an adequate methodology for any arqueo-morphological research, although it inevitably implies the existence of wide and interdisciplinary teams. On the other hand, it is reasonable to state that the improvement in Roman rural landscape studies 49 de49 Roman agricultural sources date back to the second century BC and first century AD. The most important written sources were the works by Catão (234-149BC), Varrão (116-26BC) and Columeia (circa 60AC). Catão’s work – De agricultura – is perhaps the most important literary source with regards to the agricultural features at the Roman Republic. Other than these authors, there was a collection that organized treaties from several authors from different times in a compilation known as the Corpus Agrimensorium produced in the fourth century AD. Between 1848 and 1852, these texts were organized under the generic title of Gromatici Veteres and included authors like Frontino, Agennius Vrbicvs, Balbvs, Higino Gromaticus and Siculus Flaccus. Other than this, there is another key document entitled Liber Coloniarum listing the name of cities and their legal status, the variety of agricultural Roman exploitations and the necessary jurisprudentia on behalf of technicians in defining and assessing the land. The relationship between the legal status of a territory and the required exploitation techniques was also approached by the ancient authors. Frontino defined three types of lands: agri divisi et adsignati; mensura per extremitatem comprehensi e arcifinii qui nulla mensura continentur (Castagnoli 1958: 31). The differences in terms of division were rather clear. The mensura per extremitatem comprehensi was the measure system applied to a civitas land or to the goods of a priestly college. The third case did not entail any measuring system and the boundaries were defined by geography. The territory that interest us is the agri divisi et adsignati that was traditionally attributed to cities holding a colonial status according to F. Castagnoll on the topic ‘"L'ager divisus adsignatus è proprio delle colonie come si ha da Frontino, 1". The author provided the division elements including two variables: ager limitatus subdivided by limits or ager per striges et per scamnas divisus et adsignatus. With regards to the division in non-colonial lands, the author added: “Ma oltre che nelle colonie si ha divisione e assegnazone nei municipii, praefecturae, fora, conciliabula. E inoltre nelle distribuizone viritane: il termine che ricorre in tal caso è iugeratio, adsignatio in iugeribus. E, ancora, negli agri quaestorii; in essi si hanno limites (che vengono detti anche decumani) che racchiudono laterculi di 50 iugeri, come abbiamo da Siculo Flacco (...) ma (...) nelle province, secondo Frontino (...) sono divisi per strigas et scamna. (...). Tra i moderni si ritiene la strigatio propria dei terreni non assegnati

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

manded, within the past few years, a critical overview over some evolutionary perspectives. Thus, new criteria and technical approaches have been emerging50. The need to overcome the traditional perspective strongly centred on cadastral surveys and their relationship with the rural habitats and the urban fabric is behind the current trend of developing a diachronic history of the land division morphology. Advances in this type of research and their results should be interpreted as a preliminary approach to structure a future research agenda.

3.2.1 Analysis methodologies

The conducted work respected two distinct phases. On a first phase, our aim was to exclusively understand the existence of ‘coherent alignments’ following the orientation provided by the Bracara Augusta cadastral axes in a region measuring 11,5 km N/S by 7 km E/W. Furthermore, we aimed at diagnosing the potentialities of the territory under study and list any challenges and methodologies to adopt in future works. The choice over the area to analyse was linked with two basic factors: first and foremost, this is the rural region widely considered to be directly influencing the city, appearing to us as an area suggestive enough in terms of length and of geomorphological conditions needed to perform our assessment study. Thus, we articulated aerial photography with a Geographic Information System, systematically researching all the existing aligments in the current rural land division with orientations coinciding with the detected axes of the Bracara Augusta urban perimeter (Martins et alii 1994). The aerial coverage utilized was the 1947 flight by military services in a 1:30000 scale. After the photograms collage and the consequent treatment work conducted in the Adobe Photoshop software, the photo interpretation

work brought to light the first significant alignments. 51 This data was later exported into the Geographic Information System (GISRAM) already described on the methodology chapter (Bernardes e Carvalho 1996) 52. This GIS incorporated all the data that was interpreted. The results enabled us to understand the existence of a significant group of alignments mainly located on the eastern and northern areas of the Roman city. The first diagnostic results led us to the following phase, as the studied area grew in interest although the base information was manifestly scarce. On a second phase, we enlarged the photographic and cartographic information supports. We also incorporated the results of the systematic field surveys of part of the territory under study and as possible we have also utilized available historic sources for this region, namely from medieval and modern times namely tumbles and scriptures produced by the ‘Cabido da Sé de Braga’. Moreover, we conducted further photo and cartographical interpretation. The photographic material used was at this stage aerial photography in a 1:18000 scale produced in 1938 (SPLA)53 due to its potential in terms of rural landscape conservation. Although some problems emerged in terms of image quality, the group of mentioned photograms represented the first aerial coverage for this region and enabled a type of landscape visualization impossible to achieve in more recent coverage photography. Those photograms pictured the city in the 1940s and reduced it to a single mark when compared with the current city, signalling its rural framework, where, despite the already dense settlement, remained still ample areas of cultivated fields. To the layers corresponding to these photograms we added 1:25000 cartographic analysis (sheets 56 and 70) and cadastral surveys with a 1:5000 scale, from 1968. The utilized 1:25000 cartography was produced by the military cartographic services at the time of their field 51 The followed methodology was presented in detail at the II Peninsular Archaeology Congress, in a Workshop dedicated to the application of GIS in Iberia peninsula archaeology practice alongside a poster displayed at the congress (Carvalho and Bernardes 2000:125-131 and Carvalho 2000_143-146). 52 The collage and handling of the photograms 5012 and 5160 was

optimo iure e dunque soggetti a imposte, e ciò per il fatto che la strigatio segna in singoli possedimenti (al contrario della centuriazone nella quale i confini delle centurie no corrispondono ai confini delle proprietà), evendo lo Stato interesse a delimitare con chiarezza il catasto dei terreni soggetti a imposte. (...)" (Castagnoli 1958: 31). 50 The regular chronicles by Monique Clavel-Levêque included in

executed with Adobe Photoshop 4.0. The GIS drawing was produced in Map GrafixPPC 3.5 and associated with an alphanumerical Database in 4th dimension.

the ‘Dialogues d’Histoire Ancienne’ journal enabled us to accompany the enrichment and improvement in theoretical complexity of cadastral studies. The journal is available at www.persee.fr. The Atlas Historique des Cadastres et des Mines d’Europe database performs a growing data source and may be accessed on the Institut des Sciences et des Techniques d’Antiquité site: http://ista.univ/fcomte.fr/bdd_images /index.php. Available literature on the topic is vast and there is a considerable number of sites publishing results on the topic. The preliminary nature of the results presented on this work together with the wide variety of available literature led us to mainly use the footnote-referencing system when aiming to highlight a point. We will not include several existing works, as we understand it is more important to highlight the results rather than the literature.

53 The first aerial coverage was conducted by a company named “Serviço Português de Levantamentos Aéreos” (SPLA), and its assets were acquired by the Serviços Cartográficos do Exército (SCE). The only information made available by the SCE was the fact that this aerial coverage was conducted between 1938 and 1948. The second aerial coverage was conducted in 1947 by the Royal AirForce (RAF), and had a 1:30000 scale. The comparison of some axis from the Roman city between both aerial coverages enabled us to verify that some photograms registering a 1:18000 scale were prior to the aerial coverage conducted in 1947, however the imprecision subsisted and we should be aware that when referring the 1938 coverage we are mentioning a date that must have been prior to 1947 and posterior to 1938.

104

The Shaping of Bracara Augusta Peripheries

work in 1948. They were elaborated from systematic first flights within the national territory. This was an intentional choice. In fact, this cartography was, in terms of archaeological research, an added value as it represented a substantially minor changed landscape when comparing to more recent maps. This was visible in the network of transport circuits, in human occupation areas and toponymic features that were substantially more complex in surveys conducted in 1948. The respective local authorities produced the used 1:5000 cartography for the Braga and Vila Verde municipalities. Although out of date, these surveys enabled us to generate an overview that was rather precise of the cadastre area, also providing a substantial number of toponyms and micro toponyns that were naturally not present in the Military Maps of Portugal with a 1:25000 scale. Based upon this group of data, we began to systematically assess the area under study. The base information was transferred into a digital format and systematically handled in order to gather a wider number of information with regards to the rural land division. At this stage, we utilized the AutoCAD 2007 software 54 that enabled the organization of several data layers and systematically merging the rural land division alignments. The software was later updated for the ARQGIS 9.2 system55 and the results we present here performed the first GIS migration together with tasks developed in AutoCAD. For the work presentation, we added subtitles in the cadastre hypothetical grid with a system of numbers for the N/S axes and letters for the ones running E/W (Map 12). When performing these tasks, we understood the need to conduct a systematic prospections for data control, as suggested by photo interpretation concentrated in the northern region of Braga in a well defined transection corresponding to the intersection of the 12, 13 and 14 axes with the Q to W ones, transection that included the Via XIX itinerary between Braga and the Cávado River.56 This option respected a series of conditions, namely: the topographic and geomorphological features of the region that make it plausible for an intensive prospection focused on methodologies known as ‘field walking surveys’; the great density of already known Roman remains with particular emphasis towards the archaeological area of Dume (Fontes 1993); the need to assess in detail the Via XIX itinerary in this area in particular; the fact the region performed an excellent rehearsal area in terms of research

54 The AUTO CAD work was produced by Joaquim Folhento, Alfredo Barbosa and Eurico Loureiro. 55 The migration work to ARQGIS 9.2 has been developed with the

related with landscape transformations between Roman times and the Modern Ages. Other than this transection we conducted field surveys in other areas following methodologies depending on the effective conditions of the soils and the different aims we had. In some areas, we applied an intensive prospection methodology, such as recollection, record and identification of collected materials (‘field walking survey’). In other cases, prospections focused upon the identification, photography and record of data, conservation state, and delimitation features of the axes and current use of soils. These field walking surveys took place around the S. Frutuoso de Montélios area (axis Q to R), in Monte de Cones and in the southern part of the city (axis O and N 13 to 15). The choice was related with the need to understand the detected alignments and with the possibility that all remains had been destructed due to intense urbanization of the area has been under. These events are particularly pertinent around the S. Frutuoso de Montélios area.

3.2.2 The area around the XIX itinerary, between Braga and the Cávado River

Field walking surveys conducted in the transection mentioned above (axes 12, 13 and 14 with Q to W) began from the need to assess the possibility of alternative traces with regards to the ones being mapped or described by some researchers for the outline of the Via XIX itinerary, mainly in terms of the area between Bracara Augusta and the Cávado river. This alternative followed a proposition put forward by Luís Fontes (2005). Based upon these assumptions alongside with the results from works conducted by us, we believe it is possible that the Via XIX itinerary exited the city via the north part, namely through the current Convento do Pópulo area.57 It should be notes some of the arguments sustaining the hypothesis of an exit via Cónega street (Almeida 1968) or even by a western city axis (Almeida 1979). The western exit possibility has been defended by Carlos Alberto Brochado de Almeida, who referred to its outline as ‘leaving the Maximinus door, the road followed the circuit of what is today the train station and went through Semelhe via Monte Barral and the Tourido road in S. Jerónimo de Real municipality (…)’ (Almeida 1979).58 This proposition not only suggested a western exit but also a shift on the road further west via Monte de Cones and the Tourido road. This does not seem accurate to us. 57 D. Frei Agostinho de Jesus was the archbishop of Braga between 1587 and 1609. In 1595 he donated city terrains to Stª Agostinho barefoot hermit’s monks destined to the construction of the Pópulo monastery. The construction started in 1596. 58 There are references to the finding of a ‘anepigraphed’ milestone

support of Helena Calado and António Moniz from the Geography Department of the Açores University and Eurico Loureiro from the Archaeology Unit of the Minho University, to whom we thank. 56 We utilized satellite photography in the works focused on the

in Tourido. We were able to observe the structure and we have doubts about its typology. We were inclined to believe the assumption put forward by Luis Fontes about the possibility of its being a half column used to bound the limits of parishes (Fontes 1993).

analysis of the road network system and the structuring of the land division. Furthermore, we used the same data as supporting material in conducted prospection works.

105

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

It is possible that part of this outline matched a secondary circuit, which may be justified by the need to link the city with the Semelhe villa or with the important Monte de Cones archaeological site. However, it does not correspond to the Via XIX itinerary. The road exit by Cónega and Boavista streets, going through Real and up to the Cávado through the current connection with Prado also defended by Carlos Alberto Ferreira de Almeida (1968:34) raised some questions. According to the oldest image representing the area - the Braunio Map of 1594,59 this exit only existed in the late 1600s. In fact, this map does not contain any detail of what correspond to the Cónega Street and only referred the northern exit, the same we defend as being the Via XIX itinerary. Under these circumstances, there are no elements suggesting construction works in the Boavista Street (Cónega) before the 16th century. References to Cónega in medieval documentation enabled the identification of a wide Farm or walled field surrounded by a road. The construction of the Campo da Cónega took place after the land division between ‘Mesa do Cabido’ and ‘Mesa da Mitra’ in the mid twelfth century. It seemed to have been a well-structured property between S. Pedro de Maximinos, Monte de Cones and the corner of the old Porta Nova area. The wall outline and the sidewalks encompassed two directions: towards Monte de Cones and towards S. Pedro de Maximinos. In 1389, at a time when a payment was due in the Lavandeira area, there is a reference to this location including a wall between Cónega and Monte de Cones (Nogueira 1998:94). In 1493, in a water related issue including the S. Pedro de Maximinos abbot the following was described ‘çerca da dicta egreja perenall de que corre agua e logo entra em huum campo’ (ADB, 2ª Gaveta das Propriedades do Cabido, doc. 50). In documentation from the Cabido properties dating back to the late fifteenth century, a Cónega payment was referred with regards to another one in Cones 60. The latter was bounded by a large wall made of stone that was in turn connected with the Cónega payment. At the far end, there was a sidewalk linking Cones with Cónega, as it was also referred: ’Item em Coones esta hua leira de devessa (…) que he toda de castanho e de carvalho (…) E parte a dita leira per hua parede de pedra que he grande e boa parede que vay antre ella e a devessa da Coonega que he do cabiido E vay a dita parede de fundo acima tanto quanto vay a dita leira E a dita leira jaz acarom (?) da dita 59 This illustration mentioned in the Braunio’s Map represented the city at the end of the seventeenth century and was part of one of the known atlas in Renaissance cartography – the Urban Atlas of Georg Graun and Franz Hogenberg, published in Cologne under the name of Civitates Orbis Terrarum. 60 This document was particularly detailed in the properties bounding elements. See Cunha 2004.

106

parede E da outra parte esta outra leira de devessa que foy de Lopo Fferrnandez do Penedo E antre hua e a outra esta huum valado por estrema por onde parte E da parte debaixo vay entestar na Estrada que vay de Coones pera a Conega em hua parede que esta antre ella e a dita Estrada E da parte de cima entesta em hua rriba que esta antre ellas e huum campo que he da Conega que tambem he do ditto cabiido (sic)’ (ADB, Tombo 4º do Cabido, fl. 52v.). In mid eighteenth century, the ‘Índices dos Prazos’ were developed for urban (houses) and rural (Farms) properties of the Cabido, in an operation meant to verify and manage the institution documental archive. The result was presented in the Braga Street Map (1989-1991). The map listed every street alongside the features of each house, with a thorough surveying of every façade. The indexes registered a summary that forwarded to the ‘Livros de Prazos’ existing since late 1500s the gradual deadline regeneration works for each property.61 In the urban records (the houses), the Cabido only included documentation about the Cónega street from 1607 onwards (ADB, Índice das Casas do Cabido, t. IV, fls. 267-440). It is in the rural index that one can find documents related with Cónega describing it as a Farm (ADB, Índice dos Casais do Cabido, fls. 442-442v). Later cartography, namely the Braga Map by André Soares produced in 1756, the Cónega Street was already represented and we verified the disappearance of the northern aisle that we believed matched the Via XIX itinerary fossilization. The question is then related with the ability to understand the on-going changes explaining the ‘deactivation’ of the Via XIX itinerary axis favouring the one that ended up becoming the Cónega Street and the overall layout of the current street connecting to Prado. In 1595, the donation of the future Pópulo convent terrains to the monks of Stº Agostinho conducted by Braga’s archbishop, D. Frei Agostinho de Jesus, may have activated an overall city development towards the north and was likely demanding of further urban morphological alterations in the centuries to come.

61 If we consider the group record – in the form of a Cartulary – of the properties in Braga since its regeneration at the end of the sixteenth century, under the order of the D. Pedro bishopery, we may think that the urban and rural land division was done in the long-term (Amaral 1990; 2007). We believe that future studies should privilege a sectorial study from the twelfth century up to the nineteenth century in a research presenting a long periodization. Master works in Medieval History unveiled the Cabido heritage through the study of certain circumstances cross-referenced with medieval documentation and the farms deadline books available since the sixteenth century until the nineteenth century. Similarly, these were cross-referenced with territorial archaeology. Thus, we believe it is possible to perform a geo-referenced reconstitution of the Braga land division, similarly to what Maria do Carmo Ribeiro (2008) began with the morphological study of Braga since the Roman times up to the Modern Age.

The Shaping of Bracara Augusta Peripheries

Thus, the construction of the Pópulo Convent initiated in 1596 included part of the old Via XIX itinerary and there was an appropriation of the square crossed by this itinerary and visible in the Braunio’s Map produced in 1594. The convent construction closed the square and the western area of the current Campo da Vinha. This alteration prompted the gradual deactivation of the trajectory matching to the old Via XIX itinerary and generated the possibility for a new road morphology configuration. These events were later conducted with the transformation of the ‘carreira da Cónega’ in the Cónega Street (Cfr. Braunio’s Map with André Soares Map, of 1756). With regards to the urbanization of the Cónega Street, the available and most detailed document is the ‘Braga Map Street’ (MRB) produced in 1750. This performed a document where that particular road axis was mentioned as being the Cónegas Street. The referred Map also mentioned 93 houses located on the northern site of the street and 57 houses on the south end, being all of these Cabido properties. The first core, nearby the city centre, referred house number 1 with a main façade facing Campo da Vinha and matching the house number 2 with a façade facing the Biscainhos Street and the C. Sottomayor house, an aristocrat from the Royal House.62 With regards to the topography, the steeply sloping of this axis was mentioned. In mid eighteenth century, part of the street was still underurbanized. In fact, the referred Map represented wide fields in this area, similarly to what happened with house number 67 facing the western part the number 21 field located on the Real Street (MRB: II 166-167). Some important details should also be mentioned in order to understand the road network in this space. The old Roman road was entitled at this time as ‘Cangosta do Pópulo’, and this is based upon a reference made to the house number 171, mentioned as facing the east side of the ‘Cangosta do Pópulo’. The ‘Braga’s Street Map’ and still with regards to the Cónega Street, also mentioned a sidewalk running ‘from Couturela to the S. Frutuoso convent’, passing through houses number 119 and 120 (MRB: II 166-167). This sidewalk likely matched a connecting axis of Roman origin between the city and the Couturela and S. Frutuoso Villas, boosting an outline kept throughout the Middle Ages. This layout was suggested by later cartography. It is believed that this sidewalk was the ‘caminhos dos frades’ (friar’s pathway) starting in S. Frutuoso de Montélios and it is still possible to walk this route today, or at least, part of it (the part non affected by recent urbanization regeneration processes).

62 The Casa dos Sottomayor was constructed at the end of the sixteenth century or beginning of the seventeenth century, according to Braga’s Street Map. It included the opening of the Cónega Street.

The construction of the Pópulo convent in Modern Times generated a new stage in the overall Braga urbanization process. Under these circumstances, we believe that the Cabido properties were behind the need to invest in a ‘bracarensis’ church elite. From the exit of one of the main Roman city streets, the Via XIX itinerary would then lead towards the north. The presence of graves within the Pópulo convent adjacent area was uncovered in archaeological prospection conducted by the Archaeology Unit (Martins and Delgado 1989-90). Two Milestones were found inside the Patronato and Braga Local Authority building (Fontes 1993), which seemed to sustain the argument favouring the outline proposed by us. This exit implied overcoming the steep slope between Braga and the valley and going through the current neighbourhood of S. Martinho. This would favour a less controversial circulation, one that was more suited for the existing topographic constraints, instead of going through the Cónega Street. Thus, the current S. Martinho Street, located on the west part of the Pópulo convent seemed to perform the closest path for the old Roman road, although the original outline partially incorporated in the Convent terrains should also be taken into consideration. Moreover, the conserved and visible alignments in aerial photography of the adjacent terrains of the monastery, together with the analysis of Braunio’s Map put forward an outline running alongside part of the space of the mentioned convent. This circuit would then follow the street accessing Dume and would turn left reaching S. Lorenço da Ordem chapel, referenced in some cartography as being S. Gerónimo chapel built nearby the known crossroad of the ‘quatro caminhos’. On the other hand, the fact that the boundary of the Sé borough was precisely limited by the Via XIX layout, including the Cónega Street and ending on the ‘quatro caminhos’ crossroad is also worth considering. This boundary has lasted until our days. From the S. Lourenço chapel, established in the eleventh / twelfth centuries, the Via XIX itinerary would cross Gafos watercourse. This latter reference likely got its name from the old Gafaria de S. Lourencço, from which the chapel would have been integrated in (Marques 1989). The ‘quatro caminhos’ crossroad was, until a few years ago, paved as observed and recorded by researchers from the Archaeology Unit, although urban transformation processes within the Montélios area destructed that pavement.63 Moreover, the pavement and the size of the area designated by Calçada da Ordem corresponded to the outline of the Via XIX itinerary as accounted in the field report produced by the Archaeology Unit in October 1987. The report stated that part of the circuit located between the S. Lourenço chapel and Gafos watercourse was un63 We thank Luís Fontes all the provided information and suggestions with regards to the outline of the Via XIX itinerary.

107

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

covered in an extension of around 500 meters due to intense rain. At the time, it was still possible to witness part of the road nearby Carcavelos, in Dume in a total extension of 850 meters. With around 2,20 meters medium width, the pavement made by perfectly levelled thick and wide granite slabs was adjusted and polished by their use. The circuit included supporting walls and properties land divisions at an inferior level in relation with the surrounding fields.64 After passing the Gafos watercourse, the road seemed to persist as a property division up to the area where it was cut by a recent street accessing the municipal stadium. In this square it was found, according to oral accounts, one Milestone later transported to Braga (Fontes 1993: 44). From the Carcavelos square, where Nossa Srª das Angústias chapel was located, the road followed north, probably crossing the interior of the current Quinta do Cordeiro or Brás Pereira resuming the current outline a few meters further north. The road would then cross the Sobremoure bridge (or Sobremor), that presents only one arch of rudimentary finishing and has a chronology difficult to identify due to the lack of clear features identifying a precise historical period (Fontes 1993:45). From here, the circuit would then follow a crossroad near Felgueira, following to east where one could find a cippus likely used as a boundary in the Roman land division (cfr. Point 3.3.3) Aerial photography observation led us to suggest a shift in the outline slightly to the west, in the terrain surrounding the current pathway (Carvalho 2008). From the current Felgueira crossroad site, the Roman road would follow up to the well-known Caixa de Águas site and from there to Castro Mau. The significant changes prompted by the construction of the Palmeira airfield infrastructure made part of the trajectory rather difficult to identify and/or confirm, as witnessed by 1:25000 cartography. From Castro Mau the road trajectory continued north following the current St. Estevão, Formoso and Cangosta da Cabra sites up to the Cávado River. The circuit could have included a boat crossing or a potential bridge likely located upstream. The wading crossing was viable and a regular practice until a few years ago, before the process of sand extraction started, affecting the layout and river flow. The river beaches constructed meanwhile also prevented the wading crossing of Cávado, a standard practice by both margins inhabitants’, a fact partially confirmed by the existing rural properties were not bounded by the river. Oral testimonies recollected in the area, indicated that up to the standard use of automobiles and public transportation, Dume, S. Pedro and S. Paio de Merelim’s inhabitants would use the ‘estrada velha’ (old road) to travel 64 This report was produced by Luís Fontes and proposed the classification of S . Lourenço chapel and of Calçada da Ordem, as Public Interest Buildings. It was sent to the Portuguese Institute of Cultural heritage in October 1997.

between Braga and the fields. It was a quicker and less difficult trajectory and this outline is believed to correspond to the Via XIX itinerary. Some arguments also favour the suggested layout. One of the arguments suggested that the access to the Tibães area65 would not be done via the current street, as expected if it corresponded to a Roman or medieval layout, but rather via a previously built road, meaning the Via XIX itinerary (Map 13)66. D. Henrique and D. Teresa did the first concession of the Tibães area back in 1119. The boundaries were subjected to three versions and only the third one, a consequence of the litigation with the bracaranse Mitra, ending in 1693, ensured the passage of the Braga/Prado road by the convent limits. Aurélio de Oliveira referred the following; ‘In truth, both the first and second versions bounded the area via Mastro Mau in S. Pedro de Merelim in the eastern part of the current Braga/Prado road going downwards up to the Cávado. This part of the land was included in litigation with the bracarense Mitra that only ended in 1693 and from which the territory division was originated. The eastern limit coincided with ‘Estrada do Carmo’ linking Braga to Prado via a bridge that still exists today, and they performed terrains under the Abbey jurisdiction’ (Oliveira 1979: 42-43). The references to the construction or repair of the current Braga/Prado road, together with the boundaries of the Tibães area corroborated what conducted prospections and oral testimonies indicated. The Via XIX itinerary should have followed an upstream layout in terms of the location of the current trajectory linking to Braga, as data handled by Aurélio de Oliveira seemed to suggest: ‘Prior to 1630, the Abbey and the Tibães area population collaborated in the opening and conservation of the Prado/Braga road alongside the bridge construction that still today enables the crossing of the Cávado river (…) Before the 65 The first reference to the Tibães Monastery was part of a private donation made to the Cathedral in 1077 and included Tibães properties located in the outskirts of where the monastery was established (Fontes 2005: 188). In 1110, D. Henrique and D. Teresa donated the Tibães Monastery and adjacent lands to the same entity in the Carta de Couto. In 1833/1834, the monastery was closed and its assets were included in the Fazenda Pública. In 1838 its fence and in1864 its building were acquired in public auctions. The Tibães area included an area of around 700 hectares and encompassed the totally or the majority of Padim da Graça, Mire de Tibães, Parada de Tibães, Frossos, Panóias, Merelim (São Pedro) and Merelim (São Paio) parishes. Its boundaries were mapped according to enquires conducted in 1220 by Luís Fontes, who also described its area and features as the foollowing ‘The Cávado river bounded the land on the northwest, Montélios and São Gens (current São Filipe mountain) marked it on the southwest, it would then go Northwest from Parada and Semelhe up to Fonte Cova and from there to Gandarela and Castro Mau, ending by the Cávado river’ /Fontes 2005:38). 66 The boundary of Tibães was proposed by Luís Fontes, to whom we thank.

108

The Shaping of Bracara Augusta Peripheries

opening of that road and the access to the bridge, Tibães boundaries according to the texts was made further up until the Macrome area that was the limit of Tibães Couto with Braga, where another street would lead – the old street linking Braga with Ponte da Barca, Ponte de Lima, Valença and also Viana. This was a site where a boat enabling a real passage would exist and serve all passengers. This passage stopped to be used after the opening of a new bridge and a new road further down’ (Oliveira 1979: 70-71).

boundary axes. Cippi without any inscription have also been found other than delimitation Milestones.68

Another potential trajectory was rather identical to the previous and would lead up to the site known as Cangosta da Cabra, nearby Santo Estevão and from there turning right via what is known as Cano or Bica das Moças, up to Santo Estevão Velho or Pardelhas. From there it would follow the river. Although this trajectory can still be utilized today, it seemed to correspond to an old road ensuring the connection with Santo Estevão Velho archaeological site where a villa likely existed.

Cippus 1 (ID 3031004) was located on a site known as Felgueira bounding Dume and S. Pedro de Merelim boroughs. It was a granite monument with 0,46 meters high and 0.38 meters width. It had two ‘padded’ sides with a perfectly smooth top and two lines crossing the top with a NW/SE orientation. This does not match the orientation of the current trajectory but rather the same orientation of the referred centuria, similar to the one followed by the fruit plantation of the adjacent terrain.

Field walking survey conducted in 2004 enabled us to identify what we believe are two lands surveying cippi (from here onwards named as cippus 1 and cipus 2). Both have been tailored in granitic blocks of around 0.50 meters high by 0.40 meters width. It is possible to see in one of them ‘padded’ traces on both sides. Both cippi presented perfectly smooth upper sides where a cross was engraved occupying it entirely. Both pieces have the same alignment and kept a distance between themselves of precisely one centuria (20 actus).

3.2.3 Termini and limiting devices

The operations to define boundaries were materialized within the physical limits and followed diverse morphologies: cippy without inscriptions; cippi with inscriptions that were more or less detailed; small temples and even niches. There was also the possibility to define boundaries with roads and trees among other elements. Some Latin literature specified and catalogued these physical boundary elements (Chouquer and Favory 1992 18-26). The land surveying cippi encompassed more than one function. They served to limit lots attributed to settlers, to bound limits between the territories of the cities or signalling the delimitation axes and respective sidewalks.67 Some of these cippi present in their upper part the detail of a cross - the decussis that was represented by two undefined limits, the kardo and the decumanus that would cross each other on the exact site where the cippi was located, thus indicating the orientation of those two axes. In some cases, the decussis may have included characters aiming to indicate the location in relation with wider

67 The first cippi from the Graccus period resulted from the need to bind the ager publicus, to establish the division of the land and to attribute land to the settlers. According to the Liber Coloniarum I, these first cippi were cylindrical columns with a diameter of around one foot or one and a half foot and a height of four or four and half feet. More than a dozen items of this typology are known to have existed and the vast majority is located in central and southern Italy. It is widely acknowledged that the height and the diameter of these items do not match the measures presented above (Chouquer and Favory 1992: 18-26).

68 Cippi are known to have existed in the Italic peninsula and the vast majority of these structures are located in Agrica (Castagnoll 1958: 11-12, Arino Gil, Arino Gil et alii; Decramer et alii).

109

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Figure 11. a, b, c Perspective of cippus 1, in 2004

This monument was later moved, which led us to programme an archaeological intervention in 2006. The overall aims were: to determine the size and feature of the cippus; to assess its implantation system (potential existence of a base); to determine its implantation chronology in case the monument was in situ or on the contrary, to assess potential soil mixing procedures or monument displacements. Figure 12. Perspective of cippus 1, in 2006 The overall aims also included the implantation of a 2x1m survey with the limit being the exact middle of the cippus, in order to obtain an accurate stratigraphic reading. Our hope to safeguard of the monument and avoid future displacement from its original site, determined the survey implantation and the degree of attention needed in the excavation.

110

The lack of a foundational ditch or other ancient associated artefacts did not enable a clear stratigraphic assessment of its chronology. But its already studied features and elements correlation allowed us to put forward the hypothesis of being dealing with a Roman gromatic cippus. It is likely that the cippus was later reused as a boundary unit in medieval times and functioned as a territorial boundary element until our days.

The Shaping of Bracara Augusta Peripheries

We are dealing with a granite monument of 0,47 meters high and 0,28 meters width. It presented remains of having been punctured on the sides and, similarly to cippus 1, the top was smooth and presented two engraved lines, which, at the time it was found it, had a NW/SE orientation, the same as observed for cippus 1. According to information recollected in the area, both cippi could not have been displaced from the site where they were originally located, as they bounded both parishes. These limits were carefully kept by landowners and by public services functioning in both municipalities. Figure 13. a, b Survey perspective of cippus 1, in 2006

In the territory where these monuments were located, there were disputes over the Palmeira parish boundaries in the eighteenth century. The documents used by Braga cleric Court of Appeal to define the boundaries of this particular municipality contained a transcription of the events in S. Pedro de Merelim. The following text was written in 1735:

Figure 14. Survey plans of cippus 1, in 2006

Cippus 2 (ID 3031005) was found in a field known as Pinhel establishing the limits of S. Pedro de Merelim, Dume and Palmeira parishes. It can be found some meters away from an area crossed by the Via XIX itinerary.

’(fl. 78) … na Estrada do caminho tracesso que vai da Estrada publica para Felgueiras (…) continuando a dita demarcaçam pello ditto caminho de poente a nascente the o fim quingosta que vai para o Monte de Felgueiras e dahi curtando direito por dentro da bouça (…) athe hum marco que fica no monte devizam da freguesia de Sam Pedro Frossos e Sam Martinho de Dume e deste marco contonuando a demarcaçam da freguesia de Sam Martinho com a de Sam Pedro pello Monte de Felgueiras asima para a parte do monte athe o heido (…) aonde está hum marco grande quazi enterrado que esta em (fl. 78v.) pumar (…) e continuando a demarcaçam deste marc opera a parte do monte pello caminho de dentro onde estam tres marcos todos tres direitos hun e outros todos com suas cruzes por sima’ (ADB, Relação Eclesiástica, Sentença sobre os limites de Palmeira, 17511771, pasta 19, nº 583).

In a Roman remains interrelation context, especially with regards to artefacts that may provide leads over the potential alignment of the orthogonal land division, there is a monument located on the eastern part of these cippi that must be referred. This monument was similar to some boundary representations taking place in the Roman and medieval surveying treaties (Chouquer and Favory 1992; Portet 2004). In the site known as Fonte de S. Vicente, in the slope of a hill located in western side of Monte de Montariol there is a rocky water spring where a tank has been carved that appears to be an archa petrinea (ID 3033103), an element usually referred in medieval documentation as a boundary delimitation device. Nearby, brick and tegulae were uncovered (Fontes 1003). Figure 15. Perspective of cippus 2, in 2004 111

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

relevance. Frequently appearing in medieval documentation as a boundary delimitation device, it was possible that the S. Vicente Fountain small tank served as a boundary element for a Roman centuria located in the north side of the city. The presence of water in considerable quantity helped to explain the Mars inscription dedicated by one family producing wool and tissues, an activity that needed water in considerable quantity. Given the proximity to Bracara Augusta, it may have used the city as a market site.

Figure 16. Location of the Cippi (1,2) and of the archa petrinea (3)

A votive altar was collected from a wall nearby the fountain and it was dedicated to Mars. It has been deposited in Pio XII Museum, in Braga and presented the following reading: COPORICI/MATERNI/EX VOTO/ MARTI TAR/BVCELI FV(L)LONES The members of the Coporicius Maternus family were the commentators of the monument under a vow made to Marte Tarbucelis (Santos et alii 1983:183-205). The dedication seemed to have been related with a thanking done by a family connected with the wool and tissues manufacturing.

Another relevant element associated with property boundaries was the epigraphic monument interpreted as a property limit and originally located at the Esporões churchyard. It was a monument with a ‘stela’ shape, 1,10m high and presenting a line in the upper part of the inscription where one can read the word SEVERI, which may indicate (property of ?) Severus (Tranoy 1981: 364; Fontes 1993: 73). Taking into consideration the fact that the monument in question may have been displaced, its location nearby the northern alignment next to the two cippi mentioned above is rather curious. It could actually have corresponded to another element signalling one of the land division axes bounding the territory under study (axis 12-13). The value of these boundary elements should be included within a wide diachronic framework of analysis of organizational and land division devices present in the area. The existence of landmarks on the landscape, either the Roman cippi or old pathways preserving Roman layouts visible in cartography and detected in prospection – may be related with permanence and transmission landmarks, that rescued from the Roman centuriae grid. The present work is not focused on assessing these questions, hence doubts arising from this suspicion. However, it might be relevant to suggest a few further details that in the future might be integrated in an overall equation of the subject. First and foremost, the Villa located in Dume evolved into a monastic church that then became diocese headquarters between the sixth and seventh centuries (Fontes 1999). The following statement is also pertinent: ’In the SueviVisigothic period (sixth and seventh centuries) there was a great effort in the church territorial organization and this effort put forward the role of the Martinho and S. Frutuoso bishops, who provided the rural populations with special attention’ (Fontes 1999: 39-140). They were responsible for the foundation, not long before the establishment of the Tibães Monastery, of the two older monasteries documented to have existed in the current Portuguese territory, precisely in Dume and São Salvador de Montélios. The first was built nearby the Roman Villa foundations and the latter in a small hill of around 800 meters southwest from the first monument. Both were sheltered by hills leading to Braga, the capital of the Suevi kingdom between 411 and 585.

Figure 17. Votive altar dedicated to Mars

The presence of an archa petrinea, a monument associated with a votive inscription to Mars, was of paramount 112

The boundaries of the Dume diocese in relation with the diocese in Braga likely generated the crystallization of some landmarks in the immediate outskirts of the city. The Liber Fidei, a cartulary gathered to serve Braga dio-

The Shaping of Bracara Augusta Peripheries

cese heritage in relation to Santiago de Compostela throughout the twelfth century, precisely aligned the relationships between both dioceses and accounted the presence of natural and construction boundary elements. In one of the first transcribed documents dating back to 911, the old delimitation of the Dume diocese under the order o Ordonho II, established not only the settlemts named as Villas but also used diverse boundary elements (archae, petras fictas, terrae tumeda). The epigraphic monument of S. Vicente fountain was referred in the beginning: (…) ‘Et tunc venimus nos iam supra dicti previsores ad ipsum locum, ascendimus in montem qui est inter ipsum locum et Villa quem dicunt infidias, invenimus ibidem in petra caracterem Sancti Vincenti et exinde in alia petra invenimus cruce et inde per petras fictas qui ab antiquo pro termino fuerunt constitutas et exinde in via quam dicunt de Vereda qui discurret de Bracara et invenimus ibidem congesta magna petrinea et inde per agirem et petras fictas quousque in archa qui sedet sculta in petra et exinde ad alia congesta terrinea et inde per agirem et petras fictas qui sunt secus via de Vereda quousque in termino de Pittanes et inde per agirem et archas principias qui divident inter Dumio et Palmaria et exinde per agirem Certuor qui dividet inter Dumio et Paretelias et inde per totum ipsum agirem usque in termino de Lesmiri et inde per alia via de Vereda qui discurrit de Bracara quousque inter tumeda qui fuit manum facta et exinde in archa petrinea principia ab antiquis constructa qui dividet inter Dumio et Villa Lesmiri et inde per agirem firmissimum qui dividet inter Dumio et Villa de Forozos usque ad alia archa petrinea qui dividet inter Dumio et Parata Samueli et Forozos et sic per agirem et per petras fictas usque ad congesta qui dividet inter Dumio et Culina et inde per agirem et per petras fictas usque in termino de Villa Pascasi et invenimus ibidem petra scripta ubi dicet Sancte Eolalie ubi dividet Dumio, Culina et Villa Paschasi et sic eximus per signas certas usque in montem ubi (p. 39) prius incoavimus et invenimus in ea ipsos terminos parietes quos obtinuit Spasandus et alias quos obtinuit Alamirus seu et ecclesia vocabulo Sancti Fructuosi quod dicunt Montelios. Et quod potuimus previdere in veritate sunt ipsi termini fortissimi ab antiquis constructi ipsius loci iam fati et manibus nostri confirmamus’ (Costa 1965: 39-40)

If we accept that current Dume parish inherited the old area of the Dume diocese that in turn was located on the exact site of the old Villa, we must also account the remarkable quadrangular image that results from the de113

scription that the parish priest produced as an answer to the parish survey promoted by the Crown, in 1758: ‘Fora dos suburbios da muito antiga, nobre e leal cidade de Braga, Primaz das Hespanhas, para a parte do Norte, em a espaçosa ribeira que cruza o Cavado com o dillatado giro de seu curso, se extende aprazivelmente agradavel , a freguesia de Sam Martinho de Dume, em outro tempo celebre Bispado. Está situada esta freguesia quazi em figura quadrangular; supposto que da parte que corre do meio dia a Poente e de Nascente a Norte, se lhe perceba mais alguma extenção, não hé couza que a deforme da sua quadrangular figura. Será toda a sua extençam meia legoa pouco mais ou menos, medida por cada hum dos seos quatro ventos. (…) Das fontes, não sei virtude especial; só sim que são puras, diafanas, claras e de bom sabor. E que entre commuas e privadas, tem esta freguesia mais de vinte e quatro. Não tem rio de concideração, só um piqueno regato com o nome de rio dos Gafos, cuja ethimologia hé tradição commua nesta freguesia, lhe vêm de hum hospital que antiguamente houvera em o lugar da Facha, em que curavam semelhantes achaques (…). Tem de extenção huma legoa. Nace nos lemites desta freguesia, involve-se com o Rio Torto e com elle se mette no Cavado, junto à Veiga do Argaçal em o coutto de Tibães. Corre de Nacente a Poente. Há huma ponte de pedra de hum só arco, sobre outro regato que cruza esta freguesia de Nacente a Poente e se mette também no rio Torto, a qual chamam a Ponte de Sobremoure, o qual nome há de hum lugar visinho assim chamado. Não tem montes concideraveis. O mais extenço hé o do Crasto, nos lemites que a devidem dos suburbios bracharenses.’ (Capela 2004: 172-174).

3.2.4 The characterization of the cadastre

The search for orthogonal axes indicating one centuria demanded a group of complex procedures. First and foremost, it was key to distinguish the major structural axes of land division from of the small orthogonal axes that can not display more than a singular orthogonality.69 This structuring land division axes may have corresponded to fences, walls, pathways bounding properties, or sidewalks that if observed or ‘recouped’ in several data sources performed the landscape organizational axes (Map 12). 69 The orthogonal planning is not synonymous of the Roman land division. Thus, the orthogonal axes and the regular sizes between them are part of several rural divisions presenting different chronologies. Here, we used the classic terminologies to designate the N/S and E/W axes, the kardo and decumanus respectively.

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Oriented research was conducted as a safety measure between the 15º and 20º NW.

was linked with the city foundation in Augustan times (Palet Martínez 1994:111).

Thus, the orthogonal fabric of Bracara Augusta with an orientation not entirely homogenous in the entire city was taken into consideration. The 18,5º NW was the orientation angle and included several blocks and identified building constructions.

This land division modulation may be suggested by the location of the boundaries or termini matching cippi 1 and 2 alongside the archa petrinea defining one module of 20 actus, and following the cardus orientation. On the other hand, the two cippi located on the axis 13 alongside the archa petrinea representing a potential ‘’crossing’’ of cardo 16 with an N/S orientation with a S decumanus, enabled the assumption that the orientation of the land division was 16º NW.

To this effect it was used orthogonal axes of Bracara Augusta with an orientation of 18,5º. The demand for orthogonal axes also involved research of the current orientation axes in the rural areas. We rehearsed the superimposition of a grid with 20 x 20 actus (one centuria) with its location fully respecting three variables: the most significant detected alignments within the area under study; the location of both and surveying cippi detected in Felgueira and Pinhel and also, the location of the archa petrinea. We attempted to verify the wider or smaller density of the rural land division traits according to several propositions. Equally, we conducted a research over regular equidistances in actus, particularly in 2 actus (=240 feet = squares of 70,96 cm on the side) and of 5 actus (squares of 177,4 m on the side). Finally, the assessment of results was done with regards to the archaeological shadowing conducted in the surrounding area of the city. Similarly, the detection of pathways and occupied areas was valued as the viae vicinales and privatae became an important data source in this particular research. Thus, we recovered the hypothetic layout of land registration with an orientation in 16º NNW encompassing, in grosso modo, a region of around 320 square miles with a regular orientation and modulation in actus, that was particularly visible in the conserved alignments from the 7 to 19 and M to Y axes, respectively. The cadastre would have been organized in centuriae of 20 x 20 actus.70 The recovered traits presented an NW/SE orientation were perfectly adapted to the structuring shapes of the reliefs and to the main hydrographical network found in the area included within 5 to 7 km around the city. This systematic orientation, with the actus modulation and its articulation either with the city layout or with the main regional road axes put forward the suggestion of a centuriation likely connected to the Bracara Augusta foundation and the development of the road network in the Augustan and Julio-Claudian times. This was similar, for example, with the Barcino case, where the centuriation

70 One Roman foot corresponded to 0,2960 meters (29,6 centimeters). This measure was found at the Carvalheiras insulae in Braga. This exact measure is also found in Coninbriga and was close to the pes monetatis, totalling 0,2957 meters. We then have the actus, which was the average surface measure matching 120 feet. The actus quadratus is a square of 120 feet by 120 feet, meaning, it corresponded to 35,52x35,52 meters performing an area totaling 1261,62 square miles.

114

The modulation actus was set from struturing axes including the best preserved ones in the fertile plain north of the Roman city (axes 8 and 12 to 15) and in the south thereof (12 and 13 axes). Thus, and taking the available elements into consideration, it was possible to establish one module 20 x 20 actus that performed centuriae of 200 jugera on the surface. Nevertheless, in this basic modulation several variations may have been used that need to be further researched. The best conserved cardines corresponded to axes crossing the whole territory under study, even located further away from the Cávado River. One of the cardus boosting a definition and with a persistence clearly significant was the one establishing the western limit of the city. It was also present in the northern area and corresponded in some parts to the layout of the Via XIX itinerary (13 S, T, U) running south of the city in an axis following the layout of the Via XVI itinerary (13 J to O). Another key alignment corresponded to cardus 12, running between the Cones area, in the west part of the city and Real, and partially matching the current Braga/Prado road trajectory. The axes corresponding to the decumani were more difficult to detect. In the southern part of the city a well-conserved decumanus corresponded to the axis W 13, 14 and 15 and was visible in both cartography and aerial photography. On the north, the axis identified as P12 and P15 may have corresponded to the ‘valo velho’ outline mentioned in the 4º ‘Tombo do Cabido: ‘Item acçima dos Chaaos honde descobre o cano estam bõns vij carvalhos todos juntos E tem no pee uma mea + por esta guissa ┬ Item logo abaixo quando vam para Sam Viçente, junto da porta principall do dito Sam Vicente a maão ezquerda estão iij carvalhos bõns com a dita marca dos outros de cima. Item logo abaixo de Sam Vicente esta hua leira de devessa a mayor parte dela nova que tem cem carvalhos daqueles antre grandes e pequenos E a dita leira partida ou demarcada per huum vallo velho .s. que de contra o soaão E de contra o poente parte pella estrada que vay da cidade pera Sam Martinho de

The Shaping of Bracara Augusta Peripheries

Dume E a dita leira atravessa todas as estradas . s. a que vay para Sam Martinho e a que vay pera Palmeira’ (ADB, Tombo 4º do Cabido, fls. 51 v. e segs.).

Most likely, this outline corresponded to an observed alignment in the Map of Braga produced by André Soares, in 1756. Furthermore, it is possible that the Via XX itinerary trajectory, that has until today not been the object of detailed prospection and study, corresponded to a deformation of a decumanus. Although, other modulation can exist with research emerging from this work, the proposed grid of 20 x 20 actus is attested by the clear intersection of some axes, particularly visible in the 11 to 5 alignments with the V, U, T, S, R, or in northern Cávado, the Y with 7, 8, 15, 16 and 17. Near the city, on the southern part, there were the crossing axes of O with 13 and in the northern side the P aisle with the 12, 13 ad 14 axes. Finally, the location of the archa petrinea corresponded to an intersection of the S axis with the 16, corresponding to the cadastre theoretical grid and it may correspond to the centuriated area due to Montariol elevation that likely was kept outside of this assignatio. The layers superimposition corresponding to the altimetry and hydrography with a 1:25000 scale allowed the assessment of the recovered land division occupying the valley areas with 10 to 60 meters altitude, never surpassing the level curve of 200 meters. The centuriated area seemed to correspond to a wide area presenting a rectangular shape and encompassing a territory with less elevated altitudes and good soils with a considerable agricultural potential, also crossed by Roman roads. The limits were defined by altitude elevations ranging between 250 to 350 meters, with the exception of wider elevations represented by the S. Mamede / Guizande settlement with 427 meters high and the Santa Marta da Falperra hill with 562 meters, both located in the southern part of the studied area. The centuriation was prolonged in an area with lower altitudes and located in the right margin of the Cávado River. This area was particularly visible in the lower basin of Cabanelas on the northwest and also in the region encompassing the intersection of the Homem and Cávado rivers in around 4 km further north in relation to the Cávado River. There, the limits corresponded to terrains with higher elevations in terms of topography and corresponding to altitudes ranging between 120 to 174 meters of altitude.

Figure 18. Roman Cadaster

It is also important to refer three viae privatae identified in field excavations conducted in 2001 (Lemos 2002:110122). In fact, we are dealing with the only excavated trajectories in the region and the only ones where we can recognize the sizes and type of pavements used. 71 Furthermore, the study of these roads corroborated the assumption over the existence of a dense network of viae privatae connecting the several rural establishments and of those with the city itself. With regards to the rural land division, it is also important to acknowledge that two of these viae had an identical orientation with the urban layout and with the one detected in the centuriated area.

71 Taking the example of the Via XVIII itinerary aside, an itinerary excavated in some of Gerês mountain range part, these are the only excavated trajectories presenting published results. This is particularly relevant due to the fact these were not main itineraries.

115

Chapter 5 1. The spine settlement throughout the Via Nova trajectory

The Shaping of Mountain, Valley and Coastline Settlement

Assessing settlement patterns in a predominantly mountainous region performed the key aim at this phase of our work (Map 14).

Introduction

The western expression of the conventus bracarensis settlement should be analysed considering the geomorphological diversity shaping and stimulating the human occupation of this region. This has been the assumption underlying this work from the introductory part, throughout the development of our successive approaches to the Roman settlement. In the previous chapter, we acknowledged the fact that within the vast outskirts area of Bracara Augusta, and in terms of distribution and typology of archaeological records, there were different interaction moments in terms of the landscape human occupation by the existing structuring entities. It is now time to widen our observation horizons and include two other areas that we considered to be strategically located. Their study is part of the overall aim of this work: acknowledge different correlations of settlements basic entities. The present chapter will focus on a mountainous area crossed by a Roman itinerary and on the valley and coastline areas. Due to their distinct morphologies, both areas have a relatively marginal status with regards to the study of the Roman territory. The first refers to a mountainous region that may not have interested researchers when observing the classic Romanization phenomena, namely when studying the urbanization processes or Villas modular expressions. The second is a coastline area presenting features that are rather empty in terms of human occupation. Regardless of their marginality, both areas should be reassessed due to the available new data and to decentering territorial analysis from an excessive hierarchy of ideal settlement models. The interest in both areas resides precisely in their potential and ability to prompt two observational phenomena we believe are critical to be further studied. The first one is linked with the determinant role of the road network system in reshaping mountainous settlement processes, providing areas with a ‘spin’ shape. It is possible to appreciate how the other occupational entities reacted to this predominance (point 1). A second phenomenon took place in the valley and coastline region, where, according to our research, no hegemonic impact has been registered by a structuring entity in the settlement of the area. On the other hand, we understand that the Roman occupation and development occurring in this region was helped by wider correlation solutions in terms of occupying space and exploiting resources (point 2). 116

The study of a region naturally articulating itself with Bracara Augusta took into account the fact this territory was crossed by a Roman road - the Via Nova or the Antonine Via XVIII itinerary. The circuit was built between 79 and 81 AD under the supervision of C. Calpetanus Rantius Quirinalis Valerius Festus during Titus’ and Domitian’s reigns. This road was vital in the overall organization and exploitation of the inner peninsular northwest territory and, in particular in the interior mining region. In all its extension, the Via Nova measured 215 miles, around 318 kilometres. It had 34 miles located in the Portuguese territory (around 50 kilometres), which was the same distance separating Bracara Augusta from Portela do Homem, in the Gerês mountain. Our analysis started with this exit trajectory, when leaving the Roman city. This choice was related with the intention to assess settlement sites associated with this trajectory in the city outskirts. Thus, we will attempt to understand the way this itinerary contributed towards the overall settlement organization and how these trends shaped a landscape filled with steeply reliefs and poor demography levels.

1.1 Entities and relationships

1.1.1 The XVIII itinerary or Via Nova: trajectory and associated epigraphy

The transport structure launched by Augusto after the pacification of the peninsular northwest was consolidated in the Flavian dynasty in the second half of the first century AD. The construction of the Via Nova marked this stage of the history of the peninsular northwest (Tranoy 1981)72. The Via Nova, also known by the number XVIII or by the words ‘Geira’ or ‘Jeira’, should be contextualized within the economical growth lived at the time and by the consolidation of the imperial power in the peninsular northwest. More direct than the old connection between Bracara Augusta and Asturica Augusta via Aquae Flaviae (365,5km) known as the Via XVII itinerary, the Via Nova 72 The oldest milestone of this trajectory goes back to the the Emperor Titus time (78-81).

The Shaping of Mountain, Valley and Coastline Settlement

had 318 km and crossed a region marked by rugged reliefs.

realities that also include distinct areas from a settlement organization point of view.

The trajectory was part of an Augustan project destined to provide Braga with key transport connections. It boosted a longitudinal cut penetrating a rather low demographic region integrated in the conventus bracarensis and the western part of the conventus asturicenses. The consolidation of the Roman power in a poorly populated region and the crossing of this road in the gold mining region of Bierzo were key for the Empire economy.

We limited ourselves to provide a detailed description of lesser known and promoted part of this trajectory, despite it being subjected to intensive prospection works. We are referring to the part between the city of Braga and its outskirts and all the Cávado River valley and the Paredes Secas area, in the Amares regions. This is an area located between the city exit and mile XI. The remaining part of the trajectory will only be briefly described given the recent mentioned publication (Lemos 2005). In this particular area, we will only consider settlement evidences present alongside the Via XVIII itinerary and somehow impacting upon its layout.

The area under scope corresponded to the region between Braga and Portela do Homem, presenting considerable geomorphological contrasts, shaped by rugged reliefs and with contour lines ranging between 500 and 600 meters. The area includes also valley areas near the Cávado and Homem rivers and registers altitudes reaching 100 meters contour lines. The excellent preservation of most part of this road trajectory and the abundance of associated Milestones may explain the wide interest in researching this part of the Roman itinerary, as it can be observed in several writings going back to the eighteenth century (Argote 1932-34; Capela 1985). More recently, the Via Nova has been the focus object of several projects and detailed publications (Rodriguez Colmenero et alii 2004; Lemos et alii 2005). Thus, it doesn’t seem relevant for us to provide a detailed description of its layout.73 A recent work focused on the part where this road crossed the Gêres mountain range suggested the possibility of dividing this trajectory into 10 segments (Lemos et alii 2005:19). This was linked with the trajectory features and specific challenges. The division of the Via Nova obeyed to requirements of methodological nature and was framed upon a set of questions. We could propose other layout divisions or subdivisions, all respecting other sets of questions but this would demand a shift in our observational scales. However, the division proposed by F. Sande Lemos and the way it was organized fulfilled our main analysis requirements. Furthermore, the proposed divisions corresponded to different geomorphological and methodologically operative

We believe that the city exit was done from the S. João do Souto square via Janes Street, which had a sinuous course that may have had fossilized the initial part of the Roaman road itinerary. Urban transformations carried out with the opening of Souto Street and S. Francisco square changed the extention of this trajectory that was then only recouped by the layout of Chãos Street. In fact, in 1995 when a necropolis was identified in the western part of the old Sant’Ana field, the current Central Avenue, it was then possible to confirm this layout, equally suggested by the finding of a votive altar dedicated to Lares Vialibus during an excavation conducted in the S. Francisco square. The altar was consecrated by Caius Iulius Saturninus and dates back to the first half of the second century (Cunha et alii 2005). In the Chãos Street, the trajectory went Northwest via the current São Vicente street and followed up to Infias square, Conselheiro Bento Miguel Street and Monte Arcos square (the old cemetery square). Braunio’s Map clearly illustrated this first part of the trajectory and presented a legend stating the following ‘’Via Antiqua Imperatoris Antonini in Aquas Calidas, vulgo Orense’’ that according to Francisco Sande Lemos is related with the fact that Via Nova was used as an access point to Lucus Augusti via Auria from Bande. This trajectory kept a significant permanence and became a pilgrimage route to Santiago de Compostela, via Ourense. These events corroborated the legend of Braunio’s Map and the fact that the existing pavement between Braga and Adaufe is still entitled ‘Caminho de Santiago’. Other than the Santiago de Compostela trajectory, the mentioned circuit has also been used as a pilgrimage route to São Bento da Porta Aberta, in Terras de Bouro.

73 The Via Nova recovery and research project taking place in the Portuguese territory already registers a significant amount of published works, from where we highlight the cooperation between the Archaeology Unit from the Minho University and the Parque Nacional da Penede Gerês. Between 1977 and 1992, this cooperation produced several field works conducted in the XXIX and XXXIV miles and included excavations, prospections and cleaning operations enabling the identification of new milestones, the understanding of the trajectory outline and provided the possibility for the road to be visited. Furthermore, this cooperation was behind the publication of some works and the organization of a Colloquium. Field walking surveys were conducted at a later stage by researchers from the Archaeological Unit in the entire itinerary trajectory, in Portuguese territory. Between 1998 and 2002, these works were mainly focused on the XV and XXIX miles under the supervision of Fransisco Sande Lemos. We thank the researcher the access to prospection reports and wider provided information.

Areal de Baixo, an area with 210 meters altitude, was the location of mile I. A Milestone dedicated to the emperor Constans (333-356) signalled this site. The item is currently located at the Soares dos Reis Museum (Martins 1990: 88-89). The part of the trajectory between Areal de Baixo and the Sete Fontes area raises some questions in terms of both layouts. According to Francisco Sande Lemos, the trajectory would follow northwest from Nova do Areal de Baixo 117

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Street and would follow up to Areal Square, Areal de Cima Street and Rafael Bordalo Pinheiro Street, with mile II located at the Sete Fontes. The latter is a site registering 264 meters altitude, and from here the trajectory would go downhill up to the ‘Lugar do Caminho’ via a paved sidewalk that the author concluded couldn’t be attributed to an old Roman pavement. We propose an alternative trajectory from Areal de Cima up to Montariol and running nearby the site where an inscription dedicated to Marte Tarbucelis has been uncovered. This is an artefact already analysed in the previous chapter. Moreover, this trajectory corresponded to an old paved road enabling a ‘smooth’ circuit up to the valley area where Adaúfe is located and would avoid the sharp decline between the 264 meters, altitude of the Sete Fontes chapel, and 117 meters corresponding to the height of Adaúfe. Thus, and as stated by Fransisco Sande Lemos, the decline between both sites corresponded to 147 meters in an extension of 943 meters, determining a permanent inclination of the road trajectory.74 A trajectory that would have avoided these trends, as it is the one suggested by us, would be a more suited layout for a Roman road. On the other hand, one must take into consideration the likely existence of an aqueduct located in the Sete Fontes area fossilized in the current 1700s infrastructure. For this reason, we deem more appropriate for this site to have been accessed by a secondary Roman trajectory, one that eventually was consolidated as a pilgrimage route in the times that followed. In tern, Via Nova would have followed a simpler route in terms of traffic and simultaneously, cater for the preservation of the city water supply site. The fact that one of these trajectories became a pilgrimage route, and that the other was left as an access between Braga and the valley areas naturally perform pertinent questions for which we do not have any justification at the moment. For this reason, the proposition for an alternative trajectory for this road itinerary should be faced as a simple work hypothesis. The road perhaps followed a trajectory crossing the current populated area nearby the parish church of Adaúfe, where a villa (ID 3030103) may have existed in the church surrounding area. From Adaúfe, the road would follow up north to the Cávado River and then via Redondo, Barreiro, Freire and Estrada where mile IV would have been located. From here, it would continue via Cortinhal, Souto and Salgueirinho. Mile V was located at the Quinta do Coelho, in Souto. The road would then follow up to Rio or Barca de Ancêde, where Francisco Sande Lemos suggested a potential crossing of the Cávado River existed. 74 Pioneer information by Fransisco Sande Lemos, to whom we thank.

118

The crossing of the river in boats mentioned in ‘Memórias Paroquiais’ in 1758, suggested an alternative trajectory with regards to the Cávado River passage nearby the Porto bridge, 75 which was consensual until recent prospections carried out recently. Arguments favouring an alternative crossing, also defended in the present work, are based upon the trajectory that seemed to be more credible to this itinerary, both in the south and in the north of the river. The crossing of the river in boats was practiced until recently, not only in the Ancêde area but also in other parts of the river. These events are attested by the passage of the Via XIX itinerary upstream of the Prado Bridge. The crossing of the road in Ancêde may still be defended if the photograms of 1946 by RAF are taken into consideration. We can see what no longer exists: large sandbanks that may have helped the wading crossing of the river, at least in the dyer seasons and in winter in boats. Other argument favouring the route layout via Ancêde is linked with the finding of a road in the left Cávado river margin, already destroyed. After the Cávado River was crossed, the itinerary would then follow via the current Além site in the Barreiros parish. From this location are four Milestones, although they have been scattered since being uncovered. Two of them have been linked with the Quinta de Agrolongo (ID 3010201) and one with the Quinta da Pena (ID 3010202). The later probably referred to mile VI and was originally located in Senhora das Angústias-Além. From this site, the itinerary would follow up north towards Carrezedo. Mile VII would have been located in the Feira Velha site, in the Carrazedo parish, where a low altitude Hill-fort also have existed, later occupied by a medieval manor house. According to a manuscript from the seventeenth century by the Marquês de Montebelo, twelve Milestones would have been recollected from the churchyard. Some were transported to Braga and placed in Sant’Ana Field during D. Diogo de Sousa times (Sousa 1971-72:180). We conducted field walking surveys in the same church and verified the presence of columns and identified a considerable quantity of Roman construction ceramics, common ceramics and an altar located nearby the site. The shaft of this element was completely changed as it has been reused (ID 3010702). The presence of an altar in Carrazedo added value to another piece of identic typology uncovered in Campo da Porta, namely in a field located a few meters away from the parish church. This was an altar dedicated to Lares Burici by Bloena (Sousa 1971-72: 181-182; Cunha 1975: 22-23; Tranoy 1981:303). 75 The Porto Bridge was for a long time associated with the Via XVIII itinerary in its Cávado crossing trajectory. It is a medieval construction. However it is possible that this bridge substituted a previous infrastructure of Roman origin. Nevertheless, this could only be possible if we admit that this road crossed the river on this particular location.

The Shaping of Mountain, Valley and Coastline Settlement

Ceramic remains detected in prospections are scattered around a wide area including the Carrazedo church adjoining terrains and extending themselves until the Feira Velha site and towards the area we believe was crossed by the road. Although, Francisco Sande Lemos believed a Villa could have existed on this site, the remains dispersion, the presence of two altars and wider architectonic elements recognised nearby the parish church led us to suggest the existence of a vicus, in Carrazedo. It is important to remember the existence of a low altitude Hill-fort, the Castelo do Castro site (ID 3010704), where some materials of Roman construction are still visible on the surface. Was it possible that the population of the Hill-fort moved towards an open settlement located nearby the road and for this site to have adopted mutatio functions? This phenomenon was relatively common as it has been attested by the works of F. Pérez Losada conducted in Galiza (2002). From the current Feira Velha site in Carrazedo, the itinerary would follow up north towards Barrimau and Pilar. This latter site, already located in the Fiscal parish corresponded to mile VIII, where a Milestone dedicated to the emperor Carus was uncovered (282-283). From this mile, the route went west bypassing the foothills of the Abadia mountain range, a circumstance that according to Francisco Sande Lemos enabled the itinerary to cross the watercourses existing upstraem and avoided the construction of bridges. On the other hand, this trajectory catered for a wider exploitation of the land located in the surrounding valleys of the Homem and Cávado rivers hydrographical basins, thus enabling the access to the Caldelas site. The group of recollected materials on this site suggested the potential existence of a thermal vicus due to the water features. The water from this site is still used today used for the same purposes (Martins 1995; 1996a). Based upon a proposition put forward by Francisco Sande Lemos, mile IX was located on the current Santo António site on the Besteiros parish and would then follow towards Pousadas, Caires, Paço, Paço Velho, Crasto and Tornadouro. The latter site should have housed mile X, if we consider the existence of vague references to the finding of a Milestone on site (Silva 1958-59). Mile XI coincided with the location of the Caires or Castro de Gróvios Hill-fort (ID 3010501), a site always linked with Cividade de Biscaia (Martins 1990:65). Despite the Romanization evidence present at this settlement, field surveys carried out in the valley area, mainly between the passage of the itinerary and the Freicheiro site, uncovered wall remains and abundant Roman ceramics of domestic construction. Locally, the site has been designated as Cidade de Biscaia (ID 3010503), which led us to locate a site of the vicus typology which may have encompassed mansio functions. Between miles X and XI, the trajectory seemed to have crystalized into a municipal pathway all the way up to Via Cova at the Paredes Secas parish, where mile would have been located. This part of the trajectory and the following 119

mile is very well preserved and presented pavement remains in pristine conditions in most of its circuit. In Ribeira de Pala, Dornelas (ID 3011502) at the Paredes Secas borough, a Milestone indicating mile XII dedicated to Maximino and Maximo (238) was found. The itinerary would then reach the Homem River through Portela de Santa Cruz and from then onwards, the circuit would follow the northern mountainside of Abadia, corresponding to the left margin of the Homem Rriver up to Covide. From there it would run to S. João do Campo up to the Vilarinho das Furnas crossroad, inflecting then towards the interior of the current Parque Nacional da Peneda Gerês territory. On its trajectory throughout the Homem river valley, the itinerary crossed the current Volta do Gavião, Bouça da Mó, Bico da Geira Volta do Covo and Albergaria sites (Lemos 2005; Lemos and Baptista 1995-1996). From the boundary of the Portela do Homem, the XVIII itinerary would then open up to the Lima river valley and entered the current Spanish territory running then northwest towards Aquis Originis - Banhos de rio Caldo, crossing the Lima River via the Pedrinha Bridge. Aquis Querquernis-Banhos de Bande, a key Roman settlement, was located in the northern margin of the same river (Rodríguez Colmenero 1995-96).

1.1.2 The fortified settlement

One detail never neglected by researchers when describing the Via Nova trajectory is the fact that it crossed a low populated area. This event, noted by all authors when referring to this road trajectory, cannot be explained from to lack of field surveys or further research conducted, as this is a rather well studied area. A similar situation occurs in other mountainous regions, as for example the Vieira do Minho region (Fontes and Roriz 2007). In fact, the low population rates seemed to be a common feature in most of the mountainous regions located on the border of Minho and Trás-os-Montes. In Roman times, this situation was prominent when compared to regions presenting medium and low basins of the main rivers crossing their territory. Conducted research projects in some of these regions are far from producing intensive and systematic knowledge over their studied sites. On the other hand, when studying mountainous areas it is important to adopt a diachronic perspective related with their human development in order to understand the lack of occupation processes for the chronological period under study. Our option focused on describing the references we had access to, while keeping in mind the necessary and already stated doubts and challenges. In order to facilitate the description and have a coherent reading of settlements included in A, B and C categories, we will pay attention to sites framework referring to the regions mentioned above, and leave out this group catego-

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

ries the settlements presenting distinctive topographic features and chronologies.

Hill-forts of A category Caires or Gróvios, Amares (ID 3010501) Located on a southern slope of the S. Pedro Fins mountain range and occupying a small hill separated from a rock knob end formation by a thalweg, this site presents altitudes ranging between 300 and 320 meters of altitude. Three wall lines are visible and define the site. Furthermore, there are some circular and rectangular walls that must have corresponded to houses. Archaeological literature referred the finding of metal pipes, bricks, amphorae and a potential kiln on the western hillside of the Gróvios site (Dias 1903:78). On top of these findings, a granite low-relief artefact representing a horseman with a geometric motif was uncovered (Cardoso 1972: 156). In recent field walking surveys conducted on site, fragments of indigenous ceramics from the Iron Age and fragments of common Roman pottery have also been found.

this site enabled the identification of several occupational phases that go back to the Chalcolithic period, the Bronze and Iron Ages (Jorge 1979, 281291; Jorge and Sousa 1980, 121-131). With Roman chronology are only a few ceramic fragments that do not support the idea of a significant Romanization process on site.

Monte de Caldelas, Amares (ID 3010602) Located on a hill with 220 meters of altitude, this Hill-fort had, at least, two walls. Conducted prospections supervised by Ana Bettencourt (1999: 17-218) concluded the inexistence of indigenous ceramic fragments of micaceous production and only manual ceramics fragments have been uncovered. The reason why this settlement is not excluded from this category is linked with the fact that there had been previous information suggesting otherwise (Jorge and Sousa 1980:120). In fact, the author expressly referred to having seen observed thick vegetation covering the site when visiting. We admit that, at least, one occupation stage took place in the Iron Age, as suggested by Manuela Martins (1990:66).

Espinho / S. Pedro Fins, Amares (ID 3011602) This Hill-fort occupies the southeast slope of the S. Pedro Fins mountain range and has 480 meters of altitude. Remains of a wall are visible and throughout the site fragments of bricks and iron items can also be recollected (Martins 1990: 70). Albano Belino (1909:6) mentioned the finding of worked stones, small manual grading wheels, ceramic artefacts and Roman tegulae. Conducted field walking surveys led us to suggest that the site was circumscribed to the 450 meters contour line and was not extended towards the current chapel area. In the group of Hill-forts included in the B category, the majority did not register any Roman occupation. Conducted excavations supervised by Ana Bettencourt (2000) at the Santinha mountain range enabled us to exclude Santinha setllement from the group of sites occupied throughout the Iron Age period.76

Sequeiros mountainous range, Amares (ID 3012101) This Hill-fort located at 228 meters high and had, at least, two wall lines. The uncovering of vases alongside the geomorphological features of the site enabled us to admit that this was a fortified settlement (Silva 1958_59 353-355; Martins 1990:71). However, field wlking surveys conducted on site did not identify any traces of Roman remains, events also attested by the lack of information supporting the same idea in related literature. Given the nature of the available information, we admit the site was a Hill-fort, abandoned at some point in the Iron Age period.

Hill-forts of C category

Lago, Amares (ID 3011503) (Cf. description in chapter 4)

Hill-forts of B category

Chã do Castro / Portela da Joubreia, Amares (ID 3011601) The Hill-fort occupies a small hill with 310 meters of altitude and is located on the southeast slope of the S. Pedro Fins mountain range. Excavations on 76 The ‘Monte da Santinha’ is a habitat located at 195 meters altitude and it was excavated by A. Bettencourt (2000), who acknowledged a Bronze Age occupation.

120

Castelo do Castro, Amares (ID 3010704) Located at 96 meters of altitude in a small hill overlooking the right margin of the Cávado River, this Hill-fort presented one wall and a trench although significantly destructed when the medieval a tower was constructed on site (Martins 1990: 67). Ceramics findings seemed to indicate both an Iron Age and Roman occupation. In recent field works,

The Shaping of Mountain, Valley and Coastline Settlement

it was possible to recollect some fragments of Roman tile and manual grading wheels.

Outeiro, Amares (ID 3011902) This Hill-fort is located at 175 meters high in a small hill presenting a unique platform surrounded by a slope. On the soil surface, fragments of bricks can be recollected and stone artefacts are also visible. They refer to old structures that are today in ruins. However, there are no coherent alignments on site. Apparently, this was a small and low altitude Hill-fort similar to the Lago site, in Amares. In recent prospections, fragments of tegulae and imbrex have been found.

For the group of Hill-forts located in the mountainous area, the typological and chronological assessment demanded a different set of rules that were not resumed to a simple calibration of altitudes. Nevertheless, the main issue related with characterizing Hill-forts in the region under study has been linked with the fact that most of the available information present in archaeological literature, going back to Arlindo da Cunha (1958-59) works, could not be confirmed withnew field walking surveys. This has been related with local difficulties in conducting prospections on sites covered by trees and thick vegetation layers. On the other hand, some of the sites identified as Hill-forts may actually correspond to castles or medieval watchtowers, without previous occupation.77 Although we have catalogued all the available data, we decided in some cases to exclude some sites from our cartographic analysis.

Alto dos Castelos, Amares (ID 3012002) The site is located at 750 meters of altitude, in a slope of the Santa Isabel mountain range. We highlight the findings of artificial slopes, Iron Age ceramics and some tegulae.78

Calcedónia, Terras de Bouro (ID 3100805) This refers to a site located on top of a hill boosting an excellent geostrategic location and great visibility over the Via Nova. Archaeological literature accounted for a group of structures on site, two of wide dimensions and significant extension likely to have been part of walls and remains of quadrangular and circular constructions. Fragments of tiles have also been recollected alongside common pottery described as ‘untypical’. The settlement grew towards the slope facing the Gerês River and the access to the site was likely conducted from the Freitas watercourse.

The region under study presents therefore a group of Hillforts located in a transition area, between valleys, mountain ranges and corresponding to the medium Cávado river basin. Perhaps due to the fact we have some degree of knowledge over these Hill-forts resulting from conducted excavations or because they can be included in relatively well defined typologies, the difficulties inherent to their study have been related with their chronology, the sizes of their occupational areas, with their role in Roman times, with the eventual dispersion of their inhabitants to other living habitats and consequently, with the retraction or abandonment processes the living spaces endured. On the other hand, the Hill-forts located in mountainous regions demanded no less complex studies. We have a clearly poor knowledge over the occupational areas and over the chronology for the majority of these sites. We deemed inappropriate to apply spatial organizational logics that are identical to the ones used in Hillforts located in low and medium hydrographical basin regions, as we are dealing with differentiated areas from a geomorphological point of view. This naturally determined their distinct features in terms of establishing habitats and their organization.

77 We excluded the following settlements: S. Miguel Mountain / Cidadelhe, Abadia, Bouro (Santa Marta) (ID 3011903) as we only have the information provided by Ana Bettencourt about the findings of rare fragments of indigenous ceramic and some tegulae and imbrex found on some site slopes. Without further prospections we believe we should exclude this site from our catalogue; Outeiro da Vila / Seramil settlement (ID 3012201), located on a small hill with 458 meters altitude on the eastern part of the Santa Cruz Mountain. Cónego Arlindo da Cunha referred that he observed on site some walls and recollected several ceramic artifacts. However, this information was not confirmed on further site visits. Covide castle (ID 3100807), where Arlindo da Cunha mentioned observing traces of walls and ‘stones similar to the ones of the ‘castrejas’ houses. Fragments of pre-Roman ceramics were rare but existed alongside with tegulae and imbrex’ (Cunha 1975: 513-514). Literature revealed some confusion with regards to this site and the Bouro castle. Finally, the Santa Isabel settlement at the Stª Isabel Mountain (ID 310110) occupies a hill with good visibility over the Homem and Cávado river valleys. On site, one can see potential alignments from walls on the west and northwest sides and some traces of ceramics manually produced. A visit to the site did not shed some light regarding its occupational chronology, although it seems more pertinent to state the site witnessed a medieval occupation.

The weak density of these types of sites in this area seemed to be a constant indicator and, apparently, refer to a long-term phenomenon confirming the idea that only diachronic programmed works will shed some light on the knowledge about these type of habitats, and about their geographic and historical contexts.

78 We thank Ana Bettencourt the provided information.

121

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis 1.1.3 The ‘spine’ settlement: vici, mansiones and mutationes

1973 in the Campo da Porta site opposite the Carrazedo parish church. The site is located two miles away from the Cidade de Biscaia (cfr image).

The implementation of a road trajectory demanded the planning and the establishment of sites that would function as stop areas located at predetermined distances. These sites had several typologies depending on their function, as we have already established in chapter 3. While the mutationes served as changing stations and normally corresponded to small infrastructures, the mansiones had wider dimensions and included overnight facilities. Other than the mentioned typologies, there were also tabernae or cauponae. The establishment of these types of sites obeyed the needs to accommodate imperial officials, high magistrates and imperial couriers traveling with superior supervision, and demanded the presence of storages, workshops or stables in these establishments. To these infrastructures was added the need to provide stop areas of short or long-term accommodation to all the travellers. One of the most interesting research work related with the transport network is precisely the attempt to identify the sites serving as supporting structures to circulation. In several cases, some establishments may have evolved from simple mansiones to real secondary urban conglomerates or vici. Francisco Sande Lemos estimated that the Via Nova had a mutatio every three or four miles (with each mile measuring around 1,480 meters) and one mansio in every eight or ten miles (Lemos 2002). The road trajectory between Bracara Augusta and Portela do Homem corresponded to 34 miles trajectory. Taking into consideration the available archaeological data, it is possible to identify a considerable number of these type of establishments. We will now describe them.

Cividade de Biscaia, Amares (ID 3010503) (vicus / mansio of mile X) The site is located on an upper platform of the southern slopes of a small hill of the Frecheiro site. According to on site recollected information, a stone wall, tiles and other ceramics were found at around 1 or 2 meters deep, at a time when a trench was opened to install the pipeline system. Prospections conducted, in particular in the so called Campo da Bouça area, made it possible to observe the presence of grading wheels, great quantity of construction materials, Roman common pottery and at least, three wall lines in works meant to widen the agricultural fields. The location and extension of the detected remains scattered across a smooth hill, located on a site suited for a road trajectory, and the Cidade de Biscaia toponym in this particular terrains leads us to locate a vicus / mansio on a site traditionally associated with the Caires Hill-fort. The structure is located 10 miles away from Bracara Augusta and 8 miles from Salaniana. In the Quinta de S. Vicente (ID 3010502) a statue pedestal to Genio has been recollected. It was dedicated by Quinto Sabínio Floro and dated back to the first century. According to Manuela Martins (1990), common Roman ceramics have also been collected from the site indicating the existence of a small temple perhaps associated with a vicus. Another hypothesis referred this particular item to have been displaced from the Cividade de Biscaia area.

Viriocelum / Vila Cova / Vilela, Amares (ID 3011501) (vicus / mansio of mile XI / XII?)

Bouça Alta, Braga (ID 3032701) A mutatio located just before the crossing of the Cávado River would very likely be located in Bouça Alta, in Navarra. The remains that we can attribute to this site exist in a smooth hill and are scattered around a wide area. They refer to construction ceramics and fragments of Roman common pottery.

Carrazedo / Feira Velha, Amares (ID 3010701) (mutatio of mile VIII) The site corresponds to a smooth and wellirrigated area where, according to an eighteenth century manuscript, several Milestones were uncovered. In field walking survey we found several column shafts in the churchyard area and abundant fragments of tiles which extend in contiguous properties. To these artefacts should be added a votive altar dedicated to Lares Buricis found in 122

This site, located between mile XI and XII, was probably located in the Vila (or Via) Cova / Mojeje (ID 3011501) site. It registers an altitude of around 300 and 320 meters. Artefacts found on site include manual graing wheels, worked stones, iron scoria, construction and domestic ceramics, foundation remains, architectonic elements from which a Doric capital and a coin from the third century from the emperor Constantius I Chlorus reign. The finding of a stone grave has equally been mentioned. The possibility for this site to had been a potential vicus functioning as a mansio or even a mutatio gains relevance when correlated with the findings in Vilela (ID 3012402), a site located at an inferior level where two Milestones have been found together with a statue pedestal uncovered at the graveyard nearby the churchyard and dedicated to Genio Viriocelense (Carvalho 1998). This particular item, reused as a holy water sink indicated the name of a Roman site – Viriocelum – that we

The Shaping of Mountain, Valley and Coastline Settlement

admit may correspond to the only site with significant Roman remains nearby Via Nova: the Vila Cova site. The site size is difficult to estimate, however nothing indicated with safety that it actually was extended up to Vilela79.

The difficulty in estimating area sizes of remains is of particular importance in this context, especially as it would be of great relevance in order to assess the type of settlement that seemed had been Viriocelum. Available data led us to attempt to contextualize the dedication to Génio referring to a location that was mandatory of a statue, as we are certain that the item referred to a pedestal.

Is it licit to think of a vicus, even though the site is located only a mile further from the Cividade de Biscaia mile? In truth, the only three inscriptions mentioning Génio with toponym epitets in the Northwest were all found within the bracaraugustanus conventus and may be associated with key secondary settlements as it is the case of Genius Tongobricensium mentioned in a Tongobriga votive altar (ILER 660), of Genius Tauranceaicus honored in Estorãos, Ponte de Lima (ILER 659) and Genius Laquiniensis referred to in a monument found in Caldas de Vizela, Guimarães (ILER 658). The Vilela inscription revealed that we are facing a dedication with a clear Latin onomastic (LVCRETIVS / SABINVS / GENIO VI/RIOCELENIS| (hedera?) / V(otum). S(olvit). L)ibens). M(erito).

Figure 19. Genius Viriocelensis pedestal (Carvalho 1998)

On the other hand, it is important to take into consideration the operative accountability of ideal distances between infrastructures supporting the main road itineraries and established by the cursus publicus, and adapt them to the different geomorphological realities that are crossed by the circuits. Thus, we should take into consideration other variables, namely the very own dynamic of populations organization that could group in small or wide cores and have been associated with the exploitation of agricultural, forest and pastoral resources that in turn complement and support the road trajectories.

Campo do Castelo, Terras de Bouro (ID 3100103) (mutatio of mile XV) The site is located on a hill registering 530 meters of altitude nearby the road passage. In the highest area, which is severally forested no traces of wall lines, trenches or splopes have been found. In the lower area, which was tilled at the time when we conducted research on site, there were several tile fragments alongside Roman common pottery. The site may have been a mutatio located on mile XV, which seemed to be justified by the need to have a stop over after a significant steep rise found beforehand.

79 The fact that some tegulae was found on a house nearby the church, however the fact that these remains were ‘shown’ to us rather than recollected by us prevented us from proposing its extension towards the Vilela church. The existence of these materials may be explained by land slipping that can occur on a site boosting accentuated slopes. It could also be explained by the transport of lands to the site.

S. Sebastião, Terras de Bouro (ID 3100605) (mutatio of mile XVI / XVII ?) 123

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

The S. Sevastião mutation, located at 450 meters high, was detected when road works were conducted on nearby areas. The works severely damaged the existing remains and brought to light the remains of a wall and construction ceramic fragments together with Roman common pottery and sigillata. The identified remains were similar to the ones uncovered from an infrastructure excavated in the Bouça da Mó site.

Chãs de Vilar / Salaniana, Terras de Bouro (ID 3100602) (vicus of mile XVIII) This was a Roman settlement located at 460 meters of altitude and registering wide artificial platforms and a great quantity of tegulae, tiles ad Roman common pottery. The site is located on mile XVIII, which was signalled by a Milestone from the Titus-Domitian reigns and has been catalogued as coming from Valfoios. In the Antonine Itinerary, the Salaniana mansio appeared referenced on mile XXI; however this mile was located in Travassos (ID 3101601) where a Elagabalus Milestone existed. The latter site does not present the topographic conditions required to house a site of this typology. Conducted prospections in the region led us to believe that although the discrepancy in the miles existed, the Salaniana mansio was indeed located in the Chãs de Vilar site.

Covide, Terras de Bouro (ID 3100804) (mutatio of mile XXVI) Covide marked mile XXVI. However a few Milestones were uncovered on this site, one of which was located on a porch of a private house in the area (CIL II 4810). A votive altar was also found in the property at the time when the building works of the current football field were conducted. The piece had no inscriptions. In the Barzes site (ID 3100801), in Covide, graved stones and column fragments were also found. We highlight the finding of construction stones, wall remains and column fragments uncovered in the foothill of the Calcedónia mountain range (Silva 1958: 58).

Adro Velho or Sagrado / S. João do Campo, Terras de Bouro (ID 3100302) (vicus? of mile XXVII) The site is located in S. João Veiga on a smooth area nearby the road trajectory, 10 miles away from Salaniana and 11 miles from Aquis Originis (Rio Caldo). An inscription was found on site dedicated to an indigenous deity entitled Ocaera. This item may be related with a temple or with a public building as infrastructure walls were found in excavations conducted in 1992. This archaeological work uncovered the foundations of a Christian temple, probably dating back to the low Middle 124

Ages. The site was not fully excavated making it difficult to correctly interpret it.

Bouça da Mó, Terras de Bouro (ID 3100309) (mutatio of mile XXX) The Bouça da Mó site is located on a smooth area near the river. For the majority of the year, this site is under the waters of the dam that also submerged Vilarinho das Furnas. On site, two Milestones were found one of which in situ. The excavation of the mutatio revealed a rectangular building on top of a small hill existing nearby the road and located on the left margin of the Mó watercourse. The walls were of granite. Inside the building, there were remains that suggested the existence of pillars supporting the roof. The roof was made of tegulae and imbrices if we consider the great quantity of these materials detected during the excavations. The southeast part of the building conserved the remains of a small compartment believed to have been a taberna. The materials date back to the first century and the end of the second century (cfr. Image).

Data referring to the described mountainous sites together with the cartographic analysis revealed some significant differences when compared to valley settlements associated with the Cávado and Homem rivers. In the areas registering lower altitudes we found remains of a valley occupation, which, according to epigraphic data, allowed the assumption of existing Roman settlements prior to the opening of the Via Nova. However, these remains are not abundant or sufficiently accurate to fully understand the type of exploitation units existing at the time in this area. These events may be due to prospection deficiencies. We will order the existing Roman remains starting by the ones found in the valley area.80 In the site of the current Caldelas settlement (ID 3010601), located at 150 meters of altitude, found remains support the assumption that this was an important Roman site. In the beginning of the twentieth century, several elements corresponding to a Roman necropolis were found in the Grande Hotel site. To Caldelas are also attributed some Roman architecture elements, mainly two columns shafts and three column bases in granite. Finally, in 1803, during the works to construct the current thermal complex, two votive altars dedicated to the Nymphs were found. In one case, the commemorator had clear indigenous onomastics, which seems to document an occupation of this site since the very beginning of the Roman presence in the region. 80 We excluded from this work and from cartographic information an inscription presenting a rather challenging reading. It is a finding coming from Paço located in Lago (ID 3011302) that reads: Fasn(i)u(s) Novov(i) m(erito) b(ona) v(oluntate).

The Shaping of Mountain, Valley and Coastline Settlement

Taking into account the group of mentioned remains for this particular site, together with the features of the water sources still exploited today for therapeutic aims, we consider this to have been a thermal vicus. On the other hand, it is likely that a secondary road linked this settlement with the Via XVIII itinerary, and from there linking it with Bracara Augusta. The site is located nearby the Caldelas Hill-fort (ID 3010602), although it is impossible to attest the direct displacement of the population from this settlement to a new Roman site, as there are no safe chronological indicators with regards to an occupation of the mentioned Hillfort. On a site known as Campo da Igreja (ID 3011101) located at around 90 meters of altitude, Roman remains have been found on a smooth slope area. Amongst them, there were abundant tegulae, manual grading wheels and some stone construction materials. The findings may indicate the existence of a Farm. In Cerdeirinhas (ID 3010901), on a smooth slope area with 100 meters of altitude remains from a rectangular grave covered in a stone slab has been found. Inside, there were two ceramics containers and one glass jar. In the Anjo da Guarda chapel, in Prozelo (ID 3011702) two voting altars were uncovered: one anepigraph and another consecrated to an indigenous deity, (…) / CAND /EBERO/NIO CA/ EDVRA / DIO, dating back to the first century. At the Ribeiro site (ID 301301), in Lago a funerary votive altar was found with the following inscription MARIV(S) / ADRONI (filius) / AN(norum) VI (sextium). This inscription may be associated with a potential Villa, as suggested by Manuela Martins (1990). Beyond the Hill-forts and the sites, already described, associated with the road, the only references we have in the areas of higher altitudes relate to two votive altars. One of the votive altars came from Santa Maria de Bouro (IF 3011901), however the exact location from where this item was from is unknown. It was a monument where only the following can be read PRO SALVTE / …VM / …/… The other known votive altar was found in Chã Grande (ID 3012001) in a high area of the Bouro mountain range located at around 520 meters of altitude. It was a dedication to Jupiter Optimo Maximo by Maternus, who has been mentioned in literature as a Romanised indigenous: MATE(RN)/ VS POSVI(T)/ARVDA(M) / IOVI OPTI / MO MA/ XIMO.

1.2 Limits and scales

1.2.1 The road as an organizational element of a ‘spine’ settlement network

The construction of the Via Nova nearly 70 years after the foundation of Bracara Augusta certainly mobilized part of the army, whose camping site may have been located in Aquis Querquernis. The establishment of this itinerary may have prompted the displacement of people, who were previously located in Hill-forts as for example the Torre-Carrazedo or Castro do Caires and moved towards the open settlements of Carrazedo and Cividade de Biscaia, sites that adopted wider functions than just the ones of a simple mansiones and mutationes. In fact, the areas where these settlements were located allowed their reasonable expansion, facts that suggest these sites may have housed a significant population dedicated to the several activities required by the functioning of these infrastructures, namely agricultural tasks alongside the duties supporting the road circulation. If it is true that the sites were far too close in terms of road canonic distances, they were sufficiently apart to be suited areas for the managing and exploitation of agricultural resources. In mountainous regions, simple mutationes may coexist if they were exclusively focused on supporting traffic (Bouça da Mó and probably the S. Sebastião case). Alongside performing roles of supporting circulation they may have evolved into wider settlements in areas with potential for other activities, as for example could have been the case of Vilela, Covide or S. João, in the Terras de Bouro municipality. A key assessment in future works is related with the sites distribution, which not always obeys to the logic distances required by the road support infrastructures. Thus, it is important to consider not only traffic practical conditions in steep hills areas, but also, in some cases, the existence of previous settlements housing populations that could have been displaced to open settlements without that displacement occurring under a regulated pattern of pre established distances, depending prior on the circulation supporting needs. When opening a road, it was important to establish infrastructures needed to support traffic and the cursus publicus, however, this would not necessarily demand that all the area throughout the road itinerary was ordered according to compulsory stopovers distances. In our analysis, we consulted on several occasions the works by Fermín Perez Losada (2002) for the Galicia region. The author established three basic types of agglomerates ‘between the city and the Villages’. One of those types included small habitats encompassing some houses, eventually with one fanum or small thermal associated structures. We can perhaps include in this category the S. João do Campo site, in Terras de Bouro. Some sites classified as vici do not follow the roles suggested for these types of establishments. The affirmation

125

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

that vicus were not focused on agricultural exploitations, a role that would be reserved by Villas, may eventually apply to some regions but not to all, as it likely happened in the one under study. In the case of the Entre-Douro-e-Minho region, it is possible that these events took place in areas densely populated and with a considerable number of agricultural exploitation units included in these two categories, although they are difficult to characterize. Nevertheless, it doesn’t seem that this assumption may be applied to all mountainous regions as the one we are analysing. In fact, the observed area throughout a 34 miles road trajectory allowed the verification regarding the lack of remains connected with the presence of Villa or even of units of smaller sizes, such as Farms. Under these circumstances, it is important to consider the possibility that some identified settlements may have supported the circulation traffic while exploiting the agricultural land potential of their surroundings and/or exploited wider pastoral resources. In most of its trajectory, the Via Nova conditioned the structure of the settlements. This has been evident and after listing all the data we do not have any sites located at a far distance from the itinerary. Although this situation may also emerge from research gaps, we believe that the fact this region registered a low population density and enabled the trajectory to create a type of settlement pattern cartographically similar to a ‘spine’. Despite the fact that we believe further in-depth research needs to be carried out, we believe this type of organization to be accurate considering the available data.

Alongside these mechanisms, the municipalisation of urban centres registered lesser importance in the JulianClaudian dynasties corresponded to a political and administrative strategy prompted by the economic growth of the regions that being peripheral in Augustan time became important and sensible areas of control, affirmation and cohesion, throughout less than half a century. The chosen area where to establish the Via Nova registered significant differences in terms of overall organization and chronology. We may resume them in two distinct realities. On the one hand, the valley region where settlements are scattered registered a chronology that goes back to the initial Romanization of the region. Their features are very similar to the ones found in areas closer to Braga, in the southern Cávado and in the surrounding eastern and western areas. We do not have any conclusive data regarding the chronology of most of the region settlements. In some cases, as for example the Castro Lago, it is possible that the abandonment of this centre favoured the installation of Roman habitats. In other cases, as for example the Castro de Caldelas, we do not have any credible information supporting an assumption about population transference to a potential vicus alongside the dissemination of its inhabitants to smaller farms or other agricultural establishments. We only know that nearby there would have been a Roman site linked with the exploitation of thermal waters. The other remains uncovered in the valley area only allow the assumption over a Roman type of sites that seems to have been clearly prior to the opening of the Via Nova. In this context, it seems relevant to understand the changes occurred in the Caires area.

1.2.2 The Romanization of a mountainous territory

After the Roman pacification and the administrative reorganization of Hispania conducted in Augustan times, the Flavian period is normally associated with a cohesive moment and with the consolidation of the imperial power. The concession of ius latii given to free citizens, the municipia promotion of some centres as Aquae Flaviae, Bergidum Falvium and Iria Flavia and the monumentalization programme in several Hispanic cities with different dimensions and political and administrative relevance, as for example Conimbriga, Bracara Augusta or Tarraco are some of the main features of the Flavian times. In fact, the Flavian dinasty administration developed diverse actions aiming to enforce the Roman power and consolidate the policies of their predecessors. The construction of the Via Nova should be understood within the framework of the consolidation of Roman power and should assume a similar weight to the one given to the monumentalization of cities or to the provision of ius latii to a wide number of provincial inhabitants. These events consecrated not only the need for peace but also of appeasement, integration and administrative efficiency. 126

At the site, there is a Hill-fort located on an area with easy access to the valley, and the existing remains corroborate the Romanization process. The lack of stratigraphic sequences enabled the assumption over the occupation of the site throughout the first century. This particular Hillfort may have been a reference point and had a supporting role in the opening of the Via Nova, probably of short duration due to the inevitable transformation of the surrounding spaces. In fact, to the Hill-fort of Castro do Caires will certainly succeeded an open site located on the valley region, closer to the road and likely located between the road trajectory and the Frecheiro site. Nearby, a statue pedestal dedicated to Genio was uncovered. The difficulty to correlate this group of references is aggravated by the fact that we are not able to assess with precision the extension of the site established on the foothill of the Castro de Caires.81

81 This is a difficult task to achieve, if not impossible, only with simple prospection works. Due to the existing smallholding property system and the presence of the Frecheiro village with several houses and small farms and the fact that the past few decades have witnessed

The Shaping of Mountain, Valley and Coastline Settlement

However, it seems pertinent to conclude for this region the existence of new settlement patterns of Roman origin articulated with the persistence of Hill-forts throughout the first decades of the first century. The valley area was, therefore a transition area and a ‘boundary’ of the space presumably controlled by Bracara Augusta since the beginning of the first century. The construction of the Via Nova prompted circulation and mobility processes together with the development of new habitats. The region registering higher altitudes witnessed the establishment of a group of settlements that in their vast majority are difficult to assess chronologically (Hill-forts, watchtowers or castles?), and estimate with a minimum credible occupation chronology. The small number of Hill-forts where one or two wall lines were identified and the presence of ceramic fragments from the Iron Age and Roman times (the Calcedónia case, for example), makes it very difficult to establish any type of relationship between the road trajectory and the habitats associated with this itinerary. The Calcedónia example is suggestive, as despite of the several references to its location, a simple visit to the site starting in Covide, where the road is identified, implies to win a slope too abrupt, that we can not easily combine with the necessary supporting conditions an itinerary would require. The interpretation of the group of references regarding settlements will necessarily have to be reassessed in future archaeological and historical research. Thus, we are left with a significant group of data that enables an assessment over the potentiality of what the opening of a Roman road would entail in terms of space and surrounding habitats. First and foremost, it seems evident that a possible transference of inhabitants from some Hill-forts to open settlements nearby the itinerary took place. An identical phenomenon occurred in Galicia and was studied by Férmin Perez Losada (2002). Equally, we can also observe an organized settlements pattern favouring open agglomerates, instead of agricultural exploitation units of the Villa and Farm type. Moreover, the geomorphological features of most the region may help explain this trend as these type of areas are more suited to open habitats, that were plurifunctional and where the traditional activities related with population survival corresponded to an agricultural, grazing and forestry exploration system, to which handcrafts and services provision activities related with the circulation support may be added. A region not registering these patterns is located in the valley region of Amares where it is possible that agricultural exploitation units existed. This is attested by the existence of the Cerdeirinhas necropolis

the construction of new infrastructures, prevents us from acknowledging the real size of this site.

127

(ID 3010901) and by the Campo de Igreja site (ID 3011101), believed to have been a small Farm. The majority of hill-fors seemed to have endured abandonment prior to the Roman times. In some cases, identified habitats may not even have witnessed an occupation after the Early Iron Age period. On the other hand, some sites believed to have been Hill-forts correspond to medieval watchtowers, castles and even medieval living spaces. Under these circumstances, there doesn’t seem to have been an active role of pre-Roman settlements in organizing the land in the follow up period, although in some cases a reassessment of data is needed, both in terms of typology and site chronology. The region epigraphy is overwhelmingly dominated by the group of over one hundred Milestone detected throughout the 34 miles of the road trajectory under study. To this group of items are added the ones found in the Spanish territory that have been the object of gradual publications and updates, reason why we will not list or transcribe them in the present work. However, it is important to draw the attention towards some aspects related with the quantity of the found Milestones. In fact, there doesn’t seem to exist a mile without a Milestone, and equally impressive is the quantity of Milestones found in some miles. The phenomenon has been understood as an imperial propaganda process (Encarnação 1996-1996). This idea seemed acceptable given the abundance, if not ‘redundancy’ of these markers regulating circulation. It is certain that the abundance of granite stone present in the region prompted a transformation of some areas of Via Nova into real Milestones ‘factories’, precisely in the sites where we can find most of the items. It is also true that the preservation conditions of part of Via Nova, especially in areas registering low population densities, prevented the reuse of those Milestones, preserving not only the road layout, as well as Milestones. It is not surprising that most of the monuments found in situ or ‘nearby the original site’ were located in low populated areas, in sites less suited for transport, or too far away from centres, thus not able to serve the collections of antiquities. An example would be the Milestones collection placed around Sant’Ana chapel in Braga by D. Diogo de Sousa. All these may constitute explaining arguments for the abundance of Milestones in the Via Nova, although they are not sufficient or able to explain their exceptional quantity. Could the ‘signature’ or imperial propaganda observed in the Milestones quantity be explained by the lack of urban centres over a wide territory? In the urban centers, the promotion and maintenance of the imperial power image has been inscribed almost naturally by the fact they were sites registering larger people circulation, by the existence of magistrates, by the abundance of currency or by the ability to formalize political changes in statues and other votive monuments. In a low populated territory as the one under study, one that is crossed by a road, what were the

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

ways the imperial power was manifested or how could the populations paid their homage? If the abundance of Via Nova Milestones represented a way of manifesting imperial propaganda, this also seemed to equally reflect a very specific way of promoting the imperial cult in a specific region, both in terms of population centres or the function the roads had as circulation axes for high employees serving administrative policies.

their indicators point to later chronologies than the ones normally attributed to them. Furthermore, the development of boundaries in medieval times should also be taken into consideration. Although the territory seemed intact when considering the impressive preservation of the Via XVIII itinerary layout and the state of its Milestones, the need to establish boundaries must have demanded a spatial reconstruction, which clearly was detached from the staple centred Roman perspective.

2. Valley and coastline settlement

Considering the valley and coastline areas, we verified the inexistence of urban settlements and of a network of vici that would hierarchically organize the landscape. The contrast with other regions within the transection under study is evident. Around Bracara Augusta, the key issue has been linked with the direct relationships between the city and the rural surroundings. On the mountainous areas the settlement pattern was clearly articulated with the Via XVIII itinerary. Figure 20. Via XVIII with some milestones in situ

The remaining epigraphy revealed a framework that can be compared to the one present in areas closer to the city, and included homages to indigenous deities and Roman gods. This group of monuments alongside the existing Funerary epigraphy allowed us to acknowledge the presence of indigenous items concentrated in lower populated areas, closer to Bracara Augusta and in the region of the Cávado and Homem river valleys. On the other hand, classical pieces with a later chronology were uncovered in areas near the road, with the exception of the consecrated vow to Jupiter Optimo Maximo, an artefact not possible to contextualize. Perhaps the most important conclusion we can draw, linked with the attempt to analyse available data related with the settlement types of this particular region, is our conviction that all the spaces under analysis require a long-term analysis. These were areas where the time was ‘slower’ and where landscape and habitats changing required different rhythms in comparison with the ones occurring in more urbanized areas, as for example the ones occurring in areas closer to Bracara Augusta. Furthermore, the low population density of these territories, where we include for example Vieira do Minho and Serra Amarela, demanded further diachronic studies as the one conducted by Luis Fontes for the Serra Amarela region. The marginality of the mountain areas and their boundary character increase their potencial research interest, particularly averse to too rigid chronological and historical demarcations. One of the issues that best address this challenge has been linked with the assessment of Hill-forts and the fact they seem inexistent in these regions, or perhaps due to fact 128

We may risk stating that a ‘marginal’ region status can be applied to the coastline area, although when it comes to the Romanization process, the regions considered marginal are traditionally identified as mountainous areas or the ones directly related with the mining exploitation. To this ‘marginality’, normally linked with the rarety of resources or with a structural fragility of the settlement framework is opposed the cities reality and in the lack of the latter, the infrastructures of the Villa type, normally understood as extensions of urbanitas in the ager. Research has been shifting this overview and encompassing an interest in combining other perspectives in order to surpass a compartmented and hierarchized preconcieved territory analysis, as defended by M. Ruiz del Árbol Moro (2001). Thus, there are ‘marginalities’ that are relevant to explore in their territorial sense: the ones shaped by data density and heterogeneity, interpretation not compatible with the idea of discovering one or more macroentidades that support the logic of the overall settlement logic. These are the cases of the valley and coastline regions that we will now analyse. Due to the nature of available data, both regional cases justified a need to make premises more ‘plastic’ in order to help explain the rural history. The majority of Romanization remains allowing the record of a dense and continuous settlement can be found in the valley regions. The coastline region presents, on the other hand, a settlement spaced, rarefied and without networks, that is, without centres mobilizing any type of concentration. Under these circumstances, where do we find organizational logics? Our approach sequence led us to observe the action of several factors that might have intervened within the overall rural landscape structuring. Furthermore, the most prominent image of this process has been the proliferation of a wide array of settlement solutions.

The Shaping of Mountain, Valley and Coastline Settlement

2.1 Analysis frameworks

The region under study is dominated by floodplains and alluvial soils (Map 15). The region is bounded on its northern side by the Neiva River, by the Este River on its southern part and by the Atlantic Ocean 82 on the west. It administratively corresponds to the municipalities of Barcelos, Esposende and Póvoa de Varzim. From a geomorphological point of view, it corresponds to a slightly hilled area, dominated by open valleys and erosion alveoli that explain the emergence of wide areas suited for agricultural processes and shaped occasionally by elevations that are more or less significant. The bigger reliefs in the region are located in the interior areas, with a highlight to the Serra da Franqueira (298m), the Alto da Vaia (285m), the Monte de Maio (214m), the Monte do Castro de Sequeade (285m) and the Serra de Airó (413m). A significant mark in the overall landscape structuration is the presence of the river Cávado that, up to Barcelos, runs NE towards SW. Here it shifts its course towards E/W ending on its mouth with an orientation of N/NW. Downstream from Penida, the river runs wide and open, and however on the upstreams its course is narrow. This event prompted the installation of the ‘Central Hidroeléctrica da Penida’ in 1914 that used this ‘accident’ on nascent edge of Areias de Vilar borough. This strangulation of the Cávado course prevented the river upstream navigation and shaped the end of its circulation from the sea. The varied trajectories of this river throughout the Quaternary are corroborated by the deposit remains on its terraces that present diverse altitudes in both margins of its course (Teixeira et alii 1969: 7). Furthermore, the region is covered by a network of watercourses of dense stream that explains the nature of the soils and the features of the settelement landscape. Dominated by mild slopes and alluvial soils, this region prompted excellent conditions to agricultural practices throughout time. Around 70% of the current Barcelos territory is considered flattened land. The Neiva River with a widing profile establishes the northern boundary of the region under study. It is outlined by a series of reliefs: in the western part of the region, between S. Gonçalo and Arefe and in the eastern part by the Serra de Lousado. In its narrow course, it has rocky slopes with hillside lands with terrains suited to agricultural practices. From Alto da Portela, the river runs in an open valley and presents a U shape. Between the Neiva and Cávado rivers, there is a group of reliefs facing west that bound a ‘wide depression area in Barcelos’. These reliefs have altitudes that range between 82 The Ave River is not a boundary element. Despite the fact that the Póvoa de Varzim municipality is included in this region, Vila do Conde and Vila Nova de Famalicão are not.

129

300 and 500 hundred meters of altitude. This ‘arch’ ends with the reliefs of Fonte Grande (312 meters) and the Lousado mountain range (312 meters), on its northern side. The two types of rocks presented in the region are eruptive and Silurian rocks, the former are represented by several types of granite performing 70% of this rock typology (Teixeira et alii 1969: 8). The region encompasses important mining resources including gold, tin, tungsten, silver and iron. In the Barqueiros region, the small portions of intercalated quartz enabled gold exploitation in Roman times, namely in the Lagoa Negra area. In the Airó mountain range there were perhaps gold exploits with a chronology that could go back to Roman times. Furthermore, in the coastline region there were mineral resources that prompted metallurgical activities, as for example in the Cividade de Terroso Hill-fort (Gomes 1996). Clays and kaulins perform an added value within the region, namely present in two key areas: the clay terrains covering the area between the current city of Barcelos and Prado and the wide region of Alvarões located nearby the Neiva River and corresponding to a wide sedimentary and alluvial deposit (Teixeira et alii 1972:11). Other than these wide areas, there were other smaller clays zones between the Cávado and Neiva rivers. The overall understanding of the coastline area is key in explaining the orientation and fixation sites where settlements were built and enables the interpretation of occupational gaps. Although the stability of the medium course of seawaters has around 5000 years, there are significant variations in the coastline that are also important to synthetize. The mouth of the Neiva River, referring to a low and sandy area, is bounded by a fossil cliff on the left margin, while the right margin refers to a wide higher area – the Rodanho terrace that extends itself from the Neiva castle to Viana do Castelo (Teixeira et alii 1972; Machado 938: 1-4). In some areas around the Cávado estuary it is admitted the existence of lagoons on its northern side, namely in Cepães (Marinhas) and in Moinhos River (Marinhas, S. Bartolomeu do Mar) and also on the southern part, in the region of Lagoa da Apúlia (Granja 1989). The existence of lagoons and insalubrious terrains may explain the lack of occupation processes. In southern Apúlia, the current Póvoa de Varzim municipality, there were silting processes in the coastline that transformed old areas only occupied in sandy sites. The distribution of settlements and the excavations conducted in some sites confirm this. In the Vila Mendo site, where an excavation took place, one can find a perfectly conserved Roman building in the dunes, presenting walls with two meters high (Gomes 1996). The silting process is responsible for the well known ‘’campos de masseira’’, clearing zones made on dunes by digging the sand up to the level of agricultural soil. This was a process accelerated at the end of the nineteenth century and the beginning of the

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

twentieth, when the demographic pressure demanded the conquest of new agricultural areas by the coastline population (Gomes 1996).

2.2 Entities and relationships

From the 159 references to recollected archaeological remains, we decided to work only with data we considered reliable. The selection process and that fact that we deemed paramount to deal with the disparity of records, justified our decision to assess the global nature of data statistically. Therefore, we utilized for this region a group of descriptors used by François Favory and Van der Leeuw to analyse the ancient settlement in the Reno valley (1998), although with some modifications and diversified aims. These same descriptors were also used for the region of Beaucairois (Favory, Fiches and Giradot 1987-1988: 6786) and performed a type of approach to rural settlements aiming to surpass the traditionally used methodology in the studies of agricultural landscapes. In sum, the aim of those projects was to conjugate a series of data (artefacts, furniture, activities and chronological indicators) with a site typology reduced to its minimum (wide, medium and small habitats), providing complex statistics analisis to correlate archaeological data with landscape and geomorphological information. One of the results consisted in a landscape hierarchy in terms of diffusion centres (Villas, oppida and vici) and dependent establishments, or others in any type related with this diffusion. Part of the concepts underneath these works influenced our perspective regarding the organization of data in entities, relationships, limits and scales. However, the descriptors mentioned above and due to a question of methodological viability required a different application. Moreover, the referred projects and their end results were developed in small regions; they included precise data, intensive field walking surveys and a wide and multidisciplinary group of teams. Without an intensive application of all the methodological and conceptual background used by the mentioned authors, it seemed impossible to establish suitable study comparisons, mainly when sufficient data was not available, which made it impossible to apply several of the used descriptors, especially the ones of chronological nature. For the region under study, we have some doubts related with the possibility to reduce the forms of explanation of the ancient settlements and of the rural world to a model based on relay centers as well as on the objects of this diffusion. Thus, we limited ourselves to the use of some statistical analysis, with the necessary calibrations in order to feature and highlight the nature of the available data. We began by excluding sites lacking georeferences and only kept 145 references, to which we applied descriptors 130

related with found construction materials and remains that were part of the economical and social activities of its inhabitants and corresponding to the following graphics. The more impressive result presented here has been the high percentage of sites where the only indicator was the presence of tegulae. These trends have raised several challenges in terms of identifying a site category or acknowledging the relationship between habitats. This has also been a common challenge in any work focused on this region, event also highlighted by the author Carlos Alberto Brochado de Almeida for the area under study (1996; 2003). With regards to the existence of material indicators related with the inhabitants way of life83 (i.e. common ceramics, fine wares, glass, coins, inscriptions), we verified a great percentage of sites without any of those remains followed by sites only known for their common pottery findings. The other remains (i.e. fine wares, glass/coins or inscriptions) match an almost equivalent percentage. Finally, the lack of remains with regards to the type of activity that was carried out at the time is equal to the overwhelming percentage of sites only known for the presence of tegulae. Other than this percentage, the vast majority of sites indicated activities linked with agricultural exploitation and storage. Having these results as the base, we deemed preferable to exclude references limited by the scattered presence of tegulae due to the fact that no exact chronology or site typology was easy to attribute. Hence, we dealt with 119 references.

2.2.1 The fortified settlement

The need to provide exact chronologies to Hill-forts generated, for this particular region, some paradigmatic examples. It enabled the detection of several organizational and reorganizational phases throughout the Iron Age and explained, among other things, the occupational processes. However, these processed when simply mapped prevent us from obtaining clear readings regarding the overall territorial settlement history. This first phase of attributing chronologies to sites included in the final group of Hillforts located in the region under study helped the acknowledgment about the different basic chronologies linked with the degree or intensity of the hiil-forts contact with the Roman world.

83 We chose to utilize the concept by Fraçois Favory and Van der Leeuw (1998) when referring the type of information emerging from this type of data, and we prefer it to ‘mobiliário’ (mobiller), a term provided by the authors in this chart. However, it seems dubious when used in Portuguese.

The Shaping of Mountain, Valley and Coastline Settlement

We will now present the group of Hill-forts included in categories A, B and C.84 Hill-forts of A category Outeiro do Castro, Airó, Barcelos (ID 3020501) Located on a hill dominating the confluence of the Covo and Cávado rivers, this site has 222 meters of altitude and four defensive walls. The Roman occupation on site has been corroborated by the findings of abundant ceramic fragments and of a votive altar with an inscription that cannot be read.

Crasto, Aldreu, Barcelos (ID 3020601) Located on a hill with 260 meters of altitude, this site had two wall lines and a trench. There are remains of houses and abundant ceramic fragments from the Iron Age period. Tegulae fragments, imbrices and amphorae indicate Romanization. The site also witnessed a medieval reoccupation.

Monte de Lousado, Barcelos (ID 3027402) This settlement is located on top of the Lousado mountain range, at 309 meters of altitude and presents remains from two wall lines and circular houses. On the surface, Iron Age ceramics can still be recollected alongside with some tegulae fragments.

Carmona, Barcelos (ID 3021202) This is a wide settlement located at 303 meters of altitude showing three wall lines, two trenches and a earth wall. A fourth wall line made of stone seemed to have divided the living area in two parts. Inside the acropolis, a wide building presenting an orthogonal plan has been detected on the same site from where two votive altars have been uncovered. The settlement presents a clear proto-urban organization and circular house, with the widest area located at the acropolis. On site, there are remains of grading wheels and of a well. There are abundant Iron Age and Roman pottery fragments alongside with glass and coins (one from the 1st century and some bronzes from the 4th century), artefacts, and fragments from two votive altars with inscriptions. On the foothill, there was a Roman grave, remains of tegulae and two graves excavated in rock. In Tarcisio Maciel’s (2003:114-119) opinion, this site performed the core centre of the future ‘Terra de Neiva’. Monte da Saia, Barcelos (ID 3022201)

A site located at 303 meters of altitude with three wall lines. The findings chronology indicated an occupation that goes back to the Final Bronze Age and goes all the way up to the end of the first century AD. The presence of a indigenous thermal infrastructure, and of two rocks representing human figures wearing Roman clothes, a smelter deposit and a treasure with gold pieces make this site famous within the group of regionalHill-forts.

Monte de S. Mamede, Barcelos (ID 3023202) Thos is a Hill-fort located on a hill with 409 meters of altitude. There are houses, two wall lines and several construction materials remains. The described artefacts included indigenous ceramics, tegulae, amphorae fragments, one Hispanic sigillata fragment and some bronze coins.

Castro Faria, Barcelos (ID 3023701) This settlement is located on the Northwest side of the Franqueira mountain range at 298 meters of altitude. Its occupation goes back to the Bronze Age and was active all the way through the Iron Age. In Roman times, it seemed possible that only a small sector of the site was occupied, as there are no indications of a significant reorganization of the living areas. In medieval times, the site was occupied by the Faria castle and kept this function up to the second Portuguese dynasty. There are remains of three wall lines and two trenches. From the wide variety of Roman remains, we highlight construction and domestic ceramics, ‘bracarense’ ceramics, thin walled grey pottery, Hispanic sigillata and amphorae fragments.

Citânia de Roriz, Barcelos (3025401) This is a wide settlement located at 324 meters of altitude. It has three wall lines with the possibility of a fourth having existed on the western side. The site witnessed a long occupation from the Final Bronze up to the Roman times. There are traces of streets, circular and rectangular houses and abundant ceramic fragments. On site, a Galician warrior head was found together with an indigenous thermal structure located in Stª Maria de Galegos foothill. The Romanization of this site is attested by the presence of construction ceramics, common ceramics, Hispanic sigillata and amphorae fragments (with a highlight towards Haltern 70). Two small bronzes artefacts from the emperor Constantine led some authors to assume this site was still occupied in the fourth century.

84 We excluded from this list the settlement known as Crasto located in Navais, Póvoa de Varzim (ID 13130901) that was never excavated. Its reference is based upon the existence of tegulae.

Picoto dos Mouros (ID 3028101) 131

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

A Hill-fort located at 226 meters of altitude with two wall lines and a trench. Circular and orthogonal houses are still visible. The remains consisted of indigenous ceramics, Roman common ceramics, amphorae and tegulae fragments. Conducted excavations on the site enabled the assumption over the abandonment of this Hill-fort in the first century and its reoccupation at the fourth century, at a time when rectangular houses were constructed.

ence of circular houses with a chronology between the first century BC and the beginning of the first century AD. Recollected artefacts revealed an intense Romanization of this site. Fragments of indigenous ceramics and Roman common and imported pottery were also uncovered. From the latter, we highlight fragments of amphorae and of Italic and Hispanic sigillata from the first century AD.

Srª da Aparecida, Barcelos (ID 3021201) S. Lourenço, Esponsende (ID 3061501) The Hill-fort occupies a hill with 200 meters of altitude. It has four wall lines, two trenches and one tower. The inner wall line bounds a small size and circular acropolis. Excavationss exhumed houses (circular, oblong, sub-rectangular and rectangular), circulation axes, pipelines and kils. The remains include materials indicating an occupation going back to the Bronze Age. On top of ceramics from the Bronze and Iron Ages, we highlight the finding of Greek ceramic fragments. The site Romanization has been attested by a wide group of material remains including construction ceramics, common ceramics and fragments of amphorae (Haltern 70 and Dressel 7-11), dolia, ‘bracarense’, thin walled pottery, thin grey pottery, southern Gallic and Hispanic sigillata, coins, fibulae and a votive altar consecrated to DEAE SANCTA(E) by ANICIVS. The best known occupation for this site is between the second century BC and the first and second centuries AD, although it is defended that there has been a Low Empire occupation. In the Middle Ages the site included a fortress structure.

This Hill-fort is located at 150 meters of altitude and has two wall lines and a trench. The remains included indigenous ceramics, tegulae, imbrices and some common ceramic fragments.

Crasto-Chorente, Barcelos (ID 3011301) The Hill-fort occupies a hill at 186 meters of altitude and had two wall lines. The group of detected houses has been located on the top of the hill, between walls. The remains included Iron Age ceramics, manual grading wheels and tegulae fragments.

Castro de S. Simão, Barcelos (ID 3011401) The Hill-fort is located at 236 meters of altitude and presents houses of circular shape, two wall lines and surface indigenous ceramics. The Roman presence is attested by the existence of tegulae, common ceramics and a Crispus coin dating back to 317-326. The mining resources within the region around this site included silver (Maciel 2003: 138-140).

Cividade de Terroso, Póvoa de Varzim (ID 13131201) The Hill-fort is located on a hill of a welldemarked landscape at 153 meters of altitude. It has two wall lines and an occupation chronology that goes back to the Bronze Age up to the first century AD. Excavations enabled the identification of domestic structures and a group of cist graves. Other than the abundance of Iron Age materials, there were some ceramics of Punic tradition. We highlight the existence of a metallurgical centre of relevance on site. The growth of the Hill-fort in Roman times is attested by the construction of important buildings of rectangular planning.

Hill-forts of B category

Monte do Facho, Abade de Neiva, Barcelos (ID 3020102) The Hill-fort occupies a small elevation with 170 meters of altitude and it has, at least, three wall lines. Surveys conducted on site revealed the exist132

Monte do Crasto, Barcelos (ID 3022502) The Hill-fort is located on a hill with 202 meters of altitude and it has two wall lines. Other than construction stones, one may find on the surface indigenous ceramics and some tegulae fragments.

Alto da Giesta, Barcelos (ID 3022903) A small Hill-fort located at 298 meters of altitude. A rampart is visible on site followed by a trench. The ceramic materials belong to the Iron Age period.

Alto do Senhor do Lirio, Barcelos (ID 3022901) This Hill-fort is located on a slope that has been severely destroyed registering 140 meters of altitude. Nothing exists from the defensive infrastructures and the remains included indigenous ceramics, circular grading wheels, some fragments of common Roman pottery and tegulae.

The Shaping of Mountain, Valley and Coastline Settlement

Castelos, Barcelos (ID 3022902) The hiil-fort is settled in a small elevation at 160 meters of altitude where it is still possible to witness one stonewall and two trenches. In the upper platform there are remains of circular houses and of construction ceramics with no indication of Roman or medieval occupation. The idea that this was the site where the Aguiar de Neiva medieval castle was located (Almeida 1993) has been challenged by Tarcísio Maciel (2003:141-143), who defended the castle was located instead at the Carmona Hill-fort.

er wall. Prospections carried out by Carlos Alberto Brochado de Almeida (2003) placed its foundation in the mid first century BC. The same author believed this was a ‘castro agrícola’.

Igreja /Mariz, Barcelos (ID 3024601) This is a small size Hill-fort located on a small hill at 61 meters of altitude. It has a slope wall followed by two trenches separated by another earth wall. The remains are mainly indigenous ceramics, tegulae, amphorae fragments and Roman common pottery. This was a ‘castro agrícola’ according to Carlos Alberto Brochado de Almeida (2003:173).

Stª Marinha, Barcelos (ID 3026201) A small size Hill-fort located on a small hill with 240 meters of altitude. Remains of circular houses can be seen and the small remains account for indigenous ceramics and few remains of tegulae and amphorae. In Carlos Alberto Brochado de Almeida’s (2003:173) opinion this was a ‘castro agrícola’.

Monte, Esposende (ID 3060103) The Hill-fort occupied a small hill with 90 meters of altitude and is currently destroyed. However, it endured an occupation since the Bronze Age period. Other than wall remains, the found materials were from the Bronze and Iron Ages (Maciel 2003:155-156).

Cividade de Belinho, Esponsende (ID 3060102) A small Hill-fort with two wall lines both on its northern and eastern sides. The known remains belong to the Iron Age, with the exception of some Roman amphorae fragments.

Monte de S. Felix, Póvoa de Varzim (ID 13130801) A small Hill-fort located at 208 meters of altitude. The oldest testimonies related with this site indicated the existence of three wall lines. Circular houses remains are still visible today. The surface material referred to indigenous ceramics and tegulae fragments.

Hill-forts of C category

Alto da Deveza / Giestal, Barcelos (ID 3023402)

Outeiro dos Picotos, Esposende (ID 3060701) This Hill-fort occupies a hill near the Cávado River and has been significantly altered by agricultural works and recent building constructions. On the western side, it is still possible to observe remains of a rampart, a trench and an exterior earth wall. Other than the river, a small watercourse also serves the site. The remains were mainly micaceous ceramics and Roman common pottery. In 1936, Teotónio da Fonseca referred the finding of a site with ‘tombs, bricks, ceramic vases, melting vases and coins, some of which were from the Maxentius emperor’ (1936, 49). The site chronology seemed to go all the way up to the Low Empire period. On the western foothill, there was a Roman road that crossed the Cávado River towards Barca do Lago. This was a ‘castro agrícola’ according to Carlos Alberto Brochado de Almeida (2003:173).

This Hill-fort is located on a small hill with 43 meters of altitude. This is a small size site with one wall line and a trench that is still visible today. The remains mainly account for ceramics belonging to the Iron Age and Roman tegulae fragments.

The group of presented data accounted for significant variations in terms of sizes and location topographic of the Hill-forts, which is visible on the cartography referring to the distribution of this group of sites. It is important to cross the references to these Hill-forts with further chronological available data, in order to reconfirm what has been described on chapter 3.

Picarreira / Castro de Carapeços, Barcelos (ID 3021801)

The first significant group of sites revealed abandonment processes still occurring in the Iron Age period, in the late second century BC. This group included the majority of sites in the B category, referring to the ones located on hills and slopes and registering medium to small sizes when compared to wider Hill-forts existing in the region and included in the A category. This group also included

The Hill-fort is located at 180 meters of altitude and includes a earth internally supported by anoth133

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

the Monte de Lousado Hill-fort included in the A category and the low altitude Hill-forts of Stª Marinha and Monte, in Esponsende.85

events should take into consideration diverse processes of the abandonment movements and of the population fixation phenomenon in new residential centres.

A different situation took place with Hill-forts including imported materials and registering an abandonment thought the late first century BC. We are referring to Monte do S. Mamede and Picoto dos Mouros (A category) and the remaining of Hill-forts listed in the B category namely Monte de Facho, Srª Aparecida and S. Simão, as well as, C category Hill-forts, as for example the Alto da Deveza.

The Cividade de Terroso case raised an interesting question in terms of the study of amphorae remains found on site. The fragments are dated from the second century up to the fifth86 and were articulated with the occupation of the lower platforms, namely the southern platform nearby outstanding agricultural soils located in the valley. This occupation may have resulted from the installation of the populations in this part of the hill in order to exploit agricultural and mining resources around the old settlement (Gomes 1996: 126). We are not aware, however, about the occupational organization in terms of status or the actual size of the population core.

A third group of sites presented Romanization indicators although they were incipient, as for example Roman construction remains or the existence of pottery of clear Roman typology but in small quantities. These habitats were likely abandoned in mid first century AD. The remaining Hill-forts may have been occupied until late first century or even afterwards. They share amongst themselves similar features and their differences are only related with abandonment chronologies. Some of these Hill-forts witnessed internal reorganization processes in the last century BC, an event that allowed us to consider them to have been oppida: proto-urbanization, rectangular houses and a vast array of diverse Roman materials. Moreover, they all share identical topographic locations in the crossroads of natural circulation pathways, as for example the Castro de Carmona, Faria and Monte da Saia that occupied strategic positions for the control of the coastline and watercourses areas, namely Terroso and S. Lourenço. Alternatively, they were located in key areas to control valley and road circulation in locations near Bracara Augusta, as for example is the case of Citânica de Roriz and Outeiro do Castro. In the group of Hill-forts registering an occupation after the end of the first century are included the Carmona and S. Lourenço, registering both well-known chronologies due to the fact they have been intensely excavated. Similarly to the Entre-Douro-e-Minho region, some Hillforts included in this latter group endured abandonment or at least, a significant population retraction process on the living areas in the mid or late first century AD. They may have had occasional occupation with wider chronologies in areas near the valleys. However, this situation should not be understood as a linear continuity of sites occupation as it may only mean the fixation of a smaller number of inhabitants with an economy and lifestyle integrated within the Roman patterns.

As stated in point 1.2 of chapter 3, the abandonment and dispersion process of populations towards the valleys was intensified in late first century AD. The analysis of these

The Castro de Faria may have witnessed a reduced occupation of its original area in the beginning of the Roman presence. It did not witnessed proto-urbanization processes or intense Romanization when compared to other sites included in the A category. The indigenous dispersion over the hill slopes as suggested by Carlos Alberto Brochado de Almeida (2003) corresponded to the Romanization of the Hill-fort and it was from these slopes that the land clearing of the valley took place. However, we do not fully understand how such dispersion took place in the lower platforms of the old Hill-fort of Faria. Nevertheless this dispersion demanded, as suggested in the previous chapter for the Entre-Douroe-Minho region, the organization of new habitat centres and not an effective continuity in occupation of the old Iron Age sites. The exemplification of chronological processes and the varied occupation phases may be assessed in the group of Hill-forts located close to the Neiva River, namely the Srª da Aparecida, Picoto dos Mouros, Alto do Sr. Lirio, Alto da Giesta and Carmona sites (Maciel 2003). Apparently, the Hill-forts of Alto da Giesta and Castelos have an occupation ending before the first century BC and their population may have been integrated within the Carmona settlement, as suggested by the latter internal organization. Thus, the abandonment of the referred sites and the consequent transference of the population seemed have been associated with a restructuring process of the indigenous settlement, as it is generically admitted to had occurred in the wide ‘citânias’ of the Portuguese northwest. They had wide population densities, as for example was the case of Citânia de Briteiros, Citânia de Sanfins and Cividade de Âncora (Silva 1986; Martins 1990). The Carmona Hill-fort seemed to have acquired a key relevance within the region, fact that may explain the long chronology of its occupation after the end of the first

86 The study of the Terroso amphorae identified the presence of containers with Haltern 70 shape together with a group of remains presenting varied chronologies. Included in this group are late republican findings, amphorae from late first century BC and first century AD and artifacts from the third and fifth centuries or from the second and fourth centuries (Paiva 1993).

85 In some cases, the abandonment processes may be explained by communities’ movements towards nearby habitats. Tarcisio Maciel (2003:189) put forward the hypothesis of this process having taken place at Castelos Hill-fort that registered an abandonment phase around the year 200 BC, when its inhabitants went to Carmona.

134

The Shaping of Mountain, Valley and Coastline Settlement

century AD. Thus, the site may be included within the Romanised sites category. In the fourth century it may have been reoccupied and in the Middle Ages it may have functioned as a supporting site in the ‘construction’ of Terras de Neiva. On another level, the Hill-forts of Srª Aparecida and Sra. do Lirio may have been discontinued in the late first century as they do not register any Roman remains. The Picoto dos Mouros Hill-fort was occupied up to the first century and may have witnessed a reoccupation up the fourth century. The following table lists the abandonment and reoccupation processes in some site in the Neiva river valley sites. They were studied by T. Maciel (2003) and illustrate the variability of events that may have occurred in other Hillforts of the Entre-Douro-e-Minho region, however are not thoroughly studied.

Site Picoto dos Mouros Alto da Giesta Castelos Srª Aparecida Sr. do Lírio Carmona

Occupation up to Beginning of the second century AD First century BC End of the second century BC End of the second century BC First century BC End of the first century AD

complexity demands chronological assessments that in turn, are mandatory of excavations, surveys and systematic studies of materials emerging from sites. Thus, the reoccupation processes from the fourth century onwards cannot be asses only by the numismatic findings that many times are reduced to two or three coins or to ceramics of late Roman chronology or even of PaleoChristian origin as interpreted in some case studies (Almeida 1996). It is important to study their construction background and their relevance alongside the wide historical processes of the late Roman and Suevi-Visigothic sites. The reorganization of settlement in the Middle Ages is associated with the reoccupation of some Iron Age Hillforts presenting differences in size, relevance and meaning as already suggested in chapter 3, when assessing the reoccupation processes and the correct use of the concept. In fact, some sites where previously have existed Hillforts may have functioned during medieval times as simple watchtowers or small fortresses without that demanding a real reoccupation of the previous habitats but rather a use of site areas offering the necessary topographic features to carry out new functions.

Reoccupation Fourth century?

The relevance of the Carmona Hill-fort as a core centre of the future lands of Aguiar da Neiva (Maciel 2003: 119) still needs to be further explained according to the meanings that the term ‘reoccupation’ may imply. The same may be applied to the Castro de Faria, a site where an important castle was located when the organization of the Terra de Faria took place from the twelfth century onwards. Apparently, this is a site lacking a proto-historic occupation.

Sixth century / Middle Ages

Table 10. Occupation, abandonment and reoccupation of some Hill-forts in the Neiva river valley.

The occupation of Hill-forts throughout long time periods is sometimes possible to be understood as a continuous process during all the Roman times up to the Middle Ages. However, it is mandatory of a thorough analysis. In all the studied cases, there was indication of this phenomenon including the S. Lourenço Hill-fort where this situation was also suggested (Almeida 1996). In some of the studied sites, the occupation up to the first century AC and a likely reoccupation from the fourth century onwards was suggested due to the findings of coins from the time. However, it is critical to assess whether there was an effective reoccupation of the previous sites, how it was processed and what processes may explain them. It is also important to understand if it was a temporary or partial reoccupation of the previous location or a long-term process prompting a new occupation likely to extend itself until the Middle Ages. In any case, what were the political, social or economical motivations behind the phenomenon? Could the instability lived in the late Empire prompted a partial reoccupation of these sites, in a time when Roman cities became fortified? In sum, the reoccupation process of the old Iron Age Hillforts is a complex process and entails varied reasons. This 135

Another phenomenon existing alongside the reorganization process of late first century BC and the first century AD is related with the emergence of low altitude Hillforts or ‘castros agrícolas’, as were named by Carlos Alberto Brochado de Almeida (2003). In this region, we have identified as potential ‘castros agrícolas’ the Hill-forts of Picarreira / Castro de Carapeços and Igreja-Matriz and Outeiro dos Picotos. If we admit, with all the reserves already pointed, the existence of Hill-forts with the features defined by Carlos Alberto Brochado de Almeida (2003: 173), we should only suggest that this valley Hill-forts may have witnessed chronologies surpassing the first century AD, adjusting to a different role of this sites, one that is related to the new order of factors that were clearly of Roman origin. The Outeiro do Picotos performed a Hill-fort catalogued as a ‘castro agrícola’ functionnig as a supporting infrastructure of the Roman road that run on the western part of the site and crossed the Cávado River going towards Barca do Lago. Its chronology is wide and includes micaceous ceramics and materials covering a Roman chronological spectrum up to the Low Empire (Almeida 2003b: 173). From the group of Hill-forts in this region is also important to referre a group of sites that Carlos Alberto

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Brochado de Almeida (2003: 395-406) included in a category denominated by the author as ‘late-Roman concentrated settlements’.

Under these circumstances, we were helped by the road network layout proposed by several authors, although none of the roads have been mapped on a suited scale.87

For this region, we considered as late-Roman settlements eight sites in the Barcelos municipality, Queijeiros, Abade de Neiva, Adro Velho, Barqueiros, Bouça dos Crastinhos, Chvão, Igreja, Cristelo, Agra de Além, Midões, Eira dos Mouros, Negreiros, Crasto, Ucha, Alvarão and Vila Cova, and also two sites with the same typology in Espondende: Cotovelos and Vila Chã.

The lack of written information together with constraints imposed by geomorphological features alongside with the high anthropization of the region require specific methodologies in prospection, historical data recollection, intense cartographic tasks and photo interpretation that in turn demand long and intensive work programs. In several cases the identification of the Roman road network needed to account medieval documentation in the assumption that the use of structural circulation routes lasted until the medieval times and may have, same cases, been kept in the periods to come (Almeida 1968; Almeida 1980).

The emergence of late-Roman concentrated settlements is a phenomenon we understand should be framed upon research focused on Late Antiquity studies, a period that is not under analysis in this work. Thus, the emergence of fortified habitats in a time before the medieval Villages is a phenomenon demanding a serious and specialized research that is not pertinent to analyse here.

2.2.2 Transport network and associated settlements

The lack of epigraphic information provided by Milestones, the inexistence of classic sources focused on the main itineraries and the fact that we are dealing with a profoundly anthropized landscape justify the difficulty to define precise traits of the road transport network for this region. The fact that there are no significant obstacles in terms of geography and main watercourses that would demand thorough construction works, such as pavements and bridges adds to the gap in information and to the lack of reference points that are key in assessing road trajectories. It is important to reinforce the idea that we are dealing with a secondary road network, with the strategic importance of its circulation trajectories not comparable to main itineraries. These roads were not included within the imperial program for the main ones and were not the object of constant investment in terms of pavements or bridges maintenance works. Furthermore, this network was not part of the cursus publicus supporting structure, and instead local communities maintained it. The regional road trajectories obeyed basic communication and transport imperatives and were constructed as an attempt to consolidate natural circulation courses of proto-historic origin. These natural courses crossed the region basic geographic and hydrographical structures with an E/W orientation and the valley transition circuits in an N/S orientation. The changes conducted on using old natural routes are the result from new needs emerging from the establishment of new commercial routes and new habitats of Roman foundation. The settlement network may have functioned as an activator of natural circulation corridors that naturally followed valley areas of lower altitude where a great number of Roman sites were established.

136

The stopping points supporting both land and maritime circulation may have existed in some cases due to the proximity of existing infrastructures located nearby roads or rivers, however they are difficult to identify in area where they are merely been proposed and not been confirmed. Trajectories mapped by Vasco Gil Mantas (1996) and Carlos Alberto Brochado de Almeida (2003) serving as the base to the maps we present still need further confirmation. However we have utilized secure elements to the scale in which we proposed to work on. In this group of itineraries is key in highlighting the importance of the Via XX itinerary or via per loca maritime as per the Antonine Itinerary. This road linked Bracara Augusta to the coastline in a circuit following the southern margin of the Cávado River up to Outeiro dos Picotos region, where it would go up north. Exiting Bracara Augusta on the city western side, the itinerary would follow the entitled Naia sidewalk up to Martim and from there to Areias de Vilar until Assento and Covo river. 88 This trajectory followed the southern margin of the Cávado until Agra da Vila reaching Outeiro dos Picotos where it would cross the Cávado until Barca do Lago (Morais 1998:19-20; 2005:65-68). As suggested, this river crossing may have functioned as freight transhipment site conducted in small boats from the harbour located between the coastline in southern Esposende and ‘Cavalos de Fão’. This was in fact the first crossing and transhipment site. However this does not exclude the possibility over the existence of a fluvial port closer to the Roman city (Morais 2005: 60-62).

87 When we refer a suited scale, we are referring works that enable cartography of, at least, 1:25000. 88 Although some outlines have already been proposed (Mantas 1996: 903-906), there is still the suggestion that the Via XX itinerary reached the Tamel river valley. However, it is important to acknowledge the group of potential outlines given to this trajectory, dispite all of them need more thorough researches. Nevertheless, the existence of a secondary circuit in the Tamel river valley is almost certain. This road would cross the Cávado between the Côvo river mouth and Areias de Vilae. We are, however, in agreement with Rui Morais, who defended that the trajectory of the via per loca maritime would reach the Outeiros dos Picotos / Barca do Lago region (Morais 1996: 19-20:113).

The Shaping of Mountain, Valley and Coastline Settlement

If an archaeological underwater intervention would be in Rui Morais’ opinion (2005:62) paramount in understanding the existence of an harbour between ‘Cavalos de Fão’ and the coastline, it would equally shed some light about the potential that an harbour may have had as an entry point to the Cávado River. Furthermore, this type of research would improve the overall knowledge about the shifts in configuration, both at the river mouth and in parts of its trajectory, between Roman times and our days. The existence of a road between the Outeiro dos Picotos area and the city would enable the transport of part of the merchandise coming from the Atlantic route by land, and made possible that on this part of the river ‘the transference of products to vehicles pulled by animals would be done through small ramps to shorten the distance of the load and unload of shipment and the board and land of people (…)’ (Morais 2005:106). The trajectory between Bracara Augusta and Outeiro de Picotos (ID 3060701) measured around 29 km, approximately 20 Roman miles (with an average per mile of 1.480 m). Therefore, stopping stations were mandatory and would fulfill the function of mutationes, mansiones or small size infrastructures. An estimate made for Via Nova over the existence of mutationes keeping distances of around 3 to 4 miles and of mansiones existing at every 8 to 10 miles (Lemos 2002) not always corresponded to the reality, and in this particular case we need to take other factors into consideration. First and foremost, there were no indicators that the cursus publicus distances applied to main itineraries serving the circulation of imperial servants and couriers between major cities could be applied to roads serving different functions. In the case of the trajectory under study, the existing mansiones and mutationes would mainly serve basic needs emerging from the transport and circulation of this part of Via XX itinerary. In terms of the calculation of distances needed for every stop, the orography established a trajectory that was not rugged and included watercourses of slow flows. These trends enabled a faster traffic when comparing to regions with different features. Thus, the establishment of infrastructures supporting circulation could have had longer distances between themselves, rather than the ones registered for the Via Nova region. The area crossed by this road, namely between Outeiro dos Picotos and Assento (Covo River) contrasted with the Roman remains found in Areias de Vilar region, where it is believed a fluvial port existed. Moreover, we can also observe the existence of occupational remains at regular distances varying between 4 to 5 miles between Outeiro dos Picotos /Agra da Vila (4 miles), Agra da Vila / Assento (5 miles), Assento / Martin (around 5 miles) and finally between Martin and Bracara Augusta (of around 5 miles)89.

89 It is important to highlight that these are approximate measures taken from 1:25000 cartography in a trajectory not subjected to field surveys nor to a detailed cartography. The presented measures do not

The Agra da Vila / Boavista (ID 3023401) site was located on a Cávado fluvial terrace and had altitudes ranging between 50 and 55 meters. Bricks, tegulae, dolia and Roman common ceramic fragments together with remains of old constructions have been uncovered across a wide area. It is also possible to admit the presence of Roman occupation eventually associated with ceramic kilns remains located at the Passadoiro site. These events are also suggested by the microtoponym ’Campo do Forno’ and by the existence of bricks (Almeida 1996). In Assento, Covo River (Stªa Eugénia) (ID 3026501) tegulae, some column shafts and four capitals were found. The site endured a long occupational period, witnessed by the edification of a pre-Romanesque church and a joining graveyard. Although the site has been identified as a potential Villa (Almeida 1996), we believe it is possible that due to its location and the found remains it might have been a mutatio or a mansio. The Martim church site (ID 3024701) was located on a flatten area on the southeast part of the parish church. In the current area of the church and Casa do Povo and in terrains facing east and southeast, remains of a building have been uncovered. This building seemed to have endured at least two occupational phases: one in Roman times and another one in the Suevi-Visigothic era. The building has been uncovered during one of the conducted excavations on site and its chronology was interpreted as being from the Middle Ages (Almeida 1972: 319-326). From Roman times should be the padded stones, a doorstep, fragments of opus signinum, part of a ionic capital, the base of a column and part of a podium anlongside some column shafts. Although the site has traditionally been identified as a Low Empire Villa and believed to have housed a temple, we also believe it is possible that these infrastructures belong in fact to a mansio due to its distance to Outeiro dos Picotos of around 15 miles. Its location on the road trajectory points towards it being an infrastructure supporting traffic. On the other hand, if the distance of 15 miles from Outeiro dos Picotos would justify a stop, the distance of 5 miles from the city would justify another stop from the traffic coming from Bracara Augusta. And in the eventuality of a port having existed in the Areias de Vilar area, this site would have functioned as a connection site, although this is only a possibility. The navigability of the Cávado River up to the Areias de Vilar area would effectively enable a fluvial circulation regardless of the difficulties of transhipment between a potential harbour area nearby the Cávado mouth and a port that should have been located closer to Bracara Augusta. This suggestion would enable the existence of a fluvial route ensuring the safer transport of many products supplying the Roman city (Morais 2005). The exact location of that hypothetical fluvial port is, at the moment, impossible to precise although it is likely have the rigor in calculation only done with a pedometer and were not subjected to detailed surveys, as it occurred with Via Nova or with prospections conducted on the Via XIX itinerary.

137

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

that it was located in the area near the Vilar de Frades monastery. 90 This suggestion also matches with the Afurada toponym referred by Jerónimo Contador de Argote (1732-34; 867 and 872) when mentioning the Cávado river navigation terminus. In the area nearby Vilar de Frades, the quantity of Roman remains present in Madalena (ID 3021103) is significant. This was a site where a necropolis might have existed. This structure has been referred by Albano Bellino (1090:25) for the Aveiras area (ID 3021102) (Aveleiras Farm or School Farm), who said it was located on a small hill facing the Cávado river. On site, construction ceramics and Roman pottery fragments were uncovered together with stones and traces of clay proceeding from floors and scattered around a wide area. Nearby S. Sebastião chapel (ID 3021104) and occupying a small hill, tegulae has been recolleted. Very likely related with the Aveleiras occupation is a voting altar found in S. João Baptista (ID 3021101) church and including a text that remains undecipherable. The origin site is also unknown. This is a granite altar, damaged on the upper and right sides and the inscription is of difficult to read: AILAECA / (.) IC(.)/DACOR (…)/DVREA (Almeida 1996; Martins 1990). A second important road trajectory within the region under study is represented by the north-south link mentioned in medieval documentation as Via Veteris crossing the Via XX itinerary in the Outeiros dos Picotos area (Almeida 1980). It probably corresponded to a ramification of the Via XVI itinerary on its western bound and would cross Cividade de Terroso following up to the mining area of Lagoa Negra, crossing the Cávado in Outeiro dos Picotos / Barca do Lago. From there it would run to the valley area nearby S. Lourenço Hill-fort and from there to the Neiva River. One possibility for the trajectory in this particular area entails the ramification of this itinerary following on the western bound of the S. Lourenço settlement and crossing Alto da Ponte, a site likely to have been a mutatio or mansio (Almeida 1996). In the Via Veteris trajectory, it is likely that Rates, a site located on the Terroso upstraem, may have functioned as a mutatio or mansio for this same circuit, although this assumption needs further research due to the alterations it endured in medieval times, namely in the associated supporting sites. The importance of the Atlantic route justified the existence of ports and harbours, both maritime and fluvial, and their connections with terrestrial trajectories, in order to enable the transport and distribution of products within the wide region. Although the study of ceramic materials

and glasses are concentrated in Bracara Augusta, the inventory, although incomplete of imported products such as amphorae, glass or sigillatas coming from settlements or Villas adds to the assumption of a perfect integration of the rural world with the commercial schemes connected with the Atlantic route and wider commercial circuits.

2.2.3 The exploitation of resources

2.2.3.1 Villas and associated units

The group of available data emerging from Villas enabled the consideration over the existence of 13 infrastructures referring to this typology of sites in this region. However, it is important to state that in the majority of cases we only have remains recollected from the surface. These remains can be categorized as Villa indicators as we are able to distinguish them from simple findings of tegulae and Roman common ceramic remains. Some sites from this group have been subjected to excavations and revealed rather simple structures easily associated with Farms or agricultural exploitation units of larger sizes. This would be possible if the region under study boosted other features or was located nearby an urban centre suggesting other type of correlations. Nevertheless, it is worth mentioning that some sites may effectively have corresponded to agricultural exploitation units of small sizes and and in other cases we securely identified them as villae. In the majority of cases, these establishments have been founded in Roman times. Conducted excavations however, enabled us to assess the premature emergence of agricultural exploitation units in the region. These events are well attested by the indigenous ceramics and Roman materials of High Empire chronology found in situ. This early phenomenon has also been documented both in valley and coastline areas of the northern and southern Cávado. Available data pointed towards the establishment of agricultural exploitation units, coeval of the foundation of Bracara Augusta and located in areas not directly linked with the city. Thus, the foundation and edification of the city coexisted with the reorganization and maintenance of some regional Hill-forts together with the emergence of new ways of organizing and exploiting the countryside, as was the case of Villas. We will briefly list the sites suggested as potencial Villas in the region.

90 Conducted excavation at the Vila de Frades Monastery uncovered

Alvito / Igreja, Barcelos (ID 3027401)

a voting altar presenting an inscription that despite being complex suggested the reuse of the epigraphic field in a second recording. The fact that no Roman remains were uncovered at this excavation, other than some tegulae, indicated that if the fluvial port did exist, it was not located on the monastery site nor within the vicinity. We thank Luís Fontes, the excavation supervisor, the provided information.

The site is located on a smooth hilly area nearby the parish church. On site, construction remains were found together with a column shaft and one 138

The Shaping of Mountain, Valley and Coastline Settlement

capital and Roman construction and domestic ceramics.

Paço / Vila Cova, Barcelos (ID 3028602) This Villa is located on a flattened hill on a site known as Lugar do Paço, in Outeiro. Conducted excavations identified a first occupational phase between the late first century BC and the first decades of the first century AD. After this first phase the site witnessed the construction of a first set of structures that in Carlos Alberto Brochado de Almeida’s opinion corresponded to a Roman Villa dating back from the first century / beginning of the second century AD. Throughout the High Empire the structures endured regeneration works, although the most significant one was only conducted on the fourth century, at a time when the bath infrastructure was built. Between the fourth century and the beginning of the fifth, a few remodelling works to the walls took place. The referred author admitted an occupation in the SueviVisigothic period, although it is difficult to understand the evolution of this infrastructure as ‘the found walls are scarce and they do not contributed towards the better comprehension of the site as they are located at the same level as the foundations’ (Almeida 1996, III:187). The uncovered ceramic remains revealed a predominance of Roman pottery including common pieces of ‘bracarense’, thin grey pottery and Hispanic sigillatas. Other than these, one can document Iron Age ceramics, late Roman and Suevi and Visigothic remains. This group of materials added to the belief regarding the existence of a Villa, a habitat that likely had an occupation period starting in the first century BC and the beginning of the first century AD. It endured gradual regeneration periods and was kept occupied at least, until the Suevi-Visigothic times. Linked to the Villa was a necropolis from where two rectangular graves are known made of tegulae that appeared on the southern part of the current house, at the ‘Lugar do Paço’.

Assento / Várzea, Barcelos (ID 3028301) The site is located on a lower terrain of the Covo River, particularly on an area well irrigated and with excellent agricultural lands. From the site are one hypocaust pila, two conduits and a column base other than several items classified as preRomanesque or Romanesque that were part of the primitive church. The presence of Roman elements caters for the hypothesis about the existence of a Farm or a Villa likely behind the later establishment of a Christian temple. Furthermore, a Benedictine monastery is believed to have existed on site.

139

Paço / Monte de Fralães, Barcelos (ID 3025102) The site is located between the parish church and Monte da Saia settlement, where tegulae, imbrices, Roman common ceramics and a wall have been uncovered. The potential relationship of these findings with an inscriptions found at Fralães Farm (ID 3025101) mentioning a soldier (AUR)ELIO PATRI/C(I)(i) . F(ilius) . MIL(I) (…) allowed the assumption over the existence of a Villa on site.

Assento, Barcelos (ID 3026501) The site is located on a well-irrigated area nearby fertile lands. On site, column shafts and four capitals were found alongside Roman common pottery. Moreover, it is also believed that a Roman establishment existed before the Christian temple.

Castelo / Abade de Neiva, Barcelos (ID 3020104) The site is located on a hillside terrain with rich soils that are currently occupied by the ‘’Quinta do Castelo’’. On site, a few granite column shafts were found together with an abundant presence of tegulae scattered throughout the joining property, which seemed to indicate the presence of an agricultural exploitation unit in Roman times.

Assento / Quintães, Barcelos (ID 3026202) The site is located on a small hill facing a wide valley of fertile and well irrigated lands. Abundant tegulae fragments and common Roman pottery are visible across the terrains of the Quinta do Assento lands. We also highlight the finding of a capital.

Vila Mendo, Póvoa de Varzim / Esponsende (ID 3060201) The site is located on the boundary of the Póvoa de Varzim and Esponsende municipalities and encompasses a building excavated by José Manuel Flores Gomes (Gomes 1996) and a series of remains listed by Carlos Alberto Brochado de Almeida that might correspond to the same infrastructure. The excavated building has a rectangular plan and walls reaching two meters high. Taking the found remains into consideration, we believe it corresponded to a fishing industry supporting structure with indicators of metallurgical activities. Thus, the housing area was likely located further north, very likely on an area where remains have also been uncovered and identified by Carlos Alberto Brochado de Almeida (1987) as being Villa Menendi. These findings were located on the southwest part of the current Apúlia Village and included the finding of schist walls, granite column shafts, manual grading wheels, loom weights,

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

tegulae, imbrices and other ceramics in 1950. The referred author also observed on site Roman pottery, construction material and indigenous ceramic remains that led to the assumption of a chronology for this site that goes back to the first century BC (Almeida 1987).

lower terrains; and a second area of smaller sizes on the other side of the old road on a location known as Trelopaço. The found remains inside the current Farm may have corresponded to a Villa and the Trelopaço site to a necropolis or to constructions dependent of the Villa (Almeida 1987: 98).

Marinhas, Esposende (ID 3061201)

Alto de Martim Vaz, Póvoa de Varzim (ID 13131001)

This site is located near the Lacticínios das Marinhas Factory in the southern part of the Peralta watercourse and the national route linking Porto to Viana do Castelo. One may se tegulae in the terrains nearby the itinerary, although it is more common to find these items nearby the southern part of the factory itself. We highlight the finding of a manual grading wheel. In Carlos Alberto Brochado de Almeida’s opinion, the dispersion of the findings may indicate the presence of a Villa ‘that might have had as natural boundaries, other than the sea, the Peralta watercourse on the northwest side and a secondary Roman trajectory coming from Barca do Lago on the east part, running on the foothills of the S. Lourenço area’.

Linhariça, Palmeira de Faro, Esposende (ID 3061301) In 1946 M. Boaventura reported the finding of tegulae and ceramics in the Linhariça area, a site corresponding to a small valley with lands with agricultural potential. A few meters ahead, on an area known as Chaves and Sovalo, underground walls were uncovered together with common ceramic artefacts and two fireplaces. Nearby the site, on the foothills of the Santa Eufémia mountain range, manual grading wheels and two capitals were uncovered. The dispersion and nature of the remains led the author Carlos Alberto Brochado de Almeida to come up with the hypothesis about this site having been a Roman Villa or, at least, a rich Farm.

Quinta de Belinho, Esposende (ID 3060101) The remains considered by Carlos Alberto Brochado de Almeida as ‘one of the most significant Roman occupation of the Esposende municipality’ were found inside a wide property known as Casa de Belinho, located on the foothills of the Cividade de Belinho Hill-fort, an Iron Age period site. This property is bordered by an old real road that in the author’s opinion corresponded the property boundary matching the trajectory of an ancient Roman road linking Porto to Caminha and crossing the Cávado River at Barca do Lago. The Roman remains mainly referred to construction and common ceramics that were distributed across two areas: one with over two hectares and going from the house up to a slope giving place to two further 140

In several parts of the urban area of Póvoa de Varzim, however with particular emphasis towards the Alto de Martim Vaz area, Roman ruins have been uncovered. They corresponded to a Villa with a plan that cannot be fully recoverd. On site, amphorae, sigillata and ‘bracarense’ pottery fragments have been recollected. In Carlos Alberto Brochado de Almeida’s opinion, the Villa had two occupational phases, one of them in the late first century AD and the other one a bit later. Between Banhos Street and the coastline, two tanks were found nerby remains of Roman ‘good’ construction, (Almeida 1972: 32) believed to function as a fish production unit.

Similarly to Bracara Augusta and tp some Villas around the city, indigenous communities occupied these new rural sites as the remains found in situ have attested. However, for this exact region, the infrastructures do not have a direct relationship with the area around the city or with any other urban site nearby. The Romanization of the rural areas endured diversified processes that do not necessarily implied a direct relationship between the urban centres and their ‘extension’ units existing in the countryside. On the other hand, if Villas have performed a structuring role in the space appropriation processes, mainly in areas that we may call Romanised, it is still difficult to attribute them a supremacy role as rural world Romanising structures, even though such statement doesn’t entail a subvalorisation of such entities within the construction process of a new rural landscape. Being new structures of exploiting agricultural resources, Villas had a key role as core centres of fundi, regardless of their wider or smaller sizes, and they started to dominate the new ways to exploit agricultural landscapes. Villas had a key role in the land clearance processes, made with new techniques and new or improved Farming equipment and also in the new Roman economy framework. These circumstances have been critical in the new ways of understanding and exploiting the territory. Adding to this, Villa had an important role in the private appropriation of lands and their consequent delimitation. It is still difficult to understand the ways this process occurred both on a political and practical level. However, regardless of their wider or smaller evolution and of the diverse processes occurring from region to region, land privatization became the technical and juridical mean behind taxation, which represented a real ‘revolution’

The Shaping of Mountain, Valley and Coastline Settlement

within rural space reorganization and in which the Villas had a supermacy role. We do not have, for the region under study as for the Entre-Douro-e-Minho territory, any indication that enables us to calculate the fundi of these Villas, and we are certain that any calculation for whatever area entailed basic constraints. In fact, without written documentation or clear boundary indicators, the calculation of the fundi in this region or in any other is a difficult task to estimate, mainly in terms of assessing the surface measurements as it was hardly continuous, similarly to what happened in the periods to come. The documentation since the Middle Ages up to our days, together with the observation of rural contemporary properties shed some light over the discontinuous nature of the soil use throughout time. Thus, any attempt to calculate the fundi in Roman times should take this discontinuity factor into consideration and attempt to define the joining properties boundaries in relation with the central site. In the coastline, several of these Villas may have their economy entirely based upon the exploitation of salt and fish resources, although it is difficult to estimate the economic reach of these productions. The study of Villas and associated structures enabled a basic understanding about the rural landscape patterns. However, other than the Villas, there certainly were other establishments associated with the rural world that can be guessed from the wide quantity of occupational remains scattered across the region. They may include Farms, agricultural or crafts agglomerates of unknown size and typologies that shaped the Romanization of a particular region, although their features are difficult to define. In light of what has been found thus far, it is impossible to state that Villas held the role of basic entity within the Romanization process of this particular territory and whether they assumed the leading role in the history of the rural world of this particular region. We can only contrast the number of potential Villas with the intense spread of remains that may have corresponded to smaller size exploitations units with the same agricultural or industrial character.

2.2.3.2 Small and medium size farms and artisanl units

The difference in shapes and sizes between Villas and agricultural exploitation units of smaller size still lacks further research studies and needs the application of suited methodologies in their overall analysis. Studies conducted by B. Douteyssier and F. Trément (2006-07) for the Aquitania region represented a significant advance in the identification of small and wide Villas. However, we were not able to conduct similar studies in the region under analysis. The problem was enhanced when we attempted to distinguish categories related with small rural properties, as we did not have enough data. Furthermore, the difficulty 141

related with establishing a relationship between the materials emerging in field walking surveys and the underground structures is enlarged with the lack of excavations. In the majority of cases, it was impossible to develop analysis that dealt with remains dispersion areas. Within the region, we listed 44 sites that may have corresponded to agricultural or industrial exploitation units and named them as Farms. A smaller number of cases corresponded to sites where only tegulae was found on the surface. The location of sites in hillsides, nearby agricultural soils enabled the assumption that we were, in fact dealing with Farms. The majority of sites also displayed other types of materials, namely Roman common ceramics, dolia fragments, manual grading wheels and construction stone. The early installation of some of these sites corroborated by the presence of micaceous ceramic remains of indigenous production and associated Roman materials, as was the case of Casa do Adro, in Barcelos (ID 3023001), Castro do Monte, in Barcelos (ID 3021701) where a great quantity of common and construction Roman ceramics alongside indigenous ceramics and remains of a wall have been uncovered and Outeiro site, in Barcelos (ID 302270) with tegulae remains associated with construction stone, Roman common and indigenous pottery. The description of the findings makes it difficult to list different categories for these type of sites. However, it is possible to distinguish different locations amongst them, which shed some light on the type of activities carried out on each site. Thus, alongside sites primarily focused on agricultural exploitation, there were others focused on functions linked with the production of ceramics, as it was the case of the Igreja site, in Barcelos (ID 3027701). There are indirect indicators of rural habitats perceived by the presence of necropolis, as it happened with Caride, Barcelos (ID 3021802) or Covelos, Esponsende (ID 3061502). Other indicators of potential sites require future conducted field work to enable safer classifications, as it is the example of Quinta do Lourido in Barcelos (ID 3025201), a site located on a hillside with good soils and where tegulae, imbrices, construction stone, mortar and ‘bracarense’ pottery were found . The quantity of remains is significant enough to perceive a valley landscape with an intense occupation of small exploitation units. From within the overwhelming quantity of sites where only tegulae was uncovered, some may actually have corresponded to small Farms, although we are not fully aware about their exact features. Thus, as important as the identification and study of Villas it is paramount to improve suited criteria for the interpretation of remains uncovered at sites that may correspond to small size agricultural unit, to conduct excavations that may improve the knowledge over the structures and to study of common ceramics that are mainly present in this type of sites, so important as the imported pottery to our understanding and interpretation of the Romanized rural world.

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

2.3 Limits and scales

Martinho and to its elevation to parish headquarters in the eleventh century (Almeida 1996).

2.3.1 Diffusion areas and influencing boundaries

The understanding about the existence of different distribution areas on the sites under study led us to consider developing an analysis focus on the relationship between sites, their limits and their connection to basic entities structuring the landscape. This was the case with the hydrographical network, the coastline and the geography shaping the transport network itself. On a first level, the observed distribution diversity may be partially related with the role Hill-forts detained as diffusion centres, responsible for the first attempt to fixate communities in valley areas between the end of the first century BC and the beginning of the following century. These events resulted in the dispersion of Farms, Villas and other valley habitats occupying ager areas that were previously located and integrated within exploitation lands surrounding the settlements. We will outline some of those relationships, starting with a point reference to the north of the river Neiva, more precisely where the Carmona (ID 3021202) and the Srª da Aparecida (ID 3021201) Hill-forts were located. This reference point was located in a crossroad of natural circulation route running E/W and N/S and linking the Cávado and the Lima basins through the Tamel river valley. The Roman road crossing this axis was the crystallization of a natural route that penetrated this area. Moreover, this trajectory was also the end result of the location of the existing habitats, the orientation of the hydrographical network and the excellent potential of the agricultural soils. All these may help explain the intense Roman occupation of this region, with a particular emphasis towards the areas between the Cávado and the Neiva rivers. Both Quinta das Giestas (ID 3021203) and Laje (ID 3021204) sites, alongside one Farm and one grave entirely made of tegulae and imbrices corresponded to an indigenous dispersion area linked with the Srª da Aparecida Hill-fort (Almeida 1996:52). Communities likely coming from the S. Simão (ID 3022401) Hill-fort located on the southeast side fixated themselves in sites of Casa do Souto (ID 3022403) and Eirado das Quintas (ID 3022402), both located in the region apparently crossed by a secondary Roman itinerary. These sites typology is still unclear, however the reference of mining activity remains (Maciel 2003: 138-140) and their location nearby an itinerary is relevant for any future work to develop in the region. Between S. Simão and the Monte Lousado Hill-fort, an inscription dedicated to Jupiter was uncovered in Mondim (ID 3025601). Nearby, we highlight the finding of tegulae in Ramil (ID 3025602). Mondim corresponded to an important Roman site that registered a continuous occupation leading to the construction of a temple dedicated to S. 142

The valley area is dominated by the Monte do Castro / Picarreira sites and the eastern part include an area of reliefs where the watchtower of S. Lourenço (ID 3020702) is located. This is an area where a series of Farms and other exploitation units of unknown typology attest an intense Roman occupation. The Minhotas (ID 3021803), Casalmelhe (ID 3021702), Crasto do Monte (ID 3021701) and Igreja (ID 3022601) sites amongst others also verify the intense occupation of this valley area. The dispersion of the communities previously settled in Hill-forts, the fact that this area was crossed by a Roman road and the regional hydrographical network of transversal orientation in relation with the main rivers may have played a critical influence in disseminating the occupation of this territory. We believe that it is possible for the valley region located nearby Citânia de Roriz Hill-fort that other than the indigenous thermal building entitled Stª Maria dos Galegos, the habitats of Paço / Oliveira (ID 3025403) and Igreja (D 3025402) in the northern part corresponded to a population dispersion conducted from this particular Hill-fort. We also admit that some sites near the northern side of the Cávado River may have housed communities coming from that same Hill-fort. Similar events are also believed to have occurred in Aldeia / Eidos (ID 3021001) or Assento / Manhete (ID 3024501). The communities leaving Monte do Facho may have been relocated in several other areas, namely towards the Castelo (ID 3020104) Villa in the Northwest or towards sites as Amorim (ID 3020101), Paço Velho (ID 3027303) and Vila Frescainha (ID 3027301). From S. Mamede may have come the first communities responsible for the valley land clearing at the beginning of the first century. It is admitted that this communities settled at the Paço Villa, in Vila Cova (ID 3028602), as suggested by the early site foundation revealed by the excavation works, with similar circumstances likely to have occurred with communities established at Mereces (ID 3028605). It is believed that the S. Mamede Hill-fort may have functioned as a diffusion centre for communities coming from several valley sites. In the southern part of this area there were remains from a Farm in Outeiro (ID 3022703), a site where tegulae, stones and Roman common pottery have been uncovered. The artefacts have also been associated with indigenous ceramic fragments indicating an early occupation of the Paço Villa, in Via Cova. Nearby the Outeiro site, we have verified the presence of tegulae, particularly in Carvalhal (ID 3022702) and at the Paço site (ID 3028901). Both sites are believed to have been Farms. Similar events are believed to have occurred in Cividade de Belinho (ID 3060102), a site not registering any signs of a Roman occupation. Quinta de Belinho (ID 3060101) and Agra do Relógio (ID 3060104) may be related with

The Shaping of Mountain, Valley and Coastline Settlement

the land clearing the ager previously integrated in the Hill-fort exploitation area. It is admissible that this site or Castro do Monte housed communities that lived in habitats located around the nearby area.

site (ID 3028605), where fragments of tegulae and Roman common pottery was found, it may correspond with an agricultural infrastructure associated with the Paço Villa.

On a first level, the Roman settlement located in the southern Cávado may be explained by diffusion phenomena emerging from the regional Hill-forts, namely Outeiro do Crasto (ID 3020501), Monte da Saia (ID 3022201) and Cividade de Terroso (ID 13131201).

The presence of a significant number of ceramic kilns, in particular on the northern margin of the Cávado River revealed us that they were likely associated with Villas that other then being focused on agricultural exploits may also have included structures destined to produce ceramic. This could be the case of the Castelo villa (ID 3020104) that would associate agricultural exploitation with pottery production activities located at the Quinta do Passal (ID 3028501).

Roman remains scattered around Monte da Saia, particularly on the east part, help explain the diffusion phenomenon of this particular Hill-fort. In terms of the occupation progressing towards the southern Cávado region, we can state that this diffusion should take into consideration the Via XX itinerary, as well as other roads, namely the ones crossing the Covo River and located on the eastern part of Monte da Saia. It is possible that the communities caming from Outeiro do Castro Hill-fort (ID 3020501) settled some Farms, as the cases of Quinta de Lourido (ID 3025201) and Paço (ID 3020502) in the south part, or Assento (ID 3020301) and Varzea (ID 3028301) in the western side, and even Casa do Adro (ID 3023001) and Vilarinho (ID 3023002) on the northern region. In the case of Cividade de Terroso, the settlement density around the Hill-fort may be explained by the occupation of Alto da Vinha / Beiriz de Baixo (ID 13130601) (Gomes 1996), where either a significant part of the population occupied the southern platform of the hill near the valley lands or the proximity to the sea, either because the proximity to the sea prompted Terroso inhabitants and/or Monte de S. Felix communities to fixate in Martim Vaz (ID 1313001) or in Vila Mendo (ID 3060201). This first diffusion phase of communities coming from Hill-forts is backed up by data emerging from excavations alongside found remains that go back to the first century AD. The settlement dispersion area seemed to open up the boundaries towards valley areas and included habitats corresponding to small agricultural exploitation units that in some cases may have evolved to a Villa. Subsequently, some establishments of the Villa type may have been associated with smaller units, be they kilns, fishing productions or even small Farms. Thus, the analysis of the relationships between some Villas and associated structures may enable a calculation regarding the influential or attraction areas exerted by Villas with the aim to analysing settlement areas and the establishing of property boundaries, without crossing this analysis with fundi estimation. One of the examples in which these associations seem operative corresponds to the Paço / Vila Cova Villa case. Two graves made of tegulae were uncovered on the southern part of the current house and they are believed to have been part of an associated necropolis. This helps the estimate over the property limits in the southern part. With regards to the northwest region, mainly the Mereces 143

A similar event might have occurred in Paço Velho (ID 3027303), a site that has been linked with ceramic production at the Telheiras centre (ID 3027302). Great quantities of tegulae some of them with soot particles have been recollected. However, this particular case presented us with some challenges. The archaeological remains coming from Paço Velho suggested a small to medium size agricultural exploitation unit, while the production area in Telheiras suggested a wide site. This has led towards the assumption over a potential crafts area surpassing the domestic scope. We may, in this case, admit the existence of an artisanal exploitation located in Telheiras, one that had no relationship with the Paço Velho site. Similarly, we may consider the possibility that Paço Velho and the Vila Frescaínha necropolis (ID 3027301) integrated an artisanal infrastructure with the core business being precisely in Telheiras. In the coastline areas, the salt and fish production may have functioned as an attraction factor to the installation of Villas that had in their activities a strong income source. In these cases, the dispersion of communities from the Terroso and S. Felix Hill-forts likely led to the installation of a network of villas and focused on activities linked with the exploitation of resources present in the sea. This could have been the case of Alto de Martim Vaz (ID 13131001), Vila Mendo (ID 13130701) or Caxinas (ID 13160103), sites that are currently located in silted regions, although they have in Roman times been located relatively close to the sea. These relationships have as diffusion entities the Hillforts, the potential dispersion of its inhabitants towards valley areas nearby and the proximity networks starting from some Villas. The potential limits of these relationships may be empirically established by the assessment of the potential reach of these relationships through the use of the ‘closest neighbour’ patterns. Identical rehearsals may evidently be conducted in relation to land and watercourses networks, for example in relation with the sea or even with areas presenting significant geographic reliefs. Rivers and other watercourses are key entities in attracting communities, due to their hydric potential and to the existence of animal resources that have always been of

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

paramount importance in developing economies based upon varied resources. The presence of the Cávado River is of extreme importance in the understanding of the rural occupation process of the nearby regions, both in its northern and southern regions. Alongside with its course that included boarding and crossing functions, the fishing resources and the presence of water needed in agriculture and also artisanal activities may help explain the presence of Villas, Farms and kilns on its northern margin. Both the Côvo and Tamel rivers may explain the intense habitat network located in the valley areas and revealing an orientation that is transversal to the Cávado River. Thus, the Cávado, Ave, Côvo and Tamel rivers together with some other watercourses enabled us to understand the settlement patterns that would not have existed with the simple presence of a surrounding Hill-fort. Thus, we may state that the S. Felix Hill-fort in Laundos and located in an area of unfertile lands may have witnessed its communities to be relocated in areas outside of that which would be its natural area of influence or dissemination. Similarly to the fertility level of the soils and the presence of watercourses, the forest areas provided inhabitants with important resources that played a key role over the choice of sites to establish agricultural or other type of exploitation units. Alongside basic pastoral and energy resources such as wood, these regions provided hunting opportunities and resources recollection activities, thus positively impacting upon the economy of communities located in slopes or valley areas. Thus, alongside the diffusion role of many Hill-forts and the presence of fertile lands and water, the existence of significant reliefs, with diverse resources in areas away from the coastline may explain the intensity of settlement observed. In sum, the observation of the cartography established the presence of two clearly distinct regions located within the studied subunit: the valley region with intense settlement and the coastline area where the settlements were rather scarce and disperse. A factor to take into consideration related with the rarefaction and dispersion of sites nearby the coastline may be linked with changes occurred within the coastline or with the presence of lagoon areas and unfertile soils in the region adjacent to the Cávado River, particularly in its connection with the sea. Moreover, this may also be explained by the silting processes that took place in the southern region of the Apulia area, a region that encompasses the current coastline area of the Póvoa do Varzim municipality. In the southern Cávado River the settlement include Hillforts located on the small hills existing in the region, from which Cividade de Terroso is a reference. This is an important site presenting an occupation up to the end of the first century AD that alongside with exploiting other resources was also a metallurgical centre (Gomes 1996). The first Villas at the Alto de Martim Vaz and Vila Mendo may have come from one of those Hill-forts. In both 144

cases, we are dealing with High Empire Villas that were focused on activities related with the salting and fishing industries. We admit that the silting that occurred in this region and the intense current urbanization processes prevented the finding of other identical structures. Furthermore, it is also possible that maritime resources were not significant enough to justify a dense occupational spread in the coastline area. In northern Cávado, the proto-historic and Roman settlement patterns kept an even wider distance in relation to the coastline. Hill-forts may explain the communities’ dispersion throughout the valleys nearby, however the fact that a Roman road existing in the surroundings should also be taken into consideration. Other settlements were located around this road circuit, as it was the case of Marinhas and Quinta de Belinho Villas, both infrastructures lacking the exploitation of sea resources as the reason for their establishment.

2.3.2 The Romanization of the valley landscape and coastline façade

The early influence and contact of the region under study with the Roman world, circumstances that started with D. Iunius Brutus expedition, performed a key factor in the interpretation of its settlement patterns, as it led to significant changes occurring within the Hill-forts and turning sites into changing motors up to mid first century BC. The relationships established with already pacified territories located at the southern peninsula surely anticipated the integration of this land within the Roman world. Thus, it is admissible an early foundation of rural centres in valley regions where previous Hill-forts existed. Moreover, the dispersion of communities was not conducted and based upon a single orientation or model. Instead it prompted a gradual establishment of rural exploitation units by indigenous communities in the late first century BC. These events were crucial in the overall perception of the Romanization process of the rural areas ans this was a process organized around the development of Villas and secondary aedificiae. The gradual contacts between this region and the wider Romanised world were intensified throughout the first century BC and helped develop new ways of understanding the rural space. The definite pacification of the Northwest, the foundation of Bracara Augusta, the opening of a road network and the establishment of commercial, maritime, river and terrestrial routes performed the necessary factors needed to the installation of new ways to exploit and organize the rural world. New habitats were created in order to reorganize the space and develop a rural landscape around private and delimited areas respecting to a stratified society model that was similar to the Roman one. On the other hand, primary entities such as geography and hydrography are elements of an open landscape served by

The Shaping of Mountain, Valley and Coastline Settlement

multiple resources. Geographical and hydrographical structural orientations help organizing and crystalizing natural routes in Roman times. In the area under study, the transport network respected the geographical setting and followed a N/S and E/W orientation, adapting to the basic geomorphological fractures and the main and secondary hydrographical networks. It is around these natural trajectories that were also the valleys irrigation axes, where the habitats of Roman typology occupation were established explained by land clearing of new exploitation areas and by the road network that simultaneously helped them to growth. The density of Villas, Farms or intermediate units included a wide group of resources present in the region and performed the main objectives behind their establishment. In several areas, the agricultural exploitation and the energetic resources prompted a diverse economy, where the exploitation of the ager may have been completed with pastoral activities and exploitation of mountainous energetic resources. To the group of resources linked with land and forest exploits alongside the ones present at the sea and rivers, should be added the ones behind mining and metallurgical activities. Activities linked with mining processes already existed in the Iron Age Hill-forts and in habitats located in areas nearby the Neiva River, where mining exploitation remains are still visible today. Mining richness, namely gold has been referenced to have existed in the region between Lagoa Negra and Cividade de Terroso, both in the Iron Age and Roman times. We are not aware about the size and economic relevance of the gold exploitation activities conducted in Lagoa Negra or even to identify any site linked with this particular activity. The metallurgical activity present in Terroso has been well documented. However, it is still not possible to be fully aware whether the mining richness of this region was also explored in the new emerging sites, namely the case of Alto da Vinha and in the Beiriz area. In the areas near northern Cávado and in the Alvarães region, clays and kaulins performed a key raw material resource and some of the identified sites may have corresponded to extraction units or ceramic production centres that functioned as domestic production sites or handcrafts centres focused on trading activities. In the coastline region, the distances between habitats may be explained by a great quantity of key resources coming from the sea, however the soils were less favourable to the agricultural practice. Thus, the salt and fishing industries justified the existence of some Villas, however it is almost certain that they did not attracted or were profitable enough to favour a dense occupation of the coastline. Nevertheless, it is important to keep a perspective regarding the interpretation of available data and conduct a thorough study of the coastline occupation, which is mandatory of a rigorous definition of the overall coastline settlement existing in Roman times. Moreover, the study should be accompanied by a rigorous 145

knowledge over the sizes of these exploitation units and their production potential. In southern Cávado, these types of sites located near the sea, they seemed isolated and not connected with other associated sites. In northern Cávado, the Hill-forts sites were located in higher altitude lands and away from the coastline. They were clearly associated with Roman remains that, as previously stated, suggested a displacement of indigenous communities towards the valley area, where a new human and economical landscape was structured articulated with available resources exploitation, alongside the very own needs of secondary road itineraries crossing the region. One of the challenges regarding the Roman occupational model of the valley ad coastline areas under analysis has been linked with the apparent lack of habitats of the vici type the region. This absence seemed surprising in terms of the overall territorial organization logic in Roman times. First and foremost, we are dealing with a wide area housing rather disperse settlement types that would have necessity for sites functioning as redistributors centres of imported products coming from the coastline or of products manufactured in the region, several of which would have been destined to Bracara Augusta. On the other hand, given the referred features, it doesn’t seem logic for a region of this size to be fully and directly dependent of a city, at least, in terms of certain services. We are certain that the area in question still lacks further field works that could be more detailed in certain areas, for example in the study of a wide concentration of ceramic materials or other types of artefacts that until today have been difficult to interpret. We understand that some sites should be studied in detail as they may correspond to sites of the vici type. Some of these sites are: Senhora de Águas Santas, in Santa Eulália de Rio Covo (ID 3026601), a site traditionally interpreted as a ‘Roman thermal station’ occupied in the Middle Ages according to Carlos Alberto Ferreira de Almeida (1972:126) and later classified as a Villa by Carlos Alberto Brochado de Almeida (1996); Agra da Vila (ID 3023401) classified as an agricultural exploitation unit (Almeida 1996), however interpreted by us as a potential road station; Alto da Vinha / Beiriz (ID 1310601) due to the fact that on site a Jupiter voting inscription has been uncovered alongside a potential statue pedestal (Alarcão 1988: 1/285; le Roux and Tranoy 1973:213-214). The Senhora de Águas Santas site may have corresponded to a thermal vicus and has a privilege location favouring the installation of an open agglomerate, due to its altitude and due to the fact that it is located nearby a potential Roman itinerary referenced in the Middle Ages as carraria antiqua and believed to have connected the Via XVI to the Via XX itineraries. It is believed that it also crossed the Côvo River, which was a densely populated area in Roman times (Almeida 1996). Similarly to what took place in other sites interpreted as potential thermal vici for the Entre-Douro-e-Minho region,

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

namely Caldas de Vizela, Caldelas or Caldas das Taipas, all these sites included road supporting functions and acquired, as a result, a multifunctional character. Moreover, the Senhora de Águas site in particular, may have originally had a thermal origin due to the quality of its waters. It is likely that it gradually may have been transformed into a wider settlement encompassing several other functions. On the other hand, Agra da Vila located on the Via XX itinerary may have had functions supporting circulation alongside other of economic character related with a potential market function and as an artisanal centre, as kiln remains were found nearby (Almeida 1996). The Alto da Vinha site, in Beiriz is believed to have been a potential a vicus. This assumption is based upon the type of epigraphic remains found on site and attributed to wider privileged, religious and economic functions namely linked with the distribution of salt or fish due to its location near the coastline. We highlight the fact that the region in question had a tradition of gold production since the Iron Age period, and metallurgical activities that have been attested by the findings in the Terroso and Monte de S. Felix de Laundos settlements. Under these circumstances, a potential vicus in the region could also encompass a metallurgical centre. Other potential vicus located in the region could have been a site previously referred as a fluvial port in the Cávado river, namely in the region of Areias de Vilar. The basic needs of a port demanded the existence of a populated settlement of some relevance housing several types of other activities. Although it is not possible to locate the port infrastructure, a voting altar has been found at the Vilar de Frades monastery. Another altar has been recollected from the S. João Baptista church, being of unknown origin and also of difficult reading (Santos et alii 1983:195). The northern Cávado region should be further studied and a potential reassessment should be conducted regarding the functionality of the three sites. One of them is located around S. Martinho de Alvito (ID 3027101), where an inscription not yet deciphered was uncovered together with a column shaft. There are numerous references to dispersion areas of ceramic materials however they do not provide a suited functional characterization of their origin sites. The concentration of findings on site together with their location nearby secondary road suggested the existence of sites prolonging themselves towards the north the carraria antiqua. These are indicators suggesting the potential presence of a site of the vicus type. The other site that should equally be studied in detail, although located in a region presenting lesser remains, is located around Mondim (ID 3025601), a site where a voting altar dedicated to Jupiter was found and where on the eleventh century a temple to S. Martinho de Dume was implanted. This temple is behind the origin of the parish. Moreover, the site is located nearby a valley of the Neiva river course. 146

Finally, the third area to investigate is located nearby the Neiva River mouth around the Alto da Ponte (ID 3060107), a site where Carlos Alberto Brochado de Almeida (1996) located the Neiva river crossing. In this region there are several construction and Roman common ceramics dispersion areas. The relevance of this site is related with its river crossing support and its association with the material dispersion of remains found in the surrounding area. These events may justify the existence of an important site.

Conclusion As stated in the introduction of the present study, our main aim has been centred on the analysis of the Portuguese Roman settlement patterns on the current EntreDouro-e-Minho region, the former western area of the ‘conventus bracarensis’. This performs a region that, from start, offered privileged conditions for the assessment of the cultural changing processes emerging from the integration of indigenous communities within the Roman world. The choice over the study area was dictated by the wide quantity of available data coming from research projects conducted throughout the last three decades and prompting a diachronic reading of settlement strategies of rural world reorganization and of a reassessment study of the Romanization concept when applied to a territory considered peripheral and hardly integrated within the existing Roman Empire. The wide variety of data emerging from the study of the pre-Roman settlement framework and the identification of numerous sites of Roman origin have challenged us to conduct a rehearsal in interpreting the evolution of this territory. The majority of the archaeological sites, which compose our work sources has been found in field walking surveys and they are enlisted in the inventories and site catalogues with a regional character . The difficulties in attaining our aim have been substantial and are inherent to the successive chapters included in this work. These challenges are related with the precarious nature of data, be it for its imprecision or related with the impossibility to precise chronologies, functionalities or even site features. If it is true that such limitations have led us to devalue several existing references and in other cases even to doubt the validity of our analysis exercise, we believed however to risk lead it to the end. As we gathered information, and despite of its truncated and fragmented character, we became aware about its enormous potential in generating new work hypothesis related with the research of the Romanization process in this circumscribed area of the bracarensis conventus and of the overall peninsular northwest region. In the provincial organization framework we attempted to approach the ways the territory was constructed, while seeking to highlight the particularities shaped by the geographical features and adapted by the new space organizational strategies. Thus, we tried to analyse the landscape evolutionary process of a concrete region, however wide enough to assess and acknowledge the potential models of space (re)organization. On chapter 2 we framed our research challenges within the wide peninsular northwest territory hoping to understand the diversity of the pre-Roman world, the conquest and domain strategies imposed on the existing communities by Rome and the political and administrative mechanisms ensuring the territorial control and their regional integration within the overall imperial management. 147

The difficulties felt throughout the development of this study have been greater in the approach of the EntreDouro-e-Minho region conducted on chapter 3, where the precarious nature of data impacted upon the validity of our suggestive interpretation hypothesis as oppose to what happened when analysing more restricted regions on chapters 4 and 5. Nevertheless, we deemed crucial to value a group of available data for the Entre-Douro-e-Minho region in order to, with some degree of safety, change our scale of analysis and be able to value in detail chosen territorial subunits that may have conditioned and acted as regional changing vectors. Having the Entre-Douro-e-Minho region as our main reference, we have established some main coordinates on chapter 3 that explain the territorial organizational process prompted by the Roman domain. Our analysis assumptions have been guided by two basic ideas: an attempt not to project a rupture concept imposed by the requirements of a radical conquest program and not to value a prejudice assumption related with the alleged continuity of the indigenous settlement under a long-term perspective, as it has been traditionally admitted in historiography. With the available data in mind, we attempted to compare the pre-Roman occupation reality with the one developed under the Roman rule without establishing rigid chronological readings. The Roman organization of the territory in Augustan times only performed an operative premise from which we could consider the beginning of the changing processes occurring in the region without having the available data establishing exact chronologies or the rate of these processes. Our analysis of the pre-Roman settlement reinforced the idea about a clear distinction between the Entre-Douro-eMinho region and the remaining peninsular northwest areas. Thus, the region we considered as our reference matrix offered a rather peculiar settlement pattern. Its main features were articulated with the existence of a dense network of Hill-forts that organized and developed around the hydrographical basins of the main regional rivers and offered a clear picture of potential hierarchy that is rather evident between the Lima and the Ave rivers. The referred hierarchy, already underlined by other authors, seemed to include the interrelationship between different sizes Hill-forts occupying distinct topographic locations that have been revealing a heterogeneous architectonic organization, both in terms of their defensive systems and their internal structure. Only wide settlements, located on sites favouring a wide strategic control, designated in literature as ‘citânias’ and that some authors named as real ‘oppida’, included the generally considered classic proto-urban structures with streets and quarters having allegedly in its dependence on other smaller Hillforts with differentiated features. Today, it is unquestionable that the development features of the large Hill-forts wide performed a pre-Roman phenomenon witnessing a wider social and economic complexity felt by Entre-Douro-e-Minho communities at the

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

end of the Iron Age period. It is possible that these facts corresponded to a reorganization of the settlement network operated after D. Iunius Brutus incursion. In light of the available data and the suggested sites chronology included in the B category, it is possible to admit that the development of the Portuguese northwest ‘citânias’, characterized by a considerable demographic growth and an internal reorganization to which should be added an amazing defensive system, encompassed the transference of communities from medium altitude Hillforts. Simultaneously, low altitude Hill-forts, included in the C category, became more common around the valley areas and they played key roles in exploiting the resources existing in their surroundings. Evidently, this settlement organization model occurring in the wide Entre-Douro-eMinho territory needs further validation and to be included in systematic regional studies, in order to rigorously establish the abandonment chronology of the Hill-forts listed in the B category and understand the overall reorganization of wide Hill-forts that, at this time, seemed to assume a central places. We believe there is an inherent problematic about this research agenda related with the systematic study of the communities mobility and transference, which, assuming rhythms still to be fully understood, shaped the territory. As highlighted on chapter 3, the study of the indigenous community’s mobility processes is paramount for the understanding of both the organization of the pre-Roman world and the different chronologies in terms of the Hillforts maintenance and abandonment. It doesn’t seem possible to consider the existence of a single pre-Roman settlement pattern that was common to all the peninsular northwest, due to the fact that most areas comprising this wide region witnessed different types of sites that were adapted to different geomorphological contexts and to distinct socio-economic realities emerging from rather specific evolutionary patterns. With regards to the Entre-Douro-e-Minho settlement framework, we believe that it could have been the result of the contacts this region maintained with the Mediterranean, which were fully consolidated after the campaigns of D. Iunius Brutus. These relationships prompted a wider interaction of the region with already Romanized territories from Lusitania and the Levant region, which have been corroborated by the transport of products and equally favourable to the circulation of people and inevitable innovations emerging from processes of contact and interchange. In this context, the foundation of Bracara Augusta and the development of the transport network system were part of a wider changing process related with the indigenous world structure, which had started before. These events enabled the integration of communities within the new political structure emerging from the Augustan reforms. The earliness of contacts with the southern peninsular Roman world seemed decisive to us in the overall understanding of the transition complexity of the space social organization between the late first century BC and mid 148

first century AD. Phenomena as the retraction and/or abandonment of the majority of wide ‘citânias’ located in the region, sites where certainly resided the indigenous elites, were the result of population displacement determined by the emerging new centralities. These emerged within the reorganization context associated with the founding of a city, with the structuring of the road networks and with the development of new types of habitats and new ways of economically exploiting the territory. The new settlement matrix installed to address the new needs emerging from the integration of the region within the Roman Empire have the indigenous populations as their main changing agents, which is similar to what has occurred in Gallia and Britannia. In this context, epigraphy revealed that the indigenous substrate was clearly dominant both in the urban and rural settings. Other than the city of Bracara Augusta and the new habitats focused in new ways to exploit resources, diffusely known and for convenience entitled as Villas - we registered the emergence of a network of vici that seemed to have shaped the territory from the mid first century AD onwards. The implantation of this network of new settlements seemed to obey mechanisms outside the inherent planning of cities foundation. In fact, we are facing centres displaying an organic growth and adapted to mobility trends generated by the dispersion of populations from Hill-forts towards the valleys area, alongside needs that were functionally varied in the overall territorial organization, be they economical, transport or entertainment-related or even administrative. The role carried out by some of these vici functioning as civitates headquarters does not exhaust the prominence of some of these sites in the overall (re)organization of the territory. The growing archaeological identification of these type of sites in the Entre-Douro-e-Minho region performed, without any doubts, one of the most interesting features of the Romanization process in this territory and it changed the reference framework of the settlement process. Opposed to the traditional idea related with the long survival of indigenous Hill-forts, it seems now possible to state, due to the nature of emerging data, that pre-Roman settlements were gradually abandoned from the end of the first century BC and in rare exceptions at the end of the first century AD. These trends were certainly related with the emergence of vici, which signalled a preference towards a type of habitat of clear Roman origin instead of the traditional indigenous sites. On chapters 4 and 5 we completed specific focusing of the problems handled in Chapter 3 by analyzing three 3 regional subunits: the region surrounding Bracara Augusta, the mountain region crossed by the Via XVIII itinerary and the valley and coastline regions. On a comparative analysis of the three regional subunits, we highlight the plasticity inherent to Romanization of the space. In fact, there aren’t three distinctive models of territorial occupation, but rather a different adequacy of new settlement patterns to diverse constraints.

Conclusion

We believed that a spatial analysis utilizing different scales enabled an archaeological and cartographic data cross reference, while highlighting the more influent vectors within the territorial planning. The ‘modelling’ concept was used in an attempt to correlate available data with concentration and dispersion areas in an exercise that aimed to organize the quantity and chaotic nature of data. Therefore, the end result mirrors our attempt: we believed we have isolated some different modalities in the overall organization of the Roman settlement pattern and defined some entities that can explain them. The relationship between Bracara Augusta and its surroundings analysed on chapter 4 enabled us to suggest the existence of a strong interaction between the city and the rural world, located up to 6 to 8 km around the city. It is our belief that this relationship isn’t mandatory of a preponderance of the urban world with regards to the rural one and vice versa. Thus, we refer to interaction. Although the archaeological available data revealed that the city witnessed an economic dynamic that clearly distanced it from a common concept of a city while a parasite and consumer centre of its surroundings, it is also certain that we can hardly regard it as the sole centre ‘creating’ or planning the rural space. Archaeological data related with the rural world revealed an intensive occupation of fields and corroborated that the relationships held by city encompassed different modalities. First and foremost, it has been possible to determine through the cartography of occupied areas and sites with great quantities of remains, different sub-urban areas. In the area defined by us and located between 6 to 8 km from the city, both Villas and other types of exploitation units were established in successive rings that started very close to the city and reach the northern Cávado region. They took full advantage of the wide variety of existing resources, including the great agricultural potential of soils, the dense hydrographical network and in particular the resources from the Cávado and Este rivers, including the mountain areas that were key in supplying energy and animal resources and the raw materials with a particular highlight towards clays, needed to produce ceramics or the exploitation of quarries, paramount in providing both the city and the rural world as well as the epigraphic monuments the basic construction material to ensure the continuity of the ‘granite civilization’. If it is certain that Bracara Augusta cannot be defined as a ‘parasite’ city, in the simplistic in the sense of the term, idea of a rural world ordered exclusively for the city doesn’t seem to suit the image obtained from the analysis of available data. Another important factor is the fact that some Hill-forts were maintained within new settlement pattern, assuming functions clear controlling the strategic points in the road network. The ‘construction’ of a Roman rural space around Bracara Augusta became clearly visible when we analyse the remains from Roman cadastre. With the challenges inherent to this task, we believe to have developed a wellfounded proposal over the existence of a cadastre orga149

nized in centuriae of 20 x 20 actus, as attested by numerous signs surpassing the existence of coherent and regular alignments in a wide area. If the identification of a Roman cadastre associated with a city lacking colonial status does no longer represent a problem due to the fact that research work on this particular field have shifted the idea that only colonial territories were subjected to a ‘cadastral’ delimitation, there is still another problem, more or less explicit, related with the city status, and mainly with the justifications given to the organization and delimitation of a territory housing a population not comprised by settlers or Italic immigrants. It is important to state that we are dealing with a city lacking a colonial status and, at the same time, an urban centre that, for several reasons, is very difficult to research in its rural land division. This facts required us to conduct a complex and thorough exercise in terms of substantiate evidences. The exercise was focused on identifying a cadastre on a region rather complex in terms of geomorphological features that has been subjected to radical transformations from the 1970s onwards. However, it still presents sufficient remains attesting a Roman cadastral action on the territory of a city identified by Pliny, the Elder as an oppidum. Although we are aware that the identification work and the overall understanding regarding the land division is still at a preliminary stage, we are certain that in Roman times the territory around Bracara Augusta was subjected to a cadastral delimitation. According to available data, it seemed reasonable to assume that the rural space organization and delimitation had the same intention as of any other space planning and control activities, namely encompassing fiscal control and taxation intentions, and equally serving the fixation of indigenous communities. Similarly to what happened to Bracara Augusta, the rural world primarily served the indigenous population. On chapter 5 we conjugated the analysis of two regions located in the surrounding area of Braga and defining the transection performing the ‘minhoto amphiteatre’, an expression that defines the Entre-Douro-e-Minho landscape. On point 1 we analysed the settlement patterns throughout the Via Nova and tried to assess their differences with regards to the remaining regions. The construction of the Via Nova took place on an historic period after the foundation of the most important cities in the Northwest and the organization subsequent to the conquest and pacification. It belonged to a time important to the consolidation of the imperial power and witnessed the action of the Flavian emperors in Hispania. The Via Nova somehow broke the ‘orthogonality’ of typical Augustan itineraries and presented a pathway that crossed ‘marginal’ lands including the important gold mining exploitation area of ‘El Bierzo’ and other key regions connecting the three northwest cities. Alongside a landscape that shows a settlement progressively scarcer as one move towards the mountainous region. The itinerary course organized the distribution of the populated sites that were preferentially

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

open settlements as opposed to agricultural exploitation units, such as Villas or Farms. Finally, on point 2 we set ourselves to analyse the settlement distribution pattern across the valley and coastline regions. This wide area is framed by the Neiva and Este rivers and closes the defined transection towards a ‘thinner’ analysis. First and foremost, we aimed to highlight the existence of a clear distinction between the scarce populated lands and the wide coastline area with regards to the dense settlement registered in valley regions. The articulation of areas with Roman sites with the Hill-forts areas enabled the observation of communities’ dispersion towards the valley and hills slope areas, which according to available data took place at an early Roman stage. Although Villas and other agricultural exploitation units are still not thoroughly studied and understood, it is possible to state that the transition between the exploitation of resources organized from Hill-forts and the one related with clear Roman models started to occur from mid first century AD onwards, both in valley regions and in coastline areas. We believe this process included different rhythms and modalities that still need further research. The named ‘castros agrícolas’ performed a key topic together with the gradual establishment of rural exploitation units developed by indigenous communities. These questions were critical in the overall understanding of the evolutionary process leading to the foundation of Villas and aedificiae. From a settlement and space planning point of view, it is important to acknowledge the profound assimilation of the Roman organizational models, which were inevitably translated into an acculturation process on behalf of indigenous communities detaining new power. These communities started to live on sites and in houses presenting clear Roman typologies and explored available resources according to an economical logic that was clearly Roman. According to available data regarding both the EntreDouro-e-Minho region and the selected subunits meant to provide a more detailed analytical exercise, we consider that the integration of this area within the Roman Empire translated into a profound transformation of the settlement patterns and the landscape. The nature of the registered changes surpasses the simple political and administrative framework and even an installation of basic infrastructures meant to manage the territory. So, we cannot state that this region Romanization was limited to the urban centres and that the rural world only witnessed an anachronistic indigenous action, when in truth the new settlement matrix established from Augustan times and consolidated in the Flavian and Antonin periods deeply impacted upon the overall territory and led to the transformation of the previous settlement pattern structure. The network of potential vici located in the Entre-Douroe-Minho region is still poorly studied; however, new discoveries have shed some light on the topic. These settlements follow a clear Roman typology and were settled by the indigenous population coming from Hillforts. Assuming a multifunctional character, in the majori150

ty of cases, the consolidation of the vici network, believed to have occurred in the Flavian period, is clearly opposed to the traditional image of an indigenous resistance to the Roman acculturation. Thus, the process of ‘Romanization of Hill-forts" is needed a depth research and reassessment. In practical terms, it seems that the Romanization cartography has mainly been valuing the space and ignoring the time needed for change and favoured, until recently, an undersized reading regarding the overall process. Thus, the Roman Empire cannot be seen as a simple system developed to facilitate the economical domination of the periphery countries, from where the wealth flows mainly oriented towards the centre. As other authors have highlighted, the Roman imperialism owed to the competition for the power inside oligarchies, which was the base for the exploitation and control of the provinces resources. To ensure the economic needs required to consolidate the elites’ power, who also benefited from promotion schemes to get Roman citizenship making them part of the conquered Empire, represented a paramount factor in the creation of a unified dominant class. This process has been well documented in the Western Roman Empire and we also find it well represented in the region under study, where epigraphy leaves no doubts with regards to the origin of elites living in Bracara Augusta or of the Villas owners located in the region and even of the inhabitants settling the new open habitats. The Augustan consignation of a provincial administration model providing power to indigenous elites avoided a wide centralized bureaucracy. On the other hand, the emulation of the indigenous elite power favoured a rapid assimilation of the roman lifestyle. Both public and private architecture in Bracara Augusta, identified in archaeological research, corroborated these events. Furthermore, the necessary dividends in monumentalizing the city seemed to have been ensured through the process of appropriation of private lands, which was previously communal, carried out by indigenous elites and framed upon the centuriation of the ager bracarensis. In the 1930s, Milman Parry found the key that has led to the understanding of the Homeric Poems that were until then a chaotic and unsettling (for the specialists) amount of linguistic, historical and archaeological information. The mix and confusion of different languages and dialects intricate with several historical events, where objects, archaeologically referable to different eras, ran in an amalgamation of verses, indifferent to the time they were from and to the language of each period, made the Homeric Poems to seem to be some sort of a ‘dump’, where in the apparent order reined the chaos. To this chaos, similar to a dump, were indifferent the thousands of Greeks who knew the Poems and heard them recited several times over. Also indifferent were the thousands of people, from Virgil to Camões, who from this dump withdrew the beauty, the metrics and the inspiration. However, in specialist terms, the understanding of the Poems was, in fact, a problem. Milman Parry found the key to decode this chaos when he acknowledged that the Poems were comprised and orally transmitted throughout

Conclusion

centuries before they gained a written form. The composition and the oral transmission belong to a universe that cannot be fixed and therefore not written, absorbing, correlating and mixing objects, events and language under an order that did not obey the regularity of fixed time steps, hence only producing chaos. We believe this short story (simplified) of the famous ‘Homeric question’ will explain what we believe is more attractive in the territories under study. Archaeology cannot be done without facing the shapes, the structures, the objects themselves, and face them as ‘things’ that have a ‘used density’, in order to not isolate and artificially compartment them. Thus, the ‘use density’ caters for change (in function, in use) that does not necessarily correspond to ruptures or discontinuities or continuities, but instead to recycling. This recycling may occur with objects in urban sites and yet, at the most "comprehensive" level of modeling the territory.

151

Catalogue

raphy; honorific epigraphy ; monumental epigraphy ; undetermined epigraphy).

Introduction

The field “chronology” comprises the following items: Chalcolithic; Bronze Age; Iron Age; Roman Period; High empire; Low empire; Middle Ages; High Middle Ages; Late Middle Ages. In cases where we can give another chronology of a monument, as in the case of coins or Milestones, we assign the dating.

This catalogue is the result of an intensive bibliographic research and fieldwork that we achieved in the western façade of the Conventus Bracarensis, located within the present day Portuguese region of Entre-Douro-eMinho. The standardization process inherent to existing historic and archaeological data of the studied area was accomplished with the development of a Geographic Information System, already explained in Chapter I of this work.

This catalog includes the three districts that compose the region of Entre-Douro-e-Minho- Braga, Porto e Viana do Castelo- in an alphabetical order.

For sites classification, we adopted suggested nomenclatures that have been systematized by several authors studying the Proto-historic and Roman settlements for the region under study. We also tried to synthetized the most relevant information for each of the inventoried sites and, among the more than 1.250 references identified, we considered only sites which showed a minimal secure information.

BRAGA DISTRICT

3010202, QUINTA DA PENA, Barreiros, Amares, Braga (30102). CMP 56 Typology: Milestone. Cronology: Roman Period. References: Sousa, J. J. R. 1971-1972; Alarcão 1988; Martins, M. 1990:64-65.

Thus, each site corresponds to an entry, as explained in Chapter I. 3, which contains a series of informative topics, in the following order: SítioID; identification of the site, other names that designate it; parish; county, district; national Geographic code, number of military map (scale 1: 25000); geographical coordinates and altitude (UTM_M; UTM_P; altitude); typology; chronology; references.

3010201, QUINTA DE AGROLONGO, Barreiros, Amares, Braga (30102). CMP 56 [550658.43, 4607404.00, 50]. Typology: Milestone. Cronology: Roman Period. References: Sousa, J. J. R. 1971-1972; Martins, M. 1990: 65. 3011903, MONTE DE S. MIGUEL (Cidadelhe), Bouro (Santa Maria), Amares, Braga (30119). CMP 43 [562090.23, 4613817.16, 550]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period.

Example: Sítio Id: 3010602; Identification of the site: MONTE DE CALDELAS; Other names that designate it: (Monte das Caldas; Outeiro de S. Sebastião; Castelo dos Mouros); Parish: Caldelas; County: Amares; District: Braga;National Geographic code: (30106); Number of military map-scale 1: 25000: CMP 42; Geographical coordinates and altitude- UTM_M; UTM_P; altitude in meters: [552331.14, 4613348.22, 220]; Typology: Hillfort; Chronology: Bronze Age; References. Silva 1958-59 Jorge e Sousa 1980 Martins, M. 1990: 66 Bettencourt, A.1999.

3011902, OUTEIRO, Bouro (Santa Maria), Amares, Braga (30119). CMP 43 [559413.72, 4612632.31, 175]. Typology: Hill-fort. Chronology: Iron Age. References: Martins, M. 1990:70. 3011901, SANTA MARIA DE BOURO, Bouro (Santa Maria), Amares, Braga (30119). CMP 43 Typology: Votive epigraphy. Chronology: Roman Period. References: Santos, L. A.; Le Roux, P. ; Tranoy, A. 1983; Alarcão, J. 1988; Martins, M. 1990: 70-71. 3012002, ALTO DOS CASTELOS (Fraga dos Fornelos), Bouro (Santa Marta), Amares, Braga (30120). CMP 43 [560368.76, 4615999.14, 750]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period.

In the field “typology”, settlement has been classified in the following way: watchtower, Small Farm, Medieval castle, Hill-fort, Romanized Hill-fort (generic term that includes all the Hill-forts with items of roman presence); Pottery Kiln; Medieval habitat (generic classification applied to sites that only reveal signs of a medieval occupation); roman habitat (generic classification applied to habitats that only reveal signs of a roman occupation); undetermined; roman mill; Mansio; Mutatio; Roman mines; Roman necropolis; medieval necropolis ; fluvial port; Farm ; Roman mining site; Milestone; roman bridge ; roman road; salt pans; roman coins; Roman treasure; thermal baths; vicus; Villa; Roman epigraphy (generic designation, with the following subdivisions: Funerary epigraphy ; Votive epig-

3012001, CHÃ GRANDE, Bouro (Santa Marta), Amares, Braga (30120). CMP 43 Typology: Votive epigraphy. Chronology: Roman Period (III A.D.). References: Silva, D. M. 1982; Santos, L. A.; Le Roux, P.; Tranoy, A. 1983; Martins, M. 1990:71. 3010503, CIDADE DE BISCAIA (Campo da Bouça), Caires, Amares, Braga (30105). CMP 56 [554128.77, 4610238.04, 175]. Typology: vicus; Chronology: Roman Period. 152

Catalogue

3010502, QUINTA DE S. VICENTE, Caires, Amares, Braga (30105). CMP 42 [554216.85, 4611418.56, 0]. Typology: Votive epigraphy. Chronology: Roman Period. References: Encarnação, J. 1985; Alarcão, J.1988; Martins, M. 1990:65.

3011101, CAMPO DA IGREJA, Fiscal (S. Miguel), Amares, Braga (30111). CMP 42 [550677.48, 4611732.56, 90]. Typology: Small Farm. Chronology: Roman Period. References: Martins, M. 1990: 67; Alarcão, J. 1988: 1/147.

3010501, CASTRO DE CAIRES (Castro de Gróvios), Caires, Amares, Braga (30105). CMP 56; 42 [553875.20, 4611732.56, 320]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Cardozo, M. 1972; Alarcão, J.1988: 1/150; Martins, M. 1990: 65.

3011302, PAÇO, Lago, Amares, Braga (30113). [60]. Typology: Roman epigraphy- undetermined. Chronology: Roman Period. References: Informação Arqueológica 4, 1981; Alarcão, J. 1988: 1/181 3011301, RIBEIRO, Lago, Amares, Braga (30113). CMP 56 [549058.10, 4607487.86, 50]. Typology: Funerary epigraphy . Chronology: Roman Period. References: Alarcão 1986; Martins, M. 1990; Le Roux, P., Tranoy, A. 1982; Martins, M. ; Silva, A. C. 1984.

3010602, MONTE DE CALDELAS (Monte das Caldas; Outeiro de S. Sebastião; Castelo dos Mouros), Caldelas, Amares, Braga (30106). CMP 42 [552331.14, 4613348.22, 220]. Typology: Hill-fort. Chronology: Bronze Age. References: Silva 1958-59; Jorge e Sousa 1980; Martins, M. 1990: 66; Bettencourt, A.1999.

3011503, CASTRO DE LAGO, Lago, Amares, Braga (30115). CMP 56 [549164.01, 4606889.12, 65]. Typology: Hill-fort. Chronology: Iron Age. References: Teixeira 1955-56: 1973; Silva 1958-59: 139; Sousa 1971-72: 180-181; Martins 1986: 149-184; Martins 1988; Martins 1990: 68-69.

3010601, CALDELAS, Caldelas, Amares, Braga (30106). CMP 42 [552599.19, 4613221.68, 150]. Typology: vicus. Votive epigraphy; thermal baths; Roman necropolis. Chronology: Roman Period. References: Vasconcellos, J. L. 1897,1905, 1913; Bellino, A. 1909; Cardozo, M. 1953; Silva, D. M. 1958-59; Brandão, D. P. 1962; Tranoy, A. 1981a; Martins, M. 1990.

3011502, RIBEIRA DE PALA (Dornelas; Paredes secas), Paredes Secas, Amares, Braga (30115). CMP 56 [555399.02, 4613149.92, 350]. Typology: Milestone. Chronology: 236-238, Roman Period. References: Cunha, A. R. 1961: 319-334; Martins, M. 1990: 71-72; Alarcão 1988: 1/145.

3010704, CASTELO DO CASTRO (Castro da Torre; Solar do Castro), Carrazedo, Amares, Braga (30107). CMP 56 [551445.00, 4608711.05, 96]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Leal 1873-90; Vieira 1886-87; Silva 1958-59; Azevedo 1969; Sousa 1971-72; Cunha 1975, 1977; Martins 1990: 67.

3011501, MOJEJE (Vila Cova), Paredes Secas, Amares, Braga (30115). CMP 42 [554393.23, 4712108.54, 300]. Typology: vicus/Mansio. Chronology: Roman Period. References: Silva, D. M. 1958-59; Cunha, A. R. 1961; Martins, M. 1990; Alarcão, J. 1988: 1/143; Carvalho, H. 1998.

3010701, LUGAR DA IGREJA, Carrazedo, Amares, Braga (30107). CMP 56 [551460.87, 4609121.53, 49]. Typology: Milestone; Votive epigraphy; vicus/mutatio. Chronology: Roman Period. References: Martins, M. 1990; Sousa, J. J. R 1971-1972 ; Alarcão, J. 1988: 1/200.

3011602, CASTELO DE ESPINHO (Castro de S. Pedro Fins), Portela, Amares, Braga (30116). CMP f.56 [553212.03, 4611928.27, 480]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Bellino, A. 1909: 6; Cunha, A. R. 1957; Silva, D. M. 1958-59: 135-148; Leal, P.; Ferreira, A. 1973-90; Martins, M. 1990: 70.

3010702, CARRAZEDO (Campo da Porta), Carrazedo, Amares, Braga (30107). CMP 56 [551440.98, 4609111.33, 90]. Typology: Votive epigraphy. Chronology: Roman Period. References: Sousa, J. J. R. 1971-1972; Cunha, A. R. 1975a; Tranoy, A. 1981a; Martins, M. 1990: 3010901, CERDEIRINHAS, Ferreiros, Amares, Braga (30109). CMP 56 [552540.51, 4609122.04, 100]. Typology: Roman necropolis. Chronology: Roman Period.

3011601, CHÃ DO CASTRO (Portela da Joubreia; Chã de Castros), Portela, Amares, Braga (30116). CMP56 [552534.20, 4610882.32, 310]. Typology: Hill-fort. Chronology: Chalcolithic; Bronze Age; Iron Age. References: Silva 1958-59 Jorge, S. O.1979: 281292; Jorge, S. O.; Sousa, J. J. R. 1980: 121-131; Martins, M. 1990: 69; Bettencourt 1999: 220-221.

3011102, CRUZEIRO DO PILAR, Fiscal (S. Miguel), Amares, Braga (30111). CMP 56 [551129.03, 4610307.08, 150]. Typology: Milestone. Chronology: 282-283 d.C., Roman Period. References: Brandão, D. P. 1962 ; Sousa, J. J. R. 1971-1972; Alarcão, J. 1986; Martins, M. 1990: 68.

3011702, ANJO DA GUARDA- S. MIGUEL-O-ANJO, Prozelo, Amares, Braga (30117). CMP 56 [554952.59, 4607847.01, 70]. Typology: Milestone; Votive epigraphy. Chronology: Roman Period. References: Le Roux, P. e Tranoy, A. 1973: 198-199; Tranoy, A. 1981: 271; Martins, M. 1990: 70; Alarcão 1988: 1/207.

153

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

3012101, MONTE DE SEQUEIROS (Alto ou Couto do Castelhão), Sequeiros, Amares, Braga (30121). CMP 42 [553089.91, 4614126.34, 228]. Typology: Hill-fort. Chronology: Iron Age. References: Silva 1958-59: 353-355; Martins 1990: 71; Bettencourt 1999: 222-223.

3020602, SOUTO, Aldreu, Barcelos, Braga (30206). [524260.69, 4605549.77, 95]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1993: 90; Almeida, C.A.B. 1996: 42. 3020601, CRASTO (Monte Crasto), Aldreu, Barcelos, Braga (30206). [525071.59, 4605044.40, 260]. Typology: Romanized Hill-fort; Medieval castle. Chronology: Roman Period; Middle Ages. References: Almeida, C.A.B. 1993: 89-90; Almeida, C.A.B. 1996: 39-40.

3012201, OUTEIRO DA VILA (Castro de Seramil), Seramil, Amares, Braga (30122). CMP 43 [557188.21, 4614167.64, 458]. Typology: Hill-fort? Chronology: Iron Age? References: Cunha 1961: 321 Martins 1990: 71.

3020702, S. LOURENÇO, Alheira, Barcelos, Braga (30207). [536947.06, 4605603.76, 250]. Typology: Watchtower. Chronology: Roman Period; Middle Ages. References: Almeida, C.A.B. 1996: 43-44.

3012401, RIBEIRA DAS OLIVEIRINHAS VILELA, Vilela, Amares, Braga (30124). CMP 43 [555995.75, 4613455.84, 400]. Typology: Milestone. Chronology: Roman Period. References: Vieira, J. A. 1886-1887; Capella, M. 1895: 60-105; Silva, D. M. 1981: 395; Alarcão, J. 1988: 1/142; Martins, M. 1990: 71-72.

3027101, IGREJA, Alvito (S. Martinho), Barcelos, Braga (30271). [533200.12, 4604845.31, 40]. Typology: Votive epigraphy. Chronology: Roman Period. References: Milhazes, M.C.; Sousa, M.J.C.; Pinto, P.J.C. 1993: 33-40; Almeida, C.A.B. 1996: 44-46

3012402, VILELA, Vilela, Amares, Braga (30124). CMP 43 [556201.71, 4612858.11, 260]. Typology: Votive epigraphy; Milestone. Chronology: Roman Period. References: Carvalho, H. 1998.

3027402, MONTE DE LOUSADO (Cidade Grande), Alvito (S. Pedro), Barcelos, Braga (30274). CMP 55 [534850.56, 4608122.38, 309]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Argote, J. C. 1732: 509; Azevedo, P. A. 1898: 318; Leal, P. 1886: 132; Vieira, J.A. 1886: 59; Figueiredo, A.M. 1895; Fonseca, T. 1948: 93; Silva, A.C.F. 1986; Martins, M. 1990: 73; Queiroga, F. 1992; Almeida, C.A.B. 1996: 46-48.

3020104, CASTELO, Abade de Neiva, Barcelos, Braga (30201). [530565.95, 4600433.22, 160]. Typology: Villa. Chronology: Roman Period. References: Almeida, C.A.B 1993: 67-69; Almeida, C.A.B. 1996: 36-37. 3020103, QUEIJEIROS, Abade de Neiva, Barcelos, Braga (30201). CMP 55 [528792.55, 4601063.89, 150]. Typology: Farm; Roman necropolis; Chronology: Roman Period. References: Almeida, C. A. B. 1979: 19; Costa, A. G; Faria, A. M. Carvalho, J. S. 1980: 11; Alarcão, J. 1988; Martins, M. 1990: 72- 73; Alarcão, J. 1988: 1/174; Almeida, C.A.B. 1996 III: 38-39.

3027401, ALVITO (IGREJA), Alvito (S. Pedro), Barcelos, Braga (30274). CMP 55 [534168.28, 4605744.23, 99]. Typology: Villa. Chronology: Roman Period. References: Fonseca, T. 1948: 103-104; Almeida, C.A.F. 1986: 40; Alarcão, J. 1988: 1/177; Martins, M. 1990: 78, 229; Almeida, C.A.B. et alii 1991: 18; Almeida, C.A.B. 1996: 48-50.

3020102, MONTE DO FACHO (Castro do Alto da Torre), Abade de Neiva, Barcelos, Braga (30201). CMP 55 [529484.69, 4599208.67, 170]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Fonseca, T. 1948: 55; Almeida, C. A. B. ; M. T. Soeiro 1980: 29-32; Costa, A. G; Faria, A. M; Carvalho, J. S. 1980: 12-14; Silva 1986: 77; Queiroga, F. 1992; Almeida, C.A.B. 1996: 35-36.

3021001, ALDEIA / EIDOS, Areias (S. Vicente), Barcelos, Braga (30210). [538151.75, 4601076.43, 50]. Typology: Small Farm? Chronology: Roman Period. References: Almeida, C.A.B.: 50-51 3021104, S. SEBASTIÃO, Areias de Vilar, Barcelos, Braga (30211). [538324.57, 4599072.69, 0]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 220

3020101, AMORIM, Abade de Neiva, Barcelos, Braga (30201). [530606.80, 4599436.70, 60]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1993: 64-65; Almeida, C.A.B. 1996: 36.

3021103, MADALENA, Areias de Vilar, Barcelos, Braga (30211). [536218.14, 4598630.03, 40]. Typology: Roman necropolis. Chronology: Roman Period; References: Bellino, A. 1909: 25; Almeida, C.A.B. 1996: 218

3020501, OUTEIRO DO CRASTO (Monte do Crasto; Castro de Airó; Castelo de Bastuço), Airó, Barcelos, Braga (30205). CMP 69 [538242.48, 4596913.52, 222]. Typology: Romanized Hill-fort; Roman epigraphyundetermined. Chronology: Iron Age; Roman Period. References: Fonseca, T. 1948: 13; 65-66; Soeiro, T. 1983: 67-80; Martins, M. 1990: 73; Almeida, C.A.B. 1993: 82-83; Almeida, C.A.B. 1996: 214-215.

3021102, AVELEIRAS, Areias de Vilar, Barcelos, Braga (30211). [538023.01, 4599132.78, 40]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 217 3021101, IGREJA DE S. JOÃO BAPTISTA, Areias de Vilar, Barcelos, Braga (30211). CMP 69 Typology: 154

Catalogue

Votive epigraphy. Chronology: Roman Period. References: Alarcão, J. 1986: 1/227; Santos, L. A; Le Roux, P; Tranoy, A. 1983: 195; Martins, M. 1990: 7374; Almeida, C.A.B. 1996: 220.

Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 79. 3021802, CARIDE, Carapeços, Barcelos, Braga (30218). [530242.15, 4603506.11, 117]. Typology: Small Farm; Roman necropolis. Chronology: Roman Period. References: Fonseca, T. 1948: 37; Almeida, C.A.B. 1996: 77-78.

3021204, LAJE, Balugães, Barcelos, Braga (30212). [530167.73, 4610198.25, 120]. Typology: Roman necropolis. Chronology: Roman Period- Low Empire. References: Almeida, C.A.B. 1990: 131-146; Almeida, C.A.B. 1996: 57-58.

3021801, CASTRO DA PICARREIRA (Monte do Castro; Castro; Castro de Carapeços), Carapeços, Barcelos, Braga (30218). CMP 55 [530686.75, 4604485.21, 180]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Villas-Boas S/D: 43; Silva 1986: 77; Martins 1990: 74; Queiroga, F.R. 1992: Almeida, C.A.B. 1996: 64-77; Almeida, C.A.B. 2003.

3021203, QUINTA DAS GIESTAS, Balugães, Barcelos, Braga (30212). [530212.97, 4610445.17, 90]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1990: 131-146; Almeida, C.A.B. 1996: 57-58. 3021201, SRª DA APARECIDA, Balugães, Barcelos, Braga (30212). [529472.91, 4610380.41, 150]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Azevedo, P. 1896: 262; Silva, A.C.F. 1986; Almeida, C.A.B. 1990: 134-135; Queiroga, F.R. 1992; Almeida, C.A.B. 1996: 51-53.

3022201, MONTE DA SAIA (Cidade da Citânia; Monte de Fralães), Chavão/Silveiros, Barcelos, Braga (30222). CMP 83 [534777.45, 4589679.34, 303]. Typology: Romanized Hill-fort. Chronology: Bronze Age; Ferro, Roman Period. References: Costa, A. C. 2ª ed. 1868-1869-1706-1712: 333; Torres, J. 1876-1877; Sarmento, F. M. 1933: 166; 1933-160-163; Vasconcelos, J. L. 1897, 1905; 1913: 376, 379; Cunha, A. R. 4/8/46: 3; Cardozo, M. 1957: 179-184; Cardozo, M. 1972: 159; Garcia Y Bellido, A. 1968: 33-34; Romero Masiá, A. 1976: 139, 141; Kalb, P. 1980; Centeno, R. M. S. 1987: 141; Alarcão, J. 1988: 1/231; Martins, M. 1990; Calo Lourido, F. 1991: 427-428; Queiroga, F. 1992: 148; Dinis, A. P. 1993: 34-35; Almeida, C.A.B. 1996: 233-235.

3021202,CARMONA, Balugães / Carvoeiro, Barcelos, Braga (30212). [529527.55, 4611199.92, 300]. Typology: Romanized Hill-fort; Roman epigraphyundetermined. Chronology: Iron Age; Roman Period. References: Argote, J.C. 1732: 200; Costa, C. 1888: 272; Azevedo, P. 1896: 262; Figueiredo, M. 1897: 219 e 283; Guerra, L. 1900: 2-7; Viana, A. 1926: 88-90; Viana, A. 1932: 11-34; Fonseca, T. 1948: 137-138; Paço, A.; Quesado, A.P. 1956: 80-90, 168-179; Neves, L. 1965: 172-180; Almeida, C.A.F. 1978: 33; Silva, A.C.F. 1986; Almeida, C.A.B. 1990: 131-146; Almeida, C.A.B. 1996: 53-56; Maciel, T. 2003: 114-119.

3022301, CRASTO, Chorente, Barcelos, Braga (30223). [186]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Teixeira, C. et alii 1969: 47; Almeida, C.A.B. 1996: 232-233.

3021503, ADRO VELHO, Barqueiros, Barcelos, Braga (30215). [522817.52, 4593394.02, 20]. Typology: Farm? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 224-225.

3022403, CASA DO SOUTO, Cossourado, Barcelos, Braga (30224). [530940.23, 4608104.35, 105]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996 III: 82.

3021501, LAGOA NEGRA, Barqueiros, Barcelos, Braga (30215). [522873.18, 4590464.42, 47]. Typology: Roman mines-gold. Chronology: Roman Period. References: Almeida, C.A.B. 1980: 167; Almeida, C.A.B. 1996: 213.

3022401, CASTRO DE S. SIMÃO, Cossourado, Barcelos, Braga (30224). [531886.46, 4608694.47, 236]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Almeida, C.A.B. ; Baptista, A.J. 1984: 99-100; Silva, A.C.F. 1986: nº 194; Queiroga, F. 1992 nº 21; Almeida, C.A.B. 1996 III: 79-81; Maciel, T. 2003:138-140.

3021703, IGREJA, Campo (S. Salvador), Barcelos, Braga (30217). [532598.71, 4604719.19, 40]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 63 3021701, CRASTO DO MONTE, Campo (S. Salvador), Barcelos, Braga (30217). [532201.93, 4605426.74, 130]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 5960.

3022502, MONTE DO CRASTO (Alto do Crasto), Courel, Barcelos, Braga (30225). [529236.39, 4590846.94, 202]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Fonseca, T. 1948: 138; Silva, A.C.F. 1986 nº 210; Almeida, C.A.B. 1996, III: 236-237.

3021803, MINHOTAS, Carapeços, Barcelos, Braga (30218). [530933.28, 4604341.75, 105]. Typology:

3022501, IGREJA, Courel, Barcelos, Braga (30225). [529626.20, 4590027.62, 120]. Typology: Small Farm. 155

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Chronology: Roman Period. References: Almeida, C.A.B. 1996: III: 237

tins, M. 1990: 74-75; Almeida, C.A.B. 1996 III: 242243.

3022703, OUTEIRO, Creixomil, Barcelos, Braga (30227). [527340.83, 4599946.50, 120]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1991: 129-139; Almeida, C.A.B. 1996: 84-85.

3023202, MONTE DE S. MAMEDE (Penedo do Ladrão; Guarita), Feitos / Vila Cova / Vilar do Monte, Barcelos, Braga (30232). CMP 55 [526486.71, 4601690.45, 409]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Fonseca, T. 1948: 232; Silva 1986 Nº 199: 77; Martins, M. 1990: 7; Queiroga, F. 1992: Nº 25; Almeida, C.A.B. 1996 III: 62.

3022802, PAÇO, Cristelo, Barcelos, Braga (30228). [526118.89, 4591061.38, 28]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996 III: 239-240.

3023302, S. PAIO, Fonte Coberta, Barcelos, Braga (30233). [537284.67, 4594040.22, 100]. Typology: Medieval necropolis. Chronology: Middle Ages. References: Almeida, C.A.B. 1996 III: 244

3022801, IGREJA, Cristelo, Barcelos, Braga (30228). [525581.53, 4592138.85, 48]. Typology: Farm. Chronology: Roman Period. References: Fonseca, T. 1948:141 Almeida, C.A.B. 1996 III: 237- 239.

3023402, ALTO DA DEVEZA/GIESTAL (Giestal), Fornelos, Barcelos, Braga (30234). [527379.39, 4595106.09, 43]. Typology: Hill-fort. Chronology: Iron Age; Roman Period. References: Fonseca, T. 1948:168; Almeida, C.A.B. 1996 III: 245-246.

3022905, CHÃ DE MILHOS, Durrães, Barcelos, Braga (30229). [527945.25, 4608706.94, 230]. Typology: Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996 III: 89-91

3023401, AGRA DA VILA (Boavista), Fornelos, Barcelos, Braga (30234). CMP 69 [50]. Typology: Roman habitat; Pottery Kiln. Chronology: Roman Period. References: Fonseca, T. 1948: 167; Alarcão, J. 1988: 1/224; Martins, M. 1990: 75; Almeida, C.A.B. 1996 III: 246-247.

3022904, DURRÃES, Durrães, Barcelos, Braga (30229). [ 0]. Typology: Roman treasure. Chronology: Roman Period. References: Almeida, C.A.B. 1996 III: 89. 3022903, ALTO DA GIESTA (Pau da Cabeça; Monte da Bouça da Giesta), Durrães, Barcelos, Braga (30229). [527296.96, 4608535.49, 298]. Typology: Hill-fort. Chronology: Idade do Bronze; Iron Age. References: Silva, A.C.F. 1986: 189; Queiroga, F. 1992: Nº 23; Almeida, C.A.B. 1996 III: 88-89; Bettencourt, A.M.S. 1999 II: 233-234.

3026801, SANTA MARIA DE GALEGOS, GalegosStª Maria, Barcelos, Braga (30268). CMP 55 [535877.76, 4601681.67, 130]. Typology: Balneum from the Iron Age. Chronology: Iron Age. References: Silva, A. C. F. 1986: 56-58; Martins, M. 1990: 77; Queiroga, F. 1992: Nº 26; Almeida, C.A.B. 1996 III: 94-98.

3022902, CASTELOS (Crasto; Alto dos Castelos), Durrães, Barcelos, Braga (30229). [528126.13, 4609073.42, 160]. Typology: Hill-fort? Chronology: Iron Age. References: Silva, A.C.F 1986: Nº 190; Queiroga, F. 1992: Nº 16; Almeida, C.A.B. 1993: 7677; Almeida, C.A.B. 1996 III: 86-88; Maciel, T. 2003: 141-143.

3023601, ASSENTO, Gamil, Barcelos, Braga (30236). [534234.37, 4596708.22, 52]. Typology: Medieval necropolis. Chronology: Middle Ages. References: Almeida, C.A.B. 1996 III: 247 3023701, CASTRO DE FARIA (Monte de Faria; Castelo de Faria), Gilmonde, Barcelos, Braga (30237). CMP 69 [529562.74, 4594004.58, 298]. Typology: Romanized Hill-fort; Medieval castle. Chronology: Chalcolithic; Bronze Age; Iron Age; Roman Period; Middle Ages. References: Valero, A. J. S. s/d: 11-14; Maluquer, M. J. 1948: 33-38; Santa, O. J. M. 1948: 2128; Costa, A. G.; Faria, A. M. ; Carvalho, J. S. 1980: 23-28; Almeida, C. A. B. 1982: 79-80; Almeida, C. A. B. 1985: 51; Silva, A.C.F. 1986: 77, Nº 206; Martins, M. 1990: 75-76; Almeida, C.A.B. 1996, III: 291-388.

3022901, ALTO DO SENHOR DO LÍRIO, Durrães, Barcelos, Braga (30229). [527605.22, 4609658.38, 140]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Silva, A.C.F. 1986: Nº 188; Queiroga, F. 1992: Nº 22; Almeida, C.A.B. 1996 III: 85-86. 3023001, CASA DO ADRO, Encourados, Barcelos, Braga (30230). [539024.72, 4598151.34, 70]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996: III: 240

3023802, ASSENTO, Góios, Barcelos, Braga (30238). [532079.26, 4591362.95, 110]. Typology: Small Farm. Chronology: Roman Period. References: Fonseca, T. 1948: 199; Almeida, C.A.B. 1996, III: 249

3023101, IGREJA DE FARIA; ASSENTO, Faria, Barcelos, Braga (30231). CMP 83 [528004.87, 4592168.20, 50]. Typology: Small Farm; Roman necropolis. Chronology: Roman Period. References: Fonseca, T. 1948: 160; Alarcão, J. 1988: 1/226; Mar-

3023801, PORTELA, Góios, Barcelos, Braga (30238). [531171.68, 4592037.46, 130]. Typology: Small Farm. 156

Catalogue

Chronology: Roman Period. References: Almeida, C.A.B. 1996 III: 248

3025301, EIRA DOS MOUROS, Negreiros, Barcelos, Braga (30253). [531007.43, 4587071.52, 85]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1996 III: 272-273; Almeida, C.A.B. 2003.

3024502, TELHEIRAS, Manhente, Barcelos, Braga (30245). [534825.39, 4599116.61, 35]. Typology: Pottery Kiln. Chronology: Roman Period. References: Almeida, C.A.B. 1996 III: 103-104

3025403, PAÇO, Oliveira, Barcelos, Braga (30254). [537647.22, 4604527.96, 130]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996 III: 112

3024501, ASSENTO, Manhente, Barcelos, Braga (30245). [535657.11, 4599614.13, 35]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 102-103

3025402, IGREJA, Oliveira, Barcelos, Braga (30254). [537883.14, 4603696.60, 100]. Typology: Small Farm? Chronology: Roman Period; Medieval? References: Almeida, C.A.B. 1996 III: 111.

3024601, IGREJA, Mariz, Barcelos, Braga (30246). [527440.14, 4597419.33, 61]. Typology: Roman habitat?. Chronology: Roman Period. References: Almeida, C.A.B. 1996 III: 100-101; Almeida, C.A.B. 2003: 173

3025401, CITÂNIA DE RORIZ (Monte do Facho; Cidade de Canhoane; Eira dos Mouros; Monte dos Cabrões), Oliveira, Barcelos, Braga (30254). CMP 55 [536774.15, 4603043.19, 324]. Typology: Romanized Hill-fort; Roman treasure; Iron Age warrior; Chronology: Bronze Age, Iron Age , Roman Period. References: Villas-Boas, J. S. P. 1947; Almeida, C. A. B. ; Soeiro, M.T. 1980: 33-34; Kalb, Ph. 1980: 27; Almeida, C. A. B. 1983: 81-90; Silva, A.C.F. 1986; Martins, M. 1990: 77; Queiroga, F. 1992; Almeida, C.A.B. 1996: 104-111; Bettencourt, A. 2003.

3024701, IGREJA DE MARTIM, Martim, Barcelos, Braga (30247). CMP 69 [540945.82, 4598484.17, 100]. Typology: Mansio? Chronology: Roman Period. References: Almeida, C. A. F. 1972: 319-326; Alarcão, J. 1988: 1/233; Martins, M. 1990: 76; Almeida, C.A.B. et alii 1991: 15-16; Almeida, C.A.B. 1996: III: 250254. 3024801, AGRA DE ALÉM, Midões, Barcelos, Braga (30248). [534910.73, 4595847.93, 60]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996, III: 255; Almeida, C.A.B. 2003: 173

3025601, MONDIM, Panque, Barcelos, Braga (30256). [533460.15, 4608670.78, 150]. Typology: Votive epigraphy. Chronology: Roman Period. References: Tranoy, A. 1981: 317; Santos, L. A.; Le Roux, P.; Tranoy, A. 1983: 184-185; Alarcão, J. 1988: 1/176; Almeida, C.A.B. 1996 III: 112-114.

3025001, MINHOTÃES-IGREJA, Minhotães, Barcelos, Braga (30250). [100]. Typology: Votive epigraphy. Chronology: Roman Period. References: Santos, L. A.; Le Roux, P.; Tranoy, A. 1983: 189-190; Alarcão, J. 1988: 1/288; Almeida, C.A.B. 1996 III: 257-258.

3026002, MOURIZ, Perelhal, Barcelos, Braga (30260). [526187.71, 4597661.39, 60]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996 III: 116-117

3025102, PAÇO, Monte de Fralães, Barcelos, Braga (30251). [536499.07, 4589719.80, 120]. Typology: Villa. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 266.

3026001, IGREJA, Perelhal, Barcelos, Braga (30260). [525515.18, 4597782.34, 60]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996 III: 115-116

3025101, IGREJA, Monte de Fralães, Barcelos, Braga (30251). [537080.76, 4589496.05, 168]. Typology: Funerary epigraphy. Chronology: Roman Period. References: Costa; C. 1869: 333; Cardozo, M. 1972 nº 43: 85; Alarcão, J. 1988: 1/13; Ferreira, J. 1992: 3754 ; E.E. VIII: nº 279: 504; IX, nº 526: 100; Almeida, C.A.B. 1996 III: 269-270; Revista de Guimarães, XVIII, nº 69: 69.

3026104, PAÇO, Pousa, Barcelos, Braga (30261). [540843.75, 4600959.09, 70]. Typology:Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996 III: 276-277 3026102, VERDASCA, Pousa, Barcelos, Braga (30261). [540641.01, 4599208.75, 130]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B.1996 III: 275

3025201, QUINTA DO LOURIDO, Moure, Barcelos, Braga (30252). CMP 69 [537002.78, 4594748.10, 100]. Typology: Small Farm; Roman treasure. Chronology: Roman Period. References: Fonseca, T. 1948: 278; Hipólito, M. C. 1960-1961: 16; Centeno, R.M.S. 1978: 25; Centeno, R.M.S. 1987: 70; Alarcão, J. 1988: 1/132; Martins, M. 1990:76; Almeida, C.A.B. 1996 III: 270-272

3026202, ASSENTO, Quintiães, Barcelos, Braga (30262). [529253.87, 4607357.23, 120]. Typology: Villa. Chronology: Roman Period. References: Almeida, C.A.B. 1996:120 3026201, SANTA MARINHA, Quintiães, Barcelos, Braga (30262). [529257.26, 4606524.42, 230]. Typology: Hill-fort. Chronology: Iron Age; Roman Period? 157

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

References: Fonseca, T. 1948: 347; Silva, A.C.F. 1986: nº 191; Queiroga, F. 1992: nº 29; Almeida, C.A.B. 1996: 118-120; Almeida, C.A.B. 2003

Queiroga, F. 1992 nº 24; Almeida, C.A.B. 1996 III: 118-120. 3028201, CRASTO (Monte do Castro), Ucha, Barcelos, Braga (30282). CMP 55 [540042.11, 4602567.17, 58]. Typology: Farm. Chronology: Roman Period. References: Fonseca, T. 1948: 397; Dinis 1993: 103; Silva 1986: 78, Nº 210; Silva, A.C.F. 1986: Nº 205; Martins, M. 1990: 79; Queiroga, F. 1992: Nº 32; Almeida, C.A.B. 1996 III: 128-130; Almeida, C.A.B. 2003.

3026301, TORRE DE MOLDES, Remelhe, Barcelos, Braga (30263). [533465.73, 4592510.33, 192]. Typology: Small Farm? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 277-278. 3026501, ASSENTO, Rio Covo (Stª Eugénia), Barcelos, Braga (30265). [533975.91, 4597447.17, 52]. Typology: Villa? Chronology: Roman Period? References: Almeida, C.A.B. 1996: 279-280

3028301, ASSENTO (S. Bento da Varzea), Várzea, Barcelos, Braga (30283). CMP 69 [535094.39, 4596218.91, 40]. Typology: Villa. Chronology: Roman Period. References: Rosário, A. 1973; Martins, M. 1990: 79; Almeida, C.A.B. 1996: 287-288

3026602, QUINTA DA COSTA, Rio Covo (Stª Eulália), Barcelos, Braga (30266). [534639.36, 4594428.07, 160]. Typology: Roman necropolis. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 284

3028402, QUINTA DA FONTE VELHA, Viatodos, Barcelos, Braga (30284). [535803.30, 4588820.71, 140]. Typology: Small Farm? Chronology: Roman Period. References: Fortes, J. 1905: 110-111; Fonseca, T. 1948: 379; Villas- Boas, J.S.P. 1948: 13-14; Almeida, C.A.B. 1996 III: 289-290.

3026601, SENHORA DE ÁGUAS SANTAS, Rio Covo (Stª Eulália), Barcelos, Braga (30266). CMP 69 [534963.07, 4593138.44, 75]. Typology: vicus. Chronology: Roman Period. References: Almeida, C. A. F. 1972: 126 Almeida, C. A. F. 1970: 100-101 Alarcão, J. 1988: 1/229 Martins, M. 1990: 77 Almeida, C.A.B. 1996 III: 282-283

3028401, SOUTO, Viatodos, Barcelos, Braga (30284). [536431.84, 4588422.93, 120]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996 III: 287-288

3028003, OUTEIRO, Silveiros, Barcelos, Braga (30280). [535178.12, 4591675.57, 120]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996 III: 285-286

3028501, QUINTA DO PASSAL, Vila Boa, Barcelos, Braga (30285). [531664.47, 4600663.65, 70]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996 III: 130-131

3027601, IGREJA, Tamel (S. Fins), Barcelos, Braga (30276). [531418.44, 4604590.58, 85]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996 III: 122-123

3028605, MERECES, Vila Cova, Barcelos, Braga (30286). [523287.54, 4600914.97, 60]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996 III: 142-143

3027702, TELHEIRAS, Tamel (S. Veríssimo), Barcelos, Braga (30277). [533731.15, 4600221.67, 35]. Typology: Pottery Kiln. Chronology: Roman Period. References: Almeida, C.A.B. 1996 III: 126-127

3028604, S. SALVADOR DO BANHO, Vila Cova, Barcelos, Braga (30286). [524298.68, 4598494.87, 40]. Typology: Roman necropolis. Chronology: Roman Period. References: Almeida, C.A.B. 1996 III: 137-142

3027701, IGREJA, Tamel (S. Veríssimo), Barcelos, Braga (30277). [534705.08, 4600041.25, 39]. Typology: Pottery Kiln? Chronology: Roman Period? References: Almeida, C.A.B. 1996 III: 126

3028603, CAMPO DO LOMBO, Vila Cova, Barcelos, Braga (30286). [523542.26, 4598700.83, 50]. Typology: Small Farm? Chronology: Roman Period? References: Abreu, A.A. 1989; Almeida, C.A.B. 1996 III: 132-133.

3026702, PAÇO, Tamel (Stª Leocádia), Barcelos, Braga (30267). [529547.69, 4603441.55, 150]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996 III: 125

3028601, ALVARÃO, Vila Cova, Barcelos, Braga (30286). [522661.90, 4598636.33, 40]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: III: 131-132; Almeida, C.A.B. 2003

3026701, IGREJA, Tamel (Stª Leocádia), Barcelos, Braga (30267). [529778.75, 4603565.86, 120]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 123-124

3028602, PAÇO-VILA COVA, Vila Cova, Barcelos, Braga (3028602, 30286). CMP 68 [523768.08, 4600490.30, 70]. Typology: Villa; Roman necropolis; thermal baths. Chronology: Roman Period. References: Fonseca, T. 1948: 143-414; Alarcão, J. 1988: 1/175;

3028101, PICOTO DOS MOUROS (Alto do Mouro), Tregosa, Barcelos, Braga (30281). [526889.33, 4608957.24, 226]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Fonseca, T 1948: 348; Silva, A.C.F. 1986 nº 187; 158

Catalogue

Martins, M. 1990: 79-80; Almeida, C.A.B. 1996 III: 143-195.

1981: 303; Alarcão, J. 1988: 1/214; Martins, M. 1990: 80; Fontes, L. 1994: 37.

3027303, PAÇO VELHO, Vila Frescaínha (S. Martinho), Barcelos, Braga (30273). [529801.84, 4597983.83, 50]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1991: 134; Almeida, C.A.B. 1996 III: 134.

3035102, SANTA TECLA, Braga (S. Vítor), Braga, Braga (30351). CMP 70 [549436.40, 4599694.21, 210]. Typology: Roman treasure; watchtower. Chronology: Roman Period; Middle Ages. References: Vasconcellos, J. L. 1918: 357; Alarcão, J. 1988: 1/199; Martins, M. 1990: 88; Fontes, L. 1994: 75.

3027302, TELHEIRAS, Vila Frescaínha (S. Martinho), Barcelos, Braga (30273). [531424.21, 4597959.88, 22]. Typology: Pottery Kiln . Chronology: Roman Period. References: Almeida, C.A.B. 1996 III: 133-134

3035101, AREAL DE BAIXO, Braga (S. Vítor), Braga, Braga (30351). CMP 56 [549121.46, 4601190.55, 210]. Typology: Milestone. Chronology: Roman Period (333-350 d.C.). References: Fortes, J. 1905-1908: 124; Alarcão, J. 1988: 1/192; Martins, M. 1990: 88-89.

3027301, VILA FRESCAÍNHA, Vila Frescaínha (S. Martinho), Barcelos, Braga (30273). CMP 69 [530449.30, 4598325.78, 20]. Typology: Roman necropolis Chronology: Roman Period. References: Rosário, A. 1973: nº 655; Alarcão, J. 1988: 1/228; Martins, M. 1990: 78; Almeida, C.A.B. 1996 III: 135136.

3030702, BRACARA AUGUSTA, Cividade, Braga, Braga (30307). CMP 70 [547936.04, 4599779.09, 200]. Typology: Roman town, capital of Conventus Bracaraugustanus; capital of Gallaecia. Remarks: The description and urban evolution of the roman town of Bracara Augusta is described in the first part of this work.

3028901, PAÇO, Vilar do Monte, Barcelos, Braga (30289). [527918.84, 4600713.31, 180]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996 III: 136-137

3031002, CARQUEMIJE, Dume, Braga, Braga (30310). CMP 56 [547510.08, 4602373.96, 100]. Typology: Roman necropolis; Funerary epigraphy. Chronology: Roman Period. References: Le Roux ; P. Tranoy, A. 1982 : 32-34; Martins, M. 1990: 83; Fontes, L. 1994: 44.

3020502, PAÇO, Airó, Barcelos, Braga (30205). [536882.75, 4595549.34, 85]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1993: 84-85 Almeida, C.A.B. 1996 III: 216

3031001, DUME-IGREJA (Assento; Igreja), Dume, Braga, Braga (30310). CMP 56 [547113.27, 4602070.06, 90]. Typology: Villa; Suevic church; Roman epigraphy. Chronology: Roman Period; Middle Ages. References: Hubner 1869/1982; CIL II 2414, 2426 , 2433, 2444, 2456; 2427; Dias, A. R. 1903: 133; Cardozo, M. 1972: 63 ; Tranoy, A. 1981: 202-316-329330; Le Roux, P ; Tranoy, A. 1982 :32-34; Fontes, L. 1987: 113-148; Alarcão, J. 1988: 1/191; Martins, M. 1990: 81; Fontes, L. 1991-1992: 199-230; Fontes, L. 1994: 339. 43.

3022202, BOUÇA DOS CRASTINHOS, Chavão, Barcelos, Braga (30222). [531138.32, 4589014.99, 150]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996 III: 230231; Almeida, C.A.B. 2003 3033901, MONTE DE VASCONCELOS, Adaúfe, Braga, Braga (30339). CMP 56 [551990.26, 4602921.13, 356]. Typology: Hill-fort. Chronology: Bronze Age; Iron Age . References: Martins, M. 1990: 87; Dinis, A.P. 1993: 37-38; Fontes, L. 1994: 36; Bettencourt, A.M.S. 1999: 251, vol. 2a.

3031201, ESPINHO-IGREJA (Assento), Espinho, Braga, Braga (30312). CMP 70 [553231.78, 4600032.26, 440]. Typology: Funerary epigraphy ; Villa. Chronology: Roman Period. . References: Santos, L. A.; Le Roux P.; Tranoy, A. 1983: 198; Alarcão, J. 1988: 1/216; Fontes, L. 1994: 45.

3030103, ADAÚFE-IGREJA, Adaúfe, Braga, Braga (30301). CMP 56 [550190.83, 4604200.33, 100]. Typology: Votive epigraphy; Villa. Chronology: Roman Period. References: Vasconcelos, L. 1905: 334-335; Fortes, J. 1905-1908: 479; Cunha, 1961: 319; Blázquez Martinez, J. M. 1962: 133; Encarnação, J. 1972: 103-104; Encarnação, J. 1975: 220-221; Tranoy, A. 1981: 303; Martins, M. 1990: 80-81; Fontes, L. 1994: 35-36.

3031302, MONTE DA FALPERRA (Santa Marta das Cortiças), Esporões, Braga, Braga (30313). CMP 70 [550625.68, 4596229.63, 563]. Typology: Romanized Hill-fort; Suevo-Visigothic abbey. Chronology: Bronze Age; Iron Age, Roman Period; High Middle Ages. References: Sarmento, F. M. 1933: 166; Bellino, A 1909: 15-18; Cunha, A. R. 1948: 3,6; Cunha, A. R. 21/8/49: 2; Cunha, A. R. 20/8/52: 3; Cunha, A. R. 12/3/53: 4; Cunha, A. R. 14/10/54: 1, 4; Cunha, A. R. 1953-54: 241-243; Cunha, A. R. 14/1/55: 1,3; Sousa, J. J. R. 1968-1970: 57-64; Teixeira, C., Medeiros, A.C., Macedo, J. R. 1973: 47; Cunha, A. R. 1975: 9-18, 487-

3030101, QUINTA DE SANTO ANDRÉ /MONTE DO OURO, Adaúfe, Braga, Braga (30301). CMP 56 [548977.26, 4605587.67, 73]. Typology: Funerary epigraphy . Chronology: Roman Period. References: Fortes, J. 1905-1908: 479; Blázquez Martinez, J. M. 1962: 133; Encarnação, J. 1975: 220-22; Tranoy, A. 159

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

496; Castro, J. S., Correia, S.H., Oliveira, E.P. 1980: 39; Feio, A. 1984: 99; Ponte, S. 1984: 136, nº 63; Silva, A. C. F. 1986: 79, nº 240; Alarcão, J. 1988: 1/244; Martins, M. 1990: 119; Queiroga, F. 1992: 149, nº 35; Dinis, A. P. 1993: 77-79; Fontes, L. 1994: 47-48; Bettencourt, A. 1999: 255-256.

3031901, IGREJA DE S. MIGUEL DE GUALTAR, Gualtar (S. Miguel), Braga, Braga (30319). CMP 56 [551080.30, 4602281.06, 240]. Typology: Roman necropolis; Villa. Chronology: Roman Period. References: Bellino, A. 1909: 5; Alarcão, J. 1988: 1/215; Martins, M. 1990: 84; Fontes, L. 1994. 55-56.

3031301, ESPORÕES-IGREJA, Esporões, Braga, Braga (30313). CMP 70 [548576.96, 4595686.88, 220]. Typology: Votive epigraphy; Villa. Chronology: Roman Period; Middle Ages. Santos, L. A.; Le Roux P.; Tranoy, A. 1983 : 201 Fontes, L. 1994: 46

3032002, S. MAMEDE-GUIZANDE (Monte Redondo), Guizande, Braga, Braga (30320). CMP 70 [546494.76, 4611118.25, 427]. Typology: Romanized Hill-fort; Votive epigraphy. Chronology: Iron Age; Roman Period. References: Vasconcelos, J. L. 1897, 1905, 1913: 75, 334; Fortes, J. 1905-1908: 17-18; Bellino, A. 1909: 7-15, 19-28; Teixeira, C. 1955-56: 231; Encarnação, J. 1970: 215-216; Cunha, A. R. 1975: 497-498; Castro, J. S., Correia, S.H. , Oliveira, E. P. 1980: 39; Ponte, S. 1980: 116, vol. II; Silva, A. C. F. 1986: 79, nº 239; Alarcão, J. 1988: 1/238; Calo Lourido, F. 1991: 341 – 346; Queiroga, F. 1992: 149, nº 37; Dinis, A. P. 1993: 68-70; Fontes, L. 1994: 56.

3034401, EIRAS VELHAS, Este (S. Mamede), Braga, Braga (30344). CMP 56 [554830.75, 4603990.84, 480]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Teixeira, C.; Medeiros, A.C.; Macedo, J.R. 1973: 47; Oliveira, E. P., Moura, E.S., Mesquita, J. 1982: 8-9; Silva, A. C. F. 1986: 79, nº 241; Martins, M. 1990: 87; Queiroga, F. 1992: 149, nº 41; Dinis, A. P. 1993: 38-41.

3032301, LOMAR – IGREJA VELHA (Assento), Lomar, Braga, Braga (30323). CMP 70 [547757.22, 4597677.99, 50]. Typology: Milestone; Funerary epigraphy; Villa. Chronology: Roman Period. References: Hubner 1869/1982; CIL II 4764; 2434 e 2438; Tranoy, A. 1981: 202; Alarcão, J. 1988: 1/234; Fontes, L. 1994: 57; Mantas, V. 1996: 408-411.

3034601, ALTO DO CRASTO (Crasto de Espinho), Este (S. Pedro), Braga, Braga (30346). CMP 70 [552887.27, 4600614.53, 441]. Typology: Hill-fort? Chronology: Bronze Age. References: Bellino, A. 1909: 5; Cunha, A. 1975: 31; Dinis, A.P. 1993: 85-86; Fontes, L. 1994: 51; Bettencourt, A.M.S. 1999: 264, Vol. 2a.

3032902, CASTRO MAU, Merelim (S. Pedro), Braga, Braga (30329). CMP 56 [545993.64, 4604058.14, 50]. Typology: Small Farm. Chronology: Roman Period. References: Sampaio, A. 1963: 265; Martins, M. 1990: 88; Fontes, L. 1994: 58.

3031401, MONTE DE S. GREGÓRIO (Castro de Ferreiros), Ferreiros/Maximinos / Gondizalves, Braga, Braga (30314). CMP 70 [546360.64, 4599295.08, 155]. Typology: Small Farm. Chronology: Roman Period. References: Teixeira, C. 1955-56: 7; Teixeira, C. et alii 1973: 48; Silva, A.C.F. 1986: 79, nº 233; Martins, M. 1990: 83; Queiroga, F. 1992: 149, nº 36; Dinis, A.P. 1993: 36; Fontes, L. 1994: 52, 54.

3032903, S. PEDRO DE MERELIM-IGREJA, Merelim (S. Pedro), Braga, Braga (30329). CMP 56 [545499.88, 4603453.37, 50]. Typology: Funerary epigraphy . Chronology: Roman Period (I /II d.C). References: Hubner 1869/1982; CIL, II, 2450; Vives, J. 1972, ILER, 5112; Tranoy, A. 1981: 202; Alarcão, J. 1988: 1/186; Martins, M. 1990: 88; Fontes, L. 1994: 59.

3031501, PIPE, Figueiredo, Braga, Braga (30315). CMP 70 [546388.95, 4594565.26, 230]. Typology: Roman treasure. Chronology: Roman Period (III/IV d.C.). References: Bellino, A. 1909: 7; Hipólito, M. C. 1960-1961: 20; Alarcão, J. 1988: 1/237; Fontes, L. 1994: 53.

3032904, PINHEL-PATEIRA, Merelim (S. Pedro), Braga, Braga (30329). CMP 56 [546200.25, 4603300.30, 70]. Typology: Villa. Chronology: Roman Period.

3031701, QUINTA DO OUTEIRO (Quinta do Eiteiro), Frossos, Braga, Braga (30317). CMP 56 Typology: Milestone. Chronology: Roman Period. References: Almeida, C. A. B. 1979: 72; Araújo, J. R. 1982: 135; Alarcão, J. 1988: 1/193; Martins, M. 1990: 83; Fontes, L. 1994: 53-54.

3032701, BOUÇA ALTA, Navarra, Braga, Braga (30327). CMP 56 [551664.04, 4606814.32, 85]. Typology: Mutatio or mansio. Chronology: Roman Period. References: Alarcão, J. 1988: 1/210; Martins, M. 1990: 84; Fontes, L. 1994: 61.

3031801, MONTE DAS CALDAS (S. Mamede), Gondizalves /Sequeira, Braga, Braga (30318). CMP 70 [544371.47, 4598879.1, 304]. Typology: Romanized Hill-fort. Chronology: Bronze Age; Iron Age; Roman Period. References: Bellino, A. 1909: 18-19; Cunha, A. R. 1947; Teixeira, C. 1955-56: 6; 231-254; Cunha, A. R. 1975: 499-500; Fontes 1994 : 54; Martins, M. 1990 : 83-84; Bettencourt, A. 1999 : 257-259, vol 2ª.

3032901, CASTRO DA CONSOLAÇÃO, Nogueiró, Braga, Braga (30329). CMP 70 [551338.67, 4600109.86, 340]. Typology: Hill-fort. Chronology: Iron Age; References: Costa 1706: 185; Bellino, A. 1909: 4-5; Teixeira 1936: 232; Feio 1956-1984: 99; Teixeira et alii 1973: 47; Rosário 1974: 23, 59; Cunha (1975: 485-536; Castro et alii 1980: 39; Oliveira 1982: 160

Catalogue

8-9; Silva 1986: 79; Queiroga, F. 1992:149; Dinis, A. P. 1993: 79-81.

1982: 143; Alarcão, J. 1988: 1/197; Martins, M. 1990: 87; Fontes, L. 1994: 71.

3033103, SANTO ESTEVÃO (Santo Estevão), Palmeira, Braga, Braga (30331). CMP 56 [545889.66, 4604456.96, 60]. Typology: Small Farm? Chronology: Roman Period. References: Santos, M. M. 1977: 1518; Santos, L. A.; Le Roux, P. ; Tranoy, A. 1983: 202; Alarcão, J. 1988: 1/185; Martins, M. 1990: 84-85; Fontes, L. 1993: 63.

3033701, S. FRUTUOSO DE MONTÉLIOS, Real, Braga, Braga (30337). CMP 56 [546921.04, 4601268.4, 85]. Typology: Villa? Roman treasure. Chronology: Roman; Middle Ages. References: Belllino, A. 1909: 5; Hipólito, M. C. 1960-1961: 21; Coutinho, J.M. 1978: 38; Alarcão, J. 1988: 1/195; Martins, M. 1990: 87; Fontes, L. 1994: 69-71.

3033102, SANTO ESTEVÃO, (Santo Estevão Velho/ Pardelhas), Palmeira, Braga, Braga (30331). CMP 56 Typology: Villa; Chronology: Roman Period. References: Fontes, L. 1993.

3033703, CASTRO MÁXIMO (Monte Castro), S. Vicente, Braga, Braga (30337). CMP 56 [547833.44, 4601134.34, 170]. Typology: Romanized Hill-fort? Chronology: Iron Age; Roman Period. References: Freitas, B. S. 1890: 320; Bellino, A. 1909: 5; Teixeira, C. 1955-56: 17-38; Costa, A. J. 1965: 33-35, 38-40; Castro, J. S., Correia, S., Oliveira, E. P. 1980: 37-51; Martins, M. 1990: 86; Fontes, L. 1994: 72-73.

3033101, OUTEIRO DE S. SEBASTIÃO, Palmeira, Braga, Braga (30331). CMP 56 [547987.74, 4604578.04, 115]. Typology: Villa; Roman treasure. Chronology: Roman Period. References: Santos, L. A.; Le Roux, P. ; Tranoy, A. 1983: 201-202; Alarcão, J. 1988: R. P. 1/188; Martins, M. 1990: 84; Fontes, L. 1994: 62.

3030352, ÍNFIAS/PACHANCHO, S. Victor, Braga, Braga. CMP 56 [548334.43, 4600740.96, 70]. Typology: Villa. Chronology: Roman Period. References: Costa 1965: 38-40; Oliveira 1979:174-175 e 182; Fontes, L. 1994: 74.

3033202, PADRÃO; IGREJA, Panóias, Braga, Braga (30338). CMP 56 [544808.23, 4602646.68, 55]. Typology: Milestone. Chronology: Roman Period (32/33 d.C.). References: Bellino, A. 1896: 45-48; Santos, L. A. 1979: 14; Almeida, C. A. B. 1979: 100, 101,143; Araújo, J. R. 1982: 147-148; Alarcão, J. 1988: 1/190; Martins, M. 1990: 85-86; Fontes, L. 1994: 63-64.

3033103, MONTARIOL (Fonte de S. Vicente), S. Vicente, Braga, Braga (30351). CMP 56 [548705.54, 4602785.85, 150]. Typology: Votive epigraphy; archa petrinea. Chronology: Roman Period. References: Tranoy, A. 1981: 304; Santos, L. A., Le Roux, P., Tranoy, A. 1983: 192; Alarcão, J. 1988: 1/196; Martins, M. 1990: 85; Fontes, L. 1994: 73.

3033201, QUINTA DE GERMIL, Panóias, Braga, Braga (30332). CMP 56 [545104.08, 4603049.52, 50]. Typology: Milestone; Chronology: Roman Period (3233 d.C). References: Bellino A. 1896: 45-48, 53; Dias, A. R. 1903: 133; Cardozo, M. 1972: 78; Almeida, C. A. B. 1979: 143-149; Araújo, J. R. 1982: 144-146; Alarcão 1988: 1/189; Martins, M. 1990: 85; Fontes, L. 1994: 64.

3035303, SENHOR DOS LÍRIOS -SANTARÃO, Semelhe, Braga, Braga (30353). CMP 70 [544928.29, 4600648.56, 120]. Typology: Honorific epigraphy¸ Chronology: Roman Period; Medieval. References: CIL, II, 2450; ILER, 1028; Bellino, A. 1896: 5; Dias, A. R. 1903: 278; Bellino, A. 1909: 18; Cardozo, M. 1972: 66; Le Roux, P. 1980: 155; Tranoy, A. 1981: 328; Alarcão, J. 1988: 1/194; Martins, M. 1990: 89-90; Fontes, L. 1994: 77.

3033501, LAJES, Pousada, Braga, Braga (30335). CMP 56 [554863.70, 4606746.55, 120]. Typology: Villa. Chronology: Roman Period. References: Alarcão, J. 1988: 1/212; Martins, M. 1990: 88; Fontes, L. 1994: 68-69.

3035302, MONTE DE CONES (Cide), Semelhe, Braga, Braga (30353). CMP 70 [546333.71, 4600062.89, 110]. Typology: Undetermined habitat; Milestone. Chronology: Roman Period. References: CIL, II, 4756; ILER, 1858; Capella, M. 1895: 170; Bellino A. 1909: 5; Almeida, C. A. B. 1979: 143; Alarcão, J. 1988: 1/232; Martins, M. 1990: 90; Fontes, L. 1994: 76.

3034604, CASAS NOVAS, Real, Braga, Braga (30346). CMP 56 [547200.20, 4601700.70, 70]. Typology: Undetermined. Chronology: Roman Period. 3034603, COTURELA-PARRETAS, Real, Braga, Braga (30346). CMP 70 [547099.83, 4600732.49, 130]. Typology: Villa. Chronology: Roman Period. References: Costa 1965: 38-40; Fontes, L. 1994: 69.

3035301, SANTARÃO, Semelhe, Braga, Braga (30353). CMP 70 [545625.98, 4600855.51, 80]. Typology: Villa. Chronology: Roman Period; Middle Ages References: EE, VIII, 280; ILER, 1028; Bellino, A. 1896: 5; Dias, A. R. 1903: 278; Bellino, A. 1909: 18; Cardozo, M. 1972: 66; Le Roux, P. 1980: 155; Tranoy, A. 1981: 328; Alarcão, J. 1988: 1/194; Martins, M. 1990: 89-90; Fontes, L. 1994: 76-77.

3033702, QUINTA GRANDE DE TOURIDO, Real, Braga, Braga (30337). CMP 70 [546327.68, 4600662.63, 60]. Typology: monument wrongly identified as a Milestone. Chronology: undetermined. References: Almeida, C. A. B. 1979: 98; Araújo, J. R. 161

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

3030106, PEDROSO, Adaúfe, Braga, Braga (30301). CMP 56 [550444.40, 4602994.83, 330]. Typology: Hill-fort. Chronology: Bronze Age; Iron Age. References: Fontes, L. 1990: 131; Fontes, L. 1994: 36-37; Bettencourt, A.M.S. 1999 : 252, Vol 2.

pology: Roman necropolis; Chronology: Roman Period.

Roman

habitat.

3030104, FONTELA, Adaúfe, Braga, Braga (30301). CMP 56 [549188.24, 4604490.15, 85]. Typology:Small Farm. Chronology: Roman Period. References: Fontes, L. 1994: 35.

3041402, SANTA COMBA, Refojos de Basto, Cabeceiras de Basto, Braga (30414). [585603.00, 4595849.00, 310]. Typology: Romanized Hill-fort; Iron Age warrior; vicus. Chronology: Iron Age; Roman Period. References: CIL II 2382; Martins, M., Silva, A. C. 1984: 39-40; Alarcão, J. 1988: 1/270, 1/269; Tranoy, A. 1981: 399.

3035104, LARGO DO ORFEÃO, Braga (S. Vítor), Braga, Braga (30351). CMP 70 [549327.40, 4600592.82, 180]. Typology: Milestone. Chronology: 32 d.C. References: Martins, M. 1990: 89.

3040001, MONTE DO CASTELO (Castelo), S. Nicolau, Cabeceiras de Basto, Braga (30400). CMP 72 [582974.00, 4600632.00, 550]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period.

3031402, PAU DE BANDEIRA /MONTE CRASTO, Este (S. Mamede), Braga, Braga (30314). CMP 56 [554895.17, 4601616.49, 570]. Typology: Hill-fort. Chronology: Idade do Bronze; Iron Age. References: Sarmento, F.M 1883-1884/1933: 165, 171; Macedo, M.D. 1896: 121-123; Bellino, A. 1909: 5; Teixeira, C. 1936: 232-233; Teixeira, C. et alii 1973: 47; Sousa, J.J.R. 1978: 329-336; Silva, A.C.F. 1986: 79; Queiroga; F. 1992: 150; Dinis, A.P. 1993: 87-88; Fontes, L. 1994: 50; Bettencourt, A.M.S. 1999: 262263.

3040701, ALTO DOS MOUROS, Cavez, Cabeceiras de Basto, Braga (30407). CMP 73 [591164.00, 4596211.00, 283]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. 3051701, VILA BOA DE MONTELONGO, Rego, Celorico de Basto, Braga (30517). Typology: Roman epigraphy- undetermined. Chronology: Roman Period. References: Araújo, I. 1980: 113; Alarcão, J. 1988: 1/329. 3060105, IGREJA, Antas (S. Paio), Esposende, Braga (30601). [519606.97, 4605720.61, 120]. Typology: Small Farm? Chronology: Roman Period. References: Almeida, C.A.B. 1986: 44-46; Almeida, C.A.B. 1996: Vol. IV, 15-16.

3040101, ABADIM, Abadim, Cabeceiras de Basto, Braga (30401). CMP 72 [588043.00, 4596250.00, 610]. Typology: Epigrafia honorífica. Chronology: Roman Period. References: Brandão, D. P. 1984: 367-374; Oliveira, E. P. 1978: 28; Alarcão 1988: 1/268.

3060104, AGRA DO RELÓGIO, Antas (S. Paio), Esposende, Braga (30601). [518955.53, 4606952.48, 22]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1986: 51-52; Almeida, C.A.B. 1996: 13-14, Vol. IV.

3040401, SANTA SENHORINHA DE BASTO, Basto, Cabeceiras de Basto, Braga (30404). [, 200]. Typology: Epigrafia honorífica. Chronology: Roman Period. References: Hubner 1869/1982: CIL II 2381. Alarcão 1988: 1/272.

3060102, BELINHO, Antas (S. Paio), Esposende, Braga (30601). [518773.99, 4605564.19, 137]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Almeida, C.A.B. 1986: 39-59; Almeida, C.A.B. 1996: Vol. IV, 10-12.

3040601, CIVIDADE DE CHACIM (Outeiro dos Moiros; Subidade), Cabeceiras de Basto, Cabeceiras de Basto, Braga (30406). CMP 72 [587161.00, 4598822.00, 640]. Typology: Hill-fort. Chronology: Iron Age. References: Cunha, A.R. 950; Araújo, I. 1980: 105; Silva, A.C. 1986: 82.

3060101, QUINTA DE BELINHO, Antas (S. Paio), Esposende, Braga (30601). [518264.14, 4605778.76, 62]. Typology: Villa. Chronology: Roman Period. References: Almeida, C.A.B. 1989: 98-101; Almeida, C.A.B. 1996:14-15.

3041301, ALTO DOS MOIROS, Pedraça, Cabeceiras de Basto, Braga (30413). CMP 73 [588903.00, 4593922.00, 320]. Typology: Hill-fort. Chronology: Iron Age. Inédito.

3060201, VILA MENDO, Apúlia, Esposende, Braga (30602). [518879.36, 4591840.66, 6]. Typology: Villa. Chronology: Roman Period. References: Barbosa, F. 1937; Almeida, C.A.B. 1988: 34-40; Almeida, C.A.B. 1996, vol. IV. 12-13; Gomes, J.M.F. 1996.

3041302, BOADELA (Cemitério), Pedraça, Cabeceiras de Basto, Braga (30414). CMP 72 [588043.00, 4596250.00, 400]. Typology: Roman necropolis; Achado monetário Roman Period. Chronology: Roman Period. References: Vieira, J. A. 1886: 528; Alarcão 1988: 1/271.

3060701, OUTEIRO DOS PICOUTOS (Outeiro da Felícia), Fonte Boa, Esposende, Braga (30607). [521627.13, 4595986.61, 58]. Typology: Mansio. Chronology: Iron Age; Roman Period. References: Costa, A. C. 1868; Fonseca, T. 1936; Almeida, C.A.B.

3041303, MURO, Pedraça, Cabeceiras de Basto, Braga (30413). CMP 72 [587190.00, 4594522.00, 270]. Ty162

Catalogue

1980; Almeida, C.A.B. 1988: 32-33; Almeida, C.A.B. 1996: Vol. IV; C. A. B. Almeida 2003.

4589684.12, 332]. Typology: Romanized Hill-fort; Iron Age warrior. Chronology: I a.C. - I d.C., Iron Age. References: Sarmento, M 1925: 77-80; Cuevilias, F. L., Fernandez, J. L. 1938: 103; Silva, A.C.F. 1986: 82; Martins, M. 1981; Martins, M. 1987.

3061301, LINHARIÇA, Palmeira de Faro, Esposende, Braga (30613). CMP 68 [521448.94, 4601315.71, 130]. Typology: Villa. Chronology: Roman Period. References: Almeida, C.A.B. 1989: 91; Almeida, C.A.B. 1996: Vol. IV.

3071301, SANTA MARINHA (Monte de Santa Marinha), Freitas, Fafe, Braga (30713). CMP 71 [564977.21, 4595151.83, 597]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period.

3061503, QUINTELA, Vila Chã, Esposende, Braga (30615). Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1985: 42-43; Alarcão, J. 1988: 1/173; Almeida, C.A.B. 1996: Vol. IV.

3072501, CASTRO DA PORTELA (Crasto; Penedos de S. João; Monte de Mouros), Ribeiros, Fafe, Braga (30725). CMP 72 [, 500]. Typology: Hill-fort. Chronology: Iron Age. References: Silva, A.C.F. 1986: 82; Oliveira, A.L. s. d: 95-97.

3061502, COVELOS, Vila Chã, Esposende, Braga (30615). [190]. Typology: Roman necropolis; Small Farm. Chronology: Roman Period. References: Almeida C.A.B. 1985: 41-42; Alarcão, J. 1988: 1/172; Almeida, C.A.B. 1996: Vol. IV; Almeida, C.A.B. 2003.

3072801, CIVIDADE DE PAREDES (Subidade de Paredes), S. Gens, Fafe, Braga (30728). CMP 86 [574700.00, 4589600.00, 580]. Typology: Hill-fort. Chronology: Iron Age. References: Silva, A.C.F. 1986: 82; Oliveira, A.L. s.d.: 98-100.

3061501, CASTRO DE S. LOURENÇO, Vila Chã, Esposende, Braga (30615). [520060.48, 4600818.26, 200]. Typology: Romanized Hill-fort. Chronology: Bronze Age; Ferro; Roman Period. References: Almeida, C.A.B., Neiva, M.A.P. 1982: 7; Santos, L.A., Le Roux, P., Tranoy, A. 1983: 190; Alarcão, J., 1988: 1/171; Almeida, C.A.B. 1990-92: 151; Almeida, C.A.B. 1996: Vol. IV, 46-122.

3073301, CASTANHEIRA, Travassós, Fafe, Braga (30733). CMP 71 [565386.79, 4594156.24, 495]. Typology: Roman habitat. Chronology: Roman Period. References: Alarcão, J. 1988: 1/264. 3086402, ABAÇÃO, Abação (S. Tomé), Guimarães, Braga (30864). CMP 85 [560700.00, 4584900.00, 390]. Typology: Villa; Roman necropolis. Chronology: Roman Period. References: Sarmento, M. 1933: 217; Sarmento, M. 1970: 37; Cardozo, M. 1972: 108; Alarcão, J. 1988: 1/321.

3060107, ALTO DA PONTE, Antas (S. Paio), Esposende, Braga (30601). [518654.07, 4607167.58, 47]. Typology: Mutatio. Chronology: Roman Period. References: Almeida, C.A.B. 1986: 52-53; Almeida, C.A.B. 1996: Vol. IV, 16-18. 3060106, BOUÇA DO RIO, Antas (S. Paio), Esposende, Braga (30601). [519928.59, 4606615.87, 80]. Typology: Small Farm? Chronology: Roman Period? References: Almeida, C.A.B. 1986: 51; Almeida, C.A.B. 1996: Vol. IV, 15-16.

3086401, MONTE DA LAPINHA-DEVESA ESCURA, Abação (S. Tomé), Guimarães, Braga (30864). CMP 85 [562050.00, 4585300.00, 434]. Typology: Roman necropolis. Chronology: Roman Period. References: Pina, L. 1930: 6-107; Azevedo, A.P. 1896: 3; Silva 1986: nº 293; Alarcão, J. 1988: 1/320; Queiroga, F. 1992: nº 70, p. 152

3060103, MONTE, Antas (S. Paio), Esposende, Braga (30601). [519930.04, 4606091.59, 90]. Typology: Romanized Hill-fort. Chronology: Bronze Age; Iron Age; Roman Period. References: Almeida, C.A.B. 1986: 39-59; Almeida, C.A.B. 1996: Vol. IV, 12-13; Maciel, T. 2003: 173.

3080101, ALDÃO, Aldão, Guimarães, Braga (30801). CMP 85 [560750.00, 4590800.00, 256]. Typology: Roman necropolis?. Chronology: Roman Period. References: Alarcão 1988: 1/267. 3080602, ALMAS, Barco (S. Claúdio), Guimarães, Braga (30806). CMP 70 [ 100]. Typology: Roman habitat. Chronology: Roman Period. References: Sarmento, M. 1901: 126; Alarcão, J. 1988: 1/251.

3061201, MARINHAS, Marinhas, Esposende, Braga (30612). [517968.20, 4603773.39, 8]. Typology: Villa. Chronology: Roman Period. References: Almeida, C.A.B. 1989: 92; Almeida, C.A.B. 1996: Vol. IV.

3080601, COTO DE SABROSO, Barco (S.Claúdio), Guimarães, Braga (30806). CMP 70 Typology: Roman epigraphy- undetermined. Chronology Roman Period?. References: Cardozo, M. 1972: 3; Alarcão, J. 1988: 1/247.

3070701, ALTO DA RETORTINHA (Castro de Cepães), Cepães, Fafe, Braga (30707). CMP 85 [566655.83, 4587271.03, 250]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Silva, A. C. F. 1986: 82; Oliveira, A.L. s.d: 60-61.

3084101, CITÂNIA DE BRITEIROS, Briteiros (Salvador), Guimarães, Braga (30841). CMP 71 [556961.38, 4596970.73, 336]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period.

3070901, SANTO OVÍDIO (Monte de Santo Ovídio), Fafe, Fafe, Braga (30709). CMP 85 [568031.21, 163

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

References: Cardozo, M. 1931; Cardozo, M. 1943: 247-256; Cardozo, M. 1946: 289-308; Cardozo, M. 1948: 343-348; Cardozo, M. 1949: 406-414; Cardozo, M. 1950: 518-526; Cardozo, M. 1951: 455-472; Cardozo, M. 1951: 455-472; Cardozo, M. 1952: 348358; Cardozo, M. 1953: 711-717; Cardozo, M. 1954: 412-416; Cardozo, M. 1955: 431-438; Cardozo, M. 1956: 508-512; Cardozo, M. 1957: 551-553; Cardozo, M. 1958: 455-459; Cardozo, M. 1959: 517-520; Cardozo, M. 1960: 551-554; Cardozo, M. 1961: 451454; Cardozo, M. 1965; Silva, A.C.F., Centeno, R.M.S. 1977: 277-280; Silva, A.C.F., Centeno, R.M.S. 1978: 421-430; Silva, A.C.F. 1986: 80; Queiroga, F. 1992: 152; Alarcão, J. 1988: 1/243.

3081201, COSTA, Costa (Santa Marinha), Guimarães, Braga (30812). CMP 85 [560500.00, 4588500.00, 270]. Typology: Roman habitat. Chronology: Roman Period. References: Sarmento, M. 1933: 233; Alarcão, J. 1988: 1/317. 3081301, CREIXOMIL, Creixomil (S. Miguel), Guimarães, Braga (30813). CMP 85 [557800.00, 4587800.00, 150]. Typology: Funerary epigraphy. Chronology: Roman Period. References: Cardozo, M. 1972: 111; CIL II 5554; Alarcão, J. 1988: 1/303. 3081502, MONTE DE STª EULÁLIA (Monte de Ulaia), Fermentões, Guimarães, Braga (30815). CMP 85 [556850.00, 4589400.00, 285]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Sarmento, M. 1933: 262-264; Silva, A.C.F. 1986; Queiroga, F. 1992.

3084301, SANTA LEOCÁDIA DE BRITEIROS, Briteiros (Stª Leocádia), Guimarães, Braga (30843). CMP 70 [, 150]. Typology: Roman epigraphy- undetermined. Chronology: Roman Period. References: Santos, L. A., Le Roux, P. e Tranoy, A. 1983: 194; Alarcão, J. 1988: 1/242.

3081501, PAÇO, Fermentões, Guimarães, Braga (30815). CMP 85 [556750.00, 4590700.00, 170]. Typology: Roman habitat. Chronology: Roman Period. References: Sarmento, M. 1933: 262-263; Alarcão, J. 1988: 1/262.

3080702, PENA, Brito, Guimarães, Braga (30807). CMP 84 [554100.00, 4589300.00, 120]. Typology: Roman habitat. Chronology: Roman Period. References: Sarmento, M. 1901: 117; Alarcão 1988: 1/299.

3081701, IGREJA, Gandarela, Guimarães, Braga (30817). CMP 84 [554707.81, 4582752.68, 210]. Typology: Roman habitat. Chronology: Roman Period. References: Sarmento, M. 1933: 250-251; Alarcão 1988: 1/311.

3080701, BRITO, Brito, Guimarães, Braga (30807). CMP 84 [553100.00, 4589620.00, 140]. Typology: Roman epigraphy- undetermined. Chronology: Roman Period. References: Sarmento, M. 1970: 34; Alarcão, J. 1988: 1/298.

3081901, GOMINHÃES, Gominhães, Guimarães, Braga (30819). CMP 71 [559906.37, 4592401.73, 0]. Typology: Roman habitat. Chronology: Roman Period. References: Guimarães, F. J. S. 1978: 410 e 415.

3085901, AIDRO, Caldas de Vizela (S. Miguel), Guimarães, Braga (30859). CMP 85 / 99 [557824.56, 4580684.47, 150]. Typology: Votive epigraphy. Chronology: Roman Period. References: Cardozo, M. 1972: 33; Encarnação, J. 1975: 191-192; Encarnação, J. 1970: 230-231; Alarcão, J. 1988.

3082301,IGREJA-GUARDIZELA, Guardizela, Guimarães, Braga (30823). CMP 84 [552910.57, 4582234.73, 160]. Typology: Roman habitat. Chronology: Roman Period. References: Sarmento, M. 1933: 256; Silva, A. C. F. 1986: Nº 295; Queiroga, F. 1992:153 Nº 80.

3080801, CALDAS DAS TAIPAS, Caldelas, Guimarães, Braga (30808). CMP 70 [554700.48, 4593149.21, 120]. Typology: vicus. Chronology: Roman. References: CIL II 4796; 5560; Guimarães, F. J. S. 1978: 411; Capella, M. 1895: 119; Brandão, D. P. 1854: 29-30; Alarcão, J. 1988: R P 1/257; Martins, M. 1995: 92-93.

3082501, IGREJA, Infias, Guimarães, Braga (30825). CMP 85 [557900.00, 4583300.00, 240]. Typology: Villa. Chronology: Roman Period. References: Sarmento, M. 1933: 211; Alarcão, J. 1988: 1/310. 3082703, PEDRAIS, Longos, Guimarães, Braga (30827). CMP 70 [554215.39, 4596522.45, 268]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Dinis, A. P. 1993: 86; Sarmento, F. M. 1901: 123-4; Pina, L. 1940: 509-510, 512; Silva, A. C. F. 1986: 80, nº 264; Queiroga, F. 1992: 153-154, nº 82.

3082401, MATAMÁ-IGREJA, Costa, Guimarães, Braga (30824). CMP 85. Typology: Roman habitat? Chronology: Roman Period. References: Sarmento, M. 1933: 22; Alarcão, J. 1988: 1/319. 3082402, MONTE DE STº ANTONINHO (Monte de Stº Antoinho), Costa, Guimarães, Braga (30824). [563750.00, 4588450.00, 526]. Typology: Hill-fort. Chronology: Bronze Age; Iron Age. References: Sarmento, M. 1933: 223; Kalb, Ph. 1980: 58, nº 21; Coffyn, A. 1985: 223; Silva, A.C.F. 1986: nº 284; Queiroga, F. 1992, n º75: 153.

3082702, CASA DA BÓCA, Longos, Guimarães, Braga (30827). CMP 70 [, 240]. Typology: Roman necropolis. Chronology: Roman Period. References: Museu, R.G. 45 1935: 182; Cardoso, M. 1936: 84; Guimarães, F. J. S. 1978: 409; Alarcão, J. 1988: 1/246; 1/245. 164

Catalogue

3082704, CASTRO DE SABROSO (Monte de Sabroso), Longos/S. Lourenço de Sande/S. Salvador, Guimarães, Braga (30827). CMP 70 [555079.37, 4595696.27, 278]. Typology: Hill-fort. Chronology: Iron Age. References: Sarmento, F. M. 1879 – 1933: 22-35; Sarmento, F. M. 1933: 166, 168; Sarmento, F. M. 1896; Cartailhac, E. 1888: 124-129; Fortes, J. 1904; Pinto, R. S. 1928-29; Cardozo, M. 1958: 439-446; Cardozo, M. 1930; 1980; Hawkes, C. F. C. 1958: 446453; Hawkes, C. F. C. 1971: 283-287; Romero Masiá, A. 1976; Kalb, P. 1980; Soeiro, T., Centeno, R. M. S., Silva, A. C. F. 1981: 341-350; Ponte, S. 1984: 128138; Silva, A. C. F. 1986: 265; Centeno, R. M. S. 1987: 117; Calo Lourido, F. 1991: 414-426; Queiroga, F. 1992: 153; Dinis, A. P. 1993: 88-90.

man habitat. Chronology: Roman Period. References: Cardozo, M. 1972: 55; Sarmento, M. 1933: 266; Alarcão, J. 1988: 1/259, 1/260. 3084201, IGREJA, Prazins (Santa Eufémia), Guimarães, Braga (30842). CMP 71 [, 140]. Typology: Roman habitat. Chronology: Roman Period. References: Sarmento, M. 1933: 267; Alarcão, J. 1988: 1/256. 3084801, MONTE CASTRO, Prazins (Stº Tirso), Guimarães, Braga (30848). CMP 71 [558135.30, 4593345.42, 288]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period? References: Sarmento, M. 1933: 266-268; Silva, A.C.F. 1986; Queiroga, F. 1992: 154.

3082901, PICOTO DE SANTO AMARO, Mascotelos, Guimarães, Braga (30829). CMP 85 [557010.75, 4585965.33, 309]. Typology: Romanized Hill-fort Chronology: Iron Age; Roman Period. References: Guimarães, J.S. 1970/ 80: 141-150, 381-386; Guimarães, J.S. 1971: 436-470; Guimarães, J.S. 1978: 40; Pina, L. 1940: 513; Ponte, S. s/d: 11-144; Sarmento, M. 1933: 237-239; Silva, A. C. F. 1986: nº 287; Queiroga, F. 1992: 154, nº 83.

3084001, RONFE, Ronfe, Guimarães, Braga (30840). CMP 84 [551500.00, 4587900.00, 140]. Typology: Ara divindade indígena; Ara a Júpiter. Chronology: Roman Period. References: Cardozo, M. 1972: 31; Encarnação, J. 1975: 177-179; Santos, L. A., Le Roux, P. e Tranoy, A. 1983: 185; Gomes, R. V. e Gomes, M. V. 1984: 4354; Vasconcelos 1905: 64, 329-331; Silva 1986: 81 Nº 278; Queiroga 1992: 154 Nº 93; Dinis 1993: 104. 3085201, CALDAS DE VIZELA, S. João e S. Miguel, Guimarães, Braga (30852). CMP 99 [557824.56, 4580684.47, 150]. Typology: vicus. Chronology: Roman Period. References: Hubner 1869/1982: CIL II 2408, 5538; Pereira Caldas, J. J. S. 1852: 7-15; Cardozo, M. 1972: 64, 88, 26-27; Acuña Castroviejo, F. 1974: 45-51; Encarnação, J. 1975: 143-148; Le Roux, P., Tranoy, A. 1979: 57; Costa, A. J. 1981: 138; Tranoy, A. 1981: 393; Alarcão, J. 1988: 1/315.

3083001, VEIGAS, Mesão Frio, Guimarães, Braga (30830). CMP 85 Typology: Roman habitat. Chronology: Roman Period. References: Sarmento, M. 1933: 222; Alarcão, J. 1988: 1/316. 3083102, VELA, Moreira de Cónegos, Guimarães, Braga (30831). CMP 98 [554300.00, 4581100.00, 150]. Typology: Roman necropolis. Chronology: Roman Period. References: Sarmento, M. 1970: 58; Sarmento, F. 1933: 252-255; Guimarães, F. J. S. 1978: 409; Alarcão, J. 1988: 1/314, 1/313.

3086502, VILAR, S. Torcato, Guimarães, Braga (30865). CMP 71 [, 380]. Typology: Roman habitat. Chronology: Roman Period. References: Sarmento, M. 1933: 272; Alarcão, A. M., Martins, Alina N. 1976; Guimarães, F. J. S. 1978: 410; Alarcão 1988: 1/266, 1/265.

3083201, NESPEREIRA, Nespereira, Guimarães, Braga (30832). CMP 85 [557000.00, 4584200.00, 160]. Typology: Roman habitat?. Chronology: Roman Period. References: Silva, A.C.F. 1986: nº290; Queiroga F., 1992: nº 85, p. 154; Sarmento, M. 1933: 251; Alarcão, J. 1988: 1/308.

3084902, MONTE DE S. BARTOLOMEU, Sande (S. Clemente), Guimarães, Braga (30849). CMP 70 [551627.70, 4592093.15, 421]. Typology: Hill-fort. Chronology: Iron Age. References: Ayres 1896: 388; Sarmento 1901: 119-120; Silva 1986: 80; Dinis 1993: 81-83.

3083301, OLEIROS, Oleiros (S. Vicente), Guimarães, Braga (30833). CMP 84 [550800.00, 4590400.00, 220]. Typology: Roman treasure; Thermal baths. Chronology: Roman Period. References: Alarcão, J. 1988: 1/**; Cardozo, M. 1950: 479; Chaves, L. 1936: 58; Acuña Castroviejo, F. 1974: 205; Acuña Castroviejo, F. 1974: 35.

3084901, QUINTA DA MOGADA, Sande (S. Clemente), Guimarães, Braga (30849). CMP 70 [150]. Typology: Roman habitat. Chronology: Roman Period. References: Bellino, A. 1909: 4; Alarcão, J. 1988: 1/255.

3083803, QUINTA DA RIBEIRA, Ponte (S. João), Guimarães, Braga (30838). CMP 70 [ 130]. Typology: Pottery Kiln. Chronology: Roman Period. References: Pina, L. 1928: 58-63; Zbyszewski G., Almeida, F., Ferreira, O. V. 1971: 156; Alarcão, J. 1988: 1/261.

3085801, S. MARTINHO DE SANDE, Sande (S. Martinho), Guimarães, Braga (30858). CMP 70 [140]. Typology: Milestone. Chronology: Roman Period. References: CIL II 6214; Cardozo, M. 1972: 79; Capella, M. 1895: 118; Alarcão 1988: 1/253.

3083802, S. JOÃO DA PONTE-IGREJA, Ponte (S. João), Guimarães, Braga (30838). CMP 70 [556416.84, 4591466.98, 133]. Typology: Votive epigraphy; Ro165

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

3087301, MONTE DA FORCA (Moinho de Vento), Sande (Vila Nova), Guimarães, Braga (30873). CMP 84 [552794.35, 4591145.5, 266]. Typology: Hill-fort. Chronology: Bronze Age. References: Sarmento, F. M. 1901: 118; Sarmento, F. M. 1933: 166; Pina, L. 1940: 509; Silva, A. C. F. 1986: 80; Queiroga, F. 1992: 155; Dinis, A. P. 1993: 83-85.

Hill-fort. Chronology: Iron Age; Roman Period. References: Dinis, A. P. 1993: 74-76; Sarmento, F. M. 1933: 166, 169; Sarmento, F. M. 1901: 20; Ayres, C. 1896: 388; Pinto, R. S. 1928: 194; Pinto, R. S. 1929: 194; Kalb, P. 1980; Silva, A. C. F. 1986: 81; Queiroga, F. 1992: 154. 3085101, S. FAUSTINO DE VIZELA, Vizela (S. Faustino), Guimarães, Braga (30851). CMP 99 [560850.00, 4582350.00, 279]. Typology: Votive epigraphy. Chronology: Roman Period. References: CIL II 5566; Cardozo, M. 1972: 111; Alarcão, J. 1988: 1/323.

3085402, MONTE DA SANTA, Selho (S. Jorge), Guimarães, Braga (30854). CMP 84 [553271.19, 4586245.25, 249]. Typology: Romanized Hill-fort? Chronology: Bronze Age; Ferro; Roman Period? References: Sarmento 1933: 256-259; Kalb 1980: nº 19; Silva 1986: nº 279; Queiroga 1992.

3000308, GUIMARÃES, Guimarães, Guimarães, Braga (308). CMP 85 [559198.80, 4588195.92, 170]. Typology: Votive epigraphy. Chronology: Roman Period. References: Cardozo, M. 1972: 45, 47, 56; Alarcão, J. 1988: 1/302.

3086601, SERZEDELO, Serzedelo, Guimarães, Braga (30866). CMP 84 [552950.00, 4583850.00, 145]. Typology: Votive epigraphy. Chronology: Roman Period. References: Tranoy, A. 1981: 273, 317-318; Cardozo, M. 1972: 29, 54; Encarnação, J. 1975: 160-161; Alarcão, J. 1988: 1/309.

3090301, LUGAR DA IGREJA, Brunhais, Póvoa de Lanhoso, Braga (0). CMP 57 [568086.07, 4602980.78, 240]. Typology: Roman necropolis. Chronology: Roman Period References: Carvalho, H. 1991-1992.

3086802, SILVARES, Silvares, Guimarães, Braga (30868). CMP 84 [555550.00, 4588500.00, 140]. Typology: Roman habitat; Roman necropolis. Chronology: Roman Period. References: Sarmento, M. 1933: 261; Alarcão, J. 1988: 1/301.

3090501, VILAR, Campo (S. Martinho), Póvoa de Lanhoso, Braga (30905). CMP 57 [559439.25, 4599094.98]. Typology: Roman habitat. Chronology: Roman Period.

3086801, CAMPELOS, Silvares, Guimarães, Braga (30868). CMP 84 [554950.00, 4590100.00, 120]. Typology: Roman treasure. Chronology: Roman Period. References: Sarmento, M. 1933: 266; Alarcão, J. 1988: 1/263.

3091101, LAMAÇÃS, Friande, Póvoa de Lanhoso, Braga (30911). CMP 57 [140]. Typology: Roman habitat. Chronology: Roman Period. References: Martins, M. 1990: 90. 3091301, SALGUEIROS, Garfe, Póvoa de Lanhoso, Braga 30913). CMP 71 [563237.71, 4599133.15, 180]. Typology: Roman necropolis. Chronology: Roman Perioda. References: Carvalho, H. 1991-1992.

3086201, QUINTA DO REGUENGO, Souto (S. Salvador), Guimarães, Braga (30862). CMP 71 [558479.84, 4595086.57, 0]. Typology: Roman necropolis. Chronology: Roman Period. References: Sarmento, M. 1933: 272; Guimarães, F. J. S. 1978: 410; Alarcão 1988: 1/250.

3091302, QUINTA DO BARRAL, Garfe, Póvoa de Lanhoso, Braga (30913). CMP 71 [563840.49, 4598839.3, 190]. Typology: Villa? Chronology: Roman Period.

3083601, PINHEIRO/TABUADELO, Tabuadelo, Guimarães, Braga (30836). CMP 85 [559394.85, 4583599.31, 340]. Typology: Roman necropolis. Chronology: Roman Period. References: Sarmento, M. 1933: 216; Alarcão, J. 1988: 1/307.

3091402, SANTO TIRSO, Geraz do Minho, Póvoa de Lanhoso, Braga (30914). CMP 57 [560111.53, 4606799.37, 461]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: P. M. H. Script: Nº 355; Martins, M. 1990: 91.

3087001, TAGILDE, Tagilde, Guimarães, Braga (30870). CMP 99 [560500.00, 4581200.00, 200]. Typology: Votive epigraphy. Chronology: Roman Period. References: Encarnação, J. 1975: 224-226; Cardozo, M. 1972: 36; Alarcão, J. 1988: 1/327.

3091401, INFESTA, BURGO, TORRE (Chã de S. Pedro), Geraz do Minho, Póvoa de Lanhoso, Braga (30914). CMP 57 [559274.96, 4605491.35, 100]. Typology: Roman habitat. Chronology: Roman Period. References: Cunha, A. R. 1945: 255-25; Martins, M. 1990: 91.

3087201, CAPELA DA SENHORA DO BARREIRO, Vermil, Guimarães, Braga (30872). CMP 84 [551600.00, 4588500.00, 170]. Typology: Roman habitat. Chronology: Roman Period. References: Sarmento, M. 1901: 21; Alarcão, J. 1988: 1/300.

3091502, CALÇADA (Serra do Carvalho), Lanhoso, Póvoa de Lanhoso, Braga (30915). CMP 57 [558405.08, 4604359.91]. Typology: Roman road. Chronology: Roman Period. References: Capella, M. 1895: 56; Barradas, A. L. 1956: 189.

3087202, S. MIGUEL-O-ANJO (V/O), Vermil/Oleiros, Guimarães, Braga (30872030833). CMP 84 [551456.43, 4590087.25, 385]. Typology: Romanized 166

Catalogue

3091501, VIA COVA (Cimo de Vila), Lanhoso, Póvoa de Lanhoso, Braga (30915). CMP 57 [559277.09, 4603383.96, 250]. Typology: Villa. Chronology: Roman.

3092303, CAMPO DE MARTIM, JOANE (Penedo Mogo), S. João de Rei, Póvoa de Lanhoso, Braga (30923). CMP 57 [559292.14, 4608750.52, 150]. Typology: Roman habitat. Chronology: Roman Period. References: Martins, M. 1990: 93.

3091901, CASTRO DE LANHOSO (Monte do Castelo; Monte da Senhora do Pilar; Castelo da Póvoa), Lanhoso, Póvoa de Lanhoso, Braga (30919). CMP 57 [559988.75, 4604098.96, 330]. Typology: Romanized Hill-fort; Medieval castle; Watchtower. Chronology: Chalcolithic; Bronze Age; Iron Age; Roman Period. References: Azevedo, P.A. 1902: 27-32-74-79-123128-190-192-237-240-267¬272-319-320; Azevedo, P.A. 1898: 317; Teixeira, C. 1939; Carlos, T. 1939: 117-125; A. Cardoso, M. 1941: 130; Cardoso, M. 1946: 253; Garcia e Bellido, A. 1946: 356-358; Sousa, J. M. C. 1948: 114; Santos, J., J. R. 1965: 150; Silva, J. 1966: 522; Cunha, A. R. 1975: 517-527; Almeida, C. A. F. 1978: 38; Martins, M. 1990: 91-92.

3092302, LUGAR DAS ANTAS, S. João de Rei, Póvoa de Lanhoso, Braga (30923). CMP 57 [558996.39, 4608337.67, 150]. Typology: Roman habitat. Chronology: Roman Period. 3092301, MONTE VERMELHO (Monte da Estadela; Gondim), S. João de Rei, Póvoa de Lanhoso, Braga (30923). CMP 57 [560894.03, 4609506.42, 200]. Typology: Roman habitat. Chronology: Roman Period. References: Bellino, A. 1909: 6; Sampaio, A. 1923: 107, 113; Martins, M. 1990: 93; Celestino 1994:16. 3092305, MONTE CASTRO (Castro de S. João do Rei), S. João do Rei, Póvoa de Lanhoso, Braga (30923). CMP 57 [558950.94, 4607887.35, 202]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Azevedo, P. A. 1899-1900; Bellino, A. 1909: 6; Pinto, R. S. 1929: 425; Cardoso, M. 1938: 82-83; Cardoso, M. 1938a: 82-83; Cortez, R. 1946a: 20; Kalb, Ph. 1980a: 27; Martins, M. 1990: 92; Bettencourt, A.M.S. 1993; Bettencourt, A.M.S. 1999: 291-293.

3091601, SANTA IRIA (Santa Iria), Louredo Campo, Póvoa de Lanhoso, Braga (30916). CMP 71 [558839.50, 4599088.95, 308]. Typology: Hill-fort. Chronology: Iron Age. References: Sarmento, F. M. 1970:11-72; Vasconcelos, J. L. 1901:33; Cardoso, M. 1950: 429; Sarmento, F. M. 1933: 166; Cardoso, M. 1955: 438. 3091702, MONTE DA CHAMADOURA, Monsul, Póvoa de Lanhoso, Braga (30917). CMP 57 [558254.24, 4607580.43, 128]. Typology: Roman habitat. Chronology: Roman Period.

3092802, MONTE DE SANTA IRIA, Verim, Póvoa de Lanhoso, Braga (30928). CMP 57 [557839.05, 4609997.82, 110]. Typology: Hill-fort. Chronology: Iron Age. References: Martins, M. 1990: 93.

3091701, SANTA LUZIA (Souto, Pardelhas, Devesa), Monsul, Póvoa de Lanhoso, Braga (30917). CMP 75 [558123.29, 4607679.08, 100]. Typology: Villa? Chronology: Roman Period. References: Martins, M. 1990: 91.

3092801, SAROLA (Casa das Arcelas), Verim, Póvoa de Lanhoso, Braga (30928). CMP 57 [559142.80, 4608689.03, 100]. Typology: Roman treasure. Roman habitat. Chronology: Roman Period. References: Hipólito, M. C. 1960-1961: 18; Martins, M. 1990: 93.

3091801, QUINTA DE ARREBALDE (Leira das Pedras), Moure, Póvoa de Lanhoso, Braga (30918). CMP 56 [555762.82, 4606805.59, 110]. Typology: Roman habitat. Chronology: Roman Period. References: Martins, M. 1990: 91.

3100103, CAMPO DO CASTELO, Balança, Terras de Bouro, Braga (31001). CMP 43 [557068.50, 4616125.82, 530]. Typology: Mutatio. Chronology: Roman Period. 3100102, TEIXUGOS (Pala; Devesa), Balança, Terras de Bouro, Braga (31001). CMP 43 [557577.68, 4616190.94, 460]. Typology: Milestone. Chronology: Roman Period. References: CIL II 4800; Silva, D. M. 1958: 73; Ferreira, M. 1982: 22; Alarcão, J. 1988: 139; Martins, M. 1990: 94.

3092001, VALE DA MOURA, Oliveira, Póvoa de Lanhoso, Braga (30920). CMP 57 [564884.76, 4604348.16]. Typology: Chronology: Roman Period. 3092101, SOBREDELO (Monte de Baixo, Monte de Cima, Água Levada, Cuvela), Rendufinho, Póvoa de Lanhoso, Braga (30921). CMP 75 [561335.87, 4609310.93]. Typology: Roman habitat; Roman treasure. Chronology: Roman Period. References: Martins, M. 1990: 92.

3100101, LAMPAÇAS/ CANTOS DA GEIRA (Bico da Geira; Couto de Lampaças), Balança, Terras de Bouro, Braga (0). CMP 43 [, 430]. Typology: Milestone. Chronology: Roman Period. References: Capela, M. 1895: 60, 247; Ferreira, M. 1982: 21-22; Alarcão, J. 1988: 136. Martins, M. 1990: 94.

3092304, CASA DO RIBEIRO, S. João de Rei, Póvoa de Lanhoso, Braga (30923). CMP 57 [558798.48, 4608135.74, 160]. Typology: Votive epigraphy. Chronology: Roman Period. References: Martins, M. 1990: 93; Cardoso, M. 1938a: 82-84.

3100308, FONTE DE MONÇÃO, Campo do Gerês, Terras de Bouro, Braga (310). Typology: Milestone. 167

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Chronology: Roman Period. References: CIL II 4814; Alarcão, J. 1988: 1/***.

3100604, LAGEDOS, Chorense, Terras de Bouro, Braga (31006). CMP 43 [559279.04, 4617987.54, 450]. Typology: Milestone. Chronology: Roman Period. References: Capella, M. 1895: 106, 146; Alarcão, J. 1988: 137.

3100306, LEIRAS DOS PADRÕES, Campo do Gerês, Terras de Bouro, Braga (31003). Typology: Milestone. Chronology: Roman Period. References: CIL II 4818; Alarcão, J. 1988: 1/ **.

3100603, CABANINHAS (Cabreira), Chorense, Terras de Bouro, Braga (31006). CMP 43 [559003.77, 4616535.21, 450]. Typology: Milestone. Chronology: Roman Period. References: Capella, M. 1895: 144, 184, 198; Alarcão, J. 1988: 1/135; CIL II 4801; EE VIII 226; EE VIII 227.

3100304, CRUZEIRO, Campo do Gerês, Terras de Bouro, Braga (31003). Typology: Milestone. Chronology: Roman Period. References: CIL II 4813; Capella, M. 1895: 255; Ferreira, J.M. 1982; Alarcão, J. 1988: 89.

3100602, CHÃS DE VILAR-SALANIANA (Saim Velho), Chorense, Terras de Bouro, Braga (31006). CMP 43 [558685.83, 4617331.76, 460]. Typology: Mansio. Chronology: Roman Period. References: Capella, M. 1895: 60; Alarcão, J. 1988: 134.

3100303, VOLTA DO GAVIÃO, Campo do Gerês, Terras de Bouro, Braga (31003). Typology: Milestone. Chronology: Roman Period. References: CIL II 4816, 4817, 4819; Capella, M. 1895: 174, 215; Alarcão, J. 1988; Baptista, A.M., Encarnação, J., Lemos, F.S. 1995.

3100601, VALFOIOS (Vale do Fojo), Chorense, Terras de Bouro, Braga (31006). CMP 43 [558685.83, 4617331.76, 450]. Typology: Milestone. Chronology: Roman Period. References: CIL II 4802; Alarcão, J. 1988:133.

3100302, S. JOÃO DO CAMPO, Campo do Gerês, Terras de Bouro, Braga (31003.Typology: vicus. Chronology: Roman Period. References: Vieira, J. A. 1886: 471; Encarnação, J. 1975: 253-255; Tranoy, A. 1981: 277; Alarcão, J. 1988; Baptista, A.M., Encarnação, J., Lemos, F.S. 1995.

3100803, GEIRINHA, Covide, Terras de Bouro, Braga (31008). CMP 42 [565831.71, 4621478.75, 0]. Typology: Milestone. Chronology: Roman Period. Inédito.

3100301, VEIGA DE S. JOÃO, Campo do Gerês, Terras de Bouro, Braga (31003). Typology: Milestone. Chronology: Roman Period. References: CIL II 4815; Alarcão, J. 1988: 85.

3100804, COVIDE, RUA DA CARREIRA, Covide, Terras de Bouro, Braga (31008). [565315.32, 4621147.49, 520]. Typology: Mutatio?;Mansio?; Roman epigraphy. Chronology: Roman Period. References: CIL II 4810; 4812; Capella, M. 1895: 186; Silva, D.M. 1958: 58; Alarcão, J. 1988: 91, 93.

3100307, VOLTA DO COVO, Campo do Gerês, Terras de Bouro, Braga (31003). [570067.04, 4627749.53, 0]. Typology: Milestone. Chronology: Roman Period. References: Capella, M. 1895: 62, 130, 250; Alarcão, J. 1988; Baptista, A.M., Encarnação, J., Lemos, F.S. 1995.

3100802, SÁ, Covide, Terras de Bouro, Braga (31008). [564874.49, 4621243.02, 520]. Typology: Milestone. Chronology: Roman Period. References: CIL. II 4809; Capella, M. 1895: 61-185; Alarcão, J. 1988: 90.

3100309, BOUÇA DA MÓ, Campo do Gerês, Terras de Bouro, Braga (31003). [568374.25, 4626160.06, 0]. Typology: Mutatio. Chronology: Roman Period (I-II A.D.) References: Baptista, A.M., Encarnação, J., Lemos, F.S. 1995; Azevedo, P. A. 1899-1900; Cardoso, M. 1938: 82-83.

3100805, CALCEDÓNIA (Fraga da Cidade), Covide, Terras de Bouro, Braga (31008). [566465.00, 4619633.00, 0]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Argote, J.C. 1734; Leal, P. 1842; Sousa, T. 1909; Sousa, T. 1927; Correia, A.A.M., Teixeira, C. 1946: 212-222; Leal, P. 1947; Silva, D.M. 1958; Oliveira, A.L. 1971; Moita, I. 1974; Medeiros, A.C., Teixeira, C., Lopes, J.T. 1975; Capella, M. 1895.

3100503, HERVOSA / SANTA COMBA, Chamoim, Terras de Bouro, Braga (31005). [561430.51, 4619758.71, 470]. Typology: Milestone. Chronology: Roman Period. References: Alarcão, J. 1988: 94, 95; Capella, M. 1895: 60-61, 128-129.

3100801, BARZES, Covide, Terras de Bouro, Braga (31008). CMP 43 Typology: Small Farm. Chronology: Roman Period. References: Silva, D. M. 1958: 58; Alarcão, J. 1988: 1/88.

3100501, ESPORÕES, Chamoim, Terras de Bouro, Braga (31005). [562521.49, 4620619.45, 430]. Typology: Milestone. Chronology: Roman Period. References: Capella, M. 1895: 61-260; E.E. VIII, 230; Alarcão, J. 1988: 92.

3100807, CASTELO DE COVIDE, Covide; Freitas, Terras de Bouro, Braga (31008). CMP 43 [564521.95, 4619430.20, 0]. Typology: Medieval castle? References: Correia, A.A.M., Teixeira, C. 1946: 219; Silva, D.M. 1958: 114; Cunha, A.R. 1975: 513-514.

3100605, S. SEBASTIÃO DA GEIRA, Chorense, Terras de Bouro, Braga (31006). CMP 43 [558463.59, 4616569.75, 450]. Typology: Mutatio. Chronology: Roman Period. 168

Catalogue

3101101, CASTRO DE SANTA ISABEL DO MONTE (Monte de Stª Isabel; Monte Crasto), Terras de Bouro, Braga (31011). CMP 43 [563460.00, 4618768.00, 0]. Typology: Medieval castle? Hill-fort?

3101701, BICO DA GEIRA, Vilar da Veiga ou Campo de Gerês, Terras de Bouro, Braga (31017). [569265.85, 4627063.14, 0]. Typology: Milestone. Chronology: Roman Period. References: Capella, M. 1895: 129, 187, 199; Alarcão, J. 1988: 84; Baptista, A.M., Encarnação, J., Lemos, F.S. 1995; CIL II 4821, 4823, 4822, 4807.

3101401, SANTA CRUZ, Souto, Terras de Bouro, Braga (31014). CMP 42,43 [556280.01, 4615008.22, 420]. Typology: Milestone. Chronology: Roman Period. References: Capella, M. 1895: 146, 198, 60; Martins, M. 1990: 94-95; Alarcão, J. 1988: 140; Silva, D.M. 1958: 136; Cunha, A. 1961: 320; Almeida, C.A.F. 1968: 31.

3112001, CRASTO -VILA SECA, Mosteiro / Pinheiro/ Cantelães, Vieira do Minho, Braga (31120). CMP 58 [573283.00, 4609444.00, 550]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Cunha, A.R. 1975: 485-535; Silva 1986: 79; Queiroga 1992: 158.

3101601, TRAVASSOS, Vilar, Terras de Bouro, Braga (31016). CMP 43 [560637.39, 4619100.09, 460]. Typology: Milestone. Chronology: Roman Period. References: Capella, M. 1895: 160; Silva, D.M. 1958: 170; Almeida, C.A.F 1968: 32; Alarcão1988: 153.

3110501, IGREJA VELHA, Cantelães, Vieira do Minho, Braga (31105). CMP 58. Typology: Villa . Chronology. Roman Period (High empire). References: Carvalho e Cruz 2013 a; Carvalho e Cruz 2013 b.

3101602, PODRIQUEIRAS, Terras de Bouro, Braga (31016). [559690.80, 4618791.45, 430]. Typology: Milestone. Chronology: Roman Period.

3110101, CASTRO Anissó, Vieira do Minho, Braga (31101). CMP 57 [569494.00, 4606954.00, 733]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Silva 1986: 79; Queiroga 1992: 158.

3101707, PONTE DO FORNO, Vilar da Veiga, Terras de Bouro, Braga (31017). [571520.39, 4627825.62, 0]. Typology: Roman Bridge. Chronology: Roman Period. References: Baptista, A.M., Encarnação, J., Lemos, F.S. 1995.

3110601, CASTRO DA SRª DA CONCEIÇÃO, Cova (S. João), Vieira do Minho, Braga (31106). CMP 43 [569421.00, 4614538.00, 337]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Fontes, L., Roriz, A. 2007.

3101706, PONTE DA MACEIRA, Vilar da Veiga, Terras de Bouro, Braga (31017). [571334.84, 4627916.03, 0]. Typology: Roman Bridge. Chronology: Roman Period. References: Baptista, A.M., Encarnação, J., Lemos, F.S. 1995.

3111303, MONTE DO CASTELO (Castelo de Calvos), Rossas, Vieira do Minho, Braga (31113). CMP 58 [578296.59, 4603313.43, 730]. Typology: Romanized Hill-fort; Medieval castle. Chronology: Iron Age; Roman Period. References: Cunha, A.R. 1950: 2,4; Almeida, C.A.F. 1970: 77-82; Cunha, A.R. 1975: 485535; Silva 1986: 80; Queiroga 1992: 158; Fontes, L., Roriz, A. 2007.

3101705, GERÊS, Vilar da Veiga, Terras de Bouro, Braga (31017). [569139.04, 4621078.12, 0]. Typology: Roman treasure. Chronology: Roman Period. References: Sousa, T. 1927: 171; Alarcão 1988: 100 3101704, ALBERGARIA, Vilar da Veiga, Terras de Bouro, Braga (31017). [571813.90, 4628476.31, 0]. Typology: Milestones. Chronology: Roman Period. References: CIL II 4829, 4831 , 4830 , 4832; Capella, M. 1895: 177, 187, 193, 199, 205, 216; Alarcão. 1988: 98; Baptista, A.M., Encarnação, J., Lemos, F.S. 1995.

3111302, S. PEDRO, Rossas, Vieira do Minho, Braga (31113). CMP 58 [575544.00, 4603081.00, 400]. Typology: Roman necropolis. Chronology: IV d.C., Roman Period. References: Teixeira, C. 1943-1944: 165167; Alarcão 1988: 219; Fontes, L., Roriz, A. 2007. 3111301, POMBEIRO, Rossas, Vieira do Minho, Braga (31113). CMP 58 [576390.00, 4603667.00, 460]. Typology: Small farm. Chronology: Roman Period. References: Teixeira, C. 1943-1944: 167; Alarcão 1988: 218; Fontes, L., Roriz, A.: 2007.

3101703, PONTE DE S. MIGUEL, Vilar da Veiga, Terras de Bouro, Braga (31017). [571739.40, 4628999.92, 0]. Typology: Roman Bridge. Chronology: Roman Period. References: Silva, D. M. 1958: 84; Alarcão, J. 1988: 1/97; Baptista, A.M., Encarnação, J., Lemos, F.S. 1995.

3111403, S. CRISTOVÃO, Ruivães, Vieira do Minho, Braga (31114). CMP 44 [579918.00, 4615215.00, 676]. Typology: Mansio; vicus. Chronology: Roman Period. References: Fontes, L., Roriz, A. 2007.

3101702, PORTELA DO HOMEM, Vilar da Veiga, Terras de Bouro, Braga (31017). [571226.83, 4629488.03, 0]. Typology: Milestones. Chronology: Roman Period. References: CIL II 4834, 4835, 4836, 4837,4838,4839, 4840, 4841, 4842, 4843, 4845, 4846, 4847, 4848, 6222; E.E. VIII 237; Alarcão 1988: II. 1; Baptista, A.M., Encarnação, J., Lemos, F.S. 1995.

3111402, PORTELA DE REBORDELOS, Ruivães, Vieira do Minho, Braga (31114). CMP 44 Typology: Milestones. Chronology: Roman Period. References: Barradas, L. A. 1956: A6; A7; CIL. II 4770; Alarcão, J. 1988. 169

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

3111401, BOTICA, Ruivães, Vieira do Minho, Braga (31114). CMP 44 [581025.00, 4615053.00, 660]. Typology: Milestones. Chronology: Roman Period. References: CIL., II, 4783; Barradas, L. A. 1956: 196, A19; Alarcão, J. 1988: 156.

3120801, CASTRO DO FACHO, Calendário, Vila Nova de Famalicão, Braga (31208). CMP 83 [537963.45, 4583835.74, 268]. Typology: Romanized Hill-fort. Chronology: Bronze Age; Iron Age; Roman Period. References: Teixeira, C., Medeiros, A. C. 1965: 46; Queiroga, F. 1983: 49; Queiroga, F. 1992: 157; Dinis, A. P. 1993: 53-55.

3111501, OUTEIRO DA COROA, Salamonde, Vieira do Minho, Braga. CMP 44 [450]. Typology: Hill-fort. Chronology: Roman Period. References: Fontes, L., Roriz, A. 2007: 134.

3121001, BOUÇA ALTA, Castelões, Vila Nova de Famalicão, Braga (31210). CMP 84 [547950.00, 4584600.00, 0]. Typology: Villa. Chronology: Roman Period. References: Informação Arqueológica 1979: 19; Alarcão, J. 1988: 292.

3111801, CASTRO DE ATAFONA, Tabuaças, Vieira do Minho, Braga. CMP 44 [521]. Typology: Hill-fort. Chronology: Iron Age; Roman Period. References: Fontes, L., Roriz, A. 2007: 148.

3121301, PERRELOS, Delães, Vila Nova de Fmalicão, Braga. CMP 84. Typology: vicus? Chronology: Roman Period. References: Silva, A. C. F., Dinis, A.P., Oliveira, F., Queiroga, F. 2005: 46, 47.

3120203, SANTIAGO DE ANTAS, Antas, Vila Nova de Famalicão, Braga (312020). CMP 84 [540900.00, 4583600.00, 120]. Typology: Milestone. Chronology: Roman Period. References: Mantas, V. 1996: 392-395.

3121001, CASTELÕES, Covelo, Vila Nova de Famalicão, Braga. CMP 83. Typology: Villa. Chronology: Roman Period. References: Silva, A. C. F., Dinis, A.P., Oliveira, F., Queiroga, F. 2005: 45, 46.

3120202, DEVESA ALTA / QUINTA DE PEREIRA, Antas, Vila Nova de Famalicão, Braga (31202). CMP 84 [540900.00, 4583600.00, 120]. Typology: Milestone. References: Mantas, V. 1996: 385; Capella, M. 1895: 251; EE VIII 207.

3121702, CASTRO DE PENICES, Gondifelos, Vila Nova de Famalicão, Braga (31217). CMP 83 [532826.39, 4585044.22, 99]. Typology: Romanized Hill-fort; Roman treasure. Chronology: Bronze Age; Ferro, Roman Period. References: Queiroga, F. 1985a : 3-22; Figueiral, I. 1990; Queiroga, F. 1992; Dinis, A. P. 1993: 50-52.

3120201,PORTELA DE BAIXO, Antas, Vila Nova de Famalicão, Braga (31202). CMP 84 [540950.00, 4582800.00, 110]. Typology: Milestone. Chronology: Roman Period. References: Capella, M. 1895: 149; Alarcão, J. 1988: 294.

3121701, PENICES, Gondifelos, Vila Nova de Famalicão, Braga (31217). CMP 83 [532700.00, 4584900.00, 50]. Typology: Tesouro. Chronology: Roman Period. References: Hipólito, M. C. 1960-1961: 21-22; Alarcão, J. 1988: 1/289.

3123801, SANTA MARIA DE ARNOSO, Arnoso (Santa Maria), Vila Nova de Famalicão, Braga (31238). CMP 84 [542100.00, 4591050.00, 160]. Typology: Votive epigraphy. Chronology: Roman Period. References: Santos, L. A., Le Roux, P. e Tranoy, A. 1983: 190-191; Alarcão, J. 1988: 239.

3121801, CASTRO DAS ERMIDAS, Jesufrei, Vila Nova de Famalicão, Braga (31218). CMP 83 [539902.38, 4590137.58, 209]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Brandão, A. M. 1889; 1890; 1890; 1891, 1889-1892; Queiroga, F. 1983: 49; Informacão Arqueológica 5 1985b: 56-57; Informacão Arqueológica 6 1986b: 35; Silva, A. C. F. 1986: 80; Pautreau, J. P., Queiroga, F. 1990: 44-49; Figueiral, I. 1990; Queiroga, F. 1992: 157; Dinis, A. P. 1993: 58-59.

3120702, QUINTA DE SANTA CATARINA, Cabeçudos, Vila Nova de Famalicão, Braga (31207). CMP 98 [540950.00, 4579300.00, 100]. Typology: Milestone. Chronology: Roman Period. References: Capella, M. 1895: 58, 150; E.E. VIII, 206; Alarcão, J. 1988: 1/355. 3120701, CABEÇUDOS, Cabeçudos, Vila Nova de Famalicão, Braga (0). CMP 98 [541600.00, 4580860.00, 90]. Typology: Milestone. Chronology: Roman Period. References: Capella, M. 1895: 58; Alarcão, J. 1988: 1/296.

3121901, JOANE, Joane, Vila Nova de Famalicão, Braga (31219). CMP 84 [548950.00, 4587800.00, 150]. Typology: Votive epigraphy. Chronology: Roman Period. References: Le Roux, P. e Tranoy, A. 1979: 59; Alarcão, J. 1988: 1/291.

3120802, CASTRO DE S. MIGUEL-O-ANJO (CALE), Calendário, Vila Nova de Famalicão, Braga (31208). CMP 83 [538922.24, 4582576.39, 194]. Typology: Romanized Hill-fort; Medieval castle; Roman treasure. Chronology: Iron Age; Roman Period; Medieval. References: Brandão, A. M. 1889; 1890; 1890; 1891; 1889-1892; Lima, A. P. 1940: 211-212; Costa A. J. 1959: 44; Queiroga, F. 1983: 49; Centeno, R. M. S. 1987: 112; Queiroga, F. 1992; Dinis, A. P. 1993: 55-57.

3122401, LOUSADO, Lousado, Vila Nova de Famalicão, Braga (31224). CMP 97 [538900.00, 4577800.00, 45]. Typology: Milestone. Chronology: Roman Period. References: Santos, L. A., Le Roux, P. e Tranoy, A. 1983: 201; Alarcão, J. 1988: 1/344.

170

Catalogue

3123901, CASTRO DE SANTA TECLA, Oliveira (Stª Maria), Vila Nova de Famalicão, Braga (31239). CMP 84 [549958.27, 4585296.39, 328]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Sarmento, F. M. 1901: 14-15; Queiroga, F. 1983: 49; Silva, A. C. F. 1986: 80; Queiroga, F. 1992: 157; Dinis, A. P. 1993: 72-74.

Roman Period. References: CIL II 4867; Almeida, C.A. B. 1979: 105-106; Alarcão, J. 1988: 290. 3124101, CASTRO DO CRUITO, Vale (S. Martinho), Vila Nova de Famalicão, Braga (31241). CMP 84 [544306.85, 4586924.98, 210]. Typology: Hill-fort. Chronology: Iron Age. References: Queiroga 1992: 158; Dinis 1993: 61-62.

3123001, PONTE DE SERVES, Pedome, Vila Nova de Famalicão, Braga (31230). CMP 84 [552100.00, 4585800.00, 100]. Typology: Epigrafia romana indeterminada. Chronology: Roman Period. References: Boletim de Trabalhos Históricos 1954: 102; Alarcão, J. 1988: 1/306.

3124701, CASTRO DE VERMOIM, Vermoim, Vila Nova de Famalicão, Braga (31247). CMP 84 [546117.39, 4586905.45, 356]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period; Middle Ages? References: Costa, A. C. 1706-1712, 18681869: 326; Sarmento, F. M. 1882: 81; Cardozo, M. 1932: 23-25; Sarmento, F. M. 1933: 165; Cardozo, M. 1972; Rosário, A. 1973: 59; Almeida, C. A. F. 1974b: 163; Cunha, A. R. 1975: 519-520; Queiroga, F. 1983: 48-50; Queiroga, F. 1985a: 56; Silva, A. C. F. 1986: 80; Queiroga, F. 1986a: 35; Calo Lourido, F. 1991: 614-619; Queiroga, F. 1992: 158; Dinis, A. P.1993: 6466.

3123101, CARREIRAS, Portela, Vila Nova de Famalicão, Braga (31231). CMP 84 [546650.00, 4590610.00, 0]. Typology: Milestone. Chronology: Roman Period. References: Santos, L. A., Le Roux, P. e Tranoy, A. 1983: 199; Alarcão, J. 1988: 240. 3123201, CASTRO DAS EIRAS (Alto das Eiras), Pousada de Saramagos/Joane/Ver, Vila Nova de Famalicão, Braga (31232). CMP 84 [546600.42, 4587587.04, 384]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Sarmento, F. M. 1933: 166; Sarmento, F. M. 1970, 1986, 1987/1988, 1989: 20-22; Queiroga, F. 1983: 4849; Calo Lourido, F. 1991: 649; Queiroga, F. 1992: 157; Dinis, A. P. 1993: 66-68.

3124801, VILA NOVA DE FAMALICÃO, Vila Nova de Famalicão, Vila Nova de Famalicão, Braga (31248). CMP 83 [540010.00, 4584600.00, 100]. Typology: Milestones. Chronology: Roman Period. References: CIL II 4737; CIL II 4738; CIL II 4739; Alarcão 1988: 293; Mantas, V. 1996: 395-402. 3130201, ARCOZELO, Arcozelo, Vila Verde, Braga (31302). CMP 42 Typology: Milestone. Chronology: Roman Period. References: Santos, L.A. 1979: 3-52; Araújo, J. R. 1982: 158-159; Santos, L. A., Le Roux, P., Tranoy, A. 1983: 198-199, 203; Alarcão, J. 1988: 1/127.

3123301, CASTRO DE SANTA CRISTINA, Requião, Vila Nova de Famalicão, Braga (31233). CMP 84 [544936.76, 4586404.58, 336]. Typology: Hill-fort. Chronology: Iron Age; Middle Ages? References: Queiroga 1992: 158. 3123601, CASTRO DE S. MIGUEL-O-ANJO (R/D), Ruivães/Delães, Vila Nova de Famalicão, Braga (31236031213). CMP 84 [548067.34, 4583371.47, 291]. Typology: Romanized Hill-fort; Votive epigraphy. Chronology: Iron Age; Roman Period. References: Costa, A. C. 1868-1869, 1706-1712: 329; CIL II 5561; Leal, A. P., Ferreira, A. 1873-1890: 466; Vasconcelos, J. L. I 1897, II 1905; III 1913: 327-329; Sarmento, F. M. 1933: 165, 169, 301, 420; Lima, A. C. P. 1926-1928: 289-299; Encarnação, J. 1970: 220-222; Encarnação, J. 1975: 148-151; Tranoy, A. 1981: 72, 280; Queiroga, F. 1983: 49; Silva, A. C. F. 1986: 80; Alarcão 1988: 295; Calo Lourido, F. 1991: 456-460; Queiroga, F. 1992: 157.

3130401, S. SEBASTIÃO, Atiães, Vila Verde, Braga (31304). CMP 56 [542483.87, 4608718.06, 100]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Martins, M. 1990: 95-96. 3130602, CASTELHÃO, Barbudo, Vila Verde, Braga (31306). CMP 42 [545013.75, 4611201.34, 200]. Typology: Hill-fort. Chronology: Iron Age. References: Cunha, A.R. 1975: 530; Martins, M. 1990: 96. 3130601, CASTRO DO BARBUDO (Monte do Castelo; Monte de Brito), Barbudo, Vila Verde, Braga (31306). CMP 42 [544124.12, 4611571.96, 331]. Typology: Romanized Hill-fort; Roman treasure. Chronology: Iron Age; Roman Period. References: Azevedo, P. A. 1897: 152; Oliveira, M. 1908: 667-668; Hipólito, M. C. 1960-1961: 16-17; Cunha, A. R. 1975a: 197; Alarcão, J. 1988: 131; Martins, M. 1990: 96.

3124002, CASTRO DA BOCA, Vale (S. Cosme), Vila Nova de Famalicão, Braga (31240). CMP 84 [543988.80, 4589606.17, 289]. Typology: Hill-fort. Chronology: Iron Age. References: Queiroga 1992: 158; Dinis 1993: 60-61.

3130802, LUGAR DE GAIÃO, Cabanelas, Vila Verde, Braga (31308). CMP 56 [542380.08, 4602642.25, 60]. Typology: Hill-fort. Chronology: Iron Age; Roman Period. References: Martins, M. 1990: 96-97.

3124001, COSMADE, Vale (S. Cosme), Vila Nova de Famalicão, Braga (31240). CMP 84 [544050.00, 4589160.00, 190]. Typology: Milestone. Chronology: 171

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

3130801, CABANELAS (Veiga de Cabanelas), Cabanelas, Vila Verde, Braga (31308). CMP 56 [542596.63, 4603874.02, 40]. Typology: Roman epigraphy- undetermined; Roman necropolis. Chronology: Roman Period. References: Santos, L. A., Le Roux, P., Tranoy, A. 1983: 203-204; Alarcão, J. 1988: 1/187; Martins, M. 1990: 97.

imo). References: Hipólito, M. C. 1960-1961: 17; Alarcão, J. 1988: 1/128; Martins, M. 1990: 98. 3132601, MARRANCOS, Marrancos ou Gondiães, Vila Verde, Braga (31326). CMP 42 Typology: Milestone. Chronology: Roman Period. References: Santos, L. A. 1979: 20-2; Araújo, J. R. 1982: 158-159; Alarcão, J. 1988: 1/129; Martins, M. 1990: 98

3131101, S. JULIÃO, Coucieiro, Vila Verde, Braga (31311). CMP 42 [550467.05, 4615339.62, 296]. Typology: Romanized Hill-fort; Iron Age warrior. Chronology: Bronze Age; Iron Age; Roman Period. References: Costa, A. C. 1868-1869: 211; Bellino, A. 1909: 6; Fontes, J. 1916: 198-210; Freitas, J. 1971: 133-138; Martins, M. e Silva, A. C. 1984: 29-47; Martins, M. 1986a: 159-160; Alarcão, J. 1988: 138; Martins, M. 1990: 97; Bettencourt, A. 2000b.

3132801, MOURE, Moure, Vila Verde, Braga (31328). CMP 56 [543834.18, 4610034.04, 80]. Typology: Roman treasure. Chronology: Roman Period. References: Hipólito, M. C. 1960-1961: 16; Martins, M. 1990: 98. 3132901, NEVOGILDE, Nevogilde, Vila Verde, Braga (31329). CMP 42 [543903.93, 4613034.01, 0]. Typology: Roman epigraphy. Chronology: Roman Period. References: Santos, L. A., Le Roux, P. e Tranoy, A. 1983: 203; Alarcão, J. 1988: 130; Martins, M. 1990: 99.

3131301, DOSSÃOS (Lugar de Coto; Póvoa; Lugar do Reduto), Dossãos, Vila Verde, Braga (31313). CMP 42 [544349.86, 4613818.76, 250]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Martins, M. 1990: 97-98; Vieira, J. A. 1886-1887: 404.

384/3133001, OLEIROS, Oleiros, Vila Verde, Braga (31330). CMP 56 [542716.79, 4606834.23, 0]. Typology: Milestone. Chronology: Roman Period. References: Almeida, C. A. B. 1979a: 105; Araújo, J. R. 1982: 155-157; Santos, L. A., Le Roux, P. e Tranoy, A 1983: 200-201; Alarcão, J. 1988: 1/182; Martins, M. 1990: 99.

3134501, CASTRO DE ESCARIZ (Santo Isidoro; Santos Ilus ou Idus), Escariz - S. Mamede, Vila Verde, Braga (31345). CMP 41 ou 55 [549182.37, 4610038.28, 235]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Martins, M. 1990: 101-102; Azevedo, P.A 1898: 143; Oliveira, M. 1908: 668.

3133803, PENEDOS DA PORTELA, Portela das Cabras, Vila Verde, Braga (31338). CMP 42 [542500.93, 4613653.35, 453]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Leal, P., Ferreira, A. 1873-90: 615; Regalo, H. 1986: 91; Martins, M. 1990: 99.

3131701, CASTRO DE SANTA ENGRACIA, Geme, Vila Verde, Braga (31317). CMP 42 [546900.61, 4613264.23, 217]. Typology: Hill-fort. Chronology: Iron Age. References: Martins, M. 1990: 98.

3133802, PORTELINHA, Portela das Cabras, Vila Verde, Braga (31338). CMP 42 [543323.37, 4611129.25, 0]. Typology: Roman treasure. Chronology: Roman Period. References: Martins, M. 1990: 100; Hipólito, M. C. 1960-1961: 17-18; Alarcão, J. 1988: 1/149.

3131902, SOUTELINHO, Godinhaços, Vila Verde, Braga (31319). CMP 42 [545150.97, 4618244.93, 350]. Typology: Roman habitat? Chronology: Roman Period. References: Leal, P; Ferreira, A. 1874: 282; Carvalho Da Costa 1868: 304; Vieira, J. A. 1886: 389; Costa 1959; Abreu 1963: 86.

3133801, PORTELA DE PENEDA (Portela das Cabras), Portela das Cabras, Vila Verde, Braga (31338). CMP 42 Typology: Funerary epigraphy . Chronology: Roman Period. References: Santos, L. A., Le Roux, P., Tranoy, A. 1983: 195-196; Alarcão 1988: 148; Martins, M. 1990: 99-100.

3132001, MONTE DE S. TOMÉ (Vairão), Gomide, Vila Verde, Braga (31320). [551654.23, 4617710.71, 467]. Typology: Hill-fort; Medieval castle? Chronology: Iron Age?, Middle Ages. References: Regalo, H. 1986: 89.

3132001, CASTRO DE MOEGA, Rio Mau, Vila Verde, Braga. CMP 42 [300] Typology: Hill-fort. Chronology: Iron Age. Roman Period. References: Maciel, T. 2005: 85-86.

3132104, CASTRO MAU (Monte de Santa Engrácia de Gondiães), Gondiães, Vila Verde, Braga (31321). CMP 42 [546813.40, 4613983.11, 217]. Typology: Hill-fort. Chronology: Iron Age. References: Sampaio 1963: 265; Moita, I. 1966: 533; Martins, M. 1987: 160.

3134204, BARREIRO, Prado – Santa Maria, Vila Verde, Braga (31342). CMP 56 [545496.08, 4606812.22, 0]. Typology: Roman habitat. Chronology: Roman Period. References: Martins, M. 1990: 100.

3132102, LEIRA DA CACHADA BARROCO, Gondiães, Vila Verde, Braga (31321). CMP 42 [545400.26, 4613349.07, 0]. Typology: Roman treasure. Chronology: Roman Period (Constantino I / Máx-

3134203, PONTE DO PRADO, Prado - Santa Maria, Vila Verde, Braga (31342). CMP 56 [544881.04, 4605346.51, 0]. Typology: Milestones. Chronology: 172

Catalogue

Roman Period. References: Argote, J. 1728: 536; Capela, M. 1895: 83; Cardozo, M. 1972: 77; Almeida, C. A. B. 1979a: 144; Araújo, J. R. 1982: 149-150; Alarcão, J. 1988: 184; Martins, M. 1990: 100-101. 3134202, IGREJA NOVA PRADO, Prado – Santa Maria, Vila Verde, Braga (31342). CMP 56 [544077.36, 4605738.29, 0]. Typology: Roman necropolis; vicus. Chronology: Roman Period. References: Vieira, J. A. 1886-1887: 410; Dias, A. R. 1903: 256; Abreu, L. 1963: 151; Alarcão, J.1988: 183; Martins, M. 1990: 100. 3134201, SANTIAGO DE FRANCELOS, Prado – Santa Maria, Vila Verde, Braga (31342). CMP 56 Typology: Roman epigraphy. Chronology: Roman Period. References: Alarcão, J. 1988: 182A; Martins, M. 1990: 100. 3135402, MONTE DE SANTA CRUZ (Santa Helena), Turiz, Vila Verde, Braga (31354). CMP 56 [544749.43, 4609071.61, 170]. Typology: Hill-fort. Chronology: Iron Age; Roman Period. References: Azevedo 1903: 258; Regalo, H. 1986: 92; Martins, M. 1990: 100. 3135401, IGREJA DE TURIZ, Turiz, Vila Verde, Braga (31354). CMP 56 [545426.41, 4609760.53, 170]. Typology: Roman habitat. Chronology: Roman Period. References: Costa, A. C. 1868-1869: 307; Leal 1873; Vieira, J. A. 1886-1887: 407; Alarcão, J. 1988: 1/179; Martins, M. 1990: 102. 3135701, OUTEIRINHO, Vila Verde, Vila Verde, Braga (31357). CMP 56 [547448.26, 4610353.44, 74]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Martins, M. 1990: 102.

173

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

[578300.00, 4566900.00, 240]. Typology: Mansio; vicus? Chronology: Roman Period. References: Fortes 1908: 253; Dias, L.A.T. 1995: 249.

PORTO DISTRICT 13010001, AMARANTE, Amarante, Amarante, Porto (130100). Typology: Roman habitat? Chronology: Roman Period. References: Carvalhaes, J. 1911: 109; Alarcão, J. 1988: 1/417.

13011901, PRAZOS, Lomba, Amarante, Porto (130119). [578600.00, 4567400.00, 294]. Typology: Roman necropolis. Chronology: Roman Period. References: Fortes, J. 1905-1908: 252-262; Vasconcellos 1913: 373; Abascal Palazón 1986: 259; Alarcão 1988: 1/418.

13010501, VALINHO, Bustelo, Amarante, Porto (130105). [585100.00, 4565200.00, 660]. Typology: Roman treasure. Chronology: 2ª metade do século IV. References: Hipólito, M. C. 1960-1961: 46; Alarcão 1988: 25; Dias, L.A.T. 1995: 247.

13012101, GATIÃES (SERTÃ), Lufrei, Amarante, Porto (130121). [580100.00, 4569800.00, 280]. Typology: Farm. Chronology: I-IV, Roman Period. References: Dias, L.A.T. 1995: 250.

13010901, CASTELO, Carvalho de Rei, Amarante, Porto (130109). [581400.00, 4565500.00, 526]. Typology: Hill-fort; Votive epigraphy. Chronology: Roman. References: Brandão, D. P. 1959: 909-913; Tranoy, A. 1981: 317; Alarcão 1988: 25; Dias, L.A.T. 1995: 247.

13012201, ATAÚDES, Madalena, Amarante, Porto (130122). [578500.00, 4569200.00, 203]. Typology: Roman necropolis. Chronology: Roman Period (II-III). References: Fortes 1908: 225; Dias, L.A.T. 1995: 250.

13011001, CEPELOS, Cepelos, Amarante, Porto (130110). [576500.00, 4568500.00, 80]. Typology: Villa? References: Carvalhais 1911; Dias, L.A.T. 1995: 247-248; Azevedo 1896: 252.

13010401, VILA MEÃ, Real, Amarante, Porto (130104130127). [568900.00, 4566500.00, 175]. Typology: Roman necropolis; Roman coins. Chronology: Roman Period (III-IV). References: Oleiro, J. M. 19551956: 276; Dias, L.A.T. 1995: 250; Alarcão, J. 1988: 1/414.

13011401, CAMPINHO DO MURO (CAMPELO), Freixo de Cima, Amarante, Porto (130114). [571800.00, 4574200.00, 420]. Typology: Villa? References: Dias, L.A.T. 1995: 248.

13013301, CALÇADA DA MISERICÓRDIA, S. Gonçalo, Amarante, Porto (130133). [577200.00, 4569300.00, 110]. Typology: Roman necropolis. References: Dias, L.A.T. 1995: 251.

13011601, GATÃO, Gatão, Amarante, Porto (130116). [578300.00, 4572200.00, 180]. Typology: vicus. Chronology: Roman Period. References: CIL II, 2383; Dias, L.A.T. 1995: 248.

13012902, LOUREDO DAS ALMAS, Salvador, Amarante, Porto (130129). [576500.00, 4565500.00, 300]. Typology: Roman necropolis. Chronology: Roman Period. Dias, L.A.T. 1995: 250

13010002, QUINTA DOS PASCOAIS, Gatão, Amarante, Porto (130100). [577900.00, 4571200.00, 190]. Typology: Votive epigraphy (altar dedicated by “vicani Atucausenses”) Chronology: Roman Period. References: CIL II S 6287; Cardozo, M. 1972: 50; Vives, J. 1972: 85; Tranoy, A. 1981: 316; Alarcão, J. 1988: 21; Le Roux, P. 1993: 159; Dias, L.A.T. 1995: 248.

13012901, QUINTA DO COURACEIRO, Salvador, Amarante, Porto (130129). [577000.00, 4564900.00, 224]. Typology: Roman habitat. Chronology: Roman Period. References: Dias, L.A.T. 1995: 250; Alarcão, J. 1988: 1/414.

13011703, TUBIREI, Gondar, Amarante, Porto (130117). [582900.00, 4566300.00, 418]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Dias 1995: 249.

13013001, SANCHE, Sanche, Amarante, Porto (130130). [583200.00, 4572000.00, 350]. Typology: Roman necropolis. Chronology: Roman Period. References: Dias, L.A.T. 1995: 251.

13011701, VILA LEÇA (VILELA), Gondar, Amarante, Porto (130117). [582600.00, 4566900.00, 230]. Typology: Roman necropolis? Chronology: Roman Period. References: Vasconcellos, J. L. 1895: 17; Alarcão 1988: 25; Dias, L.A.T. 1995: 249.

13013501, LABORIZ, Telões, Amarante, Porto (130135). [575700.00, 4573100.00, 200]. Typology: Roman necropolis. References: Amaral 1988-9: 111114; Dias, L.A.T. 1995: 251. 13013804, CASTRO DE COURA, Vila Caiz, Amarante, Porto (130138). [573387.06, 4564307.42, 220]. Typology: Hill-fort? References: Leal, P. 1875; Soeiro, T. 1984: 41-42; Dias, L.A.T. 1995: 251.

13011702, PANELEIROS, Gondar, Amarante, Porto (130117, 117). [581900.00, 4566500.00, 220]. Typology: Roman habitat. Chronology: Roman Period. References: Botelho, H. 1904: 99-100; Alarcão 1988: 1/25; Dias, L.A.T. 1995: 249.

13013801, VILARINHO, Vila Caiz, Amarante, Porto (130138). [573521.78, 4564802.24, 150]. Typology: Roman necropolis; Villa. Chronology: Roman Period.

13011902, LUGAR DAS PAREDINHAS E DO PARAÍSO, Lomba, Amarante, Porto (130119). 174

Catalogue

References: Fortes 1905-1908: 477; Soeiro, T. 1984: 39; Alarcão, J. 1988: 1/420; Dias, L.A.T. 1995: 252.

13020601, CRUITO, Gove, Baião, Porto (130206). [580900.00, 4554500.00, 516]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Vieira 1887: 449; Vasconcelos 1889: 177; Vasconcelos 1905-08: 670-671; Vasconcelos 1908: 672; Pereira e Gonzalez 1988: 151-158; Dias, L.A.T. 1995: 254.

13013802, VILA CAIZ, Vila Caiz, Amarante, Porto (130138). Typology: Roman treasure. Chronology: Roman Period. References: Centeno, R. M. S. 19811982: 122; Soeiro, T. 1984: 41; Alarcão, J. 1988: 24; Dias, L.A.T. 1995: 252.

13020701, CARREIRINHA, Grilo, Baião, Porto (130207). Typology: Milestone. Chronology: Roman Period. References: Figueira, J. F. 1945: 183-185.

13013803, MONTE DA SRª DA GRAÇA, Vila Caiz, Amarante, Porto (130138). [572700.00, 4566500.00, 470]. Typology: Hill-fort? References: Dias, L.A.T. 1995: 251.

13020901, LOIVOS DA RIBEIRA, Loivos da Ribeira, Baião, Porto (130209). [590000.80, 4555000.90, 250]. Typology: Roman necropolis. Chronology: Roman Period. References: Dias, L.A.T. 1995: 254.

1302, CANDORCAS, Baião, Porto (1302). Typology: Roman treasure. Chronology: Roman Period. Vieira, J. A.1887: 463-464 Alarcão 1988: 29 Dias, L.A.T. 1995: 254

13021101, GIESTA, Ovil, Baião, Porto (130211). [582700.00, 4559700.00, 550]. Typology: Roman necropolis. Chronology: Roman Period. References: Barroca, M. J. 1984: 117; Alarcão, J. 1988: 29; Dias, L.A.T. 1995: 254.

13020104, MOSTEIRÔ (QUINTA DE), Ancede, Baião, Porto (130201). [578900.00, 4549900.00, 110]. Typology: Votive epigraphy; Mutatio. Chronology: Roman Period. References: CIL II 5567; Sarmento, M.1887: 187; Sarmento, M. 1888: 11; Vasconcellos, J.L. 1889:178; Vieira, 1887: 452; Vasconcellos, J.L. 1895: 35; Sarmento, M. 1933: 310-312; Vasconcellos, J.L. 1908: 669-672; Vasconcellos, J.L. 1913: 477; Vives, J. 1972: 143; Tranoy, A. 1981: 317; Garcia, J.M. 1991: 397; Dias, L.A.T. 1995: 253.

13021102, S. JOÃO DE OVIL, Ovil, Baião, Porto (0). [530]. Typology: Funerary epigraphy . Chronology: Roman Period. References: Barroca, M. J. 1984: 117; Dias, L.A.T. 1995: 255. 13021201, PORTO MANSO, Ribadouro, Baião, Porto (130212). [577400.00, 4550100.00, 209]. Typology: Romanized Hill-fort? Mansio? References: Vieira 1887: 452; Vasconcelos 1908: 669-670; Silva 1986: 87; Dias, L.A.T. 1995: 255.

13020103, ESMORIZ (QUINTA DE), Ancêde, Baião, Porto (130201). [578600.00, 4552300.00, 295]. Typology: Roman habitat? Chronology: Roman Period. References: Dias, L.A.T. 1995: 253.

13021303, PASSAL, Santa Cruz do Douro, Baião, Porto (130213). [583700.00, 4552400.00, 230]. Typology: Farm. Chronology: Roman Period. Vasconcelos 1908: 671; Dias, L.A.T. 1995: 256.

13020102, ERMÊLO (IGREJA VELHA), Ancede, Baião, Porto (130201). [581500.00, 4556000.00, 165]. Typology: Roman necropolis. Chronology: Roman Period. References: Vasconcellos, J. 1907: 669-672; Dias, L.A.T. 1995: 253.

13021302, MANTEL, Santa Cruz do Douro, Baião, Porto (130213). [584300.00, 4553900.00, 515]. Typology: Romanized Hill-fort? Chronology: Iron Age; Roman Period References: Vieira 1887: 460; Vasconcelos 1889: 177, 671; Silva 1986: 87; Dias, L.A.T. 1995: 255.

13020101, ANCÊDE, Ancêde, Baião, Porto (130201). [579300.00, 4558000.00, 310]. Typology: Funerary epigraphy . Chronology: Roman Period. References: CIL II S 6287; Cardozo, M. 1972: 50; Vives, J. 1972: 85; Tranoy, A. 1981: 316; Alarcão, J. 1988: 21; Le Roux., P. 1993: 159; Dias, L.A.T. 1995: 253.

13021301, SANTA CRUZ DO DOURO (FIEIS DE DEUS), Santa Cruz do Douro, Baião, Porto (130213). [582400.00, 4554200.00, 709]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Vieira 1887: 449; Vasconcellos, J. 1908: 671; Alarcão, J. 1988: 29; Dias, L.A.T. 1995: 255.

13020401, FRENDE, Frende, Baião, Porto (130204). [589300.00, 4553200.00, 158]. Typology: Funerary epigraphy ; Roman necropolis. Chronology: Roman Period. References: Machado, L. S. 1920: 257; Cardozo, M. 1972: 150; Acuña Castroviejo, F. 1974: 32-34; Almeida, C. A. F. 1974: 29-33; Alarcão, J. 1988: 29; Sousa 1990: 52; Dias, L.A.T. 1995: 254.

13021401, BAIRRAL, Santa Leocádia de Baião, Baião, Porto (130214). [577100.00, 4552600.00, 375]. Typology: Votive epigraphy; Roman necropolis. Chronology: Roman Period. References: Severo, R. 1907; Brandão, D. P. 1959-1960: 76-79; Sousa, J. J. R. 1967: 181-196; Vives, J. 1972: 120; Tranoy, A. 1981: 316; Barroca, M. 1984: 116-136; Alarcão, J. 1988: 1/485; Dias, L.A.T. 1995: 256.

13020501, GESTAÇÔ, Gestaçô, Baião, Porto (130205). [589000.00, 4558800.00, 675]. Typology: Roman treasure. Chronology: Roman Period (IV). References: Hipólito, M. C. 1960-1961: 48; Alarcão, J. 1988: 29; Dias, L.A.T. 1995: 254. 175

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

13021503, BARREIRO, Santa Marinha do Zêzere, Baião, Porto (130215). [587900.00, 4555000.2, 403]. Typology: Hill-fort. Chronology: Iron Age. References: Vieira 1887: 461; Vasconcelos, J. 1889: 178; Vasconcellos, J. 1905: 286-287; Vasconcelos, J. 1908: 672; Cardozo, M. 1962:148-149; Silva 1986: 87; Dias, L.A.T. 1995: 256.

13032401 SENDIM, Sendim, Felgueiras, Porto (130324), Typology: Villa. Chronology. Roman Period (I A.D to VI A.D). References: Pinto, M.M. 2008. 13040801, MONTE DO PENOUÇO, Rio Tinto, Gondomar, Porto (130408). Typology: Roman necropolis; Roman habitat. Chronology: Roman Period. References: Severo, R. 1905-1908: 111-113, 126-127; Alarcão, J. 1988: 1/400.

13021501, SANTANA, Santa Marinha do Zêzere, Baião, Porto (130215). [586800.00, 4555500.00, 570]. Typology: Funerary epigraphy. Chronology: Roman Period. References: Brandão, D. P. 1960: 485-490; Alarcão, J. 1988: 1/490; Dias, L.A.T. 1995: 256.

13040901, MONTE CRASTO, São Cosme, Gondomar, Porto (130409). [545856.56, 4554583.66, 192]. Typology: Roman treasure; Romanized Hill-fort. Chronology: Roman Period. References: Hipólito, M. C. 1960-1961: 48; Alarcão, J. 1988: 1/451.

13021502, QUINTA DE GUIMARÃES, Santa Marinha do Zêzere, Baião, Porto (130215). [589320.00, 4554100.00, 270]. Typology: Roman necropolis; vicus. Chronology: Roman Period. References: Vasconcelos, J. 1889:179; Brandão, D. P. 1960: 486-487; Fernandes, A 1968; Alarcão, J. 1988: 29; Dias, L.A.T. 1995: 256.

13041101, QUINTA DE VILAR D'ALLEN, Valbom, Gondomar, Porto (130411). Typology: Roman treasure. Chronology: Roman Period. References: Brandão, D. P. 1967-1969: 5-8; Alarcão, J. 1988: 1/453. 13051901, SANTA EULÁLIA DE BARROSAS, Barrosas (Santa Eulália), Lousada, Porto (130519). Typology: Votive epigraphy; Roman epigraphy- undetermined; Roman habitat. Chronology: Roman Period. References: Cardozo, M. 1972: 60; Encarnação, J. 1975: 157-159; Guimarães, F. J. S. 1978: 410; Tranoy, A. 1981: 272; Alarcão, J. 1988: 1/363.

13021601, CASTRA ORESBI, Teixeira, Baião, Porto (130216). [590900.00, 4565000.2, 1150]. Typology: Roman camp? Chronology: Roman Period. References: Menedero 1983:123; Dias, L.A.T. 1995: 257. 13021901, VALADARES, Valadares, Baião, Porto (130219). Typology: Roman treasure. Chronology: Roman Period (101 b.C., 80 b.C., 55 b.C.) References: Centeno, R.M.S. 1987: 83; Dias, L.A.T. 1995: 257.

13050402, CERCA DOS VEADOS, Caíde de Rei, Lousada, Porto (130504). [564400.00, 4570200.00, 190]. Typology: Roman habitat? References: Dias, L.A.T. 1995: 257.

13031301, CAMPO DA BOUÇA NOVA, Pena Cova (S. Martinho) / Pena, Felgueiras, Porto (130313. Typology: Votive epigraphy. Chronology: Roman Period. References: CIL II 5581; Cardozo, M. 1972: 45; Alarcão, J. 1988: 1/377.

13051301, MEINEDO, Meinedo, Lousada, Porto (130513). [562200.00, 4566900.00, 220]. Typology: vicus. Chronology: Roman Period. References: Brandão, D. P. 1971: 627; Almeida, C. A. F. 1972: 117; Alarcão, J. 1988: 24; Dias, L.A.T. 1995: 258.

13031501, MONTE, Pombeiro de Riba Vizela/ Pom, Felgueiras, Porto (130315). Typology: Roman habitat? Chronology: Roman Period. References: Alarcão, J. 1988: 1/326.

13052401, EIVANDRA, Sousela, Lousada, Porto (130524). Typology: Funerary epigraphy. Chronology: Roman Period. References: Fortes, J. 1905-1908: 479480; Alarcão, J. 1988: 1/373.

13031502, QUINTA DA LOUSA, Pombeiro de Riba Vizela / Pom, Felgueiras, Porto (130315). Typology: Roman epigraphy. Chronology: Roman Period. References: Cardozo, M. 1972: 59; Alarcão, J. 1988: 1/325.

13061201, IGREJA, Avioso (S. Pedro), Maia, Porto (130612). Typology: Milestone; Votive epigraphy. Chronology: Roman Period. References: Almeida, C. A. F. 1969: 26-27; Tranoy, A. 1981: 207, 278; Alarcão, J. 1988: 1/393.

13031801, CRISTELO, Regilde, Felgueiras, Porto (130318). Typology: Votive epigraphy. Chronology: Roman Period. References: Encarnação, J. 1975: 204205; Corominas, J. 1976: 376-379; Alarcão, J.1988: 1/328.

13061101, CASTELO DA MAIA (Castro de Avioso), Avioso (Santa Maria), Maia, Porto (130611). [532455.98, 4569221.92, 108]. Typology: Hill-fort. Chronology: Roman Period. References: Alarcão 1988: 394; Almeida, C. A. F. 1969: 43; Almeida, J. A. F. 1953.

13033201, VILA FRIA, Vila Fria, Felgueiras, Porto (130332). Typology: Funerary epigraphy. Chronology: Roman Period. References: EE. VIII, 112; Alarcão, J. 1988: 1/324. 13033202, SÁ, Vila Fria, Felgueiras, Porto (130332). Typology: Funerary epigraphy; Roman sculpture. Chronology: Roman Period. References: Cardozo, M. 1972: 110-111; Alarcão, J. 1988: 1/322.

13061202, S. MAMEDE DE INFESTA, S. Mamede de Infesta, Maia, Porto (130612). Typology: Milestone. References: Mantas, V. 1996: 343.

176

Catalogue

13061301, S. PEDRO FINS, S. Pedro Fins, Maia, Porto (130613). Typology: Roman habitat. Chronology: Roman Period. References: Bellino, A. 1909: 6; Alarcão, J. 1988: 1/395.

13070302, MÓRIA, Avessadas, Marco de Canaveses, Porto (130703). [569000.00, 4557400.00, 250]. Typology: Roman necropolis. Chronology: Roman Period (IV A.D). References: Dias, L.A.T. 1995: 260.

13061501, AGRA DA PORTELA, Vermoim, Maia, Porto (130615). Typology: Roman necropolis. Chronology: Roman Period References: Almeida, C. A. F. 1969: 42; Alarcão, J. 1988: 1/396.

13070301, CASTELINHO, Avessadas, Marco de Canaveses, Porto (130703). [569700.00, 4556800.00, 344]. Typology: Roman epigraphy. Chronology: Roman Period. References: Vieira 1887: 500; Dias, L.A.T. 1995: 260.

13061601, BICAS, Vila Nova da Telha, Maia, Porto (130616). Typology: Roman necropolis. Chronology: Roman Period. References: Almeida, C. A. F. 1969: 42; Alarcão, J. 1988: 1/392.

13070701, ARCO, Folhada, Marco de Canaveses, Porto (130707). [577100.00, 4563900.00, 120]. Typology: Roman bridge. Chronology: Roman Period. References: Azevedo 1896: 191; Dias, L.A.T. 1995: 261.

13070106, FRAGA-FEIRA NOVA, Alpendurada, Marco de Canaveses, Porto (130701). [566600.00, 4552100.00, 293]. Typology: Roman necropolis. Chronology: Roman Period (IV A.D.). References: Vasconcellos 1913: 372; Alarcão 1988: 28; Dias, L.A.T. 1995: 259-260.

13070801, CARVALHEIRA, Fornos, Marco de Canaveses, Porto (130708). [572500.00, 4560100.00, 0]. Typology: Roman habitat. Chronology: Roman Period. References: Vasconcellos, J. L. 1897: 168; Fonseca 1930, 57-58; Alarcão, J. 1988: 1/424; Dias, L.A.T. 1995: 261-262.

13070105, OUTEIRO, Alpendurada, Marco de Canaveses, Porto (130701). [563800.00, 4548800.00, 150]. Typology: Roman habitat? Chronology: Roman Period. References: Lanhas, F., Brandão, D.P. 1967: 52; Dias, L.A.T. 1995: 259.

13070901, TONGOBRIGA, Freixo, Marco de Canaveses, Porto (130709). [571700.00, 4557400.00, 310]. Typology: Civitas capital. Chronology: Roman Period. References: Vasconcelos, J. L. 1900: 31; Vasconcelos, M. 1916: 323, 324; Cardozo, M. 1972: 34; Almeida, C. A. F. 1974: 149-172; Encarnação, J. 1975: 195-197; Dias, L. A. T. 1984: 86-90; Vasconcelos, M. 1884: 105-106; Alarcão 1988: 82; Dias, L.A.T. 1999: 77-107.

13070104, CASAS NOVAS, Alpendurada, Marco de Canaveses, Porto (130701). [563900.00, 4549400.00, 240]. 13070104, Typology: Roman necropolis. Chronology: Roman Period. References: Lanhas, F., Brandão, D.P. 1967: 52; Alarcão, J. 1988: 3; Dias, L.A.T. 1995: 259.

13071102, CASTILHO, Manhuncelos, Marco de Canaveses, Porto (130711). [570800.00, 4555400.00, 416]. Typology: Hill-fort. Chronology: Iron Age. References: Silva 1992: 39-45; Dias, L.A.T. 1995: 263.

13070102, MONTE DE SANTIAGO (Alto de São Tiago), Alpendurada, Marco de Canaveses, Porto (130701). [565000.00, 4550100.00, 480]. Typology: Romanized Hill-fort? Chronology: Iron Age; Roman Period? References: Costa 1706: 397; Vieira 1887: 500; Azevedo 1896: 258; Azevedo 1898: 199-200; Vasconcellos 1905: 79-80; Vasconcelos 1914-16; Aguiar 1947: 63-64; Azevedo 1941: 2-3; Lanhas, F., Brandão, D.P. 1967: 12-13; Ponte 1984: 111-114; Silva 1986: 86; Silva 1992: 23-38; Dias, L.A.T. 1995: 259.

13071101, BOI, Manhuncelos, Marco de Canaveses, Porto (130711). [569400.00, 4553700.00, 614]. Typology: Hill-fort. Chronology: Iron Age. References: Vieira 1887: 492; Azevedo 1898: 201; Sarmento 1901: 174; Vasconcelos 1901; Vasconcelos 1914-16; Silva 1986: 112; Dias, L.A.T. 1995: 262.

13070101, ALPENDURADA, Alpendurada, Marco de Canaveses, Porto (130701). [564300.00, 4548300.00, 130]. Typology: Honorific epigraphy ; Votive epigraphy. Chronology: Roman Period. References: CIL II 2376; Silva A.C. 1984: 47; Brandão, D.P. 1962: 23-51; Lanhas, F., Brandão, D.P. 1967: 34; Vives, J. 1972: 136; Tranoy, A. 1981: 317; Garcia, J.M. 1991: 396; Dias, L.A.T. 1995: 258.

13071202, URRÔ, CASA DA BABILÓNIA, Maureles, Marco de Canaveses, Porto (130712). [, 200]. Typology: Villa. Chronology: Roman Period (I-IV A.D.). References: Dias, L.A.T. 1995: 263. 13071201, MAURELES, Maureles, Marco de Canaveses, Porto (130712). [150]. Typology: Small Farm. Chronology: Roman Period. References: Dias, L.A.T. 1995: 263.

13070201, EIROZES (TAPADA DAS), Ariz, Marco de Canaveses, Porto (130702). [565900.00, 4552800.00, 200]. Typology: Roman necropolis. Chronology: Roman Period. References: Dias, L.A.T. 1995: 259; Alarcão, J. 1988: 1/480.

13072101, S. NICOLAU, S. Nicolau, Marco de Canaveses, Porto (130721). [, 100]. Typology: Votive epigraphy. Chronology: Roman Period. References: Monteiro 1946: 73-76; Alarcão, J. 1988: 1/423. 13072203, SOALHÃO, Soalhães, Marco de Canaveses, Porto (130722). [442]. Typology: Romanized Hill-fort; 177

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Roman necropolis. Chronology: Iron Age; Roman Period. References: Vasconcelos 1914-1916: 327-329; Fonseca 1930: 57-58; Silva 1986: 86; Silva 1992: 6982; Dias 1995: 265.

[190]. Typology: Villa. Chronology: Roman Period. References: Dias, L.A.T. 1995: 266. 13072701, TUIAS (Quinta de Baixo), Tuias, Marco de Canaveses, Porto (130727). Typology: Milestone; Votive epigraphy. Chronology: Roman Period. References: CIL II 2384; Sarmento, M. 1901: 181; Vasconcellos, J.L. 1905: 183; Vasconcellos, J.L. 1913: 86; Blázquez Martinez, J.M. 1962: 131; Encarnação, J. 1972: 94; Garcia 1972: 360; Vives, J. 1972: 674; Encarnação, J. 1975: 211-212; Tranoy, A. 1981: 303, 394; Dias, L.A.T. 1981; Dias, L.A.T. 1995: 266-267; Alarcão, J. 1988: 1/481.

13072202, S. TIAGO (BOGALHOS), Soalhães, Marco de Canaveses, Porto (130722). [500]. Typology: Romanized Hill-fort; Pottery Kiln. Chronology: Iron Age; Roman Period. References: Aguiar 1947: 296; Silva 1992: 82-93; Vasconcelos 1901: 177-178; Vieira 1887: 492; Dias 1995: 264. 13072201, SOALHÃES, Soalhães, Marco de Canaveses, Porto (130722). Typology: Milestone. Chronology: Roman Period. References: Cortez, F. R. 1951: 44-45; HAE, 469; Alarcão, J. 1988: 1/483.

13072901, TORRE, Várzea de Ovelha; Aliviada, Marco de Canaveses, Porto (130729). [200]. Typology: Small Farm. Chronology: Roman Period. References: Dias, L.A.T. 1995: 270.

13072302, CANAVESES, Sobretâmega, Marco de Canaveses, Porto (130723). [570254.45, 4560760.63, 50]. Typology: Roman bridge. Chronology: Roman Period. References: Miranda, A. 1944: 21; Monteiro, A. 1948: 52; Soeiro, T. 1984: 48; Alarcão, J. 1988: 1/422; Dias, L.A.T. 1995: 265.

13072801, VÁRZEA DO DOURO (Alto das Penegotas), Várzea do Douro, Marco de Canaveses, Porto (130728). [561600.00, 4547000.00, 114]. Typology: vicus; Fluvial port; Mansio. Chronology: Roman Period. References: Brandão, D.P. 1960; Lanhas, F., Brandão, D.P. 1967; Tranoy, A. 1984: 271; Centeno, R.M.S. 1987: 154; Alarcão, J. 1988: 91-92; Mantas, V. 1990: 231; Silva 1992: 165-172.

13072301, CALDAS DE CANAVESES, Sobretâmega, Marco de Canaveses, Porto (130723). [570923.68, 4561414.78, 80]. Typology: Roman necropolis; Thermal baths; vicus. Chronology: Roman Period. References: Vasconcelos 1902: 284-285; Vasconcelos, J. L. 1903: 284-285; Pinto, R. 1934: 165; Vasconcelos, M. 1935: 40; Vasconcelos, M, 1935: 2, 10; Miranda, A. 1943: 2; Cortez, F.R. 1946: 278; Soeiro, T. 1984: 4849; Alarcão, J. 1988: 422; Acuña Castroviejo, F. 1974; Mora, G. 1981: 37 e sg.; Miranda, A. 1943: 3; Dias 1995: 265; Miranda, A. 1944: 21; Monteiro, A. 1948: 52; Monteiro 1960; Acuña Castroviejo, F. 1973: 28-54; Soeiro, T. 1984: 48; Alarcão, J. 1988: 24; Dias, L.A.T. 1995: 265.

13073102, CASTRO DE QUIRES, Vila Boa de Quires, Marco de Canaveses, Porto (130731). [565092.41, 4562285.62, 512]. Typology: Romanized Hill-fort; vicus?; Medieval habitat. Chronology: Iron Age; Roman Period; Middle Ages. References: Costa 1706: 396; Leal, P. 1875; Vieira 1887: 483; Vasconcelos, J.L. 1897: 14; Vasconcelos, J.L. 1914-16; Fernandes, A. 1968; Soeiro, T. 1984: 44; Silva 1992: 55-67; Dias, L.A.T. 1995: 271. 13073101, TELHA-NECRÓPOLE, Vila Boa de Quires, Marco de Canaveses, Porto (130731). [566382.01, 4561341.18, 300]. Typology: Roman necropolis. Chronology: Roman Period. References: Soeiro, T. 1984: 44; Alarcão, J. 1988: 1/416; Almeida, C.A.F. 1980: 311.

13072601, ALVIM (Castro), Stº Isidoro, Marco de Canaveses, Porto (130726). [194]. Typology: Roman necropolis; Small Farm? Chronology: Roman Period. References: Alarcão, J. 1988: 1/421; Dias 1995: 263264. 13072401, CRASTOS, Tabuado, Marco de Canaveses, Porto (130724). [197]. Typology: Small Farm. Chronology: Roman Period. References: Silva 1992: 127132; Dias, L.A.T. 1995: 266.

13080302, FONTÃO DE ANTELA, Lavra, Matosinhos, Porto (130803). Typology: Villa. Chronology: Roman Period. References: Lanhas, F., Brandão, D.P. 1969: 304-313; Acuña Castroviejo, F. 1974: 30-32; Alarcão, J. 1988: 1/391.

13072704, PORTELA DE TUIAS (Quinta da Portela), Tuias, Marco de Canaveses, Porto (130727). [240]. Typology: Roman necropolis. Chronology: Roman Period. References: Dias, L.A.T. 1995: 267.

13080301, ANGEIRAS, Lavra, Matosinhos, Porto (130803). Typology: Garum factory. Chronology: Roman Period. References: Lanhas, F., Brandão, D. P. 1969: 318-325; Almeida, C. A. F. 1972: 6; Alarcão, J. 1988: 1/390.

13072703, PORTINHO, Tuias, Marco de Canaveses, Porto (130727). [, 40]. Typology: Fluvial port. Chronology: Roman Period. References: Dias, L.A.T. 1995: 266.

13080901, S. MAMEDE DE INFESTA, S. Mamede de Infesta, Matosinhos, Porto (130809). Typology: Milestone. Chronology: Roman Period. References: CIL II 4735; Alarcão, J. 1988: 1/399.

13072702, CASA DO OUTEIRO (Quinta Sousa Guedes), Tuias, Marco de Canaveses, Porto (130727). 178

Catalogue

13080902, QUINTA DO ALÃO, S. Mamede de Infesta, Matosinhos, Porto (130809). Typology Votive epigraphy. Chronology: Roman Period. References: Almeida, J. M. 1965: 71-73; Alarcão, J. 1988: 1/398.

necropolis. Chronology: Roman Period. References: Alarcão, J. 1988: 1/407. 13091301, CASTRO DA VILA, Penamaior, Paços de Ferreira, Porto (130913). [548573.78, 4570144.36, 419]. Typology: Roman treasure; Roman necropolis; Romanized Hill-fort. Chronology: Roman Period. References: Hipólito, M. C. 1960-1961: 45-46; Alarcão, J. 1988: 1/406; Queiroga, F. 1992.

13090201, S. ROQUE, Carvalhosa, Paços de Ferreira, Porto (130902). Typology: Roman necropolis; Pottery Kiln. Chronology: Roman Period. References: Dinis, M. V. 1962: 224, 227; Alarcão, J. 1988: 1/374. 13090301, CASTRO DO MONTE DO SOCORRO (Castro do Capelo Vermelho), Codessos, Paços de Ferreira, Porto (130903). [554045.45, 4575795.93, 423]. Typology: Romanized Hill-fort. Chronology: Roman Period. References: Coelho, J. F. 1965: 226228; Alarcão, J. 1988: 1/368; Queiroga, F. 1992.

13091401, REGUENGO, Raimonda (S. Pedro), Paços de Ferreira, Porto (130914). Typology: Roman necropolis. Chronology: Roman Period. References: Alarcão, J. 1988: 1/371. 13091501, CITÂNIA DE SANFIS, Sanfis de Ferreira, Paços de Ferreira, Porto (130915). [551495.01, 4574852.18, 570]. Typology: Romanized Hill-fort; Roman treasure. Chronology: Iron Age; Roman Period. References: Alarcão, J. 1988: 1/370; Hipólito, M. C. 1960-1961: 42-45; Silva, A. C. F. 1983; Brandão, D. P. 1963: 232-235; Queiroga, F. 1992.

13090401, EIRIZ, Eiriz, Paços de Ferreira, Porto (130904). Typology: Roman necropolis. References: Almeida, J. A. F. 1953; Guimarães, O. 1902: 265-266; Alarcão, J. 1988: 1/372. 13090701, CASTRO DE S. BRÁS, Frazão, Paços de Ferreira, Porto (130907). [ 322]. Typology: Romanized Hill-fort; Roman necropolis. Chronology: Roman Period. References: Dinis, M. V. 1976: 218-221; Alarcão, J. 1988: 1/402; Queiroga, F. 1992.

13091502, BOUÇAMONTE, Sanfis de Ferreira, Paços de Ferreira, Porto (130915). Typology: Roman necropolis. Chronology: Roman Period. References: Alarcão. J. 1988: 1/369.

13090702, BOAVISTA, Frazão, Paços de Ferreira, Porto (130907). Typology: Roman necropolis. Chronology: Roman Period. References: Dinis, M. V. 1962: 224; Alarcão, J. 1988: 1/401.

13091601, OUTEIRO DE FOGUETES, Seroa, Paços de Ferreira, Porto (130916). Typology: Roman necropolis. Chronology: Roman Period. References: Alarcão. J. 1988: 1/403.

13091001, BOUÇÓS, Meixomil, Paços de Ferreira, Porto (130910). Typology: Roman necropolis. Chronology: Roman Period. References: Dinis, M. V. 1980: 141-150; Alarcão, J. 1988: 1/376.

13100301, BALTAR, Baltar, Paredes, Porto (131003). Typology: Funerary epigraphy; Votive epigraphy. Chronology: Roman Period. References: CIL II 2378; 2379; Encarnação, J. 1975: 240; Alarcão, J. 1988: 1/456.

13091002, DEVESA GRANDE, Meixomil, Paços de Ferreira, Porto (130910). Typology: Roman necropolis. Chronology: Roman Period. References: Dinis, M. V. 1962: 224; Alarcão, J. 1988: 1/375.

13101701, PARADA TODEIA, Parada Todeia, Paredes, Porto (131017). Typology: Roman necropolis; Villa. Chronology: Roman Period. References: Correia, A. A. M. 1923-1924: 1-14; Alarcão, J. 1988: 1/461.

13091101, CIMO DE VILA (Outeiro da Vela), Modelos, Paços de Ferreira, Porto (130911). Typology: Roman habitat. Chronology: Roman Period. References: Dinis, M. V. 1981: 388-390; Alarcão, J. 1988: 1/409.

13102001, SANTA COMBA, Sobreira, Paredes, Porto (131020). Typology: Votive epigraphy. Chronology: Roman Period. References: Tranoy, A. 1981: 271; Alarcão, J. 1988: 1/467.

13091201, CASTÊLOS, Paços de Ferreira, Paços de Ferreira, Porto (130912). [552170.50, 4568195.50, 339]. Typology: Roman habitat? Chronology: Roman Period. References: Dinis, M. V.1978: 187-190; Alarcão, J. 1988; Queiroga, F. 1992.

13102002, BANJAS, Sobreira, Paredes, Porto (131020). Typology: Roman mines-gold. Chronology: Roman Period. References: Soeiro, T. 1984: 112-118; Alarcão. J. 1988: 1/465. 13110102, CASTRO DE ABRAGÃO, Abragão, Penafiel, Porto (131101). [563699.61, 4556413.61, 200]. Typology: Hill-fort. Chronology: Iron Age. References: Soeiro, T.1984: 50; Dias, L.A.T. 1995: 271.

13091202, QUINTÃS, Paços de Ferreira, Paços de Ferreira, Porto (130912). Typology: Roman necropolis. Chronology: Roman Period. References: Alarcão, J. 1988: 1/408.

13110101, PENHA GRANDE, Abragão, Penafiel, Porto (131101). [564464.96, 4556852.1, 248]. Typology: Romanized Hill-fort; Roman necropolis. Chronology: Iron Age; Roman Period. References: Soeiro, T.

13091203, PAÇOS DE FERREIRA, Paços de Ferreira, Paços de Ferreira, Porto (130912). Typology: Roman 179

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

1984: 50-51; Alarcão, J. 1988: 1/475; Dias, L.A.T. 1995: 271.

ogy: Roman necropolis. Chronology: Roman Period. References: Soeiro, T. 1984: 98; Dias, LA.T. 1995: 272.

13110201, BOLSA DO OURO, Boelhe, Penafiel, Porto (131102). [564031.25, 4553147.44, 120]. Typology: Small Farm. Chronology: Roman Period. References: Lanhas, F., Brandão, D.P. 1965: 298; Soeiro, T. 1984: 53; Alarcão 1988: 1/478; Dias, L.A.T. 1995: 271.

13111303, BOA VISTA, ABUJEFA, Galegos, Penafiel, Porto (131113). Typology: Roman treasure. Chronology: Roman Period. References: Ferreira, S.R. 19/3/1864; Ferreira, S.R. 1880: 18-19; Vasconcelos, J.L. 1895: 15-16; Machado, F.S.L. 1920: 37; Soeiro, T. 1984: 98; Alarcão, J. 1988.

13110401, JUGUEIROS, Cabeça Santa, Penafiel, Porto (131104). Typology: Roman necropolis. Chronology: Roman Period. References: Soeiro, T. 1984: 64; Alarcão 1988: 1/479; Dias, L.A.T. 1995: 275.

13111502, CASTRO DE IRIVO, Irivo, Penafiel, Porto (131115). [555941.06, 4558940.63, 157]. Typology: Hill-fort. Chronology: Iron Age; Roman Period. References: Miranda, A. 3/3/1959; Soeiro, T. 1984: 100.

562/13110504, QUINTA DA BOA VISTA, Canelas, Penafiel, Porto (131105). [557893.63, 4547822.68, 270]. Typology: Roman necropolis. Chronology: Roman Period. Miranda, A. 1943; Soeiro, T. 1984: 77, 79; Dias, L.A.T. 1995: 271.

13111501, ERMIDA, Irivo, Penafiel, Porto (131115). [556367.89, 4558049.55, 180]. Typology: Votive epigraphy. Chronology: Roman Period (II A.D). References: Miranda, A. 1944: 65-66; Soeiro, T. 1984: 100; Alarcão, J. 1988: 1/457.

13110503, PÓVOA, Canelas, Penafiel, Porto (131105). [557819.73, 4548315.53, 270]. Typology: Roman habitat. Chronology: Roman Period. References: Miranda, A. 1943; Soeiro, T. 1984: 79; Dias, L.A.T. 1995: 271.

13111605, ALTO DO CASTRO Lagares, Penafiel, Porto (131116). [554189.16, 4553314.59, 284]. Typology: Hill-fort. Chronology: Iron Age. References: Sousa, E. 1962: 128.

13110502, CHAVES, Canelas, Penafiel, Porto (131105). [557604.58, 4546000.88, 80]. Typology: Roman habitat. Chronology: Roman Period. References: Miranda, A. 1945; Dias, L.A.T. 1995: 272.

13111601, LAGARES, Lagares, Penafiel, Porto (131116). [553209.60, 4553338.24, 200]. Typology: Votive epigraphy. Chronology: Roman Period. References: Silva, A. C. F. 1984: 47; Soeiro, T. 1984: 104; Garcia, J.M. 1991: 360; Alarcão, J. 1988: 1/470.

13110801, CROCA, Croca, Penafiel, Porto (131108). [437]. Typology: Roman necropolis; Mansio; vicus? Chronology: Roman Period. References: Dias, L.A.T. 1995: 272.

13111602, PRESA DOS VASOS, Lagares, Penafiel, Porto (131116). [553744.77, 4553496.35, 230]. Typology: Roman necropolis. Chronology: Roman Period References: Soeiro, T. 1984: 104; Alarcão, J. 1988: 1/469.

13110901, ANTAS, Duas Igrejas, Penafiel, Porto (131109). [561292.44, 4559939.66, 400]. Typology: Roman necropolis. Chronology: Roman Period. References: Miranda, A. 27-29; Miranda, A. 1942: 6566; Brandão, D.P. 1961: 29; Pereira, I 1975: 185; Soeiro, T. 1984: 90, 95; Alarcão, J. 1988: 1/474; Dias 1995: 272.

13111603, AGRAS DE ORDINS, Lagares, Penafiel, Porto (131116). [581067.68, 4554149.2, 285]. Typology: Roman habitat; Roman necropolis. Chronology: Roman Period. References: Soeiro, T. 1984: 104; Alarcão, J. 1988: 1/468.

13110501, CIVIDADE DE EJA (Coto da Cividade de Eja; Castro de Entre os Rios), Eja, Penafiel, Porto (131105). [558728.50, 4548476.96, 207]. Typology: Farm; Roman necropolis. Chronology: Roman Period. References: Andrade, V. 1918: 74-76; Pinho, J. 1929: 182-187; Aguiar, J.M. 1943: 13; Aguiar, J.M. 1945: 91; Almeida, C.A.F., Lopes, F.G.A. 1981-1982: 131; Soeiro, T., 1984: 76-77; Dias, LA.T. 1995: 272.

13111604, CASTELÕES, Lagares, Penafiel, Porto (131116). [553861.35, 4553497.21, 275]. Typology: Hill-fort. Chronology: Iron Age. References: Soeiro, T. 1984: 104. 13111701, OUTEIRO, Luzim, Penafiel, Porto (131117). [563471.89, 4555794.82, 250]. Typology: Villa. Chronology: Roman Period. References: Lanhas, F., Brandão, D.P. 1965: 296; Soeiro, T. 1984: 51; Alarcão, J. 1988: 1/476; Dias, L.A.T. 1995: 272.

13111301, CASTRO DE ABUJEFA, Galegos, Penafiel, Porto (131113). [557005.48, 4556975.1, 330]. Typology: Romanized Hill-fort. Chronology: Roman Period. References: Soeiro, T. 1984: 31, 98; Alarcão, J. 1988; Dias, LA.T. 1995: 272.

13111801, MARECOS, Marecos, Penafiel, Porto (131118). Typology: Roman treasure. Chronology: Roman Period. References: Ferreira, S. R. 19/3/1864; Vasconcellos, J.L. 1895: 15-16; Soeiro, T. 1984: 31; Alarcão, J. 1988: 1/458.

570/13111302, GALEGOS-IGREJA, Galegos, Penafiel, Porto (131113). [557703.19, 4557165.62, 235]. Typol180

Catalogue

13111802, PÓVOA, Marecos, Penafiel, Porto (131118). [557077.59, 4670045.02, 210]. Typology: Sanctuary; Votive epigraphy. Chronology: Iron Age?; Roman Period. References: Encarnação, J. 1975: 244-246; Alarcão, J. 1988: 1/457.

gy: Romanized Hill-fort; Roman epigraphy- undetermined; Roman necropolis. Chronology: Roman Period. Aguiar, J.M. 1936: 116, 226-227 Miranda, A. 1942 Miranda, A. 1943 Miranda, A. 1944: 25-26 Sousa, E.F. 1962: 128 Soeiro, T. 1984: 60 Alarcão, J. 1988: 1/472 Dias, L.A.T. 1995: 273

13112401, SANTA LUZIA, Milhundos, Penafiel, Porto (131124). [559114.13, 4561371.17, 230]. Typology: Farm. Chronology: Roman Period. References: Miranda, A. s/d: 25; Mendes, J.J. 1966-1967: 57; Soeiro, T. 1984: 89; Alarcão, J. 1988: 1/411.

13112604, S. V. PINHEIRO, Pinheiro, Penafiel, Porto (131126). [559422.27, 4552121.65, 155]. Typology: Roman necropolis. Chronology: Roman Period. References: Fortes, J. 1902; Lacerda, F.S. 1920: 23; Soeiro, T. 1984: 75; Dias, L.A.T. 1995: 273.

13112101, MONTE MÓZINHO, Oldrões, Penafiel, Porto (131121). [557810.56, 4555377.76, 408]. Typology: Romanized Hill-fort; Roman necropolis; vicus? Chronology: Roman Period. References: Soeiro, T. 1984: 123-299; Alarcão, J. 1988: 1/466; Dias, L.A.T. 1995: 273.

13112801, PEDREIRA, Rãs, Penafiel, Porto (131128) [558018.92, 4558494.25, 195]. Typology: Roman necropolis. Chronology: Roman Period. References: Soeiro, T. 1984: 96; Dias, L.A.T. 1995: 273; Alarcão, J. 1988: 1/457.

13112102, CASTELO DE PENAFIEL, Penafiel, Porto (131121). [558531.56, 4715494.18, 324]. Typology: Medieval castle. Chronology: Middle Ages. References: Soeiro, T. 1984: 60.

13113201, SUVIDADE DE RECEZINHOS, Recezinhos (S. Mamede), Penafiel, Porto (131132). CMP [564812.37, 4564966.29, 285]. Typology: Hillfort. Chronology: Roman Period. References: P.M.H.: Inquisitiones 1897: 602; Soeiro, T. 1984: 46, 48.

13112201, COVELO, Paço de Sousa, Penafiel, Porto (131122). [555607.37, 4556933.49, 175]. Typology: Roman necropolis; Roman habitat? Chronology: Roman Period. References: Miranda, A. 1942: 3; Soeiro, T. 1984: 100; Alarcão, J. 1988: 1/463.

13113301, QUINTA DO CASTRO, Recezinhos (São Martinho), Penafiel, Porto (131133). [565664.92, 4563370.19, 310]. Typology: Romanized Hill-fort; Roman necropolis. Chronology: Roman Period. References: Soeiro, T. 1984: 44, 45; Dias, L.A. T. 1995: 274.

13112301, GIESTA (Tapada de Barreiros), Paredes, Penafiel, Porto (131123). [560003.27, 4552373.11, 196]. Typology: Roman necropolis. Chronology: Roman Period. References: Machado, F.S.L. 1920: 25; Soeiro, T. 1984: 64; Dias, L.A.T. 1995: 273; Alarcão, J. 1988: 1/471.

13112902, QUINTELA, Rio de Moinhos, Penafiel, Porto (131129). [560908.96, 4550036.77, 119]. Typology: Hill-fort. Chronology: Iron Age. References: Lanhas, F., Brandão, D. P. 1967: 36; Soeiro, T. 1984: 55; Dias, L.A. T. 1995: 274.

13112402, PENAFIEL, Penafiel, Penafiel, Porto (131124). Typology: Roman Sculpture. Chronology: Roman Period. References: Brandão, D. P. 1961: 1821; Alarcão, J. 1988: 1/410.

13112901, CODES, Rio de Moinhos, Penafiel, Porto (131129). [562387.75, 4551714.58, 215]. Typology: Roman necropolis; Farm. Chronology: Roman Period. References: Lanhas, F., D.P. Brandão1965: 298; Soeiro, T. 1984: 54; Dias, L.A. T. 1995: 274; Alarcão, J. 1988: 1/477.

13112903, OUTEIRO/ OUTRELA, Perozelo, Penafiel, Porto (131129). [561259.28, 4555560.06, 300]. Typology: Roman habitat. Chronology: Roman Period. References: Soeiro, T. 1984: 55; Dias, L.A.T. 1995: 274.

13113002, MONTES NOVOS, Santa Marta, Penafiel, Porto (131130). Typology: Roman habitat. Chronology: Roman Period. References: Soeiro, T. 1984: 85.

13112603, REBOLIDO, Pinheiro, Penafiel, Porto (131126). [558273.38, 4552883.4, 250]. Typology: Roman necropolis. Chronology: Roman Period. References: Aguiar, J.M. 1936: 116; Soeiro, T. 1984: 60, 64.

13113001, ESTRADA-SANTA MARTA, Santa Marta, Penafiel, Porto (131130). [562101.96, 4563369.8, 318]. Typology: Roman necropolis. Chronology: Roman Period. References: Soeiro, T. 1984: 85; Dias, L.A. T. 1995: 274; Alarcão, J. 1988: 1/415.

590/13112601, S. VICENTE DO PINHEIRO, Pinheiro, Penafiel, Porto (131126). [559397.19, 4552337.33, 150]. Typology: Thermal baths; Pottery Kiln; vicus. Chronology: Roman Period. References: Fortes, J. 1902; Correia, H.M., 1926: 60; Soeiro, T. 1984: 67-75; Alarcão, J. 1988: 1/473; Dias, L.A. T. 1995: 273.

13111304, QUINTA DO BAIRRO-ABUJEFA, Galegos, Penafiel, Porto (131113). [557142.18, 4557377.09, 215]. Typology: Roman treasure. Chronology: IV d.C. References: Soeiro, T. 1984: 85; Dias, L.A. T. 1995: 272. 13112202, CRISTELO, Paço de Sousa, Penafiel, Porto (131122). [555538.16, 4556840.44, 209]. Typology:

13112602, OUTEIRO DO DINO, Pinheiro, Penafiel, Porto (131126). [558110.15, 4552882.1, 274]. Typolo181

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Hill-fort. Chronology: Iron Age? References: Pinho, J. 1931: 10; Soeiro, T. 1984: 100.

138; Silva, A. C. F. 1986; Centeno, R. M. S. 1987: 130; Calo Lourido, F. 1991: 585-589; Queiroga, F., 1992: 168; Silva, A. C. F., Gomes, M.V. 1992; Dinis, A. P. 1993: 30-34; Gomes, M.V. 1996.

13121401, CALE (Cividade -Morro da Sé), Sé, Porto, Porto (131214). [532754.38, 4554574.87, 76]. Typology: Civitas capital. Chronology: Roman Period. References: Soares, T. S. 1962: 145-155; Brandão, D. P. 1984: 13-18; Alarcão, J. 1988: 1/450; Queiroga, F. 1992; Silva, A.C.P. 2000: 94-104; Teixeira, R.; Dórdio, P. 2000:132-133; Morais, R. 2005.

13140201, PIDRE, Água Longa, Santo Tirso, Porto (131402). Typology: Roman treasure. Chronology: Roman Period. References: Hipólito, M. C. 1974: 5-6; Alarcão, J. 1988: 1/404. 13140302, QUINTA DO PAIÇO, Alvarelhos, Santo Tirso, Porto (131403). Typology: Milestone. Chronology: Roman Period. References: Santarém, C. M. F. 1956: 69-70; Almeida, C. A. F. 1969: 27-29; Encarnação, J. 1975: 270-274; Tranoy, A. 1981: 314; Alarcão, J. 1988: 1/351; Mantas, V. 1996: 352-353.

13130601, ALTO DA VINHA, Beiriz, Póvoa de Varzim, Porto (131306). CMP 82 [523200.00, 4582600.00, 0]. Typology: Votive epigraphy. Chronology: Roman Period. References: Gonçalves, F. 1949: 225-229; Le Roux, P.; Tranoy, A. 1973: 213-214; Alarcão, J. 1988: 1/285; Gomes, M.V. 1996.

13140303, SOBRE SÁ, Alvarelhos, Santo Tirso, Porto (131403). Typology: Villa. Chronology: Roman Period. References: Santarém, C. M. F. 1954: 31-39; Silva, A. C. F. 1980: 84-86; Alarcão, J. 1988: 1/350.

13130701, VILA MENDO, Estela, Póvoa de Varzim, Porto (131307). CMP 82 [521000.00, 4588500.00, 0]. Typology: Treasure. Chronology: Roman Period. References: Almeida, C. A. F. 1972: 33; Alarcão, J. 1988: 1/** ; Gomes, M.V. 1996.

13140304, PEÇA MÁ, Alvarelhos, Santo Tirso, Porto (131403). Typology: Milestone. Chronology: Roman Period. References: Cruz, A. 1940: 210; Almeida, C. A. F. 1969: 17; Alarcão, J. 1988: 1/349; Mantas, V. 1996.

13130801, MONTE DE S. FÉLIX, Laundos, Póvoa de Varzim, Porto (131308). CMP 82 [523900.00, 4587300.00, 208]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Peixoto 1905-08b: 284; Severo, R. 1905-08: 403-412; Sarmento, M. 1933a: 166; Lopez Cuevillas 1951c: 7980; Blanco Freijeiro 1957a: 288-289; Almeida, C.A.F. 1972: 10-11; Araújo 1980: 103; Silva 1986: 263; Queiroga, F. 1992: 168; Gomes, M.V. 1996.

13140401, TORRE ALTA (Castro de Areias), Areias, Santo Tirso, Porto (131404). CMP 98 [543847.36, 4578471.94, 63]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Costa, A. C. 1706-1712, 1868-1869: 324; Pimentel, A. 1902: 64; Lima, A. C. P. 1956: 222-223; Silva, A. C. F. 1986: 83; Queiroga, F. 1992: 168; Dinis, A. P. 1993: 59-60.

13130901, CRASTO, Navais, Póvoa de Varzim, Porto (131309). CMP 82 [518750.41, 4587892.98, 84]. Typology: Hill-fort. Chronology: Iron Age; Roman Period? References: Sarmento, F.M. 1933: 165; Fortes, J. 1969: 322; Almeida, C.A.F. 1972: 10; Silva 1986 Queiroga, F. 1992: 168.

13140402, CALDAS; MONTINHO, Areias, Santo Tirso, Porto (131404). Typology: Roman habitat. Chronology: Roman Period. References: Lima, A. C. P. 1940: 9; Alarcão, J. 1988: 1/297. 13142501, TROFA VELHA, Bougado (S. Martinho), Santo Tirso, Porto (131425). Typology: Milestone. Chronology: Roman Period. References: Capella, M. 1895: 221-222, 226, 251; Almeida, C. A. F. 1969: 17; Alarcão, J. 1988.

13131001, ALTO DE MARTIM VAZ, Póvoa de Varzim, Póvoa de Varzim, Porto (131310). CMP 82 [520200.00, 4581550.00, 0]. Typology: Villa. Chronology: Roman Period. References: Fortes, J. 1969: 313-341; Almeida, C.A.F. 1972: 31-32; Alarcão, J. 1988: 1/287; Gomes, M.V. 1996.

13142502, RODRIGO VELHO, Bougado (S. Martinho), Santo Tirso, Porto (131425). Typology: Roman necropolis. Chronology: Roman Period ( III A.D.). References: Cruz, A. 1940: 203-215; Alarcão, J. 1988: 1/347.

13131201, CIVIDADE DE TERROSO, Terroso, Póvoa de Varzim, Porto (131312). CMP 82 [523424.08, 4584822.66, 153]. Typology: Romanized Hill-fort. Chronology: Bronze Age; Ferro, Roman Period. References: Costa, A. C., 1706-1712: 313; Sarmento, F. M., 1933: 166; Peixoto, R. 1905-08: 677-680; Peixoto, R 1964: 677-680; Severo, R. 1905-08: 404-405; Fortes, J. 1905-1908: 605-617, 662-665; Pinto, R. S. 1928: 311-312; Pinto, R. S. 1929: 311-312; Pinto, R. S 1964: 81-91; Gonçalves, F. 1964: 307-308; Teixeira, C., Medeiros, A.C., Assunção, C.T. 1965: 46; Freire, O. 1965: 206-209; Garcia Y Bellido, A. 1966: 9-11; Almeida, C. A. F. 1972: 15-31; Romero Masiá, A. 1976; Silva, A. C. F. 1981: 303-315; Ponte, S. 1984:

13142101, BAIRROS, Bougado (Santiago), Santo Tirso, Porto (131421). Typology: Roman necropolis. Chronology: Roman Period. References: Almeida, C. A. F. 1969: 4; Alarcão, J. 1988: 1/345. 13140601, S. SIMÃO; SANTA CRUZ, Burgães, Santo Tirso, Porto (131406). Typology: Votive epigraphy. Chronology: Roman Period. References: Encarnação, J. 1970: 225-228; Silva. 1986: 83; Alarcão, J. 1988: 1/359; Dinis, A.P. 1993. 182

Catalogue

13142601, S. MARTINHO DO CAMPO, Campo (S. Martinho), Santo Tirso, Porto (131426). Typology: Votive epigraphy. Chronology: Roman Period. References: Cardozo, M. 1972: 2; Encarnação, J. 1975: 77-78; Tranoy, A. 1981: 268; Alarcão, J. 1988: 1/365.

1956: 68; Cardozo, M. 1972: 40; Le Roux, P. 1982: 182-183; Alarcão, J. 1988: 1/358. 13143101, MONTE DOS SALTOS, Sequeirô, Santo Tirso, Porto (131431). Typology: Roman treasure; Roman habitat. Chronology: Roman Period. References: Pimentel 1902: 349; Sarmento 1940: 103; Hipólito, M. C. 1960-1961: 41; Monteiro 1985; Silva 1986: 83; Alarcão, J. 1988: 1/356; Dinis, A.P. 1993: 104.

13142602, BOCAS, Campo (S. Martinho), Santo Tirso, Porto (131426). Typology: Roman epigraphy- undetermined. Chronology: Roman Period. References: CIL II 5608; Santarém, C. M. F. 1956: 65; Alarcão, J. 1988: 1/364.

13143201, LAJE, Vilarinho, Santo Tirso, Porto (131432). Typology: Roman treasure. Chronology: Roman Period. References: Hipólito, M. C. 1960-1961: 110-111; Alarcão, J. 1988: 1/361.

13142603, CHÃOS, Campo (S. Martinho), Santo Tirso, Porto (131426). Typology: Roman epigraphy- undetermined. Chronology: Roman Period. References: Sarmento, M. 1970: 15; Alarcão, J. 1988: 1/362.

13140301, CASTRO DE ALVARELHOS (Castro de S. Marçal), Alvarelhos, Trofa, Porto (131403). CMP 97 [531860.95, 4572272.36, 222]. Typology: vicus. Chronology: Bronze Age; Iron Age, Roman Period. References: Fortes, J. 1899: 7-28; Torres, J. 1905: 381382; Ferreira, M, J. A. 1907: 285; Ferreira, M, J. A. 1923: 47; Sarmento, F. M. 1933: 165, 168; Azevedo, A. 1939: 101-102; Santarém, C. M. F. 1956: 63-64; Teixeira, C., Medeiros, A. C., Assunção, C. T. 1965: 46; Almeida, C. A. F. 1969: 31-32; Sarmento, F. M 1970, 1986, 1987/1988, 1989: 48-52; Cardozo, M. 1972: 46; Almeida, C. A. F. 1973: 207; Santarém, C. M. F. 1977: 161-170; Torres, J. 1978-79: 15-250; Soeiro, T. 1980: 237-243; Silva, A. C. F. 1980/1982: 79-90; Tranoy, A. 1981: 322; Pinto, R. S. 1984: 123; Silva, A. C. F. 1986: 83; Centeno, R. M. S. 1987: 3441, 111; Alarcão, J. 1988: 1/348; Moreira, A. 1992: 3447; Queiroga, F. 1992: 168; Almeida C. A. B. 1992; Moreira 1992; Dinis, A. P. 1993: 96-98; Mantas, V. 1996; VVAA 2006.

13141301, CASTRO DO MONTE PADRÃO, Monte Córdova, Santo Tirso, Porto (131413). CMP 98 [546246.56, 4573922.5, 413]. Typology: Romanized Hill-fort; Roman epigraphy- undetermined; vicus? Chronology: Bronze Age; Iron Age, Roman Period; Middle Ages. References: Leal, A. P., Ferreira, A. 1873-1890: 471-472; Santarém, C. M. F. 1955: 397429; Santarém, C. M. F. 1956: 66; Tomás, F. L. S. 1974: 2, 159-160; Ponte, S. 1984: 129; Martins, M. 1985: 217-230; Silva, A. C. F. 1986: 83; Centeno, R. M. S. 1987: 115-116; Alarcão, J. 1988; Moreira, A. 1991; Queiroga, F. 1992: 169. 13141402, S. CRISTÓVÃO DO MURO, Muro, Santo Tirso, Porto (131414). Typology: Roman necropolis. Chronology: Roman Period. References: Alarcão, J. 1988: 1/352. 13141401, CARRIÇA, Muro (São Cristóvão), Santo Tirso, Porto (131414). Typology: Milestone. Chronology: Roman Period. References: CIL II 4743; Santarém, C. M. F. 1956: 72; Mantas, V. 1996: 349; Alarcão, J. 1988: 1/354.

13150501, SANTA JUSTA, Valongo, Valongo, Porto (131505). Typology: Funerary epigraphy. Chronology: Roman Period. References: Fortes, J. 1905-1908: 125; Alarcão, J. 1988: 1/455.

13141901, RORIZ, Roriz, Santo Tirso, Porto (131419). Typology: Funerary epigraphy. Chronology: Roman Period. References: Santarém, C. M. F. 1956: 69; Alarcão, J. 1988: 1/367.

13150502, FOJO DAS POMBAS, Valongo, Valongo, Porto (131505). Typology: Roman mines-gold. Chronology: Roman Period. References: Domergue, C. 1970: 157; Soeiro, T. 1984: 109.

13141902, SANTA MARIA DE NEGRELOS, Roriz, Santo Tirso, Porto (131419). Typology: Funerary epigraphy; Votive epigraphy. Chronology: Roman Period. References: CIL II 5568 e 5582; Cardozo, M. 1972: 48-90; Alarcão, J. 1988: 1/366.

13150503, S. BARTOLOMEU, Valongo, Valongo, Porto (131505). Typology: Votive epigraphy. Chronology: Roman Period. References: Alarcão, J. 1988: 1/405.

13142503, S. BARTOLOMEU, S. Martinho de Bougado, Santo Tirso, Porto (131425). Typology: Votive epigraphy. Chronology: Roman Period. References: Encarnação, J. 1975: 248 Alarcão, J. 1988: 1/357

13162901, SANTA MARIA DE VILAR, Vilar, Vila do Conde, Porto (131629). CMP 97 [527710.00, 4571010.00, 55]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C. A. F. 1969: 37; Alarcão, J. 1988: 1/341.

13142201, SANTO TIRSO, Santo Tirso, Santo Tirso, Porto (131422). Typology: Votive epigraphy. Chronology: Roman Period. References: Santarém, C. M. F.

13160101, CASTRO DE ARGIFONSO, Arcos, Vila do Conde, Porto (131601). CMP 83 [528704.82, 4582312.82, 157]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: 183

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Ayres, C. 1896: 386; Azevedo, P. A. 1896-1902: 28; Sarmento, F. M. 1933: 165; Cortez, F. R. 1946: 149155; Cortez, F. R. 1949b: 270-288; Teixeira, C., Medeiros, A. C., Assunção, C. T. 1965: 46; Silva, A. C. F.1986: 83; Queiroga, F. 1992: 169; Dinis, A. P. 1993: 44-45.

References: Alarcão, J. 1988: 1/339; Severo, R. 18991903: 129-130. 13161101, MILREUS, Guilhabreu, Vila do Conde, Porto (131611). CMP 97 [530200.00, 4571530.00, 100]. Typology: Roman habitat. Chronology: Roman Period. References: Alarcão, J. 1988: 1/353.

13160102, CASTRO DE CASAIS, Arcos, Vila do Conde, Porto (131601). CMP 83 [528308.57, 4582681.34, 85]. Typology: Hill-fort. Chronology: Iron Age; Roman Period?. References: Almeida, C. A. B. 1992; Dinis, A. P. 1993: 43-44.

13161102, PAIÇO, Guilhabreu, Vila do Conde, Porto (131611). CMP 97 [530040.00, 4572520.00, 90]. Typology: Villa. Chronology: Roman Period. References: Almeida, C. A. F. 1969: 35; Alarcão, J. 1988. 13161103, VILA BOA, Guilhabreu, Vila do Conde, Porto (131611). CMP 97 [530330.00, 4572200.00, 100]. Typology: Funerary epigraphy; Roman necropolis. Chronology: Roman Period. References: Alarcão, J. 1988: 1/340; Almeida, C. A. F. 1969: 23-24; Silva, A. C. F. 1980: 82-84.

13160501, CIVIDADE DE BAGUNTE, Bagunte, Vila do Conde, Porto (131605). CMP 83 [528893.08, 4581696.78, 206]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Costa, A. C. 1706-1712, 1868-1869: 321-322; Severo, R., Cardozo, F. 1886: 137-141; Figueiredo, A. M. 1897: 218; Azevedo, P. A., 1896-1902: 28; Fortes, J. 1905-1908a: 17; Severo, R. 1905-08: 405; Sarmento, F. M. 1933: 166; Cortez, F. R. 1948: 276-277; López Cuevillas 1951: 38; Teixeira, C., Medeiros, A. C., Assunção, C. T. 1965: 46; Romero Masiá, A. 1976: 106, 109-110; Ponte, S. 1980: 116; Silva, A. C. F. 1986; Calo Lourido, F. 1991: 99-102; Queiroga, F. 1992: 169; Almeida C. A. B. 1992; Dinis, A. P. 1993: 46-48.

13161301, CASTRO DE S. PAIO, Labruje, Vila do Conde, Porto (131613). CMP 109 [522704.17, 4570264.31, 14]. Typology: Hill-fort. Chronology: Iron Age; Roman Period? References: Lanhas, F., Brandão, D.P. 1969: 302, 328-335; Silva 1986: 83; Queiroga, 1992:169; Almeida, C.A.B. 1992; Dinis 1993, 101-102. 13161801, MOSTEIRÓ, Mosteiró, Vila do Conde, Porto (131618). CMP 110 [527800.00, 4570100.00, 30]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C. A. F. 1969: 37; Alarcão, J. 1988: 1/343.

13160502, CASTRO DE SANTAGÕES, Bagunte, Vila do Conde, Porto (131605). CMP 97 [526464.48, 4578911.96, 63]. Typology: Hill-fort. Chronology: Iron Age; Roman Period. References: Costa, A. C. Braga, 1706-1712: 320-1, 1868-1869; Leal, A. P. Ferreira, A. 1873-1890: 440-1; Ayres, C. 1896: 388; Severo, R.; Cardozo, F. 1899-1903: 179; Sarmento, F. M. 1933: 165; Silva, A. C. F. 1986: 83; Queiroga, F. 1992: 169; Almeida C. A. B. 1992; Dinis, A. P.1993: 42-43.

13161802, ARÕES, Mosteiró, Vila do Conde, Porto (131618). CMP 110 [529200.00, 4570600.00, 70]. Typology: Roman necropolis. References: Almeida, C. A. F. 1969: 44; Alarcão, J. 1988: 1/342. 13162501, TOUGUINHÓ, Touguinhó, Vila do Conde, Porto (131625). CMP 83 [524700.00, 4581200.00, 40]. Typology: Roman necropolis. References: Alarcão, J. 1988: 1/286.

13160503, VILA VERDE, Bagunte, Vila do Conde, Porto (131605). CMP 97 [527720.00, 4578430.00, 54]. Typology: Roman necropolis. Chronology: Roman Period. References: Alarcão, J. 1988: 1/337; Severo, R. 1905-1908: 426-431.

13162801, VILA DO CONDE, Vila do Conde, Vila do Conde, Porto (131628). CMP 96 [521600.00, 4578300.00, 20]. Typology: Roman necropolis; Villa. Chronology: Roman Period. References: Alarcão, J. 1988: 1/338; Almeida, C. A. F. 1974: 209-222; Almeida, C. A. F.1972: 117.

13160504, VILAR, Bagunte, Vila do Conde, Porto (131605). CMP 97 [528000.00, 4579300.00, 63]. Typology: Roman habitat. Chronology: Roman Period. References: Alarcão, J. 1988: 1/336; Russell Cortez, F. 1950: 273.

13162101, CASTRO DA RETORTA, Retorta, Vila do Conde, Porto (131621). CMP 96 [523212.09, 4578746.63, 40]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Silva, A. C. F. 1986: 83; Severo, R., Cardozo, F. 18991903: 179-180; Sarmento, F. M. 1933: 165; Freitas, E. A. C. 1949: 20-27; Almeida, C. A. F. 1969: 33; Almeida C. A. B. 1992; Queiroga, F. 1992: 170; Dinis, A. P. 1993: 93-94.

13160801, CASTRO DE SANTA MARINHA DE FERREIRÓ, Ferreiró, Vila do Conde, Porto (131608). CMP 97 [530581.08, 4577972.13, 42]. Typology: Hillfort. Chronology: Iron Age; Roman Period. References: Severo, R., Cardozo, F., 1899-1903: 179-180; Dinis, A. P. 1993: 48-50; Silva, A. C. F. 1986: 83; Queiroga, F. 1992: 169; Almeida, C. A. B. 1992. 13160901, SOUTELO, Fornelo, Vila do Conde, Porto (131609). CMP 97 [529600.00, 4576100.00, 0]. Typology: Roman habitat. Chronology: Roman Period.

13162601, CASTRO DO BOI (Castro de Stº Ovídio), Vairão, Vila do Conde, Porto (131626). CMP 97 [527200.59, 4574689.71, 123]. Typology: Romanized 184

Catalogue

Hill-fort; Medieval castle. Chronology: Iron Age; Roman Period; Middle Ages. References: Sarmento, F. M. 1901: 46-47; Ferreira, M, J. A. 1907: 285; Ferreira, M, J. A. 1923; Sarmento, F. M. 1933: 165; Azevedo, A. 1939: 102; Freitas, E. A. C. 1949: 20; Almeida, C. A. F. 1969: 34; Almeida, C. A. F. 1978: 34; Silva, A. C. F. 1986: 83; Almeida C. A. B. 1992; Queiroga, F. 1992: 170; Dinis, A. P. 1993: 94-95.

33-39; Severo, R. 1904: 346; Pereira, F. A. 1910: 236; Pereira, F. A. 1915: 227; Pereira, F. A. 1929: 157; Pereira, F. A. 1938: 138; Pereira, F. A. 1939: 111; Pereira, F. A. 1941: 29 e 130, 318 e 320. 16010401, TOURAL (Outeiro do Toural), Azere or Aboim das Choças, Arcos de Valdevez, Viana do Castelo (160104). CMP 8. Typology: Roman habitat. Chronology: Roman Period. References: Pereira, F. A. 1923-1924: 276; Alarcão, J. 1988: 1/46.

13162802, CASTRO DE S. JOÃO, Vila do Conde, Vila do Conde, Porto (131628). CMP 96 [521889.35, 4578187.41, 30]. Typology: Hill-fort. Chronology: Iron Age? References: Leal-Ferreira 1873: 691-692; Vieira 1886: 262; Ayres, C. 1896: 386; Ferreira, 1906: 264; Ferreira, 1923: 11-13; Teixeira et alii, 1965: 46; Silva, 1986: 83; Amorim 1988: 11; Queiroga 1992: 170; Almeida, C.A.B. 1992; Dinis, A. P. 1993: 100.

16010601, CASTRO DAS NECESSIDADES (Castro de Cabreiro), Cabreiro, Arcos de Valdevez, Viana do Castelo (160106). CMP 8 [549583.20, 4644581.09]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Pereira, F. A.1898: 289303; Pereira, F. A. 1904: 214 e 219; Pereira, F. A. 1915: 227; Viana, A. 1926: 111; Pereira, F. A. 1938: 138; Moita, I. 1966: 533; Queiroga, F. 1992.

13170001, VILA NOVA DE GAIA, Vila Nova de Gaia, Vila Nova de Gaia, Porto (1317). Typology: Funerary epigraphy; Roman habitat. Chronology: Roman Period. References: Alarcão, J. 1988: 1/452; Silva Pinto, S. 1966: 83-94; Le Roux, P. 1982: 52,179; Silva, A. C. F. 1984: 44-48; Carvalho, T., Fortuna, J. 2000: 158; Guimarães, G. 2000: 155-158.

16010801, CASTRO DE CENDUFE (Castelo do Mau Vizinho), Cendufe, Arcos de Valdevez, Viana do Castelo (160108). [542710.14, 4628952.28, 164]. Typology: Romanized Hill-fort. Chronology: Iron Age. Roman Period. References: Azevedo, P. A. 1897: 196; Vasconcelos, J. L. 1900: 38; Pereira, F. A. 1908: 203207; Pereira, F. A 1910: 245; Pereira, F. A 1914 Pereira, F. A 1915: 1-16; Viana, A. 1926: 1; Pereira, F. A. 1938: 281; Pereira, F. A. 1941: 35-37; 326; Cardoso, M. 1946: 243-244; Gomes, C.A. 1979: 164; Queiroga, F. 1992.

VIANA DO CASTELO DISTRICT

16010102, CASTELO, Aboim das Choças, Arcos de Valdevez, Viana do Castelo (160101). CMP 8. Typology: Roman habitat? Chronology: Roman Period. References: Pereira, F. Alves 1896: 319-320; Pereira, F. Alves 1924: 276.

16011403, GIELA-IGREJA, Giela, Arcos de Valdevez, Viana do Castelo (160114). CMP 16 [549095.00, 4633749.00]. Typology: Roman necropolis? Chronology: Roman Period. References: Pereira, F. A. 1924: 251-255.

16010101, MONTE CASTRO (Castro de Eiras; Castro de Vilar; Cresto), Aboim das Choças, Arcos de Valdevez, Viana do Castelo (160101). CMP 8 [544864.34, 4642289.65, 285]. Typology: Roman treasure; Romanized Hill-fort. Chronology: Roman Period. References: Brito, M. J. C. 1908: 94-98; Pereira, F. A.1908: 202-244; Oliveira, M. 1910: 237; Pereira, F. A. 1924: 255, 275-276; Viana, A. 1932: 159; Hipólito, M.C. 1960-1961: 13-14; Gomes, C. de A. 1979: 162163; Gomes, C.A. 1979; Silva, A.C.F. 1986: 75.

16011401, QUINTA DO RIAL, Giela, Arcos de Valdevez, Viana do Castelo (160114). CMP 16 [549293.66, 4633681.7, 0]. Typology: Roman habitat. Chronology: Roman Period. References: Pereira, F. A. 1923-1924: 253.

16010201, SEARA, Aguiã, Arcos de Valdevez, Viana do Castelo (160102). CMP 16 [547569.11, 4636163.44,]. Typology: Roman necropolis? Chronology: Roman Period. References: Pereira, F. A. 1941: 256-257.

16011501, QUINTA DO OUTEIRO, Gondoriz, Arcos de Valdevez, Viana do Castelo (160115). CMP 16 Typology: Roman necropolis? Chronology: Roman Period. References: Pereira, F. A. 1924: 271-273.

16011402, GIELA, Giela, Arcos de Valdevez, Viana do Castelo (160114). Typology: Votive epigraphy. Chronology: Roman Period. References: Santos, L. A., Le Roux, P., e Tranoy, A. 1983: 193-194.

16011602, MONTE DAS CRUZES (Alto da Pedrada; Castro de Grade), Grade, Arcos de Valdevez, Viana do Castelo (160116). CMP 16 [552155.01, 4637409.45, 212]. Typology: Romanized Hill-fort? Chronology: Iron Age; Roman Period? References: Pereira, F. A. 1924: 261-262; Pereira, F. A. 1929: 155-157; Pereira, F. A. 1933: 227-228.

16010402, MONTE DE S.MIGUEL-O-ANJO (Castelo de S.Miguel-o-Anjo), Azere, Arcos de Valdevez, Viana do Castelo (160104). CMP 16 [549786.37, 4634386.45, 180]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Pereira, F. A. 1895: 161-175; Rocha, A. S. 1895: 262-264; Pereira, F. A. 1898: 231-238, 289-303; Vasconcelos, J. L. 1900: 185

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

16011601, GRADE, Grade, Arcos de Valdevez, Viana do Castelo (160116). CMP 16. Typology: Funerary epigraphy. Chronology: Roman Period. References: Pereira, F. A. 1904: 74-81; Alarcão, J. 1988: 1/49.

A. 1929-68: 50, 109; Pinto, R. S. 1932: 81-91; Cardoso, M. 1976. 6020103, FORTE DO CÃO, Âncora (Stª Maria), Caminha, Viana do Castelo (160201). [510569.21, 4627721.92, 5]. Typology: Salt pans. Chronology: Iron Age; Roman Period. References: Lemos, F.S. 1982: 21-48; Almeida, C.A.B. 1996: 48-49.

16013501, TAVAREZ, Padreiro (Salvador), Arcos de Valdevez, Viana do Castelo (160135). CMP 28. Typology: Roman necropolis? Chronology: Roman Period. . References: Brito, M. J. C. 1933: 83 e 92

16020102, CÚTERO, Ancora (Stª Maria); Afife, Caminha, Viana do Castelo (160201). [513180.74, 4625752.42, 270]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Figueiredo, A.M. 1895: 143; Viana, A. 1932: 154-167; Corrêa, M. 1928: 143; Sarmento, F. 1933: 63-67, 171; Meira, A. 1945; Paço, A. e A. Q. Paço, 1956: 172-173; Cardozo, M. et alii 1959: 522-546; Hawkes, C.F.C. 1971: 283-286; Alves. L. 1980: 105-141; Moreira, M. 1982: 31-96; Hawkes, C.F.C. 1984; Silva, A.C.F. 1986; Queiroga, F. 1992; Almeida, C.A.B. 1996: 13-15; Viana, A. 247-270.

16012801, PELOURINHO, Prozelo, Arcos de Valdevez, Viana do Castelo (160128). CMP 16. Typology: Roman necropolis. Chronology: Roman Period. References: Pereira, F. A. 1923-1924: 273-274; Alarcão, J. 1988: 1/47. 16013103, IGREJA, Rio de Moinhos, Arcos de Valdevez, Viana do Castelo (160131). CMP 16 [545552.78, 4637842.47, 0]. Typology: Roman necropolis? Chronology: Roman Period. References: Pereira, F. A. 1924: 258. 16013102, ANTRE-OS CASTROS, Rio de Moinhos, Arcos de Valdevez, Viana do Castelo (160131). CMP 16. Typology: Hill-fort. Chronology: Iron Age. References: Vasconcelos, J.L. 1900: 38; Pereira, F. A. 1903: 58; Pereira, F. A. 1924: 260-261; Viana, A. 1926: 111.

16020101, CIVIDADE DE ÂNCORA (Monte da Subidade de Âncora; Cividade de Afife), Âncora (Stª Maria); Afife, Caminha, Viana do Castelo (160201). [512210.11, 4626367.52, 187]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Sarmento, F.M. 1880; 1933, p. 63-67; Sarmento, F.M. 1883-1884: 9-11, 17-19, 25-26, F.M. 1933, p. 165-172; Viana, A. 1926: 88-90 ; Viana, A 1936: 111113; Viana, A. 1932: 154-167; Cardozo, M. 1959: 522546; Viana, A. 1960-1961: 247-270; Viana, A. 1963: 167-178; Hawkes, C.F.C. 1971: 283-286; Hawkes, C.F.C 1984; Silva, A.C.F. 1986; Queiroga, F. 1992; Almeida, C.A.B. 1996: 8-13.

16013101, CASTRO DE REBOREDA (Castro de Anhó), Rio de Moinhos (Santa Vaia), Arcos de Valdevez, Viana do Castelo (160131). CMP 16 [544252.78, 4636929.65, 300]. Typology: Romanized Hill-fort; Votive epigraphy. Chronology: Roman Period. References: Pereira, F. A. 26: 257-260; Vasconcelos, J.L. 1924: 258; Viana, A. 1926: 111; Lambrinho, S. 1963-1964: 125-127; Encarnação, J. 1975: 156-157; Tranoy, A. 1981: 271-272; Alarcão, J. 1988: 1/44, 45.

16020602, PAÇO, Azevedo, Caminha, Viana do Castelo (160206). [515428.93, 4632295.30, 80]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 19.

16013001, CASTRO DE RIO FRIO, Rio Frio, Arcos de Valdevez, Viana do Castelo (160130). CMP 16 [543685.97, 4634624.65, 0]. Typology: Romanized Hill-fort? Chronology: Iron Age. References: Peixoto, R. 1898: 332; Vasconcellos, J. L.1900: 38.

16020601, BARROCAS, Azevedo, Caminha, Viana do Castelo (160206). [516097.08, 4632605.19, 60]. Typology: Hill-fort. Chronology: Iron Age. References: Vieira, J.A. 1886: 176; Almeida, C.A.B. 1996: 17-18.

16014401, RAMALHOSA, Senharei, Arcos de Valdevez, Viana do Castelo (160144). Typology: Roman treasure. Chronology: Roman Period. References: Hipólito, M. C. 1960-1961: 14-15; Alarcão, J. 1988: 1/43.

16020801, CIVIDADE, Cristelo, Caminha, Viana do Castelo (160208). [513096.30, 4634017.76, 65]. Typology: Farm. Chronology: Roman Period. References: Viana, A., 1926: 88-90, 111-113; Silva, A.C.F. 1986; Almeida, C.A.B. 1996: 19-21; Almeida, C.A.B. 2003.

16014901, COTO DA PENA (Alto de Penacova; Penedos Grandes), Vale, Arcos de Valdevez, Viana do Castelo (160149). CMP 16 [550892.98, 4633697.87, 0]. Typology: Hill-fort. Chronology: Iron Age. References: Pereira, F. A. 1902: 97, 197; Pereira, F. A. 1903: 56-58; Cardoso, M. 1903: 270-271; Cardoso, M. 1905-8: 322, 499, 500; Cardoso, M. 1915: 130-131, 231-233, 244, 247; Pereira, F. A. 1915: 224-258; Fontes, J. 1916: 5; Pereira, F. A. 1924: 256; Pereira, F.

16020901, BOUCINHA DO CASTRO (Castro do Germano), Dem, Caminha, Viana do Castelo (160209). [519419.65, 4632058.32, 279]. Typology: Hill-fort. Chronology: Iron Age; Roman Period. References: Alves, L. 1980: 105-141; Alves, L. et alii 1985: 255; Silva, A.C.F. 1986; Queiroga, F. 1992; Almeida, C.A.B. 1996: 21-23.

186

Catalogue

16021001, IGREJA, Gondar, Caminha, Viana do Castelo (160210). [518318.01, 4629927.34, 70]. Typology: Votive epigraphy. Chronology: Roman Period. References: CIL II 2463; Tranoy, A. 1981: 314; Almeida, C.A.B. 1996: 23-24.

16021801, Nª SRª DO CRASTO, Vilar de Mouros, Caminha, Viana do Castelo (160218). CMP 14 [516889.65, 4638775.52, 130]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Viana, A. 1932: 154-167; Sarmento, F.M. 1933: 63; Viana, A. 1955: 25-115; Kalb, Ph. 1980: 25115; Silva, A.C.F. 1986; Queiroga, F. 1992; Almeida, C.A.B. 1996: 43-44.

16021101, MONTE GÓIS, Lanhelas, Caminha, Viana do Castelo (160211). [518638.09, 4639859.21, 344]. Typology: Romanized Hill-fort. Chronology: Bronze Age; Iron Age; Roman Period. References: Sarmento, F. 1933: 166; Kalb, Ph. 1980: 25-115; Silva, A.C.F. 1986; Almeida, C.A.B. 1996: 24-27. 16021201, BARROCA, Moledo, Caminha, Viana do Castelo (160212). [512706.01, 4633060.97, 90]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Almeida, C.A.B. 1996: 2728.

16021901, ALTO DO COTO DA PENA, Vilarelho, Caminha, Viana do Castelo (160219). [513924.29, 4635098.79, 79]. Typology: Romanized Hill-fort. Chronology: Bronze Age; Iron Age; Roman Period; Middle Ages. References: Fortes, J. 1902: 102-104; Viana, A. 1932: 154-167; Alves, L. et alii 1985: 441; Silva, A.C.F. 1986: 29 e 38; Martins, M. 1990: 125; Jorge, S.O. 1990: 285; 292; Queiroga, F. 1992; Calo Lourido, F. 1993: 104; Almeida, C.A.B. 1996: 44-48.

16021302, CABANAS, Orbacém, Caminha, Viana do Castelo (160213). [519521.79, 4628450.14, 140]. Typology: Small Farm? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 31.

16030301, IGREJA, Chaviães, Melgaço, Viana do Castelo (160303). CMP 1. Typology: Roman habitat. Chronology: Roman Period. References: Nunes, H.B., Oliveira, E.P. 1978: 24.

16021301, IGREJA, Orbacém, Caminha, Viana do Castelo (160213). [519429.06, 4628604.09, 100]. Typology: Votive epigraphy. Chronology: Roman Period. References: CIL II Nº 5248; Pereira, F. 1907: 36-52; Le Roux, P.; Tranoy, A. 1973: 251-256; Tranoy, A. 1981: 322-323; Almeida, C.A.B. 1996: 28-31.

16031101, PADERNE, Paderne, Melgaço, Viana do Castelo (160311). CMP 4 [560121.53, 4660183.12, 0]. Typology: Roman epigraphyundetermined. Chronology: Roman Period. References: Vasconcellos, J. L. 1907: 275-281.

16021402, MONTE DE STº AMARO, Riba de Ancora, Caminha, Viana do Castelo (160214). [514581.87, 4629209.37, 100]. Typology: Hill-fort. Chronology: Iron Age. References: Sarmento, F. 1933: 63; Silva, A.C.F. 1986; Queiroga, F. 1992; Almeida, C.A.B. 1996: 33-34.

16031102, CIVIDADE DE PADERNE (Couto Grande; Castro da Cividade de Paderne), Paderne, Melgaço, Viana do Castelo (160311). CMP 4 [559515.67, 4661150.84, 159]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Vasconcellos, J.L. 1993: 31-45; Viana, A. 1926: 89; Lópes Cuevillas, F. Fernandez, J. 1938: 100; Pereira, F.A. 1941: 37.

16021401, PICOTO DOS MOUROS (Corucho dos Mouros), Riba de Ancora; Vile, Caminha, Viana do Castelo (160214). [514740.20, 4630751.77, 302]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Almeida, C.A.B. 1996: 3233.

16040101, PENHA DA RAINHA (Penha da Rainha, Penaguda), Abedim, Monção, Viana do Castelo (160401). CMP 8 [540060.41, 4648892.1, 472]. Typology: Romanized Hill-fort? Medieval castle. Chronology: Iron Age; Roman Period; Middle Ages. References: Marques, J. A. T. M. 1985: 84; Guerra, L. F. 1899-1900: 134-136.

16021501, CRASTO, Seixas, Caminha, Viana do Castelo (160215). [515257.90, 4636766.97, 35]. Typology: Hill-fort. Chronology: Iron Age; Roman Period. References: Viana, A. 1932: 154-167; Alves, L. et alii 1985: 157-159; Almeida, C.A.B. 1996: 34-35; Almeida, C.A.B. 2003.

16040201, CIVIDADE DE ANHÕES (Monte da Cividade, Lugar da Cividade), Anhões, Monção, Viana do Castelo (160402). CMP 8 [546157.208, 4649083.37, 453]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Viana, A. 1926: 89; Viana, A. 1932: 15; Marques, J. A. T. M. 1985: 44.

16021601, IGREJA, Venade, Caminha, Viana do Castelo (160216). [515471.10, 4634115.05, 33]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1996: 347.

16040301, CASTRO DA SENHORA DA GRAÇA (Muro, Torre, Outeiro (S e SO); Eira dos Mouros (NE), Badim; Sá, Monção, Viana do Castelo (160403160425). [552816.33, 4657149.36, 315]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Marques, J. A. M. 1984a; Marques, J. A. M. 1984c; Marques, J. A. T. M. 1985: 49.

16021701, FORTE DA LAGARTEIRA, Vila Praia de Âncora, Caminha, Viana do Castelo (160217). [510912.65, 4629511.37, 5]. Typology: Salt pans. Chronology: Iron Age; Roman Period? References: Almeida, C.A.B. 1996: 37-38. 187

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

16040401, CASTRO DA ASSUNÇÃO (Lugar do Crasto; Buraca da Moura), Barbeita, Monção, Viana do Castelo (160404). CMP 3 [549120.18, 4656505.85, 265]. Typology: Romanized Hill-fort. Chronology: References: Marques 1984c; Marques 1985: 51-55, 111-136.

Chronology: Iron Age; Roman Period. References: Marques, J. A. T. M. 1985: 98. 16042402, QUINTENDA (Monte da Ladeira, Cótoros), Ribeira de Mouro, Monção, Viana do Castelo (160424). CMP 3 [554622.18, 4655466.61, 499,5]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Marques, J. A. T. M. 1985: 96.

16040701, CASTRO DE CAMBESES (Monte Crasto; Srª da Boa Morte, Cambeses, Monção, Viana do Castelo (160407). CMP 3 [544626.38, 4654656.29, 247]. Typology: Hill-fort. Chronology: Iron Age. References: Vasconcelos, J.L. 1903: 287; Viana, A. 1926: 89; Marques, A.M. 1985: 47-48.

16042403, COROA (Coroa, Fonte Boa, Cavenca, Calçadinha), Ribeira de Mouro, Monção, Viana do Castelo (160424). CMP 4 [557062.17, 4652247.17, 925]. Typology: Hill-fort. Chronology: Iron Age. References: Marques, J. A. T. M. 1985: 94.

16041001, CASTRO DE LARA (Crasto; Paços; Aldeia), Lara, Monção, Viana do Castelo (160410). CMP 2 [539410.82, 4653977.63, 209]. Typology: Hillfort? Chronology: Iron Age; Roman Period? References: Marques 1985: 2-83.

16042404, CIVIDADE DE RIBA DE MOURO (Lugar da Cividade, Veiga), Ribeira de Mouro, Monção, Viana do Castelo (160424). CMP 3 [555319.09, 4654546.69, 259]. Typology: Romanized Hill-fort. Chronology: Iron Age. References: Marques, J. A. T. M. 1985: 58.

16041101, CASTRO DE SÃO CAETANO, Longos Vales, Monção, Viana do Castelo (160411). CMP 3 [546145.89, 4654172.65, 342]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Marques, J. A. T. M. 1985: 56; Viana, A. 1926: 89; Neves, L. Q. e Viana, A. 1959: 389-398.

16042601, MONTE DAS BARROCAS (Barrocas, Paço), Sago, Monção, Viana do Castelo (160426). CMP 3 [545897.14, 4653708.38, 352]. Typology: Romanized Hill-fort; Watchtower? Chronology: Iron Age; Roman Period. References: Marques, J. A. T. M. 1985: 78.

16041401, CORTES, Mazedo, Monção, Viana do Castelo (160414). Typology: Roman necropolis. Chronology: Roman Period. References: Marques, 1985: 142-147.

16042801, CRASTÊLO (Crastêlo, Paço), Tangil, Monção, Viana do Castelo (160428). CMP 3 [553261.36, 4652988.69, 127]. Typology: Hill-fort? Watchtower? Chronology: Iron Age; Roman Period. References: Marques, J. A. T. M. 1985: 80; Viana, A. 1926: 89.

16041501, BUSTAVADE (Muro, Bustavade), Merufe, Monção, Viana do Castelo (160415). CMP 3 [547976.09, 4652334.05, 734754]. Typology: Hillfort? Chronology: Iron Age. References: Marques, J. A. T. M. 1985: 90.

16042802, CASTRO DA Sª DA VISTA (Souto), Tangil; Podame, Monção, Viana do Castelo (160428160422). CMP 3 [551882.63, 4652824.24, 244]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Marques, J. A. T. M. 1985: 62-63; Marques, J. A. M. 1984c.

16041701, REIRIZ, MONÇÃO, Monção, Monção, Viana do Castelo (160417). Typology: Votive epigraphy; Roman habitat. Chronology: Roman Period. References: Marques, J. A. T. M. 1985: 138-140; Cortez, F. R. 1945: 12-15.

16042901, TROPORIZ, Troporiz, Monção, Viana do Castelo (160429). Typology: Roman bridge. Chronology: Roman Period. References: Almeida, C. A. F. 1968: 42; Marques, J. A. T. M. 1985: 140-141.

16041801, CASTRO DO OUTEIRO DA TORRE (Cidade, Paço, Torre), Moreira, Monção, Viana do Castelo (160418). CMP 3 [542457.12, 4652144.61, 131]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Marques, J. A. T. M. 1985: 86.

16043001, MONTE REDONDO (Monte Redondo, Mourisco), Troviscoso, Monção, Viana do Castelo (160430). CMP 3 [544260.47, 4658015.98, 73]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Marques, J. A. M. 1984c; Marques, J. A. T. M. 1985: 71.

16042001, MONTE DOS PENEDOS (Lapa; Eido), Pias, Monção, Viana do Castelo (160420). CMP 8 [541315.21, 4650749.91, 198]. Typology: Hill-fort. Chronology: Iron Age. References: Marques, A.M. 1985: 73-77.

16043002, CRISTÊLO, Troviscoso, Monção, Viana do Castelo (160430). CMP 3 [543430.21, 4990905.55, 78]. Typology: Hill-fort. Chronology: Iron Age; Roman Period? References: Marques, J. A. T. M. 1985: 69; Viana, A. 1926: 89; Viana, A. 1932: 21.

16042401, COTOS AGUÇADOS (Paço, Portela do Marco, Costa da Ponte.), Ribeira de Mouro, Monção, Viana do Castelo (160424). CMP 8 [552949.78, 4648513.94, 585]. Typology: Romanized Hill-fort.

16043101, COUTO DA TRUTE (S. Paio, Portal, Campo Redondo), Trute, Monção, Viana do Castelo (160431). CMP 8 [544311.58, 4649842.69, 425]. Ty188

Catalogue

pology: Hill-fort. Chronology: Iron Age. References: Marques, J. A. T. M.1985: 88.

Cunha, N. A. 1979: 398; Almeida, CA.B. 1979; Silva, A.C.F. 1986: 55; Queiroga, F. 1992: 318; Silva, M.F.M. 1992: 37; Silva, M.F.M. 1994: 29; Almeida, C.A.B. 1996: 15-16.

16040202, CASTRO DO MENDOIRO (Monte Redondo; Lameiras; Furnas; Cotinho), Anhões, Monção, Viana do Castelo (160402). CMP 8 [548878.77, 4648176.42, 794]. Typology: Hill-fort. Chronology: Iron Age. References: Moita: 227; Marques, 1995 : 92-93.

16050602, MONTE DE S. SEBASTIÃO (Monte do Crasto; Castro de Cristelo), Cristelo, Paredes de Coura, Viana do Castelo (160506). CMP 15 [538314.63, 4639783.77, 493]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Santos, L.A.; Viana, A. 1926: 88-90, 111-113; Viana, A. 1932: 11-24; Oliveira, A.C. 1976: 23; Cunha, N.A. 1979: 390; Le Roux, P., Tranoy, A. 1983: 16; Silva, A.C.F. 1986: 32; Silva, M.F.M, 1992: 37-45; Queiroga, F. 1992: 321; Silva, M, 1994: 482; Almeida, C.A.B. 1996: 17-20.

16050101, PESO, Agualonga, Paredes de Coura, Viana do Castelo (160501). CMP f. 15 [531644.00, 4637283.58, 250]. Typology: Small Farm. Chronology: Roman Period. References: Silva, M.F.M. 1994: 28; Almeida, C.A.B. 1996: 10. 16050205, TELHADO, Bico, Paredes de Coura, Viana do Castelo (160502). [540165.49, 4638436.90, 561]. Typology: Small Farm. Chronology: Roman Period. References: Azevedo, P.A. 1897: 309; Cunha, N.A. 1979: 379; Almeida, C.A.B. 1996: 11.

16050601, CRISTELO, Cristelo, Paredes de Coura, Viana do Castelo (160506). CMP 15 [538536.64, 4639670.81, 380]. Typology: Votive epigraphy. Chronology: Roman Period. References: Rosário, A. 1973; Santos, L.A., Le Roux P., Tranoy, A. 1983: 194195; Alarcão, J. 1988: 1/32; Garcia, J.M. 1991: 574; Silva, M.F.M, 1992: 43; Almeida, C.A.B. 1996: 20-21.

16050204, IGREJA, Bico, Paredes de Coura, Viana do Castelo (160504). CMP 15 [515]. Typology: Medieval habitat? Chronology: Middle Ages? References: Almeida, C.A.B. 1996: 14.

16050603, SOUTELO / MÓ, Cristelo / Parada, Paredes de Coura, Viana do Castelo (160506). [538958.06, 4640095.70, 420]. Typology: Small Farm. Chronology: Roman Period. References: Silva, M.F.M. 1992: 38; Almeida, C.A.B. 1996: 22.

16050201, CHELA (Crasto), Bico, Paredes de Coura, Viana do Castelo (160502). CMP 15 [539821.90, 4638064.85, 540]. Typology: Hill-fort. Chronology: Iron Age; Roman Period. References: Alarcão, J. 1988: 1/38; Oliveira, A.L. 1976: 18; Silva, M.F.M. 1994: 28; Almeida, C.A.B. 1996: 12-13.

16050704, PAÇO, Cunha, Paredes de Coura, Viana do Castelo (160507). [533324.77, 4637661.36, 320]. Typology: Roman habitat. Chronology: Roman Period; Middle Ages? References: Almeida, C.A.B. 1996: 28.

16050202, ESTEVE, Bico, Paredes de Coura, Viana do Castelo (160502). CMP 15 [539217.44, 4639017.62, 460]. Typology: Small Farm. Chronology: Roman Period. References: Alarcão, J. 1988: 1/36; Silva, M. F. M. 1994: 28-29; Almeida, C.A.B. 1996: 14.

16050702, MONTEZELO (Ventuzelo; Rechio da Cidade; Cidade), Cunha, Paredes de Coura, Viana do Castelo (160507). CMP 15 [534251.73, 4638327.89, 465]. Typology: Hill-fort. Chronology: Iron Age; Roman Period? References: Azevedo, P.A. 1897: 240; Cunha, N.A. 1979: 24; Oliveira, A.L. 1976: 24-25; Silva, A.C.F. 1986; Queiroga, F. 1992; Silva, M.F.M. 1994: 29; Almeida, C.A.B. 1996: 23-25; Almeida, C.A.B. 2003.

16050203, CASCALHAL, Bico, Paredes de Coura, Viana do Castelo (160502). CMP 15 [539881.01, 4639854.07, 540]. Typology: Roman habitat? Chronology: Roman Period. References: Alarcão, J. 1988: 1/42; Silva, M. F. M. 1994: 28; Almeida, C.A.B. 1996: 13-14.

16050703, IGREJA, Cunha, Paredes de Coura, Viana do Castelo (160507). [533830.04, 4638033.85, 393]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 25-26.

16050402, IGREJA, Cossourado, Paredes de Coura, Viana do Castelo (160504). [530453.11, 4640886.94, 230]. Typology: Mutatio. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 17.

16050701, LISOUROS-CAPELA DE SANTO ESTEVÃO, Cunha, Paredes de Coura, Viana do Castelo (160507). CMP 15 [536046.64, 4637211.98, 490]. Typology: Votive epigraphy. Chronology: Roman Period. References: CIL II 5069; Vasconcelos, J.L. 1913: 196-199; Tovar, A., Navascués, J.M. 1950: 178-191; Blazquéz Martinez, J.M. 1962; Encarnação, J. 1975; Cunha, N.A. 1979: 136-138; Tranoy, A. 1981: 175; Garcia, J.M. 1991: 364; H.E. 1994; Almeida, C.A.B. 1996: 26-28.

16050401, FORTE DA CIDADE (Monte da Cividade; Alto da Cidade; Crasto), Cossourado, Paredes de Coura, Viana do Castelo (160504). CMP 7 [530200.89, 4640608.29, 376]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Argote, J.C. 1732: 610; Leal, P. 1884: 270; Vieira, J.A. 1886: 122; Capella, M. 1895; Pereira, F.A. 1924: 281; Viana, A. 1926: 88-90; Viana, A. 1932: 11-24; Sarmento, M. 1933: 84; Cunha, A.R. 1956; Viana, A. 1960-61; Cunha, A. 1975; Oliveira, A.L. 1976: 21-22; 189

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

16050804, LOURIDO (Barrosas; Côto da Bouça; Fornos), Ferreira, Paredes de Coura, Viana do Castelo (160508). CMP 7 [534813.68, 4644237.97, 520]. Typology: Small Farm. Chronology: Roman Period. References: Cunha, N.A. 1979: 473; Silva, M.F.M. 1994: 30; Almeida, C.A.B. 1996: 32-33.

Viana 1926: 89; Viana 1932: 161; Cunha 1909 [1979]: 131; Silva 1986: 71; Alarcão, J. 1988a: 18; Silva, M.F.M. 1994: 127; Almeida, C.A.B. 1996: 33-34. 16050905, REIRIGO, Formariz, Paredes de Coura, Viana do Castelo (160509). [536593.80, 4642858.99, 520]. Typology: Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 39-40; Almeida, C.A.B. 2003.

16050801, MOIMENTA, Ferreira, Paredes de Coura, Viana do Castelo (160508). CMP 7 [533832.40, 4644237.97, 305]. Typology: Farm; Roman necropolis. Chronology: Roman Period. References: Vieira, J.A. 1886: 124; Cunha, N.A. 1979: 121-125; Alarcão, J. 1988: 1/16; Silva, M.F.M, 1994: 29; Almeida, C.A.B. 1996: 31-32.

16051003, PAÇO DO OUTEIRO, Infesta, Paredes de Coura, Viana do Castelo (160510). [533543.61, 4640129.79, 317]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 40.

16050803, BOUÇA DO CROAS, Ferreira, Paredes de Coura, Viana do Castelo (160508). CMP 7 [536566.00, 4643784.13, 500]. Typology: Small Farm? Chronology: Roman Period. References: Cunha, N.A. 1979: 134-135; Alarcão, J. 1988: 1/15; Silva, M.F.M, 1994: 30; Almeida, C.A.B. 1996: 30-31.

16051102, CIDADE (Paço), Insalde, Paredes de Coura, Viana do Castelo (160511). [538911.38, 4644413.58, 500]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 44-45. 16051101, CASTELINHO, Insalde, Paredes de Coura, Viana do Castelo (160511). CMP 7 [539881.03, 4643956.18, 513]. Typology: Hill-fort. Chronology: Iron Age; Roman Period?. References: Vieira 1886:128; Viana 1926: 89; Cunha 1909 [1979]: 489; Oliveira 1976: 30 e 32; Silva 1986: 71; Queiroga 1992; Silva, M.F.M. 1994: 483; Almeida, C.A.B. 1996: 4244; Almeida, C.A.B. 2003.

16050805, MONTE DA MADORRA (Forte da Madorra; Alto da Madorra), Ferreira / Linhares, Paredes de Coura, Viana do Castelo (160808). CMP 7 [532771.52, 4642686.12, 351]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Viana, A. 1926: 88-90, 111-113; Viana, A. 1932: 11-24; Oliveira, A.L. 1976: 23-24; Silva 1986: 71; Queiroga, F. 1992; Silva, M.F.M. 1994: 483; Almeida, C.A.B. 1996: 29-30.

16051203, CASTRO DE BORZENDES (Castro de Brutendes), Linhares, Paredes de Coura, Viana do Castelo (160512). CMP 7 [532937.03, 4641761.61, 336]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Leal, P. 1874: 170; Vieira, J.A. 1886: 123; Viana, A. 1926: 88-90, 111-113; Viana, A. 1932: 11-24; Cunha, N.A. 1979: 89 e 497; Oliveira, A.L. 1976: 23-25; Silva, A.C.F. 1986; Queiroga, F. 1992; Silva, M.F.M. 1994: 483; Silva, M. F. M. 1994; Almeida, C.A.B. 1996: 45-46.

16050904, MONTE DO CRASTO (Monte de Venade; Monte da Giesteira), Formariz, Paredes de Coura, Viana do Castelo (160509). CMP 7 [536544.56, 4643475.59, 603]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Leal, P. 1874: 215; Vieira 1886: 125; Viana, A. 1926: 88-90, 11-113; Viana, A. 1932: 11-24; Cunha, N.A. 1979: 135; Silva, A.C.F. 1986; Queiroga, F. 1992; Silva, M.F.M. 1994: 482; Almeida, C.A.B. 1996: 34-37.

16051201, CRASTINHO, Linhares, Paredes de Coura, Viana do Castelo (160512). [532198.70, 4642035.81, 285]. Typology: Hill-fort. Chronology: Iron Age; Roman Period. References: Almeida, C.A.B. 1996: 46-47.

16050901, Nª SRª DO LIVRAMENTO, Formariz, Paredes de Coura, Viana do Castelo (160509). CMP 7 [534418.25, 4641628.42, 322]. Typology: Villa? Chronology: Roman Period. References: Oliveira, A.L. 1976: 97; Cunha, N. A. 1979: 456; Alarcão 1988; Silva, M.F.M. 1994: 30; Almeida, C.A.B. 1996: 37-38; Alarcão, J. 1988: 1/21.

16051202, CABECEIRAS / VINHAS, Linhares, Paredes de Coura, Viana do Castelo (160512). CMP 7 [532199.67, 4641819.92, 250]. Typology: Small Farm? Chronology: Roman Period. References: Alarcão, J. 1988: 17, 22; Silva, M.F.M. 1994: 32; Almeida, C.A.B. 1996: 48-49.

16050902, PONTE DE MANTELÃES (Regato dos Freis), Formariz, Paredes de Coura, Viana do Castelo (160509). CMP 15 [535405.76, 4640940.72, 310]. Typology: Roman treasure. Chronology: Roman Period. References: Cunha 1979: 12; Oliveira 1976: 97; Alarcão, J. 1988a : 23; Silva, M.F.M, 1994: 31; Almeida, C.A.B. 1996: 38.

16051304, AFE, Moselos, Paredes de Coura, Viana do Castelo (160513). CMP 7 [536532.31, 4641378.22, 360]. Typology: Roman habitat. Chronology: Roman Period. Viana, A. 1932: 11-24; Oliveira, A.L. 1976: 23-24; Silva, M.F.M. 1994: 32; Almeida, C.A.B. 1996: 49-50.

16050903, CIDADE (Castelo), Formariz, Paredes de Coura, Viana do Castelo (160509). CMP 7 [535677.70, 4641836.52, 380]. Typology: Hill-fort. Chronology: Iron Age. References: Pereira, F.A. 1903: 133-134;

16051301, MOSELOS-IGREJA, Moselos, Paredes de Coura, Viana do Castelo (160513). CMP 7 [536915.09, 190

Catalogue

4641853.63, 380]. Typology: Roman habitat? Chronology: Roman Period. References: Alarcão, J. 1988: 1/20; Silva, M.F.M. 1994: 32.

16052010, RUBIÃES, Rubiães, Paredes de Coura, Viana do Castelo (160520). Typology: Milestone. Chronology: Roman Period. References: Rosário, A., 1973: 29; Almeida., C.A.B. 1979:120-124; 146-147; Santos, L.A. 1979: 21-26; Araújo, J.R. 1982: 200-201; Santos, L.A., Le Roux, P., Tranoy, A. 1983: 200; Alarcão, J. 1986: 1/40; Silva, M. F. M. 1994: 33; Almeida, C.A.B. 1996: 74-75.

16051302, FAVAIS, (Cruzes), Moselos, Paredes de Coura, Viana do Castelo (160513). CMP 7 [380]. Typology: Roman habitat? Chronology: Roman Period. References: Silva, M. F. M. 1994: 32; Almeida, C.A.B. 1996: 50-51.

16052006, BOUÇA DO LIMA-ANTAS, Rubiães, Paredes de Coura, Viana do Castelo (160520). CMP 15 [528525.60, 4638997.57, 150]. Typology: Pottery Kiln; Small Farm. Chronology: Roman Period. References: Alarcão, J. 1986: 1/29; Silva, M. F. M. 1994: 33; Almeida, C.A.B. 1996: 76-77.

16051602, PADORNELO, Paredes de Coura, Paredes de Coura, Viana do Castelo (160516). [537936.12, 4641632.31, 360]. Typology: Roman habitat?. Chronology: Roman Period? References: Almeida, C.A.B. 1996: 51. 16051701, CERDEIRAIS, (Cerqueiras), Porreiras, Paredes de Coura, Viana do Castelo (160517). CMP 7 [537391.40, 4644466.97, 514]. Typology: Small Farm. Chronology: Roman Period. References: Alarcão, J. 1988: 14; Silva, M.F.M. 1994: 32; Almeida, C.A.B. 1996: 51-52.

16052002, S. BARTOLOMEU DAS ANTAS/Rubiães, Paredes de Coura, Viana do Castelo (160520). CMP 15 [529443.82, 4639278.87, 200]. Typology: Milestones. Chronology: Roman Period. References: Almeida, A. 1866; Costa, A.C. 1868: 232; CIL II 6225; Capella, M. 1895; Bellino, A. 1895: 82; Brandão, C. 1904; Sarmento, F. M. 1933: 83-85; Melo, M.F.S. 1967: 124; Cunha, N.A. 1979: 101-108; Almeida, C.A.B. 1979: 93-94; Santos, L.A. 1979: 32; Tranoy, A. 1981: 210216; Alarcão, J. 1986: 1/28; Almeida, C.A.B. 1996: 6674.

16051801, IGREJA, Resende, Paredes de Coura, Viana do Castelo (160518). CMP 15 [536887.75, 4639467.80, 488]. Typology: Roman habitat; Small Farm? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 52.

16052003, IGREJA DE RUBIÃES, Rubiães, Paredes de Coura, Viana do Castelo (160520). CMP 15 [531250.07, 4638597.41, 220]. Typology: Funerary epigraphy; Milestone. Chronology: 213¬217, Roman Period. References: Capella, M. 1895: 143-144; Capella, M. 1903: 132; Capella, M. 1920: 241-270; Melo, M.F.S. 1967; Almeida, C.A.F. 1978: Santos, L.A. 1979: 10; Araújo, J.R. 1982: 218-221; Almeida, C.A.B. 1979: 119; Alarcão, J. 1988: 1/37; Almeida, C.A.B. 1996: 63-65.

16051904, IGREJA, Romarigães, Paredes de Coura, Viana do Castelo (160519). [531099.60, 4635276.44, 235]. Typology: Roman habitat? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 58. 16051903, CRASTO, Romarigães, Paredes de Coura, Viana do Castelo (160519). [530754.72, 4635059.04, 249]. Typology: Roman habitat? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 57-58. 16051901, ALTO DA CIDADE, (Couto d' Ouro; Penedo do Curral das Éguas; Poço do Ouro; Cividade), Romarigães, Paredes de Coura, Viana do Castelo (160519). CMP 15 [530520.60, 4635890.78, 280]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Costa, A.C. 1868: 232; Leal, P. 1878: 242; Vieira, J.A. 1886: 137; Viana, A. 1926: 88-90, 111-113; Viana, A. 1932: 11-24; Oliveira, A.L. 1976: 42-43, 96; Cunha, N.A. 1979: 133; Silva, F.M. 1992: 37; Silva, F.M. 1993: 39; Silva, M. F.M. 1994: 281-302; Silva, M.F.M. 1994: 483; Almeida, C.A.B. 1996: 53-57.

16052004, PONTE VELHA DE RUBIÃES, Rubiães, Paredes de Coura, Viana do Castelo (160520). CMP 15 [531142.02, 4639396.03, 160]. Typology: Roman bridge? Chronology: Roman Period. References: Brandão, M.J.C. 1904; Pereira, F.A. 1923-1924: 281; Almeida 1968: 35; Alarcão, J. 1988: 1/33; Silva, M. F. M. 1994: 34; Almeida, C.A.B. 1996: 77-79.

16052009, ANTAS, Rubiães, Paredes de Coura, Viana do Castelo (160520). [529215.86, 4639247.11, 200]. Typology: Milestone. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 59.

16052005, CRASTO, Rubiães, Paredes de Coura, Viana do Castelo (160520). CMP 15 [530668.28, 4639068.17, 179]. Typology: Milestones. Chronology: Roman Period. References: Viana, A. 1926: 88-90, 111-113; Cunha, N.A. 1979: 559; Almeida, C.A.B., 1979: 120-121; Santos, L.A., 1979: 8; Tranoy, A., 1981: 210-216; Araújo, J.R., 1982: 222-225; Alarcão, 1988: 1/33; Silva, M. F. M., 1994: 34; Almeida, C.A.B. 1996, VI: 59-62.

16052007, VILA CHÃ, Rubiães, Paredes de Coura, Viana do Castelo (160520). CMP 15 [530551.21, 4639499.46, 222]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 7576.

16050505, EIRA DA MOURA-LOUSADO (Castro de S. Martinho), S. Martinho de Coura, Paredes de Coura, Viana do Castelo (160505). CMP 15 [527983.37, 4636250.46, 325]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: 191

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Oliveira, A.L. 1976: 22-23; Cunha, N.A. 1979: 131; Silva, A.C.F. 1986; Queiroga, F. 1992; Silva, M.F.M. 1994: 29; Almeida, C.A.B. 1996: 81-82.

Chronology: Roman Period. References: Peixoto, R. 1903: 807; Marçal, H. 1958: 706; Baptista, A. M. 1985: 14-15.

16050504, PORTELA DA BUSTARENGA, S. Martinho de Coura, Paredes de Coura, Viana do Castelo (160505). CMP 15 [529256.64, 4634929.37, 221]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Costa, A.C. 1868: 232; Cunha, N.A.1979: 132-133; Silva, M.F.M. 1994: 29; Silva, M.F.M. 1994: 484-485; Almeida, C.A.B. 1996, VI: 83-84.

16061205, TORRÃO DOS GALEGOS, Lindoso, Ponte da Barca, Viana do Castelo (160612). CMP 17 [567214.00, 4635564.00, 0]. Typology: Roman necropolis. Chronology: Roman Period. Fontes, L. 2012 16061204, CABEÇO DE LEIJÓ, Lindoso, Ponte da Barca, Viana do Castelo (160612). CMP 17 [565694.00, 4634595.00, 0]. Typology: vicus. Chronology: Roman Period. Fontes, L. 2012

16050501, FONTE DO OLHO, S. Martinho de Coura, Paredes de Coura, Viana do Castelo (160505). CMP 15 [529107.46, 4637581.18, 160]. Typology: Milestone. Chronology: Roman Period. References: Capella, M. 1895: 248; Cunha, N.A. 1979: 107; Santos, L.A. 1979: 9; Araújo, J.R. 1982: 226; Alarcão, J. 1988: 1/30; Silva, M.F.M. 1994: 29; Almeida, C.A.B. 1996: 84-85.

16061203, VILARINHO, Lindoso, Ponte da Barca, Viana do Castelo (160612). CMP 17 . Typology:Small Farm. Fontes, L. 2012. 16061202, CIDADELHE (Cidade; Veiga de Vão da Lage), Lindoso, Ponte da Barca, Viana do Castelo (160612). CMP 17 [562739.00, 4634750.00, 210]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Almeida, J.A. 1866: 191; Vasconcellos, J.L. 1913: 57; Fontes, L. 2012

16050502, BARREIROS, S. Martinho de Coura, Paredes de Coura, Viana do Castelo (160505). CMP 15 [529181.15, 4636471.16, 154]. Typology: Milestone. Chronology: Roman Period. References: Santos, A. L. 1979: 26-30; Almeida, C.A.B. 1979: 120; Araújo, J.R. 1982: 228-229; Alarcão, J. 1986: 1/29, 1/31; Silva, M.F.M. 1994: 29; Almeida, C.A.B. 1996: 79-81.

16061201, S. MAMEDE DE LINDOSO (Lugar do Castelo; Castelo do Lindoso), Lindoso, Ponte da Barca, Viana do Castelo (160612). CMP 17 [566437.00, 4635245.00, 0]. Typology: Votive epigraphy. Chronology: Roman Period. References: Almeida, C. A. B. 1981-1982: 167-171; Fontes, L. 2012

16051001, IGREJA (Cascalhal; Roriz), Infesta, Paredes de Coura, Viana do Castelo (160510). CMP 7 [533661.72, 4639513.49, 325]. Typology: Roman habitat. Chronology: Roman Period. References: Alarcão, J. 1988: 1/42; Silva, M.F.M. 1994: 31; Almeida, C.A.B. 1996: 40-41.

16061902, CÔTO DA ASSUREIRA, Sampriz, Ponte da Barca, Viana do Castelo (160619). CMP 29 [551314.00, 4626403.00, 360]. Typology: Hill-fort? Chronology: Iron Age.

16051002, RAPADEIRA, Infesta, Paredes de Coura, Viana do Castelo (160510). CMP 7 [533540.94, 4621069.37, 345]. Typology: Roman habitat; Small Farm? Chronology: Roman Period. References: Silva, M.F.M. 1994: 31; Alarcão, J. 1988: 1/34; Almeida, C.A.B. 1996: 33-34.

16061901, CRASTO, Sampriz, Ponte da Barca, Viana do Castelo (160619). CMP 29 [550467.00, 4626274.00, 310]. Typology: Roman treasure; Romanized Hill-fort? Chronology: Roman Period. References: Brito, P.M.J.C. 1908: 94-982; Viana, A. 1926: 111; Hipólito, M. C. 1960-1961: 15.

16060201, CRASTO, Boivães, Ponte da Barca, Viana do Castelo (160602). CMP 29 [545881.00, 4623013.00, 311]. Typology: Hill-fort. Chronology: Iron Age?

16062401, CASA DA POUSADA, Vade (S. Tomé), Ponte da Barca, Viana do Castelo (160624). CMP 29 [549367.40, 4626384.82, 0]. Typology: Funerary epigraphy. Chronology: Roman Period. References: Alarcão, J. 1988: 1/82 ; Fontes, L. 2012.

16060501, PENEDO DE S. MARTINHO, Crasto (S. Martinho), Ponte da Barca, Viana do Castelo (160605). CMP 29 [547379.36, 4625265.10, 180]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Viana, A. 1932.

16070102, IGREJA, Anais, Ponte de Lima, Viana do Castelo (160701). CMP 28 [539825.85, 4616660.48, 280]. Typology: Roman habitat?. Chronology: Roman Period? References: Almeida, C.A.B. 1996: 16-17.

16060701, CRASTO (Outeiro de S. Miguel), EntreAmbos-os-Rios, Ponte da Barca, Viana do Castelo (160607). CMP 30 [557435.00, 4630102.00, 109]. Typology: Hill-fort? Chronology: Idade de Ferro; Roman Period?

16070101, MONTE CRASTO (Rodinha do Crasto; Eirinha do Crasto), Anais / Calvelo, Ponte de Lima, Viana do Castelo (160701). CMP 28 [537862.07, 4616464.67, 365]. Typology: Hill-fort. Chronology: Bronze Age; Iron Age. References: Viana, A. 1926: 88-90, 111-113; Viana, A. 1932: 11-24; Almeida,

16060801, BILHARES, Ermida, Ponte da Barca, Viana do Castelo (160608). CMP 30 [562968.00, 4630162.00, 660]. Typology: Funerary epigraphy; Roman habitat. 192

Catalogue

C.A.B., Baptista, A.J. 1981: 287-309; Silva, A.C.F. 1986; Queiroga, F.1992; Almeida, C.A.B. 1996: 15-16.

Almeida, C.A.F. 1968; Almeida, C.A.B. 1979: 113; Almeida, C.A.B. 1990: 158-160; Alarcão, J. 1988: 1/70.

16070304, IGREJA, Arcos (S. Pedro), Ponte de Lima, Viana do Castelo (160703). CMP 28 [528910.01, 4623732.37, 50]. Typology: Medieval habitat? Chronology: Middle Ages? References: Almeida, C.A.B. 1996: 21.

16070406, CASTRO DE STº. OVÍDIO, Arcozelo / Sá, Ponte de Lima, Viana do Castelo (160704). [532918.88, 4625723.58, 248]. Typology: Romanized Hill-fort; Roman treasure¸ Medieval castle. Chronology Chalcolithic; Bronze Age; Iron Age; Roman Period; Middle Ages. References: Costa, P.A.C. 1706: 201; Oliveira M. 1905/1908; Brito, P.M.J.C. 1908: 95; Hipólito, M.C. 1960/1961: 15-16; Viana, A. 1962; Reis, A.P.M. 1980: 9; Almeida, C.A.B., Baptista A. J. 1981: 299, 303; Almeida, C.A.B. 1990: 155-157; Melo, M.F.S.: 61/62.

16070303, QUINTA DA LAGE, Arcos (S. Pedro), Ponte de Lima, Viana do Castelo (160703). CMP 28 [529349.89, 4623456.59, 20]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 20.

16070504, ESTREIGE, Ardegão, Ponte de Lima, Viana do Castelo (160705). [534190.07, 4610894.78, 140]. Typology: Medieval habitat? Chronology: Middle Ages? References: Almeida, C.A.B. 1996: 44-45.

16070302, CASTELO DA FORMIGA (Castro do Formigoso; Alto de Estorãos), Arcos (S. Pedro), Ponte de Lima, Viana do Castelo (160703). CMP 28 [528119.19, 4625178.77, 388]. Typology: Hill-fort; Medieval castle. Chronology: Iron Age; Middle Ages. References: Costa, A. C. 1706: 206; Leal, P. 1873: 233; Vieira, J.A. 1887; Pereira, F.A. 1907: 149; Cortez, F.R. 1946: 7; Melo, M.F.S. 1967: 60; Almeida, C.A.B. A.J. Baptista, 1981; Almeida, C.A.B. 1982; Silva, A.C.F. 1986; Almeida, C.A.B. 1990: 164-166; Almeida, C.A.B. 1996: 17-19.

16070503, MARACÃES, Ardegão, Ponte de Lima, Viana do Castelo (160705). [533519.49, 4610829.93, 95]. Typology: Villa? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 44. 16070502, COSTEIRA; ABELHEIRA, Ardegão, Ponte de Lima, Viana do Castelo (160705). [533267.16, 4610366.14, 100]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 43.

16070301, CASA DE PENTIEIROS, Arcos (S. Pedro) /Estorãos, Ponte de Lima, Viana do Castelo (160703). CMP 28 [528997.25, 4624997.23, 80]. Typology: Small Farm. Chronology: Roman Period. References: Melo, M.F.S. 1967: 121; Teixeira, C., Medeiros, A. C. 1972: 39; Silva, A.C.F. 1986; Almeida, C.A.B. 1990: 174; Almeida, C.A.B. 1996: 22.

16070501, EIRA VEDRA, Ardegão, Ponte de Lima, Viana do Castelo (160705). [533450.81, 4610675.41, 97]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 42-43 16070607, CRASTO, Bárrio, Ponte de Lima, Viana do Castelo (160706). [534137.31, 4631620.12, 120]. Typology: Hill-fort. Chronology: Iron Age; Roman Period? References: Viana, A. 1926: 88-90; Almeida, C.A.B., Baptista, A.J. 1981; Silva, A.C.F. 1986; Almeida, C.A.B. 1990: 153; Queiroga, F. 1992.

16070401, PONTE DA GEIRA, Arcozelo, Ponte de Lima, Viana do Castelo (160704). [534298.82, 4626809.54, 20]. Typology: Roman bridge. Chronology: Roman Period. References: Araújo, J.R. 1962: 100-102; Melo, M.F.S. 1967: 95; Reis, A.P.M. 1978: 18-19; Almeida, C.A.B. 1979: 114-115; Almeida, C.A.F. 1987: 25; Alarcão, J.1988: 1/68; Almeida, C.A.B. 1990: 161-162; Almeida, C.A.B. 1996: 41-42.

16070701, QUINTA DO PAÇO, Beiral do Lima, Ponte de Lima, Viana do Castelo (160707). [542960.99, 4606500.73, 100]. Typology: Villa. Chronology: Roman Period. References: Viana, A. 1961: 6; Melo, M.F. 1967: 141-153; Sousa, J.R. 1979: 293-304; Gonçalves, A.A.H.B. 1979: 209-225; Almeida, C.A.F. 1980: 312; Pereira, J.C. 1983: 213-224; Almeida, C.A.F. 1986; Barroca. M. 1987: 85-86; Almeida, C.A.B. 1990: 139-141; Almeida, C.A.B. 1996: 49-53.

16070402, QUINTA DO ANTEPAÇO, Arcozelo, Ponte de Lima, Viana do Castelo (160704). [534263.30, 4624588.75, 13]. Typology: Milestone. Chronology: Roman Period. References: Almeida, C.A.B. 1990: 158-160; Araújo, J.R.1982: 179-182. 16070405, CABANAS, Arcozelo, Ponte de Lima, Viana do Castelo (160704). [533250.94, 4628778.47, 70]. Typology: Small Farm. Chronology: Roman Period. References: Reis, A.P.M. 1980: 40; Almeida, C.A.B. 1990: 157-158; Almeida, C.A.B. 1996: 33.

16070702, EIDO DA RENDA, Beiral do Lima, Ponte de Lima, Viana do Castelo (160707). [542850.23, 4625005.18, 100]. Typology: Roman necropolis. Chronology: Roman Period. References: Almeida, C.A.B. 1990:133-139; Viana, A. 1961: 6; Lanhas, F. 1969: 249-260; Melo, M.F.S. 1967: 141-152; Almeida, C.A.F. 1980: 312; Gonçalves, A.A.H.B. (s/d).

16070403, PONTE DE ARCOZELO, Arcozelo / Ponte de Lima, Ponte de Lima, Viana do Castelo (160704). [534309.93, 4624496.45, 10]. Typology: Roman bridge. Chronology: Roman Period. References: Pereira, F.A. 1912; Machado, A.S. 1962; Araújo, J.R. 1962: 87-96;

16070703, CASTRO DE VINHÓ, Beiral do Lima, Ponte de Lima, Viana do Castelo (160707). [543082.97, 4624698.15, 220]. Typology: Romanized hillfort. 193

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Chronology: Iron Age; Roman Period. References: Viana, A. 1926: 60; Viana, A. 1932: 160; Almeida, C.A.F. 1975; Alarcão, A. M., Martins, A. N. 1976: 91109; Soeiro, T. 1981; Almeida, C.A.B. 1990: 124-133; Almeida, C.A.B. 1996: 47-49; Almeida, C.A.B. 2003; Soeiro, T. (s/d): 97-120.

16071301, PAÇO DE CALHEIROS, Calheiros, Ponte de Lima, Viana do Castelo (160713). [536145.06, 4626880.33, 200]. Typology: Roman habitat; Pottery Kiln. Chronology: Roman Period. References: Almeida, C.A.B. 1990: 157. 16071303, CASTELO PEQUENO, Calheiros / Brandara, Ponte de Lima, Viana do Castelo (160713). [536555.64, 4627838.51, 274]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Almeida, C.A.B. 1996: 60-61.

16071004, PAÇO DO CARDIDO, Brandara, Ponte de Lima, Viana do Castelo (160710). [536210.40, 4627651.70, 160]. Typology: Roman habitat? Chronology: Roman Period? Almeida, C.A.B. 1996: 56-57 16071003, QUINTA DO OUTEIRO, Brandara, Ponte de Lima, Viana do Castelo (160710). [535933.76, 4627588.62, 110]. Typology: Small Farm? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 55-56.

16071302, CASTELO DE GENÇO (Castelo de Calheiros), Calheiros/Brandara, Ponte de Lima, Viana do Castelo (160713160710). [536831.95, 4627963.31, 297]. Typology: Romanized Hill-fort; Watchtower? Chronology: Iron Age; Roman Period. References: Costa, A.C. 1706: 202; Calheiros, P. A. 1910:180; Melo, M.F.S.1967: 57; Almeida, C.A.F. 1971: 293296; Almeida, C.A.B., Baptista, A.J. 1981; Almeida, C.A.B. 1990: 154-155; Almeida, C.A.B. 1996: 61-64.

16071002, CANADELO, Brandara, Ponte de Lima, Viana do Castelo (160710). [535294.85, 4626105.02, 40]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1996: 55.

16071405, QUINTA DE MERECES, Calvelo, Ponte de Lima, Viana do Castelo (160714). [537688.66, 4614274.01, 170]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 80-81.

16071001, IGREJA, Brandara, Ponte de Lima, Viana do Castelo (160710). [535540.33, 4627802.55, 100]. Typology:Small Farm? Chronology: Roman Period/Baixo Império. References: Almeida, C.A.B. 1996: 54.

16071404, POUSADA, Calvelo, Ponte de Lima, Viana do Castelo (160714). [537044.73, 4613592.14, 220]. Typology: Roman habitat; Small Farm? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 7879.

16071103, PASSAIS / PAÇO, Cabaços, Ponte de Lima, Viana do Castelo (160711). [535115.27, 4615648.81, 320]. Typology: Roman habitat; Small Farm? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 59-60. 16071102, MOURELHE, Cabaços, Ponte de Lima, Viana do Castelo (160711). [535103.93, 4613243.12, 213]. Typology: Small Farm. Chronology: Roman Period. References: Leal, P. 1874: 6; Almeida, C.A.B. 1996: 58-59.

16071403, IGREJA, Calvelo, Ponte de Lima, Viana do Castelo (160714). [537297.02, 4613994.39, 200]. Typology: Roman habitat? Chronology: Roman Period. References: Melo, M.F.S. 1967: 129; Reis, A.P.M. 1978: 32; Santos, L.A. 1979: 15; Araújo, J.R. 1982: 122-292; Almeida, C.A.B. 1996.

837/16071101, CARIDE, Cabaços, Ponte de Lima, Viana do Castelo (160711). [535808.35, 4615744.76, 265]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B., Baptista, A.J. 1981: 287-309; Almeida, C.A.B. 1996: 57-58.

16071402, VALADARES, Calvelo, Ponte de Lima, Viana do Castelo (160714). [537732.63, 4614706.02, 202]. Typology: Hill-fort. Chronology: Iron Age; Roman Period? References: Almeida, C.A.B. 1996: 74-75 Almeida, C.A.B. 2003

16071306, PINHEIRO, Calheiros, Ponte de Lima, Viana do Castelo (160713). [536365.67, 4628886.18, 250]. Typology: Roman habitat? Chronology: Roman Period. References: Morais, A.T. 1990: 9; Almeida, C.A.B. 1996: 59.

16071401, CADÉM, Calvelo, Ponte de Lima, Viana do Castelo (160714). [538749.79, 4614742.29, 180]. Typology: Hill-fort. Chronology: Iron Age; Roman Period. References: Costa, A.C.1869: 235; Almeida, C.A.B., Baptista, A.J. 1981: 287-309; Almeida, C.A.B. 1996: 72-73.

16071305, NOVA, Calheiros, Ponte de Lima, Viana do Castelo (160713). [535050.23, 4628910.44, 140]. Typology: Roman habitat? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 67-68.

16071501, CRASTO DE CEPÕES, Cepões, Ponte de Lima, Viana do Castelo (160715). [534046.22, 4631372.95, 122]. Typology: Romanized hillfort.. Chronology: Roman Period. References: Viana, A. 1932: 163; Almeida, C.A.B. 1990: 154; Almeida, C.A.B. 2003.

16071304, CABANA, Calheiros, Ponte de Lima, Viana do Castelo (160713). [536208.05, 4628114.32, 170]. Typology: Medieval habitat? Chronology: Roman Period? Middle Ages?. References: Morais, A.T. 1990: 9; Almeida, C.A.B. 1996: 66-67. 194

Catalogue

16074602, SÃO JOANES, Serdedelo, Ponte de Lima, Viana do Castelo (160746). [539797.04, 4621841.63, 350]. Typology: Medieval habitat? Chronology: Middle Ages? References: Almeida, C.A.B. 1996: 82-83.

Chronology: Iron Age; Roman Period? References: Viana, A. 1926: 90; Viana, A. 1932: 161; Almeida, C.A.B., Baptista, A.J. 1981; Almeida, C.A.B. 1990: 100-105.

16074601, IGREJA, Serdedelo, Ponte de Lima, Viana do Castelo (160746). [538339.05, 4622357.98, 232]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1996: 82-83.

16072801, CASTRO DE VILAR, Couto de Gondufe, Ponte de Lima, Viana do Castelo (160728). [541606.17, 4624535.2, 210]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Almeida, C.A.B. 1990: 122-124; Melo, M. F. S. 1967: 6; Almeida, C.A.B., Baptista, A.J. 1981: 294.

16071613, QUINTA DE BARREIROS, Correlhã, Ponte de Lima, Viana do Castelo (160716). [533790.53, 4622026.63, 35]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 275.

16071701, SOCRASTO, Estorãos, Ponte de Lima, Viana do Castelo (160717). [526999.12, 4628320.3, 420]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1990: 176.

16071612, QUINTA DO OUTEIRO, Correlhã, Ponte de Lima, Viana do Castelo (160716). [532846.37, 4621405.40, 55]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 275.

16071702, IGREJA PAROQUIAL DE ESTORÃOS, Estorãos, Ponte de Lima, Viana do Castelo (160717). [529937.64, 4626481.51, 30]. Typology: Votive epigraphy. Chronology: Roman Period. References: Almeida, C.A.B. 1990:170-172; Pereira, F.A. 1907: 29, 36-52; Vasconcelos, J. L.1913: 199-200; Cortez, R. 1946: 7; Neves, F.F. 1949: 39-47; Almeida, D.F. 1953: 209; Almeida, D.F. 1956: 111; Blazquez, J.M. 1975: 102; Moreira, M.A.F. 1982: 43-47; Alarcão, J. 1974: 91; Tranoy, A. 1981: 302.

16071610, TESIDO, Correlhã, Ponte de Lima, Viana do Castelo (160716). [532083.12, 4621617.83, 20]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 102. 16071609, PREGAL, Correlhã, Ponte de Lima, Viana do Castelo (160716). [532474.26, 4621958.85, 20]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 101-102.

16071703, BOUÇA DO CASTRO, Estorãos, Ponte de Lima, Viana do Castelo (160717). [529982.38, 4626820.96, 33]. Typology: Hill-fort; Roman mining site? Chronology: Iron Age; Roman Period. References: Viana, A. 1932: 161; Cortez, R.1946: 6; Centeno, R.M.S.1984: 190; Soeiro, T.1984: 21; Almeida, C.A.B. 1990: 167-169.

16071608, MOURELO, PAÇO, Correlhã, Ponte de Lima, Viana do Castelo (160716). [532496.66, 4622113.16, 30]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 100-101.

16071804, QUINTA DO PAÇO VELHO, Facha, Ponte de Lima, Viana do Castelo (160718). [529766.06, 4617783.49, 110]. Typology: Villa. Chronology: Roman Period. References: Almeida, C.A.B. 1990: 72-75, 281-338.

16071606, BARRÔ, Correlhã, Ponte de Lima, Viana do Castelo (160716). [532565.95, 4622113.48, 20]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1996: 100.

16071806, AGRAS DE GONDIM, Facha, Ponte de Lima, Viana do Castelo (160718). [529783.78, 4619078.9, 135]. Typology: Roman necropolis; Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1990: 72-73; Alarcão, J.1988: 1/123.

16071601, Nª SENHORA DA CONCEIÇÃO, Correlhã, Ponte de Lima, Viana do Castelo (160716). [533348.95, 4622610.56, 54]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Almeida, C.A.B. 1990: 107-108; Almeida, C.A.B., Baptista, A.J.,1981; Costa, A. C. 1706: 197; Vilela, S. 1877: 30; Melo, M. F. S. 1967: 64-65.

16071808, JORGES, Facha, Ponte de Lima, Viana do Castelo (160718). [530984.00, 4619392.41, 35]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1990: 68.

16071602, S.JOÃO, Correlhã, Ponte de Lima, Viana do Castelo (160716). [533449.11, 4620945.58, 250]. Typology: Roman treasure; Romanized Hill-fort; mutatio? Chronology: Iron Age; Roman Period. References: Viana, A. 1926: 90; Viana, A. 1932: 161; Hipólito, M.C. 1960-1961: 15; Reis, A.P.M. 1978: 36; Almeida, C.A.B., Baptista, A.J. 1981; Almeida, C.A.B. 1990: 106-107.

16071810, CHÃ DE BERGONHA, Facha, Ponte de Lima, Viana do Castelo (160718). [531678.27, 4619148.71, 125]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1990: 68. 16071813, CHÃ, Facha, Ponte de Lima, Viana do Castelo (160718). [532312.16, 4616930.96, 165]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1990: 68.

16071607, CASTRO DO EIRADO /ANTA, Correlhã, Ponte de Lima, Viana do Castelo (160716). [531901.36, 4620938.51, 30]. Typology: Romanized Hill-fort?. 195

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

16071814, SOPENA, Facha, Ponte de Lima, Viana do Castelo (160718). [532149.67, 4617084.44, 170]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1990: 71.

1912: 84-85; Almeida, C.A.B. 1979: 84-85, 111; Araújo, J.R. 1982: 166-172; Capela, M. 1895: 173; Almeida, C.A.B. 1990: 119-121. 16072302, IGREJA, Freixo, Ponte de Lima, Viana do Castelo (160723). [534414.15, 4612407.07, 180]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1996: 169.

16071821, CASTRO DE STº ESTÊVÃO DA FACHA, Facha, Ponte de Lima, Viana do Castelo (160718). [529783.53, 4619140.58, 142]. Typology: Romanized Hill-fort. Chronology: Chalcolithic; Bronze Age; Iron Age; Roman Period. References: Almeida, C.A.F., Soeiro, T., Almeida, C.A.B., Baptista, A.J. 1981; Almeida, C.A.F., Soeiro, T., Almeida, C.A.B., Baptista, A.J. 1982: 79; Almeida, C.A.B. 1990: 3-58.

16072303, PAÇÔ, Freixo, Ponte de Lima, Viana do Castelo (160723). [533258.16, 4612309.09, 131]. Typology: Farm. Chronology: Roman Period. References: Almeida, C.A.B. 2003.

16071823, CAPELA DE S. CIPRIANO, Facha, Ponte de Lima, Viana do Castelo (160718). [532994.92, 4619247.18, 320]. Typology: Roman habitat. Chronology: Iron Age; Roman Period. References: Almeida, C.A.B. 1990: 53.

16072301, S. CRISTOVÃO, Freixo (S. Julião), Ponte de Lima, Viana do Castelo (160723). [533494.20, 4611292.43, 265]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Almeida, C.A.B. 1996: 164-166.

16071826, CASTRO DE TRÁS-CIDADE, Facha/Vitorino dos Piães, Ponte de Lima, Viana do Castelo (160751). [533194.77, 4616009.76, 450]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Almeida, C.A.B., Baptista, A.J. s/d: 294; Viana, A. 1932; Almeida, C.A.B. 1990: 51-52.

1607243, PACINHO, Friastelas, Ponte de Lima, Viana do Castelo (160724). [536025.08, 4614018.73, 210]. Typology: Roman necropolis; Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 172-173. 16072401, SALAMONDE, Friastelas, Ponte de Lima, Viana do Castelo (160724). [535055.11, 4613767.19, 195]. Typology: Roman necropolis. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 169171.

16071901, QUINTA DA POSA, Feitosa, Ponte de Lima, Viana do Castelo (6079). [534852.86, 4622062.57, 80]. Typology: Milestone. Chronology: Roman Period. References: Melo, M. F. S. 1967: 124; Reis, A.P.M. 1978: 31; Almeida, C.A.B. 1990: 119121.

16072402, MONTE DE CASAIS, Friastelas, Ponte de Lima, Viana do Castelo (160724). [534826.72, 4613180.09, 197]. Typology: Medieval habitat. Chronology: Roman Period; Middle Ages. References: Almeida, C.A.B. 2003.

16072001, CALVÁRIO, Fojo Lobal, Ponte de Lima, Viana do Castelo (160720). [534737.61, 4617250.72, 350]. Typology: Hill-fort. Chronology: Iron Age. References: Almeida, C.A.B. 2003.

16072502, IGREJA, Gaifar, Ponte de Lima, Viana do Castelo (160725). [536943.24, 4610846.74, 150]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 175.

16072101, MONTE DE S. CRISTÓVÃO, Fontão, Ponte de Lima, Viana do Castelo (160721). [529168.93, 4622530.61, 88]. Typology: Watchtower? Chronology: Roman Period. References: Almeida, C.A.B. 1990: 172; Almeida, C.A.B., Baptista, A.J. s/d: 174.

16072501, BARALDE, Gaifar, Ponte de Lima, Viana do Castelo (160725). [536982.67, 4612173.12, 160]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 174.

16072204, FORNELOS-IGREJA, Fornelos, Ponte de Lima, Viana do Castelo (160722). [538140.23, 4620660.62, 220]. Typology: Milestone. Chronology: Roman Period. References: Reis, A.P.M. 1978: 27-28; Araújo, J.R. 1982: 162-164; Almeida, C.A.B. 1990: 119-121.

16072603, CASAIS, Gandra, Ponte de Lima, Viana do Castelo (160726). [528602.04, 4625674.15, 85]. Typology: Villa? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 176-177.

16072201, ANQUIÃO, Fornelos, Ponte de Lima, Viana do Castelo (160722). [536401.00, 4621946.91, 140]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 157.

16072602, PAÇO, Gandra, Ponte de Lima, Viana do Castelo (160726). [528928.48, 4624873.58, 130]. Typology: Roman habitat; Pottery Kiln. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 175-176.

16072202, STº AMARO, Fornelos; Feitosa, Ponte de Lima, Viana do Castelo (160722). [535698.83, 4623793.89, 100]. Typology: Mutatio/Vicus. Chronology: Roman Period. References: Guerra, L.F.

10672601, MADORNO, Gandra, Ponte de Lima, Viana do Castelo (106726). [541154.10, 4618826.8, 72]. Typology: Roman habitat. Chronology: Roman 196

Catalogue

Period. References: Viana, A. 1961: 8; Almeida, C.A.B. 1990: 141.

nology: Roman Period? References: Almeida, C.A.B. 1996: 241.

16072701, FREIRIZ, Gemieira, Ponte de Lima, Viana do Castelo (160727). [539197.94, 4625786.10, 50]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 177-178.

16073102, IGREJA, Mato (S. Lourenço), Ponte de Lima, Viana do Castelo (160731). [535916.72, 4612568.64, 200]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1996.

16072908, PÓVOA, Labruja, Ponte de Lima, Viana do Castelo (160729). [533461.01, 4633189.86, 160]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1996: 187.

16073101, QUINTA DE ARCELOS, Mato (S. Lourenço), Ponte de Lima, Viana do Castelo (160731). [536101.13, 4612692.93, 180]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 100-101.

16072901, PORTELA DA CÂMBOA, Labruja, Ponte de Lima, Viana do Castelo (160729). [531801.42, 4632997.22, 380]. Typology: Milestone. Chronology: Roman Period. References: Araújo, J.R. 1982: 198; Almeida, C.A.B. 1990: 162-163.

16073206, PARADA, Moreira de Lima, Ponte de Lima, Viana do Castelo (160732). [531256.33, 4625808.64, 55]. Typology: Small Farm? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 195.

16072902, FREITA, Labruja, Ponte de Lima, Viana do Castelo (160729). [533463.18, 4632727.24, 140]. Typology: Milestone. Chronology: Roman Period. References: Aráujo, J.R. 1982: 198; Almeida, C.A.B. 1990: 162-163.

16073204, BOUDILHÃO, Moreira de Lima, Ponte de Lima, Viana do Castelo (160732). [531027.89, 4625252.49, 25]. Typology: Villa?; Roman necropolis, Pottery Kiln. Chronology: Roman Period. References: Almeida, C.A.B. 1990: 175; Almeida, C.A.B. 1996: 192-193.

16072903, CAPELA DE S. SEBASTIÃO, Labruja, Ponte de Lima, Viana do Castelo (160729). [533632.18, 4631124.25, 60]. Typology: Milestone. Chronology: Roman Period. References: Aráujo, J.R. 1982:183, 192 e 198; Almeida, C.A.B. 1990: 162-163.

16073201, CIVIDADE DE MADORNA, Moreira do Lima, Ponte de Lima, Viana do Castelo (160732). [531484.19, 4626488.16, 85]. Typology: Roman habitat? Chronology: Roman Period. References: Almeida, C.A.B. 1990: 176.

16072904, ESPINHEIRO, Labruja, Ponte de Lima, Viana do Castelo (160729). [533624.62, 4632728.00, 170]. Typology: Milestone. Chronology: Roman Period. References: Araújo, J.R. 1982: 197; Almeida, C.A.B. 1990: 162; Alarcão, J. 1988: 1/41.

16073205, BOUÇA DO CAVALO, Moreira do Lima, Ponte de Lima, Viana do Castelo (160732). [531027.89, 4625252.49, 25]. Typology: Roman habitat? Chronology: Roman Period. References: Almeida, C.A.B. 1990: 175.

16072905, CASTRO DE CIMA, Labruja, Ponte de Lima, Viana do Castelo (160729). [534066.35, 4631989.89, 133]. Typology: Romanized Hill-fort? Chronology: Iron Age; Roman Period. References: Viana, A. 1926: 90; Almeida, C.A.B., Baptista, A.J. 1981; Almeida, C.A.B. 1990: 153-154.

16073202, CANADELO, Moreira do Lima, Ponte de Lima, Viana do Castelo (160732). [528997.25, 4624997.23, 80]. Typology: Small Farm? Chronology: Roman Period. References: Melo, M.F.S. 1967: 59; Almeida, C.A.B. 1990: 173-174; Teixeira, C., Medeiros, A.C. s/d: 39.

16072906, CASTRO DE BAIXO, Labruja, Ponte de Lima, Viana do Castelo (160729). [533975.26, 4631742.72, 96]. Typology: Romanized Hill-fort? Chronology: Iron Age; Roman Period. References: Almeida, C.A.B., Baptista, A.J. 1981; Almeida, C.A.B. 1990: 151-153; Almeida, C.A.B. 2003.

16073203, QUINTA DA BOAVISTA, Moreira do Lima, Ponte de Lima, Viana do Castelo (160732). [532058.77, 4627045.87, 170]. Typology: Roman habitat? Chronology: Roman Period. References: Melo, M.F.S.1967: 59; Almeida, C.A.B., 1996: 173.

16072907, PORTELA GRANDE (Cidade Murada?), Labruja, Ponte de Lima, Viana do Castelo (160729). [531730.86, 4633305.33, 442]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Viana, A. 1932: 162; Almeida, C.A.F. 1978: 33; Almeida, C.A.B. 1979: 125,131-132; Cunha, N.C.A. 1979: 546-548; Almeida, C.A.B; Baptista, A.J. 1981; Almeida, C.A.B. 1990: 147-149.

16073301, SANTA MARINHA, Navió, Ponte de Lima, Viana do Castelo (160733). [532742.01, 4613910.46, 150]. Typology: Roman habitat?. Chronology: Roman Period? References: Almeida, C.A.B. 1996: 198. 16073404, SABUGUEIRO, Poiares, Ponte de Lima, Viana do Castelo (160734). [531058.53, 4612854.38, 93]. Typology: Farm? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 281; Almeida, C.A.B. 2003.

16073103, RABIÃO, Mato (S. Lourenço), Ponte de Lima, Viana do Castelo (160731). [535338.07, 4612658.29, 180]. Typology: Roman habitat? Chro197

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

16073403, S. SEBASTIÃO, Poiares, Ponte de Lima, Viana do Castelo (160734). [530948.35, 4611589.42, 80]. Typology: Roman epigraphy. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 199-200.

16073703, BOUÇA DA SEPULTURA DE S. SIMÃO, Refojos do Lima, Ponte de Lima, Viana do Castelo (160737). [536651.59, 4627129.64, 161]. Typology: Roman habitat. Chronology: Iron Age; Roman Period. References: Calheiros, P. A. 1927: 180-181; Almeida, C.A.B., Baptista, A.J. 1981; Almeida, C.A.B. 1990: 144-145.

16073402, PAÇOS, Poiares, Ponte de Lima, Viana do Castelo (160734). [531829.93, 4611007.29, 90]. Typology: Roman habitat? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 199.

923/16073701, CASTRO DE S. SIMÃO, Refojos do Lima, Ponte de Lima, Viana do Castelo (160737). [536838.15, 4626760.5, 180]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Almeida, C.A.B. 1990: 144; Viana, A.1926: 90; Viana, A. 1932: 163; Almeida, C.A.B., Baptista, A.J. 1981.

16073401, LAJES, Poiares, Ponte de Lima, Viana do Castelo (160734). [530598.63, 4612235.58, 100]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1996: 198. 16074702, IGREJA, Rebordões (Souto), Ponte de Lima, Viana do Castelo (160747). [536530.18, 4619264.34, 108]. Typology: Roman necropolis. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 207.

16073906, OUTEIRO, Ribeira (S. João), Ponte de Lima, Viana do Castelo (160739). [538253.75, 4625349.20, 50]. Typology: Roman habitat? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 228-229.

16074402, SOLREUS, Rebordões (Stª Maria), Ponte de Lima, Viana do Castelo (160744). [534838.62, 4620242.84, 128]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 206.

16073905, AGROS, Ribeira (S. João), Ponte de Lima, Viana do Castelo (160739). [537499.33, 4623926.47, 100]. Typology: Roman habitat? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 228.

16074401, PAÇOS, Rebordões (Stª Maria), Ponte de Lima, Viana do Castelo (160744). [535095.77, 4619627.27, 100]. Typology: Small Farm? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 205.

16073904, IGREJA, Ribeira (S. João), Ponte de Lima, Viana do Castelo (160739). [537240.19, 4624912.06, 40]. Typology: Roman habitat? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 227-228.

16073702, QUINTA DO VAL DO VEZ /S. GIÃO, Refoios do Lima, Ponte de Lima, Viana do Castelo (160737). [539659.13, 4625881.18, 160]. Typology: Roman necropolis; Roman habitat. Chronology: Roman Period. References: Calheiros, P. A. 1927: 181; Almeida, C.A.B., 1990: 145-147.

16073903, PAÇO, Ribeira (S. João), Ponte de Lima, Viana do Castelo (160739). [538670.48, 4625135.55, 70]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 226-227. 16073902, CASTELINHO, Ribeira (S. João), Ponte de Lima, Viana do Castelo (160739). [535926.96, 4624350.18, 56]. Typology: Hill-fort. Chronology: Iron Age. References: Almeida, C.A.B. 1996: 225-226.

16073704, CAPELA DE STª. EULÁLIA, Refoios do Lima, Ponte de Lima, Viana do Castelo (160737). [538913.19, 4627203.29, 75]. Typology: Votive epigraphy; Roman necropolis; Roman habitat. Chronology: Roman Period. References: Barreiros, P.M.A.1926: 4548; Santos, L.A., Le Roux, P., Tranoy, A. 1983: 191; Almeida, C.A.B. 1990: 147-149.

16074003, BOUÇA DA CASTANHEIRA, Sá, Ponte de Lima, Viana do Castelo (160740). [530732.27, 4624202.59, 16]. Typology: Roman habitat?. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 232-233.

16073708, QUINTA DO PAÇO, Refojos do Lima, Ponte de Lima, Viana do Castelo (160737). [537321.26, 4627071.42, 70]. Typology: Villa? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 213214.

16074002, CASTANHEIRA, Sá, Ponte de Lima, Viana do Castelo (160740). [531908.91, 4624423.64, 16]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 230-231.

16073707, CASA SÓ, Refojos do Lima, Ponte de Lima, Viana do Castelo (160737). [537223.47, 4628119.54, 150]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 212-212.

16074001, LUGAR DE LOUREDO, Sá, Ponte de Lima, Viana do Castelo (160740). [531122.12, 4624821.12, 17]. Typology: Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1990: 173.

16073706, PAÇO VELHO, Refojos do Lima, Ponte de Lima, Viana do Castelo (160737). [537226.21, 4627595.24, 150]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 210-212.

16074101, SANDIÃES, Sandiães, Ponte de Lima, Viana do Castelo (160741). [536234.97, 4609085.18, 140]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1990: 233-234. 198

Catalogue

16074302, PAÇO, Santa Cruz do Lima, Ponte de Lima, Viana do Castelo (160743). [543027.50, 4626239.93, 50]. Typology: Roman habitat? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 239-240.

nology: Roman Period. References: Almeida, C.A.B. 1990: 82. 16075002, ALMUINHAS / ALDEIA, Vitorino das Donas, Ponte de Lima, Viana do Castelo (160750). [530288.22, 4620006.26, 25]. Typology: Roman necropolis; Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1990: 80-82.

16074301, IGREJA, Santa Cruz do Lima, Ponte de Lima, Viana do Castelo (160743). [543258.71, 4626179.64, 70]. Typology: Votive epigraphy; Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 239.

16075003, Nª SRª DO BARCO, Vitorino das Donas, Ponte de Lima, Viana do Castelo (160750). [529246.03, 4620618.75, 30]. Typology: Farm. Chronology: Roman Period. References: Costa, A. J. 1958: 144; Melo, F.S. 1967: 64; Almeida, C.A.B., Baptista, A.J. 1981; Almeida, C.A.B. 1990: 80; Almeida, C.A.B. 2003.

16074501, VILARINHO, Seara, Ponte de Lima, Viana do Castelo (160745). [531833.43, 4620629.79, 28]. Typology: Farm. Chronology: Iron Age; Roman Period. References: Viana, A. 1926: 90; Viana, A. 1932: 163; Almeida, C.A.B., Baptista, A.J. 1981; Almeida, C.A.B. 1990: 98; Almeida, C.A.B. 2003.

16075004, QUINTA DA TORRE, Vitorino das Donas, Ponte de Lima, Viana do Castelo (160750). [527634.02, 4619286.17, 20]. Typology: Roman habitat? Chronology: Roman Period. References: Almeida, C.A.B. 1990: 42.

16074202, CARRO, Stª Comba, Ponte de Lima, Viana do Castelo (160742). [532117.68, 4624208.68, 20]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1990: 236.

16074903, CIVIDADE, Vitorino de Piães, Ponte de Lima, Viana do Castelo (160749). [536512.51, 4631755.25, 556]. Typology: Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 274.

16074201, STª COMBA, Stª Comba, Ponte de Lima, Viana do Castelo (160742). [531773.52, 4623713.68, 13]. Typology: Hill-fort? Chronology: Iron Age? References: Almeida, C.A.B. 1990: 166-167.

16075110, PAÇO, Vitorino de Piães, Ponte de Lima, Viana do Castelo (160751). [533476.63, 4615054.98, 150]. Typology: Small Farm? Chronology: Roman Period? References: Almeida, C.A.B. 1996: 273.

938/16074803, QUINTA DO CASAL, Vilar das Almas, Ponte de Lima, Viana do Castelo (160748). [537383.74, 4610663.97, 165]. Typology: Roman necropolis; Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 253-254.

16075109, OUTEIRO DE S. PEDRO, Vitorino de Piães, Ponte de Lima, Viana do Castelo (160751). [533566.80, 4615548.86, 160]. Typology: Roman epigraphy- undetermined. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 271-273.

16074802, ASSENTO, Vilar das Almas, Ponte de Lima, Viana do Castelo (160748). [537893.01, 4610604.96, 170]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1996: 100-101.

16075108, OUTEIRO, Vitorino de Piães, Ponte de Lima, Viana do Castelo (160751). [531750.70, 4613227.50, 108]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 270.

16074801, MONTE DOS OUTEIROS (Monte do Hillfort; Monte de Santozilos), Vilar das Almas, Ponte de Lima, Viana do Castelo (160748). [538698.89, 4611318.61, 234]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Almeida, C.A.B. 1996: 250-251.

16075107, FIGUEIRÓ, Vitorino de Piães, Ponte de Lima, Viana do Castelo (160751). [532459.50, 4615019.46, 140]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1996: 270.

16074902, REGUEIRA, Vilar do Monte, Ponte de Lima, Viana do Castelo (160749). [537549.83, 4631884.01, 550]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1996: 254.

16075106, FIDALGA, Vitorino de Piães, Ponte de Lima, Viana do Castelo (160751). [530707.67, 4613778.10, 120]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1996: 269.

16075007, PESSEGUEIRO, Vitorino das Donas, Ponte de Lima, Viana do Castelo (160750). [530562.70, 4620655.09, 20]. Typology: Small Farm. Chronology: Roman Period? References: Almeida, C.A.B. 1996: 255-257.

16075105, REGUEIRA, Vitorino de Piães, Ponte de Lima, Viana do Castelo (160751). [530771.09, 4615166.22, 190]. Typology: Roman habitat? Chronology: Bronze Age; Roman Period. References: Almeida, C.A.B. 1996: 100-101.

16075005, ALDEIA, Vitorino das Donas, Ponte de Lima, Viana do Castelo (160750). [530402.43, 4620315.28, 27]. Typology: Roman necropolis. Chro199

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

16075104, CANCELAS, Vitorino de Piães, Ponte de Lima, Viana do Castelo (160751). [530938.91, 4613779.09, 145]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1996: 100-101.

Viana do Castelo (160728160746). [541390.81, 4621850.68, 609]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Melo, M.F.S.1967: 71; Almeida, C.A.B., Baptista, A.J. 1981; Almeida, C.A.B 1990: 109-110.

16075103, S. SIMÃO, Vitorino de Piães, Ponte de Lima, Viana do Castelo (160751). [531837.04, 4614615.74, 176]. Typology: Hill-fort. Chronology: Iron Age. References: Almeida, C.A.B. 1996: 267-268.

16071704, CASAIS, Estorãos, Ponte de Lima, Viana do Castelo (160717). [529732.84, 4625771.29, 57]. Typology: Roman Mines. Chronology: Roman Period. References: Almeida, C.A.B. 2003: 113-115.

16075102, CRESTO, Vitorino de Piães, Ponte de Lima, Viana do Castelo (160751). [531809.82, 4615540.86, 220]. Typology: Hill-fort. Chronology: Iron Age. References: Viana, A.1926: 90; Viana, A.1932: 164; Baptista, A.J. 1981: 294-295; Almeida, C.A.B. 1990: 48-51.

16071802, FOJO, Facha, Ponte de Lima, Viana do Castelo (160718). [529253.59, 4618768.3, 130]. Typology: Roman mill? Chronology: Roman Period? References: Almeida, C.A.B 1990: 78. 16071805, FOJO VELHO, Facha, Ponte de Lima, Viana do Castelo (160718). [529783.78, 4619078.9, 135]. Typology: Roman necropolis? Chronology: Roman Period? References: Almeida, C.A.B 1990: 72-73.

16070404, QUINTA DE FALDEJÃES, Arcozelo, Ponte de Lima, Viana do Castelo (160704). [534786.85, 4626133.37, 60]. Typology: Milestone. Chronology: Roman Period. References: Argote, J.C. 1732: 614; CIL II 4871, 4872, 4873; Capella, M. 1895: 128; Araújo, J.R. 1962: 17-18; Araújo, J.R. 1982: 122-292; Melo, M.F.S. 1967: 108-111; Reis, A.P.M. 1978: 2333; Tranoy, A. 1981: 216; Almeida, C.A.B. 1990: 160161; Almeida, C.A.B. 1996: 37-41.

16071809, PEDRA, Facha, Ponte de Lima, Viana do Castelo (160718). [531776.03, 4617946.33, 85]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B 1990: 68. 16071801, CIVIDADE; FREI; FORNO; JUNCAL; LOURINHO, Facha, Ponte de Lima, Viana do Castelo (160718). [531340.03, 4617235.06, 60]. Typology: Farm? Chronology: Roman Period. References: Almeida, C.A.B 1990: 68.

16070704, PEDREGUEIRA, Beiral do Lima, Ponte de Lima, Viana do Castelo (160707). [542630.01, 4623215.01, 250]. Typology: Small Farm? Chronology: Roman Period. References: Baptista, A.J. 1982: 295-311; Almeida, C.A.B. 1996: 53-54.

16071812, MARIA VELHA, Facha, Ponte de Lima, Viana do Castelo (160718). [531687.54, 4617051.54, 100]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B 1990: 68.

16071611, PEREIRA, Correlhã, Ponte de Lima, Viana do Castelo (160716). [532542.99, 4622082.53, 30]. Typology: Pottery Kiln. Chronology: Roman Period. References: Almeida, C.A.B. 2000: 103-106. 16071604, QUINTA DA AGRA, Correlhã, Ponte de Lima, Viana do Castelo (160716). [533511.77, 4622364.59, 20]. Typology: Milestone. Chronology: Roman Period. Almeida, C.A.B. 1979: 113; Capella, M. 1895: 171; Almeida, C.A.B 1990: 118.

16071818, PRAZIL, MENDE, MANGAS, TIANDES, TELHEIRO, Facha, Ponte de Lima, Viana do Castelo (160718). [531331.94, 4619085.51, 60]. Typology: Villa? Chronology: Roman Period. References: Almeida, C.A.B 1990: 68-70; Almeida, C.A.B 1996: 129132.

16071605, PAÇO / TRAVASSELAS, Correlhã, Ponte de Lima, Viana do Castelo (160716). [532471.88, 4622483.16, 20]. Typology: Villa. Chronology: Roman Period. References: Almeida, C.A.B 1990: 118-119; Almeida, C.A.B 1996: 95-97.

16071820, QUINTA DO PAÇO, Facha, Ponte de Lima, Viana do Castelo (160718). [530061.97, 4618864.17, 70]. Typology: Roman necropolis; Pottery Kiln; Iron age house. Chronology: Iron Age; Roman Period. References: Almeida, C.A.B 1990: 58-62.

16071603, IGREJA, Correlhã, Ponte de Lima, Viana do Castelo (160716). [533308.21, 4621438.39, 75]. Typology: Epigrafia romana indeterminada; Villa? Chronology: Roman Period. References: Almeida, C.A.B 1990: 119; Almeida, C.A.B 1996: 98-100.

16071822, STº ESTEVÃO-MINAS, Facha, Ponte de Lima, Viana do Castelo (160718). CMP [529783.78, 4619078.9, 140]. Typology: Roman mines? Chronology: Roman Period? References: Teixeira, C., Ferreira, A. M. 1972: 35; Almeida, C.A.B 1990: 77-78.

16072803, SERÉM, Couto de Gondufe, Ponte de Lima, Viana do Castelo (160728). [541657.74, 4623610.23, 260]. Typology: Small Farm? Chronology: Roman Period. References: Almeida, C.A.B 1990: 182.

16071824, CHÃ DA TORRE-MINAS, Facha, Ponte de Lima, Viana do Castelo (160718). [529948.76, 4618308.55, 60]. Typology: Roman mines. Chronology: Roman Period. References: Almeida, C.A.F., Soeiro, T., Almeida, C.A.B., Baptista, A.J. 1981; Almeida, C.A.B 1990: 77-78.

16072802, S. LOURENÇO DA ARMADA (Hill-fort Mau), Couto de Gondufe; Cerdedelo, Ponte de Lima, 200

Catalogue

16071825, MONTE DA NÓ (MINAS), Facha, Fojo Lobal, Vitorino de, Ponte de Lima, Viana do Castelo (160718). [533191.20, 4616780.77, 450]. Typology: Roman mines. Chronology: Roman Period. References: Almeida, C.A.B 1990: 77- 78.

Roman Period. References: Viana, A. 1926: 90;Viana, A. 1932: 161; Cardoso, A.P. 1966; Melo, M.F.S. 1967: 71; Almeida, C.A.B., Baptista, A.J. 1981; Almeida, C.A.B. 1990: 109-112; Paço, A., Paço, A.Q. s/d: 176. 16075001, CIVIDADE DAS DONAS, Vitorino das Donas, Ponte de Lima, Viana do Castelo (160750). [530447.32, 4620623.76, 30]. Typology: Hill-fort. Chronology: Iron Age; Roman Period. References: Almeida, C.A.B 1990: 82- 98; Almeida, C.A.B. 2003.

16071906, MONTE DOS MEDOS (Campo Raso), Feitosa; Correlhã, Ponte de Lima, Viana do Castelo (160719). [533855.44, 4622952.2, 25]. Typology: Roman necropolis. Chronology: Roman Period. References: Melo, M.F.S. 1967: 65; Baptista, A.J. 1981; Almeida, C.A.B. 1990: 116-117.

16075006, PAÇO DAS DONAS, Vitorino das Donas, Ponte de Lima, Viana do Castelo (160750). [529478.95, 4620157.08, 30]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B 1990: 80; Almeida, C.A.B 1996: 261-262.

16072203, QUINTA DO OUTEIRO, Fornelos, Ponte de Lima, Viana do Castelo (160722). [538126.98, 4618810.05, 200]. Typology: Small Farm? Chronology: Roman Period. References: Alarcão 1988: 1/79; Almeida, C.A.B 1990: 117-118.

16075101, ALTO DE VALADAS, Vitorino de Piães, Ponte de Lima, Viana do Castelo (160751). [532962.89, 4616162.89, 347]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Viana, A. 1932: 164; Almeida, C.A.B., Baptista, A.J. 1981: 294-295; Almeida, C.A.B 1990: 51.

16072205, MONTE DE STª MARIA MADALENA (Monte das Santas), Fornelos, Ponte de Lima, Viana do Castelo (160722). [536397.08, 4622717.95, 256]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Paço, A.Q. s/d: 174-178; Viana, A. 1932: 161; Melo, M.F.S. 1967: 70-71; Baptista, A.J. 1981; Almeida, C.A.B 1990: 108-109.

16073801, ALTO DA PENA, Rendufe/Labrujó, Ponte de Lima/P. de Coura, Viana do Castelo (160738). [537486.67, 4635122.13, 746]. Typology: Hill-fort. Chronology: Bronze Age; Iron Age. References: Almeida, C.A.B., Baptista, A.J. 1981: 287-309; Silva 1986; Almeida, C.A.B. 1990: 150; Queiroga, F. 1992; Almeida, C.A.B. 1996: 219-221.

16072206, QUINTA DO CRASTO, Fornelos, Ponte de Lima, Viana do Castelo (160722). [538700.34, 4619615.01, 163]. Typology: Hill-fort. Chronology: Iron Age; Roman Period. References: Araújo, J.R. 1962: 97; Almeida, C.A.B. 1979: 109; Almeida, C.A.B., Baptista, A.J. 1981; Almeida, C.A.B. 1990: 113-114; Almeida, C.A.B. 2003.

16100201, IGREJA, Candemil, V. N. Cerveira, Viana do Castelo (161002). [525978.43, 4642411.30, 294]. Typology: Roman habitat? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 21/22.

16073501, PONTE DE LIMA-LIMIA (Quinta da Graciosa; Sobral; Av. dos Plátanos; Sangarinhal), Ponte de Lima, Ponte de Lima, Viana do Castelo (160735). [534864.37, 4624437.45, 35]. Typology: vicus. Chronology: Roman Period. References: Almeida, C.A.B 1990: 118.

16100301, CIVIDADE, Cornes, V. N. Cerveira, Viana do Castelo (161003). [526586.40, 4645047.83, 62]. Typology: Roman habitat. Chronology: Roman Period. References: Vieira, J. A. 1886; Almeida, C.A.B. 2000: 22, 23.

16074701, CIVIDADE DE REBORDÕES, Rebordões (Souto), Ponte de Lima, Viana do Castelo (160747). [536256.34, 4618584.44, 160]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B 1990: 118; Almeida, C.A.B., Baptista, A.J. 1981.

16100603, MONTE DO CRASTO, Gondarém, V. N. Cerveira, Viana do Castelo (161006). [520931.62, 4640792.1, 1210]. Typology: Romanized Hill-fort? Chronology: Iron Age; Roman Period? References: Pereira, F. A. 1924; Oliveira, E.P. 1994: 72. 16100801, FORTE DE LOBELHE, Lobelhe, V. N. Cerveira, Viana do Castelo (161008). [521503.36, 4644709.53, 46]. Typology: vicus. Chronology: Iron Age; Roman Period. References: Azevedo, P. A. 1897; Oliveira, E.P. 1994: 74; Almeida, C.A.B. 2003.

16073705, CASTRO DO SOUTINHO (Castro da Pousada; Castro do Espadanal), Refojos do Lima, Ponte de Lima, Viana do Castelo (160737). [538470.83, 4627910.26, 160]. Typology: Hill-fort. Chronology: Iron Age. References: Callheiros, A. 1910: 98; Teixeira et alii 1972: 39; Silva 1986; Almeida, C.A.B 1990: 143-144; Queiroga, F. 1992; Almeida, C.A.B 1996: 210-211.

16100901, CASTRO DE MENTRESTIDO, Mentrestido, V. N. Cerveira, Viana do Castelo (161009). [527694.64, 4639364.39, 253]. Typology: Hill-fort. Chronology: Iron Age; Roman Period. References: Vieira, J. A. 1886; Oliveira, E.P. 1994: 76; Almeida, C.A.B. 2003.

16073901, CASTELO DA ERMIDA (Castro Suberoso), Ribeira (S. João), Ponte de Lima, Viana do Castelo (160739). [538886.15, 4623687.15, 443]. Typology: Romanized Hill-fort. Chronology: Iron Age; 201

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

16101101, CASTRO, Reboreda, V. N. Cerveira, Viana do Castelo (161011). [522484.29, 4645406.11, 44]. Typology: Hill-fort. Chronology: Iron Age; Roman Period. References: Cunha, A. R. 1975; Silva, A.C.F. 1986: 69; Oliveira, E.P. 1994: 78; Almeida, C.A.B. 2003.

16080501, TELHÕES, Fontoura, Valença, Viana do Castelo (160805). CMP 7 [530068.47, 4644678.94, 130]. Typology: Small Farm. Chronology: Roman Period. References: Leal, P. 1874: 212; Vieira, J.A. 1886: 111; Almeida, C.A.B. 1996, VI: 26. 16080502, PAÇO, Fontoura, Valença, Viana do Castelo (160805). CMP 7 [531100.50, 4645608.66, 85]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 21-25.

16101203, CAPELA DE S. BRÁS, Sapardos, V. N. Cerveira, Viana do Castelo (161012). [529283.21, 4639679.18, 180]. Typology: Milestone. Chronology: Roman Period. References: Nunes, J. C. 1957; Pereira, F. A. 1903: 132-136; Pereira, F. A. 1924; Almeida, C. A. B. 1979; Oliveira, E.P. 1994: 79.

16080603, IGREJA, Friestas, Valença, Viana do Castelo (160806). [535770.91, 4655315.41, 62]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1996: 29.

16100402, TORRE, Covas, V.N. Cerveira, Viana do Castelo (161004). [525785.01, 4638586.16, 169]. Typology: Hill-fort? Chronology: Iron Age? References: Viana, A. 1932; Silva, A.C.F. 1986: 69; Oliveira, E.P.1994: 70.

16080602, QUINTA DO CRASTO, Friestas, Valença, Viana do Castelo (160806). [536619.61, 4655689.87, 24]. Typology: Hill-fort. Chronology: Iron Age; Roman Period. References: Almeida, C.A.B. 1996: 28-29.

16080201, PAÇO, Boivão, Valença, Viana do Castelo (160802). [538539.90, 4653448.49, 170]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1996: 11.

16080601, EIRADO, Friestas, Valença, Viana do Castelo (160806). [537288.22, 4655323.24, 60]. Typology: Roman habitat? Chronology: Roman Period? References: Viana, A. 1926: 88-90,111-113; Viana, A. 1932: 154-167; Oliveira, A.L. 1987: 23; Queiroga, F. 1992; Almeida, C.A.B. 1996: 27.

16080303, PAÇOS, Cerdal, Valença, Viana do Castelo (160803). [530286.87, 4647455.71, 50]. Typology: Hill-fort. Chronology: Iron Age. References: Costa, A.C. 1868: 244; Leal, P. 1874: 245; Vieira, J.A. 1886: 10; Viana, A. 1926: 88-90, 111-113; Viana, A., 1932: 154-167; Oliveira, A.L. 1978: 23-24, 30; Silva, A.C.F. 1986; Queiroga, F. 1992; Almeida, C.A.B. 1996: 1316; Almeida, C.A.B. 2003.

16080702, PAÇO, Gandra, Valença, Viana do Castelo (160807). [532563.85, 4652801.63, 150]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996, VI: 31-32. 16080701, CRESTOS, Gandra, Valença, Viana do Castelo (160807). CMP 7 [532107.11, 4652090.15, 178]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Almeida, C.A.B. 1996: 29-31.

16080302, BACELAR, Cerdal, Valença, Viana do Castelo (160803). CMP 7 [140]. Typology: Hill-fort. Chronology: Iron Age; Roman Period. References: Leal, P. 1874: 245; Vieira, J.A. 1886: 109; Almeida, C.A.B. 1996: 16-18; Almeida, C.A.B. 2003.

16080801, ALTO DO FACHO (Cristelo; Crastos), Ganfei, Valença, Viana do Castelo (160808). [531637.33, 4654277.90, 90]. Typology: Hill-fort. Chronology: Iron Age; Roman Period. References: Viana, A. 1926: 88-90, 111-113; Viana, A. 1932: 154167; Silva, A.C.F.1986; Queiroga, F. 1992; Almeida, C.A.B. 1996: 32-34.

16080301, MONTE CRASTO, Cerdal, Valença, Viana do Castelo (160803). [533736.92, 4647933.92, 220]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Leal, P. 1874: 109; Vieira, J.A. 1886: 110; Viana, A. 1926: 88-90, 111-113; Viana, A. 1932: 154-167; Oliveira, A.L. 1978; Silva, A.C.F. 1986; Queiroga, F. 1992; Almeida, C.A.B. 1996, VI: 12-13

16080802, MOSTEIRO DE GANFEI, Ganfei, Valença, Viana do Castelo (160808). [531474.60, 4654678.14, 50]. Typology: Small Farm? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 34-35.

16080503, ALTO DO CORGO (Monte do Cónego; Monte de Grove), Fontoura, Valença, Viana do Castelo (160805). CMP 7 [531610.25, 4644870.69, 120]. Typology: Romanized Hill-fort; Roman treasure. Chronology: Iron Age; Roman Period. References: Leal, P. 1874: 212; Vieira, J.A. 1886: 111; Viana, A. 1932: 154-167; Hipólito, M.C. 1960-1961: 1-166; Centeno, R. 1977: 94 e 99; Oliveira, A.L. 1978: 24 e 32; Silva, A.C.F. 1986; Centeno, R. 1987: 99-101; Alarcão, J.1988: 1/13; Queiroga, F. 1992; Almeida, C.A.B. 1996: 18-21.

16080901, COUTO/CRUZ, Gondomil, Valença, Viana do Castelo (160809). [50]. Typology: Hill-fort; Roman necropolis. Chronology: Iron Age; Roman Period. References: Viana, A. 1926: 88-90, 111-113; Viana, A. 1932: 154-167; Silva, A.C.F. 1986; Oliveira, A.L. 1987: 35; Queiroga, F. 1992; Almeida, C.A.B. 1996: 36-42; Almeida, C.A.B. 2003. 16081101, S. SEBASTIÃO, S. Julião, Valença, Viana do Castelo (160811). [527839.13, 4643652.05, 250]. 202

Catalogue

Typology: Undetermined habitat. Chronology: Iron Age. References: Almeida, C.A.B. 1996: 43.

16090104, BAGANHEIRAS, Afife, Viana do Castelo, Viana do Castelo (160901). [511752.10, 4624146.19, 40]. Typology: Villa; Roman treasure. Chronology: Roman Period. References: Alarcão, J. 1988: 1/61; Moreira, M.A.F., 1982: 31-96; Almeida, C.A.B. 1996: 29.

16081403, BRAGUEIRO/JUNCAL, S. Pedro da Torre, Valença, Viana do Castelo (160814). [527907.10, 4649790.00, 20]. Typology: Small Farm? Chronology: Roman Period. References: Vieira, J.A. 1886: 108; Almeida, C.A.B. 1996: 45-46.

16090101, CASTRO DE AGRICHOUSA, Afife, Viana do Castelo, Viana do Castelo (160901). [513135.87, 4625042.99, 140]. Typology: Hill-fort. Chronology: Iron Age. References: Meira, A.R. 1945; Paço, A., Quesado, A.P. 1956; Silva, A.C.F. 1986; Queiroga, F. 1992; Almeida, C.A.B. 1996, II: 21.

16081402, CRASTOS -IGREJA, S. Pedro da Torre, Valença, Viana do Castelo (160814). [528208.16, 4649297.71, 30]. Typology: Undetermined habitat. Chronology: Iron Age? References: Viana, A. 1926: 88-90, 111-113; Viana, A. 1932: 154-167; Silva, A.C.F. 1986; Oliveira, A.L. 1987: 23, 40-41; Almeida, C.A.F. 1987: 165-166; Queiroga, F. 1992; Almeida, C.A.B. 1996: 44-45.

16090103, MONTE DO CASTRO, Afife, Viana do Castelo, Viana do Castelo (160901). [513436.41, 4624827.66, 160]. Typology: Hill-fort? Medieval castle? Chronology: Iron Age; Middle Ages? References: Meira, A.R. 1945: 29; Silva, A.C.F. 1986; Queiroga, F. 1992; Almeida, C.A.B. 1996 II: 28-29.

16081401, CHAMOSINHOS (Igreja de S. Miguel de Fontoura), S. Pedro da Torre, Valença, Viana do Castelo (160814). CMP 7 [526669.80, 4648428.16, 0]. Typology: Milestone. Chronology: Roman Period. References: Vieira, J.A. 1886: 108; Santos, L.A. 1979: 3-52; Almeida, C. A. B. 1979: 123-124; Araújo, J.R. 1982: 239; Almeida, C.A.F. 1987: 165-166; Alarcão, J. 1988: 1/10; Almeida, C.A.B. 1996: 46-47.

16090102, CASTRO DE STº ANTÓNIO, Afife, Viana do Castelo, Viana do Castelo (160901). [511703.30, 4625749.85, 68]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Viana, A. 1932: 154-167; Viana, A. 1937: 75-83; Meira, A.R. 1945: 26-29; Viana, A. 1955: 525-528; Paço, A., Quesado, A.P. 1956; Viana, A. 1963: 167178; Neves, L.Q. 1965: 172-180; Moreira, M.A.F. 1982: 31-96; Silva, A.C.F. 1983-1984; Silva, A.C.F. 1983: 100-147; Silva, A.C.F. 1986: 149-150, 194; Queiroga, F. 1992; Almeida, C.A.B. 1996: 22-23.

16081301, OUTEIRO DA MADORRA (Outeiro da Cruzeira), Silva, Valença, Viana do Castelo (160813). [528522.89, 4645412.79, 55]. Typology: Hill-fort. Chronology: Iron Age; Roman Period. References: Vieira, J.A. 1886: 111; Viana, A. 1926: 88-90, 111113; Viana, A. 1932: 154-167; Oliveira, A.L. 1978: 24, 41, 43; Silva, A.C.F. 1986; Queiroga, F. 1992; Almeida, C.A.B. 1996: 47-49; Almeida, C.A.B. 2003.

16090203, PAÇO, Alvarães, Viana do Castelo, Viana do Castelo (160902). [521026.54, 4611090.88, 60]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 38.

16081502, VALENÇA, Valença, Valença, Viana do Castelo (160815). [529432.19, 4653744.06, 60]. Typology: Roman treasure; Funerary epigraphy. Chronology: Roman Period. References: Leal, P. 1882: 122126; Vieira, J.A. 1886: 97; CIL II 2465; Roldan Hervás, J.M. 1974: 201 e 449; Tranoy, A. 1981: 223; Le Roux, P. 1982: 222-223; Alarcão, J. 1988: 1/4; Almeida, C.A.B. 1996, VI : 53-54

16090202, IGREJA, Alvarães, Viana do Castelo, Viana do Castelo (160902). [521976.10, 4610723.63, 70]. Typology: Votive epigraphy. Chronology: Roman Period. References: Cepa, M.M. 1937: 73; Rosário, A. 1973; Tranoy, A. 1981: 311; Santos, L., Le Roux, P., Tranoy, A. 1983: 192; Almeida, C.A.B. 1996, II: 35-37.

16081501, ARINHOS, Valença, Valença, Viana do Castelo (160815). CMP 2 [529730.60, 4653868.68, 12]. Typology: Milestone. Chronology: Roman Period (44 A.D). References: CIL II nº 4875; Argote, J.C. 1732: 570; Leal, P. 1880: 134; Vieira, J.A. 1886: 99-100; Capella, M. 1895: 98; Guerra, L.F.1919: 43-44, 71-72; Melo, M.F.S. 1967; Almeida, C.A.B. 1979: 61-163; Tranoy, A. 1981: 210; Araújo, J.R. 1982: 122-246; Alarcão, J. 1988: 1/3; Almeida, C.A.B. 1996, VI: 50-52.

16090201, OUTEIRO (Eiteiro), Alvarães, Viana do Castelo, Viana do Castelo (160902). [520869.01, 4609579.23, 58]. Typology: Farm; Medieval necropolis. Chronology: Roman Period; Middle Ages. References: Cepa, M.M. 1939: 19-20; Silva, A.C.F. 1986; Queiroga, F. 1992; Almeida, C.A.B. 1996: 3235; Almeida, C.A.B. 2003. 16090302, AMONDE, Amonde, Viana do Castelo, Viana do Castelo (160903). [520682.48, 4626078.62, 120]. Typology: Small Farm; mutatio? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 40.

16081601, EIRADO, Verdoejo, Valença, Viana do Castelo (160816). [533429.76, 4654502.04, 183]. Typology: Undetermined habitat. Chronology: Iron Age? References: Viana, A. 1926: 88-90, 111-113; Viana, A. 1932: 154-167; Oliveira, A.L. 1978: 47; Silva, A.C.F. 1986; Queiroga, F. 1992; Almeida, C.A.B. 1996, VI: 54-55.

16090301, ALTO DA COROA, Amonde, Viana do Castelo, Viana do Castelo (160903). [520564.66, 4626910.99, 212]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: 203

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Silva, A.C.F. 1986; Queiroga, F. 1992; Almeida, C.A.B. 1996: 38-40.

16090803, IGREJA, Carreço, Viana do Castelo, Viana do Castelo (160908). [511408.85, 4622202.65]. Typology: Iron age treasure. Chronology: Iron Age. References: Severo, R. 1905-1908: 403-412; Teixeira, C. 1944: 154-161; Teixeira, C. 1946: 169-174; Lopez Cuevillas 1951: 80-81; Blanco Freijeiro, A. 1957: 289; Parreira, R. 1980; Perez Outeiriño, B. 1982: 43-45; Silva, A C.F. 1986: 262; Almeida, C.A.B. 1996: 65-66.

16090402, FOZ DO RIBEIRO DE ANHA, Anha, Viana do Castelo, Viana do Castelo (160904). [514706.09, 4613665.75, 5]. Typology: Salt pans. Chronology: Iron Age; Roman Period? References: Almeida, C.A.B. 1996: 43-46. 16090401, PAÇO, Anha, Viana do Castelo, Viana do Castelo (160904). [517274.04, 4613178.03, 60]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 42-43.

16090802, MONTEDOR (Gândara), Carreço, Viana do Castelo, Viana do Castelo (160908). [510923.54, 4622386.93, 75]. Typology: Romanized Hill-fort; Roman treasure; Salt pans; Iron age treasure. Chronology: Bronze Age; Iron Age; Roman Period. References: Almeida, C.A.B. 1996: 59-64.

16090501, IGREJA, Areosa, Viana do Castelo, Viana do Castelo (160905). [512270.03, 4618348.96, 14]. Typology: Small Farm? Chronology: Roman Period. References: Almeida, C.A.B. 1990: 239; Almeida, C.A.B. 1996: 50-51.

16090801, CRASTOS VELHOS (Croa), Carreço, Viana do Castelo, Viana do Castelo (160908). [511847.74, 4622172.52, 112]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Almeida, C.A.B. 1996: 57-59.

16090502, CANCELA DA AREOSA, Areosa, Viana do Castelo, Viana do Castelo (160905). [513649.39, 4622114.18, 40]. Typology: Romanized Hill-fort Chronology: Iron Age; Roman Period. References: Viana, A., Oliveira, M.S. 1954; Almeida, C.A.B. 1990: 224; Almeida, C.A.B. 1996: 49-50.

16090905, SUB-RIBES; CORTINHAL, Carvoeiro, Viana do Castelo, Viana do Castelo (160909). [527853.35, 4610404.78, 50]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1996: 69.

16090503, MONTE CRASTO, Areosa, Viana do Castelo, Viana do Castelo (160905). [512730.34, 4619398.36, 160]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Vasconcelos, J.L. 1903: 17; Guerra, L.F. 1909: 59-60; Viana, A. 1926: 90; Sarmento, F.M. 1933: 168; Viana, A. 1955: 70; Viana, A. 1962: 81; Silva, A.C.F. 1986; Almeida, C.A.B. 1990: 223; Queiroga, F. 1992; Almeida, C.A.B. 1996: 46-48.

16090904, CONVENTO, Carvoeiro, Viana do Castelo, Viana do Castelo (160909). [529030.98, 4610964.58, 110]. Typology: Undetermined habitat. Chronology: Roman Period? Middle Ages? References: Almeida, C.A.B. 1996: 69. 16090903, REBOLEDO/TELHEIRAS, Carvoeiro, Viana do Castelo, Viana do Castelo (160909). [527597.21, 4610835.57, 67]. Typology: Pottery Kiln? Chronology: Roman Period? References: Almeida, C.A.B. 1996: 68.

16090601, PAÇO, Barroselas, Viana do Castelo, Viana do Castelo (160906). [526508.24, 4611294.07, 80]. Typology: Undetermined habitat. Chronology: Roman Period? References: Almeida, C.A.B. 1996: 51.

16090902, LAJE, Carvoeiro, Viana do Castelo, Viana do Castelo (160909). [527823.52, 4612131.76, 140]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 67-68.

16090702, PAÇO, Cardielos, Viana do Castelo, Viana do Castelo (160907). [521514.81, 4617969.83, 22]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 56.

16090901, ALTO DOS MOUROS, Carvoeiro, Viana do Castelo, Viana do Castelo (160909). [528166.71, 4613089.17, 385]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Almeida, C.A.B. 1996: 66-67.

16090701, TERRONHA, Cardielos, Viana do Castelo, Viana do Castelo (160907). [520981.73, 4618492.55, 119]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Almeida, C.A.B. 1996: 54-55.

16091002, IGREJA, Castelo de Neiva, Viana do Castelo, Viana do Castelo (160910). [517515.61, 4608953.46, 112]. Typology: Votive epigraphy. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 78-79.

16090705, CASTRO DE S. SILVESTRE, Cardielos, Viana do Castelo, Viana do Castelo (160907). [521210.90, 4619140.89, 281]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Viana, A. 1926: 90; Couto, A., Viana, T.S., Araújo, J.R. 1935: 27-30; Neves, L.Q. 1965; Moreira, M.A.F.1982: 31-96; Silva, A.C.F. 1986; Almeida, C.A.B. 1990: 183-184; Queiroga, F. 1992; Almeida, C.A.B. 1996: 52-54; Paço, A., Quesado, P. A. s/d: 88.

16091001, MONTE DO CASTELO-MOLDES (Castro de Moldes), Castelo de Neiva, Viana do Castelo, Viana do Castelo (160910). [517889.68, 4607412.36, 183]. Typology: Romanized Hill-fort; Medieval castle. 204

Catalogue

Chronology: Iron Age; Roman Period; Middle Ages. References: Almeida, C.A.B. 1996: 70-75.

16091301, IGREJA, Deocriste, Viana do Castelo, Viana do Castelo (160913). [525219.72, 4615915.57, 0]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1990: 240; Almeida, C.A.B. 1996: 94-95.

16094001, MONTE, Chafé, Viana do Castelo, Viana do Castelo (160940). [519226.10, 4609636.32, 30]. Typology: Pottery Kiln. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 79-80.

16091501, COVA ALTA /OLAS, Lanheses, Viana do Castelo, Viana do Castelo (160915). [527742.38, 4621137.07, 26]. Typology: Roman mines. Chronology: Roman Period. References: Moreira, M.A. 1981: 342-395; Almeida, C.A.B. 1990: 205; Almeida, C.A.B. 1996: 104-105.

16091104, QUINTA DO CARTEADO, Darque, Viana do Castelo, Viana do Castelo (160911). [517940.98, 4614660.02, 70]. Typology: Roman habitat? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 85.

16091502, BOUÇA DO MOISÉS, Lanheses, Viana do Castelo, Viana do Castelo (160915). [526238.84, 4621686.54, 44]. Typology: Roman mines. Chronology: Roman Period. References: Almeida, C.A.B. 1990: 205; Almeida, C.A.B. 1996: 105-106.

16091103, MONTE ARCULO, Darque, Viana do Castelo, Viana do Castelo (160911). [517410.10, 4614288.62, 169]. Typology: Hill-fort. Chronology: Iron Age. Almeida, C.A.B. 1996: 81-83. 16091101, ALTO DO GALEÃO (Faro de Anha), Darque, Viana do Castelo, Viana do Castelo (160911). [517132.56, 4614318.8, 166]. Typology: Watchtower; Medieval castle. Chronology: Iron Age; Roman Period; Middle Ages. References: Neves, L. Q. 1965: 178; Paço, A., Quesado, A.P. 1956: 86-90; Almeida, C.A.F. 1978: 290; Almeida, C.A.B. 1990: 30-31; Almeida, C.A.B. 1996: 83-85.

16091503, ALTO DAS MOURAS, Lanheses, Viana do Castelo, Viana do Castelo (160915). [526465.83, 4622766.82, 65]. Typology: Roman mines. Chronology: Roman Period. References: Almeida, C.A.B. 1990: 205; Almeida, C.A.B. 1996: 106. 16091504, TABONEIRA, Lanheses, Viana do Castelo, Viana do Castelo (160915). [526240.31, 4621285.61, 35]. Typology: Pottery Kiln. Chronology: Roman Period. References: Costa, A.C. 1706: 194; Vieira, J.A. 1887: 229; Almeida, C.A.B. 1990: 204-205; Almeida, C.A.B. 1996: 103-104.

16091102, CAPELA DE S. LOURENÇO, Darque, Viana do Castelo, Viana do Castelo (160911). [514125.19, 4615083.26, 5]. Typology: Roman epigraphy; Mutatio. Chronology: Roman Period. References: Moreira, M.A.F.1982: 68; Almeida, C.A.B. 1990: 30; Almeida, C.A.B. 1996: 86-88.

1064/16091505, QUINTA DA CHAMUSCA, Lanheses, Viana do Castelo, Viana do Castelo (160915). [526537.63, 4622088.58, 60]. Typology: Roman necropolis. Chronology: Roman Period. References: Gonçalves, G. 1986; Almeida, C.A.B. 1990: 204; Almeida, C.A.B. 1996: 101-102.

16091201, CIVIDADE DE DEIÃO (Subidade de Deião), Deião, Viana do Castelo, Viana do Castelo (160912). [525607.30, 4617428.16, 21]. Typology: Hill-fort. Chronology: Iron Age; Roman Period. References: Viana, A. 1926: 90; Viana, A. 1932: 161; Almeida, C.A.B. 1990: 38; Almeida, C.A.B. 1996: 89-91; Almeida, C.A.B. 2003.

16091506, PAÇO, Lanheses, Viana do Castelo, Viana do Castelo (160915). [526658.86, 4620546.96, 16]. Typology: Pottery Kiln. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 102-103.

16091202, IGREJA, Deião, Viana do Castelo, Viana do Castelo (160912). [526254.10, 4617523.03, 10]. Typology: Small Farm. Chronology: Roman Period. References: Vieira, J. A. 887: 231; Almeida, C.A.B. 1990: 36-38; Almeida, C.A.B. 1996: 91-92.

16091507, CIVIDADE DE LANHESES, Lanheses, Viana do Castelo, Viana do Castelo (160915). [527511.38, 4621136.17, 45]. Typology: Hill-fort; Roman necropolis. Chronology: Iron Age; Roman Period. References: Costa, A.C.1706: 194; Viana, A. 1932: 162;; Delgado, M. 1971: 403 e segs.; Castro, M. F. 1980; Almeida, C.A.F., Soeiro, T.; Almeida, C.A.B., Baptista, A.J. 1981: 45; Gonçalves, G. 1986; Almeida, C.A.B. 1990: 189-203; Almeida, C.A.B. 1996: 95-101; Almeida, C.A.B. 2003.

16091303, ALDEIA, Deocriste, Viana do Castelo, Viana do Castelo (160913). [525242.62, 4615977.33, 40]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 94. 16091302, Nª SRª DO CRASTO, Deocriste, Viana do Castelo, Viana do Castelo (160913). [525661.151, 4615300.32, 248]. Typology: Hill-fort. Chronology: Iron Age. References: Viana, A. 1926: 90; Paço, A., Quesado, A. P. 1957: 80-90; Neves, L.Q. 1965: 180; Teixeira, C. et alii 1972-1975; Moreira, M.A.F. 1982: 31-96; Silva 1986; Almeida, C.A.B. 1990: 36; Queiroga, F. 1992; Almeida, C.A.B. 1996: 92-94.

16091701, IGREJA VELHA, Meadela, Viana do Castelo, Viana do Castelo (160917). [504113.03, 4617568.60, 20]. Typology: Roman epigraphy- undetermined. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 107-108.

205

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

16091801, VALE DAS COVAS /MATA DAS CORTAS, Meixedo, Viana do Castelo, Viana do Castelo (160918). [522952.44, 4623679.92, 100]. Typology: Roman mines. Chronology: Roman Period? References: Moreira, M.A.F. 1981: 408; Almeida, C.A.B. 1990: 205; Almeida, C.A.B. 1996: 113.

References: Costa, A.C. 1706: 194; Viana, A. 1926: 90; Viana, A. 1932: 163; Couto, A., Viana, T.S., Araújo, J.R. 1935: 27-30; Paço, A., Quesado, P. A. 1957; Almeida, C.A.F. 1978: 40; Almeida, C.A.F., Soeiro, T., Almeida, C.A.B., Baptista, A. J. 1981; Almeida, C.A.B. 1996: 119-121; Almeida, C.A.B. 1990: 182-183; Paço, A., Quesado, P. A. s/d: 88-89.

16091802, BALTEIRO, Meixedo, Viana do Castelo, Viana do Castelo (160918). [524894.09, 4623131.28, 70]. Typology: Roman treasure; Iron Age warrior. Chronology: Iron Age; Roman Period. References: Guerra, L.F. 1900: 175-177; Vasconcellos, J.L. 1910: 50; Pereira, F.A. 1915:1-16; Hubner 1881: 40-42; Silva, A.C.F. 1984: 29-47; Centeno, R. 1987: 63; Almeida, C.A.B. 1990: 179-182; Almeida, C.A.B. 1996: 109-112.

16092502, PAÇO, Outeiro, Viana do Castelo, Viana do Castelo (160925). [517252.48, 4622121.97, 100]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 123. 16092501, REGO DO CRASTO, Outeiro, Viana do Castelo, Viana do Castelo (160925). [518511.83, 4617437.15, 0]. Typology: Roman habitat? Chronology: Roman Period? References: Paço, A., Quesado, P. A. 1956: 90; Almeida, C.A.B. 1990: 217; Almeida, C.A.B. 1996: 122-123.

16092101, VILA NOVA, Moreira de Geraz do Lima, Viana do Castelo, Viana do Castelo (160921). [526253.31, 4617738.91, 30]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 113-114.

16092605, VILA MEÃ, Perre, Viana do Castelo, Viana do Castelo (160926). [518090.60, 4619502.41, 25]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 132.

16092202, PAÇO, Mujães, Viana do Castelo, Viana do Castelo (160922). [523911.10, 4613197.08, 0]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1996: 118.

16092604, MOLHADOIROS DO GALO (Minadoiros do Galo), Perre, Viana do Castelo, Viana do Castelo (160926160917). [515685.28, 4620576.18, 312]. Typology: Hill-fort. Chronology: Iron Age. References: Viana, A. 1926: 90; Viana, A. 1932: 163; Paço, A., Quesado, P. A. 1956: 86-87; Almeida, C.A.B. 1990: 221-222; Almeida, C.A.B. 1996: 126-127.

16092201, CASTRO DE SANTINHO (Castro de Roques), Mujães; Vila Franca; Subportela, Viana do Castelo, Viana do Castelo (160922). [523029.36, 4614119.43, 269]. Typology: Romanized Hill-fort; Roman treasure. Chronology: Iron Age; Roman Period. References: Sarmento, F.M. 1888:150; Guerra, L.F. 1900: 2-7; Guerra, L.F. 1910; Vasconcelos, J. L. 1917: 116-118; Couto, A., Viana, T.S., Araújo, J.R. 1935: 2730; Cunha, A. R., 1945: 264-268; Neves, L.Q. 1942; Neves, L.Q. 1948: 56-61; Neves, L.Q. 1959: 381-387; Neves, L.Q. 1965: 172-180; Almeida, C. A. F. 1978: 40; Nunes, H. M.B. 1979; Moreira, M. A. F. 1982: 54; Silva, A.C.F. 1983-1984: 124; Almeida, C.A.B. 1996: 114-118; Almeida, C.A.B. 1990: 32-35.

16002601, CASTELHÃO, Perre, Viana do Castelo, Viana do Castelo (160026). [517212.89, 4619376.89, 50]. Typology: Farm; Medieval habitat. Chronology: Roman Period; Middle Ages. References: Viana, A. 1926: 90; Viana, A. 1932: 163; Neves, L.Q. 1948: 5661; Paço, A., Quesado, P. A. 1956: 87-88; Paço, A., Quesado, P. A. 1957: 5-11; Almeida, C.A.B. 1996: 131-132; Almeida, C.A.B. 1990: 218-220; Almeida, C.A.B 2003.

16092301, Nª SRª DO CRASTO, Neiva (S. Romão de), Viana do Castelo, Viana do Castelo (160923). [517073.40, 4610031.81, 106]. Typology: Hill-fort. Chronology: Iron Age. References: Viana, A. 1926: 88-90, 111-113; Viana 1932: 154-167; Almeida, C.A.F. 1978: 40; Moreira, M.A.F. 1982: 31-96; Almeida, C.A.B. 1982: 13-33; Silva 1986; Queiroga 1992; Almeida, C.A.B. 1996: 135-137.

16092602, VIEITO, Perre, Viana do Castelo, Viana do Castelo (160926). [517189.49, 4619500.2, 122]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Viana, A. 1926: 90; Viana, A. 1932: 163; Moreira, M.A.F. 1982: 54; Teixeira, C., Medeiros, A.C. 1972: 39; Paço, A., Quesado, P. A. 1956: 84-86; Paço, A., Quesado, P. A. 1957: 5-11; Almeida, C.A.B. 1990: 217-218; Almeida, C.A.B. 1996: 128-129.

16092401, RASAS, Nogueira, Viana do Castelo, Viana do Castelo (160924). [521551.55, 4621115.74, 170]. Typology: Roman mines. Chronology: Roman Period? References: Almeida, C.A.B. 1990: 206; Almeida, C.A.B. 1996: 122.

16092603, CALVÁRIO, Perre, Viana do Castelo, Viana do Castelo (160926). [518623.78, 4618825.28, 53]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Azevedo, P.A. 1901: 110; Viana, A. 1926: 90; Viana, A. 1932: 163; Cardozo, M. 1937; Paço, A., Quesado, P. A. 1956: 81/82; Almeida, C.A.B. 1996: 124-126, 212-215.

16092402, CASTELO DE S. MARTINHO (Aguieira), Nogueira / S. Marta de Portuzelo, Viana do Castelo, Viana do Castelo (160924). [520553.82, 4622654.86, 470]. Typology: Romanized Hill-fort; Medieval castle. Chronology: Iron Age; Roman Period; Middle Ages. 206

Catalogue

16092703, PAÇÔ, Portela de Susã, Viana do Castelo, Viana do Castelo (160927). [525758.16, 4614036.18, 250]. Typology: Roman habitat? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 135.

tat? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 143.

16092702, IGREJA, Portela de Susã, Viana do Castelo, Viana do Castelo (160927). [525112.02, 4613663.80, 160]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 134-135.

16092906, COUTADA /PEROPESCOÇO, Stª Leocádia de Geraz do Lima, Viana do Castelo, Viana do Castelo (160929). [529087.50, 4614141.44, 250]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1990: 40; Almeida, C.A.B. 1996: 141.

16092701, CRESTO, Portela de Susã, Viana do Castelo, Viana do Castelo (160927). [525762.59, 4612802.56, 200]. Typology: Hill-fort. Chronology: Iron Age; Middle Ages. References: Almeida, C.A.B. 1996: 133-134.

16092907, AGROS, Stª Leocádia de Geraz do Lima, Viana do Castelo, Viana do Castelo (160929). [527374.06, 4614751.64, 350]. Typology: Small Farm? Chronology: Roman Period. References: Almeida, C.A.B. 1990: 40; Almeida, C.A.B. 1996: 142.

16092801, MONTE DOS CASTELOS, S. Marta de Portuzelo, Viana do Castelo, Viana do Castelo (160928). [520537.36, 4620341.73, 177]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Viana, A. 1926: 90; Viana, A. 1932: 163; Paço, A., Quesado, P. A. 1956: 88; Almeida, C.A.B. 1990: 215; Almeida, C.A.B. 1996: 153-154.

16093001, PAÇO, Stª Maria de Geraz do Lima, Viana do Castelo, Viana do Castelo (160930). [527063.59, 4617371.83, 41]. Typology: Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1990: 46-48; Almeida, C.A.B. 1996: 152; Almeida, C.A.B. 2003. 16093002, IGREJA, Stª Maria de Geraz do Lima, Viana do Castelo, Viana do Castelo (160930). [527596.60, 4617003.77, 40]. Typology: Villa. Chronology: Roman Period. References: Almeida, C.A.B. 1990: 42; Almeida, C.A.B. 1996: 146-152.

16093201, IGREJA, Serreleis, Viana do Castelo, Viana do Castelo (160932). [520706.57, 4617751.57, 30]. Typology: Hill-fort. Chronology: Iron Age; Roman Period. References: Almeida, C.A.B. 1990: 241; Almeida, C.A.B. 1996: 154-155; Almeida, C.A.B. 2003.

16093301, PAÇO, Subportela, Viana do Castelo, Viana do Castelo (160933). [523789.13, 4615108.81, 55]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 155-156.

16092901, IGREJA, Stª Leocádia de Geraz do Lima, Viana do Castelo, Viana do Castelo (160929). [527409.69, 4617527.36, 36]. Typology: Roman necropolis; Roman epigraphy- undetermined. Chronology: Roman Period. References: Moreira, M. A. F. 1982: 68; Almeida, C.A.B. 1990: 46; Almeida, C.A.B. 1996: 144-146.

16090001, CITÂNIA DE SANTA LUZIA (Cidade Velha de Santa Luzia), Viana do Castelo, Viana do Castelo, Viana do Castelo (160900). [514005.49, 4617180.2, 226]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Silva, J.P. 1877; Guerra, L.F. 1878; Caldas, J. 1884: 333-335; Cardoso, F. 1897; Guerra, L.F. 1900; Fortes, J. 1904: 3; Fortes, J. 1905: 16/17; Vasconcelos, J.L. 1910: 507-508; Vasconcelos, J.L.1913: 507-508; Pereira, F.A. 1914; Corrêa, A.M. 1924; Déchelette, J.M. 1927: 173; Vitorino, P. 1929: 448; Sarmento, F.M. 1933: 166; Viana, T.S. 1935: 36-37; Lopez, C.F. 1953; Viana A.; Oliveira, M.S. 1954: 5, 12, 19, 36-37; Viana A.; Oliveira, M.S. 1955: 25, 25, 33, 34, 66, 68, 71-78, 541, 547; Viana A. 1955; Balil, A. 1972: 9; Romero, M.A. 1976; Viana A.; Almeida, C.A.F. 1977: 37; Kalb, Ph. 1980: 25-115; Ponte, S. 1981: 149-151; Almeida, C.A.F. 1983: 72; Silva, E.J.L., Marques, J.A.T.M. 1984: 36, 38; Almeida, C.A.B. 1996: 167221; Almeida, C.A.B. 1990: 224-239.

16092902, PESO, Stª Leocádia de Geraz do Lima, Viana do Castelo, Viana do Castelo (160929). [528711.36, 4615681.98, 183]. Typology: Romanized Hill-fort. Chronology: Bronze Age; Iron Age; Roman Period. References: Soeiro, T. 1981; Almeida, C.A.B., Abreu, A.A., Baptista, A.J. 1985; Almeida, C.A.B. 1990: 42-44; Almeida, C.A.B. 1996: 137-141. 16092903, CHOUSO DE MONDIM, Stª Leocádia de Geraz do Lima, Viana do Castelo, Viana do Castelo (160929). [527454.73, 4617835.94, 40]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1990: 40; Almeida, C.A.B. 1996: 144. 16092904, BALTAR, Stª Leocádia de Geraz do Lima, Viana do Castelo, Viana do Castelo (160929). [528711.36, 4615681.98, 0]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1990: 40; Almeida, C.A.B. 1996: 143-144.

16093804, QUINTA DA PORTELA, Vila de Punhe, Viana do Castelo, Viana do Castelo (160938). [522062.47, 4612728.55, 130]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 229.

16092905, PAÇO, Stª Leocádia de Geraz do Lima, Viana do Castelo, Viana do Castelo (160929). [528429.79, 4616729.45, 60]. Typology: Roman habi-

16093802, IGREJA, Vila de Punhe, Viana do Castelo, Viana do Castelo (160938). [522272.63, 4612081.54, 207

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

110]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 227.

[525319.81, 4620295.5, 53]. Typology: Roman mining site; vicus. Chronology: Roman Period. References: Costa, A.C. 1706: 194; Leal, P. 1873-1890: 799; Vieira, J.A. 1887: 229; Viana, A. 1926: 90; Viana, A. 1932: 164; Araújo, J.R. 1934: 9; Couto, A., Viana, T.S., Araújo, J.R. 1935: 27-30; Araujo, J.R. 1959: 85-90; Moreira, M.A.F. 1981: 31-89, 395-423; Moreira, M.A.F. 1982: 54 e segs.; Almeida, C.A.B. 1990: 186187; Almeida, C.A.B. 1996: 159-161; Almeida, C.A.B. 2003.

16093801, COTORINHO, Vila de Punhe, Viana do Castelo, Viana do Castelo (160938). [522894.46, 4612916.20, 226]. Typology: Romanized Hill-fort. Chronology: Iron Age; Roman Period. References: Almeida, C.A.B. 1996: 225-227 . 16093803, QUINTA DO MONTE, Vila de Punhe, Viana do Castelo, Viana do Castelo (160938). [522617.62, 4612699.43, 130]. Typology: Small Farm. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 228-229.

16093901, COVAS (Poço Negro; Poço do Inferno), Vilar de Murteda, Viana do Castelo, Viana do Castelo (160939). [525491.44, 4623997.07, 150]. Typology: Roman mines. Chronology: Roman Period? References: Almeida, C.A.B. 1990: 206; Almeida, C.A.B. 1996: 166-167.

16093502, FONTÃO, Vila Franca, Viana do Castelo, Viana do Castelo (160935). [520947.78, 4614298.09, 60]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 224; Almeida, C.A.B. 2003.

16093903, FOLGADOURO, Vilar de Murteda, Viana do Castelo, Viana do Castelo (160939). [521840.27, 4624910.09, 270]. Typology: Roman mines. Chronology: Roman Period. References: Almeida, C.A.B. 1990: 205; Almeida, C.A.B. 1996: 166.

16093501, IGREJA, Vila Franca do Lima, Viana do Castelo, Viana do Castelo (160935). [521963.82, 4614733.03, 0]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1990: 240; Almeida, C.A.B. 1996: 228-229.

16100202, IGREJA, Candemil, Vila Nova de Cerveira, Viana do Castelo (161002). [525817.10, 4642441.56, 260]. Typology: Milestone. Chronology: Roman Period. References: Almeida, C.A.B. 2000: 21.

16093601, SABARIZ, Vila Fria, Viana do Castelo, Viana do Castelo (160936). [520833.60, 4613804.3, 110]. Typology: Hill-fort. Chronology: Iron Age. References: Viana, A. 1932: 164; Paço, A. 1933: 272-276; Neves, L.Q. 1965: 178; Almeida, C.A.B. 1990: 31-32; Almeida, C.A.B. 1996: 157-159.

16100404, STª LUZIA, Covas, Vila Nova de Cerveira, Viana do Castelo (161004). [524800.37, 4636793.82, 118]. Typology: Villa? Chronology: Roman Period. References: Almeida, C.A.B. 1996: 23-25.

16093701, RASAS, Vila Mou, Viana do Castelo, Viana do Castelo (160937). [525086.62, 4620911.37, 40]. Typology: Roman mines. Chronology: Roman Period. References: Almeida, C.A.B. 1990: 206; Almeida, C.A.B. 1996: 164-165.

16100403, PAGADE (Pias), Covas, Vila Nova de Cerveira, Viana do Castelo (161004). [526527.12, 4637324.36, 160]. Typology: Hill-fort. Chronology: Iron Age. References: Almeida, C.A.B. 1996: 20-21. 16100401, COUÇO DO MONTE FURADO, Covas, Vila Nova de Cerveira, Viana do Castelo (161004). [522771.82, 4636848.73, 90]. Typology: Roman mines. Chronology: Roman Period. References: Oliveira, E.P. 1994: 71-72; Léwis, P.R., Jones, G.D.B. 1970: 169-185.

16093706, MATA DE VILA MOU, Vila Mou, Viana do Castelo, Viana do Castelo (160937). [525390.22, 4619987.21, 35]. Typology: Roman mines. Chronology: Roman Period. References: Almeida, C.A.B. 1990: 206; Almeida, C.A.B. 1996: 165.

16100503, PAÇO, Gondar, Vila Nova de Cerveira, Viana do Castelo (161005). [526515.74, 4640377.70, 260]. Typology: Roman habitat. Chronology: Roman Period. References: Almeida, C.A.B. 1996, Vol. V: 3031.

16093703, IGREJA /PAÇO, Vila Mou, Viana do Castelo, Viana do Castelo (160937). [525090.08, 4619924.47, 30]. Typology: Votive epigraphy. Chronology: Roman Period. References: Guerra, L. F. 1900: 155-177; Pereira. F.A. 1906: 202-209; Araújo, J.R. 1959: 85-90; Almeida, C.A.B. 1996: 161-164; Almeida, C.A.B. 1990: 188-189.

16100502, IGREJA, Gondar, Vila Nova de Cerveira, Viana do Castelo (161005). [527044.79, 4640595.58, 255]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1996, Vol. V: 29

16093704, QUINTA DA TORRE, Vila Mou, Viana do Castelo, Viana do Castelo (160937). [525321.12, 4619925.28, 20]. Typology: Roman necropolis. Chronology: Roman Period. References: Almeida, C.A.B. 1996: 187-188; Araújo, J.R. 1959: 85-90; Almeida, C.A.F.1978:80; Moreira, M.A.F.1982: 63.

16100501, SOBREIRO, Gondar, Vila Nova de Cerveira, Viana do Castelo (161005). [526169.72, 4640499.78, 250]. Typology: Hill-fort. Chronology: Iron Age; Roman Period. References: Almeida, C.A.B. 1996, Vol. V: 28-29.

16093705, CIVIDADE DE VILA MOU, Vila Mou, Viana do Castelo, Viana do Castelo (160937). 208

Catalogue

16100802, LOUREDO, Lobelhe, Vila Nova de Cerveira, Viana do Castelo (161008). [521665.47, 4644401.60, 38]. Typology: Roman necropolis. Chronology: Roman Period. References: Almeida, C.A.B. 1996, Vol. V: 34. 16100701, IGREJA, Loivo, Vila Nova de Cerveira, Viana do Castelo (161007). [520866.67, 4641993.51, 38]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1996, Vol. V: 34. 16101102, IGREJA, Reboreda, Vila Nova de Cerveira, Viana do Castelo (161011). [522882.30, 4645484.89, 65]. Typology: Roman habitat. Chronology: Roman Period. References: References: Almeida, C.A.B. 1996, Vol. V: 40. 16101302, S. SEBASTIÃO, Sopo, Vila Nova de Cerveira, Viana do Castelo (161013). Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1996, Vol. V: 46. 16101301, MATA DOS CRASTOS, Sopo, Vila Nova de Cerveira, Viana do Castelo (161013). [522443.68, 4638574.84, 258]. Typology: Hill-fort. Chronology: Iron Age. References: Almeida, C.A.B. 1996: 44-46. 16101401, MONTORROS, Vila Meã, Vila Nova de Cerveira, Viana do Castelo (161014). [526254.68, 4648673.35, 15]. Typology: Roman habitat? Chronology: Roman Period? References: Almeida, C.A.B. 1996.

209

Appendix I- Maps

Map 1. Entre-Douro-e-Minho region and delimitation of selected areas 210

Appendix I- Maps

Map 2. Hillforts in Entre-Douro-e-Minho region 211

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Map 3. Hillforts of A category

212

Appendix I- Maps

Map 4. Hillforts of B category 213

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Map 5. Hillforts of C category 214

Appendix I- Maps

Map 6. Iron Age hillforts (categories A, B and C) 215

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Map 7. Road network system, vici and road stations 216

Appendix I- Maps

Map 8. Road network system and vici 217

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Map 9. Villae 218

Appendix I- Maps

Map 10. Bracara Augusta peripheries: archeological sites and roman roads 219

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Map 11. Main influence area around Bracara Augusta 220

Appendix I- Maps

Map 12. Roman cadastre, roman roads and archeological sites 221

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Map 13. Roman cadastre with “Couto de Tibães” delimitation

222

Appendix I- Maps

Map 14. Framework of the mountain region

223

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Map 15. Valley and coastline settlement 224

portuguesa, Actas do I Congresso da Associação Portuguesa de Estudos Clássicos. Coimbra, 17-32. Alarcão, J. (2000a). «As paróquias suévicas do território actualmente português». In: Religión, Lengua y Cultura prerromanas de Hispania. Salamanca: Ediciones Universidad Salamanca, 30-59. Alarcão, J. (2000b). «Sobre cantões proto-históricos do Noroeste de Portugal». Santo Tirso Arqueológico, 2ª série, 2-3: 45-50. Alarcão, J. (2003). «A organização social dos povos do Noroeste e Norte da Península Ibérica nas épocas pré-romana e romana». Conimbriga, 42: 5-115. Alarcão, J. (2004). «Notas de arqueologia, epigrafia e toponímia - II». Revista Portuguesa de Arqueologia, 7 (4):193-216. Alarcão, J. (coord.) (1990b). Portugal das origens à romanização. Lisboa: Editorial Presença. Albertos Firmat, M. L. (1974). «El culto de los montes entre Galaicos, Astures y Berones y algunas de las deidades mais significativas». Estudios de Arqueologia Alavesa, 6: 147-157. Albertos Firmat, M. L. (1975). Organizaciones suprafamiliares en la Hispania Antigua.Valladolid (Studia Archaeologica, 37). Albertos Firmat, M. L. (1977). «Perduraciones indígenas en la Galicia romana: los castros, las divinidades y las organizaciones gentilitas en la epigrafía». In: Bimilenario de Lugo. Lugo, 1727. Alcock, S. (1994). Graecia capta: the landscape of Roman Greece. Cambridge. Alcorta Irastorza, E. J. (2007). «Muros, torres y escaleras. Aproximación al modelo constructivo de la muralla romana de Lucus Augusti». In: Actas del Congreso Internacional Murallas de ciudades romanas en el Occidente del Imperio. Lucus Augusti como paradigma, edited by A. Rodríguez Colmenero, and I. Rodá de Llanza. Lugo, 285-311. Alfenim, R. (1995). «A barragem de Aquae Flaviae». Aquae Flaviae, 13. Allan, J. C. (1970). Considerations on the Antiquity of mining in the Iberian Peninsula. London: Royal Anthropological Institute of Great Britain and Ireland. Allen, K. M. S.; W. G. Stanton, and B. W. Zubrow, (eds.) (1990). Interpreting Space: GIS and archaeology. London. Almeida, C. A. B. (1979). «A rede viária do Conventus Bracaraugustanus - via Bracara-Asturicam Quarta». Minia, 1ª série, 2 (3): 61-163. Almeida, C. A. B. (1980). «Via Veteris. Antiga Via Romana?». Revista de Guimarães, vol. III (Actas do Seminário de Arqueologia do Noroeste Penínsular): 151-169. Almeida, C. A. B. (1987). «Carta Arqueológica do Concelho de Esposende». Boletim Cultural de Esposende, 9/10.

References Acuña Castroviejo, F., and M. Cavada Nieto (1971). «Noticias arqueológico-numismaticas del castro Lupario (Rois-Brion, La Coruña)». Cuadernos de Estudios Gallegos, 26 (80): 265-277. Aira Rodríguez, M. J., and P. Ramil Rego (1992). «Contribución al conocimiento de la vegetación tardiglaciar y holocena en el Extremo Norte de la terra chá. Galicia, España». Nova Acta Cientifica Compostelana. Bioloxía, 3: 49-58. Aira Rodríguez, M. J., P. Sáa Otero, and C. Taboada (1989). «Estudios paleobotánicos y edafológicos en yacimentos arqueológicos de Galicia». Arqueoloxía/Investigación, 4. Alarcão, J. (1973). Portugal Romano. Lisboa. Alarcão, J. (1976). Fouilles de Conimbriga.VI. Céramiques diverses et vers. Paris: Mission Archéologique Française au Portugal / Musée Monographique de Conimbriga / De Boccard. Alarcão, J. (1980). «Os problemas da origem e da sobrevivência das villae romanas do Norte de Portugal». Revista de Guimarães, III (Actas do Seminário de Arqueologia do Noroeste Peninsular): 171-193 Alarcão, J. (1988). Roman Portugal. I: Introduction. II: Gazetteer. Warmister: Aris & Phillips, Alarcão, J. (1990a). «Identificação das cidades da Lusitânia Portuguesa e dos seus territórios». In: Les Villes de Lusitanie Romaine. Paris: CNRS, 21-24. Alarcão, J. (1992). «A evolução da Cultura Castreja». Conímbriga, XXXI: 39-71. Alarcão, J. (1995). «Aglomerados urbanos secundários romanos de Entre Douro e Minho». Biblos, LXXI: 387-401. Alarcão, J. (1995-96). «As civitates do Norte de Portugal». Cadernos de Arqueologia, Série II, 12-13: 25-30. Alarcão, J. (1996). «Aglomerados urbanos secundários romanos de Entre-Douro e Minho». In: A Cidade e o Mundo: Romanización e Cambio Social, edited by S. Reboreda Morillo, and P. López Barjá. Xinzo de Limia: Concello de Xinzo de Limia, 169-180. Alarcão, J. (1998a) Domínio romano em Portugal. Lisboa: Europa-América. Alarcão, J. (1998b). «Paisagem rural romana e altomedieval em Portugal». Conímbriga, XXXVII: 89-119. Alarcão, J. (1999a). «As cidades capitais do Norte de Portugal na época romana». In: Los origenes de la ciudad en el Noroeste hispânico. Actas del Congreso Internacional (Lugo, Maio 1996). Lugo, 429-437. Alarcão, J. (1999b). «Populli, castella e gentilitatis». Revista de Guimarães, volume especial (Actas do Congresso de Proto-História Europeia, I): 133-150 Alarcão, J. (1999c). «Raízes históricas da paisagem portuguesa». In: Raízes greco-latinas da cultura 225

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Almeida, C. A. B. (1990) Proto-História e Romanização da bacia inferior do Lima. Estudos Regionais 7/8. Viana do Castelo: Centro de Estudos Regionais. Almeida, C. A. B. (1996). «Povoamento romano do litoral entre Cávado e Minho», PhD dissertation, Faculdade de Letras da Universidade do Porto. Almeida, C. A. B. (1998). «Povoamento Romano do Litoral Minhoto entre o Cávado e o Minho: Esposende». Boletim Cultural de Esposende, 20. Almeida, C. A. B. (2000). A Arqueologia Protohistórica e Romana do Concelho de Vila Nova de Cerveira. Vila Nova de Cerveira: Câmara Municipal de Vila Nova de Cerveira. Almeida, C. A. B. (2003a). «Alterações no povoamento indígena no início da romanização. Ponto da situação no Conventus Bracaraugustanus». Boletín Auriense, 33: 77-93. Almeida, C. A. B. (2003b). «Povoamento romano do litoral minhoto entre Cávado e Minho», PhD dissertation, Edição do Autor. Almeida, C. A. B.; Maia, M. I. S; Moreira, M. M. L.; Baptista, A. J. (1994). «A estação do Bronze Final da Regueira. Vitorino de Piães – Ponte de Lima». Revista da Faculdade de Letras. História, 2ª série, 11: 547-565. Almeida, C. A. B.; Soeiro, T.; Barroca, M. J. (1995). «Estação arqueológica do Castelo de Fraião (Boivão, Valença)». Portugália, Nova Série 16: 311-322. Almeida, C. A. F. (1968) As vias medievais Entre Douro e Minho, (policopiado), Faculdade de Letras do Porto, Porto. Almeida, C. A. F. (1972). «Achados arqueológicos de Martim (Barcelos)». Revista da Faculdade de Letras do Porto, Série História, 2: 319-326. Almeida, C. A. F. (1978) Arquitectura românica de Entre Douro e Minho, Dissertação de Doutoramento (policopiado), Faculdade de Letras do Porto, Porto. Almeida, C. A. F. (1983). «Cultura castreja. Evolução e problemática». Arqueologia, 8: 35-42. Almeida, C. A. F.; Soeiro, T.; Almeida, C.A.B.; Baptista, A.J. (1981). «Escavações arqueológicas em Santo Estêvão da Facha». Arquivo de Ponte de Lima. Álvarez Asorey, R.; Carreño Cáscon, M. C.; Gonzálvez Fernández, E. (2003). Aqua Urbi: Historia do Abastecemento de Auga à cidade de Lugo (época romana – século XX). Lugo (Traballos de Arqueoloxía - Concello de Lugo, I). Alves, F. M. (1934-38) Memórias ArqueológicasHistóricas do Distrito de Bragança, 9 (1934), 10 (1938), 11 (1948), Porto. Amaral, L. C. (1990). «O património fundiário da Sé de Braga entre 1071 e 1108. In: Actas do Congresso Internacional sobre o IX centenário da dedicação da Sé de Braga. Braga: Universidade Católica Portuguesa – Faculdade de Teologia, vol. I, 513-550.

Amaral, L. C. (2007) Formação e desenvolvimento da diocese de Braga no período da reconquista (século IX-1137), Dissertação de Doutoramento em História (policopiado), Faculdade de Letras da Universidade do Porto. Amaral, L. (2007). As moedas das Carvalheiras. Braga: UAUM / NARQ. Amaral, P. (1999). Ocupação e organização do território em torno de Aquae Flaviae». In: Actas da Mesa Redonda Emergência e desenvolvimento das cidades romanas no Norte da Península Ibérica. Porto: Escola Profissional de Arqueologia / IPPAR, 109-113. Araújo, J. R. (1982). «Os Milestones da estrada romana de Braga a Tuy». O Distrito de Braga, 2ª série, 5 (9): 121-246. Arce, J. (1982). El ultimo siglo de la España Romana. Madrid. Argote, J. C. (1732-34) Memórias para a História Eclesiástica do Arcebispado de Braga. Primaz das Hespanhas, Lisboa Occidental, Lisboa, Vols. I-IV. Arias Vilas, F. (1998) “Lucus Augusti e o río Miño: as termas y o seu contorno”, in Rodríguez Colmenero, A. (coord). Actas del Congreso Internacional Los orígenes de la ciudad en el Noroeste hispânico, Lugo, pp.1209-1223. Arias Vilas, F., and Vega Rodríguez (1996). «Las termas romanas de Lugo». In: Actas I Congreso Termalismo Antiguo (La Rioja 1996). Madrid. Arias Vilas, F. e Villa Valdés, A. (2005). «El poblamiento romano en el território de los galaicos lucenses. In: III Coloquio Internacional de Arqueología en Gijón, Unidad y diversidad en el Arco Atlântico en época romana, edited by C. Fernández Ochoa, and P. García Díaz. Gijón. 297-308. Arias Vilas, F., P. Le Roux, and A. Tranoy (1979). Inscriptions romaines de la province de Lugo. Paris: Centre Pierre Paris. Ariño Gil, E. (2005). «La Hispania citerior occidental y la Lusitania Septentrional entre Augusto y los Flavios: el ager per extremitatem mensura comprehensus». In: L’Aquitaine et l´Hispanie septentrional à l’époque julio-claudienne. Organisation et exploitation des espaces provinciaux, Colloque Aquitania,Saint, 11-13 Septembre, 2003. Édition de la Fédération Aquitania, 95-112. Ariño Gil, E., J. M. Gurt i Esparraguerra, and M. A. Martin-Bueno (1994). «Les cadastres romains d`Hispanie: état actuel de la recherche». In: Structures rurales et societés antiques, Actes do Colloque de Corfou (1992), edited by P. N Doukellis, and L. G. Mendoni. Besançon: Annales Littéraires, 309-328 Ariño Gil, E.; Gurt i Esparraguerra, J. M. ; Palet Martínez, J. M. (2004). El pasado presente. Arqueología de los paisajes en la Hispania romana, Barcelona. Publicacions Edicions de 226

References

Universitat de Barcelona (Acta salmanticensia. Estudios históricos & geográficos; 122). Arroyo-Bishop, D. (1995). «Organisation de l`analyse et de l`interprétation en archéologie ». In: Aplicaciones informaticas en arqueologia: teorias y systemas. Bilbao, 1-22. Baena, J., C. Blasco, and F. Quesada (eds.) (1997). Los S.I.G. y el análisis espacial en arqueología. Madrid. Barker, G. W. (1995). «Landscape archaeology in Italy - Goals for the 1990s». In: Settlement and economy in Italy, 1500 BC – AD 1500, edited by N. Christie. Oxford, 1-11. Barker, G. W. e Hodges, R. (eds.) (1981) Archaeology and Italian society. Prehistoric, Roman and medieval studies. BAR, International Series 102. Oxford: BAR. Barker, G.W. (1991). «Aproaches to archaeological Survey». In: Roman Landscapes. Archaeological Survey in the Mediterranean Region, edited by G. W. Barker, and J. A. Lloyd. London, 1-9. Barradas, L. (1956). «Vias romanas das regiões de Chaves e Bragança». Revista de Guimarães, 66 (1-2): 159-241. Barret, J.C. (1997). «Romanization: a critical comment». In: Dialogues in Roman Imperialism. Power, discourse and discrepant experience in the Roman Empire, edited by D. J. Mattingly. JRA Supplementary Series, 23. Portsmouth, 5154. Belino, A. R. (1906). Novas inscrições romanas de Braga. Braga. Bello Diéguez, J. M. (coord.) (1994) Ciudad y torre, Roma y la Ilustración en A Coruña. A Coruña: Ayuntamiento de A Coruña. Bernardes, J. (2007). A ocupação romana na região de Leiria. Faro: Centro de Estudos de Património, Departamento de História, Arqueologia e Património, Universidade do Algarve. Bernardes, J.P., and H. Carvalho (1996). «Aplicação de um S.I.G. à arqueologia romana dos territórios de Bracara Augusta e Collippo». In: Actas do II Congreso Peninsular de Arqueologia. Zamora. Bettencourt, A. M. S. (2000c). Estações da Idade Do Bronze e inícios da Idade do Ferro da Bacia do Cávado (Norte de Portugal). Braga: Instituto de Ciências Sociais, Universidade do Minho. Bettencourt, A. M. S. (1999). «O Homem e a Paisagem no vale do Cávado entre o II e o I milénio a.C.», PhD dissertation, Universidade do Minho. Bettencourt, A. M. S. (2000a). «O vale do Cávado dos finais do III milénio aos meados do I milénio AC. Sequências cronológico-culturais. In: Actas do III Congresso Peninsular de Arqueologia.Vila Real, vol. IV, 79-83. Bettencourt, A. M. S. (2000b). O povoado de S. Julião, Vila Verde, Norte de Portugal, nos finais da Bronze Age e na transição para a Idade do Ferro. Cadernos de Arqueologia. Monografias 10. Braga.

Bettencourt, A. M. S. (2003). «Manifestações simbólicas da Proto-História do Norte de Portugal: a estátua sedente de Roriz (Barcelos)». Conímbriga, XXXXII: 141-151. Bettencourt, A. M. S. (2005). «El Bronce Final y la primera Edad del Hierro en el noroeste de Portugal». In: Actas do II Coloquio de Arqueología en la Cuenca del Navia. La Génesis del Habitat Fortificado en el Norte Peninsular: los Castros en el Tránsito de la Edad del Bronce e la Edad del Hierro. Navia: Parque Histórico del Navia, Servicio de Patrimonio Histórico e Consejería de Cultura del Principado de Asturias. Bettencourt, A. M. S., A. Dinis, C. Cruz, I. S. Silva, A. Alves, and D. Pereira (2003). «A reconstituição paleoambiental do Entre Douro e Minho durante o Quaternário Recente (III e II milénios BC): um projecto multidisciplinar». Ciências da Terra, 5. Bettencourt; A. M. S., and I. Figueiral (In press). «Estratégias de exploração do espaço no EntreDouro e Minho desde os finais do IV aos meados do I Milénios a.C.». In Simpósio “Paisagem, Recursos e estratégias de exploração do espaço no Noroeste peninsular da Pré Historia à Idade Média. Actas do IV Congresso Peninsular de Arqueologia. Faro. Bettencourt; A. M. S., I. Figueiral, and A. P. Dinis (In press). «A ocupação do território e a exploração de recursos durante a Pré-História recente do Noroeste de Portugal». In Simpósio “Paisagem, Recursos e estratégias de exploração do espaço no Noroeste peninsular da Pré Historia à Idade Média. Actas do IV Congresso Peninsular de Arqueologia. Faro. Blázquez Martínez, J. M. (1957) «Aportaciones al estudio de las religiones primitivas de España». Arquivo Español de Arqueologia, 30 (15): 15-86. Blázquez Martínez, J. M. (1962). Religiones primitivas de Hispania, I. Fuentes literarias y epigráficas. Roma. Blázquez Martínez, J. M. (1970). «Las religiones indígenas del área Noroeste de la Península Ibérica en relación con Roma». In: Legio VII Gemina. Léon, 63-77. Blázquez Martínez, J. M. (1977). «La Romanización del NO de la Península Ibérica». In: Bimilenario de Lugo. Lugo, 65-81. Bosch-Gimpera, P. (1921). «Los Celtas y la civilización céltica en la Península Ibérica». Boletin de la Sociedad Española de Excursiones, Ano 29: 248-300. Bouhier, A. (1979). La Galice. Essai géographique d’analyse et d’interprétation d’une vieux complexe agraire. Vendée: La Roche-Sur-Yon. Braga, C. (2010). «Rituais funerários em Bracara Augusta: o novo núcleo de necrópole da Via XVII». Master dissertation, Universidade do Minho. http://hdl.handle.net/1822/13913. 227

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Braga, M. A. C. (1988). «Arenas e depósitos Associados da bacia de Drenagem do rio Cávado (Portugal) – Contribuição para o estudo da arenização», PhD dissertation, Universidade do Minho. Broise, P. (1976). «L’urbanisme vicinal aux confins de la viennoise et de la Séguanaise». Aufstieg und Niedergang, 2-3. Brulet, R. (1994). «Les agglomérations secondaires de Wallonie et du Grand Duché de Luxemburg». In: Les agglomérations secondaires. La Gaule Belgique, les Germanies et l’Occident romain, Actes du Colloque de BliesbruckReinheim/Bitche (21-24 Oct. 1992), edited by J.P. Petit, M. Mangin, and Ph. Brunella. Paris, 122-135. Buchsenschutz, O. (1994). «Les habitats groupés a La Tène moyenne et finale». In: Les agglomérations secondaires. La Gaule Belgique, les Germanies et l’Occident romain, Actes du Colloque de Bliesbruck-Reinheim/Bitche (21-24 Oct. 1992), edited by J.P. Petit, M. Mangin, and Ph. Brunella. Paris, 149-152. Burnham, B. C. e Wacher, J.S. (1990). The small towns of Roman Britain. Berkeley-Los Angeles: University of California Press. Burón Alvarez, M. (1997). «El trazado urbano en las proximidades del Foro en Asturica Augusta. La casa del pavimento de opus signinum». Memorias. Arqueología en Castilla y León, 2: 289-312. Burón Alvarez, M. (1998). «El Trazado Urbano de Asturica Augusta. Génesis y Evolución. In: Nuevos Elementos de Ingeniería Romana. III Congreso de las Obras Romanas. Astorga: Junta de Castilla y Léon – Colegio de Ingenieros T. de O. P., 289-312. Burón Alvarez, M. e García, V. (2000). «Las Termas Menores de Asturica Augusta». In: Actas del Congreso Internacional: Los Orígenes de la ciudad en el Noroeste Hispánico, Lugo 15-18 de Mayo de 1996. Lugo, 207-214. Caamaño Gesto, J.M. (1977-78). «Las Mansiones de la via 18 en su tramo orensano». Gallaecia, 3/4: 109-136. Caamaño Gesto, J.M. (1995-96). «El trazado de la via 18 del Itinerario de Antonino en Galicia». Cadernos de Arqueologia, Série II, 12-13: 4587. Capela, M. (1895). Miliários do Conventus Bracaraugustanus em Portugal. Porto. Carballo Arceo, J. (1986). Povoamento castrexo e romano da Terra de Trasdeza. A Coruña: Xunta de Galicia. Carballo Arceo, J. (1989). Catálogo de los materiales arqueolóxicos do Museo do Castro de Santa Trega: Idade do Ferro. Pontevedra. Carballo Arceo, J. (1990) “Los Castros de la cuenca media del río Ulla y sus relaciones con el medio físico”, Trabajos de Prehistoria, 47, pp. 161-199.

Carballo Arceo, J. (1998). «A agricultura en Castrovite (Orazo, A Estrada) durante a Idade do Ferro». A Estrada. Miscelánea Histórica e Cultural,1: 925. Cardoso, M. (1929). «Jóias arcaicas encontradas em Portugal». Nós, 11 (72): 207¬218. Cardoso, M. (1932). «A última descoberta arqueológica na citânia de Briteiros e a interpretação da “Pedra Formosa”. Revista de Guimarães, 42 (3-4): 7-25. Cardoso, M. (1934). «A Pedra Formosa da citânia de Briteiros e a sua interpretação arqueológica». Brotéria, 18 (3). Cardoso, M. (1937). «Citânia de Briteiros. Alguns aspectos etnográficos e sociais da nossa Protohistória». Revista de Guimarães, 47 (3-4): 228240. Cardoso, M. (1938). «Citânia de Briteiros. Alguns aspectos etnográficos e sociais da nossa Protohistória». Revista de Guimarães, 48 (1-3) (4): 161-169, 218-236. Cardoso, M. (1942). «Uma notável peça de joalharia primitiva». Anais da Faculdade de Ciências do Porto, 27 (2): 89-100. Cardoso, M. (1944). «Novo achado de jóias préromanas». Revista de Guimarães, 54 (1-2): 1928. Cardoso, M. (1946a). «Arquitectura citaniense. (O problema das casas com tecto em abóbada)». Minia, 1 (3-4): 243-257. Cardoso, M. (1946b). «O “monumento funerário” da citânia». Revista de Guimarães, 56 (3-4): 289308. Cardoso, M. (1952). «A organização defensiva dos castros no Norte de Portugal e a sua romanização». Arquivo Histórico Militar, 22: 112-152. Cardoso, M. (1959). «Die “Castros” in Norden Portugals». Revista de Guimarães, 69 (1-2): 417-439. Cardoso, M. (1962). «Alguns problemas da cultura dos castros no Norte de Portugal». XXVI CLEPC, 2 (5): 391-423. Cardoso, M. (1965). «A romanização do noroeste da Península Ibérica». Ethnos, 4: 72-82. Cardoso, M. (1966). «La romanizzazione del NorOuest della Penisola Ispanica». In: Atti VI Congreso International delle Science Pre e Protostoriche, 3. Roma, 53-58. Cardoso, M. (1972). Catálogo do Museu Martins Sarmento. Secção de Epigrafia e de Escultura Antiga. Guimarães. Caro Baroja, J. (1970). «Organización social de los pueblos del Norte de la Península Ibérica en la Antigüedad». In: Legio VII Gemina, 9-62. Carreté, J-M., S. Keay, and M. Millett (1995). A Roman Provincial Capital and its Hinterland. The Survey of the territory of Tarragona, Spain, 1985-1990. JRA, suppl. 15. Ann Arbor. Carrocera Fernández, E. (1995). «El territorio de los astures: los castros». In: Astures. Pueblos y 228

References

culturas en la frontera del Imperio Romano. Gijón, 53-65. Carvalho, H., and J. Bernardes (1995a). «The use of GIS in archaeological site analysis and classification: two case-studies in Portugal». In: Angewandte Geographische Informationsverarbeitung (AGIT 95). Salzburg. Carvalho, H., and J. Bernardes (1995b). «GIS and archae-ology: two case-studies in Portugal». In: European Association os Archaeologists-First Meeting. Santiago de Compostela. Carvalho, H., and J. Bernardes (2000). «O povoamento romano de Bracara Augusta e Collippo: exemplos de aplicação de um S.I.G.». In: Actas do 3º Congresso Peninsular de Arqueologia. X. Sistemas de Informação Arqueológica-SIG's aplicados à Arqueologia da Península Ibérica. Porto, 125-131. Carvalho, H. (2000). «Parcelamento rural de Bracara Augusta: questões metodológicas». In: Actas do 3º Congresso Peninsular de Arqueologia. VI. Arqueologia da Antiguidade na Península Ibérica. Porto, 143-146. Carvalho, H. (2008). «Organisation cadastrale autour de Bracara Augusta, Braga (Portugal)». Dialogues d’ Histoire Ancienne, 34/1: 155-160. Carvalho, H. (2008b). «O povoamento romano na fachada ocidental do Conventus Bracarensis», PhD dissertation, Universidade do Minho, II vols, http://hdl.handle.net/1822/8755 Carvalho, H; e Ribeiro, M.C. (2009). «Aménagement et exploitation des espaces rural et urbain: Bracara Augusta et son territoire (un questionnaire préalable)». In: Aménagement et exploitation des zones humides depuis l’Antiquité. Approches comparées en Europe méditerranéenne et continentale, Clermont-Ferrand: Maison des Sciences de l’Homme), (in print). Carvalho, H., and F. A. Mendes (2009-2010) «Centuriazione e evoluzione degli assetti agrari intorno alla città di Bracara Augusta (Hispania Citerior Tarraconensis)». Agri Centuriati, vol 6: 193-204. Carvalho, H. (2010b). «Paysages ruraux dans le Conventus Bracaraugustanus: bilan et perspectives». In: Colloque Ager IX Paysages ruraux et territoires dans les cités de l’Occident romain, Barcelona, 25-27 Mars, 2010. Carvalho, H. (2012). «Marcadores da paisagem e intervenção cadastral no território próximo de Bracara Augusta (Hispania Citerior Tarraconensis)»., Archivo Español de Arqueología, 85: 129-146 Carvalho, H. (2010a). «A construção do espaço ibérico em época romana. Conceitos, escalas e modelos de desenvolvimento». Revista de História das Ideias, vol. 31: 7-23. Carvalho, H. (2013) «Modelos de povoamento romano no Noroeste da Península Ibérica: marcadores de desenvolvimento, escalas, processos”,

Simpósio da MAE-USP, Universidade de S. Paulo, 28 de Novembro de 2013 (in print). Carvalho, H., and M. Cruz, (2014). «Peuplement et organisation du paysage dans la partie occidentale du Conventus Bracarensis». In: Atti del IV Convegno Internazionale di Studi Veleiati (Veleia-Lugagnano Val d’Arda, 20-21 Settembre 2013), edited by Dall’Aglio, Franceschelli, Maganzani. Bologna, Ante Quem, p. 487-494. Carvalho, H., and M. Cruz (2013). «Igreja Velha de Cantelães, Vieira do Minho, Portugal. Une nouvelle énigme du 1er siècle ap. J.-C. dans le nord-ouest de l’Hispanie». In: XVIIIth International Congress of Classical Archaeology- Centre and Periphery in tne ancient world, Mérida 13-17 May 2013 (in print). Carvalho, H. (2010). «A construção do espaço ibérico em época romana. Conceitos, escalas e modelos de desenvolvimento». Revista de História das Ideias, vol. 31: 7-23 Carvalho, H.. 2010 (in print). «Paysages ruraux dans le Conventus Bracaraugustanus: bilan et perspectives». In: Paysages ruraux et territoires dans les cités de l’Occident romain, Colloque Ager IX, Barcelona, 25-27 de Março 2010. (no prelo) Carvalho, H., J. Encarnação, M. Martins, and C. Cunha, (2006). «Altar romano encontrado em Braga», Forum, 40: 31-42. Carvalho, H., and M. Martins. «Marcadores da paisagem e organização do território». In: Congresso comemorativo do nascimento de Alberto Sampaio, Guimarães, 2008, (in print) Carvalho, J. R. (1997). «Comportamentos morais e estruturas sociais numa paróquia de antigo regime (Soure, 1680-1720). Reconstituições, interpretações e metodologias», PhD dissertation, Faculdade de Letras da Universidade de Coimbra. Carvalho, P. C. (2006). «Cova da Beira. Ocupação e exploração do território na época romana», PhD dissertation, Faculdade de Letras da Universidade de Coimbra. Casal Garcia, R., and P. Blanco Sanmartín, (2005). «O mundo romano: Galicia romana». In: Arte e cultura de Galicia e Norte de Portugal: Arqueoloxía, edited by Idalgo Cuñarro. Vigo: Nova Galicia, vol. II, 109-148. Castagnoli, F. (1958). La ricerche sui della centuriazone, Roma. Cavada Nieto, M. (1972). «Hallazgos monetários de Galicia». BSAA, 28: 211-248. Centeno, R. (1976-77). «O tesouro monetário do castro de Romariz (Portugal)». Sautuola, 2: 209-219. Centeno, R. (1977). «O tesouro de denarii do Alto do Corgo (concelho de Valença)». Conímbriga, 16: 93-99. Centeno, R. (1978). «O tesouro de Aurei do Norte de Portugal», Nummus, 2ª série, 1: 37-98. Centeno, R. (1987). Circulação monetária no Noroeste de Hispânia até 192. Anexos Nummus, 1. Porto. 229

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Chastagnol, A. (1995). La Gaule romaine et le droit latin. Recherches sur l`histoire administrative et sur la romanisation des habitants. Scripta varia 3. Lyon. Chavarría Arnau, A. (2007). El final de las villae en Hispânia (siglos IV-VII D.C.). Turnhout: Brepols Publishers,. Chevallier, R. (1997). Les voies romaines, Paris. Chouquer, G. (1995). «Parcellaires, cadastres et paysages». Actualité de l’Archéologie. Revue Archéologique du Centre de la France, 205-230. Chouquer, G. (2000). L`étude des paysages. Essais sur leurs formes et leur histoire. Paris: Errance. Chouquer, G. (dir.) (1996). Les formes du paysage. II. Paris: Errance. Chouquer, G. (dir.) (1997). Les formes du paysage. III. Paris: Errance. Chouquer, G., M Clavel-Levêque, F. Favory, and J.-P. Vallat (1983). Structures agraires en Italie centro-méridionale. Cadastres et paysages ruraux. Coll. De L´École Française de Rome nº100. Paris-Rome: EFR-De Boccard. Chouquer, G. e Favory, F. (1991) Les paysages de l`Antiquité. Terres et cadastres de l`occident romain. Paris: Errance. Chouquer, G., and F. Favory (1992). Les arpenteurs romains. Théorie et pratique. Paris: Errance. CIL -Corpus Inscriptionum Latinarum , II, supplementum. Berlim (Hübner 1892). Clavel-Levêque, M. (1983) (éd.). Cadastres et espaces rurales. Approches et réalités antiques. Actes de la table ronde de Besançon (mai 1980). Paris: CNRS. Clavel-Levêque, M. (1992). Centuriation, géométrie et harmonie. Le cas du Biterrois». In: Mathématiques dans l`Antiquité, edited by J.-Y. Guillaumin. Saint-Étienne: Publ. Univ. SaintÉtienne, 161-184. Clavel-Levêque, M. (1995). Atlas des cadastres de Gaule. 1. Le réseau centurié Béziers B. Besançon: Annales Littéraires. Clavel-Levêque, M. (1998). «Le réseau colonial: Béziers C». In: Atlas historique des cadastres d`Europe, edited by M. Clavel-Levêque, and A Vignot. Luxembourg: OPOCE. Clavel-Levêque, M., I. Jouffroy, and A.Vignot (éd.) (1994). De la terre au ciel.I. Paysages et cadastres antiques. Besançon: Annales Littéraires, Clavel-Levêque, M., and R. Plana Mallart (éd.) (1995). Cité et territoire. 1er colloque européen (Béziers 1994). Besançon: Annales Littéraires. Clavel-Levêque, M., and A. Vignot (éd.) (1998). Atlas historique des cadastres d`Europe. Luxembourg: OPOCE. Costa, A. J. da (ed.) (1965-1990). Liber Fidei Sanctae Bracarensis Ecclesiae. Braga: Junta Distrital de Braga, 3 tomos. Costa, L. (1979). Igreja de S. Vítor. Elementos para a história deste templo. Braga.

Criado Boado, F. (1993). «Visibilidad e interpretación del registro arqueológic». Trabajos de Prehistoria, 50: 39-56. Cruz, M. R. M. D. (2001). «Vidros romanos de Bracara Augusta», Master dissertation, Universidade do Minho. Cunha, A. C. (2004). «O património rural do Cabido da Sé de Braga nos finais do século XV», Master dissertation, Universidade do Minho. Cunha, A. R. (1975). «Trepando aos montes». O Distrito de Braga, 1 (3-4): 485-535. Cunha, A., J. Encarnação, and F. S. Lemos (2005). «Ara aos Lares Viales, de Bracara Augusta». Forum, 37: 147-155. Cunha, R. (1634-35). História Eclesiástica do Arcebispado de Braga. Braga, 2 vols. Curchin, L. A. (1991). Roman Spain. Conquest and assimilation. London and New York: Routledge. Curchin, L.A. (1986). «Vici and pagi in Roman Spain». Revue des Etudes Anciennes, 87 (1-2): 327-343. David, P. (1947). Études historiques sur la Galice et le Portugal du VI au XII siècle. Lisboa. Decramer, L.; Elhaj, R.; Hilton, R. ; Plas, A. (2007). «Approche géométrique des centuriations romaines. Les nouvelles bornes du Bled Segui». Histoire et Mesure, 17 (1-2). http:// histoiremesure.revues.or/document903.html Delgado, M., and F. S. Lemos (1985). «Zona das Carvalheiras. Notícia das campanhas de escavação de 1984 e 1985». Cadernos de Arqueologia, II, 2: 159-176. Delgado, M., and F. S. Lemos (1986). «Zona das Carvalheiras: notícia das campanhas de escavação de 1985». Cadernos de Arqueologia, II, 3: 151-167. Dias, A. R. (1903). «Notícias arqueológicas extraídas do “Portugal Antigo e Moderno” de Pinho Leal, com algumas notas e indicações». Boletim da Real Associação dos Arquitectos e Arqueólogos Portugueses. Dias, L. A. (1997). Tongobriga. Lisboa: Instituto Português do Património Arquitectónico. Dias, L. A. (1999). «A urbanização do Noroeste Peninsular: o caso de Tongobriga». In: Emergência e Desenvolvimento das cidades romanas no Norte da Península Ibérica. Actas da Mesa Redonda. Porto, 77-107. Diáz-Fierros Viqueira, F. (1996). «Clima e solos de Galicia na época romana». In: Lucus Augusti. I. El amanhecer de una ciudad, edited by A. Rodríguez Colmenero. Lugo: Fundación Pedro Barrié de la Maza, 9-42. Diáz-Fierros Viqueira, F., M. Aira Rodríguez, and P. Ramil Rego (1994). História e arqueologia da paisagem. Serra do Bustelo (Braga, Portugal). Relatório final, (policopiado). Diego Santos, F. (1975). «Die Integration Nord und Nordwestspanien als römischen Provinz». In: Der Reichspolitik des Augustos. Aufstieg und Niedergang der römischen Welt, 2. Berlin, 523571. 230

References

Dinis, A. P. (1993). «Ordenamento do território da bacia do Ave no I milénio a. C.», Master dissertation, Faculdade de Letras da Universidade do Porto. Dinis, A. P. (2001). O povoamento da Idade do Ferro do Crastoeiro (Mondim de Basto, Norte de Portugal). Cadernos de Arqueologia – Monografias, 13. Braga: Universidade do Minho. Domergue, C. (1970a). «Introduction à l’étude des mines d’or du Nord-Ouest de la Péninsule Ibérique dans l’antiquité». In: Legio VII Gemina. León, 255-286. Domergue, C. (1970b). «Les exploitations aurifères du Nord-Ouest de la Péninsula Ibérique sous l’occupation romaine». La minería hispana e iberoamericana, 1: 151-193. Domergue, C. (1986). «Dix-huit ans de recherche (1968-1986) sur les mines d’or romaines du nord-ouest de la Péninsule Ibérique». In: I Congreso Internacional Astorga Romana II. Astorga, 7-101. Domergue, C. (1987). Catalogue des mines et des fonderies antiques de la Péninsule Ibérique. Madrid: Publication de la Casa de Velásquez. Domergue, C. (1990). Les mines de la Péninsule Ibérique dans l’antiquité romaine. Roma: École Française de Rome Domergue, C., and G. Herail (1978). Mines d’or romaines d’Espagne. Le district de la Valduerna (Léon). Toulouse: Université de Tolouse. Domergue, C., and Th. Martin (1977). Minas de oro romanas de la provincia de León II. Excavaciones Arqueológicas en España 94. Madrid. Domergue, C., and P. Sillières (1977). Minas de Oro romanas en la provincia de León I. Excavaciones Arqueológicas en España 93. Madrid. Dopico Cainzos, M. D. (1986). «Los conventus iuridici. Origen, cronología y naturaleza histórica». Gerion, 4. Dopico Cainzos, M. D. (1988). La Tabula Lougeiorum. Estudios sobre la implantación romana en Hispania. Vitoria. Dousteyssier, B., and F. Trément (2006-2007). «Des "grands" et des "petits" domini? "Grandes" et "petites” villae en Gaule Aquitaine. Le cas de la cité des Arvernes"». Revue archéologique du Centre de la France, Tome 45-46. http://racf.revues.org/index711.html. Drinkwater, J.F. (1983). Roman Gaul. The Three Provinces, 58 BC-AD 260. London. Dyson, S.L. (1983). The Roman Villas of Buccino. Wesleyan University excavation in Buccino, Italy 1969-72. BAR IS 187. Oxford. Eco, H. (1999). Kant e o Ornitorrinco. Lisboa: Difel. Elena Garrido, A., R. Mar, and M. Martins (2007). A Fonte do Ídolo: análise, interpretação e reconstituição do santuário. Bracara Augusta.

Escavações Arqueológicas, 4. Braga: UAUM/NARQ. Encarnação, J. (1970). «Lápides e divindades indígenas no Museu de Guimarães». Revista de Guimarães, 80 (3-4): 207-238. Encarnação, J. (1975). Divindades indígenas sob o domínio romano em Portugal. Lisboa. Encarnação, J. (1995). «Panorâmica e problemática geral da epigrafia rupestre em Portugal». In: Saxa Scripta (Inscriciones en Roca). Actas del Simposio Internacional Ibero-Itálico sobre epígrafia rupestre, edited by A. Rodríguez Colmenero, and L. Gasperini. Santiago de Compostela y Norte de Portugal, 29 de junio a 4 de julio de 1992. Anejos de Larouco, 2. A Coruña, 261-278. Encarnação, J. (1995-1996). «Miliários da Geira: informação e propaganda». Cadernos de Arqueologia, 12-13: 39-43. Esparza Arroyo, A. (1986). Los castros de la Edad del Hierro del Noreste de Zamora. Zamora: Diputación Provincial de Zamora. Estefania Alvarez, M.D. (1960). «Vias romanas da Galicia». Zephyrus, 11: 5-103. Étienne, R. (1958). Le culte impérial dans la Péninsule Ibérique d’Auguste à Dioclétien. Paris. Fabião, C. (1992). «O povoamento proto-histórico e romano». In História de Portugal, edited by J. Mattoso. Lisboa: Círculo dos Leitores, vol. I, 76-299. Fabião, C. (2001). «Mundo indígena, romanos e sociedade provincial romana e sociedade provincial romana: sobre a percepção arqueológica da mudança». Era. Arqueologia, 3: 108-131. Fabião, C. (2005). «Caminhos do Atlântico Romano: evidências e perplexidades». In: III Coloquio Internacional de Arqueología en Gijón. Unidad y diversidad en el Arco Atlântico en época romana, edited by C. Fernández Ochoa, and P. García Díaz. Gijon, pp. 83-86. Fabião, C. (2006). A Herança Romana em Portugal, Lisboa: CTT Correios de Portugal. Fabre, G. (1970). «Le tissu urbain dans le nord-ouest de la Péninsule Ibérique». Latomus, 29: 314-339. Favory, F. (1983). «Propositions pour une modélisation des cadastres ruraux antiques». In: Cadastres et espaces rurales. Approches et réalités antiques. Actes de la table ronde de Besançon (mai 1980), edited by M. Clavel-Levêque. Paris: CNRS, 51135. Favory, F., G. Chouquer, and M. Clavel-Levêque (1982). «Cadastres, occupation du sol et paysages agraires antiques». Annales. Économies, Societés, Civilisations, 37 (5): 847-882. Favory, F., and J.-L. Fiches (éd.) (1994). Les campagnes de la France méditerranéenne dans l`Antiquité et le haut Moyen Âge: études microrégionales. Paris: Éd. De la Maison des Sciences de L´Homme. 231

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Favory, F. ; J.-L. Fiches, J.-L. Girardot (1987-88). «L’analyse des donnés appliquée à la typologie des sites gallo-romains dans le Beaucairois (Gard): matériel de prospection et environnement paysager. Essai méthodologique». Gallia, 45: 67-85. Favory, F.; A. Gonzales, J.-Y Guillaumin, and Ph. Robin (1995). «Témoignages antiques sur le bornage dans le monde romain II». Revue Archéologique du Centre de la France, 34: 263265. Favory, F., A. Gonzales, J.-Y. Guillaumin, and Ph. Robin (1996). «Témoignages antiques sur le bornage dans le monde romain III». Revue Archéologique du Centre de la France, 35: 203216. Favory, F., A. Gonzales, J.-Y. Guillaumin, and Ph. Robin (1997). «Témoignages antiques sur le bornage dans le monde romain III». Revue Archéologique du Centre de la France, 36, 203209. Favory, F., A. Gonzales, and Ph. Robin (1994). «Témoignages antiques sur le bornage dans le monde romain». Revue Archéologique du Centre de la France, 33: 214-238. Favory, F. e Van der Leeuw, S. (1998). «Archaeomedes, la dynamique spatiotemporelle de l’habitat antique dans la vallée du Rhône: bilan et perspectives». Revue Archéologique de Narbonnaise, 31: 257-298. Fernandes, A. A. (1968). Paróquias suevas e dioceses visigóticas. Separata do Arquivo do Alto Minho, 14-16. Viana do Castelo. Fernandes, A. A. (1997). Paróquias suévicas e dioceses visigóticas. Arouca. Fernández Ochoa, C. (1982). Asturias en época romana. Madrid. Fernández Ochoa, C. (1983-84). «Poblamiento rural en Asturias durante la época romana». Portugalia, IV-V: 231-243. Fernández Ochoa, C. (1995b). «Astures y Roma: la configuración del território». In: Astures. Pueblos y culturas en la frontera del Imperio Romano. Gijón, 99-111. Fernández Ochoa, C. (1997). La muralla romana de Gijón (Asturias). Madrid. Fernández Ochoa, C. (2003). El lenguaje de las piedras. La recuperación del patrimonio arqueológico de Gijón. Gijón: G.E.A. Distribuciones Gráficas 2000 / Gran Enciclopedia Asturiana. Fernández Ochoa, C., and F. Gil Sendino (2004). «La Villa romana de Veranes. El complejo rural Tardorromano y propuesta de estudio del territorio». Archivo Español de Arqveología, 77, 197-219. Fernández Ochoa, C., P. García Díaz, and F. Gil Sendino (1996). «La investigación sobre Gijón y su consejo en época romana». In: Los finisterres altânticos en la Antigüedad. Época Prerromana y Romana. Actas del Coloquio

Internacional, edited by C. Férnandez Ochoa. Gijón: Electa, 157-164. Fernández Ochoa, C., and A. Morillo Cerdán (1999). La tierra de los astures. Nuevas perspectivas sobre la implantación romana en la antigua Asturia. Ed. Trea, S.L. Fernández Ochoa, C., A. Morillo Cerdán, M. Zarzalejos Prieto, and V. García Entero (1997). «Las termas romanas de Hispania: Balance historiográfico y perspectivas de investigación». In: Termalismo Antiguo. I Congreso Peninsular, edited by Pérex Agorreta. Madrid, 381-389. Fernández Ochoa, C., P. Garcia Díaz, and F. Gil Sendino (2003). «Gijón, enclave marítimo en la ruta comercial cantábrica. Evidencias arqueológicas e hipótesis sobre el puerto romano y los embarcaderos antíguos». In: Gijón Puerto Romano. Navegación y comercio en el Cantábrico durante la Antigüedad, edited by C. Fernández Ochoa. Gijón, 97-117. Fernández Ochoa, C., A. Morillo Cerdán, M. Zarzalejos Prieto (2000). «Grandes conjuntos termales públicos en Hispania». In: Actas Coloquio Internacional: Termas Romanas en el Occidente del Imperio. Gijón, 59-72. Fernández-Posse, M.D. (1998). La investigación Protohistórica en la Meseta y Galicia. Madrid: Síntesis. Fernández-Posse, M. D., and F. J. Sánchez-Palencia (1996). «Consideraciones sobre la estrutura social y el territorio en la Asturia prerromana y romana». In: Los finisterres altânticos en la Antigüedad. Época Prerromana y Romana. Actas del Coloquio Internacional, edited by C. Férnandez Ochoa. Gijón: Electa, 171-180. Ferreira, N., G. Dias, and C.A.P. Meireles (2000) Carta geológica de Portugal (1/50.000). Notícia explicativa da Folha 5-D, Braga. Lisboa: Departamento de Geologia, Instituto Geológico e Mineiro. Figueiral, I. (1990). «Le Nord-Ouest de Portugal et les modifications de l’écosystème, du Bronze Final à l’époque romaine, d’après l’anthrocoanalyse des sites archéologiques», PhD dissertation, Université de Montpellier II. Figueiral, I. (1993). «Charcoal analyses and the vegetational evolution of North – West Portugal». Oxford Journal of Archaeology, 12 (2), 209-222. Figueiral, I. (2000). «O povoado de S. Julião (Vila Verde, Braga): contributos da antracologia». In: O povoado de S. Julião, Vila Verde, Norte de Portugal, nos finais da Bronze Age e na transição para a Idade do Ferro, edited by A. Bettencourt. Cadernos de Arqueologia, Monografias, 10. Braga, 149-156. Figueiral, I., and A. M. S. Bettencourt (2004). «Middle/Late Bronze Age plant communities, and their exploitation, in the Cávado Bassin (NW Portugal) as shown by charcoal analysis: the significance and co-occurrence of Quercus (de232

References

ciduous)–Fabaceae». Vegetation History and Archaeobotanic, 13: 219-232. Figueiral, I., and C. Carcaillet, (2005). «A review of Late Pleistocene and Holocene biogeography of highland Mediterranean pines (painus taype sylvestris) in Portugal, based on wood charcoal». Quaternary Science Review, 24: 2466-2476. Finley, M. (1977). Economy and society in ancient Greece. London. Fontes, L., and A. Roriz, (2007). Património Arqueológico e Arquitectónico de Vieira do Minho. Roteiros. Vieira do Minho: Município de Vieira do Minho. Fontes, L. (1990). «Escavações arqueológicas na antiga igreja de Dume». In: Actas do Congresso Internacional sobre o IX centenário da dedicação da Sé de Braga, Universidade Católica Portuguesa – Faculdade de Teologia. Braga, vol. I, 147-169. Fontes, L. (1991-92). «Salvamento Arqueológico de Dume (Braga). Resultados das Campanhas de 1989-90 e 1991-92». Cadernos de Arqueologia, Série II, 8-9: 199-230. Fontes, L. (1992). «O Norte de Portugal no Período Suevo-Visigótico. Elementos para o seu estudo». In: Actas XXXIX Corso di Cultura Sull’Arte Ravennate e Bizantina (Ravena, 6-12 Aprile 1992). Ravena, 217-248. Fontes, L. F. O. (1993). «Inventários de Sítios Arqueológicos de Concelho de Braga». Mínia, 3ª série, 1: 31-88. Fontes, L., F. S. Lemos, and M. Cruz (1997-98). «“Mais Velho” que a Sé de Braga. Intervenção arqueológica na catedral bracarense: notícia preliminar». Cadernos de Arqueologia, 14/15, série II: 137-164. Fontes, L., B. Vilar, H. Carvalho, and J. Ribeiro (2009). Colecção de epigrafia e de Arquitectura Antigas (séculos I a.C.-VII d.C.). Braga, II vols. Fontes, L. (2005). São Martinho de Tibães: um sítio onde se fez um mosteiro. Ensaio em arqueologia da paisagem e da arquitectura. Lisboa: IPPAR. Fontes, L. (no prelo). «O Período Suévico e Visigótico e o Papel da Igreja na Organização do Território». In. Minho. Traços de Identidade, edited by Paulo Pereira. Braga. Conselho Cultural da Universidade do Minho. Fontes, L; M. Martins, M. C. Ribeiro, and H. Carvalho (2009). «A cidade de Braga e o seu território nos séculos V-VII». In: Espacios Urbanos en el Occidente Mediterráneo (s. VI-VIII), edited by A. Garcia, R. Izquierdo, L. Olmo, and D. Peris. Toledo: Toletvm Visigodo, 255-262. Fontes, L., M. Martins, and F. Andrade (2010a). Salvamento de Bracara Augusta: quarteirão dos CTT - interligação Túnel Avenida da Liberdade (BRA 09 CTT-ITAVL). Relatório final. TAUM: Memórias. 2. Braga: UAUM. http://hdl.handle.net/1822/10143. Fontes, L., M. Martins, J. Sendas, and S. Catalão (2010b). Salvamento de Bracara Augusta:

ampliação do túnel da Avenida da Liberdade (BRA 08-09 TAVL). Relatório final. TAUM: Memórias. 7. Braga: UAUM. http://hdl.handle.net/1822/11048. Fontes, L.(2012) Arqueologia, povoamento e construção de paisagens serranas : o termo de Lindoso, na Serra Amarela, Braga. http://hdl.handle.net/1822/20982 Forni, G. (1970). «L’occupazione militaire romana della Spana Nord-ocidentalle: analogie e parelleli». In: Legio VII Gemina. León, 207-225. Freitas, J. (1971). «Citânia de S. Julião de Caldelas». Arqueólogo Português, 3ª série, 133-138. García Entero, V. (2001). Los Balnea de las villae Hispanorromanas. Provincia Tarraconense. Monografías de Arquitectura Romana, 5, Serie Termas I. Madrid. García Marcos, V., and J. M. Vidal Encinas (1998). «Asturica Augusta: de asentamiento militar a urbs magnifica». In: Los orígenes de la ciudad en el Noroeste hispânico. Actas del Congreso Internacional, Lugo 15-18 de Mayo 1996, edited by A. Rodríguez Colmenero. Lugo, 911-944. García Marcos, V., A. Morillo, and E. Campomanes (1997). «Nuevos planteamientos sobre la cronología del recinto defensivo de Asturica Augusta (Astorga)». In: Congreso Internacional sobre la Hispania de Teodosio (Segovia, 1997).Valladolid/Segovia, 515-531, Gil, F. (1984). Mimésis e negação. Lisboa: INCM. Gomes, A. M. (2000). «Cerâmicas pintadas de época romana: tecnologia, morfologia e cronologia», Master dissertation, Universidade do Minho. Gomes, J. M. F. (1996). «Cividade de Terroso e Vila Mendo. Aspectos da Proto-História e Romanização do Litoral Minhoto», PhD dissertation, Faculdade de Letras da Universidade do Porto. González Fernández, E., and M. C. Carreño Gascón (1998). «La capital del extremo noroeste hispánico: Lucus Augusti y su tejido urbano a la luz de las ultimas intervenciones arqueológicas». In: Los orígenes de la ciudad en el Noroeste hispânico. Actas del Congreso Internacional, Lugo 15-18 de Mayo 1996, edited by A. Rodríguez Colmenero. Lugo, 1171-1208. González Fernández, M. L. (1995a). «La forma urbis: planimetria de una ciudad roman». In: Lucus Augusti, Urbs Romana. Los orígenes de la ciudad de Lugo. Lugo: Ayuntamiento de Lugo/Fundacion Caixa Galicia, 53-54. González Fernández, M. L. (1999). «De acampamento a ciuitates. La primera fortificación urbana de Asturica Augusta (Astorga, León)». Numantia, 7: 95-115. González Ruibal, A. (2003). «Arqueología del Primer Milenio en el Noroeste de la Península Ibérica», PhD, Universidad Complutense de Madrid. González Ruibal, A. (2006-07). Galaicos. Poder y comunidad en el Noroeste de Península Ibérica 233

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

(200 a.C.-50d.C.). Brigantium 18-19. A Coruña: Museu Arqueológico e Histórico, , Tomos I-II. Goodman, P. J. (2007). The roman city and is periphery. From Rome to Gaul. London and New York: Routledge. Gorges, J. G. (1979). Les Villas hispano-romaines. Paris. Granja, H. M. L. P. (1989). «A História Geológica recente da Área de Paisagem Protegida do Litoral de Esposende e o seu Futuro». In: Actas do 2º Congresso de Áreas Protegidas. Lisboa. Grau Lobo, J., and J. L. Hoyas (ed.) (2001). El bronce de Bembibre: un edicto del emperador Augusto del año 15 a.C. Valladolid. Hauschild, Th. (1976). «El Faro romano de La Coruña». In: Actas del Bimilenario de Lugo. Lugo, 131136. Hingley, R. (1989). Rural Settlement in Roman Britain. London: Seaby. Höck, M. (1986). «Studien zur sogennanten CastroKultur in Nordportugal». PhD dissertation, Philipps-Universität, Marburg. Hopkins, K. (1980). «Taxes and trade in the Roman empire». Journal of Roman Studies, 70, 101-125. Hübner, E. (1869). Corpus Inscriptionum Latinarum. II. Supplementum (=CIL). Berlin. Hutter, S. (1973). Der Leuchtturm von La Coruña. Madrider Beiträger, 3. Mainz: P. von Zabern. Iglesias Gil, J. M. (1976). Epigrafia cántabra. Estereometria. Decoratión. Onomástica. Santander. Jones, R. F. J. (1972). «Roman Gold-Mining in NorthWest Spain II: workings on the rio Duerna». Journal of Roman Studies, 62: 59-74. Jones, R. F. J. (1987). «A false start? The Roman urbanization of Western Europe». World Archaeology, 19, 47-57. Jorge, S. O. (1979). «Escavações arqueológicas no povoado da Chã do Castro (Amares, Braga)». Revista de Guimarães, 89: 281-292. Jorge, S. O. (1980). «Escavações da mamoa 1 de Outeiro de Gregos». Portugália, 1: 9-28. Jorge, S. O. (1987) O povoado da Bouça do Frade (Baião), no quadro do Bronze Final do Norte de Portugal. Monografias Arqueológicas, 2. Porto: GEAP. Jorge, S. O. (1988) “Reflexões sobre a Pré-História recente do Norte de Portugal”, Trabalhos da Sociedade Portuguesa de Antropologia e Etnologia, 28 (1-2), pp. 85-112. Jorge, S. O. (ed.) (1998). Existe uma Bronze Age Atlântico? Trabalhos de Arqueologia, 10. Lisboa: Instituto Português de Arqueologia. Jorge, S. O., and J. J. R. Sousa (1980). «Resultados preliminares de uma sondagem na estação arqueológica da Chã do castro (Amares, Braga)». In: Seminário de Arqueologia do Noroeste Peninsular. Guimarães, 121-131. Kristiansen, K. (1998). Europe before History. Cambridge: Cambrige University Press.

Lambert, C., and J. Rioufreyt (1982). «Habitats indigènes, Villas gallo-romaines et structures agraires antiques dans le Maine”. In: Actes du Colloque: La Villa romaine dans les provinces du nord-ouest. Caesarodunum, nº XVII. Paris, 141-188. Launaro, A. (2004). «Concerning landscape». Agri centuriati: International Journal of Landscape Archaelogy, 1: 31-41. Le Roux, P. (1974). «Les stèles funéraire de Braga. Remarques sur une nouvelle inscription en vers». Archeologica Opuscula, 1: 41-48. Le Roux, P. (1994). «Bracara Augusta, ville latine». Trabalhos de Antropologia e Etnologia, 34 (12) (Actas do I Congresso de Arqueologia Peninsular): 229-241, Le Roux, P. (1995). Romains d`Espagne. Cités et politique dans les provinces II siècle av. J.-C. – IIIe siècle ap. J.-C. Paris: Armand Colin. Le Roux, P., and A. Tranoy (1973). «Rome et les indigènes dans le NO de la Péninsule Ibérique. Problèmes d'épigraphie et d'histoire». Mélanges de la Casa Vélazquez, 9: 177-231. Le Roux, P., and A. Tranoy (1974). «Contributions à l’étude des régions rurales du NO hispanique au Haut-Empire: deux inscriptions de Penafiel». In: III Congresso Nacional de Arqueologia. Porto, 249-258. Le Roux, P., and A. Tranoy (1975). «Problèmes épigraphiques de la province d’Orense». BAUR, 5: 271-279. Le Roux, P., and A. Tranoy (1982). «Nouvelles inscriptions du Nord du Portugal». Minia, 6: 3137. Le Roux, P. e Tranoy, A. (1984). «Le mot et la chose. Contribution au débat historiographique». Lucerna (Homenagem a D. Domingos de Pinho Brandão): 239-255. Leite, F. S. (1997). «Contribuição para a localização do centro de produção da cerâmica “dita” Bracarense», Master dissertation, Universidade do Minho. Lemos, F. M. S., and A.M. Baptista (1995-1996). «Estudo de um troço da via XVIII do Itinerário de Antonino». Cadernos de Arqueologia, II Série, 12-13: 118¬138. Lemos, F. S. (1993). «Povoamento romano de Trás-osMontes Oriental», PhD dissertation, Universidade do Minho. Lemos, F. S. (1995) «Caminhos e Muralhas de Braga». Forum, 18: 49-62. Lemos, F. S. (2000). «A via romana entre Bracara Augusta e Asturica Augusta, por Aquae Flaviae (Contributo para o seu estudo)». Revista de Guimarães, 110: 15-52. Lemos, F. S. (2001). «Arredores de Bracara Augusta – escavações arqueológicas na necrópole de S. Vítor, no contexto da via romana para Aquae Flaviae». Forum, 29: 9-38.

234

References

Lemos, F. S. (2002). «Bracara Augusta - A Grande Plataforma viária do Noroeste Peninsular». Forum, 31: 95-127. Lemos, F. S. (2003). «Bracara Augusta – periferia imediata». Forum, 33: 11-25. Lemos, F. S. (2005). «O mundo romano. Romanização do norte de Portugal». In: Arte e cultura da Galiza e norte de Portugal. Arqueologia, edited by J. M. Hidalgo Cuñarro. Setúbal: Marina Editores, vol. 1, 148-207. Lemos, F. S., and G. Cruz (2007a). Citânia de Briteiros. Povoado Proto-Histórico. ProtoHistoric Settlement (Guia). Guimarães: Ed. Sociedade Martins Sarmento. Lemos, F. S., e J. M. F. Leite (2000). «Trabalhos Arqueológicos no logradouro da Casa Grande de Santo António das Travessas». Forum, 27: 15-38. Lemos, F. S., and P. Morais (2004). «Vias Augustas e Mineração Aurífera». Forum, 36: 15-56. Lemos, F. S., C. Carvalho, and C. Meireles (2005). «Conjunto Mineiro Romano em Montalegre». In: Actas do XVI Jornadas sobre a função social do Museu. Montalegre: C. M. Montalegre. Lemos, F. S., and G. Cruz, (2007b). «Citânia de Briteiros: trabalhos arqueológicos recentes», AlMadan, IIª série, n.º 15, Adenda Electrónica. Lemos, F. S., and G. Cruz (no prelo). «Citânia de Briteiros. Perspectivas recentes sobre a Romanização». In: Proceedings of the XV Congress International Union for Prehistoric and Protohistoric Sciences. Lisboa. Leveau, Ph. (1983). «La ville antique et l'organisation de l'espace rurale: Villa, ville, Village". Annales, ESC, 4: 920-942. Leveau, Ph. (1984). Caesarea de Maurétanie. Une ville romaine et ses campagnes. Coll. École Française de Rome, 70. Rome. Leveau, Ph. (1997). «Temps, espace et structuration des paysages». In: Les formes du paysage. 3. L'analyse des systèmes spatiaux, edited by G. Chouquer. Paris: Ed. Errance. Leveau, Ph., and Ch. Goudineau (1983). «La ville antique, “ville de consomation?”». Études rurales, 89-91: 275-289. Lopes, A. B. (2003). «Proto-história e romanização do Baixo Minho», PhD dissrtation, Faculdade de Letras da Universidade do Porto. Lopes, M. C. (2003). A cidade romana de Beja. Percursos e debates acerca da "civitas" de Pax Ivlia. Conimbriga. Anexos 3. Coimbra: Instituto de Arqueologia, Faculdade de Letras da Universidade de Coimbra. Lopes, M.C., P. C Carvalho, and S. M. Gomes (1998) Arqueologia do Concelho de Serpa. Serpa: Câmara Municipal de Serpa. Lopez Cuevillas, F. (1933). «A área xeográfica da cultura norte dos castros». In Homenagem a Martins Sarmento, 99-107.

Lopez

Cuevillas, F. (1946-47). «Armeria posthallstática del Noroeste hispânico». Cuadernos de Estudios Gallegos, 2 (8): 543-589. Lopez Cuevillas, F. (1947). «Las cabanas de los castros». Anuarios y Memorias de la Sociedad Española de Antropologia, Etnologia y Prehistoria, 22 (1-4): 141-146. Lopez Cuevillas, F. (1953). «Cerámicas castrexas de la ciudad y provincia de Lugo». Boletín de la Comisión Provincial de Monumentos de Lugo, 5: 253-259. Lopez Cuevillas, F. (1958). «Cerámicas inéditas del castro de Vigo». Cuadernos de Estudios Gallegos, 13 (41): 321-328. Loureiro, E., H. Carvalho, N. Botica, and M. Martins (2010). «Sistemas de Informação Geográfica aplicados à Arqueologia: o caso de Bracara Augusta e do território envolvente». Comunicação apresentada no VIII Encontro de Utilizadores de ESRI- Portugal 2010, Lisboa, 3 e 4 de Março de 2010. Maciel, T. (1997). «Padrões de povoamento protohistórico no vale do Neiva», PhD dissertation, Faculdade de Letras da Universidade do Porto. Maciel, T. (2003). O Povoamento Proto-Histórico do Vale do Neiva. Esposende: Rio Neiva / Associação de Defesa do Ambiente. Mangin, M. (1983). «Agglomérations gallo-romaines du Centre-Est et acculturation sous le HautEmpire». In: La patrie gauloise d’Agrippa au VIème siècle (Actes du Colloque, Lyon 1981). Lyon, 39-54. Mantas, V. G. (1996). «A rede viária da Faixa Atlântica entre Lisboa e Braga», PhD dissertation, Faculdade de Letras da Universidade de Coimbra. Mantas, V. G. (2000). «A via romana Bracara AugustaCale: traçado, funções e influência no povoamento regional». Revista de Guimarães, 110: 53-87. Mantas, V.G. (1987). «As primitivas formas de povoamento em Portugal». Povos e Culturas, 2: 13-55. Mapa das Ruas de Braga (1989-1991). Mapa das Ruas de Braga. Barga: Arquivo Distrital de Braga/Universidade do Minho, 2 vols. Mar, R. (ed.) (1993). Els monuments provincials de Tàrraco. Noves aportaciones al seu coneixement. Documents d’Arqueologia Clàssica, I. Tarragona. Mar, R. e Pensabene (2001). «El urbanismo de Tarraco». In: Tarraco puerta de Roma, Catálogo de la exposición. Tarragona, 37-45. Marques, J. (1989). «A Assistência no Norte de Portugal nos finais da Idade Média». Revista da Faculdade de Letras do Porto – História, II Série, VI: 11-93. Marques, J.A.M. (1984). «Inventário arqueológico do concelho de Monção: estado da questão». Revista de História, 1. 235

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis Martinez Temuje, J. (1975). La via romana “per loca maritima” por el Baixo Miño y costa atlantica. Pontevedra. Martins, C. M. B. (2005). A Exploração Mineira Romana e a Metalurgia do Ouro em Portugal, PhD dissertation, Faculdade de Letras da Universidade de Porto. Martins, M. (1988). O povoado fortificado do Lago, Amares. Cadernos de Arqueologia-Monografias, 1. Braga. Martins, M. (1990) O povoamento proto-histórico e a romanização da bacia do curso médio do Cávado. Cadernos de Arqueologia, Monografias 5. Braga. Martins, M. (1991) O povoado de Santo Ovídio (Fafe). Resultados dos trabalhos realizados em 19801984. Cadernos de Arqueologia, Monografias 6. Braga. Martins, M. (1995). «A ocupação romana da região de Braga: balanço e perspectivas de investigação». In: Actas do Congresso Histórico Comemorativo dos 150 anos do Nascimento de Alberto Sampaio. Guimarães, 73-114. Martins, M. (1996). «Povoamento e habitat no Noroeste português durante o 1º milénio a. C.». In: De Ulisses a Viriato. O primeiro milénio a. C., Museu Nacional de Arqueologia. Lisboa, 118-133. Martins, M. (1997-1998). «A zona arqueológica das Carvalheiras. Balanço das escavações e interpretação do conjunto». Cadernos de Arqueologia, 14-15: 23-45. Martins, M. (1999). «A urbanização do Noroeste peninsular: o caso de Bracara Augusta». In: Actas da Mesa Redonda. Emergência e desenvolvimento das cidades romanas no Norte da Península Ibérica. Porto: Escola Profissional de Arqueologia. Martins, M. (2000) Bracara Augusta cidade romana. Braga: Unidade de Arqueologia da Universidade do Minho. Martins, M. (2004). «Urbanismo e Arquitectura de Bracara Augusta. Balanço dos contributos da Arqueologia Urbana». In: Simulacra Romae. Roma y las capitales provinciales del ocidente europeu. Estudios Arqueológicos, 149¬173. Martins, M. (2005). As termas romanas do Alto da Cividade. Bracara Augusta. Escavações Arqueológicas 1. Braga: UAUM-NARQ. Martins, M. (2006). «Bracara Augusta: a Roman town in Atlantic area». In: Early Roman Towns in Hispania Tarraconensis, edited by L. Abal Casal, et al., Portsmouth, Rhode Island, 213222. Martins, M. (no prelo). «A romanização». In: Minho. Traços de identidade. Braga: Universidade do Minho. Martins, M., and M. Delgado (1989-90a). «História e Arqueologia de uma cidade em devir: Bracara Augusta». Cadernos de Arqueologia, 6/7: 11-38.

Martins, M. e Delgado, M. (1989-90b). «As necrópoles de Bracara Augusta. A. Os dados arqueológicos». Cadernos de Arqueologia, Série II, 6/7: 41-186. Martins, M. e Delgado, M. (1999). «Bracara Augusta: uma cidade na periferia do Império». In: Los finisterres altânticos en la Antigüedad. Época Prerromana y Romana. Actas del Coloquio Internacional, edited by C. Férnandez Ochoa. Gijón: Electa, 121-127. Martins, M., F. S. Lemos, and F. Pérez Losada (2005). «O povoamento romano no território dos galaicos bracarenses». In: III Coloquio Internacional de Arqueología en Gijón. Unidad y diversidad en el Arco Atlântico en época romana, edited by C. Fernández Ochoa, and P. García Díaz. Gijon, 279-296. Martins, M.; Ribeiro, J.; Magalhães, F. (2006). «A arqueologia em Braga e a descoberta do teatro romano de Bracara Augusta». Forum, 40: 9-30. Martins, M., Carvalho, H. (2010). «Bracara Augusta and the changing of rural landscape». In: Changing Landscapes. The impact of roman towns in the Western Mediterranean. Proceedings of the International Colloquium, edited by C. Corsi, and F. Vermeulen. Bologna: AnteQuem, 281-298. Martins, M., L. Fontes, C. Braga, J. Braga, F. Magalhães, and J. Sendas (2010). Salvamento de Bracara Augusta: quarteirão dos CTT Avenida da Liberdade (BRA 08-09 CTT). Relatório final. TAUM: Memórias. 1. Braga: UAUM http://hdl.handle.net/1822/13345 Martins, M; M. C. Ribeiro, and J. M. Baptista (2011). «As termas públicas de Bracara Augusta e o abastecimento de água da cidade romana». In: Aquae Sacrae: Agua y Sacralidade n la Antigüedad, edited by A. Costa, L. Palahí, and D. Vivó. Girona: Institut de Recerca Històrica de la Universitat de Girona, 69-101. Martins, M; Ribeiro, M. C.; (2012). «Gestão e uso da água em Bracara Augusta. Uma abordagem preliminar». In: Caminhos da água. Paisagens e usos na longa duração, edited by M. Martins, I. Freitas, and I. Valdivieso. Braga: CITCEM, 951. Martins, M; Ribeiro, J.; Magalhães, F; Braga, C. (2012), “Urbanismo e arquitetura de Bracara Augusta. Sociedade, economia e lazer”. In: Evolução da Paisagem urbana. Sociedade e Economia, edited by M. C. Ribeiro, and A. Melo. Braga: CITCEM, pp. 29-68: Martins, M., M. C. Ribeiro (2013). «Em torno da Rua Verde. A evolução urbana de Braga na longa duração». In: Evolução da Paisagem Urbana. Transformação morfológica dos tecidos históricos, edited by M. C. Ribeiro, and A. Melo. Braga: CITCEM, 11-44. Martins, M; Ribeiro, J; Magalhães, F; Mar, R. (in press) «A construção do teatro romano de Bracara Augusta». In: História da Construção. 236

References

Arquiteturas e técnicas Construtiva, edited by A. Melo, and M. C. Ribeiro. Braga:CITCEM. Mattingly, D. J. (1997). Dialogues in Roman Imperialism. Power, discourse and discrepant experience in the Roman Empire. JRA Supplementary Series, 23. Portsmouth. Maya González, J. L., and Cuesta Toribio, F. (2001). El castro de la campa Torres. Período preromano. Serie Patrimonio, 6. Gijón. Maya, J. L., and F. Cuesta (1995). «Estratigrafía e interpretación histórica de la Campa Torres (1991-1994)». In: Excavaciones arqueológicas en Asturias, 1991-1994. Oviedo: Principado de Asturias, 105-116. Menéndez de Luarca, J. R., and Navia Osorio (dir.) (2000). A construção do território. Mapa histórico do Noroeste da Península Ibérica. Madrid/Barcelona. Metzler, J. (dir.) (1994). «Table Ronde nº 2, 1ère partie: Les origines des agglomérations secondaires». In: Les agglomérations secondaires. La Gaule Belgique, les Germanies et l’Occident romain. Actes du Colloque de Bliesbruck-Reinheim/Bitche (21-24 Oct. 1992), edited by in J. P. Petit, M. Mangin, Ph. Brunella. Paris, 249-254. Millett, M. (1979). «An approach to the functional interpretation of pottery». In: Pottery and the Archaeologist, edited by M. Millet. Londres, 35-48. Millett, M. (1982). «Town and country: a review of some material evidence». In: The RomanoBritish countryside, edited by D. Miles. BAR, 103. Oxford, 421-431. Millett, M. (1985). «Field survey calibration: a contribution». In: Archaelogy from the Ploughsoil: Studies in the Collection and Interpretation of Field Survey Data, edited by C. Haselgrove, M. Millett, and I. Smith. Sheffield: J. R. Collis, pp. 31-37. Millett, M. (1990). The Romanization of Britain. An essay in archaeological interpretation. Cambridge: Cambridge University Press. Millet, M. (2000). «Dating, quantifying and utilizing pottery assemblages from surface survey». In: The Archaeology of Mediterranean Landscapes 5: Extracting Meaning from Ploughsoil Assemblages, edited by R. Francovich, H. Patterson, and G. Barker. Oxford, 53-59. Millett, M., F. Queiroga, F., K. Strutt, J. Taylor, and S. Willis (2000). «The Ave valley, Northern Portugal: an archaeological survey of Iron Age and roman settlement». Internet Archaelogy, 9. htt://intarch.ac.uk/journal/issue9/millett_index.h tml. Misurare la terra (1983). Misurare la terra: centuriazione e coloni nel mondo romano. Modena. Moatti, C. (1993). Archives et partage de la terre dans le monde romain (IIe siècle avant- Ier siècle après J.-C.). Coll. École Française de Rome,

173). Paris-Roma: École Française de Rome De Boccard. Moita, I. (1962). «A cultura castreja na área galaicoportuguesa». In: Associação Portuguesa para o Progresso das Ciências. Porto, 453-460. Moita, I. (1971). «Povoados, redutos e santuários castrenses (ensaio de classificação)». In: II Congresso Nacional de Arqueologia, Coimbra, 271-281. Montenegro Duque, A., J. M. Blázquez Martinez, and J. M. Solana Sáinz (1986). Historia de España. 3.España Romana. Madrid: Ed. Gredos. Morais, R. M. L. (1998). As ânforas da zona das Carvalheiras. Contributo para o estudo das ânforas romanas de Bracara Augusta. Cadernos de Arqueologia, Monografia 8. Braga: Unidade de Arqueologia da Universidade do Minho. Morais, R. M. L. (2001). «Breve ensaio sobre o anfiteatro de Bracara Augusta». Forum, 30: 5576. Morais, R. M. L. (2005). Autarcia e comércio em Bracara Augusta no período Alto-Imperial: contribuição para o estudo económico da cidade. Bracara Augusta. Escavações Arqueológicas 2. Braga: UAUM/NARQ, 2 vols. Morais, R. M. L. (2007). «Ânforas da Quinta da Ivanta - um pequeno ´habitat` mineiro em Valongo». Conímbriga, XLVI: 267-280. Morel, J.P. (1994). «Les agglomérations secondaires dans l’Italie Péninsulaire». In: Les agglomérations secondaires. La Gaule Belgique, les Germanies et l’Occident romain. Actes du Colloque de Bliesbruck-Reinheim/Bitche (21-24 Oct. 1992), edited by J. P. Petit, M. Mangin, and Ph. Brunella. Paris, 153-162. Morel, J. P. (1981). «La produzione della ceramica campana: aspetti economici e sociali». In: Società romana e produzione schiavistica. Ii. Merci, mercati e scambi nel Mediterraneo, edited by A. Giardina, and A. Schiavone. Roma/Bari, 81-98. Morgan, M., and P. Brusegard (1989). «Spatial analysis using GIS». In: URISA`89 Preconference workshop, Urban and Regional Information Systems Association. Boston. Morillo Cerdán, A. (1995). «Cántabros y astures, pueblos prerromanos de norte de la Península». In: Astures. Pueblos y culturas en la frontera del Imperio Romano. Gijón, 41-51. Morillo Cerdán, A. (2003). «Lucernas romanas de Astorga». Astorga, IV: 9-632. Naveiro López, J. (1991). El comercio antiguo en el NW Peninsular. Monografías Urxentes 5. A Coruña: Museu Arqueoloxíco da Coruña, Nicolet, C. (1984). Roma y la conquista del mundo mediterráneo (264-27 a.C.). 2. La génesis de un Imperio. Barcelona Ed. Labor. Nicolet, C. (1988). L’inventaire du monde. Géographie et politique aux origines de l’Empire Romain. Paris: Fayard. 237

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Nogueira, A. M. M. (1998). «Formação e defesa do património do Cabido de Braga nos finais da Idade Média», Master dissertation, Universidade do Minho. Olesti i Vila, O. (1995-1996). «Actuaciones catastrales y romanización en el territorio del Maresme en época republicana: la contribución del estudio arqueológico». Studia Historica. Historia Antigua, 13-14: 105-124. Oliveira, A. (1979). «A Abadia de Tibães 1630/801813. Propriedade, exploração e produção agrícolas no vale do Cávado durante o Antigo Regime», PhD dissertation, Faculdade de Letras da Universidade do Porto. Oliveira, E. P. (1978). «O salvamento de Bracara Augusta. IV. Os apontamentos arqueológicos de Braga de José Teixeira». Minia, Nº. 2, série 1(1): 20-44. Orejas Saco del Valle, A. (1996). Estrutura social y território. El impacto romano en la Cuenca Noroccidental del Duero. Anejos de Archivo Español de Arqueología, 15. Madrid. Orejas Saco del Valle, A. (2001) Atlas Historique des zones minières d’Europe. I. Bruselas: Comisión Europea, Orejas Saco del Valle, A. (2005). «El poblamiento romano en los distritos mineros del Noroeste». In: III Coloquio Internacional de Arqueología en Gijón. Unidad y diversidad en el Arco Atlántico en época romana, edited by Fernández Ochoa, C e García Díaz, P. (eds.). Gijón, 309319. Orejas Saco del Valle, A., and F. J. Sánchez-Palencia (2002). «Mines, territorial organization and social structure in Roman Iberia». American Journal of Archaeology, 106 (4): 581-601. Paiva, M. B. C. (1993). «Ânforas Romanas de Castros da Fachada Atlântica do Norte de Portugal», Master dissertation, Faculdade de Letras da Universidade do Porto, Porto. Palet i Martínez, J.M. (1994). Estudi territorial del Pla de Barcelona. Estructuració i evolución del territorio entre l’època iberoromana i l’altmedieval segles II-I a. C. / X-XI d.C. Estudis i Memòries d`Arqueologia de Barcelona. Barcelona: Centre d’Arqueologia de la Ciutat. Panerai, M. C. (1983) «Le misure romane». In: Misurare la terra: centuriazione e coloni nel mondo romano. Modena, 122-124. Pastor Muñoz, M. (1974). «El culto imperial en el “Conventus Asturum». Hispania Antiqua, 4: 203-233. Pastor Muñoz, M. (1976). «La religión romana en el “Conventus Asturum”». Hispania. Revista española de historia, 36 (134): 489-524. Peña Santos, A. (1996). «A secuencia cultural do mundo castrexo galego». In: A cultura castrexa galega a debate, edited by J. M. Hidalgo Cuñarro. Vigo: Instituto de Estudos Tudenses, 65-103.

Pereira Menault, G. (1982). «Los Castella y las comunidades de Gallaecia». Zephyrus, 34-35: 249-267. Pereira Menault, G. (1983). «Los castella y las comunidades de Gallaecia». In: II Seminario de Arqueologia del Noroeste. Madrid, 169-192. Pereira, F. A. (1933). «Os vestígios das habitações citanienses». In: Homenagem a Martins Sarmento, 27-30. Pereira, F. A. (1938-41). «Páginas inéditas». Revista de Guimarães, 48 (1-3): 137¬150; 48 (4): 277-289; 49 (1-2): 35-41; 49 (3-4): 103-111; 51 (1-2): 2740; 51 (3-4): 312-330. Pereira, I., J. P. Bost, and J. Hiernard (1974). Les Monnaies, FC – III. Paris: Diffusion du Boccard, Pérez Losada, F. (1987). «Sobre el concepto de “Villa” no mundo romano». Cadernos de Arqueologia, 4: 79-110. Pérez Losada, F. (1990-91). «A Villa romana de Noville (Mugardos - A Coruña) Estructuras e estratigrafía arqueológicas. Campañas 1988 e 1989». Castrelos, III-IV: 245-281. Pérez Losada, F. (1991). Os asentamentos na Galicia romana”, in Galicia. Historia I (Prehistoria e Historia Antiga). A Coruña: Hércules Ed., 403441. Pérez Losada, F. (1995). Arqueoloxía e arte no mundo rural: Habitat e arquitectura das villae galaicorromanas. Monografias del Museo Arqueolóxico e Histórico de A Coruña, 7, 165188. Pérez Losada, F. (1996). «Hacia una definición de los asentamientos rurales en la Gallaecia: poblados (vici) y casas de campo (villae)». In: Los finisterres altânticos en la Antigüedad. Época Prerromana y Romana. Actas del Coloquio Internacional, edited by C. Férnandez Ochoa. Gijón: Electa, 189-197. Pérez Losada, F. (1998). «Cidades e aldeias na Galiza romana. Uma proposta de classificação hierárquica do habitat galaico-romano». O Arqueólogo Português, série IV, 16: 157-174. Pérez Losada, F. (2000). «Sistema viário e núcleos agrupados romanos no sector oriental galego do convento bracarense». Revista de Guimarães, 110: 119-161. Pérez Losada, F. (2002). Entre a cidade e a aldea. Estudio arqueohistórico dos "aglomerados secundários" romanos en Galicia. Supl. Brigantium, vol. 13 A Coruña. Pérez Losada, F., and A. Acuña Piñeiro (1998). «Tude (Tui, Pontevedra): un núcleo secundario na Galicia romana”». In:. Los orígenes de la ciudad en el Noroeste Hispânico. Actas del Congreso Internacional, edite by A. Rodríguez Colmenero, Lugo, 871-890. Pérez Losada, F., J. L. Naveiro López, F. Boval Galán, L. Castro Pérez, J. M. Vasquéz Varela, C. Fernández Rodríguez (1992). «Estudio do material arqueolóxico procedente da Villa 238

References

romana de Noville (Mugardos - A Coruña)». Minius, I : 57-87. Peyras, J. (2006). «Organisation et marqueurs des paysages d’après les géomètres de l’antiquité tardive“, in Marqueurs des Paysages et Systèmes socio-économique». In: Proceedings of le Mans COST Conference. Rennes: Presses Universitaire de Rennes, 23-34. Pimenta, J.R. (1997). «O “Lugar do Passado” em Martins Sarmento. Geo-historiografia do programa de investigação arqueológica da “Cultura castreja” (1876-1899)», PhD dissertation, Universidade do Minho. Pinto, R.S. (1932). «A cividade de Terroso e os castros do Norte de Portugal». Revista de Guimarães, 42 (1-2): 81-91. Pinto, M.M. (2008). Villa Romana de Sendim-Roteiro e Exposição Permanente. Felgueiras. Portet, P. (2004). Bertrand Boysset, La vie et les œuvres techniques d’un arpenteur médiéval (v. 1355 – v.1416). Paris: Le Manuscrit. Provost, M. (1993). Le Val de Loire dans l’Antiquité. Gallia, 52e supplement. Paris. Pucci, G. (1983). «Pottery and trade in the Roman period». In: Trade in the ancient economy, edited by P. Garnsey, K. Hopkins, ad C.R. Whittaker. London, 105¬117. Queiroga, F. (1992). «War and Castros», PhD dissertation, Oxford. Rabanal, M. (1988). Astorga romana: burócratas, administradores y funcionários. Astorga. Ramil Rego, P. (1993). «Paleoetnobotánica de yacimientos arqueológicos holocenos de Galicia (N.O Cantábrico)». Munibe, 45: 165-174. Rasmussen, T. (1991). «Tuscania and its territory» In: Roman Landscapes. Archaeological Survey in the Mediterranean Region, edited by G. Barker, and J. Lloyd. London, 106-114. Reece, R. (1997). «Romanization: a point of view». In: The Early Roman Empire in the West, edited by T. Blagg, and M. Millett, Oxford: Oxbow Books, 30-34. Redentor, A. (2011). «A cultura epigráfica no conuentus Bracaraugustanus (pars occidentalis): percursos pela sociedade brácara da época romana», PhD dissertation, Faculdade de Letras da Universidade de Coimbra. Regalo, H. J. L. A. (1986). «Levantamento arqueológico do concelho de Vila Verde Notícia preliminar». Mínia, 8: 70-110. Regalo, H. J. L. A., and M. A. Brito, (1991). Carta arqueológica da Póvoa de Lanhoso». In: Actas das IV Jornadas Arqueológicas (Lisboa 1990), Lisboa: Associação dos Arqueólogos Portugueses, 117-122. Reilly, P.; and S. Rahtz (eds.) (1992). Archaelogy and the information age. One World Archaeology 21. London. Ribeiro, M. C. F. (2008) «Braga entre a época romana e a Idade Moderna. Uma metodologia de análise para a leitura da evolução da paisagem urbana»,

PhD dissertation, Instituto de Ciências Sociais, Universidade do Minho. Richards, J. D., and N. S. Ryan (1985). Data processing in archaeology. Cambridge Manuals in Archaeology. Cambridge. Richardson, J. S. (1996). «The new Augustan edicts from northwestern Spain». Journal Roman of Archaeology, 15. Rodríguez Colmenero, A. (1979). Augusto e Hispania. Conquista y organización del Norte Peninsular. Cuadernos de Arqueología de Deusto, 7. Bilbao. Rodríguez Colmenero, A. (1995). «Cidades e urbanismo na Galicia romana». In: Arqueoloxía e arte na Galicia Prehistórica e Romana, edited by F. Pérez Losada, and L. Castro Pérez. Monografias, 7. A Coruña: Museo Arqueolóxico e Histórico da Coruña, 87-103. Rodríguez Colmenero, A. (1995-96). «Mansiones y mutationes en la Via Nova (XVIII del Itinerário de Antonino». Cadernos de Arqueologia, II Série, 12-13:89¬112. Rodríguez Colmenero, A. (1996a). «Integración administrativa del Noroeste peninsular en las estructuras romanas». In: Lucus Augusti I. El amanecer de una ciudad, edited by A. Rodríguez Colmenero. A Coruña, 265-299. Rodríguez Colmenero, A. (1996b). «La Tabula hospitalitatis de la civitas Lougeiorum. Documento genuino o falsificacion?». In: Lucus Augusti I. El amanecer de una ciudad, edited by A. Rodríguez Colmenero. A Coruña, 301-315. Rodríguez Colmenero, A., and E. J. Alcorta Irastorza, (1998). «Aquae Flaviae (Chaves), ciudad romana. Balance y possibilidades» In: Los orígenes de la ciudad en el Noroeste Hispânico. Actas del Congreso Internacional (Lugo 15-18 de Mayo 1996), edited by A. Rodríguez Colmenero, Lugo, 779-806. Rodríguez Colmenero, A., and M. C. Carreño Gascón (1999). «Lucus Augusti, capital romana del Finisterre Hispánico». In: Emergência e Desenvolvimento das cidades romanas no Norte da Península Ibérica. Actas da Mesa Redonda. Porto: 115-132. Rodríguez Colmenero, A., M. C. Carreño Gascón, E. González Fernández, F. Hérves Rraigoso, S. Ferrer Sierra, E. Alcorta Irastorza, R. Casal García, and J. Naveiro López (1995). Lucus Augusti, Urbs Romana. Los Orígenes de la ciudad de Lugo. Lugo: Ayuntamiento de Lugo. Rodríguez Colmenero, A., S. Ferrer Sierra, and R.D. Alvarez Asorey (2004). Callaeciae et Asturiae Itinera Romana. Miliarios e outras inscricións viárias romanas do Noroeste Hispânico (Conventos Bracarense, Lucense e Asturicense). Santiago de Compostela. Rodriguez, J. (1970). «Vias militares romanas». In: Legio VII Gemina. León. Roldán Hervás, J. M. (1974). Hispânia y el ejército romano. Salamanca. 239

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Roldán Hervás, J. M. (1976). «El ejército romano y la romanización de la Péninsula Ibérica». Hispania Antiqua, 6: 125-145. Romero Masiá, A. (1992). A Torre de Hércules, revisión bibliográfica. Fuentes para el estúdio de la Torre de Hércules. A. Coruña: Instituto José Cornide de Estudios Coruñeses. Ruiz del Arbol, M. (2001). «Organización y explotación del territorio en el Nordeste de Lusitania en época imperial», PhD dissertation, Madrid. Ruiz-Gálvez Priego, M. L. (1998). La Europa Atlántica en la Edad del Bronce: un viaje a las raízes de la Europa Occidental. Barcelona: Crítica Sampaio, A. (1979). Estudos económicos. I. As vilas do Norte de Portugal. Lisboa: Vega. Sánchez Palencia Ramos, F .J. (1983). «La explotación prerromana del oro del Noroeste de la Península Ibérica». Boletín Auriense, 13: 31-67. Sánchez Palencia Ramos, F. J. (1980). «Alguns vasos cerâmicos inéditos do Museu de Antropologia do Porto». Arqueologia, 1: 12-19. Sánchez Palencia Ramos, F. J. (ed.) (2000). Las Medulas (Léon). Un paisage cultural en la Asturia Augustana. Diputación de León: Instituto Leonés de Cultura. Sánchez-Palencia Ramos, F. J., F. Domingo Plácido, A. Orejas Saco del Valle, and M. D. FernándezPosse (1999). «Minería y metaluria: de la Protohistoria a La España romana». Studia historica. Historia antigua, 17: 263-298. Sánchez-Palencia Ramos, F. J. (1979). «Las explotaciones romanas auríferas de El Caurel». In: Actas del XV Congreso Nacional de Arqueología. Saragoça. Sánchez-Palencia Ramos, F. J., and M. D. FernándezPosse (1993). «Poblamiento y minería prerromana y romana en el Noroeste Peninsular». In: Indígenas y romanos en el Norte de la Península Ibérica, edited by. J. Santos. San Sebastián, 201-225. Sánchez-Palencia Ramos, F. J., and M. D. FernándezPosse (1998) «Las comunidades campesinas en la cultura castreña». Trabajos de Prehistoria, 55 (2) 127¬150. Sánchez-Palencia Ramos, F. J., M. D. Fernández-Posse, J. Fernández Manzano, Y. Álvarez González, and L. López González (1990). «La Zona Arqueológica de las Médulas (1988-89)». Archivo Español de Arqueología, 63: 249-264. Santa Cruz, M. S. C. (2003). «El Puerto de Brigantium y la navegación romana en el atlántico norte?». In: Gijón Puerto Romano. Navegación y comercio en el Cantábrico durante la Antigüedad, edited by C. Fernández Ochoa. Gijón, 121-134. Santos Yanguas, J.N. (1977). Estruturas indígenas del Noroeste peninsular y los cambios de las minas (I a. C. – II p. C.). Oviedo.

Santos, L. A. (1979). «Miliários inéditos da via romana de Braga a Tuy». Arquivo do Alto Minho, 24: 352. Santos, L. A., P. Le Roux, and A. Tranoy (1983). «Inscrições romanas do Museu Pio XII em Braga». Bracara Augusta, 37 (83-84): 183-205. Sarmento, M. (1901). «Materiaes para a archeologia do concelho de Guimarães». Revista de Guimarães, 18 (3-4): 117-135. Sastre Prats, I. (1998). Formas de dependencia social en le Noroeste Peninsular (Transición del Mundo Prerromano al Romano y Época Altoimperial) Pontferrada: Instituto de Estudios Bercianos. Sastre Prats, I. (2001). Las formaciones sociales rurales de la Asturia romana. Madrid: Ediciones Clásicas. Sastre Prats, I. (2002). Onomástica y relaciones políticas en la epigrafía del conventos asturum durante el Alto Imperio. Anejos de Aespa XXV. Madrid. Schülten, A. (1943). Los Cántabros y Astures y su guerra con Roma. Madrid. Sevillano Fuertes, A., and J. M. Vidal Encinas (2002). Urbs Magnífica. Una aproximación a la Arqueología de Astúrica Augusta, Astorga, León. Museo Romano (Guía-Catálogo). Astorga. Shennan, S. (1985). Experiments in the collection and analysis of archeological survey data: the East Hampshire survey. Sheffield. Silva, A. C. F. (1986) A cultura Castreja no Noroeste Português. Paços de Ferreira. Silva, A. C. F. (1999). «A ocupação do território do Noroeste Peninsular aquando da chegada dos romanos». In: Emergência e Desenvolvimento das cidades romanas no Norte da Península Ibérica. Actas da Mesa Redonda. Porto, 39-52. Silva, A. C. F. (2000). «Proto-História e Romanização do Porto». Al-madan, IIª série, 9. Silva, A. C. F. (2007). A cultura castreja no Noroeste de Portugal. Paços de Ferreira: Câmara Municipal de Paços de Ferreira, Museu Arquelógico da Citânia de Sanfins / Centro de Arqueologia Castreja e Estudos Célticos. Silva, D. M. (1958-59). Entre Homem e Cávado. Amares. Silva, M. F. M. (1994). O povoamento proto-histórico e a romanização da bacia superior do Rio Coura: estudo, restauro e divulgação. Cadernos de Arqueologia, Monografias 2. Paredes de Coura: Câmara Municipal de Paredes de Coura. Soeiro, T. (1984). Monte Mózinho. Apontamentos sobre a ocupação entre Sousa e Tâmega em época romana. Penafiel. Sousa, J. J. R. (1971-72). «Nova ara dedicada aos Lares no conventus Bracaraugustano». Bracara Augusta, 25-26, 59-62 (71-74): 179-184. Suárez Otero, J. (1998). «Cerámicas e cultura na Idade do Bronce en Galicia». In: A Idade do Bronce en Galicia: novas perspectivas, edited by R. 240

References

Fábrages Valcarce. Cadernos do Seminário de Sargadelos 77. A Coruña: Ediciós do Castro. Syme, R. (1970). «The Conquest of North-West Spain». In: Legio VII Gemina. León, 83-107. Taboada Chivite, J. (1946). «Via romana del valle del Tamega». Revista de Guimarães, 55 (3-4): 1-12. Taboada Chivite, J. (1968). «Los Tomagani y su romanización». XI CAN: 677-684. Taboada Chivite, J. (1976). «Nuevos testimonios del culto a los Lares Viales en la Galicia». Gallaecia, 2, 193-201. Tarpin, M. (2002). Vici et pagi dans l’occident roman. Coll. École Française de Rome, 299. Rome. Tassaux, F. (1994). «Les agglomérations secondaires de l’Aquitaine Romaine: morphologie et réseaux». In: Les agglomérations secondaires. La Gaule Belgique, les Germanies et l’Occident romain. Actes du Colloque de BliesbruckReinheim/Bitche (21-24 Oct. 1992), edited by J. P. Petit, M. Mangin, and Ph. Brunella. Paris, 197-214. Teixeira, C. (1955-56). «Subsídios para o estudo da arqueologia bracarense – I. O Monte do Castro (Castrum Maximum)». Anais da Faculdade de Ciências do Porto, 21 (4): 231-254. Teixeira, C. e Medeiros, J. R. M. (1973). Carta geológica de Portugal (1/50.000). Notícia explicativa da Folha 5-D, Braga. Lisboa: SGP. Teixeira, R. (1996). «De Aquae Flaviae a Chaves: povoamento e organização do território entre a Antiguidade e a Idade Média», Master dissertation, Faculdade de Letrasa Universidade do Porto. Torres Rodríguez, C. (1948). «Galicia en las guerras cantabras: la tragedia del Monte Medulio». Boletín de la Universidad de Santiago, 51-52: 53-72. Torres Rodríguez, C. (1951-52). «Conquista de Galicia por los romanos, antes las guerras cântabras». Boletín de la Universidad de Santiago, 56:. 79110. Torres Rodríguez, C. (1976). «La conquista romana de Galicia. La Romanización de Galicia». Cuadernos del Seminario de Estudios Cerámicos de Sargadelos, 16: 11-30 Tranoy, A. (1982). «Agglomerations indigènes et villes augustéenes dans le Nord-Ouest Ibérique». In Actes du Colloque de Aix-la-Provence, 1980. Marseille, 125-137. Tranoy, A. (1983). «Remarques sur la permanence et les mutations dans la Galice antique: le rôle des villes». In: Seminario de Arqueologia do Noroeste. Madrid, 193-202. Tranoy, A. (1977). «À propos des “Gallaeci” de Pline. Epigraphis et peuplement». Bracara Augusta, 31: 225-233. Tranoy, A. (1980) “Religion e societé à Bracara Augusta (Braga), au Haut Empire romain”, Revista de Guimarães, vol. III, pp. 67-83, (Actas do Seminário de Arqueologia do Noroeste Península).

Tranoy, A. (1981). La Galice romaine: Recherches sur le Nord-Ouest de la Péninsule Ibérique dans l'Antiquité. Paris: Diffusion du Boccard. Tranoy, A. (1986). «Immigration et émigration. Les mouvements de population dans le conventus d’Astorga». In: Actas del Congreso International Astorga Romana. Astorga, 123-144. Tranoy, A. (1995-96). «La route, image et instrument du pouvoir impérial dans le nord¬ouest ibérique». Cadernos de Arqueologia, II Série, 1213: 31-37. Tranoy, A. e Le Roux, P. (1982). «Nouvelles inscriptions du nord du Portugal». Mínia, V: 31-37. Tranoy, A. e Le Roux, P. (1989/90). «As necrópoles romanas de Bracara Augusta – Les inscriptions funéraires». Cadernos de Arqueologia, 6-7, 183-230. Trément, F., J. Argant, E. Brémon, H. Carvalho, H. Cubizolle, B. Dousteyssier, J.A.López-Sáez, G. Massounie, P. Rigaud, A. Veron (In press). «The Mynedor project. Archaeological and Paleoenvironmental characterization of the ancient Arvernian gold mines of Haute-Combraille (Auvergne, France)». In: The 22nd V.M Goldschmidt Conference. Earth in Evolution, 24-29 Jun., Montreal, 2012. Trément, F., H. Carvalho (2013). «Romanisation et développement: approche comparée des territoires des Bracari (Tarraconaise) et des Arverni (Aquitaine). Une perspective de longue durée (IIe s. av.J.C.-IIe s. ap. J.C.)». In: Simposi Internacional L’Ager Tarraconensis. Paisatge, poblament, cultura material i historia (27-28 Outubro de 2010). ICAC: 249-269. Trément, F. (Dir.) (2013). «Les Arvernes et leurs voisins du Massif Central à l’époque romaine. Une archéologie du développement des territoires». Revue d’Auvergne, 1-2. Van der Leeuw, S., F. Favory, J.-L. Fiches (dir.) (2003) Archéologie et systèmessocio-environnementaux. Études multiscalaires sur la vallée du Rhône dans le programme ARCHAEOMEDES. CRA Monographies, 27. Paris: CNRS. Viana, A. (1926). «Através do Minho: II. A exploração metódica dos nossos castros». Gente Minhota, 6: 88-90; 7: 111-113. Viana, A. (1930) Notas Históricas, Arqueológicas e Etnográficas do Alto Minho. Viana do Castelo. Vidal Encinas (1986). «Arqueologia urbana en Léon: precedentes y aportaciones recientes». Archivos leoneses. Revista de estúdios y documentación de los reynos hispano-occidentales, 79-80: 365380. Vigil, M. (1963). «Romanización y permanencias de estruturas sociales indigenas en la España septentrional». Boletin de la Real Academia de Historia, 152: 225-233. Vigil, M. (1973). Historia de España.I. Edad Antiga. Madrid. 241

The Roman Settlement Patterns in the Western Façade of the Conventus Bracarensis

Villa Valdés, A. (2002). «Periodización y registro arqueológico en el castros del occidente de Asturias». In: Los poblados fortificados del Noroeste de la Península Ibérica: Formación y desenrollo de la Cultura Castreña, Coloquios de Arqueología en la Cuenca del Navia, edited by M.A. Blas Cortina, and, e A Villa Váldés. Navia, 159-187. Villaronga, L. (1970). «Emision monetaria augustea com escudo atribuible a P. Carisius y a la zona norte de Hispânia». In: XI Congreso Nacional de Arqueologia. Zaragoza, 591-600. Vita Finzi, C. (1975). «Related territories and alluvial sediments». In: Palaeoconomy, edited by E S. Higgs. London: Cambridge University Press, 225-231. Vita Finzi, C. (1978). Archaeological Sites in their setting. London: Thames and Hudson. Vita Finzi, C., and E.S. Higgs (1970). «Prehistoric economies in the Mount Carmel area of Paletine: site catchment analysis». Proceedings of the Prehistoric Society, 36: 1-37. VVAA (2005). Roteiro Geira/Via Nova na Serra do Gerês. De Bracara Augusta a Asturica Augusta,. Projecto “ A Geira na Serra do Gerês” – Programa Interreg III A. VVAA (2006). Guia dos Castros da Galiza e Noroeste de Portugal. Vila Nova de Famalicão: Adrave. Wacher, J. (1974). The towns of Roman Britain. London. Wahl, J. (1988). «Três Minas, Vorberiht über die archäeologischen. Ausgraben im Bereich des römischen goldberwerks 1986/87». Madrider Mitteilungen, 29: 221-244. Wahl, J. (1988-1989). «Resultados das pesquisas arqueológicas, efectuadas entre 1998 e 1999, na zona de mineração romana de Três Minas de Jales (conc. Vila Pouca de Aguiar, distr. Vila Real)». Report (relatório dactilografado).

Wahl, J. (1998) «Aspectos Tecnológicos da Indústria Mineira e Metalúrgica Romana de Três Minas e Campo de Jales (Concelho de Vila Pouca de Aguiar)». In: Actas do Seminário Museologia e Arqueologia Mineiras. Lisboa: Ed. Museu do IGM, 57-68. Weber, M. (2004). Historia agraria romana. Madrid: Akal Ediciones. Whittaker, C.R. (1990). «The consumer city revisited: the vicus and the city». Journal of Roman Archaeology, 3: 110-118. Wiegels, R. (1978) «Das Datum der Verleihung des Ius Latii na die Hispaniae: zur Personal und Municipal politik in den ersten Regierung Yahren Vespasians». Hermes, 106: 196-213. Wightman, E. M. (1986). «Le vicus dans de contexte de l’administration et de la societé galloromaine: quelques reflexions». In: Actes du Colloque Le vicus gallo-romaine. Ed. Errance, 59-64. Xavier, C; H. Carvalho ; H. Loureiro (In press). «O parcelamento de Agras Velhas (Tibães, Braga) e o cadastro rural de Bracara Augusta. Estudo das linhas de continuidade». In: XVIII Congreso Internacional de Arqueología Clásica. Centro y perifería en le mundo clásico. Mérida, Maio 2013 Zabaleta Estévez, M. M (2000). «Hallazgos Numismáticos de los comienzos de Bracara Augusta». In: Actas do 3º Congresso de Arqueologia Peninsular, VI, Arqueologia da Antiguidade na Península Ibérica. Porto: Adecap, 395¬399. Zabaleta Estévez, M. M. (1999). «Breve síntese sobre o estudo das moedas de Bracara Augusta». Forum, 26: 80-85.

242