The Western Roman Atlantic Façade: A Study of the Economy and Trade in the Mar Exterior from the Republic to the Principate 9781407307060, 9781407337067

Table of contents :
Front Cover
Title Page
Copyright
Table of Contents
FOREWORD
List of Illustrations
List of Tables
INTRODUCTION
Part One THE NATURE OF THE ROMAN TRADE IN THE ATLANTIC
Part Two THE ATLANTIC ROUTE
Part Three MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD
Part Four THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE
Part Five THE ATLANTIC TRADE ELITES
CONCLUSIONS: THE NEW ATLANTIC FRONTIER
BIBLIOGRAPHY

Citation preview

BAR S2162 2010

The Western Roman Atlantic Façade

CARRERAS & MORAIS (Eds)

A study of the economy and trade in the Mar Exterior from the Republic to the Principate Edited by

C. Carreras R. Morais

THE WESTERN ROMAN ATLANTIC FAÇADE

BAR International Series 2162 2010 B A R

The Western Roman Atlantic Façade A study of the economy and trade in the Mar Exterior from the Republic to the Principate Edited by

C. Carreras R. Morais

BAR International Series 2162 2010

ISBN 9781407307060 paperback ISBN 9781407337067 e-format DOI https://doi.org/10.30861/9781407307060 A catalogue record for this book is available from the British Library

BAR

PUBLISHING

Table of Contents Foreword .............................................................................................................................. iii List of Illustrations ................................................................................................................. v List of Tables ........................................................................................................................ xi INTRODUCTION (C. Carreras & R. Morais) ....................................................................... 1 Part One THE NATURE OF THE ROMAN TRADE IN THE ATLANTIC 1.1 An archaeological perspective (C. Carreras) .................................................................... 7 1.2 Trade and the means of transportation (R. Morais) ........................................................ 21 1.3 The costs of transport in Hispania (P. de Soto) .............................................................. 31 Part Two THE ATLANTIC ROUTE 2.1 The Latin Sea and the Ocean conquest (II-I BC) (C. Aranegui) .................................... 47 2.2 Formal Romanisation and Atlantic projection of amphorae from the Guadalquivir valley (E. García Vargas) ........................................................... 55 2.3 Rome and whale fishing – Archaeological evidence from the Fretum Gaditanum (D. Bernal Casasola)......................................................... 67 2.4 Seaports and fluvial harbours in Portuguese territory – the options for ancient harbour activities within a changing nautical landscape (M.L. Blot)........................................... 81 2.5 Looking towards the North (R. Morais) ......................................................................... 91 2.6 Roman lighthouses on the Atlantic coast (C. Fernández Ochoa and A. Morillo) ........ 109 Part Three THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD 3.1 The structure of Roman military supply: an explanatory model (C. Carreras) ............ 119 3.2 Supplying armies in the Iberian Peninsula during the Republic (P. Edrkamp) ............ 135 3.3 Garrisons, military logistics and civil populations in the Late Republic: Africa and Hispania (T. Ñaco)...................................................................................... 145 3.4 Octavian’s reforms (C. Carreras) ................................................................................. 151 3.5 Demand and military supply in the northwest of Hispania throughout the Early Empire (A. Morillo) ...................................................................................... 157

i

3.6 Hispania: a particular supply network (C. Carreras) .................................................... 173 Part Four THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE 4.1 The incorporation of the Baetican hinterland into the western supply during the Late Republic – A reading based on the distribution of Guadalquivir’s minority amphora types (R.R. de Almeida) ................................................................................ 191 4.2 The Western African amphorae (II BC – I AD) (C. Aranegui) .................................... 197 4.3 A case study from the Haltern 70 amphorae (C. Carreras)........................................... 201 4.4 Bracara Augusta (R. Morais) ....................................................................................... 213 4.5 The mining area of Quinta da Ivanta (R. Morais) ........................................................ 223 4.6 Rías Baixas and Vigo (Vicus Eleni) (A. Fernández) .................................................... 229 4.7 The example of Astorga and León (C. Carreras).......................................................... 239 4.8 The hybrid market system as suggested by Haltern 70 amphorae (R. Morais) ............ 245 Part Five THE ATLANTIC TRADE ELITES 5.1 The Gaditan elites and the figure of L. Cornelius Balbus (C. Carreras)....................... 249 5.2 The Atlantic traders: the Gaditan and Lusitanian elites (R. Morais) ............................ 255 5.3 Was there a true Atlantic vocation among the first emperors? (R. Morais) ................. 259 CONCLUSION: The new Atlantic frontier (C. Carreras & R. Morais) ............................. 261 Bibliography....................................................................................................................... 265

ii

FOREWORD This book, written by leading experts in their particular fields and edited by two of the contributors, César Carreras and Rui Morais, is a welcome addition to an unfortunately slim reading list that deals with the development of the Atlantic Zone through time. The importance of the western seaway as one of the major maritime routes has long been recognized, although much of the evidence has often to be sought in journals which are too frequently difficult to obtain and which are published in a variety of languages. The regions of the Atlantic Zone cover the north-western coastal area of Africa, western southern Spain, Portugal, north-western Spain, north-western France, south-western Britain and Ireland. This long and broad corridor, covering some 4000 kilometres of coastline, faces fierce westerly winds and heavy gales, often making winter sailing extremely hazardous. In spite of the vast geographical breath of this region, it does possess a certain functional unity in two main respects. Firstly, it provided trading networks within and along the different regions that make up the Atlantic corridor and secondly, this impetus encouraged an expansion of local industries which could take advantage of this ‘shipping corridor’ on their doorstep. The Atlantic Zone also offered major routes of entry through the Straits of Gibraltar into the Mediterranean Sea and through the English Channel to the coasts of southern England, northern France and Germany and also into the Irish Sea. The Atlantic corridor provided major exchange networks and proved to be a hinge in the long-distance commerce between north and south; a mobile exchange and mart where goods brought both from within the region and also from outside, could, for example, be despatched for onward transit to the harbours of southern Britain or the Rhineland. At the same time it was both a distribution (‘middle man’) region as well as a creative centre of original production. The period covered by this book, from the Roman Republic to the death of Augustus, was a particularly active one in Atlantic studies, when the advance of the Roman legions westwards into Spain and Portugal and northwards into France and Germany, brought about the interplay of Iron Age and Roman traditions, the progress of Romanization and the exploitation of local resources. Together with a new communications system and the greatly changed political geography of conquest, this all had a dramatic effect on both regional and long distance trade. At this time, the volume of material passing along the Atlantic route greatly increased, partly due to Roman merchants, or negotiators, intent on developing lucrative northern markets and partly due to the legions themselves, which consumed a wide range of commodities that had to be constantly supplied. The present book indicates the very considerable research potential of this region, on the western outskirts of the early Roman Empire, as an area worthy of the same kind of attention as has been enjoyed in the past by other regions of the Mediterranean. The great advantage of this volume is the breath of scope that has been used to address a wide range of topics that cover large areas of the Atlantic corridor and to see them not in isolation but as part of the context of the Atlantic route as a whole. Although inevitably there are case studies of particular areas, industries and amphorae production, iii

these are essentially dealt with in terms of the economy and trading networks of the Atlantic as a whole. This book is bound to be a major source of reference and it is to be hoped that it will also act as a stimulus to further field-work and research all along the Atlantic Zone. David Williams Southampton April 2009

iv

List of Illustrations Figure 1.1.1 Gauloise 4 amphora distribution (densities) .................................................... 11 Figure 1.1.2 Dressel 20 amphora distribution (densities) ..................................................... 12 Figure 1.1.3 Simulation model of transport cost (ARC/INFO) ............................................ 13 Figure 1.1.4 Transport cost simulation from Roman Greece ............................................... 14 Figure 1.1.5 Transport cost simulation from Roman Italy (Campania)................................ 15 Figure 1.1.6 Transport cost simulation from Tarraconensis ................................................. 15 Figure 1.1.7 Transport cost simulation from Narbonense .................................................... 16 Figure 1.1.8 Transport cost simulation from Aquitania ....................................................... 16 Figure 1.1.9 Transport cost simulation from Baetica ........................................................... 17 Figure 1.1.10 Transport cost simulation from Africa ........................................................... 18 Figure 1.1.11 Transport cost simulation from Lusitania ...................................................... 19 Figure 1.2.1 Commercial routes and communication roads in NW Pensinsula (Beltran, 2008)................................................................................................................ 22 Figure 1.2.2 Carruca from Bosceaz (Switzerland) .............................................................. 23 Figure 1.2.3 Carruca with skin from a Pompei’s painting.................................................... 23 Figure 1.2.4 Inscription of urinatores (Ostia) ...................................................................... 24 Figure 1.2.5 Lighthouses ...................................................................................................... 25 Figure 1.2.6 Downloading cargoes (saccarii) ...................................................................... 26 Figure 1.2.7 Port arrival (mosaic from Rimini) .................................................................... 27 Figure 1.2.8 Mosaic with mensores...................................................................................... 27 Figure 1.2.9 African mosaic with a ship loaded with amphorae .......................................... 28 Figure 1.2.10 Pompei. Graffiti of a ship entitled Europa done by sailors on the wall of shop in Via Dell`Abundanza .................................................................... 28 Figure 1.2.11 Oberstimm boat’s model ................................................................................ 29 Figure 1.2.12 Reconstruction of Xanten’s Roman boat ....................................................... 29 Figure 1.3.1 Example of an ordering diagram with simple braches to the network ................................................................................................................. 32 Figure 1.3.2 Example of a graphs diagram (Social Network) rating nodes according its degree of centrality (Network Degree) ...................................................... 32 v

Figure 1.3.3 Main communication routes in Pre-Roman times ............................................ 34 Figure 1.3.4 Main communication network from Flavian period onwards. Main ports in the Cantabric coast: 1. Brigantium 2. Noega 3. Gigia 4. Portus Vereasuecae 5. Portus Blendium 6. Portus Victoriae Iulobrigensium 7. Portus Amanum 8. Oiasso ................. 35 Figure 1.3.5 Map with the main communication land and river network in the NW ........... 37 Figure 1.3.6 Reproduction of the Peutingerian tabula .......................................................... 38 Figure 1.3.7 Map where the main Roman communication axis turned up and the conventos capitals .............................................................................................. 38 Figure 1.3.8 Transport cost map of distributing products coming from the Guadalquivir valley (Carreras, 1996) ............................................................... 39 Figure 1.3.9 Map of the NW Peninsula with the main nodes and edges of the resulting graphs diagram ............................................................................................................... 40 Figure 1.3.10 Map of nodes interpolation from graph diagram. In white, there are zones with little communication and grey is the opposite ............................... 40 Figure 1.3.11 Map with the main communication axes ....................................................... 41 Figure 1.3.12 Graphs diagrams representing the main nodes of the Iberian Peninsula ........ 42 Figure 1.3.13 Interpolation maps of nodes in the Iberian Peninsula .................................... 43 Figure 2.1.1 Map of the Western Mediterranean ................................................................. 49 Figure 2.1.2 Map of the Strait of Gibraltar........................................................................... 51 Figure 2.1.3 Sculpture of Hercules-Anteo............................................................................ 52 Figure 2.2.1 Location of the roman towns mentioned in the text ......................................... 56 Figure 2.2.2 Late republican and Augustean amphorae from Guadalquivir Valley: 1. Dressel 1 A; 2. Dressel 1 B-C; 3. Dressel 2-4; 4. LC 67; 5. Haltern 70; 6. Urceus typus; 7. Oberaden 83/Dressel 20; 8. Haltern 71/Dresel 20; 9. Pellicer D. (after Berni 1998; García Vargas and Bernal 2008 and Morais 2007) ............................................................................................................ 57 Figure 2.2.3a: Ceramic kilnsites at Carmona (ancient Carmo) (after Rodríguez Rodríguez 2001) .................................................................................. 58 Figure 2.2.3b Amphorae productions from “13-15 Dr. Fleming St.” kilnsite (Carmona) 1. Haltern 70, 2. Pellicer D. (after Conlin et al., forthcoming) ....................................... 58 Figure 2.2.4 Amphorae production (except n. 6) from Mesones st. Kilnsite: (Alcalá del Río, ancient Ilipa): 1-2. Haltern 70; 3-4. Haltern 71/Dressel 20; 5. Dressel 28; 7. Haltern 70; 8. Haltern 71/Dresel 20; 9. Dressel 2/4 Rhodian? (local imitation), 10. Haltern 70; 11. Haltern 71/Dressel 20........................................... 59 Figure 2.2.5 Republican deposit from Italica (Santiponce, Sevilla): 1. Locally produced Dressel 1B; 2. Locally produced Dressel 1C; 3. LC 67; 4. Class 24/Oberaden 83?; 5. Oberaden 83/Dressel 20. Sc. 1:10. (After García Fernández 2004) ....................................................................................... 60 Figure 2.2.6 Sevilla (ancient Hispalis). Location of archaeological excavations mentionen in the text. Grey dotted line indicates the (hypothetical) Roman republican wall-circuit ....................................................................................... 61 Figure 2.2.7 Amphorae from n. 25, Alemanes st. (Sevilla): Context 164, n. 1-9; context 161, n. 10-16 ...................................................................................................... 62 Figure 2.2.8 Amphorae from n. 8. Fabiola St. (Sevilla): Level 18, n. 1; level 16, n. 2-9; level 14, n- 10-19 .................................................................................. 63 Figure 2.3.1 Idealized cetacean hunt by Basque fishermen in Greenland, performed with the kind of equipment described by Oppian, in a 1715 print (Cazeils, 2000, 52) ................................................................................. 69 vi

Figure 2.3.2 Cetacean vertebra from Baelo Claudia, showing cut marks, from which we can infer the used it was put to (courtesy of I. García Jiménez) ..................... 70 Figure 2.3.3 Cetacean vertebra used as an anvil, showing cut marks on the cranial joint face, from the Industrial Complex I, numbers 3-5 San Nicolás Street, in Traducta/Algeciras...................................................................... 70 Figure 2.3.4 Stratigraphic section (A) and view from above (B) of Trench 4, number 3, África Square, including the strata (SS.UU. 4018 and 4042) in which the burned cetacean rib (C) and other bone remains (D) were found (Illustrations from Bernal, Lorenzo, Sáez and Bustamante, 2007 b) .............................. 71 Figure 2.3.5 Clays disc from Tamuda, depicting a harpoon armed figure riding a hippocampus, surrounded by big sea creatures, a shark and a whale among them (Fumadó, 2006, 2013, figs. 2 y 3).......................................................................... 73 Figure 2.3.6 Circular vats from the Industrial Complex I in Baelo Claudia (A), and drawings of the preserves factory at Torre Vindicari (B), south of Syracuse (Púrpura, 1989, fig. 9) .................................................................................................... 74 Figure 2.3.7 Fishing harpoons from the western Mediterranean A. Bronze harpoon from San Martí d’Empuries, VIth century BC (Castanyer, 2006, 22) B. Small bone harpoon from Traducta, early VIth century AD (Bernal, 2009, eds., chapter 26) C. Bone double-headed harpoon from Castrum Perti, early Byzantine period (De Vingo and Fossati, 2001, 659, fig. 95,1) ............................ 75 Figure 2.3.8 Chained hook from the suburbs of Pompey (A; Stefani, 1991, 14), and double anchor from Pisa’s harbour (B; Bigagli, 2000, 97, fig. 4) ............................ 75 Figure 2.3.9 Hippocaustum from the preserves factory of Tahadart –A- (Ponsich, 1988, 144, fig. 76), and draining structures from Gigia’s cetaria – B- (Fernández Ochoa, 1994, 143, fig. 21), potentially linked to cetacean by-products exploitation ................................................................................................. 76 Figure 2.3.10 Archaeological sites related to whaling in Antiquity, due to faunal evidence (1-4 and 7), facilities potentially used in whaling (3, 5, 8 and 9), and iconographic depictions of cetaceans (6): 1.- Manilva Castle; 2.- Traducta; 3.- Baelo Claudia; 4.- A Lanzada; 5.- Gijón; 6.- Tamuda; 7.- Septem Fratres; 8.- Cotta; 9.- Tahadart ..................................................................... 78 Figure 2.4.1 Ancient coast and estuaries during the maximum of the Eustatic sea-level rise (Ribeiro, Lautensach, Daveau, 1987) ....................................................... 82 Figure 2.4.2 Ancient Tagus and Mondego estuaries and their Phoenician settlements. (Bronze Age sites lie close to the upper parts of the paleogeography of Mondego river) (Priego, 1998) ................................................................................... 82 Figure 2.4.3 Punic importations (Fabião, 2001) ................................................................... 83 Figure 2.4.4 Ancient lead anchor stocks as they were recorded before 1990 (Alves et al., 1988-1989) ................................................................................................ 83 Figure 2.4.5 Fluvial routes according to medieval accounts (Priego, 1998) ........................ 83 Figure 2.4.6 Iberian mining places in ancient times (Priego, 1998) ..................................... 84 Figure 2.4.7 Navigable Western Iberian rivers according to Pomponius Mella, Pline and Strabo texts (apud Parodi Álvarez, 2001) ....................................................... 84 Figure 2.4.8 Roman viae and ancient urban centres working as nautical termini (Blot, M.L., 2003 apud Daveau, 1995)........................................................................... 86 Figure 2.4.9 (from left to right): A. Phoenician/Punic importations and settlements; B. Nautical pre-roman and roman underwater finds (anchor stocks and sounding leads) corresponding to mooring sites, harbours and presumable shipwrecks; C. Pre-roman and roman amphorae found under the sea and under fluvial waters (mooring contexts and presumable shipwrecks) (Blot, M.L., 2004)........................................................................................................... 87 vii

Figure 2.4.10 Seaports and fluvial harbours in the ancient Southwestern Iberia (Portugal) overlapping Justino’s cartography of the navigable parts of the Portuguese rivers showing the regional nautical transportation still active in the XIXth century (Blot, 2003a apud Justino, 1988): 1 Atrium (Viana do Castelo); 2 Portus; 3 Talabriga; 4 Aeminium (Coimbra); 5 Eburobritium; 6 Sellium (Tomar); 7 Scallabis (Santarém); 8 Olissipo (Lisboa); 9 Caetobriga (Setúbal) 10 Tróia; 11 Salacia (Alcácer do Sal); 12 Sines (seaport of Mirobriga); 13 Lacobriga (Lagos); 14 Ipses (Alvor); 15 Arade estuary (P. Hannibalis?); 16 Silves; 17 Ossonoba (Faro); 18 Balsa (T. de Ares/Tavira); 19 Baesuris (Castro Marim); 20 Myrtilis (Mértola)....................... 88 Figure 2.4.11 Seaports and fluvial harbours in the ancient Southwestern Iberia studied as city contexts-containing archaeological memories of the corresponding harbours (Blot, M.L., 2004a) ............................................................ 89 Figure 2.5.1 Map of the Mediterranean with the location of Olisipo. Places where the amphorae found in Castelo de S. Jorge came from: 1. Tirrenica Coast in the Italian Peninsula; 2. Adrian Coast; 3. City of Brindisi; 4. Tripolitania; 5. Ebusus; 6. Area of the Strait of Gibraltar (in Pimenta 2005) ...................................... 94 Figure 2.5.2 View of the entrance of the bar of the Tejo River and its estuary (in Pimenta 2005) ........................................................................................................... 95 Figure 2.5.3 Portas do Sol – Alcáçova de Santarém (in Viegas 2003) ................................. 95 Figure 2.5.4 Geomorphologic evolution of Peniche’s coastline (in Blot 2003) ................... 96 Figure 2.5.5 Plan of the location of the old Roman pottery of Morraçal da Ajuda in the present city of Peniche (in Cardoso e Rodrigues 2005) ........................................ 97 Figure 2.5.6 Plan of Porto (1833) – Academia Real de Ciências ......................................... 98 Figure 2.5.7 Traces of wharves of the Modern Era by the mouth of the Douro River, on Gaia’s shore (Afurada)................................................................... 98 Figure 2.5.8 Photography of the old building of Aljube (Porto) .......................................... 99 Figure 2.5.9 Limit of the basin of the Leça River .............................................................. 100 Figure 2.5.10 Photography from the early 20th century, showing the mouth of the Leça River before the construction of Leixões Seaport ...................................... 100 Figure 2.5.11 Sunset view in the mouth of the Minho River from Santa Tecla Hill .......... 101 Figure 2.5.12 Photograph of the mouth of the Minho River .............................................. 101 Figure 2.5.13 General plan of Santa Tecla ......................................................................... 102 Figure 2.5.14 Underwater findings recovered in Ria de Vigo ............................................ 103 Figure 2.5.15 Plan with the location of Praia do Areal ...................................................... 104 Figure 2.5.16 Location of the port enclave in La Coruña................................................... 104 Figure 2.5.17 Reconstruction of the lighthouse in Roman times (T. Hauschild) ............... 105 Figure 2.5.18 Flat Base amphorae, urceus type, from Gadalquivir .................................... 107 Figure 2.5.19 Lusitanian Ovoid amphorae ......................................................................... 108 Figure 2.6.1 Recreation of the lighthouse of Brigantium in Roman times (E. and P. Cabarcos, Factoría Gráfica) ......................................................................... 110 Figure 2.6.2 Eastern lighthouse of Dover, near St. Mary-in-Castro Church (A. Goy Diz) ................................................................................................................. 110 Figure 2.6.3 Reconstruction of the Eastern lighthouse of Dover (G. de la Bédoyère, 1991) ............................................................................................. 110 Figure 2.6.4 Probable location of the Roman lighthouse of Gijón (FMC and UP Gijón) .................................................................................................... 111 Figure 2.6.5 Calpurnius’ plaque (9-10 BC) from the Campa Torres.................................. 111

viii

Figure 2.6.6 Excavations by Manuel Reguerra in the Campa Torres in 1783 .................... 112 Figure 2.6.7 Campa Torres (Gijón). Drawing of the remains of the tower after the excavations undertaken in the year 1884 according to Nemesio Martinez ............. 113 Figure 2.6.8 Roman lighthouses and maritime routes in the European Atlantic coast (Fernández Ochoa et al., 2005) .................................................................................... 114 Figure 3.1.1 Distribution map of beneficiarii ..................................................................... 122 Figure 3.1.2 Simulacra Haltern .......................................................................................... 125 Figure 3.1.3 Legionaries in the Trajan’s column ............................................................... 126 Figure 3.1.4 Castro Urdiales patera .................................................................................... 129 Figure 3.1.5 Picture of Blanco Cela 21 (Astorga) .............................................................. 131 Figure 3.4.1 Map of republican military camps (Morillo) ................................................. 152 Figure 3.4.2 Map of the early division of Spanish provinces (27 BC) ............................... 153 Figure 3.4.3 Map of the late division of Spanish provinces (16-13 BC) ............................ 154 Figure 3.5.1 Distribution of the Roman camps in the north-western area of the Iberian Peninsula during the Cantabrian Wars (according to A. Morillo) .............. 159 Figure 3.5.2 Moneta militaris with reverse of caetra found at Astorga (courtesy: Imagen MAS) .............................................................................................. 160 Figure 3.5.3 Distribution of the Roman camps in the north-western area of the Iberian Peninsula during the Armed Peace (19/15 BC-10/20 AD) (according to A. Morillo).............................................................................................. 161 Figure 3.5.4 Lamp of the type Dressel 4 (Vogelkopf-lampe) from the sector San Millan (Herrera de Pisuerga), manufactured in the military workshop of legio IIII Macedonica in this town (Photograph: A. Morillo) .................................. 163 Figure 3.5.4 Base of container of local terra sigillata of Italic tradition signed by C. Licinius Maximus, potter linked to the Augustan camp of the legio VI victrix (Photograph: V. García Marcos)................................................................................... 166 Figure 3.5.6 Lamp derived from Dressel 3 (“Andujar type”) found in the excavation in Casa Pallarés (León) (Photograph: A. Morillo) ............................ 167 Figure 3.5.7 Local production of thin-walled pottery of the “Melgar de Tera type” (courtesy Imagen MAS) ............................................................................................... 169 Figure 3.5.8 Military stamp of legio VII gemina found in León (courtesy Museo de León) ............................................................................................ 171 Figure 3.6.1 Distribution of beneficiarii et al..................................................................... 174 Figure 3.6.2 Map of densities of Dressel 20....................................................................... 177 Figure 3.6.3 Castro Urdiales patera .................................................................................... 178 Figure 3.6.4 Mapa Hispania amb els llocs de producció d’oli ........................................... 179 Figure 4.1.1 Guadalquivir’s late Republic minority types ................................................. 193 Figure 4.1.2 Map – Widespread of the minority types on the atlantic façade .................... 196 Figure 4.2.1 T-7.4.3.3. Amphorae; Haltern 70; Sala 1/LC67; with almond-shaped rim; Dr. 8 y Dr. 9 present in Lixus between 50 BC and 10 AD.................................... 198 Figure 4.2.2 Lixus. Trade of amphorae between 10 and 50 AD ........................................ 199 Figure 4.2.3 Lixus. Dr. 20A; Dr. 2-4; G4; cylindrical and Cos amphora. 10-50 AD ...................................................................................................................... 200 Figure 4.2.4 Findings in Khedis workshop, near Salé (courtesy of H. Najdi) .................... 200

ix

Figure 4.3.1 Haltern 70 amphorae from Cala Culip VIII ................................................... 202 Figure 4.3.2 Haltern 70 tipology of Cala Culip VIII .......................................................... 203 Figure 4.3.3 Workshop production of Haltern 70 – “Simpósio Rota Altântica (Peniche, Portugal)” ..................................................................................................... 204 Figure 4.3.4 Imitations: 3 and 4, Verulamium 1908; 5, Haltern 70 similis; 6, London 555............................................................................................................... 205 Figure 4.3.5 Production areas of Baetican Haltern 70 ........................................................ 207 Figure 4.3.6 Quantified (cg/m2) distribution of Haltern 70 (Carreras et al., 2005) ............ 209 Figure 4.3.7 Quantified (cg/m2) distribution of Haltern 70 (2007) .................................... 210 Figures 4.4.1a & 4.4.1b Graphs of amphorae finds (in Carreras Monfort, 1996) .............. 214 Figures 4.4.2 Hispania – Costs of transportation from the Baetica (in Carreras Monfort, 1996).......................................................................................... 217 Figure 4.4.3 Location of Bracara Augusta in the context of the road and sea route networks ................................................................................................. 218 Figure 4.4.4 Origin by areas and centers of production of the Early Empire products imported by the Roman city of Bracara Augusta .......................................... 220 Figure 4.5.1 View of the excavation (northeast/southeast area) ......................................... 224 Figure 4.5.2 Underground gallery ...................................................................................... 224 Figure 4.5.3 Military chart 1:25.000, Page 123 (41º11’15’’ Latitude North; 8º 29’4’’ Longitude West (Greenwich) ............................................... 225 Figure 4.5.4 The mines in the Mountains of Santa Justa and Pias in the context of the Northwest and of the Iberian Peninsula ....................................... 225 Figure 4.5.5 Chart of the imported amphorae from Quinta da Ivanta ................................ 226 Figure 4.5.6 Amphorae and sigillata from Quinta da Ivanta .............................................. 228 Figure 4.6.1 Location of the Rias Baixas and Vigo (Vicus Eleni) in the Iberian Peninsula ................................................................................................ 230 Figure 4.6.2 1: Dressel 1C, underwater find (Photo: Archive of the Municipal Museum Quiñones de León, Vigo); 2, 4 and 5: Lomba do Canho 67, Beiramar tunnel, Vigo (Cortegoso, 2004); 3: LC 67 Castro de Montealegre (González Ruibal et al., 2007) ...................................................................................... 229 Figure 4.6.3 1 and 2: T-7 of Beiramar tunnel, Vigo (Cortegoso, 2004); 3 and 4: T-7 of Castro de Montealegre (González-Ruibal et al., 2007)........................ 232 Figure 4.6.4 1: Oberaden 83 of Beiramar tunnel, Vigo; 2: Haltern 70 of Beiramar tunnel, Vigo (Cortegoso, 2004) ................................................................ 233 Figure 4.6.5 1: Almagro 50C of the villa of Toralla; 2: LRA 1B of Vigo (Photos: Archive of the Municipal Museum Quiñones de Leon, Vigo) ....................... 235 Figure 4.6.6 Amphorae from the Rias Baixas. 1. Dress. 1A; 2. Dress. 1B-C; 3. Dress. 2-4; 4. Lamb. 2; 5. Amphora of Apuleya or Brindisi; 6. Pellicer D; 7. Haltern 70; 8. LC 67; 9. T-12; 10. T-7; 11 y 12: Dress. 7-11; 13. Haltern 70. 14. Oberaden 83; 15. Dress. 20; 16 y 17. Dress. 7-11; 18. Pascual 1; 19. Dress. 2-3; 20. Dress. 14/Beltrán I; 21. Beltrán IIA; 22. Beltrán IIB; 23. Almagro 50C; 24. Regional II; 25. Africana II; 26. Keay 35; 27. Keay 62; 28. Keay 61; 29. LRA 3; 30. LRA 4; 31. LRA 1B; 32. LRA 2 .................................... 237 Figure 4.7.1 View of Gijon bay including the site of Campa Torres ................................. 242 Figure 4.7.2 Percentages of Haltern 70 and Baetican fish-sauce amphorae in the NW ..... 243 Figure 4.7.3 Interpolation of Haltern 70 densities in the Iberian Peninsula ....................... 244 Map of the main sites mentioned in the text....................................................................... 263 x

List of Tables Table 1.1.1 Amphorae percentages in early contexts (Augustan or Julio-Claudian period) ................................................................................ 9 Table 1.1.2 Amphorae percentages of 2nd century AD contexts ............................................ 9 Table 1.1.3 Amphorae percentages of Vth century contexts ................................................... 9 Table 2.3.1 Histogram of whale fishing evidences in the Ancient World ............................ 77 Table 3.1.1 Quantified table of the amphorae from Blanco Cela, 21 (Astorga) ................. 131 Table 3.4.1 Categories of the cities in Roman Spain from Pliny the Elder ........................ 154 Tabla 3.6.1 Densities of amphorae in different sites in Hispania ....................................... 176 Table 4.4.1 Amphora found in Braga in the Early Empire period ..................................... 215 Table 4.4.2 Amphorae from Astorga (in Carreras Monfort e Piero Berni, 2003) .............. 216 Table 4.7.1 Amphora quantities from Asturica Augusta (Carreras and Berni, 2003) ........................................................................................... 240 Table 4.7.2 Amphora quantities from Legio – León (umpublished report) ........................ 241 Table 4.7.3 Amphora quantities from Campa Torres (Carreras, 2001) .............................. 242 Table 4.7.4 Amphora quantities from Petavonium (umpublished report) .......................... 242 Table 4.7.5 Amphora quantities from Tiermes (umpublished report) ................................ 243 Table 4.7.6 Amphora quantities from Uxama Argaela (umpublished report).................... 243 Table 5.1.1 List of tria nomina present on Baetican Dressel 7-11 (Etienne and Mayet, 1998; Etienne, 2002; Lagóstena, 2002-2003).............................. 252

xi

INTRODUCTION César Carreras & Rui Morais It is not our intention in this introduction to put forward a summary of all the issues presented by the different contributors of this work. Each author deals with specific problems even though they are all obviously interrelated from a common denominator that led to the title of this work: The Western Roman Atlantic façade: a study of economy and trade in the Mar Exterior from the late Republic to the Principate.

The role of the army must, however, be duly pointed out in what regards the reorganization and establishment of new frontiers. Even though the literary and epigraphic sources tell us that numerous amounts of military contingents participated in these warlike episodes, the truth is that the information we have about these contingents is scarce; for example, we do not know their setting places or the war fronts in which they participated.

In this brief introduction we will try to approach the issue of two of the main routes of circulation in the peninsular territory: the Atlantic coastal circuit, which is part of the Outer Sea, and the overland route through the ancestral Silver Route.

Actually, the written evidences about the development of the military campaigns during Augustus’s period in the north of the Peninsula are scarce and fragmented. These evidences are usually isolated, sometimes even anecdotic, passages taken from Suetonius, Horace, Veleius Paterculus or from the Anthologia Palatina and they are also mentioned in Floro’s, Dio Cassius’s and Orosius’s works. This reality is, however, contrary to the different interpretative scenarios created by the history on the Cantabrian Wars. Such excessive interpretations, solely based on arguments of toponymic nature resulted in a series of “unquestionable truths” and as often as not archaeology was only resorted to when it could be used to corroborate some statement.

As you will see throughout this work, the study of the economy and of the trade on the Atlantic façade is at first closely connected to the conquest process of a territory that culminates with the Astur-Cantabrian Wars at the end of the Ist century BC. These wars were justifiable taking into account the mineral richness of the Northwest in tin, and especially in gold, which the Romans were well aware of because of the torques and other jewellery elements the native warriors wore.

The insufficiency of these arguments regarding the Cantabrian War issue has led many people to try to approach the subject through the archaeology. A good example is the critical study written by Enrique Gutiérrez Cuenca and José Ángel Hierro Gárate, entitled “La Guerra Cantábrica: de ficción historiográfica a realidad arqueológica” (Gutiérrez Cuenca and Hierro Gárate, 2001: 71-96), where this question is analysed. This study points out the poor results of a traditional historiography solely based on supposedly archaeological and toponymic arguments in order to emphasize the importance of a methodological updating based on systematic archaeological excavations.

This conflict has also been understood as a declared political propaganda since the victory over these peoples gave Augustus a remarkable fame as the peace builder and guardian of the Peninsula’s prosperity, if not of the Empire. On the other hand, as Chic Garcia points out (1997: 1218), the constitutional pact of 27 BC, which aimed to occupy the whole north of the Iberian Peninsula, still had an Atlantic dimension that was not well stressed. It should unite, under Rome’s direct rule, the Northwest territory with the Gaulish one, which would in the future favour the maritime communications in the north-south axis of the Empire and, in the mid run, a permanent contact with the Baetica province.

The result of these new researches based on excavations has, in fact, given us fundamentally different data from 1

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

the ones provided by the traditional historiography, which let us revaluate the knowledge so far acquired (Fernández Ochoa and Morillo Cerdan, 1999).

As we will find out throughout this work, one of the key issues for the study of the economy and trade in the Atlantic façade is closely connected to the subject of the military contingents’ movement in the northern region of the Peninsula. It was necessary to organise extensive supply networks for the troops, from products of first need to manufactured ones. Depending on the political and military situation of the moment, varied processes and different solutions were adopted.

A last and major contribution to this issue is included in the II Congreso de Arqueologia Militar Romana en Hispânia organised in Leon in 2005 which, together with the I Congreso held in Segovia in 1998, allow us to reconsider this subject under a different perspective.

As Ángel Morillo Cerdán (2006, 47) stated, only the amphorae allow us to know which are the productive regions and the supply routes for the Augustan camps in the north of the Peninsula. As a matter of fact, as it has been recently demonstrated (Morais e Carreras, 2004, 93112) regarding the “consume geography of the Haltern 70 amphorae”, one has verified an overwhelming presence of these amphorae in the Northwest Peninsular context, which can only be explained through the supply of the coastal populations and of the armies in campaign (id. ibidem).

Indeed, a set of places, directly or indirectly related to these episodes, has been identified for the routes of Roman military penetration in the territory of the former Cantabria (vd. Gutiérrez Cuenca e Hierro Gárate, 2001; Morillo Cerdán, 2006, p. 33-74). On the other hand, little is known, and has been discussed, about the area of penetration of one of the three columns that set off from the coastal area of Lusitania ad moved northwards, reaching Galician territory. The references to a tripartido exercitu by Floro (II, 33, 48) or, in the words of Orosius (VI, 21, 4), to a tribus agminibus, allude to the 26 BC campaign carried out by Antistius Vetus from his military camp in Sesigama (Sasamón, Burgos). The emperor himself participated in this campaign while, simultaneously, the Gaulish armada from Aquitaine completed the attack on the Cantabrian coast (Strabo, III, 4, 18; Floro, II, 33, 46; Orosius, VI, 21, 4).

Even though the Haltern 70 amphorae are an unambiguous indicator of the circulation of these products through the Atlantic route, the same phenomenon is well documented for other type of amphorae and other ceramic goods that accompanied them (vd. Naveiro, 1991; Morais, 2006). However, if the supply of the armies on the coastal area and on the areas that benefited from the river system was carried out naturally through the Atlantic route, as we have previously mentioned, the situation is more complex when it gets to the supply of inland places with strategic interest, such as the Asturian camps.

However, though the 26 BC campaign turned out to be victorious, it was not enough to guarantee the Northwest occupation. In 25 BC the operations were thus resumed from Gallaecia and Asturias and culminated in the assault to Mons Medullius. Publius Carisio commanded this expedition from a military camp still undiscovered by the archaeology. Schulten (1943, 174) suggested that its location might have corresponded to Bracara Augusta.

There are two different perspectives regarding the supply of these inland places: one defends the supply through the Atlantic trade circuit (vd. Morais e Carreras, 2004, 93112) whereas the other is for the overland route through the ancestral Silver Route (vd. Morillo, 1999, 325; 2006, 56).

The location of a camp in the place where this Roman town would be founded cannot be proved though. Indeed, despite the growing number of pre-Roman evidences so far collected in Bracara Augusta (Morais, 2005, p. 125138), none of them can be considered clear evidence of the existence of such a camp. An exception has to be made for some twenty metal objects gathered out of context belonging to soldiers’ equipment (ornaments, buttons, pieces of belt, hooks) as well as parts of their weaponry, especially concentrated in the western and older sector of the town.

Nevertheless, we will see that both routes are of the utmost importance for the supply of the armies in the Peninsular Northwest despite being dependent on the political and military situations of the moment. Regarding the Atlantic route, two fundamental aspects stand out: a strong concentration of imported goods on the coastal areas and areas benefiting from the river network and a minor concentration of these same goods in the inland regions.

It is, however, possible that there might have been a base of operations in the Northwest, to the south of the Minho River.1 If this theory is to be proved, the treasures collected in Citânia de Sanfins and in the so called “Castro de Alvarelhos” indicate the existence of military operations in the region.

Besides the coastal areas and the ones benefiting from the river networks, some other inland urban nuclei stand out, such as Bracara Augusta. The economic and commercial importance of this town has already been duly highlighted (vd. Morais, 2006), but we still have to mention its importance in the context of the overland circulation, acting as a road junction (XVI, XVII, XVIII, XIX e XX) only comparable in the Peninsula to Emerita Augusta and Caesaraugusta.

1 It can not be highlighted too often that the first major Roman incursions in the Peninsular Northwest seem to have already occurred with D. Junius Brutus. They are frequently associated to the destructions that took place in Sabroso I (Guimarães), Cividade de Terroso (Póvoa do Varzim) and Coto da Pena (Caminha) (vd. Silva, 1983-84; 1986).

2

INTRODUCTION

same happens in the route between Seville and Cadiz with the consequent distribution of goods through the river and sea routes (id. ibidem). In fact, as we have personally found out for the Roman city of Bracara Augusta (Morais, 2006), these products are inclusively well documented in relatively inland places but that directly benefit from the circulation of goods through the Atlantic route.

The Silver Route was, as we have already said, equally important for the supply of the armies stationed in the Asturian lines of combat. This was an exclusively overland route that went from Emerita and linked this Lusitanian capital with the Asturian territory through northern Lusitania (Iter ab Emeritam Asturicam) (vd. Morillo, 1999, 325; 2006, 56). This is indeed an ancestral route, very active ever since the VIIth-VIth century BC, used by the Tartessian “intermediaries” to obtain tin to be used in the trade with the eastern peoples and subsequently intensified by the Turdetans to obtain that and other minerals (Naveiro Lopez, 1991, 140).

Nevertheless, this does not devalue the importance of the Silver Route. The about 42 derived Dressel 3 documented oil lamps in Astorga and Léon are a solid proof that this route was used for commercial purposes. In this context, we should also take into consideration Ángel Morillo Cerdán’s suggestion (2006, 47) regarding the supply of food through this route. The author says (id. ibidem) that the practical absence of Baetican containers for the transport of olive oil (fundamental in the food diet and essential as fuel) suggests the use of non ceramic containers, such as wineskins and leather pellejos (cullae), whose light weight would make their transport from Emerita to the Asturian camps through that route easier. As a matter of fact, this hypothesis helps us explain the presence of the above mentioned oil lamps.

As Ángel Morillo Cerdán (2006, 33-34) has well pointed out, the displacement of a considerable military contingent in the northern region of the peninsula from the reign of Augustus onwards brought about many supply problems, which became one of the major problems the Roman military administration had to deal with. As far as this issue is concerned, the foundation of the Roman colony of Emerita Augusta in the centre of Lusitania, after the official conquest of the Northwest in 25 BC (Dio Cassius, Hist. Rom. LIII, 25, 8-26, 1), was vital since it was a place of redistribution and of regional trade of consumer and manufactured goods. According to that author (id. 55), it would be precisely in this colony that the negotiatores or mercartores would get their products from to sell in Asturica or to supply the military troops stationed in Herrera de Pisuerga, León and Rosinos de Vidriales.

And yet we have to agree with Fº Germán Rodríguez Martín (2006, 201) when he says that, with the pacification and the transition of some settlements to towns, such as Astorga, the Baetican olive oil may have been preferably transported using the maritime routes. However, what is at stake here should not be which of the two routes is more or less important. As a matter of fact, we are in possession of many data that unequivocally testify the existence of a relation of privileged contacts between the town and the northern peninsular region, which we could call of “colonial” type. As we know, the foundation of Emerita as a colony in 25 BC is related to the context of pacification after the surrender of the Cantabrian and Asturian resistance as a result of the campaigns carried out by Augustus and his generals. As a reward for the victory, a first contingent made up of veterans of the V alaudae and X gemina legions settled in Merida under the authority of Augustus’s legate (legatus), Publius Carisius.

However, as we know, this question is only valid for the period of Tiberius. Indeed, as Fº Germán Rodríguez Martín (2006, 189) has well pointed out, using the data so far collected in Merida, the town does not grow significantly up to the period of Tiberius. If from the Tiberirus period onwards, the oil lamps and the thin walled ware from Emerita Augusta are good indicators of an overland route through the ancestral Silver Route (Morillo Cerdán, 2000, 629-630; 2006, 3374), it does not necessarily mean that this was the route of greater impact for the supply of the armies camped in the Asturian region.

In 16 BC, when he begins his second trip to the West, Augustus draws a new reform that will delineate a new Hispania. As a consequence, between 16 and 13 BC, Emerita gains a new importance becoming provincial capital.

As a matter of fact, these products are not quantitatively significant to value this route over another one, namely over a supply carried out from the coastal area of the Atlantic façade. On the contrary, if we follow the ceramic records so far collected in the peninsular territory, we will see that the largest bulk of products from Emerita is found towards the sea routes, as Fº Germán Rodríguez Martín (2006, 197) has highlighted.

The relation of privileged economic and commercial ties between the north and the south of the Peninsula through the Silver Route is, however, equally well documented in the opposite direction, at least from the mid Ist century. We refer to the well known abundance of Tricio’s potters (La Rioja), Valernus Paternus and Lapillius, in Emerita, which has led people to think that their ware was produced in Merida (vd. Mayet, 1970, 1-13). But it is still within the scope of the Hispanic terra sigillata produced in the region of Tricio that one can document the

As the author points out (id. 198), one comes across a large number of ceramic ware from the workshops from Merida in the paths that link Olisipo, Salacia, Scallabis and Italica-Hispalis and along all the Atlantic façade. The 3

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

On the other hand, the recent discovery of a small production centre in Merida, which may have been active from the beginnings of the Ist to the IInd century, is an indicator of the importance of the contacts established with the Guadalquivir River and the Baetican coastal area. This centre is said to have produced, among other materials, amphorae similar to the Haltern 70 ones and smaller variants of this form (ovoid type) (Alba Calzado e Méndez Grandes, 2002, p. 375-409).

important role Emerita has played in the production activities and in the redistribution of these products. Just like Françoise Mayet (1984, 197-198) has put forward, these examples are visible in terra sigillata stamps with the letters C.I.A.E.F. documented in the Museum of Merida, which, together with a set of identical stamps found in a tegullae and a lead tube there deposited and with an interesting inscription found in Cabezo de Pilas, seem to suggest we are dealing with letters that match the Lusitanian capital itself: C(olonia) I(ulia) A(ugusta) E(merita).

These are the data we present as being suggestive of a complementarity of both routes, the maritime, via Atlantic route, and the overland, via the Silver Route, essential for the economic growth of Emerita. If, on the one hand, it benefited from the proximity of the river harbour of Hispalis and of the harbour and anchorage of Cadiz to sell and redistribute its products, on the other hand, it maintained privileged relations with the communities of the north of the Peninsula through the ancestral overland Silver Route.

According to the author (Mayet, 1984), these three products used by Merida’s inhabitants possess a similar stamp to the town’s initials, which makes us assume, at least as far as the stamps in terra sigillata are concerned (from the production centre of Bezares), that the town of Merida received official orders from the artisans of Tritium (unless the town herself possessed a workshop in this large handicraft sector).

4

Part One THE NATURE OF THE ROMAN TRADE IN THE ATLANTIC

THE NATURE OF THE ROMAN TRADE IN THE ATLANTIC

1.1 AN ARCHAEOLOGICAL PERSPECTIVE1 C. Carreras The word ‘trade’ identifies a part of the Roman economy that has lately been a focus of attention amongst scholars, both historians and archaeologists. This particular concern is due to the recognized key role of commerce in the economic structure of the Roman world, where commodities moved from one side to the other of the Empire in remarkable quantities. There is no doubt that the Roman world witnessed a high development of exchanges, but it is still hard to assess the real volume and scale (Finley, 1973; 1985; Hopkins, 1980; Giardina, 1986; Duncan-Jones, 1990; 1994; Jongman, 1991; Whitakker, 1993). Archaeological evidence documents clearly the movement of goods between provinces and inside them; however such remains do not manifest the real significance of trade in the Roman economy. 1

The present paper attempts to explain such exchange mechanisms and use archaeological evidence as a way to test some possible hypotheses. ARCHAEOLOGICAL DISTRIBUTIONS: SOME CLUES Lately, archaeological distributions have been analysed in a more systematic way so as to distinguish patterns that could give insights into the structure of exchange. Normally the study of such distribution requires classification and quantification of an archaeological object from a good number of sites. After this, such quantities should be related to the distance to its production centre or original source (Hodder and Orton, 1976, 100-126). Application of geographical theory is a way to interpret such distributions.

Testimonies of long-distance exchange are, at least, noteworthy and provide some grounds to wonder whether Rome was the first large-scale economy, in which provinces took part in an integrated market (DuncanJones, 1990, 30-47) that could be defined as a world system (Woolf, 1990). However, it is extremely complicated to answer such a question with only the support of few economic texts, mainly papyri, and partial archaeological data. This kind of evidence cannot be compared to any later historical period.

One of the key issues in this approach is the way in which archaeological artefacts are quantified. Due to its fragmentary nature, archaeological evidence is difficult to relate to overall populations (Shennan, 1988, 288-289). In other words, a few fragments of a style of Samian ware cannot tell us the volume of trade in such merchandise over a span of time. Methods for quantifying archaeological objects are well known (Orton, Tyers and Vince, 1993; Arcelin and Tuffreau-Libre, 1998; Carreras, 2006b). However a problem arises when one wishes to compare data collected by different archaeologists, since the quantification method selected may affect results.

Despite this gloomy overview, there have been some breakthroughs in the study of the nature of Roman trade. Those advances are in relation to mechanisms of exchange responsible for such movement of commodities. Bearing in mind the constraints in flow of information, including transport costs, in Roman times (Mitchell, 1993, 132-134; Carreras, 1994; Lawton, 2004), exchange mechanisms were vital for making contacts between traders and potential consumers.

Many archaeologists prefer to count by weight as it is a non-variable measure, while others recommend the method of MNI (Minimal Number of Individuals) and the number of sherds, which can change according to the expertise of the archaeologist and the time invested (Carreras, 2006b). Nowadays, we have equivalence tables to covert number of sherds to weight using a measure of average sherd weight for different ceramic typologies.

1 This paper was published in 1999 in the Münstersche Beiträge zur antiken Handelsgeschichte, Bd. XVIII, H.2.S.87, pp.87-114. The present article is an updated version of the paper published in 1999.

7

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

province are absent in the other, and vice versa. Geographical determinism looks even more acute when ceramic types and commodities are produced locally, and then imports of similar products are seemingly absent, such as the case of North-African olive-oil in Sabratha (Dore and Keay, 1989, 36-45). As mentioned before, this regional distribution pattern corresponds to the principle of least cost and effort in goods transport in modern economies (Dicken and Lloyd, 1990, 16-17).

Therefore quantities could be transformed to different measurement units. Despite progress towards a common practice in quantification methods, comparisons are normally limited to percentages. Using percentages is prone to error since they establish a relationship between artefacts of similar characteristics (i.e. pottery) that is not so obvious. For instance every amphora typology may contain a different product, so only the ones with the same merchandise should be compared by percentages. Local productions can also normally alter percentages, making difficult any kind of interpretation.

Were Roman traders aware of such principles of minimum cost and maximum profit in their trade ventures? As far as contemporaneous writers are concerned, Roman merchants were perfectly aware of business mechanisms (Aubert, 1994). Quoting the words of Salonius, bishop of Vienne (AD V):

A solution to such hindrances may come from standardization according to area excavated and the site’s occupation length (Carreras, 2000, 45-57). Following this approach, archaeological distribution patterns can be easily visualized and interpreted. Such complexity in generating raw archaeological data for economic interpretation is a weakness that has been overstressed by Classical historians who disregard this evidence as misleading (Finley, 1985, 33; Jongman, 1991, 19).

“A merchant is called institor, because he always tools in order to acquire merchandise and to increase his profit. His ship loaded with various commodities available in greater quantities in his homeland, the institor sails across the sea to sell his cargo in foreign places before returning home with goods of a higher value”

In this context of raw archaeological data for economic interpretation should be included amphora studies. Amphorae from distant regions distributed in variable quantities all over the Roman Empire demonstrate the nature of Roman trade. Amphora research has shown that not only luxurious goods were traded all over the Roman Empire but also agricultural commodities borne in such containers. The way in which amphorae are quantified and later interpreted may shed some light on the essence of Roman commerce.

Second, studies of amphorae assemblages also reveal changes over the time in the main suppliers of particular regions and markets. For instance proportions of imports from Lyon and Vienne changed in less than a generation (30/20 BC to AD 15/20) (Desbat and Lemaitre, 2000), and shifts were also more remarkable in Rome over two centuries (Rizzo, 2003). Step by step, producers took control of a commodity market (e.g. wine from Gallia, African olive-oil) generally taking advantage of lower costs in production and distribution.2 Spatial and temporal variability of amphora imports at different Roman settlements can be observed in the following tables summarising their percentages in some key markets.3 Table 1.1.1 reveals that Baetican Dressel 20 was the common oliveoil container in Central and Northern Europe in the Augustan-JulioClaudian periods; whereas Berenice preferred the local olive-oil (i.e. Tripolitana). Moreover, Gaulish wine (Gauloise 4) in this early stage was mainly consumed in Gallia, without accessing other markets, as revealed by the percentages from Vienne and Lyon.

At first sight, quantifications of amphorae types seem to demonstrate control of local and regional markets by a single type. For instance Baetican olive-oil amphorae were common in all the provinces of the western Roman Empire during the Principate, while Istrian olive-oil amphorae dominated the eastern provinces. Likewise, Gauloise amphorae controlled the Gaulish markets from AD 25 onwards, as well as markets in Germania and Britannia, while they had little share in markets such as Hispania. Such distribution patterns appear to be related in modern geography with transport and production costs within a standard market system.

The second century AD (table 1.1.2) provides a different picture, since Gaulish wine (Gauloise 4) controlled the main markets, except the ones from the African provinces (Berenice and Sabratha). Furthermore, Baetican olive-oil was still being imported in good quantities, although it could hardly access African markets, where local oils were more popular. For instance, Tripolitana amphorae were the most common olive-oil containers at Sabratha

However it is hard to argue from this scarce evidence that a similar system operated in Roman times (Hopkins, 1980; Finley, 1985; Giardina, 1986; Duncan-Jones, 1990; Jongman, 1991), although it is likely that it was in place. In addition, distribution of other pottery types such as Oxfordshire Ware or Black-Burnished 2 seems to reveal identical control of areas close to production centres (Hodder and Orton, 1976, 100-126).

2

Public contracts may have also affected the distribution costs of some products, as may have happened with the annona and military supply (copia) (Remesal, 1986; Carreras, 2000). 3 The use of percentages is prone to error, but it is still the only data available for making such comparisons at the moment. The location of all markets mentioned in the table appear in Figure 1 (1: Colchester; 2: Rödgen; 3: Lyon; 4: Vienne; 5: Arles; 6: Marseille; 7: Turris Libisonis; 8: Ostia; 9: Rome; 10: Sabratha; 11: Berenice; 12: Barcelona).

Detailed analyses of ceramic assemblages from Cyrenaica and the Tripolitanian provinces based on percentages have revealed that both regions imported pottery from different sources, and in fact there were two commercial circuits. Some typologies present in one

8

THE NATURE OF THE ROMAN TRADE IN THE ATLANTIC

Table 1.1.1 Amphorae percentages in early contexts (Augustan or Julio-Claudian period) Sites

Dressel 20 (Baetica)

Gauloise 4 (Gallia)

Tripolitana

Africa

Lyon

5%

4%

–

–

Ostia

2%

3%

2%

–

Colchester

38%

1%

–

–

Rödgen

38%

–

–

–

Vienne

10%

7%

–

–

Rome

12%

0.3%

0,3%

3,2%

Barcelona

13%

0.2%

–

14%

Berenice

1%

–

10%

5%

Dressel 20 (Baetica)

Gauloise 4 (Gallia)

Tripolitana

Africa

Table 1.1.2 Amphorae percentages of IInd century AD contexts Sites Turris Libisonis

9.3%

21.3%

5.2%

10,5%

Sabratha

1%

–

80%

9,5%

Rome

12%

13%

0.3%

7,2%

Ostia

–

14%

–

15%

Lyon

15%

31%

–

–

–

1%

5%

1%

Berenice

Table 1.1.3 Amphorae percentages of Vth century contexts Sites

Dressel 23 (Baetica)

Late Roman (Oriens)

Tripolitana

Africa

Arles

5%

39%

3%

25%

Marseille

3%

61%

2%

30%

Turris Libisonis

–

6.4%

–

59%

Sabratha

–

–

66%

31%

Berenice

–

45%

–

–

Barcelona

0,1%

15%

–

44%

Rome

2%

20%

5%

32%

Lyon

2%

–

–

15%

(80%), while Africana types covered the regional market of Italy (Ostia, Rome), Sardinia (TurrisLibisonis) and, of course, the home African provinces (Sabratha, Berenice).

was still important in the markets close to its production areas, such as Sabratha. These percentages provide only a general picture of amphorae distribution. At least they shed some light into amphora supply according to market location and context date. Again the temporal evolution of amphorae quantities may be explained by economic factors such as production and transportation costs. The latter are implied with the appearance of new production areas closer to the potential markets (e.g. Gaulish wines in the Limes versus Italian and Tarraconense wine, or African olive-oil versus Istrian and Baetican oils).

Finally, the origin of suppliers and amphorae typologies changed again in the Late Empire as can be observed in table 1.1.3. Eastern Mediterranean wines carried by a series of Late Roman amphora typologies are recorded in high percentages in western provinces (Arles, Marseille) and eastern markets (Berenice).4 With regards to oliveoil, Baetican imports in Dressel 23 amphorae diminished, while Africana typologies wielded control over the markets of the western provinces. Only Tripolitanian oil

Although archaeological distributions hint at possible reasons for the nature of Roman trade, their data is not enough to confirm any hypothesis. Therefore it is

4

It seems that amphorae were not the only containers for Mediterranean wines in this period: barrels and skins were potential alternatives for trade in local wines (i.e. Tarraconense wines).

9

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

on the information provided by Diocletian’s Price Edict (AD 301),5 which gives the following ratios for transport in Roman times:

necessary to try and recreate the economic conditions of Roman times and then evaluate these results so as to understand their logic. That is why simulation models are required as prediction tools (Jongman, 1991, 19; Carreras, 1994).

1 sea shipping: 3.4 river boats (downstream): 6.8 river boats (upstream): 43.4 pack animals: 50.72 wagons The simulation model in ARC/INFO calculates the shortest path, or the least-cost route from a chosen point to the rest of the potential destinations.6 A final clear composition can be seen on a contour map created by interpolation minimum costs from a single point to all destinations in a region (see figures 1.1.1, 1.1.2 and 1.1.41.1.11).

SIMULATING TRANSPORT COSTS: A PREDICTION TOOL FOR ARCHAEOLOGY Observations of archaeological distributions, not only amphorae but any kind of pottery (i.e. Samian ware), suggest an economic logic behind the movement of goods. Such patterns can be seen even clearer in longdistance exchange. It has been also pointed out that the transport principle was responsible for the quantities recorded at any particular market. Therefore a simulation model of transportation costs in Roman times would be a method to evaluate if they really determined distributions. Simulating a transport network and its possible outcomes is the object of research of many geographical and logistic studies (e.g. the shortest path, the best allocation), also applicable to archaeology. Technically computer applications of Geographic Information Systems (GIS) will enable us to obtain the least-cost itinerary between two points or between one point and a region. Both functions are based on the principle of maximization of benefits and cost reduction in any commercial venture (Dicken and Lloyd, 1990, 16-17).

Some initial experiments were carried out in Roman Britain with two well-known local productions (i.e. Black-Burnished 2 and Oxfordshire ware), with distinctive patterns of distribution also studied by other means (Hodder and Orton, 1976). Simulation maps generated by optimized transport costs calculated from the production centres appeared to identify the two patterns of distribution. In addition, such maps reinforced the idea that a kind of market system existed in Roman times. Nevertheless, simulation maps failed to recognize the distribution pattern of another archaeological artefact, Dressel 20 amphorae (Carreras, 1994). Such exceptions to the optimized transport cost model challenged assumptions of a market system in place. The question was now whether the Dressel 20 distribution was due to another exchange mechanism in place in Roman times.

Accepting the existence of a market system in Roman times it is possible to generate a simulation model of transport costs for any chronological period. With a previous knowledge of Roman transport infrastructures (roads, tracks, rivers, maritime routes and ports) and the costs of every means of transport (wagon, ship, river boat, pack animals), simulations can be generated and run (Carreras, 1994; 2000). As a result the model creates a series of least-cost markers calculated from any chosen point or production area (i.e. amphora workshop, mine, etc.).

Further experiments with other amphorae distributions in Britain, such as Gauloise 4 or Campanian Dressel 2-4 (Carreras, 2000), managed to demonstrate that the optimized transport cost model still worked quite well and perfectly identified those patterns of distribution. Most of these distribution maps show the same traits, with high densities in the southern and coastal areas of Britain and low densities inland, due to the high cost of land transport. Of course, there were some minor differences which probably reflect other variables such as population density, market hierarchy and ethnic groups, but they do not affect the general view. However, Dressel 20 amphora distribution seemed to behave in a different way.

The model requires digitalization of every transport infrastructure in a defined territory (i.e. Roman Britain, the Roman Empire as a whole) normally in a CAD/CAM program. Later, the resulting files can be imported into a GIS program with network analysis facilities. In our first application in 1991 the model was implemented in SPANS, a GIS from IBM, with limited facilities (Carreras, 1994). A second version in 1994 employed ARC/INFO and its module Network analysis, which allowed us a more accurate calculation of transport costs (Carreras, 2000, 63-70).

As previously noted, Dressel 20 amphora distribution in Roman Britain appears to be an exception to the rule, since high amphora densities are recorded in places with higher transport costs and more distant from the English Channel. Baetician olive-oil contained in the Dressel 20 amphorae is said to have been part of a redistributive

The model works in a very straightforward way. Once the transport infrastructure is imported into the GIS at the right scale, every segment in the network (km) can be multiplied by the suitable coefficient according to the means of transport (e.g. costs, speed). In other words, multiplying this cost coefficient by a route length in km would give us the actual cost of a particular journey in Roman times. Cost coefficients were calculated basically

5

Similar ratios can be worked out from papyri collected by Drexhage (1991, 327-351), although they come from different periods and regions in Egypt. 6 Simulations were generated from an initial point located in either the production centres (e.g. Black-Burnished 2 from Colchester and Oxfordshire Ware from Oxford) or a midpoint in the transportation route (e.g. Continental amphorae from the middle of the English Channel), as happens with Dressel 20 amphorae.

10

THE NATURE OF THE ROMAN TRADE IN THE ATLANTIC

Figure 1.1.1 Gauloise 4 amphorae distribution (densities)

in the SW (i.e. Exeter, Gloucester) and other urban settlements in central Britain (i.e. Leicester, Chester, York, Carlisle). Such typical trends can be singled out in the distribution of amphora typologies such as Campanian Dressel 2-4, Baetican Haltern 70, Gauloise 4 and the Spanish fish-sauce vessels (Carreras, 2000), which were the most common amphorae in Britain apart from Dressel 20.

system tagged with the name of annona. It is believed that the Roman State was responsible for distributing olive-oil to the plebs of Rome (van Berchem, 1937; Pavis d’Escurac, 1976; Sirks, 1991) and perhaps also the Roman army (Remesal, 1986). The predominant military distribution of Dressel 20 amphorae in Roman Britain appears to support this second theory, since high densities are recorded in the north, covering a broad area from the Antonine Wall to an imaginary line linking Chester with Lincoln (Carreras and Funari, 1998).

All these distribution maps look closer to the one generated by the simulation of transport costs in a market system (see figure 1.1.1) sustaining the idea that cost factors were responsible for amphorae movement in the province. In other words, the lower the transport costs, the higher the density of amphorae recorded. The principle of maximization lies here, linked to the existence of a market system. If such statements bear logic this means that the simulation model of transport optimization may predict densities of amphorae at each

Distribution maps of different amphorae of Roman Britain were generated by interpolating quantified data (densities – cg/m2) from 104 sites, in which all the areas from the province were represented. Most amphora distributions reveal high amphora densities in the southern areas, mainly in the SE sector (i.e. London, Colchester, Verulamium), where there was also a high density of population. Moreover, there is a good sample 11

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 1.1.2 Dressel 20 amphora distribution (densities) 80% of amphorae in any amphora assemblage in the province. Apart from high densities in Exeter and London, most Dressel 20 amphorae were recorded in the military area of northern Britain (Carreras and Funari, 1998). Therefore the simulation model was adapted to recreate a redistributive model in which the costs of maritime routes towards military ports were half the price of normal maritime transport fees.7 Despite the high cost of accessing inland military sites, the resulting simulation map of a redistributive model (see figure 1.1.2) appears to more or less identify Dressel 20 amphora distribution.

location, and therefore the model could foresee a potential density that could be later compared to theactual results from excavations. The model does not only predict distributions but also shows the logic that lies behind them. It must be borne in mind that the cost of transport was not the only factor affecting movement of goods within in a market exchange system. Other causes such as purchase power, ethnicity or population density may also alter normal distribution. Such factors are well known in modern economies and cover the concept call demand variability (Dicken and Lloyd, 1990, 181-184).

As a conclusion, simulation models of transport costs have managed to explain distributions of single archaeological artefacts, such as pottery or amphorae,

Although most amphorae distributions were identified by the model of transport optimization, Dressel 20 distribution was an exception. Dressel 20 vessels were quite important in Britain, accounting for between 60%-

7

The Spanish fleet to America provides an example of high costs to the free market with traders charged 40% in taxes, whereas public cargoes did not have to pay anything (Lorenzo-Sanz, 1979, 462).

12

THE NATURE OF THE ROMAN TRADE IN THE ATLANTIC

Figure 1.1.3 Simulation model of transport costs (ARC/INFO)

Lloyd, 1990, 121-141). In this case the location of each production centre with regards to consumption areas is the most relevant feature that affects final venture success.9 Such principles were already applied to the study of archaeological distribution in what has come to be known as spatial archaeology (Hodder and Orton, 1976, 187-197), but the complexity of some calculations deterred possible inferences.

according to transport costs within particular exchange mechanisms (i.e. market or redistribution). Nevertheless, simulation models applied to a single province such as Britain do not solve complex phenomena such as competition between amphorae containing similar commodities or variations over time. Temporal variations observed in tables 1 to 3 show the production and distribution costs of each amphora typology, going on to compare results between potential competitors. Perhaps such cost comparisons may explain the presence or absence of particular amphora types in each settlement.

It is believed that such hindrances can be overcome by comparing results and maps obtained from the simulation models of transport costs. However a major change should now be introduced: the scale of the model. Instead of applying it to an individual province, the transport model should be generated from the Roman Empire as a whole. Therefore such a complete model would include the transport infrastructure of the Roman Empire as a whole, but with less detail. Figure 1.1.3 illustrates a possible infrastructure layout based on navigable rivers and Roman roads, to which the main maritime routes have been added.

Since calculations of production costs are still impossible due to the lack of written testimonies,8 comparisons can only be drawn from distribution costs. Therefore the application of simulation models of transport optimization from production areas to final consumption centres may enhance our current view about amphorae trade. COMPETITION AMONGST SUPPLIERS: TRAVEL EXPENDITURE If two regions produce the same commodity with a similar quality and analogous production costs, the success of their respective productions in distant markets will rather depend on transportation costs (Dicken and

The simulation model was then generated, taking some production regions as starting points – such as Baetica, Tarraconense, Aquitania, Narbonense, Lusitania, Africa, Italia and Greece. The results obtained are commented on below in relation to each commodity produced in those regions carried by amphorae.

8 Nowadays we are aware of the importance of production costs in our global world since transport costs no longer affect to such a degree the consumption of distance goods (i.e. Chinese merchandise).

9 Gravity models and substitution analyses are applied to understand such relationships (Isard, 1956; Haggett, 1965; Dicken and Lloyd, 1990).

13

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 1.1.4 Transport cost simulation from Roman Greece

for the transport cost simulation appearing in figure 1.1.5. The high cost coefficient of land transport justifies the extraordinary rates illustrated by the resultant contour map, which could have produced lower rates if a coastal city had been selected instead. Nevertheless, the choice of Capua still underlines that high quality of Campanian wines, which are documented by amphorae in distant markets such as Britain, Germania, Raetia and India, despite transport costs.

Wine Wine is probably one of the most difficult products to study since there was a wide variety of producers and qualities of wine, according to age, production origin, etc. (Tchernia, 1986). Greek and Italian wines enjoyed a deserved fame amongst Romans, who considered them luxurious produces. Greek wines have an easy access to all Eastern Mediterranean markets (e.g. Cyrenaica, Egypt), as is revealed in the transport cost simulation map (see figure 1.1.4) that shows lower values than 210 kg/wheat per ton. The simulation map confirms the pattern observed by Fulford (1989) at Berenice, as well as the proportions of amphorae recorded at Athens (Robinson, 1959). Transport costs suggest that Greek wine had a complete monopoly of the Eastern Mediterranean region.

If distributions of high quality beverages are hard to compare, those for table wines are, on the contrary, quite straight forward. Italian wines carried in Dressel 1 or Dressel 2-4 vessels dominated western markets until the early Augustan period, when Tarraconense wines carried in Pascual 1containers took over (Revilla, 1993). This is recorded at many sites in Gaul (Aquitania, Belgica), Raetia and, where Tarraconense amphorae (mainly Pascual 1, Oberaden 74 and Dressel 2-4) arrived in good numbers, more or less replacing Italian wine. According to the simulation of transport costs (see figure 1.1.6), Tarraconense wines travelled at lower costs than Italian ones. The coastal region of Tarraconense, therefore, took advantage of its better location to access distant markets, such as Gaul or Germania at that time.

Greek amphorae are also recorded in good numbers in western places such as Lyon, Augst (Martin-Kilcher, 1994) and Britain (Carreras, 2000), where transport costs reached higher values. The only way that such products could have covered these higher costs was by having luxury prices to the end consumer. That is why Greek wines imported in western markets were considered as expensive goods, as the ancient sources also reveal (Tchernia, 1986, 100-107).

However, the second half of the Ist century AD witnessed a further change in the control of western markets, when Gaulish wines outstripped their competitors, either Tarraconense or Campanian. From the Julio-Claudian period onwards, Gauloise amphorae took control of the main wine markets in Gallia, Raetia, Germania and

Other praised wines came from Campania, not only from the coastal strip (i.e. Falernum, Surrentinum), but also from inland (i.e. Caulium) (Tchernia, 1986). Capua, an inland community, was chosen here as a point of origin 14

THE NATURE OF THE ROMAN TRADE IN THE ATLANTIC

Figure 1.1.5 Transport cost simulation from Roman Italy (Campania)

Figure 1.1.6 Transport cost simulation from Tarraconensis Britannia (Carreras, 2000). Again, the better location of Gaulish production centres appears to justify such an overwhelming presence (Laubenheimer, 1985). The simulation of transport costs from Narbonense (see figure

1.1.7) shows the better position of Gaulish exporters, closer to the Rhone valley and the German limes. It is even more interesting to observe the resulting map from Aquitania (see figure 1.1.8) that points to an over15

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 1.1.7 Transport cost simulation from Narbonense

Figure 1.1.8 Transport cost simulation from Aquitania whelming control over Britain and the Atlantic provinces: this is confirmed by most of the amphora assemblages of Britain (Carreras, 2000). The highly favourable location of Gaulish vineyards contributed to the predominance of Gaulish wines in the western markets.

Transport costs, therefore, seem to explain the evolution of the wine market in Roman times, following a market system framework implemented in the simulation model. The model also demonstrates the existence of a kind of market system in other products, 16

THE NATURE OF THE ROMAN TRADE IN THE ATLANTIC

Figure 1.1.9 Transport cost simulation from Baetica to debate (Remesal, 1986; Whittaker, 1989; 1993), but such a system was in place.

such as Samian ware (Marsh, 1981). Despite the lack of sources of information and transport constraints, the Roman economy manifested clear signs of the existence of a market exchange mechanism for long-distance trade.

Other olive-oil suppliers were African provinces and Italy (i.e. Istria, Brindisi). Some oils, such as Istrian and Tripolitanian, were favoured in the eastern and central Mediterranean, each with its own distribution area. Oils from African Byzacena and Proconsularis, however, seem to have been distributed in the western and central provinces. Transport costs from the coastal town of Africa Proconsularis was similar to those associated with Baetican olive-oil. So the predominance of Baetican Dressel 20 amphorae cannot be explained only in terms of transport costs. Again, a comparison of African and Baetican olive-oil distribution during the Principate reveals the existence of a redistributive exchange system that could have altered the possible balance between equal competitors (Mattingly, 1998).

Olive-oil The case of olive-oil is rather different, as is suggested by the distribution of Dressel 20 amphorae in Britain (Carreras and Funari, 1998). Furthermore, this typology is basically found in great numbers in the city of Rome and the western borders in the military camps. In theory, the Dressel 20 distribution pattern fits better in a redistributive system than in a market one, because travelling to some of these final destinations involved further costs. The simulation model run from Baetica (see figure 1.1.9), the origin of these olive-oil vessels, shows high transport rates for the German, Raetian and British limes. Nevertheless these provinces record high concentrations of Dressel 20 amphorae, reaching 60%-80% in some assemblages. Such a distinctive pattern can only be explained by a redistributive system with direct State intervention (Carreras, 2000). In other words, only a public administration could afford to mobilise goods that incurred such travel costs that not even private traders could bear. As mentioned above, a market system is based on maximization principles – minimal expenses and maximum benefit and Baetican olive-oil distribution does not seem to indicate such principles. Whether the redistributive system responsible for Baetican olive-oil movement was called annona militaris or not is still open

Since African olive-oils became more common in Rome and western markets from the second century AD onwards (Keay, 1984), one may wonder whether the African products that sometimes arrived from Imperial estates could also have formed part of this redistributive system.10 Coastal African amphora workshops were in a competitive location to access Mediterranean markets, whereas inland centres could hardly emulate them. In this case, the simulation model was based on Thugga (Dougga), an inland town, and the results appear in figure 1.1.10. The overhead transport costs to access any 10 Excavations at Monte Testaccio (Rome) reveal an important presence of African amphorae from the middle of second century onwards, where Baetican Dressel 20 are still in the majority (Blázquez and Remesal, 1999; 2001; 2003).

17

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 1.1.10 Transport cost simulation from Africa

There are many possible explanations, such as the qualities of the different products, as is the case with wine (Curtis, 1991). However it is obvious that fish-sauce amphora distributions do not match simulated transport costs.

African outgoing port explain the high values on the map. Archaeologically, olive-oil presses and some amphorae workshops are recorded inland, where land transport costs were extremely high. However, such olive-oil production seems to have been aimed at exports, so a redistributive system should also have been in place.

An alternative explanation comes from the relationship between Baetican Dressel 20 and fish-sauce amphorae (Carreras, 2000) which sometimes shared the same ship cargoes. If a redistributive system were responsible for the movement of Dressel 20, then fish-sauce amphora may have taken advantage of such low-cost public supplies. Therefore the distribution of Baetican salazones may have benefited from a ‘parasitic’ exchange mechanism (Carreras, 2000, 167-168).

Further research is required to analyse in more detail the olive-oil trade in the Roman period, including other production regions such as Palestine, Gaul, Greece, Syria and Tarraconense, as well as the complexity of exchange mechanisms. What appears to be obvious from the Dressel 20 amphora distribution is the existence of nonmarket system mechanisms capable of moving large amounts of goods to distant destinations. It is believed that the origin of such distribution was the presence of a redistributive system, whose features do not match with the rules of our simulation model.

There are still many grey areas in the distribution patterns of many amphorae types, since quantifications are not available. Further research is required to assess whether particular distributions identify market exchange behaviour or other alternative mechanisms. Other archaeological artefacts, such as fine ware, may generate similar patterns and draw similar conclusions.

Fish-sauces (salazones) A third commodity, fish-sauces, is also widely represented by amphora distributions, although little attention has been paid to them so far. The main production regions of fish-sauces were North Africa, Lusitania and the Baetica coastal strip (Curtis, 1991). Transportation costs from these three regions were quite similar according to the model (see figures 1.1.9, 1.1.10 and 1.1.11), which should reflect an equal shares of the market. However the reality was different since Baetican fish-sauces dominated the western markets in the Principate, whereas Lusitanian and North-African products were more common in the Late Empire.

FINAL COMMENTS The present survey of amphora distributions has allowed us to record temporal and variability factors in the consumption of some products. It was argued that transportation cost was one of the main reasons for such variability whenever new production centres close to potential markets were identified. This principle, documented in Roman times, identifies maximization 18

THE NATURE OF THE ROMAN TRADE IN THE ATLANTIC

Figure 1.1.11 Transport cost simulation from Lusitania

of Baetican fish-sauces, constituting, in a way, a distinctive mechanism of exchange of its own (‘parasitic’ exchange).

behaviour of any market systems, which can be easily modelled. That is why a simulation model of transportation routes was created in GIS in order to test if amphora distributions fulfilled market system rules. The simulation not only identified the distribution patterns of many types of amphora but also proposed some expected quantities for non-excavated sites. Consequently the model is a guide for interpretation and future research.

The importance of Roman trade stemmed from all these exchange mechanisms in place and acting in complementary way. Transportation costs and the flow of information limited the capacity of the market system to match demand with supply, and Roman state involvement developed to respond to these hindrances by generating an alternative mechanism exchange system known as redistribution. More work is required to unveil the complexity of Roman trade, but little by little fresh evidence is shedding light on the more obscure areas and lines of research. The combination of textual information and archaeological results may enhance our current knowledge of this particular side of the Roman economy. Archaeological data may result in an even more complex image of the Roman economy, but attempting to reveal its nuances will be a rewarding task.

A few amphora distribution patterns do not fit in the cost model provided by the simulations and therefore hint at the existence of alternative exchange mechanisms in the Roman economy. For instance the distribution pattern of Dressel 20 amphorae suggests state intervention in their supply to the city of Rome and the armies, in what is known as a redistributive system. In the same way, North-African olive-oil appears to have circulated through public networks. Abuses of such public networks seem to be a more likely explanation for the distribution

19

THE NATURE OF THE ROMAN TRADE IN THE ATLANTIC

1.2 TRADE AND THE MEANS OF TRANSPORTATION R. Morais countries; consequently, it is an exceedingly wretched place to live in.”

THE TRADE “Una de las actividades económicas que ha supuesto un mayor contacto entre distintas culturas, provocado un mayor desarrollo técnico de las comunicaciones y el transporte y favorecido la estabilidad es sin duda el comercio.” (Carreras, 2000a, 17).1 We could not agree more. The Latin word Commercium meant any activity related to business, the exchange or the buying and selling of goods which accepted the risks that came with the transaction (Will, 1987). As Le Goff points out (Le Goff, 1982, 7) trade, besides causing some conflicts, also works as one of the best links between geographical areas, civilisations and peoples.

This situation is not completely true, though. In fact, despite the isolation of these settlements, which is the result of the lack of overland roads, we cannot forget that different archaeological remains were found that prove the existence of an intense trade in these regions, in particular using the river and the Atlantic routes. We know that the peoples of the Northwest practised two types of trade since the pre-Roman era: an internal trade, carried out in the area of the fortified camp (castreja) culture; and another one, a foreign one, that linked this area with other geographic regions.

No society can avoid consuming nor can it thus stop producing and distributing. In this way the whole economic activity, centred in the movement of production, distribution and consumption of goods is a dynamic and circular process. As far as the circulation of goods is concerned, distribution provides the essential connection. In this way, all lasting relations of a given society imply transactions, exchanges and, in a more complex degree, commerce (Aubet, 1997, 93). Nevertheless, when we do not have enough written documentation it is extremely hard to recreate the exchange mechanisms in the ancient world. Such is the case of the Peninsular Northwest (figure 1.2.1).

The internal trade took place in the fora, which possibly worked as tribal or gentile markets, where the exchange products would be perishable goods, objects and clothes, or products such as gold, copper, tin and lead. In these same fora some other products could be exchanged, such as exotic pottery and glass beads from the foreign markets. The foreign market with other peninsular areas, or extra peninsular ones, was determined by tin and certainly encouraged by the creation of some favourable conditions such as the provision of services, protection, freedom of circulation and exemption of payments. This trade was carried out by water and land roads. The latter were, however, in most of cases, hard to precise, except for an inland path from the south of the Peninsula up to the Northwest (the so-called Silver Way in Roman times). Contrary to this situation, the most significant examples of the commercial routes in the pre-Roman period are represented by the coastal fortified settlements, situated in important places and in the estuary of important rivers, such as the settlements of Coaña, Sta. Tecla, Sta. Luzia, Terroso, Bagunte and Vila Nova de Gaia (vid. Naveiro López, 1991b, 24). We should also emphasize the importance of another smaller settlement situated in a perched position by the mouth of the Cávado River. We

A superficial reading of the sources can induce us to admit an almost total isolation of the Northwest and North of the peninsula settlements. This isolation is especially enhanced at the beginning of Strabo (STR. 3.1.2) when he says that this region “in addition to its ruggedness, not only is extremely cold, but lies next to the ocean, and thus has acquired its characteristic of inhospitality and aversion to intercourse with other 1 Trade is unquestionably one of the economic activities that has fostered a wider contact between different cultures, caused the development of the communications and transports and favoured stability.

21

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 1.2.1 Commercial routes and communication roads in NW Peninsula (Beltran, 2008) southeast and the east. This orientation of the different Roman camps, as Blot has well pointed out (Blot, 2003, 34), curiously matches the orientation of the different settlements and emporiums of Phoenician times in our territory, according to a typically Mediterranean criterion.

are talking of the settlement of S. Lourenço, situated to the NE of Esposende, in a nearer position relatively to Bracara Augusta, which can be inserted in the context of the maritime-river trade of the Atlantic, as a result of the archaeological remains so far found – among which there is an Attic red-figure ceramic fragment from the IVth century BC (Arruda, 1997, 82; Almeida e Cunha, 1997, 14).

With the eventual annexation of these territories by the Romans, trade, together with agriculture and mining, becomes one of the major factors of economic development; trade was only conditioned by the borders rights of each province and consequent distribution, which strongly simplified its development. Roman merchants were frequently organized in voluntary associations called collegia. The data provided by the epigraphy, which consist largely of funerary and honorary epigraphs, show that these activities were socially and religiously important in the communities’ life. Membership of a collegium and participation in its rituals gave the merchants, whether freedmen or freeborn, a sense of identity and a certain status in the community (Meijer e von Nijf, 1992, 74).

The work done by Pinto (Pinto, 1997) on the fortified settlements situated in high places of the Peninsular Northwest is very elucidatory. The author presents the reasons that led to the human settling throughout the coastal strip and reports the human involvement with the water environment from the respective geographic position (maritime, coastal and estuary). As a complement to this study, a recent one carried out by Struut (Struut, 2000) proves, from the spatial analysis of the distribution of proto-historic and Roman settlements in the Ave’s region, that the human behaviours regarding the choice of habitat were not only different in these two periods but they also obeyed to different options. The fortified settlements of “castrejo” type found in high places were frequently situated overlooking the main river courses, whereas the Roman occupation privileged the low slopes and the alluvium plains, corresponding to the river terraces and the coastal plains, which suggest an immediate connection with the water resources and the choice of lands fit for farming (2000, p. 126). Another interesting aspect emphasized by the author (2000, 127; 132) has to do with the type of orientation adopted by the Roman camps which – unlike the ones of castrejo type, whose slopes are orientated in all directions – are preferably placed in the slopes exposed to the south, the

Much less frequent are the epigraphic references regarding the economic activities of the collegia, while association. We know, however, that these worked as true commercial businesses that aimed to defend the respective hegemony in the economic activities they dedicated themselves to (Meijer e von Nijf, 1992, 74). Nevertheless, we should not forget that in a society based on a subsistence economy, as all the societies in the ancient world were, the most important mercantile branch was not, naturally, the commerce of luxury commodities, 22

THE NATURE OF THE ROMAN TRADE IN THE ATLANTIC

but rather the exchange of basic needs, vital for the population’s supply of the different urban nuclei.

As a matter of fact the use of the ox, as main draught animal in the ancient times, of the mule or donkey, was determined not only by the slowness but also by their natural voracity. We should also point out that the horseshoe was still unknown, which made it impossible to prevent damage to legs (Chic, 1997a, 46). The cost would thus be higher, especially in long distance journeys according to the amount of goods to transport and the number of animals and carriers used. Actually, the weakness of the traction of the old yoke determined that in the IVth century AD carts could not transport more than 429 kilograms (Foraboschi, 1990, 820).

Thus, the Roman State provided the provinces with a main web of roads, linking the hub centres that encouraged the consumption of basic needs and the expansion of the urban life. To get a deeper and contextualised knowledge of the economy and trade of a city in the Roman world we should also take in consideration the costs of transportation in the ancient times and its relation with the means of transportation used in a certain geographical context.

On the contrary, transport by water – maritime or fluvial – was not only faster, safer and more efficient, but also less expensive.

THE MEANS OF TRANSPORTATION A first study of the transportation cost of goods and its relation to the means of transport used was written by Ducan-Jones (1974, 366), who established the following values: 1 sea; 4.9 river and 34 to 42 roads.

Forbes (1996, 160) calculated that the prices of land transportation made the cereal price double every 100 miles (142.2 km) and the same probably happened with those goods whose value was small in relation to their volume or weight.

Afterwards, Deman (1987, 81) calculated from the established prices for cereals found in Diocletian’s Edictum de pretiis rerum venalium (7, 1) that the cost proportion between the several means of transportation would be the following: sea 1; river 5 to 8; roads 39.

Finley (1986, 175-81), taking into account these average costs of transportation in the ancient world, pointed out the existing relation between the transportation capacity and the development of the urban life, stressing the fact that the traction systems were expensive and not very efficient, taking into consideration the heavy wooden carts pulled by draught and pack animals.

More recently, in a more detailed study, Carreras (1994, 32) calculates the following costs taking into account the different means of transport: 1 sea shipping; 3.4 riverboats (downstream); 6.8 river boats (upstream); 43.3 pack animals and 50.72 wagons. Gianfrotta’s and Pomey’s study (1981) on the advantage of transportation by water is very elucidating: 375 wagons with a 400 kg capacity would be necessary to carry out the transport of a cargo equivalent to that of a vessel with an average capacity of about 3000 amphorae, i.e. 150 tons. According to these authors’ calculations (Gianfrota and Pomey, 1981, 51), from the already referred to price list established at the time of Diocletian, it would be less expensive to send a load of cereals by waterway to the remotest parts of the Empire than to transport it in wagons by draught animals over a distance of 100 km.

Figure 1.2.2 Carruca from Bosceaz (Switzerland)

Figure 1.2.3 Carruca with skin from a Pompei painting 23

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

a magpie in the rigging was bad luck, as were glimpses of wreckage on the shore, the uttering of certain words or expressions. Some dreams were also considered bad omens: for example to dream of turbid waters, or a key, or an anchor, or certain animals (especially goats, wild boars, bulls, owls and other night birds, seagulls and other sea birds) (Casson, 1994).

Despite the obvious advantage of transport by waterway, we should bear in mind that even that kind of transport, especially the maritime, had many risks. This was due to the fact that there were not many navigational instruments, which made pilots use their instincts and stars (guiding themselves by the sun during the day and by the stars during the night). For that reason, most of the sea travel was done using the cabotage system, not only because of the lack of nautical equipment, with primitive directional and propulsion systems, but also due to the meteorological phenomena characteristics of each climatic season, which made maritime traffic difficult or even impossible and often leading to the loss of human lives and cargo.

There was still another factor which anyone choosing between travelling by land or sea had to keep in mind: ship passage was not available all year round. If we take into consideration the fact that for a third of the year ships were tied up in harbour, and that during the other third navigation was considered dangerous, we may conclude that sailing only took place during the four summer months (Pounds, 1987, 41). Such restrictions to navigation were due to severe winter storms and the increase in clouds between autumn and spring. The Roman administration established a period of mare clausum (November 12th to March 10th) and another of ‘open sea’ (May 27th to September 14th), as well as two other periods of caution (March 10th to May 27th and September 14th to November 12th) (Chevalier, 1988a, p. 119).

Ancient authors referred to the mere finger’s width of plank separating a sailor from drowning, and the farewell poems addressed to friends departing overseas sometimes read like elegies on their certain death (Casson, 1994, 150). When a vessel was in difficulties near the coast and able to anchor it should throw overboard its cargo rather than sink. This situation happened so often that there were even specialized salvage divers, such as the urinatores, asocumented in an inscription found in Ostia, dedicated to the Emperor Antonine Pius (Beuderley, 1991, 14-15).

This fact did not stop, however, that in some occasions navigation took place during the less advisable months: for example, Claudius tried, inter alia, to encourage the merchants specialized in cereal transport to import them to Rome during the winter months according to a specific regulation (SUET. Cl. 18.4.24). That regulation established that in case of shipwreck or abandonment of part of the cargo because of the winter storms these merchants were to be paid compensation worth the value they had lost, even granting them special privileges (SUET. Cl. 18.4.19) (Cfr. Sirks, 1991, 41; 61-62). We also know of journeys carried out exceptionally in less advisable periods by transport troops that were obliged to deal with an emergency (Casson, 1994, 149). As a result of these limitations the period reserved for navigation was not very long and, as we have already mentioned, navigation was plotted by the stars, in particular the Polar star. This star, which the Greeks called pheniké, marked not only the north but also the altitude of the observer (Tinoco, 2001, 198). The navigators also sent a watchman to the mast or followed bird migration; at times, they would also set loose birds – doves and crows – to observe how distant they were from the coast (Martínez Maganto, 1990, 70).

Figure 1.2.4 Inscription of urinatores (Ostia)

We also know of the existence of night lighting in the vessels, by means of perforated amphorae that fixed to the bow. Inside these amphorae a lamp was kept burning that kept night contact between the fleets (Tinoco, 2001, 194; 199). The archaeological proof of this theory is documented by the discovery of a Dressel 1 A amphora found in a shipwreck dated from the IInd century BC. Inside were traces and characteristics that helped identify it as a vessel lamp (Martínez Maganto, 1990, 69-70). Lighthouses were also essential features. They were used, as they are still today, to mark harbours and safe

On many days of the year the religious calendar forbade business of any sort, and this included the departure of ships. August 24th, October 5th and November 8th, and even the last days of the month, were considered illomened for navigation (Casson, 1994, 155). On the favourable days for sailing, before embarkation, animal sacrifices would be made, often a sheep or a bull. If the wind was favourable, if there was nothing wrong with the date and if the sacrifice had gone off as desired, superstition still left a gamut of bad omens to be run: a sneeze as you went up the gangplank was bad, a crow or 24

THE NATURE OF THE ROMAN TRADE IN THE ATLANTIC

Figure 1.2.5 Lighthouses

(Martínez Maganto and Carreras, 1993, 102). This lighthouse, which Ptolomeus situates in the great port of the Gallaecians and the people of Lugo, and that he calls Farum Precantium (it should be read as Pharim Brigantium; Cfr. Franco, 2001, 227, endnote 25), was certainly related to a statio. This was in turn connected to the important economic (commercial) role of that port centre, which was responsible for the gathering of products, such as the portorium, by Roman servants in charge of that task (exactores), as documented in the inscriptions from Coruña (Arias, 1992b, 55). However, we should not overvalue the importance of this equipment in the context of Brigantium. As Pérez Losada (2002, 135) says: Brigantium é polo tanto un punto de especial interesse estratéxico e administrativo a nivel estatal (ou imperial se se quere), concretado na prestación de servicios a este alto nivel que confiren ó enclave un character próprio e extremamente singular dentro do mundo galaicoromano (Brigantium is a place of special strategic and administrative interest at a state level – or even at an imperial one – realized in the provision of work at this high level that gives the enclave a very particular character inside the Gallaecian Roman world). In fact, as the author also says (id. ibidem), the archaeological remains are not very abundant when compared to other coastal contexts (for example Vigo or Iria), what makes us think of a “trafico «de paso» de mercadorías (abastecendo da Galia e Britania desde o Mediterráneo) que propriamente un centro relevante de consume, producción ou almacenamento / redistribución masiva de bens importados cara ás terras do interior” (“passing through” traffic of commodities – supplying Gallia and Britain from the Mediterranean – rather than a centre of consumption, production or storage / redistribution of imported goods for the inland regions.)

anchorages. The used a signal system, which dates from at least the Greek era, based on established methods of coastal fires and military watching towers (Martínez Maganto, 1990, 67-69; 85). It is also probable that lighthouses were also important during in the day; they used signals using the sun’s rays on polished surfaces, which enabled an intermittent signal that could be seen from a great distance. Columns of smoke were also employed, as wells as mechanical sound signals during thick fog or storms (Martínez Maganto, 1990, 75-76; 85). In the context of the Atlantic strip of present-day Portugual, we only know of one tower (used as a lighthouse) from Strabo (STR. 3.3.1). This tower might have existed in a place not yet determined, probably at Espichel Cape or in Outão (Alarcão, 2004, 317-325), and, like many others that certainly existed along the Atlantic strip, it was of the utmost importance for the viability of navigation along the coast. It is possible that there were other lighthouses or towers (Alarcão, 2004, 317-25). The best known lighthouse along this coast is situated in Coruña. Known since ancient times as the ‘Artabrian port’ (STR. 3.154), or altissimum pharum (OROS. Hist. 1.2.71), this lighthouse served as a lookout for vessels that went or came from Britannia. Also known as the Tower of Hercules, it keeps the architectural tradition of the famous ancient lighthouse built in Alexandria, in Egypt (Hauschild, 1977, 136). Brigatium, which was situated in the vicinity of Coruña, was the last of Hispania’s great port centres and a tin trade centre since the pre-Roman era. It also capitalized, since the Roman era, on the commercial traffic of the Cantabric Sea and the Atlantic, and controlled the route to Britannia 25

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 1.2.6 Offloading cargoes (saccarii)

According to Naveiro López (1991a, 133; 154), this lighthouse probably not only had an economic function but also a particular strategic role on a political and administrative level. We know from Orosius (OROS. Hist. 1.2.71) that this lighthouse was supposed to warn vessels from the British Isles at night of the proximity of the peninsular coast. This was the only continental port for boats heading for the English Channel or St. George’s Channel, normally relying on fast Atlantic currents (Millet, 1998, 70). This lighthouse might have been built in the IInd century as we can deduce from some materials found by Monteagudo in the excavations carried out near the tower (vid. Martínez Maganto, 1990, 80). A votive inscription was found carved into the natural rock, situated about 15 meters south of the tower. The inscription is dedicated to Mars (Marti Aug(usto) sacr(um)) by one ‘G. sevius Lupus Aeminiensis Lusitanus)’, who as an architect (architectus), possibly an architectus militaris, according to Le Roux (Le Roux, 1990).

THE JOBS ASSOCIATED WITH THE TRANSPORTATION AND STORAGE OF COMMODITIES Roman vessels had simply a commercial function. There were no such things as passenger vessels, so travellers did as they were to do until the packet ship made its appearance in the XIXth century: they went to the waterfront and asked around until they found a vessel scheduled to sail in a direction they could use. Having selected his sailing, the traveller arranged to book a passage with the magister navis (ship master), the officer charged with the business side of the voyage, the maintenance of the ship and similar matters. The handling of the vessel under way was left to the sailing master, the gubernator (Casson, 1994, 153). The vessels build during the Imperial period belonged to the naviculari, well known both through inscriptions and judical texts preserved in the Digesto. These naviculari, representing rich individual owners or powerful sea- transport societies, were in charge of the supply of the differrent regions of the empire and, in particular, of the city of Rome, through the known system of annona. In addition to these representatives there existed in the different ports of the empire a whole naval transport industry, which was, in turn, at the service of the merchants.

Other lighthouses or torches, like the Fariña example, might have existed at Arousa bay, making the entrance by the Ulla River easier (A Lanzada, O Grove, St. Tomé do Mar, Torres do Oeste). The old Torch of Donón in Mozarro, where several tombstones dedicated to a Gallaic-Roman deity were found, had possibly the same function (Arias, 1992b, 55).

The epigraphy explains the different professional categories of navigation: the organization and specific occupations (navicularii, nautae, ratiarii, utricularii …)(Izarra, 1993, 13). For example, we know of numerous boatmen’s guilds connected to the Tiber – scaphari, caudicarii, lenuncularii and lintrarii –, associated with the type of vessels in service (Rougé, 1978, 18-19). The urinatores, who had a very particular job, can be classed with these boatmen. As already mentioned, the urinatores belonged to a special class of divers who tried to salvage, whenever was possible, commodities that had fallen into the sea at various depths (id.: 19). A text of the Digesto (Gassend, 1978, 30) mentions a trader that would have hired urinatores to salvage a shipwreck that went down in a storm.

In a passage from Pomponio Mela (MEL. 3.11) we find a reference to a turris Augusti situated in the estuary of the Sar River, near Iria Flavia (Forum Iriensis). The existence of a possible lighthouse on the outskirts of the city is rather interesting since this city is, together with Brigantium, the main nucleus along the shore of presentday Gallicia (Pérez Losada, 2002, 87). The lighthouse thus had a clear harbour and commercial function given its proximity to the only one so far documented by the archaeological remains. Among these was found a set of imported materials that clearly prove, at least for the early empire period, the existence of a storage and redistribution centre in that region (Naveiro, 1991a, 139; 153-154; Pérez Losada, 2002, 97; 103-104; 108).

26

THE NATURE OF THE ROMAN TRADE IN THE ATLANTIC

Figure 1.2.7 Port arrival (mosaic from Rimini) responsible for the registration of goods in and out of the warehouses; where these functionaries worked in port warehouses they were known as horrearii (Rougé, 1978, 19; Chic, 1984, 41; Izarra, 1993, 190). Quality control was also of importance. We know that, for example, each cereal load was accompanied by a sealed leather bag that contained a sample to check if the product was in a good condition; in case of wine, this had to be accompanied by sealed bottles (ampullae) and professional tasters to check if the quality of wine matched the sample (Izarra, 1993, 190). VESSELS USED IN SEA TRADE Roman commercial fleets were made up of extremely varied vessels of small and medium cabotage, or of large deep-sea vessels. This variety of vessel occurs because each region of the Mediterranean had its own traditional types. Unfortunately, it is not yet possible to determine with certainty the origin of the main types of vessels. The different written sources tell us, however, the names given to some of them, such as Isis, Corbita, Ponto, Cladivata, Gaulus, Rotundae. These names are also sometimes associated with the specific transports or with the volume of goods they could transport, such as the ones referred to by Strabo (STR. 3.3.1) when he described the Tejo as a place of large port capacity – the muriagogoi, which, according to this author, could transport about 10,000 amphorae. There were other names associated with specific transport functions, such as the naves onerariae, with a capacity of more than a thousand tons. They were used only for the transport of building materials and grain for the annona and could anchor only in the large ports of Ostia, Alexandria, Antioch, Massilia, Carthage and Piraeus (Meijer and van Nijf, 1992, 152).

Figure 1.2.8 Mosaic with mensores

There were other jobs related to the transport of commodities, such as the saccarii, directly connected to the transport of commodities in the ports and places of consumption. The jobs related to the storage of goods in ports are better known and among these we can identify the role of mensores (machinarii or ponderatores) and tabularii. The former had to guarantee the volumes and weights of the commodities; for this they used previously established and carefully checked measures, resorting at times to double-pan scales (machinae). Tabularii were

For the study of these vessels of commercial character we have at our disposal, besides the salvaged remains from shipwrecks, an important iconographic list (vital for the knowledge of the different types of tools used) and a group of texts, essentially literary and, rarely, technical ones, which allow a more or less accurate approach. 27

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 1.2.9 African mosaic with a ship loaded with amphorae

Figure 1.2.10 Pompei. Graffiti of a ship entitled Europa done by sailors on the wall of shop in Via Dell`Abundanza

More and more nautical archaeological studies have provided important material for the study of tonnages and naval building techniques, although they cannot accurately explain riggings or handling. To accomplish this scholars have frequently resorted to the numerous representations of Roman daily life that exist – reliefs, mosaics, murals, grafitti, terracotta plaques and coins.

to the bow, which gives it a symmetrical profile (Pomey, 1978, 20);  the third design can be distinguished from the others by the shape of its concave bow that ends in the shape of a ‘spur’ at keel level; the stern resembles type 1 (Pomey, 1978, 20). In the majority of these vessels we find some common characteristics. For example most cabins were situated in the stern and only rarely in the centre. We also know that they were equipped with two lateral rudders, protected by a sort of lateral box that came from the front of the ship (Pomey 1978, 22-23).

These allow – apart from the particular design of each representation – three fundamental types of hull in profile, which can be identified by their extremities:  the first, far from being the most common, may be considered as the cargo ship par excellence: it has a convex bow and a powerful stern (Pomey, 1978: 20);

The existence of a round sail, perpendicular to the axis of the ship, was also a feature, even though it was not particularly large and was mostly used to improve the

 the second, apparently less well-known, also possesses a convex bow but presents a shallower stern and similar 28

THE NATURE OF THE ROMAN TRADE IN THE ATLANTIC

Figure 1.2.12 Reconstruction of Xanten’s Roman boat

Among the several types of craft documented in the written sources, epigraphic monuments and other features (mosaics, reliefs, etc.) – in addition to the ponto (a large vessel used in commerce) – there were two types of large river barge, the linter and the rates, which could navigate in the meanders of the waterways.

Figure 1.2.11 Oberstimm model

The linter (also known as the lynter or lintris; Bonsor, 1989, p. 89), seldom mentioned as a maritime vessel (Izarra, 1993, 112), was typically a riverboat or lacustrine vessel used in shallow waters; it was manoeuvred with an oar and had keel, bridge and sail (Chic, 1984, 34; Parodi, 2001, 30-32). The length of the craft (it did not have a flat bottom as a simple barge) gave it excellent mobility although it made it unstable (Grenier, 1934, 592). Cicero (CIC. Brut. 60.216), mocking an orator who swayed from right to the left when he spoke, said that it seemed he was talking from inside a linter. The most primitive craft were carved from a single trunk, but later became more complex and larger as required (Izarra, 1993, 111). This type of vessel was meant for the transport of people, animals and goods; it could shelter up to six individuals (Izarra, 1993, 111). Caesar indicates the presence of these small boats in the Sona and Sena rivers; Livius (LIV. 21.26) in the Rodanus; and Ovid (OV. Fast. 6.779) in the Tiber (cfr. Grenier, 1934, 591). These boast were also used (like the rates) for the building of ‘boat bridges’.

balance of the vessel at speed and to help with change of direction (Pomey, 1978, 24). Except for the large three-masted vessels, which from the IInd century onwards could support large sails, the original masts were made from a single trunk or branch and thus prevented the overlapping of round sails (Pomey, 1978, 24). Powered by sail, speed depended on the direction and intensity of the winds. For day sailing, taking into account departure and boarding manoeuvres, speeds could vary between 2 knots with lateral winds (the knot corresponding to one sea mile per hour), and 4 knots with stern winds (Naveiro, 1991a, 121). Since there were fewer manoeuvres and a better use of wind and current, in the open sea navigation speeds could rise to 5, 6 or more knots as long as the winds were favourable (Naveiro, 1991a, 121).

RIVER BOATS

The rates (also known as rates silve rataria) were simple rafts that were built by connecting trunks with ropes or nails (Chic, 1984, 35; Parodi, 2001, 31); they were solid and sturdy even though they were more primitive and of simpler construction than the linter (Grenier, 1934, 59293). They were powered by long pole, oar or sail, according to their size and the nature of river they were on (Parodi, 2001, 32). Like the linter they could be used as a barge or aid for loading and unloading goods (Grenier, 1934, 592-93; Parodi, 2001, 32).

Although river boats (naves fluminales) are known for their precariousness and lack of speed, they were widely used in the Roman world. Unlike sea vessels, which were only powered by oars and sails, river boats had the additional benefits of being able to rely on downstream currents and make use of shallow water (Izarra, 1993, 157). Use depended on the features of each river and on the season of the year, in particular autumn and early spring when the rivers presented a larger volume of water (Izarra, 1993, 76-77).

The ponto was a high-tonnage vessel powered by sail; sharp-bowed, it had two masts and a solid hull that formed a volute in the stern. It also had a rather prominent spur and lateral reinforcements that were used to hook ropes. This ship was particularly suited for trade use on rivers and larger waterways. It could also be used in sea navigation as a large cargo ship, or for military purposes (being included in the ‘naves onerariae’ (Izarra,

River boats could be of several different types according to function and the nature of the river. Almost all were powered by oars, although occasionally there was a small sail which helped the oarsmen, as with larger barges. The majority of riverboats, when they were not simply called nautae, acquired specific names (Izarra, 1993, 38). 29

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

pulling the barges with a rope tied to a mast that was fixed to the bow (Chic, 1984, 37; Parodi, 2001, 25-26). This system required the existence of appropriate tow paths (pedestro iter) (Chic, 1984, 37). Scaphae, barges powered by oars, could also be employed to pull sail craft upstream. The existence of these barges is especially documented in Narbona’s naval iconography (Izarra, 1993, 56) and in inscriptions found in Seville (Bonsor, 1989, 89), where there are references to boatmen (schapharii) that traded with the Colonia Romula Hispalis (Seville). These boatmen are said to have erected several monuments to the Emperors Antonine Pius and Marcus Aurelius, or to eminent citizens such as L. Castricius Honoratus and Sex. Iulius Possessor. An inscription found in the necropolis east of Cordova mentions a female slave who belonged to the corporation of the portonarii (Chic, 1990, 29).

1993, 115; Parodi, 2001, 27-30)). Caesar referred to this kind of vessel in Gallia (CAES. Gal. 3.13.1). The three above-mentioned models represent the main types of riverboat, but there were numerous varieties of craft in the context of trade and transport of goods, for example: the stlatta, round of hull, flat-bottomed, powered by oars and meant for trade; the vegeiia, a special vessel, also powered by oars even though it could also transport more passengers or goods; the naves codicariae, smaller covered vessels with round hull, hold, and a mast especially designed for towing (vid. Grenier, 1934, 598-99; Izarra, 1993, 108-109; Parodi, 2001, 2425); and the scapha, another type of smaller vessel used in inland navigation and corresponding to barges powered by oars. It also had a round hull and was used for the transport of goods and passengers; they were also useful for towing and manoeuvring (Parodi, 2001, 33-34).

We know that in several places the Guadalquivir was adapted for the navigation of large barges: a series of dams (portus) was constructed to retain and raise the water level (Chic, 1988b, 11). Similar adaptations are also documented in a text by Pliny (PLN. Nat. 3.9.1), concerning the Tiber. In it he refers to the existence of natural pools (piscinae) to hold the water in summer and make navigation lower down easier. Rome’s general interest in keeping rivers navigable, without reference to a specific place, is also documented in the legal dispositions compiled in the Digesto, i.e. Book XLIII, Chapter XV (Chic, 1978; Escacena and Padilla Mongue, 1992, 73).

In later times, Ausonius also mentions other types of vessel. In an extract (AUS. Epist. 22.1.7) he says that he wrote to a man called Paulinus of Nola to ask him to supply to his procurator, Philon, a nausus or nausum, a well-known trading vessel, so he might transport his stored goods. He also refers to different types of vessels that his procurator might use to sail the Tarn and the Garona rivers (cfr. Grenier, 1934, 598; Izarra, 1993, 114). The tow rope was the method used to pull boats upstream. On certain occasions a set of oxen might have been used, but it was most commen to see men (herciarii)

30

THE NATURE OF THE ROMAN TRADE IN THE ATLANTIC

1.3 TRANSPORTATION COSTS IN NW HISPANIA P. de Soto Modern studies on the configuration of communication and transport networks have become popular over recent years thanks to the innovation of GIS. This new approach to the study of transport networks derives in part from Hagget (1969). He and others demonstrated that network morphology provides understanding of economic and commercial structures across a wide range of territory. With regards to a British case study, Haggett (1969) concluded that the communications network gave primacy to London as a main vertex in the network. Later on, in the 1970s, scholars realized that this primacy stemmed from an earlier period – Roman times.

INTRODUCTION To understand the importance of the Atlantic façade and its close relationship with the NW Peninsula it is necessary, of course, to consider the question of transport in Roman times. Ringrose (1972) stated that peninsular stagnation between 1750 and 1850 was due mainly to a relative inability to adapt transportation to the requirements of that period. In Roman times the physical and cost conditions shaped transport systems and reflected the landscape and patterns of human settlement. The close relationship between these two historic moments is based on the continuity of road networks from the Roman period to the XVIth century AD (Uriol, 1985). Up to the XXth century the costs of land transport hindered contacts between the centre and periphery of the Iberian Peninsula, with the only exception being the areas located near rivers, which benefited from the low costs of water transport (Carreras, 1997, 164).

Dicks (1972) established that the communications network of the Roman province of Britannia was organised according to a form of ‘tree shape’, in which Londinium was located in a central position. His study analyzed the whole Romano-British road network in Roman times, with roads the simple branches of the network; like a river, the more tributaries a road had the more important it was. Following this method, he reached a conclusion that the communication network was organised according to three large axes that related the capital, Londinium, with three other centres: Ebucarum (York) through Ermine Street (Fens-Lincoln), Isca (Caerleon) through Silchester-Gloucester and Deva (Chester) through Watling Road (St. Albans-Wroxeter). According to Dicks (1972) this road pattern responded to the first military requirements of the Roman occupation, although such axes were developed later to generate the internal economic relationships of Britannia.

Communication and trade in antiquity relied on three main systems of transport. As the ancient sources relate, the most economic form of transport – the fastest and with most carrying capacity – was maritime navigation.1 This means of transport was especially employed in commercial relations of the Iberian Peninsula. The second most favoured inland transport was riverine; the Iberian Peninsula rivers such as the Ebro, Tajo, Duero, Guadiana, Guadalquivir, Miño and its tributaries stood out.2 Finally, the third means of communication was by land. Rome laid out a complex network of roads, approximately 90,000 km of paved roads and an overall network of 300,000 km (Chic, 1993). In the case of Hispania, the road networks were established sometimes along pre-existing pathways, for instance the Via Heraclea or the Via de la Plata (Silver way). However sometimes the roads were also built ex novo based on economic, military or political interests.

In the present work, the methodology developed by these British scholars will be applied to study how the NW of the Iberian Peninsula organized its communication network. From the particular shape of communications in this geographical region, an attempt will be made to recognize its key features and how they affect regional political and socio-economic aspects. Another important aspect in the study of the NW road network will be to focus on the centrality of the different urban centres in the territory. Such studies come from the concept of

1 On the importance, use and consequences of navigation in antiquity see Rougé (1975); Pomey (1997). 2 There is recent monograph on classical resources, history and rivers of the Iberian Peninsula in Roman times (Parodi, 2001).

31

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 1.3.1 Example of an ordering diagram with simple braches to the network

Figure 1.3.2 Example of a graph diagram (Social Network) rating nodes according to its degree of centrality (Network Degree)

Centrality Degree, employed in a so-called Social Network Analysis (Freeman, 1979). Such analyses give values to each ‘node’ (or point to which more than one roads converge) according to the number of total connections linked to it. Therefore the more connections received by a single node the more important it is. In other words it has a higher degree of accessibility or communicability.3 The origin of those analyses stem from “graph theory”, applied not only in mathematics but also in computer science. Graphs are understood as a collection of points (i.e. cities, people…) identified as vertices or nodes, connected by arcs or lines, which identify relationships such a distance, cost (i.e. economic, social…).

In our study, a methodological parallel becomes evident: the more transport infrastructures are associated with a particular population centre, the better the accessibility will be for its people and merchandise. The complexity of the relationship increases when a correct network is defined from the archaeological and historical data, and when trying to rate the diverse types of transport employed. This will probably not affect the concept of the “degree of centrality” when arriving at either a secondary road or the existence of maritime port of interprovincial trade.4 From our study, it is believed that the key urban centres in the NW transportation network in Roman times will stand out, as well as the territorial organization in terms of mobility.

Despite the widespread dissemination of these applications nowadays, the theory was developed by Leonhard Euler in 1736. He applied his theory to solve the communication problems of the inhabitants of Köningsberg, who built 7 bridges to optimise mobility. Afterwards, in the middle of XIXth century, new intellectual challenges related to transportation (i.e. Chinese postmen, travelling businessman) were solved by graph theory. In the XXth century, sociology and social network studies have helped develop the latest applications of graph theory. Those applications have helped us to understand interpersonal relationships, and they have even been employed to analyse Internet communities.

As mentioned, our first step was to analyse the degree of centrality amongst different cities, as well as their layout, using the tree-branch theory proposed by Dicks (1972). It was attempted to find as many historical and archaeological details as possible in order to understand the communications layout (graphs theory) on which to base further study.5 TRANSPORT INFRASTRUCTURES (pre-Flavian period) Proper communications were set up in the NW peninsula in the Flavian period, however their origins stemmed from previous periods. The development of Roman roads in this region is obviously closely related to Roman state interests. However, in most cases they took advantage of

3 This system of centrality is interpreted as the most intuitive way of connectivity analysis (Freeman, 1979, 220) and was introduced by Shaw (1954). But Nieminem (1974) was the one who defined a mathematic algorithm: Cd(Pk) = Σ a(Pi,Pk). Later, other types of analysis were established, such as Betweenness Degree or Closeness Degree. All are used in the study of social networks due to those functional components (influences and functions of individuals according to the degree of closeness). Because of the functional and economic components of our study, the analysis chosen is more related to the potential of communicability amongst cities.

4 In our case we added a different value to each point according to the mean of transport and the road category (between 1 and 4); a node with only one point (rated 4) could have more accessibility than a node with three linked points (rated 1). 5 It is not intended here to look in depth at the historic and archaeological data of the NW territories.

32

THE NATURE OF THE ROMAN TRADE IN THE ATLANTIC

the primitive organization of the territory coming from pre-Roman periods.

continued along the Bay of Biscay (Atlantic route). This route is referred to mainly by Caesar (CAES. Gal. 3.26.6), Strabo (STR. 3.4.18) and Orosius (OROS. Hist. 6.21.4), who describe contacts, exchanges and transport of troops between Cantabrians and the territory of Aquitania.

With regards to the maritime communications, Atlantic navigation is known from the Neolithic period, in which some cultural influences such as megaliths or metallurgy are well identified (Naveiro, 1991, 135; Ruiz-Galvez, 1986, 24). The discovery of a monoxile boat near Viana do Catello (Morais, 2006) appears to justify an early dating for its use (XIth century BC). The location of hillforts on the NW Peninsula coast close to the mouths of large rivers defines the importance of such maritime contacts. In addition they provide some clues to the possible pre-Roman communication network in the territory (Morais, 2007), together with the distribution of staple products and manufactured objects (Naveiro, 1991, 135). In the Bronze Age, communities such as the Tartessian or Phoenicians (documented from the Peripli of Hannon or Himilcion) developed regular maritime contacts with the Atlantic communities to obtain gold, ivory, cooper, tin and amber. Arruda (1999-2000) demonstrates in her study of the estuaries of Sado, Tagus and Montego, as well as the Algarve coast, that eastern materials are significant, which suggests regular contacts with Phoenician from the VIIth BC onwards.

River routes in pre-Roman periods were also often employed. The use of flat-bottomed boats, made of skins, trunks or wooden-planks or monoxile craft allowed northern peoples to navigate through most rivers in the NW (Naveiro, 1991, 138). The coastal features allowed shelter and a place for loading and offloading ships in the river mouths. This favoured the development of activities and settlements in rivers during antiquity (Morais, 2007). Due to such geographic morphology, a series of statios were developed along the Atlantic and Cantabrian coasts; some sites existed in pre-Roman times and were consolidated in the Roman period. The best description of river navigation in the NW Peninsula appears in the work of Strabo (STR. 3.3.4), in which the geographer refers to navigation along the Douro and Miño rivers (upriver navigation of 800 stadia – 150 km.). “After these two is the Durius, which, coming from afar, flows by Numantia and many other settlements of the Celtiberians and Vaccaenas, and is navigable for large boats for a distance of about eight hundred stadia inland. Then come other rivers. And after these the River of Lethe, which by some persons is called Limaeas, but by others Belion; and this river, too, rises in the country of the Celtiberians and the Vaccaenas, as also does the river that comes after it, namely the Baenis (others say "Minius"), which is by far the greatest of the rivers in Lusitania — itself, also, being navigable inland for eight hundred stadia. Poseidonius, however, says that the Baenis rises in Cantabria. Off its mouth lies an island, and two breakwaters which afford anchorage for vessels. The nature of these rivers deserves praise, because the banks which they have are high, and adequate to receive within their channels the sea at high tide without overflowing or spreading over the plains. Now this river was the limit of Brutus’ campaign, though farther on there are several other rivers, parallel to those mentioned.” (STR. 3.3.4)

By the VIth century BC, Atlantic navigation and trade in the NW Peninsula was developed further by the Gades populations, as can be inferred from Strabo (STR. 3.5.11). They kept control of the tin trade and developed trading ventures with Carthage and the Phoenicians in the Mediterranean (Morais, 2007). Information recorded by classical geographers that has survived show that navigation across the Atlantic, between the Mediterranean and the NW coast of the Peninsula, Gallia, Britannia or Germania was relatively frequent. Strabo in the Ist century BC, and Avienus in the IV c. AD (Morais, 2007), reveal the itineraries used to navigate these coasts. Because geographical knowledge was gathered over centuries, the conclusion has been reached that such commercial contacts were in place from the VI c. BC. Nevertheless, some differences and significant errors in the sources and locations are detected in our ancient sources. In the Roman period, the NW Peninsula still maintain its relationship with the Roman World through Gades mediation. Thereby, the port of Gades and the NW reinforced relationships through stops on the route from Britannia to Rome.6 There were already some strategic points on the peninsular coast, which, thanks to their location, became calls on the supply routes. It is also quite feasible that the relationship between the Cantabrian coast and the Aquitanian and Breton coasts was

The descriptions of the river routes of this territory and their navigability are of significant interest, so they may have shared a common means of inland transport.7 There is even the possibility that rivers such as the Douro made the Atlantic coast accessible to important land routes such as the Ruta de la Plata (Silver route) (Balil, 1974, 221; Naveiro, 1991, 139). Furthermore, NW territories received more attention by ancient geographers in the description of rivers, although they did not always describe their navigability (Parodi, 2001, 227228).

6

Rome appeared to adopt immediately the Gaditan Atlantic policy as its own. From the start, Roman sailors attempted to obtain key information to navigate across the Atlantic North (data about courses and calling ports and exchange points). Later, Romans will wage wars to pacify and assimilate hostile territories in order to reach the Northern Gallia and Britannia (Naveiro, 1991, 130-131; Parodi, 2001, 232-233).

7

Only a few years ago, the transport of wine inland towards the port of Gaia-Porto was done along the Douro over a long distance.

33

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 1.3.3 Main communication routes in Pre-Roman times

via Vía XVI of the Itin. Ant. that crossed along the Atlantic foreshore was the probable result of a preRoman axis. Likewise, a multitude of secondary routes, tracks and natural features may have had a pre-Roman origin.

Finally, the land routes had also a clear pre-Roman history before the establishment of the Roman road network. It seems feasible that the natural features and routes of the territories were employed in antiquity. Most Roman sites with indigenous precedents were located in strategic places to control natural thoroughfares. The location of stone cairns (amilladoiros) at cross-roads discovered in the NW presupposes the existence of indigenous track infrastructures. These were identified by such stone cairns and later preserved in medieval times to identify such crossroads8 (Rabanal-García, 1996, 278, Naveiro, 1991, 141-142).

Even before, but mainly in Augustus’s reign, the communication network started to grow. The creation and consolidation of a new provincial project, the establishment of new conventus capitals and new cities and colonies forced the administration and Roman army to set up a new communications layout. The Roman general Agrippa was responsible for such project when he arrived on the peninsula in 19 BC (Rodà, 1998; 2007). In this way the main cities of the Peninsula were defined not only as strategic military sites but also as economic and communication hubs. Cities such as Tarraco, Clunia, Caesaraugusta, Lucus Augusti, Asturica Augusta and Brácara Augusta, were favoured by this active policy in infrastructures (Blázquez, 1997). Furthermore, Agrippa was key to achieving peace in the NW, with a policy of integrating the newly conquered territories. Developing an intensive infrastructure network and the employment of indigenous personnel for local municipal posts were tasks that faced the Roman administration in terms of the process of integration (Rodà, 2007, 27).

Following ancient sources, such as Strabo (STR. 3.4.10), it seems likely that two important routes were employed in pre-Roman times. The first route connected the coastal areas of the Mediterranean near Tarraco (the Iberian Kesse) with the Cantabric coast in the Bay of Biscay near Oyarzun, and perhaps continued further west (STR. 3.3.8; Fernández Ochoa and Morillo, 1994; Yanguas, 2001; Urteaga, 2003). This route was probably directed towards Pompaelo, following the way north of the Ebro River. There was also a parallel E-W axis linking the territories of the northern Douro Valley with places such as Caesaraugusta. The flow of Roman coinage from Spanish mints in the Julio-Claudian period probably came from this route (Naveiro, 1991, 141). The second route coincided with that known as the Ruta de la Plata (Silver route), which was in full use in the Augustan period. It is believed that Tiberius kept roads in good condition, and the lack of milestones from the emperors Trajan and Hadrian appears to indicate that such landmarks were still visible (Blázquez, 1997, 6-7). It was begun in the last third of the IInd c. BC, in the Viriatus revolts. Its function was to link the Baetican territories and the Cantabric zone, where the main conflict took place at the end of the Ist c. BC. Other land routes were in place in pre-Roman times, the

ROMAN ROADS (from Flavian onwards) The physical features of the Iberian Peninsula, surrounded by the sea, favoured contacts between coastal towns with the rest of the Empire; inland movements were more difficult. In the inland regions, not all the territories suffered the same problems of accessibility and mobility. The geographical morphology favoured inland contacts along the main rivers and in this context Strabo (STR. 3.1.1) detailed a total of 2000 km of river networks in the Peninsula (Yanguas, 2001). Apart from their rivers, the important regions of the Peninsula

8

It seems clear according from several scholars that amilladoiros had a Pre-Roman origin. They were known in Roman times as “Mercury Mounts” (Caamaño, 1979, 284-285; Acuña, 1971; Bermejo, 1978, 77117).

34

THE NATURE OF THE ROMAN TRADE IN THE ATLANTIC

Figure 1.3.4 Main communication networks from the Flavian period onwards. Main ports along the Cantabric coast: 1. Brigantium 2. Noega 3. Gigia 4. Portus Vereasuecae 5. Portus Blendium 6. Portus Victoriae Iulobrigensium 7. Portus Amanum 8. Oiasso were pacified. Tarraco, Valentia, Gades, as well as other urban centres, were developed during the IInd-Ist centuries BC, in the coastal region, fostering military, political and economic control desired by Rome.

lacked good land communications, mainly because of the mountain barriers (Cantabric, Galician, etc.). Geographical hindrances to communication and economy, basically in land transport, remained unchanged until the XIXth century (Ringrose, 1972). The high cost of transport and low quality infrastructure reduced contacts between the centre and periphery, except for the Ebro and Guadalquivir valleys. Certain forms of agriculture, limited by mobility, tend to lead to self sufficiency (Carreras, 1997, 164). The close relationship between the Roman road infrastructure and the road system in place in the XVIth century meant that the problems of communication between centre and periphery still remained a main factor of the country’s relative underdevelopment and the changes later introduced in the XVIth and XVIIIth centuries did not significantly improve the situation (Uriol, 1985).

This was not the case for the central zone of the Peninsula or the NW; urbanism of those territories by the Roman government was promoted during the Principate, stimulated by Augustus’s policy (Rodà, 1998). Similarly, the construction of the complex network of roads in this region can be dated to later periods for some coastal axes, as can be recognised in the classical sources (Solana, 2001) and the many milestones still seen today (Rabanal, García, 1997). (Because many dates on these milestones refer to restorations, they indicate not construction dates but repair.) In the NW region, territories were determined by orography. The mountainous ranges of the northern coast, such as Picos de Europa or the Cantabric range, represented major obstacles for any communications between the maritime stationes and inland regions. In the Galician-Portuguese case these difficulties were overcome, as we can see from the location of cities with excellent communications in inland territories, i.e. Lucus Augusti and Bracara Augusta. In addition, one of the main factors that affected communication layout was the urban policy aimed at developing the old military camps from the Cantabric war into flourishing municipalities. In this sense the choice of different cities did not depend solely on economic reasons but was related to strategic decisions from previous conflicts in the area.

There was a development eventually in the land transport system when the Royal roads increased traffic. However the road layout was too biased towards Madrid and its periphery, leaving entire regions with no communication at all, such as Valladolid (Uriol, 1985).9 During the XVIIth and XVIIIth centuries there were numerous projects set up by public companies to solve the problem of communication between the centre and periphery (new waterways, new roads…). Nevertheless, problems remained due to a lack of resources and bad planning (Ringrose, 1972). In the Roman period, geography determined in many cases the economic exploitation of territories near the coast. Territories in the NE, Levante or Baetican coast were exploited and urbanised earlier, when those regions

Geographical location and military interests came together in Cantabric territory. It seems feasible that the Roman interest in the conquest and pacification were influenced by other factors. The exploitation of metal ores, known from antiquity, was a key factor and one which fostered the need to ensure a network of stable ports around the north Atlantic, including the territories of Gallia and Britannia (Parodi, 2001, 233).

9 Uriol (1977, 1985) highlighted different road itineraries and the required time to cover some distances, which did not differ much from the Roman period. The only technical evolution in land transport, basically wagons and carriages took place in the XIXth century (Lawton, 2006, 426-454).

35

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

central role among the northern ports. Its proximity to Lucus Augusti and its relationship with the Via XX allowed it to maintain good communications with the western territories of the Cantabric coast. Flaviobriga, the ancient Portus Amanum, also enjoyed special status, since it was the only maritime city that received the title of colonia in the year AD 74 (PLN, Nat 4.110). Perhaps its importance stems from the change of era, as not much attention is given to it by earlier authors such as Strabo. Its location favoured seagoing communications between the NW Peninsula and Atlantic Gaul, as well as being on one of the access routes towards the inland territory. This helps explain why it may have generated a special interest for traders. Portus Victoriae offered excellent communication facilities for inland transport, facilitating foreign products to reach markets such as Iulobriga and Segisama. Finally, the port of Oiasso, on the estuary of the Bidasoa River, also benefited from a position that enabled it to maximise the distribution of goods inland, using not only the river but also the land roads following the Pyrenees valleys (Urteaga, 2003, 106).

MAIN ROUTES OF COMMUNICATION As explained above, the main communication network in the NW was defined according to its maritime routes. This means of transport was vital for bringing staples and was determined by the cost and speed of the ships employed. The position was enhanced by the special configuration of the Atlantic coastline and its main river estuaries that allowed upstream navigation for commercial ships (Sillières, 2003, 22). The main coastal itineraries included the established long-distance one connecting the northern territories with the port of Gades, as well as the regular coastal-traffic route calling in at many of the ports en route. According to Naveiro (1991, 125-127), the whole Cantabric and Atlantic coasts established a series of ports and harbours linking all the commercial trade. Ships were kept safe in bays and river estuaries, from which they ventured to sail on the open sea or closer coastal navigation. Gades represented the first call for vessels coming from the Mediterranean and travelling to the Atlantic provinces (PLN. Nat. 2.57.167). It was probably the departure point for Baetican products. Seagoing routes towards the Atlantic had likely calls at Olisippo, Portus Cale, and Brigantium before reaching their finals destinations. Navigation could continue farther on to provinces such as Britannia, northern Gallia (Belgica, Aquitania) and Germania (Naveiro, 1991; Morais, 2007).

River communications in the NW created an unbalanced infrastructure in the communications network. Long rivers such as the Ebro, Guadalquivir, Tajo, Duero and Miño made accessibility inland easier, thanks to their navigability upstream. The navigational features of the Tagus made it one of the main axes towards inland territories. Many cities grew on its banks, such as Olisipo, Scallabis, Lebura and Toleto. Further north, navigation of the Douro was described by classical geographers, who pointed out its use as a means of communication for as long as 800 stadia (STR. 3.3.4). Finally, in the extreme NW, there was the Miño, with its tributary the Sil, provided greater penetration inland. The location of Lucus Augusti, only 70 km from the navigable rivers according to Strabo (800 stadia – 150 km), seems to be related to it (Parodi, 2001, 228).10 Similarly the location of Bracara Augusta on the river Nebis (present-day Cavado) appears to demonstrate navigation along other Atlantic rivers.

These routes avoided the Bay of Biscay, as can be inferred from the archaeological material documented in the main ports of the Cantabric region. Ceramic assemblages from the Cantabrian ports looks quite different from the assemblages recorded in the Atlantic centres. Archaeological assemblages from Cantabrian sites appear closer in composition to the ones from the Garonne region, the middle Ebro valley and La Rioja (Iglesias, 2003, 108). This material implies certain commercial contacts by sea between the Cantabrian coast and territories near Burdigala. Cantabrian navigation was affected by the particular features of this sea, similar to those found from the Bay of Biscay to Galicia. It is a typical straight coast, of recent formations, with continuous cliffs and a few lower coastal strips and numerous estuaries and anchorages (Iglesias, 2003, 108). Despite this apparent homogeneity (Fernandez Ochoa and Murillo, 1994), only some locations went on to become important ports due to their strategic positions close to natural inland ways or mineral resources (gold mines).

Along the Cantabrian coast the many estuaries permitted the establishment of anchorages and regular exchanges between boats and ships of different sizes. Cantabric vessels offloaded cargoes to small boats that were hauled upstream (Iglesias, 2003, 108). It seems feasible, according to the recent archaeological finds, that the Deva and Nervión rivers were employed as water routes inland. The use of Nervión was especially linked to the city of Oiasso.

Centres such as Brigantium, Niega-Gigia, Portus Vereasuecae, Portus Blendium, Portus Victoriae Iulobrigensium, Portus Amanum-Flaviobriga and Oiasso enjoyed special associations with the Roman maritime communication network, although with some singular differences (Fernandez Ochoa and Morillo, 1994, 165170). Among the major centres Flavium Brigantium appears to have enjoyed a prominent role. The construction of a large lighthouse, known as the Tower of Hercules, as well as the possible presence of a statio (Fernández Ochoa, Murillo, 1994, 165), indicate its

Another case is the River Ebro. Its navigation is recorded by the classical sources11 and allowed communication between the NW territories and the Mediterranean coast, near Tarraco. Together with navigability, mobility in 10 Rodríguez Colmenero (1996, 419-424) refers to the importance of the River Miño for the transport of building materials to the city; good river communications was one of the main factors for the city’s foundation. 11 For ancient sources quoting the River Ebro, see Parodi, 2001.

36

THE NATURE OF THE ROMAN TRADE IN THE ATLANTIC

Figure 1.3.5 Map showing the main communication land and river networks in the NW

region with the other side of the Pyrenees near Burdigala, allowing access to Aquitania through northern Meseta, the plains of Alava and Pamplona. It became the basic thoroughfare for the administration of this territory and a link for the land and river routes that continued to the coastal ports (Yanguas, 2003, 50).

inland territories was complemented by an extensive road network following the Douro valley and the northern Meseta on the fringes of the Cantabria range. As well as their use of rivers, the Romans consolidated their power through a sophisticated road network (Naveiro, 1996) and river valleys were also used as land tracks. The main roads were described in the well-known Itineraria, such as the Antonine itinerary, the Peutingerian tabula and Rabean anonymous.12

Another important axis for communication was Via XVII (Item a Bracara Asturicam), XVIII (Item alio itinere a Bracara Asturica) and XIX (Item a Bracara Asturicam), which created a triangle-shaped network that linked cities such as Bracara Augusta, Lucus Augusti and Asturica Augusta.

These Itineraria included most, but not all, roads that crossed the territory. The main roads were still marked thanks to the conservation work of later emperors. These conservation works are recorded in the milestones preserved today, showing the imperial policy in the territory (Rabanal-García, 1996). In the E-W direction, the main road from the Itineraria corresponds to that known as Via I, and XXXII-XXXIV (Via Tarraco ad litus Oiarsonis) (Solana, 2001, 89-90). It covers the route from Caesaraugusta to Lucus Augusti, following the axis of Vareia, Segobriga and Asturica. It can be observed that this great axis, parallel to the Cantabric range, is the one that has retained most milestones (Rabanal-García, 1996, 280-281). This road (Via XXXIV) also linked this

In the S-N direction, the main axis covering the region was the Via de la Plata (Silver road). The road has a military origin and was built between 139 and 78 BC. The lack of milestones from the Flavian period suggests that the route was well maintained (Blázquez, 1997).13 In the times of Trajan and Hadrian the road was enhanced to preserve it and improve this vital means of communication. The axis represents the only important route between north and south of the Peninsula, and linked the capital of the Lusitanian province, Emerita Augusta, with all the northern territories.

12 There are many sources on Roman roads that have survived. Vicarello vases or Valencia tegula are two good examples, as well as the ones quoted before, which help to complete the Itineraria from the Iberian Peninsula.

13

The lack of milestones has also been interpreted in some cases as a decrease in the importance of their use (Carreras, 1997).

37

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 1.3.6 Reproduction of the Peutingerian tabula

Figure 1.3.7 The main Roman communication links and conventus capitals

communication network, as well as showing differences in mobility in the diverse regions studied (territories in which many of the cities have a high degree of communicability are identified as either “well-” or “poorly-linked”.

The Itinerari Antonini also included the road that linked Bracara with the territories of the Lusitanian coast, mainly Olisipo. This road (XVI) ran parallel to the coast, crossing towns such as Aeminio and Coninbriga. Special mention should be made of the Via per loca maritima, known also as Via XX. The importance of this route was its role in linking the diverse maritime and riverside settlements of the extreme NW of the Peninsula. The itinerary of this road linked the conventus capitals (Bracara, Lucus and Asturica), but with the intervention of different means of transport. For instance, some secondary parts of its network were river routes (Franco, 2000).

Part of this study stems from studies on social networks defined above and also studies on transport and mobility carried out some years ago. Carreras (1994) defined a transportation cost map of the different means used in Roman times based on the analysis of communication network in the Roman Empire and the different transport costs (figure 1.3.8). Thanks to a thorough study of Diocletian’s Edict, he established a series of cost ratios amongst means of transport: seagoing navigation = 1; downstream river navigation = 3.4; upstream river navigation = 6.8 and land transport = 43.3.

ANALYSIS OF THE TRANSPORT NETWORK With the aim of analysing communications in the Iberian Peninsula, it was thought sensible to study the role and importance of different Roman cities. In order to define importance, each city (node) was rated according to its level of accessibility, analysing for each one its degree of centrality within the network. The degree of centrality obtained offers a vision of hierarchy within the

From the resulting map it became evident that lower transport costs were recorded in coastal areas in the Peninsula, as well as along the valleys of rivers such as the Ebro, Guadalquivir, Tajo and Duero. The map reveals higher values in inland region as a result of inaccessibility problems.

38

THE NATURE OF THE ROMAN TRADE IN THE ATLANTIC

Figure 1.3.8 Transport cost map of distributing products coming from the Guadalquivir valley (Carreras, 1996) settlements. There are also significantly lower values in specific regions, such as the Cantabric range.

In the present study it was decided to analyse the communication network from another viewpoint and the communication network is here analysed without taking into account transport costs. Then, using the network and the different means of transport, the nodes in the network were rated according to their accessibility. The classification of these differing nodes was undertaken following studies on the degree of centrality (Freeman, 1979), by quantifying the different axes (or edges) which link each node, and also by distinguishing edges according to the means of transport. In other words, values were given to the edges depending on whether they were linked by maritime, river or land routes.

From this diagram other graphical representations can be generated to help with its interpretation. By interpolating the values of each node, the map reflects the accessibility in the territory as a whole (mobility of people and merchandise). As can be observed (figure 1.3.10), there are clear differences in the degree of centrality in the diverse territories studied. In the Roman periods there were areas with little communicability, such as north of Lucus Augusti, between the Eo and Besaya rivers and the zone between the Nervión and the Pyrenees border. Other regions with little infrastructure was the area of southern Salmantica, as well as the mountainous regions of the Cantabria range and the Iberian and Moncayo ranges, amongst others.

One of the most complex issues was defining the actual values for each means of transport. Following studies on communication transport in the Roman period by Carreras (1994) and the cost studies described above, it was decided that a maritime port should be rated 4, a river port 3, each main road 2, and secondary roads as 1. Adding these edge values to each node will give “a degree of centrality” rate.

As a contrast to these data, there are singular points with high rates in connectivity. The points with the highest connectivity correspond to the conventus capitals such as Caesaraugusta, Clunia, Bracara Augusta, Lucus Augusti and Asturica Augusta. Another feature is the location of important nodes along the main road axes inland, e.g. the axis defined by Lucus Augusti, Bracara Augusta and Asturica Augusta.

In the resulting graph diagram (figure 1.3.9) each node in the network received a size and colour grading according to its degree of centrality. The importance goes from green (lowest) to red (highest) and the larger the node on the diagram, the more accessibility it has. The resulting diagram shows interesting features in the network of the NW Peninsula, including the relative importance of the conventus capitals, the influence of maritime and river ports, as well as main roads affecting particular

The triangle formed by these three conventus capitals records a great number of milestones, chiefly from the IIIrd and IVth centuries AD, following the main roads XVII, XVIII and XIX in the Antonine Itinerary. This was the main focus of gold exploitation in the NW and thanks to the triangular communication network it could 39

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 1.3.9 Map of the NW Peninsula with the main nodes and axes of the resulting graph diagram

Figure 1.3.10 Interpolation map from the graph diagram, showing zones with little communication in white, and darker areas with most

followed initially the lowlands of the Douro valley before reaching Asturica Augusta, where it was obliged to follow a detour towards the north.

communicate to the axis that ended in the Ebro valley with two parallel itineraries. According to the study results, the best route was the northern one, closer to the Bay of Biscay and crossing places such as Grachurris, Calahurris, Virovesca and Regio, to reach the nearby region of Asturica. This itinerary is of extra interest in that it became part of the Santiago pilgrimage route in medieval times. This itinerary also had secondary routes to the Cantabric coast through some natural features, normally following a river. This is the case with the secondary roads that communicate with the Cantabric coast, such as Iruña through Deba, Deobriga towards the Cantabric coast, and the eastern routes from the Ebro valley or La Rioja (Yanguas, 2001, 712-713; Urteaga, 2003).

The upper axis was probably the route most employed for trade purposes linking those peninsular regions, taking advantage of particular geographical points that connected with the coast; its proximity to the military front in the campaigns against the Astures was another reason for its busy use. Apart from those axes of communication that linked eastern regions near the Ebro valley with the extreme NW of the Peninsula, the road going from the north towards the centre and south also stands out as a prominent one. This axis, traditionally known as Via de la Plata (Silver Road), linked the capital of Lusitania, Emerita Augusta, with the NW territories, especially with Austurica Augusta, crossing towns such as Norba Caesarina, Salmantica and Brigaecium.

The second axis running parallel to the Cantabric coast linked minor sites such as Uxama, Numantia, Clunia, Tella and Brigeco. This road corresponds to the Via XXVII recorded in the Antonine Itinerary. This road 40

THE NATURE OF THE ROMAN TRADE IN THE ATLANTIC

Figure 1.3.11 Map showing the main communication axes ports such as Tarraco, Valentia, Cartago Nova and Gades. In addition, the potential of different river routes generated hubs of communication: Caesaraugusta, Hispalis and Bracara Augusta can be considered as examples of such hubs. Finally, there is a series of key interior centres such as Emerita Augusta, Salmantica and Austurica Augusta. These centres attracted many roads (axes in the graph diagram) as a result of their political influence, as well as their central location in the peninsular urban landscape.

One of the most outstanding aspects of those axes was that they linked the key political centres in Hispania. The two E-W axes, as already described, linked the Atlantic cities with Caesaraugusta and Tarraco, the most important cities; the N-S roads linked this NW region with Lusitania and Betica, the other key political and economic powers in Hispania. With regards to the beginnings and modifications of these roads, the classical sources, archaeology and remains of milestones all indicate that they were initially built in the Ist century AD. Despite having less milestones preserved, it is seems feasible to think that this century was the time of the creation of infrastructures that consolidated political, administrative and economic control of the NW after Agrippa’s pacification process (Rodà, 1998), to which must be added the special requirements of the mineral (gold) interests. In terms of the IInd century, the number of recovered milestones is even lower, although those erected by the Roman emperors stand out. Of most interest is the number of milestones from the IIIrd and IVth centuries, but they do not normally register distances in great detail. Rabanal-García has proposed (1997) that they are evidence of imperial propaganda.

The data from the main nodes, following the same process described before for the NW, produced some interesting (figure 1.3.12) findings, for example the importance of coastal areas. The eastern coast, the southern Baetican coast (Malaka and Gades), Tajo estuary (near the city of Olisipo) and Porto-Bracara Augusta all appear as areas of high communicability, indicating the excellent conditions for maritime transport. The existence of these ports was linked to the requirements of the Roman State to assure a series of safe ports on the Mediterranean coast as well as the Atlantic routes towards other destinations such as Gallia and Britannia. The anchorages that developed an important network of communications inland (either by river boat or road) were the ones that became attractive ports for seagoing traffic. Ports such as Tarraco (provincial capital), Gades and Brigantium consolidated themselves as ‘compulsory’ stops for such activities.

To understand NW communications, they should be analysed within the framework of the infrastructures of the Iberian Peninsula as a whole. Therefore the same methodology has been employed for the whole province with some interesting results (figure 1.3.11). As mentioned above, the geography of the Peninsula favoured communications by sea routes. Some of the most important towns in Hispania were also maritime

One of the better linked areas was the lower and middle Ebro, in which Caesarugusta assumed a relevant role as hub of communications. From this city there was a major 41

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 1.3.12 Diagram representing the main nodes within the Iberian Peninsula

because of the River Guadalquivir and a complementary network of roads.

road network that linked all the corners of the Peninsula. The city was joined by roads with the NW towns, Lusitania and its capital, Emerita Augusta, and with the NE coast and its most powerful city, Tarraco. Furthermore, thanks to the river Ebro’s navigability, it was linked to other towns in the upper (Vareia) and lower Ebro, such as Dertosa. This latest town was also a maritime port for Rome and the rest of the Empire. Cities such as Ilerda, Osca and Pompaelo were organised in the northern regions with other intermediate points between the Pyrenees (and on the other side) and the river Ebro. Similarly, Bilbilis, Numantia, Gracchurris and Calagurris represented key transit points on the southern side and also acting as intermediate points between the river valley and other inland territories. The road network and the navigability of the Ebro and its tributaries made these regions some of the best linked in the Peninsula.

Finally, the results also reveal the importance of the N-S axis – the Via de la Plata (Silver Road). The axis started in the Baetican territories, linking cities such as Emerita Augusta, Norba and Salmantica towards the NW Peninsula, and consequently features with a high level connectivity on our map (figure 1.3.13). This may have been the result of a long tradition, having been well used since Republican times. As a conclusion, it is important to distinguish the theoretical character of the present study with regards to relationships between the network and its use in Roman times. In other words one factor is the theoretical explanation of how this network was created, and the other is its actual use, which may differ. Mobility between the main urban sites in Roman period has been studied here, favouring contacts between the three main conventus capitals. However, we do not know whether its use for military, state or trade purposes was exactly the same.

Another region with enormous mobility potential was the Guadalquivir valley. From Gades (located further away on the coast), to Hispalis, Italica, and as far as Corduba, there were excellent conditions for communications,

42

THE NATURE OF THE ROMAN TRADE IN THE ATLANTIC

Figure 1.3.13 Interpolation map of major points in the Iberian Peninsula

43

Part Two THE ATLANTIC ROUTE

THE ATLANTIC ROUTE

2.1 THE LATIN SEA AND THE OCEAN CONQUEST (II-I BC) C. Aranegui … The need to take cultural goods of the other to use them in one’s own benefit… Clara Gallini, Che cosa intendere per ellenizzazione century BC, the maritime control of the western Mediterranean basin was the key to consolidate the position of the two hegemonic powers and for this reason they intervene in specific places of the Mediterranean coast to keep their maritime hegemony. It is well known (LIV. XXIII, 2,6) that Asdrúbal prepared a squadron in 216 BC to defend the Balearic islands, the Pitiusas and the coast close to the islands of the Roman attack. Moreover, Ebusus enjoyed at that time a good port and powerful walls (D.S. V,16,1-3) and it is recorded, on the other hand, in a Roman-Carthaginian treaty of 202 BC, as being limited to 10 war vessels authorised from Carthage. However, the Punic capital built a commercial and military port in this period with a capacity, confirmed by excavations (Hurst 1994), of 220 vessels according to Apianus (Lib. 96). In other words it was quite an impressive port, and the reason that may have influenced the decision to destroy Carthage in 146 BC. (Medas 2000).

INTRODUCTION The Mediterranean as an integrated whole is, despite Braudel (1964), only a common environment at particular moments. It becomes an integrated space during the Roman Empire (and in the times of Philip II, as the cited author suggests), when there was relative harmony between the different countries on both sides, submitting hierarchically to a large metropolis. Therefore, it was far from a real integration. From this perspective we will deal with an especially interesting period of the ancient history in the Mediterranean, after the defeat of Carthage (146 BC). We will pay attention to particular strategies that reveal the western populations’ attitudes traditionally opened to external contact and leave behind military history, understood as the process of conquest. The idea is to examine particular facts from post-colonial approaches (Van Dommelen 1996, 25-45), derived from the structuralisms and socio-economic positions of a privileged scenario such as the western Mediterranean and the Atlantic façade. We will try to move away from conventional categories of ‘winners and losers’, or hegemonic and subsidiaries, according to Marxist terminology.

From the Roman side it is known, thanks to the archaeology, that in the last few years of the IIIrd century Scipio’s army removed from Iberia any obstacle to his control of the Lower Ebro (Noguera 2002). Meanwhile, Publius Cornelius Scipio created in Campania the colony of Puteoli (194 BC), south of Naples, to regulate tolls from maritime exports, as a confirmation of the Roman supremacy over the famed Punic traders. So essential for Rome was it to emulate Carthage in its control of the searoutes, that Aemilianus Scipio chose Polibius, after the fall of Carthage, to lead an expedition for the reconnaissance of the Atlantic coastline of North Africa (PLIN. Nat. V, 9-10), repeating the voyage previously carried out by Hannon (Mederos, Escribano 2000, 77107), with light vessels and cargo ships, including experts in different fields, geographers and expert seamen, necessary for a successful expedition. The expedition appears to have reached the eastern Canary Islands, which were not included in regular sea-routes at that time. A little later, Iunius Brutus, consul in the year 138

When life is altered in a community (between people of different ethnos, coastal and inland populations, farmers and traders, between indigenous and foreigners…), a readjustment takes place, affecting all the instances involved. The first Roman phase in Hispania offers models with a rich archaeological interpretation as a key to redefine cultural identities scarcely explored in common historic reflections. In addition, the Atlantic conquest involved the meeting of societies, so different as to be reciprocally admired, with the consequent contact between them. This is the paradigm or starting point from which we develop our arguments. A major change took place in our case when the coastline became more than a landscape. Between the III and II

47

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

route was protected by small fortresses in IVth (Badie et al. 2000) and IIIrd century BC (Bendala 1990,25-42), at least near Tossal de Manises (Olcina 2006, 105-117). There were also fortresses on the rivers, such as Castellet de Banyoles de Tivissa (Asensio et al. 2005), and preexistent sites that became fortresses. However, in some places, the relationship with Qart Hadast is different, since some oppida resisted falling under its control. Perhaps some of those oppida enjoyed from Rome preferential treatment as allies, as history records. As a consequence of Sagunt’s restitution by Scipio, an embassy from Sagunt was invited before the Roman Senate, where they gave, in 205 BC, the following speech acknowledging Iberian debt towards the winner: … Senators, even if it is impossible to increase the evil we suffered to keep an inviolable faithfulness towards you, your successes and those of your Generals have been greater for us than our sorrow (LIV, XXXVIII, 39) […senadores, aunque sea imposible aumentar los males que hemos sufrido para mantener una fidelidad inviolable hacia vosotros, vuestros méritos y los de vuestros generales han sido para nosotros más grandes que nuestros pesares (LIV, XXXVIII, 39)].

BC, when Valentia was founded (LIV. Per. 55), directed a punitive operation against the Galician tribes (Guerra 2000; Morais 2005, 25-26), following the Portuguese coast with the maritime support of Gaditans. The Gaditans, from their Phoenician tradition, were expert seamen and traders. The operation concluded in the battle of the River Lethes (river Leça, in Guifões), and the victory was commemorated with the construction of an altar in the Roman Campus Martius. This confirms that the sea, from finis terrae to the southern latitude near cape Juby, was included in the Roman orbit after the fall of Carthage. With the fall of Greece, the Mediterranean became a Latin sea (although the Eastern littoral maintained Greek as the common language) to be navigated by ships of larger tonnage than those prior to the Second Punic War. Half of the ships documented between the IInd century BC and first quarter of the Ist century BC fulfil this rule. They are ships that could reach 40, 50, or even more metres in length (Lazareto, Escombreras 1, Grand-Congloué 1, Albenga, Mandrague-de-Giens…). Such vessels were instrumental to the Roman foundation in the IInd century BC of ports with privileged status, such as Carteia, Valentia, Pollentia, Palma, Narbona… All these foundations are strategic ports linked to the sea-routes going from Narbon to the Strait, as well as light traffic towards Ebusus, with excellent sailing conditions, to connect the Numidia coast, which received painted pottery at that time (Sennequier, Colonna 2003, 52, 54, 63 y 72), or even western Sicily. And, of course, those sea-routes were related to the itinerary that crossed the islands through the Bonifacius Strait towards Rome.

Although for many years this emotive text, as well as others referring to the Second Punic War, was not related to any early contact between the Roman capital and the city of Arse-Saguntum, the regional archaeology of Sagunt has revealed a growing urban presence at the beginning of the IInd century BC, while other indigenous sites in the Camp de Túria disappeared violently. Renewal of the port infrastructures of the Grau Vell, the construction of a towered wall in the Castell, as well as the construction of a large civic space around a temple of square proportions, a three-part structure and a well in the porch, all these elements denote not only an important identity element – not a Roman official cult element which only appears later – distant from the Iberian tradition, but also probably indicates the setting of a group of publicani (Aranegui 2004). This took place together with the coinage, which was initially related to Massilia and Emporion weights (and closer to southern Italian models), changing to follow Roman patrons with Iberian legends (Ripollès, Llorens 2002, 110-112).

RESTITUTION OF SAGUNT (200-175 BC) AND THE DESTRUCTION OF CARTHAGE (146 BC) AND NUMANTIA (133 BC) These seas were navigated by fleets that consisted of long, light ships with oarsmen, such as galleys, that could land on beaches when no ports were available. There were also heavy vessels with less crew and more carrying space, known in general as naves onerarias (ships fitted for this purpose). These naves onerarias were propelled by square sails like cistern-ships, with less than 3000 litres of capacity. Therefore those merchant ships were suited to carry large amounts of food and amphorae, with a greater risk of shipwreck if deposits broke. That is why improvements of the coasts were needed in the IInd century BC, since natural ports and bays, traditionally used as refuges for ships, were improved by specific port areas, such as Grau Vell (Sagunt), and reinforced vantage points such as Penyal d’Ifac (Calp), in order to ensure better conditions for mooring and major areas for loading and offloading cargoes.

It is even more interesting to detect that transformations are not an isolated phenomena, although pioneering in the Edetanian and Constetanian area. Sagunt, frequently dealt by historians as a hápax, cannot be so from different points of view because defensive and port infrastructures worked within its particular systems and operated in a framework of interrelated realities. According to the data available today, the Contestan coastline grosso modo, with more fortresses in the IIIrd century BC, appeared to be under Barquid control more than Edetan, in which Arse (Sagunt) did not retain clear remains of port defence before Roman intervention after the Punic War. Since more than 15 seasonal excavations have been undertaken in the Grau Vell, the ancient port of Sagunt, whose occupation began in 500 BC, becomes a

In the Iberian Peninsula, Qart Hadast (Cartagena), capital of the Barka family, with powerful walls near the Iberian Mastia, enjoyed an exceptional role as natural port (Ramallo, Ruiz 2002, 113-122), pinpointing interest in continuing the Strait route towards the east. This coastal

48

THE ATLANTIC ROUTE

Figure 2.1.1 Map of the Western Mediterranean

Lucentum (Tossal de Manises), the end of La Serreta (Alcoi, Cocentaina, Panàguila) (Olcina 2005, 147-178), La Escuera (Sant Fulgenci) (Abad, Sala 2001, 205-266), the sanctuary of Coimbra del Barranco Ancho (Jumilla) (García Cano et al. 1997, 239-256), the Iberian sanctuary of La Encarnación (Caravaca) (Ramallo, Brotons 1997, 257-268), the sanctuary of Collado de los Jardines (Sta. Elena), the temple construction of Cerro de los Santos (Montealegre), or the end of the necropolis of Albufereta (Alacant), etc. These changes suggest breaking socioeconomic networks, with parallel adaptations to the first hegemony of Rome. The situation is now difficult to understand without an innovative approach looking for interactions between indigenous communities (not only scenarios of Roman-Punic conflict) and later Roman ones.

reference for the whole gulf of Valencia. It seems that in 200-175 BC, a tower on the coast, whose basement levels include Punic coins, was responsible for keeping watch and communicated by signs to the Castell. A little later a breakwater, perpendicular to the marine current was created. This acted as a dock, with an estimated length of 130 m and width of 10 m (Aranegui, De Juan, Fernández 2005, 75-100). Such constructions cannot be justified only for fishing or local commercial activity, but they seem logical if the location is considered as a port of call along an official Roman route. Some differences can be seen between north and south of the coastline of Valencia; perhaps they provided support for C. Metelus’ conquest of the Balearic isles, where they were repelled by slings. More to the north, or south, of this point the situation over the coast was not immune to these dynamics of the hinterland connected with it. Violent clashes are recorded in different sites in the Camp de Túria, confirmed by recently (Bonet, Mata 2002), as well as, at the beginning of the IInd century BC (Kelin/Los Villares), destruction of

Imported ceramics, such as amphorae, increased in number within Iberian settlements at the beginning of the IInd century BC (Sanmartí et al. 2004). However, they retained typologies of Punic traditions destined to transport fish-sauces, mostly from Ebussus, together with 49

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

of Cronos, the commander heads (Per pl., 6-8) for the River Lixus (Taradell, 1950, 56-57; Rebuffat, 1988, 7786) without referring to the eponymous city. I do not believe this silence is justified by the lack of archaeological documentation provided by Lixus in the IVth and Vth centuries, since the city was occupied at that time (Aranegui, ed., 2005, 29-33), but rather by political reasons (Mederos, Escribano, 2000, 77-107).

the Greco-Italian amphorae and Dressel 1 for wine. This activity combined an important Italian commercial traffic with local traditions documented by Punic vessels and Iberian kalathoi (Bencivenga 1984, 20-33). Such local distribution does not appear before and it coincides with the production of workshops such as Fontscaldes (Lafuente 1992, 47-77). Therefore, it was an heterogeneous traffic, with different ports of call, that modified cargoes before reaching final destinations, something that was to become more common later.

The location of the river and its people has been debated; some scholars locate them in the region of Draa, in the south of the country, but in general, it is believed that Hannon refers to the present zone of Larache and the river Lucus. The periplus says that the people of Lixus belonged to a group of nomad shepherds who enlisted in the Carthaginian expedition as interpreters and guides, which confirms that there was a meeting point between the Mediterranean navigators and the inhabitants of the Atlantic coast of Morocco, holders of a centuries-old tradition of exploration along the southern route, who had already given a name to some of its geographical features (Per pl. 14, 16). It is logical to suppose that such a point would have some port facilities. According to the text (Per pl. 11), the island of Cerné was twelve days’ navigation away, with poor accessibility (Guerrero, e.p.), which modern investigations relate to Mogador and which had already been visited by the Phoenicians (López Pardo, 2000); considering the features of Cerné, it is unlikely to have been the operation centre for Hannon’s expedition. In this way the people of Lixus, natives of the country, are acknowledged as experts in the geography of the Moroccan Atlantic, more than any other, which makes them, historically, inheritors of knowledge gathered from the Late Bronze Age and during the Phoenician colonization, eras with archaeological indicators in and around the city of Lixus.

In this period, production is diversified but transport is exclusively Roman. This is the reason for the facies of the seaborne supply of food for the Roman troops during the Celtiberian wars (155-133 BC), as is recorded in the material from the Numantia camps (Luik 2002). This fact reveals that the sea routes continued through river and land routes. Rome, after the creation of Roman provinces (197 BC), could neither tolerate limits in circulation nor the opposition of the Arevacos, who were unequivocally subdued after the siege of Numantia in 133 BC. But transit on inland routes may have not been suitable for normal wagons. Wheel marks from Oral (0.9-1.1 m), La Escuera (1.1-1.26 m) (both at Sant Fulgenci), Llíria (1.4 m), Sagunt (1.35 m), Meca (Ayora) (1.15 m)… present different widths, which raises the question whether the route via Heraclea (Aranegui 2006, 73-87) was suitable for standard land transport. This route was the axis of Iberian territories during the Second Punic War and may have been only suitable for horses and pack animals. The route was central in providing access to the mining district, one of Rome’s key objectives (land transport was 25 times more expensive than by sea according to Andreau 2001, 303-317) but it proved less suitable for the transportation of goods or, when required, troops. However, Roman presence did not affect all, limiting its influence to strategic facts. There was a change in the program of ethnic adscription to the territory as recognized by the destruction, abandonment or transformation of many sanctuaries from the beginning of the IInd century BC, as well as the disappearance of pebeters shaped with the head of Demeter at the same time (García Cano, Page 2004). It seems that the symbolism of the previous period changed, recognizing the importance of holy places in the way territories were organised.

This priority is confirmed by other Classical writers, who always link Lixus with the outside, rather than the other cities of western Mauritania. Pseudos Scilax (112) says that in Lixus (here a Phoenician city close to a Libyan city) wine is made and exported by the Phoenicians, and even Strabo (STR. XVII, 3, 3), who, following Posidonio, gives it territorial and commercial control of the lake (geo-morphologically well defined thanks to recent works (Carmona, 2005, 5-11), or acknowledges the hippoi as humble fishermen from Cadiz, who arrived there after two days’ navigation (STR. II, 3, 4) and had to hand over their catches to the cetariae of Lixus (Aranegui, Rodriguez, Rodrigo, 2007), the largest in the area. Something similar is pointed out by the Pseudo Aristotle, referring to the Phoenicians from Cadiz, who, navigating four days beyond the Columnas, with an east wind, arrived at a deserted and muddy coastline stocked with large and abundant tuna (De mir. Ausc. 136), although there is not in this quote any specific mention to any city.

THE MAURITENEAN ATLANTIC COAST The Hannon periplus is one of the best studied and discussed descriptions of the Atlantic coastline of Morroco (Desanges 1978, 35-85; Gómez Espelosín, 2000, 145-155; López Pardo, 1990, 59-70; Tejera, Chávez, Montesdeaoca, 2006, 35-45; Medas 2007, etc., etc.). It is accepted that the copy of the text preserved at Heidelberg transcribes the account of the Carthaginian admiral who led the reconnaissance expedition and foundation of colonies on the other side of Strait, equipped with sixty penthaconterae in the IVth century BC (Vth?, IIIrd?). After entrusting himself to the sanctuary

The religious reputation of Lixus (PLIN. Nat. V, 2; XIX, 63) is probably the most evidence in favour of its role in Atlantic navigation, both from a cultural point of view – considering the role of sanctuaries in the compilation of travel news – and from a commercial point of view,

50

THE ATLANTIC ROUTE

Figure 2.1.2 Map of the Strait of Gibraltar

monarchy in the western area of the kingdom, practising policies that would strengthen collaboration with the northern centres of the Circle of the Strait. Both Cadiz and Lixus increase their traditional production and commerce between the II and III Punic wars, and they do so in order to contribute to the interests of Rome; it is through its networks that their products are distributed, since they are necessary for the maintenance of their new dominions. This phenomenon has also been noticed in other cases, such as Ebusus, not only through the increase in the production of Punic amphorae of the low period, but also through the importance gained by the sanctuary of Cova d’Es Cuieram (Ramón, 2005, 261-313; Aranegui e.p. b), exponent of the type of strategy used by Rome for the integration of ethnic groups different to their own, in Republican times.

as they offered neutral spaces to guarantee agreements and transactions, as well as the fulfilment of ritual obligations of navigators, with all the symbolic material this provides. The sanctuary of Lixus was dedicated to Hercules-Melgart (Bonnet, 1988, 198-200), the same as in Cadiz, and, even though its exact location is not known, some sources distinguish the urban element of the altar which is often found at the outskirts of the town (Bonnet, 1992, 123-129). Archaeologically, Lixus reveals a fact that can be checked against its relevance on the Atlantic route in Republican times. The city was redesigned between 200/175 BC (Bonet et al., 2005, 87-153), at the beginning of its Mauritanian cultural phase. Through a generalised reconstruction, the scope of possibilities for building increase dramatically, reflected in the large number of commercial containers found in the excavations. This demographic and economic increase can be linked on one hand to Cadiz, Roman from the year 206 BC, as it coincides with Italic materials from the beginning of the II century BC, which must have come this way, and, on the other hand, to the interest shown by the Mauritanian

There are many signs that indicate increasing Roman explorations of the Atlantic after the fall of Carthage (146 BC) (Morais e.p.), trying to consolidate for themselves several routes that had been out of their control until then, with the help of local experts. The expedition Aemilianus Scipio commissioned to Polibius (PLIN Nat. V, 9-10) to

51

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

large water tanks and small compartments accessed from a corridor and related to the needs generated by port traffic. The displacement of a service construction like the one mentioned above, besides obeying the functional reorganization of the urban sectors, could be due to the redevelopment of a new port under the western slope of the city for boats of deeper draught than those that would have access to the coastal lagoon where the cetariae were placed. This latter area, at just over 3 metres deep, did not allow the anchoring of ships with great tonnage that would need to draw 5 metres; a little closer to the ocean the water is deeper. This plausible enlargement of infrastructures takes us to the last stage of the Mauritanian monarchy which demonstrates, in the expedition of Juba II to the Canary Islands (PLIN. Nat. 201-205), testimony of the culmination of discoveries carried out by Polibius. The control of the peninsular Atlantic had been a requirement for the pacification of the Galician area, as indicated by the campaigns of Q. Servilio (138 BC) and I. Brutus (137-136 BC), victorious in the Lethes river (Guerra, 2000) and commemorated in Campus Martius. Also, the establishment of finis terrae, always with the naval support of the Gaditans, meant the first access to the north-western goldfields of the peninsula. Both circumstances seem to have left the exploration of Atlantic Africa temporarily on hold.

Figure 2.1.3 Sculpture of Hercules-Anteo

It is proper to remember that, some time later, Juba probably accompanied Octavious in the Cantabrian campaigns (D.C. LI, 15,6), making sure of the provisioning of the troops and making contact with their commanders. Supposing that this experience, together with the literary erudition of Juba II of Mauritania (Sirago, 1996, 303-317), had encouraged his initiative to explore the southern route again, the monarch would have organised a state enterprise supported by the port bases of the country so as to guarantee the best expeditionary apparatus with the ambition of beating all his predecessors and giving a boost to his reign, risking the advanced nautical knowledge of Augustus’ time and the expertise of navigators of the Strait circle and, particularly, the citizens of Lixus.

examine the African Atlantic coast is set within a program which, well organised technically and scientifically, with all means and equipment necessary, seemed to have reached latitudes rarely explored until then – such as the eastern Canary Islands (Medas, e.p.). At the same time, not only Lixus but also other cities around it – Kuass (Kbiri Alaoui 2007), Dchar Jdid (Kbiri Alaoui 2004, 195-214), Tamuda (El Khayari, 1996), Banasa (Arharbi, Lenoir, 2002, 1-45), Thamusida, etc. – show a development linked to the usage of these routes. And, as far as is known, Lixus retains the primacy suggested by its archaeology; its western city wall (Tarradell, 1959, 59-60; Lenoir, 1992, 289-298), its salted fish factories, and its monumental landmark of the platform culminating in the southern slope (Aranegui, e.p. b), place it in an incomparable urban category in the Moroccan scene previous to the change of era.

Without forgetting the influence of Juba on the different Mauritanian cities that claim it (Jodin, 1987) – and there are political reasons for that influence to exist – we cannot ignore the fact that the culmination of the monumental program of Lixus around the change of era is contemporary with the expedition to the Fortunate Islands. Also, the exquisite artistic and decorative objects associated with it are notorious in the Moroccan context and, in some cases, clearly refer to the Ocean and the works of Hercules in the extreme west (PLIN. Nat. XIX, 63).

Recent studies have identified warehouses of amphorae adjoining the houses between 150/130 BC and the Ist century (Aranegui, Kbiri Alaoui and Vives-Ferrándiz 2004, 366-378), as well as, and most importantly of all, public warehouses at the high area of the site. In this way we can interpret the structures of the sector known as “cámaras Montalbán” (Aranegui, Hassini and Tarradell, e.p.) within the south-eastern side of the monumental landmark, from IVth-IIIrd centuries until its transformation into a semi-basement of a residential complex at the middle of the Ist century BC. At this time a quadrangular building was erected. It was made of regular, huge ashlars (Hassini 2005, 9), a feature of the Lixus archaeological landscape from the nineteenth century, and incorporated

Some of the coins minted by Boccus and Juba II in the cities of western Mauritania, and particularly those minted with the Punic sign mqm sms, choose for their reverses a representation of Baal Amón, identifiable with 52

THE ATLANTIC ROUTE

The first Roman task was directed towards taking control of the coast and the ways of communication; maintaining the preceding demographic patterns but changing the centrality of some oppida by a process that hierarchized the functions of the city and its territory according to their own thinking, and which finds its best examples in the defensive architecture and port facilities, both prerogatives of an urban rank.

oceans, which is not common in the area, keeping Punic toponyms and terms associated to them. Avoiding the controversy of whether lks and mqm sms correspond to a double city (Mazard, 1955), to two different and distant cities (Alexandropoulos, 2000, 426), or else to a city with a royal mint that coins different signs (Manfredi, 1993, 95-102), in which case the ‘who’ would be more important than the ‘where’, the truth is that the reference to a divinity of the sea fits with the Mauritanian Atlantic vocation made specific in the expeditionary ability of Juba II. The legend surrounding the life and work of this monarch has put a question mark before many of his actions, but, if there is an enterprise, out of the many attributed to him, which deserves credibility it is the maritime one, as the undertaking of an intellectual and vital political ambition which places him among the great discoverers of the world. It seems logical for that expedition to have taken place relying on the most venerable city in all of Mauritania in terms of navigational expertise.

In the religious aspects archaeologically documented, a duality was established without creating a dilemma. While Roman foundations followed plans appropriate to those times, such as the case in the surroundings of Valencia or of Pollentia (Alcúdia) in Mallorca, the indigenous groups witnessed the introduction of constructive typologies unheard of in traditional sacred places, and with ornaments which, in many cases, tried to reflect local subjects, as can be seen in some late votive offerings of bronze, limestone, ceramics and funerary monuments: all examples of cultural hybridization. The memories of autochthonous roots were already very altered in the IInd and Ist centuries BC, and, therefore, when relatively unskilled craftsmen recreated such paintings, ceramics or small statues, themes and compositions appear which are difficult to associate with previous centuries, although it is likely they were added to allow local minorities to evoke feelings of belonging. This is important because it suggests a consent to change by people of different countries, essential, for instance, for the articulation of Hispania. This harmony would be, in my opinion, the best contribution of the Latin Sea to our historic sources.

A CHANGE THAT CONTEMPLATES THE FEELING OF BELONGING The penetration of Roman culture in the west reached, in the II Punic War, a decisive milestone. Despite having been a conflict in which the Iberians played only a secondary role, the war unleashed retaliation and punishment, bearing in mind that Carthage remained active until 146 BC, and it was necessary to strongly impede their rearmament.

53

THE ATLANTIC ROUTE

2.2 FORMAL ROMANISATION AND THE ATLANTIC PROJECTION OF AMPHORAE FROM THE GUADALQUIVIR VALLEY1 E. García Vargas Fernández Cacho, 2005; Bernal & Giménez Camino). In both areas, amphorae inscriptions show that these phenomena are nothing but a material reflection of a more general trend towards the introduction of romanised systems for pottery production and management, at least from the second third of the Ist century BC.

ROMANISATION OF THE PRODUCTION RELATIONS AND AMPHORAE DISTRIBUTION IN THE GUADALQUIVIR VALLEY Throughout the Ist century BC and the early years of the Ist century AD, the SW of the Iberian Peninsula witnessed the emergence of a regional amphorae repertoire, including containers for olive oil, wine and salted products, of Roman, or a strongly romanised, typology (García Vargas, 1998; 2001). During this period, the pottery workshops on the Atlantic coast, the Strait of Gibraltar and the mouth of the Guadalquivir, experienced a significant growth in size and number, showing evidence, or an important reorganization of the production relations to suit the new economic needs. This is the crucial period for the romanisation of the economic structures in the region and, although this formal and structural process of romanisation extends until the final decades of the Ist century AD, the morphologic repertoire of Baetican amphorae had already taken its final shape by the final years of the Augustan period or the early years of the Julio-Claudian dynasty (García Vargas, forth-coming). 1

In the wide estuary of the river Baetis, this romanisation of pottery production becomes clear only slightly later. The marks printed over tegulae give evidence of the spread, from the coast to the Guadalquivir valley inland, of Roman building techniques during the civil wars and immediately after (Chic & García Vargas, 2004). The production of Roman-type amphorae intensifies in this same region over the last quarter of the Ist century BC, after the formative period of the Republican repertoire, dated after the second third of the century. A recent reassessment of the amphoric material from the Republican period found in Seville (Colonia Iulia Romula Hispalis), also points to an early romanisation of amphorae distribution relations from the mid Ist century BC (García Vargas, forthcoming).

In the Bay of Cadiz, where imitation of Italian amphorae shapes begins in the early IInd century BC (García Vargas, 1996; 1998; Bustamante & Martín Arroyo, 2004), we can observe a generalized relocation of pottery workshops with the earliest evidence for rural pottery making (García Vargas 1998; Lagóstena, 1996) spreading throughout the extensive hinterland of the bay, and which shows an increasingly romanised amphora shapes repertoire (García Vargas, 1996; 1998).

For a long time, we lacked any archaeological evidence for these processes in the estuary of the Guadalquivir, but the urban excavations carried out in recent years in cities of the low course of the river (figure 2.2.1) have begun offering data, which systematization allows for the definition of a productive and commercial scenario which, in the late Republic and the early Augustan period, is still related, as far as pottery production goes, to the development of workshops around cities (figure 2.2.2).

In the bay of Gibraltar, this romanisation process equally brings the emergence of Italian shapes, such as Dressel 1 and 21-22, and the creation of workshops beyond the cities’ immediate area of influence (Sotomayor, 1969;

Carmo In the city of Carmona, a Roman municipium from the Augustan period and located in the plain of the Corbones river, the oldest evidence for amphorae production relates to the Turdetan pottery workshops in the “Arbollón” (Rodríguez Rodríguez, 2001), a natural stream bed at the

1 With special thanks to: D. Govantes Edwards (translation), M. Casado Ariza, J. Vázquez Paz, C. Maestre Borge and M. Furnier Pulido (materials and drawings).

55

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 2.2.1 Location of the roman towns mentioned in the text

13-15, Dr. Fleming Street (Conlin Hayes et al., forthcoming), gives evidence for the production, during the early Augustan period, of these types of amphorae (figure 2.2.3b) along with an only partially romanised formal repertoire of plain and painted wares. This is complemented with the find of an Oberaden 83 amphora with a grooved handle in example 25, Dr. Fleming Street.

NE foot of the hill were the original oppidum lay (Lineros, 2005). From the mid Ist century, pottery production concentrates at the opposite end of the city on the NW, clearly in connection with the roads running towards Hispalis, to the E (Via Augusta), and with the Guadalquivir, to the N. Evidence for pottery making has been found at least in six locations (figure 2.2.3a) within this area (Rodríguez Rodríguez, 2001), in strata dating from the last third of the Ist century BC, and which productions include building material presenting the stamp ALBI, plain pottery of Turdetan and Roman typologies and amphorae of traditional (Pellicer D) and romanised (Haltern 70) typologies. The material found in the waste area associated with the kilns found in numbers

Ilipa The ancient Ilipa, modern Alcalá del Río, and Roman municipium from the Augustan period, is located on the Guadalquivir, around 20 km. north of Seville. It is placed at the last location upstream affected by the Atlantic tides 56

THE ATLANTIC ROUTE

Figure 2.2.2 Late republican and Augustean amphorae from Guadalquivir Valley: 1. Dressel 1 A; 2. Dressel 1 B-C; 3. Dressel 2-4; 4. LC 67; 5. Haltern 70; 6. Urceus typus; 7. Oberaden 83/Dressel 20; 8. Haltern 71/Dresel 20; 9. Pellicer D. (after Berni 1998; García Vargas and Bernal 2008 and Morais 2007)

57

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 2.2.3a: Ceramic kilnsites at Carmona (ancient Carmo) (after Rodríguez Rodríguez 2001)

Figure 2.2.3b Amphorae productions from “13-15 Dr. Fleming St.” kilnsite (Carmona) 1. Haltern 70, 2. Pellicer D. (after Conlin et al., forthcoming)

and was the uppermost point navigable by ships of a certain size in antiquity. Rescue archaeology has found the ancient city’s harbour at the eastern end of the

oppidum, outside the city walls, where the Caganchas stream joins the Guadalquivir. In this area, along the Caganchas, which runs embracing the north end of the 58

THE ATLANTIC ROUTE

Figure 2.2.4 Amphorae production (except n. 6) from Mesones st. Kilnsite: (Alcalá del Río, ancient Ilipa): 1-2. Haltern 70; 3-4. Haltern 71/Dressel 20; 5. Dressel 28; 7. Haltern 70; 8. Haltern 71/Dresel 20; 9. Dressel 2/4 Rhodian? (local imitation), 10. Haltern 70; 11. Haltern 71/Dressel 20

ancient city, at least two amphorae production areas are attested, of which the best known (Mesones Street) offers a typology of amphorae (figure 2.2.4) that can be

generically dated to the Augustan period, including Haltern 71/Dressel 20, Haltern 70, Dressel 28, and shapes related to Dressel 2-4. In other areas SE of the city, pre59

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 2.2.5 Republican deposit from Italica (Santiponce, Sevilla): 1. Locally produced Dressel 1B; 2. Locally produced Dressel 1C; 3. LC 67; 4. Class 24/Oberaden 83?; 5. Oberaden 83/Dressel 20. Sc. 1:10. (after García Fernández 2004)

centuries BC, are known. Archaeological works carried out in the Guadalquivir floodplain, to the W of the ancient city, which served as a harbour area in antiquity, found a deposit of amphorae dating from the middle of the Ist BC century, probably, a period in which the city was already a Roman municipium. This deposit includes one find of Dressel B1, without base, and another with no neck, both locally produced. In addition there was one complete Lomba do Canho 67, one neckless Type 24/ Oberaden 83 along with a complete neck from a different amphora of the same type, and one Gaditan 7.4.3.3

Augustan consumption contexts predominantly bear local Haltern 70, also including 7.4.3.3 type, from Cadiz, local LC 67, Oberaden 83/Dressel 20, and other small amphorae similar to the urceus type (Cervera Pozo et al., 2007; Izquierdo Montes, 2007). Italica The earliest Roman foundation in the Iberian Península (206-205 BC) is located over a Turdetan settlement, of which several pottery kilns, dated to the IIIrd and IInd

60

THE ATLANTIC ROUTE

Figure 2.2.6 Sevilla (ancient Hispalis). Location of archaeological excavations mentionen in the text. Grey dotted line indicates the (hypothetical) Roman republican wall-circuit

without neck or mouth. The deposit gives evidence for the reception in the city during the Republican period (figure 2.2.5) of local wines and oils and also Gaditan amphorae of late-Phoenician typology containing salted products Hispalis

on the spot where seafaring gave way to river navigation. For that reason Hispalis functioned over as long period as an important trading post, an emporium (Str. 3.2.1), channelling goods from and to the valley’s hinterland. The large onenaria ships could reach it easily by sea, and therefore, in all respects it can be considered the major sea harbour in the region (figure 2.2.6).

The Caesarian colony of Hispalis was located on the Guadalquivir river, about 15 km. from the ancient mouth,

The available evidence on amphorae production in Seville is dated to the late Julio-Claudian and the Flavian 61

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 2.2.7 Amphorae from n. 25, Alemanes st. (Sevilla): Context 164, n. 1-9; context 161, n. 10-16

periods (García Vargas, 2001; 2003). Recent excavations have allowed for systematization of amphorae circulation in the city from the Turdetan period to Late Antiquity (García Vargas, 2007), showing a predominance of Italian (Dressel 1 and B-C, Lamblogia 2) and Gaditan (T.9.1.1.1 and T.7.4.3.3) imports, but also including local productions of Turdetan typology (Pellicer D) during the late IInd century and early Ist century BC (excavations carried out in the Abades and Argote de Molina

streets, within the oppidum limits: Jiménez Sancho et al., 2006). The mid Ist century BC excavations in the Alemanes (in the harbour area, outside the citywalls to the SW. (figure 2.2.7) and Fabiola (at the SE end of the Republican Hispalis: figure 2.2.8) streets, give evidence for amphorae of different provenances (García Vargas, forthcoming), predominantly imports from the Bay of Cadiz (types 7.4.3.3 and Dressel 7-11), the Bay of Gibraltar (Dressel 7-11 and Dressel

62

THE ATLANTIC ROUTE

Figure 2.2.8 Amphorae from n. 8. Fabiola St. (Sevilla): Level 18, n. 1; level 16, n. 2-9; level 14, n- 10-19

1c stamped S.C [.G]) and the Guadalquivir valley (Pellicer D, Haltern 70, LC 67, Type 24 (?) and Urceus type).

ATLANTIC CONNECTIONS This period of economic growth in the Guadalquivir valley has frequently been explained with the development, from the Augustan period onwards, of the Mediterranean trade routes which linked the recently created province of Hispania Ulterior Baetica with the annonary ports in southern Gallia and the Tyrrhenian coast of Italy. The amphorae dating from the JulioClaudian and, especially the Flavian periods, found in Seville do not run against this conclusion (García Vargas, 2007), for, against a background always dominated by the regional productions, only the Gaulish types G4 and 5 and, in a lesser degree, the Italian, Ibizan, and Greek productions, are present in significant proportions. But even at such a late date, the weak, but constant, presence of the Dressel 14 type, originating in the pottery

In general, the regional productions amount to around 85% of the amphorae dating later than the mid Ist century BC. This contrasts vividly with the situation during the late IInd and early Ist centuries BC in which Italian imports predominate, and in which the regional productions are almost limited to the Gaditan imports and old Turdetan productions of the Pellicer D type (García Vargas, forthcoming). Furthermore, this new scenario offers a better match to what we know about the following, late Augustan and Julio-Claidian, periods, which can now be better understood as the continuation of a process of economic growth begun much earlier.

63

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

families and their dependents (familiares, in the Roman sense). The Catabrian Wars’ veterans would join after the Augustan colonisation of the region, allowing a general rearrangement of the agricultural resources of the Guadalquivir valley, which would progressively place more weight on the production of exportable agricultural produce on the non-coastal areas of the valley, the triangle formed by Hispalis, Astigi (Écija, over the Singilis-Genil river) and Corduba.

workshops in the Tajo-Sado area, show the survival of trade links with the Lusitan coast which were apparently opened at a much earlier date. In the present volume, Rui Morais argues in favour of a true “Atlantic spirit” in the policy carried out by the early emperors. We understand, along the same lines suggested several years ago by G. Chic García, that this spirit originates in a Caesarian policy which can be traced back to the period in which the future dictator held a magistracy at the Ulterior (ca. 61 AC). It is clear, and F. Chaves Tristán’s (2005) study on the Caesarian money issues during the civil wars in the Peninsula rules out every possible doubt, that this interest in Atlantic affairs aimed at the exploitation of the Ulterior’s mining resources, of which, those located around the Tinto and Guadiana rivers seem to have become active precisely during the Caesarian or triumviral period. It is also quite clear that access to Atlantic tin appears to have had an influence in this matter, both in the west (with Caesar’s campaigns in the Lusitania) and in the east (with the role played by Cornelius Balbus the Elder in the conquest of the Gallia Comata, despite playing against the commercial interests of the Gaditan-Phoenician trading oligarchies).

Indeed, the Cantabrian Wars were the culmination of the Augustan Atlantic policy, not only because they placed the rich gold mines of the Spanish NW under Roman control, but also, as stated in the Res Gestae, because they gave the Romans a total grip over the coast. It offered a western route to the North Sea, complementing the Rhone route, in use since the foundation of Lyon in 43 BC, through which amphorae containing wine and salted products were exported from the Guadalquivir valley until the late Imperial period (Desbat & Martin-Kilcher, 1989). It does not appear that the reorganisation of the mining areas of the Spanish SW and, the reorganisation of the territorial layout and the road system in the Guadalquivir valley, and the rearrangement of the sea routes for military supplies, both from the Atlantic and the Mediterranean through the axis Rhone-Rhine, all measures carried out by Augustus, are unconnected. Although amphorae containing products coming from the Guadalquivir valley are attested in relatively distant areas of the Alto Alemtejo (Mataloto, 2002), in the Huelva mining areas (Pérez Macías & Delgado Domínguez, 2007), the pre-Augustan contexts in Tarraco (Gebelli & Díaz, 2001), Lyon (Desbat & Lemaître, 2001), or Saint Roma-en-Gall (idem) from the third quarter of the Ist century BC; these amphorae, belonging to types such as Haltern 70 and early Dressel 20 (Haltern 71/Dressel 20), are most commonly found in these areas during the late Augustan or Tiberian periods, that is, after the stabilisation of the military borders along the German limes, and after the creation (Morillo Cerdán, 2002) of several military settlements in the northern Iberian Peninsula (Lugo, Astorga) and their civilian counterparts and suppliers along the coast (Braga), and after the deductio of veteran’s colonies such as Astigi (Écija, Sevilla) or Tucci (Martos, Jaén) in the Guadalquivir valley.

Fabião (2001) has suggested that the presence of amphorae produced in the Guadalquivir in late Republican Portuguese settlements such as Mesas do Castelinho or Lomba do Canho, from the mid Ist century BC, reflects a growing process of regionalisation of military supplies in the Lusitania, favouring the economic growth of neighbouring regions. Fãbiao also wonders (2001) about the causes lying behind the early connections between the Guadalquivir valley and the Algarve, especially around the Guadiana river (Castro Marim, Silves), a phenomenon which, from our point of view, is nothing but the westernmost manifestation of the increasingly important role played by the low Guadalquivir in supplying the mining settlements of the pyritic areas in Huelva, as recently shown by Aurelio Macías and Aquilino Domínguez (2007) for Riotinto in the Augustan period. We can assume that the progressive shift of mining activities from the sub-Baetica (central Sierra Morena) towards the pyritic areas between the Tinto and the Guadiana (Chic, 2007), between the Sertorian and the Augustean periods, would give the cities in the mouth of the Guadalquivir a growing role in the supply of basic foodstuffs, temporarily undermining that of the cities in the middle Baetis valley around Corduba.

The Roman colony of Astigi (Colonia Augusta Firma) played a major role in supplying the western areas of the empire with oil. It was founded around 14 BC, on the spot where the Singilis (Genil) becomes navigable, and the Via Augusta, built around 2 BC ran through the city. The earliest attested rural pottery workshops producing early Dressel 20 (Haltern 71/Dressel 20), marked with a PSAVIT stamp, date on the Tiberian period (García Vargas, 1997; Sáez et al., 2001), and it is therefore not too risky to suggest that the peak moment for Baetic amphorae exports in the late Augustean and the Tiberian periods coincides with the emergence of rural pottery making in the region, a sector which experienced a

The fact that all settlements considered here enjoyed full rights, especially municipia, leads us to pursue the influence of the Roman elements in these cities, which seemed to have played a major role in the basic organization of agricultural land. The agricultural properties held by the urban aristocracies had not yet appeared as the foundation of the official food supply because these aristocracies were still forming out of the regional client networks around the old Italian aristocratic

64

THE ATLANTIC ROUTE

1. Formation period: between the Sertorian period and the civil wars (80/70-50/45 BC): production of containers of Turdetan tradition, of the Pellicer D type; first evidence for production of Dressel 1A-B imitations; first production in the Guadalquivir of LC 67, Haltern 70 and Type 24/Oberaden 83 amphorae (Molina, 2001; Fabião, 2001).

sustained growth, both in number of workshops and productive capacity until the late IInd century, although in these pages we will not consider its development after Tiberius’ reign due to a change in the economic conditions from the mid Ist century onwards. CONCLUSIONS

2. Consolidation period: Triumviral and early Augustean periods (45/40 BC-20/15 BC): persistence of Pellicer D and Dressel 1 B-C amphorae production; earlier mass Haltern 70, LC 67 and Oberdaden 83/Dressel 20 amphorae production, attested so far in pottery workshops around Carmo or in regional distribution contexts related to Italica, Ilipa and Hispalis. Regionalisation of consumption in the production areas; exports to the coast of the Algarve and western Sierra Morena (including the mining estates in the Alto Alemtejo), and also to the Spanish eastern coast (Molina, 2001; Pascual y Ribera, 2001) and the axis Rodan-Rhone.

During the period between the beginnings of Caesar’s Atlantic policy (ca. 61/45 AC) and the Augustan colonisation of the Guadalquivir (ca. 14/2 AC), the southern Ulterior went through a period of consolidation and early growth of a fully-romanised agricultural economy. The production and circulation of amphorae give testimony of these parallel processes of romanisation and economic growth, supported by the supply needs of the mining estates around the Tinto and Guadiana rivers and of the military detachments in the Atlantic shores, in campaign first and then as more or less stable garrisons in the northern Spanish and German limites after.

3. Expansion period: late Augustan and Tiberian period. Emergence of the earliest rural pottery workshops producing Haltern 70/Dressel 20 amphorae around the cities in the middle Guadalquivir valley and, especially in the area around the mouth of the Guadalquivir and its surrounding wetlands, with mass production of the Haltern 70 type (Carreras, 2001). Emergence and consolidation of other regional typologies: Dressel 2/4, Dressel 28, urceus type amphorae; disappearance of the types in imitation of the Dressel1 type and the Turdetan Pellicer D shapes. Massive exports towards the Atlantic, probably in relation with the supply of the military settlements in the NW of the Iberian Peninsula and the Mediterranean, the Roman civilian market and especially with military supply in the limes.

Recent excavations in cities in the low Guadalquivir valley, such as Alcalá del Río, Carmona, Itálica and Seville itself, are beginning to show material evidence for this process of economic growth and of effective romanisation, which from the point of view of amphorae production brings the consolidation of a regional repertoire based on Roman shapes. Although the record is yet scant, we think it suffices to suggest several features in the transition between the emergence of this differentiated repertoire and its consolidation and mass “international” commercialisation around the turn of the Ist century BC to the Ist century AD. In a very simplified way, and with all precautions due to the initial stage of our research, we suggest the following outline for the periodisation of amphorae production and trade in the late Republican and early Imperial Guadalquivir valley:

65

THE ATLANTIC ROUTE

2.3 ROME AND WHALE FISHING – ARCHAEOLOGICAL EVIDENCE FROM THE FRETUM GADITANUM1 D. Bernal Casasola [email protected] Facultad de Filosofía y Letras. Avda. Dr. Gómez Ulla s/n, 11003 Cádiz A DIVERSIFED MARINE ECONOMY1

(Arévalo and Bernal, 2007; Bernal and Arévalo, 2008) and Carteia and Traducta, both in the bay of Gibraltar, for Late Antiquity (Bernal et al., 2003 and 2008b). A few examples will suffice to show the potential gain derived from reviewing these rescue excavation reports; the excavation carried out by the University of Cadiz on the cetariae/preserves factories found in Iulia Traducta (Algeciras), has shown that economic activities in the Fretum Gaditanum included, among others, oysterfarming, the production of fish flour and other byproducts, made by the milling of fish bones (exclusive of fish meat milling) in rotating mills – some of them were probably even powered by water – and also the production of shellfish preserves. All this evidence dates from the second half of the Vth and the early VIth century AD (Bernal et al., 2003; Bernal, 2009, ed.), a period of alleged economic stagnation according to traditional interpretations (Bernal, 2008). Apart from this, other activities which hitherto were scarcely known in the ancient world are beginning to offer some very significant archaeological evidence, including shellfish harvesting and purple production; furthermore, the evidence also shows that all these activities were interrelated, as recently shown, at least in the IVth century AD town of Carteia, (Bernal et al., 2008 b), the site of the first textile dyers’ workshop documented to date in the Baetica (García Vargas, 2004). Also the excavation of the first “rotting” deposit of tuna known in the Mediterranean, found in Baelo Claudia, and the remains of mixed preserves, based on fish but which also incorporated land mammals and even snails, all within a IInd century BC context. In this period, the late Phoenician fisheries became heavily influenced by Italian colonization, active in this area since Carteia’s deductio in 171 BC (Bernal and Arévalo, 2008).

Archaeology offers very poor evidence of the kind of goods subject to trade in Antiquity, and the picture shown hardly corresponds with the variety of natural resources offered by the fertile lands and the rich waters of the Roman Empire (an interesting and recent overview in Curtis, 2001; Alcock, 2001). Due to the perishable nature of most of the goods, except in privileged contexts with exceptional conditions for preservation, such as Pompey/Herculaneum, Egypt or the Red Sea, the evidence is in most cases indirect, and needs to be taken in combination with the iconographic and textual evidence to allow for a basic reconstruction of trade links, sea routes and, thus, the outlines of so-called “provincial inter-dependence” during the first centuries of the empire. The reductionism forced upon us by the remarkable difficulty in recognizing these goods in the records will only be progressively overcome by the use of new techniques and thorough analysis of available ecofacts. This has been our line of research during the past few years, including first-hand study of new stratigraphic sequences and attempts to tackle the problematic issue of non-ceramic evidence in fishing sites in our area of interest, the Gibraltar Strait (Bernal, 2007). Our work is, to a substantial degree, based on the analysis of the rich unpublished material offered by rescue archaeology. This includes sites on the African shores of the strait, such as the ancient municipium of Septem Fratres (modern Ceuta)2 and especially Andalusia, including Baelo Claudia for the late republic and the early empire 1 This work is set in the framework of development of research project SAGENA (HUM-03015) supported by the Consejería de Innovación, Ciencia y Empresa de la Junta de Andalucía, and Research Group HUM-440, IV Plan Andaluz de Investigación, Desarrollo e Innovación de la Junta de Andalucía. 2 Studies made possible thanks to two research grants funded by Instituto de Estudios Ceutíes in 2006 and 2007.

An analysis of these practices often relies, due to the aforementioned scarcity of direct evidence, on mosaic iconography – especially common in Africa Proconsu67

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

observations on the animal kingdom and is of little scientific interest but enormous literary value, thanks to his characteristically Attic sarcasm (Díaz Regañón, 1984, 9-10). His texts include several references to whales and other cetaceans (IX, 49; XI, 37, XVI, 18), but their hunting is not explicitly mentioned, although these animals are described as a “terrible and invincible enemy” (IX, 49, 8-9), which could be taken as indirect evidence of their perception as potential prey.

laris – and written sources; this situation leads to excessive generalisation and, especially, the inability to account for the regional variations that we know existed between the different Mediterranean and Atlantic regions during the Roman period and Late Antiquity, beyond the international fame of garum and other derived products (muria, liquamen, allec….). Our work, which has a twofold aim, stands in this context. First, we will try to gather and show several pieces of evidence, of miscellaneous nature but mainly archaeological and zooarchaeological, regarding the hunting and economic exploitation of whales, fin whales and other cetaceans that populated the Atlantic shores and the Gibraltar strait in the Roman period. Second, we suggest that whaling and the economic exploitation of whale by-products was an important activity in the Roman period, trying to go beyond the most optimistic statements to date, which limit this activity to the exploitation of aground individuals; we suggest, therefore, that whaling began in the Fretum Gaditanum at least in the Roman period. As we will see, Spanish literature on the matter dates the origin of this activity in the Late Medieval period in the Cantabrian Sea, and its generalisation in modern times. The evidence available for these later periods is indeed far more abundant, especially documents and studies on historical retrospective (i.e. the search for the origins of whaling harbours known in the XIXth and XXth centuries). With this we aim to develop a research agenda already outlined in a previous work (Bernal, 2007, 97-99), with the addition of the new archaeological evidence available.

On the other hand, Oppian, a Greek author from Cylicia, has left us a poem called Haliéutica or On Fishing, dated between 177 and 180 AD, which is the first text which treats this matter extensively – the Spanish translation of C. Calvo Delcán (1990, 9 & 14) has been used. In the first place, there is an explicit reference to the presence of whales in the western Mediterranean shores, as shown in the following quotation: “often too – referring to sea monsters – they bring terror to ships when they meet them in the Iberian Sea in the West, where chiefly, leaving the infinite waters of the neighbouring Okeanos, they roll upon their way, like unto ships of twenty oars” (Oppian, V, 56-60). That is: major cetaceans travelled across the Fretum Gaditanum, on their entrance into the western Mediterranean waters after leaving the Atlantic. According to Rougé’s definition of the geography of the Mediterranean (1975), this Mare Ibericum broadly corresponds with the sea of Alborán, to the south of Cartagena. Regarding their capture, Oppian explicitly says that “Often also they stray and come at night to the beach where the water is deep inshore; and there one may attack them” (Oppian, V, 60-62). That would obviously be an indirect practice of economic exploitation.

WHALING IN ROME – A PROMISING YET INSUFFICIENTLY CONSIDERED FIELD

Of the long passage that Oppian dedicates to whaling (V, 114-358), we are especially interested in underlining the following:

In the following paragraphs we will display several pieces of evidence, both historical and archaeological, which show the common practice of whaling in the Roman world.

 it is a collective fishing practice, compared to the siege of a city (V, 115-120).

Oppian’s eloquent literary account. First detailed description of whaling?

 the first step of the process is the determination of the weight and size of the animal, based on the way it swims (V, 125-130).

Whaling in the Mediterranean is mentioned in several scattered passages from a variety of cultural contexts, some dating back to proto-historic chronology, as shown by the mention of a nakhiru, a cetacean offered by a group of rulers of the Phoenician coast as a gift to the Assyrian king Ashur-nasir-pal II (Giammellaro, 2004, 452).

 the fishing equipment used is explicitly described: “For these monsters, the line is fashioned of many strands of well-woven cord, so thick as the forestay of a ship, neither too large nor too small, and of a suitable length. The well-wrought hook is rough and sharp with barbs projecting alternately on either side, strong enough to take a rock and pierce a cliff... a coiled chain is cast about the butt of the dark hook – a stout chain of beaten bronze... in the midst of the chain are set round wheels close together, to stay his wild struggles and prevent him from straightway breaking the iron in his bloody agony, as he tosses in deadly pain, but let him roll and wheel in his fitful course (Oppian, V, 132-147)…. “and the fishermen, as if waging war, carry strong tridents and harpoons and heavy axes and other metal weapons”

Pliny, in his Natural History, offers rich information on cetaceans and on their presence in Gaditano oceano (IX, 5), and although references to whale hunting are minimal, the passage on Ostia proves that they were indeed captured (IX, 4-6). Unfortunately, Aelian’s On the Nature of Animals – De Natura Animalium, dated in the mid IIIrd century, also lacks explicit references to whaling. This Italian author, born in Praeneste, spent a good deal of his life in Rome, chaotically transcribing his own and other’s 68

THE ATLANTIC ROUTE

Figure 2.3.1 Idealized cetacean hunt by Basque fishermen in Greenland, performed with the kind of equipment described by Oppian, in a 1715 print (Cazeils, 2000, 52)

(V, 150-154). The specific equipment used seems to depend on the size of the animal: “for those with smaller arms, the weapons are appropriate for the prey; the cords are finer, the hooks are smaller and the bait is scarcer and, instead of goatskins, emptied dry marrows tied to the cord pull the beast’s body to the surface” (V, 352-357).

This brief outline of Oppian’s references neatly describes the traditional whaling practices so common prior to the introduction of mechanized harpoon (Cazeils, 2000): collective fishing, multiple hooks, ropes and bait, use of skins/dry pumpkins to act as floats, tying of the rope to the shore and close distance killing with all sort of weapons – harpoons, tridents, sickles and axes. The practice, therefore, involves a heavy investment in equipment, which explains the common recurrence of the topos among seamen (figure 2.3.1). The texts show that hunting beached whales near the shore was also a common practice. It should be stressed that, when referring to cetaceans, the only geographical indication made in the Halieutica is to mention the presence of these animals in the western waters of the Iberian sea, into which they occasionally entered from the Atlantic through the Gibraltar strait. That reference shows an extensive knowledge of the life cycle of these animals, as proven in recent works on the Gibraltar strait waters (García y Ocaña, 2006).

 the process includes the use of bait, as it is said that “For the fatal banquet, the hook is baited with part of a bull’s black liver or a bull’s shoulder, suitable for the guest’s jaws” (Oppian, V, 148-150).  to make the captures easier and help recover the equipment, “they [the fishermen] let go with him into the water large skins filled with human breath and fastened to the line. And he, in the agony of his pain, heeds not the hides but lightly drags them down, all unwilling and fain for the surface of the foamy sea. But when he comes to the bottom with wavering heart, he halts and spits up abundant foam” (V, 178-183).  once the prey is tired, “one of the whalers, rowing at speed, leads the vessel to land and ties the cord to a rock on the coast, returning hurriedly” (V, 224-227).

These references, dated to the late IInd century but obviously based on earlier sources (Calvo Delcán, 1990, 14-17), show that this form of whaling was already practiced in the Antonine period, and probably much earlier, as can be inferred from Pliny’s Natural History from the Ist century. Cetacean hunting in Rome must, therefore, at least date back to that period.

 when finally exhausted and pulled up by the inflated skins, the animal resurfaces, the killing begins: “then one carries the sharp pointed trident, another the spear, others the sickle. There is work for them all; they finally bludgeon the beast to surrender” (V, 255259).

Rather surprisingly, such explicit sources have been overlooked in the literature about fishing in the ancient world; to our knowledge, this issue has been subject to no specific analysis, neither in the Gibraltar strait area (AA.VV., 2006; Bernal, 2006) nor, obviously, in other areas of the Mare Nostrum or the Black Sea, in which these cetaceans were either absent or very seldom found (Bekker-Nielsen, 2005).

 sea-water is poured into the open wounds: “they [the fishermen] take water and pour it into his wounds, and the salt sets him on fire” (V, 279-282).  once the task is finished, “he is then tied and brought to land”, “and the fishermen... sing their song to speed up the oars” (V, 290-294). 69

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

First zooarchaeological evidence in Hispania: A promising beginning

late Vth and early VIth centuries AD (Bernal et al., 2003; Bernal, 2009, ed.), offering new evidence.

M. Ponsich, in the earlier references about zooarchaeological evidence, explicitly mentions whale bones among the remains found in several Roman preserves factories, but without giving specific locations (Ponsich, 1988, 39 and 43). In the case of Baelo Claudia, he mentions several vertebrae “alrededor de un pilar de mayores dimensiones y un volumen muy escaso, cosa que no se encuentra en ninguna otra factoría de salazón conocida hasta hoy (around a pillar, bigger in size, but with a smaller volume, something not found elsewhere)” (Ponsich, 1988, 39). Ponsich explicitly argues that these big pseudo-conical vats could have been used to process whale meat (1988, 40). Unfortunately, the detailed archaeological recording of the remains found in the excavation of this factory is not preserved, so these references cannot be zooarchaeologically tested today. Nevertheless, the size of the bones in question makes it likely that the identification was accurate.

The abandonment levels of the factory known as Industrial Complex I, in numbers 3-5 San Nicolás Street, provided a whole vertebra, which had been used as a butcher’s anvil, as shown by the abundant cut marks on one of its joint faces (figure 2.3.3). This find could be understood as evidence for the exploitation of cetaceans in these factories, as I argue elsewhere (Bernal, 2007, 9799), although its isolation could also mean that it was a piece brought ex profeso to be used in the fish cleaning/butchering process. Subsequent zooarchaeological analysis show that the vertebra, grosso modo 32-36 cms in diameter, belongs to a major cetacean; its fragmentary state does not allow for an accurate taxonomic identification, but size points to a major cetacean, above 8 meters long which, considering the two most common species in the Mediterranean, could be a fin whale (Balaenoptera physalus) or a sperm whale – Physeter catodon – (Morales and Roselló, 2009), both common in the Gibraltar strait (García and Ocaña, 2006, 51).

In fact, the collection in the Baelo Claudia archaeological complex includes a whale vertebra of unknown provenance which could have originated in this excavation or come from a different archaeological context. Regardless of its unknown provenance, it is a significant piece of evidence, especially so because it shows abundant cut marks in its joint faces, suggesting its probable use as a working surface (figure 2.3.2). Recent zooarchaeological analysis carried out in several contexts belonging to the industrial areas of Baelo Claudia have drawn no further evidence in this direction (Cáceres, 2007; Morales and Roselló, 2007).

Figure 2.3.3 Cetacean vertebra used as an anvil, showing cut marks on the cranial joint face, from the Industrial Complex I, numbers 3-5 San Nicolás Street, in Traducta/Algeciras

A third batch of new evidence comes from the preserves factory of Septem Fratres, on the strait’s African Shore. A recent excavation carried out in late 2006 in number 3 África Square uncovered a stratigraphic sequence running from the mid-imperial period to the present day (Bernal et al., 2007 b; Sáez et al., 2009). No material from this excavation has been published, other than the report and the cited works, but we can mention that the pre-Islamic remains belong to the major preserves factory previously identified in this area of Ceuta’s urban centre. The factory was active between the IInd and the VIth centuries AD (Bernal and Pérez, 1999), as shown by the abundant remains of fish exploitation activities, including fishing equipment, and the proximity to several salting facilities, such as those found in Gómez Marcelo Street or in Paseo de las Palmeras Street. The excavation of the significant

Figure 2.3.2 Cetacean vertebra from Baelo Claudia, showing cut marks, from which we can infer the used it was put to (courtesy of I. García Jiménez)

On the other hand, recent excavations uncovered a significant proportion of the industrial quarters of Iulia Traducta (San Nicolás Street, Algeciras) (JiménezCamino and Bernal, 2007), including five preserves factories/cetariae, active during the late republic and the

70

THE ATLANTIC ROUTE

Figure 2.3.4 Stratigraphic section (A) and view from above (B) of Trench 4, number 3, África Square, including the strata (SS.UU. 4018 and 4042) in which the burned cetacean rib (C) and other bone remains (D) were found (Illustrations from Bernal, Lorenzo, Sáez and Bustamante, 2007 b)

71

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

these two moments (IInd and late Vth/early VIth centuries), although the most likely hypothesis is that the activity took place continuously over the five centuries in between.

pre-Islamic levels of this wide sequence (figure 2.3.4, A and B) produced several contextualized bone remains that could belong to cetaceans.3 On one side, a fragment, more than 15 cm long, belonging to the joint face of a rib, was found in the Stratigraphic Unit (S.U.) 4018. The fragment shows obvious traces of thermo-alteration in its inner face (figure 2.3.4 C). S.U. 4018 is one of the latest levels within the Late Roman sequence excavated in number 3 África Square, dating to the late Vth or early VIth centuries, as shown by several shapes of African sigillatas (Hayes 99, late varieties of Hayes 61, Hayes 91 B, etc.), imported amphorae (Keay XXXV, Keay LIII, etc.), slow wheel pottery, two bronze hooks, and a significant deposit of muricidae, apparently intentionally milled, maybe evidence for a purple workshop. The stratum corresponds with the demolition of the factory’s structures and the levelling of the area. In one of the earlier levels (S.U. 4042), five further fragments of bone were recovered. They all were of considerable size, in some cases more than 10 cm in length, and were heavily weathered, none preserved any of the outer surfaces, and presented a spongy consistency, due to the extreme humidity conditions caused by a high water table, which even forced some interruptions in the excavation process. The final anatomical identification is not yet ready, but according to the preliminary assessment by the zooarchaeologists, based on dimensions, it is likely that they were vertebrae or ribs belonging to cetaceans. This context, which also includes other marine remains – especially shellfish – is dated to the IInd century AD, probably the later half, by the presence of African cooking pots (among others, Lamboglia 9 B and 10 A cooking pans, Hayes 197, etc.), and the amphoric remains (evolved Beltrán II A, Gauloise 4 and Dr. 20).

Other factories in the Gibraltar strait area have also drawn similar evidence. That is the case with the recent excavations in Manilva Castle (Málaga), where a complete vertebra, of remarkable dimensions, has been found in a mid/late imperial context. The evidence is currently under analysis.4 It is likely that the publication of the excavation results, including the find of several cetariae, related with a macellum and several housing and bathing complexes, will increase our available evidence. Outside the strait area, the pre-medieval evidence is almost inexistent, according to our bibliographical survey and consultations made with several zooarchaeologists. To our knowledge, the only occurrence comes from the Galician castro of A Lanzada, where a vertebral disc belonging to a cetacean was recently found in a preRoman context of unknown date – but probably earlier than the IInd century BC – and interpreted as evidence of secondary exploitation of a beached animal (Fernández Rodríguez, 2003, 50, foto 1). Regardless of its apparent isolation, the importance of this piece of evidence is twofold. On the one hand, it is the first evidence for cetacean exploitation (or hunting?) ever found in the northern Atlantic, far away from the Gibraltar strait. This area would later play, along with the Cantabrian area, a leading role in whaling (Cazeils, 2000). On the other hand, it is the first evidence of cetacean exploitation prior to the Roman presence, confirming the importance of the exploitation of marine resources before Italic colonization.

The outstanding importance of the excavation in number 3 África Square lies, therefore, in that it sheds light into three essential questions. First, it offers confirmation for the presence of whale bones within an industrial context of preserves processing, as shown by the spatial relationships with other cetariae septenses of similar date and by the sort of archaeological evidence found (multiple remains of marine fauna, salted products/ preserves amphorae and fishing implements): that is, its presence within this productive context is not incidental, but shows a practice of cetacean exploitation in Ceuta’s fishing industry. That would explain the traces of thermoalteration shown by the rib in a context otherwise devoid of evidence of fire and full of other unburned faunal remains: the traces are due to warming incurred in the course of food processing. Second, the detection of whale bones in two different stages of the site’s life span is particularly important. Their chronology also coincides with pivotal moments for this economic activity in Septem Fratres, between the IInd century (S.U. 4042) and approximately 500 AD (U.E. 4018): the evidence unmistakably shows that cetaceans were exploited in this small municipium of the Fretum Gaditanum at least in

Much of the evidence offered by seaside sites in Hispania’s Cantabrian and Atlantic coasts probably remains unpublished; a comprehensive review of such evidence would require a monographic study, well beyond the scope of this paper. Further archaeological evidence from the Fretum Gaditanum The presence of cetaceans in Atlantic and Mediterranean waters during Classical Antiquity can be also drawn from other indirect archaeological sources. Among them, the iconographic representation of several cetaceans on a clay disc found in Tamuda (figure 2.3.5), in Mauritania, dated between the late IIIrd and the Ist century BC, and known from old, but recently re-studied (Tarradell, 1950; Fumadó, 2006). Interpreted as a baker’s stamp, its interest lie in the idealized scene it depicts,

3 We thank Drs. M. Soriguer, J. Hernando and C. Zabala for providing us with the identification of this material, currently under study in Cádiz University, and of which only a draft publication has been produced (Soriguer, Zabala and Hernando, 2007).

4 We thank J. Suárez, from Arqueotectura, for making a visit to this recent excavation possible, and C. León, archaeologist in charge of Manilva, for showing us all these recently excavated contexts, including the one in which the cetacean remains were found.

72

THE ATLANTIC ROUTE

Figure 2.3.5 Clays disc from Tamuda, depicting a harpoon armed figure riding a hippocampus, surrounded by big sea creatures, a shark and a whale among them (Fumadó, 2006, 2013, figs. 2 y 3)

VI is dated in the late imperial period on the basis of stratigraphy, standing upon parts of Casa del Oeste (Bernal et al., 2007 a). Therefore, the chronology matches with that of the evidence listed above, dating back to the IInd century but peaking during Late Antiquity. We can even go as far as to suggest that this factory could have specialized in the processing of cetacean meat during the late empire. Regarding this hypothesis, we should keep in mind that a cetacean vertebra bearing evidence of reuse was found in this cetaria.

some character armed with a double harpoon riding a double-tailed hippocampus, according to the most recent iconographic interpretation. The animals represented, traditionally interpreted as dolphins, have been reassessed as sharks or, in the case of the larger one, a whale (Fumadó, 2006, 2014-2015), interpretations with which we agree. The lack of iconographic parallels in other Mediterranean contexts is of utmost importance, proving its singularity, especially in a geographical context such as the Mediterranean Mauritania Tingitana, in which the capture of these animals is a common activity. Its chronological relevance should also be underlined, being the earlier iconographic representation of its type in the Gibraltar area. It should be interpreted as a representation of an “heroic scene” of whaling, aimed at decorating cakes for a community of fishermen: maybe used on the occasion of a successful fishing season, or of the capture of a major cetacean, it is yet further evidence of how well known these mammals were in the strait area during Antiquity.

The only other known examples of circular salting vats in the Roman Mediterranean are in several cetariae in eastern Sicily (Portopalo and Torre Vindicari), near Syracuse (Purpura 1989, 26 and 30-31, figs. 2 and 9). They are not well known, for they remain unexcavated, so they must be handled with care (figure 2.3.6 B). Furthermore, their location in the central Mediterranean places them far away from the routes followed by cetaceans in their life cycle.

Regarding iconography, the episode of Jonah and the whale should not be forgotten. Although the theme reaches its floruit during the medieval period, and no premedieval iconographic depiction is known for the area, the story grows from ancient predecessors, particularly in the Late Roman period.

Specific fishing arts? The scarcity of bronze hami catenati and harpoons As shown in the first section, whaling was carried out with specialized equipment, of which the archaeological record bears little, at least well known, evidence. Regarding harpoons, only a handful have been published. In Hispania, the only known example of a certain size was found in San Martí de Empuries – Gerona (Castanyer, 2006, 22). It is dated to the VIth century BC, and is evidence of the use of this sort of fishing implement in pre-Roman dates (figure 2.3.7 A), although given the area where it was found, it was probably used for hunting other major species (sharks or others). No comparable specimens have been found in the area around the Gibraltar strait or the Atlantic coast. Our evidence is limited to small harpoon heads made of bone, for example, the one found in the factories in Traducta

The previously mentioned pseudo-conic salting vats should also be brought here. The only known examples have been found in Baelo Claudia (figure 2.3.6 A), and therefore, they must be regarded as an exclusive feature of preserves processing during Antiquity. Their capacity is remarkable, 16 and 18 m3 for the containers numbers 9 and 8 in the Industrial Complex VI in Baelo (Bernal et al., 2007 a, 168), almost doubling the capacity of other such facilities. As already suggested by Ponsich, that makes their potential use as salting containers for cetacean meat an appealing hypothesis for the future (1988, 40, fig. 14). This area of the Industrial Complex

73

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 2.3.6 Circular vats from the Industrial Complex I in Baelo Claudia (A), and drawings of the preserves factory at Torre Vindicari (B), south of Syracuse (Púrpura, 1989, fig. 9)

(Algeciras), of early VIth-century date (figure 2.3.7 B) and obviously too small to have been used for larger cetaceans. This kind of harpoon is paralleled elsewhere in

the Mediterranean; a similar specimen dated on the VIth/ VIIth centuries was found in Castrum Perti (Figure 2.3.7 C), the known Byzantine settlement on the Ligurian coast 74

THE ATLANTIC ROUTE

Figure 2.3.7 Fishing harpoons from the western Mediterranean A. Bronze harpoon from San Martí d’Empuries, VIth century BC (Castanyer, 2006, 22) B. Small bone harpoon from Traducta, early VIth century AD (Bernal, 2009, eds., chapter 26) C. Bone double-headed harpoon from Castrum Perti, early Byzantine period (De Vingo and Fossati, 2001, 659, fig. 95,1)

Figure 2.3.8 Chained hook from the suburbs of Pompey (A; Stefani, 1991, 14), and double anchor from Pisa’s harbour (B; Bigagli, 2000, 97, fig. 4)

in archaeological sites, as illustrated by the specimen shown in figure 2.3.8 B, recently excavated in Piazza San Rossore, in Pisa, and dated in the Ist/IInd centuries – area 2, US 78- (Bigagli, 2000, 96-97, fig. 4). This identification helps to clarify some references otherwise obscure, as the mention of iron hooks, an item almost inexistent in the record, or the reference to the chain attached to the end “of the dark hook” (Oppian, V, 135), a dark colour that could be explained by the use of iron. A future reassessment of single anchors – that is, not the threepieced classical type – found in Atlantic and Mediterranean waters around the Gibraltar strait might prove rewarding because, probably, whaling practices lie behind many of them.

(De Vingo and Fossati, 2001, 659, fig. 95, 1). No examples, unfortunately, of the large metal harpoons, as the first one shown, have been found in the Fretum Gaditanum. Regarding hooks attached to chains, mentioned by Oppian, the evidence is minimal. The known hami catenati of Roman date are very few (and none come from Hispania) including an example found in a villa in Asciutta, south of Pompey (Stefani, 1991, 14, nº 4229 a); other examples are currently under study, and remain unpublished. In any case, these hooks, of no more than 10 cm in height (figure 2.3.8 A), were obviously not used for whaling, but for smaller species such as tuna, sword-fish or even sharks, around 1 meter in length and 20/40 kilos in weight. They cannot, therefore, correspond with the hooks described by Oppian (V, 135-145), with their “terrible curve” capable of “piercing a cliff”.

Meat and what else?: the importance of cetacean byproducts in Antiquity Apart from the meat itself, cetaceans could be exploited in a variety of other ways: their fat, skin, teeth, etc., could

We think that the “hooks” mentioned by Oppian must have been very similar to the single anchor so often found

75

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 2.3.9 Hippocaustum from the preserves factory of Tahadart –A- (Ponsich, 1988, 144, fig. 76), and draining structures from Gigia’s cetaria – B- (Fernández Ochoa, 1994, 143, fig. 21), potentially linked to cetacean by-products exploitation

be used for food, fuel, or wood-working (Cazeils, 2000, 41-43). Given the state of our knowledge on whaling in Antiquity, it goes without saying that the issue of the exploitation of their by-products has been largely ignored. The exploitation of “grey amber” is well attested for the Canary Islands in historical times, at least from the XVIth century. This substance, produced in the intestines, was excreted along with undigested remains, floating

towards the beaches in considerable quantities. “Grey amber” is known to have been used for the production of perfumes – due to its own smell, and for its capacity to retain other scents. Also used were the fat, for burning, and the so-called “espermaceti”, an oily substance found in great quantities in the head of sperm whales, for oiling precision instruments (Sánchez Pinto, 2004, 214216).

76

THE ATLANTIC ROUTE

Table 2.3.1 Histogram of whale fishing evidences in the Ancient World YACIMIENTO

EVIDENCIAS

CRONOLOGÍA

A Lanzada

Disco intervertebral

¿ss. IV-III a.C.?

Tamuda

Medallón de arcilla decorado con cetáceos

III-I a.C.

–

Haliéutica de Opiano

s. II d.C.

Factoría de Septem (Plaza de África nº 3, Ceuta)

Costilla termoalterada (U.E. 4018)

s. II d.C.

Restos óseos diversos (U.E. 4042)

Finales s. V – principios s. VI

Restos óseos citados por Ponsich

¿Bajoimperial?

Vértebra del Conjunto Arqueológico

Indeterminada

Factoría del Castillo de Manilva

Una vértebra al menos

¿Bajoimperial?

Factoría de Traducta (c/ San Nicolás 3-5)

Vértebra

Finales s. V – principios s. VI

Factoría de Baelo Claudia

So far there is no evidence for their exploitation in Antiquity although, as argued elsewhere, we find it very likely (Bernal, 2007, 98-99); it is through future archaeometric residue analysis that we might achieve some progress. Why waste such valuable, and prized, fats and oils? Could the heating devices found in some cetariae in the Tingitana, for example Cotta or Tahadart, have been used to process these by-products? It would be an alternative explanation to their interpretation as facilities to obtain salt, by heating sea water (Hesnard, 1998). In historical times, the whaling factories in the Gibraltar strait were equipped with big “caldrons” for processing fats, oils, and other by-products obtained from whales (Vargas, 2005). Some of these facilities, dating to the XIXth century, have been archaeologically excavated, for example, in Australia (Jacomb, 1998). We should bear in mind that all suspensurae found in Roman cetariae – Cotta and Tahadart – are located along the Atlantic coast between Larache and Tanger, an area which is particularly relevant for whaling (figure 2.3.9 A). This interesting fact, pointed out a few years ago (Bernal, 2007, 99), should find future confirmation with new archaeological evidence.

CONCLUSIONS AND FUTURE PERSPECTIVES First, we would like to underline the importance of Oppian’s Halieutica, which would in itself be enough justification for this work: indeed, whales were subject of specialized hunting, which followed well established procedures and used specific equipment, during the Roman period. It seems safe to say that the hunting of major cetaceans was already on its way by the IInd century BC. The literary evidence, the Halieutica and the archaeological record, lets remember the thermo-altered rib attested for the IInd century in the preserves factory of Septem Fratres (Number 3 África Square excavation), seem to point in that direction. However, we still lack enough evidence as to know whether these activities began taken place earlier in time or if they actually were inaugurated during the Antonine period. The logical assumption is that this activity must have pre-dated this period. The indirect references in Pliny’s Natural History and the economic significance of fishing practices and other derived activities in the Fretum Gaditanum from the Phoenician period are in support of this idea. Indirectly, Tamuda’s discs’ iconography and the finds in A Lanzada are also supportive of the idea of cetacean hunting on the strait and the Atlantic. It seems, however, that fishing intensified on a significant degree in the strait during the IInd century, intensification which becomes especially apparent in cases such as Septem Fratres, in which a major factory is built (Bernal, 2006), remaining active until the end of the Late Antiquity. Therefore, more evidence is needed to date the beginnings of these activities which, as said before, were well on their way by the IInd century.

In addition, the excavation of the preserves factory found in Marqués Square, in Gijón, in the Cantabrian coast, and dated in the late imperial period (IIIrd-IVth centuries), uncovered an interesting network of pipes connected to a series of circular containers and a cistern (Fernández Ochoa, 1994, 26-27). This structural arrangement, to our knowledge exceptional in fish processing factories, could have been used to decant these by-products, which are produced in sufficient quantities as to justify the construction of such facilities (figure 2.3.9 B). The location of Gijón in the Cantabrian Sea makes this find even more suggestive. Notwithstanding, no cetacean remains have been found in this preserves factory (Roselló and Cañas, 1994), so no empirical evidence can be offered; nevertheless, it is possible that these large mammals were butchered on the beach, so only the meat and other semi-solid matters – skin, fat, bones, entrails, etc. – reached the factories.

It would seem that most of the evidence groups around the Late Roman period, as shown by the faunal remains in Manilva and, indirectly, by the previously mentioned evidence from Gijón, Baelo Claudia (including the dates in which the Industrial Complex VI or “factory of the pseudo-conical vats” was active), Cotta and Tahadart. The cetacean faunal remains found in Traducta and 77

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 2.3.10 Archaeological sites related to whaling in Antiquity, due to faunal evidence (1-4 and 7), facilities potentially used in whaling (3, 5, 8 and 9), and iconographic depictions of cetaceans (6): 1.- Manilva Castle; 2.- Traducta; 3.- Baelo Claudia; 4.- A Lanzada; 5.- Gijón; 6.- Tamuda; 7.- Septem Fratres; 8.- Cotta; 9.- Tahadart

Septem Fratres in late Vth-/early VIth-century contexts are especially relevant, for they show the persistence of the activity until the end of Antiquity. Thus, we must conclude that these fishing activities were carried out at least over a period of 350 years (150-500 BC), although is likely that new evidence will push back the initial date even further.

Admittedly, our survey for sites with faunal remains has not produced a long list, returning only five names (Manilva, Baelo, Septem, Traducta and A Lanzada) after almost two years of research. Although we are sure that the future will provide us with further evidence, especially around the strait and the Atlantic and Cantabrian areas, we are not very optimistic on the quantitative side, for strictly methodological reasons: the butchering of the whales must have taken place, for obvious reasons, on the beaches, so most of the remains must have been left there or buried near the shore. Only some parts would reach the factories; for example, the ventral parts or the skeleton, as illustrated by the case of Septem Fratres. This problem of archaeological “visibility” would explain the under-representation, or complete absence, of faunal remains in the Spanish preserves factories (we only know the cases of Portus Illicitanus, Baelo, Septem, Traducta, Gijón….).

Regarding the sites from which this evidence comes from, we must not forget that all of them, apart from A Lanzada and Tamuda, were preserves production centres, so the interpretation seems self-evident. The most incontrovertible evidence comes from the cetariae of Septem Fratres: a thermo-altered rib of which interpretation cannot be doubted. For the others, the only available evidence so far consists of a variety of reused vertebrae. From a geographical point of view, as shown in figure 2.3.10, the sites group mainly around the strait (nº 1-3 and 6-9), with the exception of some Atlantic locations (nº 4 and 5): that is, a distribution logically determined by the life cycle of cetaceans. In this respect, the emergence of some evidence in waters east of this area, as illustrated by the case of Manilva Castle, is interesting, for it gives archaeological support to Oppian’s mentions of “incursions” of cetaceans into the waters of the Mare Ibericum.

The reuse of vertebrae as anvils for butchering fish seems to be an exclusive feature. We know of at least two examples (Traducta y Baelo), and possibly of a third in Manilva. The shape of these large vertebrae makes them ideal for creating butchering surfaces; they were rigid but not too hard, so no damage could be done to the metal instruments used upon them, and their light weight also made them very versatile.

78

THE ATLANTIC ROUTE

seems to no coincidence that all three locations mentioned have their own preserves factory from the Roman period, Traducta for Getares (in addition, more of this facilities can be found in Caetaria, in the mouth of the Pícaro river), Septem Fratres for Benzú5 and the recently documented preserve factory/purple workshop of Metrouna, active during the IInd century, for Sidi Abdeselam del Behar (Bernal et al., 2008 a). As argued elsewhere (Bernal, 2007, 97-99), everything points to a close relationship between whale exploitation at the cetariae, in the absence of specifically devised whaling facilities in the Roman period; the finds of cetacean faunal remains in some of these factories (Baelo, Traducta and Manilva Castle) also supports the idea.

The exploitation of cetacean by-products has already been commented on; fat, oils, bones, and “grey amber/espermacite”, etc. It is difficult to dig deeper into the issue without reference to archaeometric studies, and this is one of the main topics to be addressed in the future; more evidence is needed in the form of equipment/ structures – e.g. the examples from Gijón or Cotta/Tahadart – likely to have been used for the exploitation of said by-products. Another issue to be addressed in the future is the commercial projection of these products. Whale meat was probably salted, and perhaps Baelo Claudia offers a ready example of the salting facilities (Industrial Complex VI or “factory of the “pseudo-conical” vats). How was it marketed? We think that it must have been sold in amphorae, among other kinds of containers, due to the high quantity of meat; it was, however, no delicatessen to be sold in smaller quantities. Amphorae containing cetacean meat is an appealing notion, not demonstrated so far in the absence of specific tituli picti and physical remains. Some recent readings of inscriptions written over Italian Dr. 21/22 have identified the formula “CET(vs)” followed by numbers, that is, a reference to a large-sized species and the number of slices contained (Botte, 2007, 445). Is this reference ceti – a large species of fish – pointing to cetaceans? These Dr. 21/22 were produced in a big scale in central Italy – especially in Cumae – and additionally in Sicily and, “suspiciously”, in El Rinconcillo, in Algeciras, during the late republican period and the early empire (Bernal y Jiménez-Camino, 2004). Further developments require for this sort of evidence to emerge in consumption contexts. Regarding the amphorae types employed, we favour the Beltrán IIB and Keay XVI types for the IInd/IIIrd centuries, and the Keay XIX and Almagro 51c for the IVth century, although this later type, due to a narrow neck, was more suited for fish pastes than for salsamenta. It may be worth inquiring into some Atlantic productions, “exclusively” Lusitanian, with broad necks, such as the Lusitana 8 and 9 types (Fabiao, 2008, 728, fig. 2); the former’s shoulders characteristically inscribed with numbers, maybe allusive to the content in pieces of salsamenta, flat based the later. It is an interesting suggestion which needs further evidence for confirmation. However the storage of unusual products in amphorae should not surprise us; thanks to the tituli picti, we know of oyster preserves – as indicated by Aelian in the IIIrd century – stored in tailor-made amphorae manufactured in the Danubian provinces (Dyzcek, 2008, 518, fig. 4), among other examples (Bernal, 2007).

We have already mentioned that whales would be butchered on the beaches, due to their size, before their transport to the factories. It is therefore interesting to note that the contemporary factories were always equipped with ramps, from which the captures were lifted. These ramps are well attested in two of the relevant sites from the Roman period. Traducta (Algeciras) had a ramp connecting the urban industrial quarters, located on a hill, with the embankment area of Río de la Miel. The ramp, excavated in Méndez Núñez Street, was covered in the VIIth century, and was interpreted as embankment and access to the industrial area (Bernal, Iglesias y Lorenzo, 2009). In Baelo Claudia, the harbour area has been recently interpreted as a stone ramp with wooden dykes (Alonso, Menanteau, Gracia and Ojeda, 2007), which could also have been used for partially lifting the captures to be butchered. Apart from the ramp, no other specific equipment was required for the processing of these animals, apart from an open area for butchering and caldrons; the settlements were often of a temporary nature and the structures were built with perishable materials – the only outstanding feature would be the watchtowers – as recently shown by archaeological projects carried out in Australia for XIXth and XXth century examples (Lawrence and Staniforth, 1998). These pages have set the basis for the study of whaling in the Roman Empire, an almost forgotten topic by modern economic historians. Although its historiographic origin is set in the Middle Ages, partially due to the popularity of Basque seamanship in the North Atlantic, now we know that the initial date must be pushed at least back to the Early Roman Empire. This is one of the earliest studies addressing the issue, so many topics are still awaiting further development. First, we must hope that more faunal studies will be developed in the future, for so far only Traducta has produced such evidence.

Finally, some considerations must be made bearing in mind recent traditional whaling, which in the strait area remained “officially” active until 1954. The only two whaling factories in the area were Bahía de Getares in Algeciras (“Ballenera del Estrecho”) and Benzú in Ceuta (“Industrial Marítima”), as recently shown (Vargas, 2005, 99 and 100). Additionally, some smaller factories existed in the Moroccan coast, as in the mouth of the Martil River, near Tetuan, in Sidi Abdeselam del Behar. It

5 We thank our colleague A. Bouzouggar from the INSAP of Rabat for the reference to the find of cetacean bones in some caves around the Moroccan area of Benzú Bay, in association with Neolithic remains, currently under study.

79

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

It is also too soon to draw a theory on the volume of commercialization and the real impact of these products on the Atlantic-Mediterranean trade, for we even ignore the containers in which they were traded. This issue must also be addressed in the future.

Another potential line of research must point to the extension of the chronological/cultural framework; for it is surprising that these practices did not begin in the Fretum Gaditanum until the IInd century (did they not apply in the Phoenician period?). Most of the known evidence dates in the late empire and Late Antiquity, so we could be tempted to think that the “systematic” capture of cetaceans could be a complementary measure in the face of the exhaustion of Mediterranean fishing grounds and the shortage of tunids, as some scholars have suggested for the IIIrd century onwards. We believe that the archaeological and faunal evidence available is yet too thin to support this hypothesis, although it is well worth pursuing.

Finally, this is one of the links between the area around Gades and the Fortunatae Insulae, because many of the contemporary whaling routes either whirled around the Gulf of Cadiz (between the capes of San Vicente and Cantín – Vargas, 2005), or set out for waters near the Canary Islands. Maybe, evidence for the knowledge of the Canary Islands, and even of links between them and Hispania, will follow from the study of these fishing practices.

80

THE ATLANTIC ROUTE

2.4 SEAPORTS AND FLUVIAL HARBOURS IN THE PORTUGUESE TERRITORY – THE OPTIONS FOR ANCIENT HARBOUR ACTIVITIES WITHIN A CHANGING NAUTICAL LANDSCAPE M.L. Blot MAKING A “BRIDGE” BETWEEN LAND AND UNDERWATER ARCHAEOLOGICAL DATA

PARKER’S SILENCE ABOUT THE PORTUGUESE COAST

The archaeological underwater discoveries which have occurred during the last 3 decades correspond to both maritime and fluvial contexts and deal directly with the reality of ancient navigation and transportation according to the human use of geographical shelters along the coast, as well as inside the estuaries and rivers of Portugal.

A.J. Parker’s maps showing Roman underwater archaeological sites in the Atlantic waters, from northwestern Europe to southern Spain and the Straits of Gibraltar (Parker, 1992), leave an uncomfortable silence along the coast of Portugal. Nevertheless, the western sea lanes, including the Portuguese coast, have both historical and archaeological evidence of ancient traffic (Naveiro Lopez, 1991; Wooding, 1996).

We have in mind the first harbour essays in the early Mediterranean civilisations studied by several generations of archaeologists, as well as the well-known spread of navigation in the Mediterranean basin as early as the Bronze Age.

As a matter of fact, by the time of Parker’s publication many underwater finds had already come to light, and the Portuguese inventory of underwater archaeological sites had started 7 years before, having C. Westerdhal’s published methodology and analysis as the main guidelines (Westerdhal, 1991; Blot and Blot, 1990-1992). The number of cases registered in the mentioned inventtory led us to study the precise case of the Portuguese littoral as a new territory to be explored in the field of nautical archaeology, more specifically concerning the ports and harbour conditions on our Atlantic coast.

We recently applied this same theme to the case of the Atlantic Portuguese territory, to bridge the land and water archaeological data under a common scope: to make these two territories converge to a better understanding of the human relationship with nautical spaces, which means human and environmental interactions. As a departure from such an investigation we obviously had to consider some interdisciplinary contributions, such as the ones brought about by geomorphological studies, in order to create some hypothesis of the location of ancient natural ports, considering ancient harbour activities as an independent reality from any built structures, including the material presence of ship remains. We have to deal with the opacity of these kinds of sites, and trace maritime and fluvial traffic without ships, but rebuilding instead a maritime cultural landscape which shows the links between water and land (Westerdhal, 1991, 1993; Blot and Blot, 1990-1992; Teigelake, 2003). In addition we also have to understand the modifications of the seaboard which were responsible for the different interactions between men and environment.

Through such an investigation we could recognize the importance of the ancient western landscape of the sea lanes of Portugal in terms of human contacts between land and sea, namely through the previous geomorphologic features of the estuaries, which played an important role on the development of the urban centres whose proto-shape is known as having had ancient contacts with Mediterranean cultures (Arruda, 1993, 1997a, 1997b, 1997c, 2001; Barros et al., 1993; Blot, M.L., 2004a; Pereira, 1993). From such evidence we have to admit that, as J. Wooding says (Wooding, 1996), the societies living along these western sea lanes, and those sailing from elsewhere and having chosen these port sites, did certainly find some solutions to help adapt marine

81

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

skills to the natural conditions of the local geography. The study of harbours and sea routes along the western sea lanes has to consider both marine topography and ship technology. THE ANCIENT GEOMORPHOLOGICAL CONDITIONS OF THE PORTUGUESE SEASHORE LANDSCAPE The nautical conditions in the main estuaries and in the lower fluvial water bodies were such as the last transgression phase had left the littoral and lower parts of the rivers: wide estuaries and a few littoral/marine lagoons giving to the south-western part of the Iberian Peninsula a very complex shape due to the drowned landforms – submerged valleys transformed into rias (Freitas and Andrade, 1998; Dias et al., 1997; CasteloBranco, 1957, 1989). These water bodies allowed nautical communications through fluvial penetration towards the hinterland. A good example of these conditions is suggested by Ribeiro, Lautensach and Daveau’s map concerning the paleoshapes of the estuaries and rivers (figure 2.4.1). The same is proposed by M. Priego in her work concerning the Atlantic Bronze Age (Priego, 1998) (figure 2.4.2).

Figure 2.4.2 Ancient Tagus and Mondego estuaries and their Phoenician settlements. (Bronze Age sites lie close to the upper parts of the paleogeography of the Mondego River) (Priego, 1998) HUMAN/ENVIRONMENTAL INTERACTIONS SUGGESTING SEAFARING AND HARBOUR ACTIVITIES IN THE PORTUGUESE TERRITORY In order to suggest a few reflections regarding the study of human occupation of fluvial banks and maritime coasts,1 we aim in this paper to share our understanding of human behaviour on the south-west part of the Iberian Peninsula over the last 2,800 years, dealing with the neighbourhood of the Atlantic, as well as coastal modifications, in terms of land/water connections. These human/environmental interactions within harbour activities have been a matter of scholarly conjecture, but never really put together as two different parts (land and water) of a unique human phenomenon, such as human interaction according to changes in water level, until we recently took this subject as a necessary bridge to settle certain issues between land and water archaeology (Blot, 2003). Dating from the late Iron Age (Amaro, 1993, 1995; Arruda, 1999-2000; 1993, 1997a, b, c, 2001; Barros et al., 1993; Mayet, Silva, 1993, 1994, 2001; Pereira, 1993, 1997) (figure 2.4.3), and going through Roman times as well as the medieval and modern eras (figure 2.4.5 and 2.4.6), the archaeological data found on land archaeological sites led us to explore quite a few cases of nautical and harbour activities whose archaeological

Figure 2.4.1 Ancient coast and estuaries during the maximum of the Eustatic sea-level rise (Ribeiro, Lautensach, Daveau, 1987)

1 The only roman structures found under the Portuguese sea correspond to hypothetical harbour structures on the ancient coast of the Algarve, deeply transformed by specific coastal changes (Simplício et al. 2000).

82

THE ATLANTIC ROUTE

clues lay on the fluvial and maritime beds (figure 2.4.4). Such finds reveal these activities have often stretched over more than 2 millennia. In many cases they lasted until the XVIIth and XVIIIth centuries, with fluvial transportation only coming to an end with the XIXthcentury construction of the railway network, which replaced the ancient water routes and impacting on bays for mooring, estuaries and rivers (Blot, M.L., 2003).

Figure 2.4.3 Punic importations (Fabião, 2001)

Figure 2.4.5 Fluvial routes according to medieval accounts (Priego, 1998) We have to consider that there is evidence that climate conditions, including prevailing winds on the western coast of Portugal before the Roman period, during the Roman era and even until 1000 AD, were quite different from today’s strong north wind or “nortada” (Soares, 1987). Since this publication, A. Soares’ study was recently presented by the author as an academic thesis, and it indicates how the evidence of much milder wind conditions could help us better understand these conditions of sailing in ancient times, as well as the fact that the Atlantic sea lanes had specific conditions, with tides bringing important changes in sea-level and tidal streams so different from the Mediterranean. Early mariners certainly understood very quickly the necessity of waiting for high tide before entering an estuary through the typical sandbars and shallows, so

Figure 2.4.4 Ancient lead anchor stocks as they were recorded before 1990 (Alves et al., 1988-1989) 83

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 2.4.6 Iberian mining places in ancient times (Priego, 1998)

linked to fishing activities, allowed in early times the production of salted fish (Edmondson, 1987).

characteristic of the personality of these western shores at river mouth. These mooring places, naturally sheltered by high cliffs, are still today’s deep and safe mooring sites while waiting to enter an estuary, or until the contrary winds change direction. Underwater surveys in the Berlenga waters (Blot, 2003, 2004, 2005) have shown the diachronic utilisation of such natural marine shelters as mooring places. We share J. Wooding’s opinion of how both the willpower and ability of ancient mariners sailing over long distances, even in unknown waters, cannot be underestimated (Wooding, 1996). Today we have sufficient evidence to understand how seafarers of the past were able to navigate out of sight of land, including their reliance on safe harbours (Blot, 2003a, b, c; 2004a, b, c, d). ANCIENT AUTHORS, RIVERS, AND THE ARCHAEOLOGICAL FINDS ON LAND AND UNDERWATER The navigability of the main rivers of the Iberian Peninsula (figure 2.4.7) seems to come to a definitive point of understanding from the well-studied classical texts of Pomponius Mella, Pline and Strabo (Parodi Álvarez, 2001). These authors added precious elements to our understanding of the ancient utilisation of such water routes, and their role in communication and transport, as well as the shipping of local natural resources – ores from mining activities, such as copper, silver, pewter and gold (figure 2.4.6). The development of salt exploration,

Figure 2.4.7 Navigable western Iberian rivers according to Pomponius Mella, Pliny and Strabo (apud Parodi Álvarez, 2001) 84

THE ATLANTIC ROUTE

Diogo and Alves, 1988-1989; Diogo and Cardoso, 1992, 1999; Diogo et al., 2000; Diogo and Cardoso, 2000), while other finds correspond to different kinds of imports dating from the pre-Roman period (Cardoso, 2001).

The underwater archaeological data include nautical artefacts such as lead anchor stocks (Alves et al., 19881989, 2001; Blot, J-Y., 2002b and a), showing their Mediterranean origin, and sounding leads. They also include those elements that, despite a lack of nautical evidence from Portuguese waters to date, as containers of imported and exported products, provide important information about the capacity of overseas transport, and of social, economic and cultural factors relating to marine activity. Among the underwater evidence are the Punic North African amphorae, as well as Roman examples coming from Baetica and Gallia and the Italian Peninsula, with their contents of olives, olive oil, wine, and fish products. They correspond to the types Dressel 2-4, Lusitana 2, Haltern 70, Mañá A, A4, C2, C, C2b (Diogo, 1987, 1988-89, 1992, 1999, 2000). In addition, elephant tusks (Loxodonta africana cyclotis) (Cardoso, 2001) have been found at the underwater sites of the main mooring places and inside estuarine harbours (both river and coastal seas), from north to south all along the Portuguese coast. Some of these cases seem to correspond to Roman shipwrecks. One of them, on the southern coast of Peniche, is the shipwreck site of Cortiçais (Peniche) with dates from the turn of the era, and studied by J-Y Blot (in press).

All of them suggest a well-established maritime trade network between the Mediterranean harbours and the Iberian natural ports of the Atlantic coast.

UNDERSTANDING THE VARIETY OF PAST NAUTICAL ROUTES We have seen that Portuguese territory offered a variety of circulation routes through estuaries, marine lagoons and rivers. This observation affords comparison with the assorted nautical routes studied by C. Westerdhal to establish the archaeological heritage corresponding to a cultural nautical landscape (Westerdhal, 1991, 1993, 2000). During the initial phase of our research (Blot, 2003a), previous to the discovery of a few harbour sites on the banks of the Tagus in the Lisbon urban context, we paid special consideration to the archaeological evidence of Mediterranean imports during the period corresponding to Phoenician/Punic contacts revealed by the archaeological excavations along the seaboard and estuaries of Portugal, from Castro Marim, at the mouth of the southern Guadiana river (Arruda, 2001) as far as the Mondego estuary (Pereira, 1997). All the studied cases correspond to strategic points located in the ancient features of the Portuguese coast, with special focus on the following estuaries, from north to south: Mondego (Pereira, 1997) Tagus (Amaro, 1993 e 1999; Barros, Cardoso e Sabrosa, 1993), Sado (Mayet e Silva, 1993, 1994, 2001; Paixão, 2001), Arade (Diogo and Cardoso, 1992; Diogo et al., 2000; Teichner, 1997) and Guadiana (Arruda, 2001).

We are facing evidence that implies the presence of Mediterranean ships in the western seas of Iberia since remote times and extending through the Roman period, showing imports and types of containers. All these elements suggest not only harbour activities, but also quite a long list of harbour complexes available in the antiquity and surviving until the XVIIth century, at least, and there is value in distinguishing the hierarchy of such important sites (Molina Vidal, 1997). We have studied such possibilities according to existing witnesses who suggest the ancient maritime transport of heavy stone artefacts dating from the Roman era (Maciel et al., 2002). Such experiences were the natural continuity of earlier contacts, such as pre-Roman “Orientalising” artefacts dated from the Iron Age that were directly imported from the Mediterranean Phoenician/Punic harbours and found at Portuguese archaeological sites located on estuarine areas, such as at Castro Marim (Arruda, 1997 a,b,c, 2001), Abul (Mayet and Silva, 1993, 1994, 2001), Alcácer do Sal (Paixão, 2001), Almada (Barros et al., 1993), Santarém (Arruda, 1993), Santa Olaia (Pereira, 1993, 1997), as well as beneath the historic centres of Portuguese cities such as Lisbon, Alcácer do Sal, and Porto (Amaro, 1993, 1995; Paixão, 2001; Silva, 2000).

All the mentioned sites showed evidence of nautical contacts pointing to a Mediterranean focus, using informal landing places according to their natural features (Blot, M.L., 2003; 2004). The Roman era appeared to us to have utilised the earlier attempts to establish harbouring solutions and options.

EVIDENCE OF THE MOST IMPORTANT NAUTICAL TERMINI (RIVERS AND ESTUARIES) BEHIND THE STUDY OF THE MAIN CITIES

The underwater archaeological data seems to correspond to the nautical circulation between the estuaries, where these important land sites have been excavated, and the ocean. As we have seen, the amphorae and lead anchor stocks found in rivers, as well as off the coast, include not only presumed shipwreck sites but also some important mooring sites dating from the pre-Roman period as well as the Roman era (figures 2.4.4 and 2.4.9) (Alves et al., 1988-89, 2001; Blot, J-Y, 2005; Diogo, 1987;

The main cities of the Roman occupation were established within natural lagoons, as well as on river mouths and inside the estuaries. We have also noticed that the main roman viae going from south to north were inter-fluvial routes, traced exactly in order to link these cities (figure 2.4.8), which means a mixed network of 85

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

The archaeological finds of the last five years uncovered beneath the waterfront of Lisbon show a variety of wooden solutions for harbour structures corresponding to the XVIIth and XVIIIth century. Historical geography, as well as ancient sources, can lead us to apply a similar understanding to other cases where an ancient natural harbour has left no manmade evidence.

roads combining maritime routes and oceanic termini with inner fluvial termini (Blot, 2003a).

For antiquity, we can base our understandings on the conjectures amassed from land and underwater archaeological data to provide associations with the geographical and geomorphological data that are relevant in terms of harbour activities. From the underwater archaeological discoveries made since the 1970s, recorded in the National Inventory of the Underwater and Nautical Archaeological Heritage,2 we can recognize the associations of such finds with entrances to the main rivers, as well as with estuaries and the major mooring places, such as Cape Espichel and Berlenga Island, off the coast of Peniche (figures 2.4.9B and 2.4.9C). This in turn aids interpretation of the Phoenician/Punic sites (figure 2.4.9A) located within the estuaries and close to the underwater sites that have provided finds, including Lusitanian, Mediterranean and North African amphorae (figure 2.4.9C). In addition, the city centres themselves, their development and decay, are main indicators of the navigability of rivers and estuaries and subsequent silting action (and all its consequences), as well general coastal degradation (Blot, M.L., 2003a). In cases of both urban success or decay, the archaeological evidence found beneath medieval and modern city centres suggests ancient river transportation of seafaring trade routes coming from the Mediterranean. Their termini were mainly small harbours within naturally sheltered features that offered mooring facilities as well as modest landing sites for the small beaching craft that provided the easiest links with the land for both the loading and unloading of goods. Cargoes would be transhipped to smaller natural harbours that acted as elements within complexes that could assist each other whenever silting or other factors required it (Blot, M.L., 2003a).

Figure 2.4.8 Roman viae and ancient urban centres working as nautical termini (Blot, M.L., 2003 apud Daveau, 1995)

In both cases, this network of mixed roads was mainly linking those harbours whose facilities were the main reason for the emergence of important centres.

These elements suggest an ancient economy based on both agriculture and fishing and to which sea trade was added. A special role was given to the seaports and fluvial harbours of our Atlantic façade during Roman times (Naveiro Lopez, 1991); Roman sea trade injected new life to earlier natural harbours, creating centres that became, in turn, new commercial destinations (Blot, M.L., 2003a). We might suggest an overlap of the location of such commercial termini and natural nautical conditions as they were still available in the XIXth century, as shown in figure 2.4.10.

One good example, among others that fit our research (Christophersen, 1999), illustrating this hypothesis is Lisbon. The geoarchaeological data studied in the context of the understanding of the human occupation of the early Tagus estuary (Blot, J.-Y., 2004) indicate that the main ancient estuarine ports lay buried beneath the medieval and modern city centre, under the main streets of downtown Lisbon, as well as under today’s urban waterfront, and in places corresponding to the former fluvial beach.

2

Divisão de Arqueologia Náutica e Subaquática – IGESPAR, Lisbon, Portugal.

86

THE ATLANTIC ROUTE

Figure 2.4.9 (from left to right): A. Phoenician/Punic importations and settlements; B. Nautical pre-roman and roman underwater finds (anchor stocks and sounding leads) corresponding to mooring sites, harbours and presumable shipwrecks; C. Pre-roman and roman amphorae found under the sea and under fluvial waters (mooring contexts and presumable shipwrecks) (Blot, M.L., 2004)

understanding of how wood is used as a specific way of building near a variety of waterfronts is essential for the comprehension of such sites, e.g. Lisbon (Santos, 2006) and similar examples in the Atlantic ports of northern Europe (Bill and Birthe, 1999).

Comparing the urban topography with the available cartography and iconography of the last four centuries, we are able to understand how long the natural coastal features have been used as informal harbours, allowing contacts between ships staying at safe mooring places and land. This situation ran parallel to artificially constructed structures, and in both cases we need to understand it within the clear specialisation of harbour areas – shipyards, dockyards, quays, installations where boats could be kept and repaired, and naval storehouses as well as commercial warehouses, etc.

Such wooden remains of nautical structures can correspond to wooden, waterfront structures required to contain and preserve the shape of the land, and wooden structures containing detritus or stones for the functions of quays, both allowing direct landing from ships. The presence of wooden artefacts can also mean ship timbers, either corresponding to abandoned or sunken ships, shipyards, or even to re-used ship timbers inserted in harbour structures (Blot, M.L., in press).

DIGGING HARBOURS BENEATH THE CITIES As a final suggestion we would like to point out some main considerations to be kept in mind for the study of archaeological data relating to harbours of the past.

Next, the specialisation of the spaces connected to past harbour activities must be carefully observed in order to understand the functions of these spaces. In addition, the interdisciplinary study of waterfront geomorphological changes, related to the archaeological data from the floodplains is an essential way to better understand human interactions with environmental changes (Angelucci, 2003; Brown, 1997).

Discounting smaller locations with no material evidence of harbour activities, such as structures, archaeologists working in cities along riversides, or other kinds of waterfronts, might expect to find evidence of ancient wooden structures much more often than stone ones. An

87

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 2.4.10 Seaports and fluvial harbours in the ancient Southwestern Iberia (Portugal) overlapping Justino’s cartography of the navigable parts of the Portuguese rivers showing the regional nautical transportation still active in the XIXth century (Blot, 2003a apud Justino, 1988): 1 Atrium (Viana do Castelo); 2 Portus; 3 Talabriga; 4 Aeminium (Coimbra); 5 Eburobritium; 6 Sellium (Tomar); 7 Scallabis (Santarém); 8 Olissipo (Lisboa); 9 Caetobriga (Setúbal) 10 Tróia; 11 Salacia (Alcácer do Sal); 12 Sines (seaport of Mirobriga); 13 Lacobriga (Lagos); 14 Ipses (Alvor); 15 Arade estuary (P. Hannibalis?); 16 Silves; 17 Ossonoba (Faro); 18 Balsa (T. de Ares/Tavira); 19 Baesuris (Castro Marim); 20 Myrtilis (Mértola)

The continuity of long-distance seaborne traffic and commerce received an undeniable stimulus during the Roman presence. Such evidence crossed the following centuries, going sometimes through periods of decline,

but still active. This remains the explanation for the success of most Portuguese cities (figure 2.4.11), even if for anthropic interactions many of them decreased as their former nautical facilities disappeared.

88

THE ATLANTIC ROUTE

Figure 2.4.11 Seaports and fluvial harbours in the ancient Southwestern Iberia studied as city contexts-containing archaeological memories of the corresponding harbours (Blot, M.L., 2004a)

89

THE ATLANTIC ROUTE

2.5 LOOKING TOWARDS THE NORTH R. Morais Strabo’s reference (STR. 3.5.11) to the existence of a port that obtained tin all the way from the Cassiterides, seems to be corroborated by later references from the Visigothic Age that show that there were intense relations between Ireland, Britannia and Hispania, or between the latter and the Gallias, that must naturally go back to pre-history (apud. Balil, 1968 p. 336, notes 61 and 62).

The Atlantic Route was situated on one of the natural circuits of navigation and, consequently, in an area of confluence of cultures from the Bronze Age up to preRoman times. To prove it we have Hanno and Himilco’s circumnavigation, which, from Cádiz, marked the North and South routes of the Atlantic, in search of gold, ivory, cupper, tin and amber. This last circumnavigation has been interpreted as representative of the reopening of the tin route up to the Oestrimnides or Cassiterides (Tin Islands), with the purpose of renovating its trade, which had been under the control of native groups since the Late Bronze Age and had thus been interrupted in the VIth century BC (López Castro, 1995, p. 72). The Oestrimnides or Cassiterides are frequently placed off the coast of Gallicia, in Brittany or even in the British Isles (PLIN. Nat. 4. 119). Regardless, it is likely that the Tartassian from Huelva developed regular contacts by sea with the NW Peninsular during the Late Atlantic Bronze Age as the signs of Phoenician regular navigation throughout the Portuguese coast since the mid VIIth century BC seem to corroborate (Aubet, 1997, p. 251; Arruda, 2002; González Ruibal, 2004, 287-317).

From the VIth century BC onwards, trade with the Peninsula’s northern coast was definitely re-established by the Gaditanians (STR. 3.5.11) in order to obtain tin and trade it with Carthage and the Phoenicians in the ambit of Mediterranean commerce. During Roman times the maritime connection of the Northwest with the rest of the Empire’s commercial world was still essentially done from the harbour of Cádiz, which served as a link to this distant part of the Empire. This port was also connected to the supply of Britannia through the Atlantic route and to the Mediterranean basin through the port of Ostia. Since the foundation of Cádiz, this prosperous emporium attracted large quantities of products (especially tin) from the Peninsular Atlantic coast, using the route long known and already carried out by the Tartessic, from the Gallician Rias Baixas (estuaries) up to the south of England and Ireland.

The navigation in the Atlantic, from the north of Hispania up to Britannia, is also documented by other ancient travels, such as those of Pytheas, Timeo or Posidonius (IVth-Ist century BC), as the texts of Avienus and Strabo testify. The wider of these itineraries is represented by Avienus (Rufius Festus Avienus) in the IVth century AD, from a Greek text, probably written in the Ist century BC, known as Seashore. According to Schulten this journey could have been carried out around 530, between the Battle of Alalia and the first Roman Carthaginese Treaty of 509. By copying ancient sailing accounts this itinerary turns out to be extremely interesting, since it not only reveals the knowledge that the Greek sailors and explorers had of the peninsular coast but also provides geographic references to cities, ports, geographic faults and peoples that inhabited Iberia in the year 600 BC.

Actually Cádiz controlled the access to the Strait of Gibraltar and influenced a large part of the maritime, Mediterranean and Atlantic circuit. For Mª Eugenia Aubet (Aubet, 1997, 172), the control of the Gaditanian archipelago meant, among other things, direct access to one of the richest western territories in metalliferous resources, that is, a doorway to Atlantic mineral trade. In fact, it was from Cádiz that a process of Atlantic expansion was organised, whose consequence was the systematic exploitation of all types of resources (vid. Blázquez, et al., 1999, 364).

91

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

In 137 BC Brutus crosses the Douro River and enters the territory of the Gallecians up to the Minho River. In this punitive expedition the passing of the mythical Lima or Leça River, called in the Ancient Times Lethes (Guerra, XXX, 147-61), similar to the river of forgetfulness of classical mythology, became famous.

The vitality of this trade is documented in Strabo (STR. 3.5.11) when he says that before the arrival of the Romans the Gaditanians commercialized with these islands, exchanging salt, bronze utensils and pottery for furs, tin and lead. The Romans were so conscious of this region’s legendary richness in mineral resources that from early times it was coveted by Roman generals on campaign.

The classic sources state that this was a trick of the Gallaecian tribes to hold back the advance of Roman troops, however another reading can be made; the Roman troops, tired of these punitive campaigns and at the prospect of having to face alien guerrilla tactics, said that there was in those lands a ‘lethal’ river (Martínez Mera, 2001, 306, endnote 33).

THE CONTRIBUTION OF GADIR TO ROMAN CAMPAIGNS IN THE NORTHWEST FAÇADE OF THE PENINSULA Gadir’s direct intervention in the conquest of the NW Peninsular has not been specifically registered in the written sources, nor is it possible to say how and under what circumstances it contributed to the Romans. We do know, however, that such a help existed and that all the support within Gadir’s reach was given to the Romans: help in material resources, and above all, the use of its naval fleet and the supply of Roman troops.

This conflict, which must have taken place in July IIIrd 137 BC (OV. Fast. 6.669) – a date which would be celebrated in Rome with the building of a temple to Mars (Tranoy, 1981, 128) in Campus Martius two or three years later (Blanco Freixeiro, 1992, 97) – must have caused the death, according to Orosius (who certainly gives exaggerated numbers), of about 50,000 Gallaecians and made 6,000 prisoner (only 4,000 could escape). From that moment on, the Gallaecians were under the theoretical control of Hispania Ulterior.

The dependency of the city in relation to Rome was well pointed out by Cicerus (CIC. Balb.24; 39) when he tells us that, in virtue of the foedus of 206 BC, the city would have participated in numerous campaigns in favour of the Romans. According to him (CIC. Balb. 40), the city would have provided support to several Roman generals in their campaigns, quoting, in particular, those of Cipianus, Brutus, Metellus, and others that are more questionable, such as Horatians, Cassius and Crassus.

P. Licinius Crassius It is tempting to associate one of the names of debatable interpretation in Cicerus’ Pro Balbo to P. Licinius Crassius, who, as Proconsul, led a second military expedition to the northwest from 96 to 94 BC. This expedition, which came at the time of his intervention in Lusitania forbidding human sacrifices and controlling Lusitanian unrest, had a unique meaning in the context of Roman trade because it allowed the negotiatores of Rome to open new horizons in terms far of mineral exploitation in this rich area of the Peninsula (Tovar, 1975, 80-81; Tranoy, 1981, 130). In fact, as Strabo (STR. 3.5.11) points out, Crassus was responsible for the “teaching of” this route for the supply of great amounts of ore. This expedition made it possible for Crassus to get rich and from that moment onwards his family became one of the most wealthy and influential families in Rome (Martínez Mera, 2001, 307).

From this list we can easily conclude that the Gaditanians associated themselves with the Romans and had a major role in the gradual conquest of the territory. Among the different Roman parties that directly or indirectly resorted to Gaditanian support within the context of the campaigns that took place in the Atlantic façade, we should very briefly highlight the following: Q. Servilius Cipianus It is possible that the Gaditanians gave logistical and naval help to Proconsul Q. Servilius Cipianus in the context of the Lusitanian wars and in his first expeditions to Gallaecia. This Proconsul organized a joint attack against Viriathus from the south and north of the Peninsula, devastating much of the Gallaecian territory.

M. Perpena There is no direct reference regarding Gaditanian participation in the military expedition promoted by M. Perpena, probably in the year 74 BC, during the Sertorian Wars that led to the conquest of Cale and to a hypothetical uprising of the tribes of Gallaecia. It is, however, possible that the Gaditanians (or any other Phoenician-Punic city) might have cooperated in this campaign. Their valuable naval fleet was certainly indispensable for providing logistical support and transporting supplies and equipment.

D. Junius Brutus Much more certain seems to be the support given by the city of Cádiz, as a naval base, to D. Junius Brutus, the Gallaic, in his expedition along the Atlantic façade up to Gallaecia. This campaign, carried out between 138 and 136 BC, required the support of Gaditanian vessels, so that the base in Olisipo, at the mouth of the Tejo River, could be provisioned. 92

THE ATLANTIC ROUTE

places in this analysis thus serve as a guide to a much more fragmented and difficult path.

C. Julius Caesar The most decisive moment and one that unquestionably associates Gaditanian support to the Roman conquest of the NW Peninsular is directly related to the two characters that most contributed to this process – C. Julius Caesar and L. Cornelius Balbus.

MORÓN: THE POSSIBLE CAMP OF D. JUNIUS BRUTUS IN CHÕES DE ALPOMPÉ “This island is situated by the city of Móron, which rises on a hill near the river, at 500 stadia of distance from the sea, and is surrounded by a fertile region. The navigation up there is easy, even for boats of large tonnage along a good part of the way, which is then made by river boats. After Móron the navigation is even longer. Brutus, named the Callaicus, used this city as a basis for his operations when he fought and subdued the Lusitanian. Then he walled in Olisypón, on the banks of the river, to have free access to navigation and provisions. These are also the largest cities that you can find by the Tejo River.” (STR. 3.3.1).

They met for the first time in 68 BC when Caesar first visited Hispania as a quaestor of the propraetor Caius Antistio Vetus. Later, around May 15th 61 BC, Caesar disembarks in Corduba (Ferreiro, 1988, 363) as a proconsul, during the rule of the Hispania Ulterior, after a long voyage from Rome that followed the continental route (PLU. Caes. 11.3 and following). There he prepares his troops and around June 10th, or perhaps a little earlier, sets off with his legions to Lusitania. However he felt that controlling banditry in Lusitania was not enough, so he embarked on a war (that ends before the spring of 60 BC) and expanded it up to the north of the Douro River.

This passage by Strabo is perhaps one of the most important references about the military garrisons used in the first military campaigns in the extreme west of the Peninsula.

This risky expedition was vital to fulfil Caesar’s wishes: the army hailed him imperator and he could thus count on the solid support of his military. The numerous spoils of war were generously divided among his soldiers, but he did not omit to set aside a certain amount to pay his debts and another large sum for the Treasury in Rome to justify the expenses of his war.

The identification of Chões de Alpompé (Vale da Figueira Santarém) with the city of Móron by A. Girão and Bairrão Oleiro (Girão, Oleiro, 1953), based on its geographical context and on the gathering of Campanian A wares, has been gaining support among researchers (Fabião, 1989). Countless excavations carried out in the past few years have also contributed by revealing a considerable amount of material that demonstrates their antiquity, in particular ceramic ware of continental shape from the Iron Age together with imported Campanian A ware fragments, and also very significant and heterogeneous assemblage of amphorae of different types and fabrics (vd Zbyszewski, Ferreira, Santos, 1968, 4957; Diogo, 1982; Fabião, 1989; Diogo and Trindade, 1993-94, 263-81).

THE DATA OF THE MATERIAL CULTURE – SOME PLACES REFERRED TO IN THE WRITTEN SOURCES The support of the city of Cádiz was fundamental, as we can infer from the sources regarding the campaigns carried out by the Romans along the Atlantic façade from the last third of the IInd century BC onwards, and which affected directly the territories north of the Douro River. As already said, the direct intervention of the Gaditanian elite was continuously devalued by the Roman sources, contemporary or later, since, naturally they only give importance to Roman deeds.

Among these materials the amphorae stand out. They are mainly represented by wine amphorae from the Tirreno coast of type Graeco-Italic and of type Dressel 1, olive oil amphorae of type Brindisi and of type old Tripolitanian, amphorae of unknown content of type 4.2.2.5, and fish amphorae of Baetic production, represented by the forms Maña C2b (= Type 7.4.3.2 and Type 7.4.3.3) and Type 9.1.1.1 (Diogo, 1982; Fabião, 1989; Diogo, 1993; Diogo and Trindade, 1993-94).

However, if we take a close look at the critical analysis of the sources and compare them with the data that archaeology provides, we easily realise that Gaditanian intervention was far from limited to just occasional help or mere obligation. If we outline a simple itinerary of some places or areas referred to by the sources and then analyse the data of the material culture found by archaeology, we get a very curious picture about the evolution of this process and of its most direct participants.

Although these materials were collected in surface excavations, the consumption profile here represented is identical to the amphorae collected in Castelo de São Jorge, in Lisbon, where wine amphorae from the Tirreno coast and the Gaditanian fish amphorae stand out (Pimenta, 2005).

The places that follow do not, naturally, represent all the settlements involved in this process, because the topic is so large that it is impossible to explore it all here. The

Regardless of whether this is the Móron of Strabo or not, or at least the military garrison of Junius Brutus, what is really important is its privileged situation on a hill near

93

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 2.5.1 Map of the Mediterranean with the location of Olisipo. Places where the amphorae found in Castelo de S. Jorge came from: 1. Tirrenica Coast in the Italian Peninsula; 2. Adrian Coast; 3. City of Brindisi; 4. Tripolitania; 5. Ebusus; 6. Area of the Strait of Gibraltar (in Pimenta 2005)

the river, some 500 stadia from the sea. Such a position has determined the occupation of this place, at least over the last three centuries BC (ending probably in the mid last century), and has made it possible for the place to have enjoyed the redistribution of provisions meant for the army.

value to fully understand the trade development and the strategic importance of the city in the Republican period. We are talking about sets collected in well-dated stratified layers of the old settlement that existed there (Pimenta, 2005), or from contexts of Islamic chronology but that have in common the same representative sample of materials with similar typologies and chronologies.

FROM AVIENUS’S OPHIUSSA TO STRABO’S OLISYPÓN

The careful study and analysis of the assets/remains in stratified layers has made it possible to include them in a single stage, the third quarter of the IInd century BC (150125 BC), more precisely between 140-130 BC (Pimenta, 2005, 129).

As Guerra has emphasised in a work about the peninsula of Lisbon in the Ist millennium BC (Guerra, 2005, 11928), this region is subject to two factors – its opening to the ocean and the presence of the Tejo River. Both mean that this region had considerable potential, since the Atlantic, which is only referred to as “the ocean” in the written sources, works as a link between the extreme west and the Mediterranean world; and the Tejo River as “a major route to the inland, has the advantage of being a sea way into a vast territory” (Guerra, 2005, 122).

Besides the fine wares, mostly represented by the Campanian A wares, we should also emphasise the amphorae, represented by different productions and typologies. Despite their heterogeneity, the wine amphorae from the Tirreno coast (58%) mainly represented by the Dressel 1 amphorae (31%) are predominant, followed by the Graeco-Italic amphorae (24%) (Pimenta, 2005, 48). From this group we should likewise stress the significant presence of fish amphorae (14%), represented by imitations of Graeco-Italic amphorae, Maña C2b amphorae (= Type 7.4.3.2 and Type 7.4.3.3) and Type 9.1.1.1, with characteristics of fabric that resemble the Gaditanian productions (Pimenta, 2005, 71-79).

As far as the area of Lisbon is concerned, the studies carried out by Arruda (2002) reveal that there was an active trade with the Mediterranean, at least from the IVth century BC. The studies so far performed by Pimenta concerning the amphorae collected in Castelo de São Jorge are of great

94

THE ATLANTIC ROUTE

River (Alarcão, 1991), and it was only surpassed at a political and economic level by Olisipo (Tranoy, 1990, 20; Alarcão, 1994, 60; Mantas, 1994, 71).

The list of amphorae mentioned here is rather significant. The presence of these types of Italic origin and their association with products of Gaditanian origin should not only be regarded as a result of food supply needs. Their presence is especially important because they are part of the institutional circuits of supply of the armies on campaign (Fabião, 1989).

This city, mentioned in several classical texts (Pliny, Ptolomeus, Antonine Itinerary), but apparently ignored by Strabo, has revealed a large diachrony of occupation. Among the significant remains of public architecture that have been discovered in the highest and most protected area (Alcaçova) of the city a Roman temple dated to the late Republican period stands out; this area probably corresponds to the capitol of the city (Arruda and Viegas, 1999, 194-95; 201).

As Pimenta has duly highlighted (2005, 24-25 and 129130), the chronological context revealed in the excavations of the Castle is contemporary with the first big Roman military campaign in the extreme west of the Iberian Peninsula – the already referred to campaign of D. Junius Brutus launched in 138 BC with the aim of pacifying the last centre of Lusitanian rebellion. It is also significant that the best represented amphorae correspond to Italic and Gaditanian products. In Chões de Alpompé, the data provided by this sample corroborate Strabo’s text (STR. 3.3.1) when he mentions the help provided by Gaditanian boats to D. Junius Brutus, with the intent of supplying his garrison in the city that had been fortified precisely to receive the provisions arriving from the river.

Figure 2.5.3 Portas do Sol – Alcáçova de Santarém (in Viegas 2003)

The importance of this city is, however, unquestionably represented by the amount and diversity of materials that have been dug up. We thus know that the first occupation of the place, where the Roman city would later be established, dates to the VIIIth century BC and is certainly related to a Phoenician commercial presence in the estuary of the Tejo River (Arruda, 1993; 1994).

Figure 2.5.2 View of the entrance of the bar of the Tejo River and its estuary (in Pimenta 2005)

The material remains related to the Roman presence are contemporary with the remains registered in Chões de Alpompé and in Olisipo (especialy in the Castle’s highest and most protected area – Alcaçova). We find similar importatant material in Scallabis, together with some Campanian A wares, wine amphorae from the Tirreno coast of type Graeco-Italic and of type Dressel 1 (Fabião, 1998; Arruda and Almeida, 1999), Gaditanian fish amphorae that fit the Types 9.1.1.1 (Arruda, 2002, 210), Maña C2b (Arruda and Almeida, 1998, 206; Fig. 11, nº 99), and an imitation of Dressel 1 (Arruda and Almeida, 2000, 715, nº 17-18).

SCALLABIS PRAESIDIUM IULLIUM – AN IMPORTANT MILITARY POST IN SANTARÉM AT THE TIME OF CAESAR Following Julius Caesar’s campaigns in 61 BC, an important military post called Scallabis Praesidium Iullium was established. The strategic importance of this area, situated about 80 km of the mouth of the Tejo River – already documented by the establishment of a camp near Móron by Decimus Junius Brutus – results from the seaworthiness of this river and its sea access.

Later, around the middle Ist century BC (c. 60 to 30 BC), we witness a peak in imports, which is very likely related to the juridical status Scallabis gains during Caesar’s praetorship. These importats are represented by Campanian wares (in particular the B and the B-ovoid fabric) associated with considerable amounts of Dressel 1

This colony, founded by Caesar during the third quarter of the Ist century BC, was originally a military headquarters. Later, in the period of Augustus (Dopico, 1986), it became a caput of a conventus, whose territory would stretch southwards to the left bank of the Tejo

95

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

wine amphorae from the Tirreno coast (Arruda, Almeida, 1999, p.317-19), and amphorae of Gaditanian production of the type Maña C2b (= Type 7.4.3.2 and Type 7.4.3.3) (Arruda and Almeida, 1998, 201-31).

According to Schulten (1940, 81), one of the most emblematic episodes of Julius Caesar’s campaign would have taken place there while he was governor of Hispania Ulterior in 61 BC.

The presence of these and other amphorae – with special attention to the Haltern 70 finds (Arruda and Almeida, 2000, 703-715) – from the middle of this century and the Augustan period, together with the Italic fine wares (Arruda and Sousa, 2003, 703-715), seem to suggest that this city benefited from a privileged commercial relationship with the Baetica province, assuming that the majority of the products came, directly or indirectly, from the goods stored in the port of Cádiz. Later, from the period of Augustus onwards, these products are associated with a larger diversity of types of amphorae and Italic terra sigillata (Viegas, 2003).

The reference to these events by Dio Cassius (D.C. 37.52-55) is worth quoting (Schulten, 1940, 12-13; apud., Alarcão, 1991, 25): “Meanwhile he learned that the inhabitants of the Herminian Mountains had withdrawn and were intending to ambush him as he returned. So for the time being he withdrew by another road, but later marched against them and, being victorious, pursued them in flight to the ocean. When, however, they abandoned the mainland and crossed over to an island, he stayed where he was, for his supply of boats was not large; but he put together some rafts, by means of which he sent on a part of his army, and lost a number of men. For the man in command of them landed at a breakwater near the island and disembarked the troops, thinking they could cross over on foot, when he was forced off by the returning tide and put out to sea, leaving them in the lurch. All but one of them died bravely defending themselves; Publius Scaevius, the only one to survive, after losing his shield and receiving many wounds, leaped into the water and escaped by swimming.” (D.C. 37.52-55)

If we compare Scallabis with the rest of the sites we have been analysing, we easily realise the importance of the Gaditanians in Santarém, both in the conquest process and in the establishment of a praesidium. Their importance was also significant in its establishment as a colony and later promotion to capital of conventus. As we have seen, the Gaditanians must have had a twofold role: first they provided logistical help to the permanence and movement of the legions, and then they assumed a permanent trading role, which was secured and controlled from the city of Cádiz.

However, Caesar took his lesson from this disastrous experience. He sent messengers to Cádiz, ordering the population to send him a fleet to take his army to the island. Balbus’s firm help made it possible for the fleet, of about a hundred transportation boats, to be ready in a short time and so the indigenous populations’ resistance was promptly subdued.

PENICHE – THE OLD ISLAND BY THE “OCEAN” Thanks to the description of the location attributed to the crusader Osberno and many other documents (Calado, 1994), we know that Peniche, probably known as Londobris (PTOL. Geog. 2.5.7) or Lanobris (MARC. 3.13) (Guerra, 2005, 125), was an island in ancient times.

As Alarcão suggested (Alarcão, 1991, 26) this campaign might have been aimed at defeating the indigenous

Figure 2.5.4 Geomorphologic evolution of Peniche’s coastline (in Blot 2003) 96

THE ATLANTIC ROUTE

Figure 2.5.5 Plan of the location of the old Roman pottery of Morraçal da Ajuda in the present city of Peniche (in Cardoso e Rodrigues 2005)

from the Gaditanian area, and that also have similar parallels in specimens of Lusitanian production gathered in the present Portuguese territory and Galicia (Morais, 2003, 36-40).

populations so that the region between the Tejo and the Douro Rivers might be later occupied with stable military posts. It was also useful to demonstrate that any intervention carried out along the coast would require boat support and a detailed knowledge of the coast, which was only in the hands of the Gaditanians and other cities of “Phoenician-Punic” tradition from the southern coast of the Peninsula.

Regardless of the identification of Peniche with the island mentioned by Dio Cassius, the existence of these productions, especially the latter one, seems to document a relation with the southern peninsular world that would imply a movement of Baetic potters to the west of the Peninsula (Morais and Fabião, forthcoming).

If the identification of Peniche with these events is correct, we can understand the reason why the remains of effective Roman occupation are only documented from the Augustan period onwards. In the place known as Morraçal da Ajuda, a potter’s complex has been discovered,where a set of kilns was found which might have been used to produce amphorae and other wares (Cardoso et al., 1998, 178-79; Cardoso and Rodrigues, 2002, 6; Cardoso and Rodrigues, 2005, 83-102).

The geo-strategic and economic importance of this coastal region in the ambit of the Atlantic circulation (vd Blot, 2003, 229-231; 2004, 465-80) is suggested by the existence of the above-mentioned remains, together with a large number of leaded Roman anchors (Alves, et al., 1988-89, 120); other recent underwater works (Diogo et al., 2005; Blot et al. 2005, 31-55) and archaeological remains found (Bugalhão and Lourenço, 2005, 57-63) on the Berlenga island and on the southern coast of Peniche (Sítio dos Cortiçais) also stress the importance of the area. Here members of the team of the CNANS are studying a shipwreck of Haltern 70 amphorae dated from the Augustan period.1

Among these materials we are particularly interested in the amphorae, because they are the oldest productions so far collected in contexts of the production within the present-day Portuguese territory. These amphorae are still not fully characterised but they present a close affinity to the Baetic forms datable from the change of the era, especially with Dressel 7-11 and Haltern 70 amphorae. We are also of the opinion that we are in the presence of a production context related to the ovoid amphorae known

1 We should add a possible shipwreck off the island of Farilhão, where a Dressel 9 Baetic amphorae was salvaged.

97

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 2.5.6 Plan of Porto (1833) – Academia Real de Ciências

CALE – FROM A FRONTIER IN THE PROCESS OF CONQUEST TO A FRONTIER OF THE TRANSDURIAN PROVINCE

Nova de Gaia), on the southern bank, and the urban nucleus of Cale (present city of Porto), on the northern bank.

The conquest of Cale as a result of M. Perperna’s expedition in 74 BC and Caesar’s later expedition, that, according to Dio Cassius, went up to Gallaecia because he did not stop when he reached the Douro River but kept chasing the fleeing enemy, tell us that the Douro River was considered at the time as a frontier. There is another document, known as “The Bronze of Bembibre”, which also points to that idea. This document, written in the year 15 BC, just when the wars against the Cantabrians and the Asturian had finished, mentions a Transdurian province, which, regardless of its meaning, considers the Douro River a boundary.

The latter Cale, situated on the knoll of Oporto’s Cathedral, also called Penaventosa, and was one of the most important anchorages in the NW Peninsular. The importance of the Cale – not yet truly highlighted – has been recently recognised by the different archaeological remains documented along the right bank of the Douro River.

The Douro (Durius) River, which has the longest hydrological basin of all the Peninsular rivers (98 375 km2), could be sailed by large vessels (STR. 3.3.4) for eight hundred stadia (corresponding to about 25 leagues, just under 150 km). The first settled point of navigation could be found in the indigenous city of Acontia “… which belonged to the Vaqueus”, according to Strabo (STR. 3.2.3). This limit corresponds to Barca d’Alva, which is situated in the present border between Portugal and Spain. The rest of the distance was done in rowed barges helped by sails (APP. Hisp. 91). Around its mouth there were two large inhabited nuclei,2 the maritime enclave of Portus Cale (the present city of Vila

Figure 2.5.7 Traces of wharves of the Modern Era by the mouth of the Douro River, on Gaia’s shore (Afurada) Among these remains we should emphasize the find of funerary steles on both banks of the Douro River that led

2

Probably Porto was Cale and Gaia Portus Cale; only in Suevic times did they call Porto as Portus Cale.

98

THE ATLANTIC ROUTE

of the amphorae material) (Morais, Carreras, 2003, 36-40), associated to other amphorae, namely the Lusitanian ovoid amphorae (Morais, 2003, 36-40). The presence of these materials demonstrates once again the importance of the Atlantic route in the supply and establishment of regular contacts, especially with the Baetic province. Strong proof is the significant group of common wares and imported amphorae lids found, which have a strong affinity with the fabric of the amphorae from the Baetica.

Alain Tranoy (Tranoy, 1995) to stress the role of communication and convergence carried out by the river. A recent find (out of context) of an altar by the Gabinete de Arqueologia Urbana da Câmara Municipal do Porto (Urban Archaeology Study of Oporto City Hall) corroborates this position. The granite altar was found in one of the foundations of a proto-Roman temple at the time of the archaeological intervention carried out in 1987 in Castelo de S. João da Foz. This stele, of problematic reading as a result of severe wear and tear of the epigraphic field, might have been dedicated to water deities (Aquis Magaudiis), given its votive character documented by the final formula, L(ibens) V(otum) S(olvit) (Osório and Silva, 1994, 91; Silva, 2000, 102103).

MONTE CASTÊLO (GUIFÕES) IN THE MOUTH OF THE LÉTHES RIVER The identification of the Leça River with the famous Léthes River, recently advanced by Guerra (1996, 147161), allows us to say that D. Junius Brutus’s famous crossing of the “River of Forgetfulness” with his troops might have occurred by the mouth of that river.

Besides these epigraphic remains, which unquestionably prove the commercial importance of the Cale, we also have different traces of pre-Roman and Roman occupation found in Rua D. Hugo (nº 5), in Largo do Colégio (nº 9-12), in Rua Penaventosa (nº 45 and 39), in Rua S. Sebastião and in Casa Museu Guerra Junqueiro, as well as traces of exclusive Roman occupation found in Casa do Infante (Real, et al., 1985-86, 34; Silva, 2000, 99).

There we can find one of the most important anchorages of the NW Peninsular thanks to the existence of exterior reefs that work as natural breakwaters. These exterior reefs, presently known as Leixões in the mouth of the Leça River, are nowadays used as a port centre because the maritime coast was previously adapted to receive the present sea port.

However, among all the interventions carried out so far (vd. Silva, 1999, 138), the excavations done in the building of the old Aljube (prison), built in the midXVIIIth century, in 1749, situated in Rua S. Sebastião, are of special importance (vid. Carvalho, et al., 1996, 199203).

The characteristics of this place, with a natural shelter configuration, together with the Cavalos de Fão, in Esposende and the Islet of Caminha, were already duly stressed by J.O. Boléo in 1943, and, more recently, by Blot (2003, 48; 61). Quoting Daveau (1995), this last author (Blot, 2003, 113) refers to the importance of this place, emphasizing the transfer of the main port functions to Leixões to the detriment of the river port of the Douro. This transfer was carried out because the latter had a small estuary, an uncertain sandbar and an insecure river port in case of floods (which is contrary to current patterns). From these data we infer that the ideal port had already been planned in the rocks known as “leixões” (steep rocks), which afforded, as we have already seen, a natural shelter off the small mouth of the Leça River. Although there was no mooring place, the larger vessels must have been able to put into port in Leixões and wait for entrance to the sandbar of the Douro. The smaller vessels could unload directly onto the sand on the south bank.

Figure 2.5.8 Photography of the old building of Aljube (Porto)

According to a transcription of a work dated from 1666, published by Felgueiras in 1958 in Monografia de Matosinhos (Felgueiras, 1958, 751), we know that the Leça River was navigable at least as far as the old Guifões bridge, near the settlement of the same name:

The excavations carried out here, under Joel Cleto, revealed deposit tips from the time of the construction of this building. In them were many different Roman materials, especially thin-walled wares from Etrury, Italic terra sigillata and different types of amphorae, datable to the late Republican and Augustan periods. In this set of materials the Haltern 70 amphorae stand out, with about three hundred specimens (about 80%

“3. Navegase nos tempos antepassados da sua fóz atê a Ponte de Guifoēs, que nos fica mais assima: mas como esta passagem devassava o nosso recolhimento, a prohibiraõ os Reis. Ficou depois impedida com o assude das azenhas, que se fizeraõ abaixo, cujas

99

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 2.5.9 Limit of the basin of the Leça River

Figure 2.5.10 Photography from the early 20th century, showing the mouth of the Leça River before the construction of Leixões Seaport

condições, que nos importavaõ muito, nunqua foraõ bem guardadas. Saõ agora do Colégio da Sagrada

Companhia de Jesu no Porto, com o qual fizemos este concerto; que avendo de meter pera serviço dellas na 100

THE ATLANTIC ROUTE

When making this route, the Proconsul caught a glimpse of the Atlantic sunset. This had a strong impact on him and made him give up his campaign to the north (Schulten, FHA, IV, p. 139-40 and 334-35), into a territory that would later be described by Strabo (STR. 3.1.2) as inhospitable and “not very welcoming”.

sua calseira barco, andará nelle hŭ padre, qué nos vigie, & defenda a clausura: mas nem elle se ha mister pera isso, nem convém que os Padres a essa conta padeçaõ tanto trabalho.” (“In the old times we could navigate from its mouth up to the Bridge of Guifões, which is situated a bit further up: but, since this passage disturbed our cloister, the kings forbade it. It was later obstructed by a watermill that was built a bit further down, whose conditions, which were of our interest, were never observed. They now belong to Colégio da Sagrada Companhia de Jesus in Porto, who we did an agreement with; in case a boat needs to come this way, a priest will sail along on it to watch and defend our seclusion: but there is no need for the Priests to suffer so much work”). (Felgueiras, 1958, 751)

However Brutus must have noticed a settlement situated exactly at the mouth of the river and where we can, from the acropolis, see one of the most beautiful landscapes in the region.

Situated there is the largest of three settlements we can find on the course of this river, known as Monte Castêlo or Castro (defended hill-top settlements) of Guifões. This settlement, founded on the left bank of the Leça River, is situated very near an area that was still an estuary before the building of the Leixões Port. According to Joel Cleto and Manuel Varela (2000, 142-143), this privileged position made commercial activities there easier, since it was the main emporium for other important settlements in the basin of the river and also a “central place” for other small settlements in the surrounding area.

Figure 2.5.11 Sunset view in the mouth of the Minho River from Santa Tecla Hill

With an occupation that goes back to the Iron Age, this settlement gained special importance at the end of the Republican period and throughout the entire Imperial period. This fact is corroborated by the group of Roman materials so far identified, especially by the amphorae. Among these we should highlight the predominance of Haltern 70 amphorae, with about 230 specimens (about 75% of the total %), one of the largest concentrations in the entire NW Peninsular (Morais and Carreras, 2003, 93112). The presence, among other materials, of Campanian wares datable to the IInd century BC, and of Maná C2b amphorae (= Type 7.4.3.2), are strong evidence that this region kept regular commercial contacts with the south of the Iberian Peninsula and the Mediterranean world, even before its full integration into the Roman empire. Monte Castêlo might have thus worked as one of the main redistributive centres at a regional scale for other important settlements in the basin of the Leça River, namely the settlements of Padrão (Santo Tirso) and Citânia de Sanfins (Paços de Ferreira), the largest known settlement in the NW Peninsular.

Figure 2.5.12 Photograph of the mouth of the Minho River Knowledge of these places, and certainly known to the Gaditanians in charge of the logistics of this expedition, can take us to other interpretations of the facts besides the unquestionable lack of security of these lands situated at the end of the world. According to Ovid, D. Junius Brutus would have already fought a battle there that witnessed the deaths of 50,000 Gallaic fighters and 6,000 taken prisoner.

SANTA TECLA – A BAD OMEN GLIMPSED IN AN ATLANTIC SUNSET Despite the boldness demonstrated by D. Junius Brutus by crossing with his troops the famous “river of forgetfulness”, the expedition to the distant lands of the northwest was still a difficult psychological barrier to overcome.

Despite these numbers, certainly exaggerated, Junius Brutus had already accomplished one of the greatest military deeds in those regions and certainly gathered considerable 101

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 2.5.13 General plan of Santa Tecla

In the specific case of Santa Tecla, as well as the several fragments of these typologies deposited in the reserve of its museum, we should stress the existence of a restored specimen of type Dressel 1. This find is exhibited in the Santa Tecla museum, with a fragment of a neck and handles of the same typology, but of a fabric that can be attributed to the imitations of Gaditanian productions.

spoils of war, which would guaranteed him the nickname “Gallaic”. This triumph against the Gallaic was also celebrated in Rome with the building of a temple to Mars (Tranoy, 1981, 128) in Campus Martius two or three years later (Blanco Freixeiro, 1992, 97). The presence of D. Junius Brutus in this region, together with the Gaditanians’ logistical support or, at best, the opening of this region to the traders who followed the military campaigns, is indirectly documented in the set of materials that have been so far recovered in the coastal settlements, or by navigable rivers situated along the north-western coast of the present Portuguese territory (vd. Paiva, 1993), and even in the settlement of Santa Tecla.

As one would expect, the strategic and commercial importance of this settlement increased from the mid Ist century BC onwards, reaching its peak of imports in the Augustan period. To prove it there is a large amount of fine table wares, represented by the Italic terra sigillata and Etrurian fine wares, associated with oil-lamp fragments, glasses and coins, in comparison with the small percentage of metal productions and local wares.

In these settlements (and including Santa Tecla) the amphorae stand out both in their number and diversity. Even though these amphorae are found in smaller percentages here than in the other sites we have been analysing (situated between the Douro and the Tejo), there are also amphorae that are directly related to the military campaigns – wine amphorae from the Tirreno coast, Dressel 1 amphorae and fish amphorae from the Gadiatan bay, mainly represented by the Maña C2b forms (=Type 7.4.3.2.2 and 7.4.3.3).

However, the Haltern 70 amphorae stand out among these materials. Their minimum number of individuals of about five hundred specimens is only surpassed by the amounts gathered in the settlement of Vigo and in the Roman city of Bracara Augusta (Morais, Carreras, 2003, 93-112). Just like other places situated in the present Portuguese territory and Galicia, these amphorae are found together with the known Dressel 7-11 amphorae of Gaditanian origin and ovoid amphorae of Lusitanian fabric. 102

THE ATLANTIC ROUTE

Figure 2.5.14 Underwater findings recovered in Ria de Vigo

This commercial and military route was much helped by the Gallaic coastline, penetrated by large and deep estuaries, which enabled a thriving marine livelihood and promoted the existence of port areas and locations for the transfer of products.

ON THE WAY TO BRIGANTIUM – THE CASTRO OF VIGO AND THE BIRTH OF AN ENCLAVE BY THE “ZONA DO AREAL” (BEACH) (VICUS ELANEI?) The Gaditanians’ logistical support was certainly fundamental to the campaigns carried out by P. Licinius Crassus, M. Perperna and Julius Caesar in the NW Peninsular. Unlike the first two expeditions, Caesar’s campaigns are well documented. As we have already mentioned, Caesar made use of the Gaditanian fleet to defeat the Lusitanian and Gallaic populations in northwest Portugal. To accomplish this Caesar counted on the indispensable help of the Gaditanian Balbus, giving him the post of praefectus fabrum. But Caesar, as Dio Cassius reports, did not stop when he reached the Douro River, the real limit of the province, but crossed over, chasing those who had run away and entering Gallaic territory. The success so far reached, and with Gaditanian naval resources, led Caesar to try a new sea expedition against the people to the north of the Douro River, arriving at the western end of the peninsula of Brigantium (Betanzos, La Coruña).

In spite of not being used as ports, the same is true for the settlements situated near the coast, because they were in the proximity of quays or anchorages. One of those privileged places is the Castro of Vigo. The several excavations carried out there have revealed large amounts of imported wares and coins, which prove its importance as a place of exchange and reception of products manufactured in different parts of the Roman Empire. As one would expect, the amphorae, represented by a diversity of types and fabrics, stand out. They follow the different stages of conquest of this territory and are associated with different chronological events. The oldest amphorae, from the first military campaigns, are represented by wine amphorae from the Tirreno coast, especially type Dressel 1, and fish amphorae from the Gaditanian coast, mostly represented by the Maña C2b forms (= Type 7.4.3.2 and 7.4.3.3).

The data supplied by the sources seem to indicate that Caesar just forced the Gallaic tribes to recognise his sovereignty. In fact, as well as by the considerable amount of spoils that he got in these campaigns, the material culture gathered in the settlements by the coast show that there was a healthy trade in charge of the societates publicanorum. In other words, there were tradesmen who distributed products from Italy and South of Spain throughout the Atlantic, preceding the mechanism known, in Imperial times, as annona militaris (Carreras, 2000).

The rest of the amphorae correspond to types dated after Caesar’s campaign to Brigantium, with a peak of circulation in the Augustan period. Among these we should highlight the presence of the following: Haltern 70 and Dressel 7-11 Baetic amphorae and less typified amphorae that fit in the Lusitanian ovoid shapes (Morais, 2003, 36-40) and flat bottomed amphorae, of

103

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

el cabotaje y pesca locales? No nos muestra el mundo antiguo que los grandes faros se hallan siempre asociados a grandes puertos o, al menos, zonas de gran tráfico marítimo?. Puede creerse que la presencia continuada, al menos en la segunda mitad del s. II, de tabularii de la administración imperial en La Corunha estaría sobradamente justificada por las necesidades de la presión fiscal sobre la población de un puerto pesquero?” (Balil, 1974, 215-16).

urceus type, probably from Gadalquivir (Morais, forthcoming). As we had the opportunity of reporting (Morais and Carreras, 2003, 93-112), the Haltern 70 amphorae stand out in this set, with a minimum number of individuals of about a thousand specimens. The intensification of the urban excavations from the 1990s onwards in the city of Vigo (area of the Areal and “Casco Vello”) has revealed the existence of an important trade port that carried out a storage and redistribution function that corresponds to a Vicus Elanei (vd. Pérez Losada, 2002, 239-66). A considerable amount of amphorae, dated to the IInd century BC to the Ist AD, has been discovered here and were probably unloaded on the well-known Areal beach.

Figure 2.5.15 Plan with the location of Praia do Areal Figure 2.5.16 Location of the port enclave in La Coruña

The importance of this region was predictable because of the appearance of wares of Punic and Meridional origin (settlements of Toralla and Castriño), dated from between the Vth and IInd century BC, and because of the discovery of underwater remains – situated in the present port of Bouzas, near “punta de San Gregorio” (Coia) and in the proximities of the R.C. Nautico sports pavilion.

These questions, wisely answered by Balil by presenting a group of exogenous materials recovered up to then from different NW sites, are still current. As we have seen, regardless of the large amount of material gathered in this region, the existence of the lighthouse under discussion presumes that the Atlantic coast was part of a natural circuit of navigation and consequently an area of convergence of cultures from Bronze Age to pre-Roman times.

BRIGANTIUM’S ALTISSIMUM PHARUM In 1974, when A. Balil analysed the maritime relationships of the NW Peninsular during Imperial times, he posed a series of very important questions:

As we have also reported, this region was later coveted by Roman generals. Once again the sources tell us of D. Junius Brutus and his punitive expedition to the Northwest lands to find out about the gold exploitations of this region (STR. 3.176). What we can certainly accept

“puede creerse que una construcción como el faro de La Coruña conocido con el nombre de “Torre de Hércules” pudo nacer sin otro propósito que facilitar 104

THE ATLANTIC ROUTE

Figure 2.5.17 Reconstruction of the lighthouse in Roman times (T. Hauschild) is that from that time onwards the regions to the north of the Douro River seem to have opened themselves to trade with the south of the Peninsula and with the Roman world as we can deduce from the presence of “IberianPunic” amphorae and of type Italic Dressel 1 and GraecoItalic amphorae throughout the Atlantic façade (Pimenta, 2005, 118-24), the Galician coast, and as far as the bay of La Coruña/Brigantium (Naveiro, 1981, 117-126; 1982, 63-74; 1991, 63-66; 69). To these we should also add other types of wares and late Republican coins found in the the settlements of Sanfins and Alvarelhos (Paço and Jalhay, 1955; Torres, 1978-79, 15-250; Martins, 1990, 166), and in Alto da Cividade, in Braga (Zabaleta, 2000).

TWO STAGES OF THE SAME ITINERARY This small “itinerary” of places and areas referred to in the sources, and their contextualization according to data from the material culture that have been lately studied, is enlightening of Gaditanian importance in the conquest process of this territory, situated in the extreme west of the Peninsula. If we assess the data as a whole we can distinguish two different stages at least. The first stage The first stage, situated between the third quarter of the IInd century BC and the mid Ist century BC, is directly related to the first Italic military campaigns in the Atlantic façade and consequent hegemony of Italic trade products.

The presence of these materials is also due to the late expeditions of P. Licinius Crassus, M. Perperna, and, in particular, to Caesar’s campaign to the Northwest. According to known sources (APP. Civ. 2.8; D.C. 37.52-53; PLU. Caes.12), this campaign was mainly launched to obtain precious metals, probably gold, so as to get the necessary funds to pay his considerable debts in Rome and enrich the treasury (Blázquez, 1975, 198).

The exponential increase of Italic imports is visible, in particular, in the wine amphorae from the Tirreno coast, type Dressel 1 and Graeco-Italic, which are mostly concentrated along the coastline of the present Portuguese and Gallaic coast, especially in the estuaries of large rivers. This privileged coastal concentration is directly related to the supply circuits of the armies on campaign, which shows there is clear breaking off with pre-Roman traditions (Fabião, 1989; 1998).

As we have seen, such an expedition, carried out with the support of the Gaditanian, Balbus, and helped by the cargo vessels of that city, led Caesar up to the city of Brigantium (Betanzus – Corunha) and to the surrender and eventual submission of the native peoples of the Northwest (D.C. 37.52; 37.53.4; PLU. Caes. 12).

Together with these products, but in smaller percentages, we can also find fish amphorae and oil amphorae, even though the latter is only residually represented.

Later a lighthouse was built there known in ancient times to sailors as the “Artabrian port” or altissimum pharum (OROS. Hist. 1.2.71), which served as a lookout for vessels that went or came from Britannia.

As we have previously mentioned, fish sauces played a major role in the imports situation, even though they are numerically inferior to the Italian wine products. Such imports, represented by the “Iberian-Punic” or “neo105

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

wine amphorae, of type urceus (= similar to the type Gauloise 5 and 7), made in Gadalquivir (Morais, 2007, 401-415), while the second corresponds to the ovoid amphorae of Lusitanian fabric (Morais, 2003, 36-40).

Punic” shapes (i.e. Type 9.1.1.1. and Maña C2b) and by the Graeco-Italic imitations with a fabric that can be attributed to the Gaditanian bay, show that there was an early awakening of Hispania and of the gradual importance this region will gain in the last decades of the late-Republican period.

To this hegemony of products from the Baetica we should also add a significant number of common wares from the Gaditanian area and from Guadalquivir, which were used as a complement to the amphorae trade. The studies that have been carried out on these wares in the present Portuguese territory (Morais, 2000; Pinto, 2003; 2004; Pinto and Morais, 2007, 235-254; Arruda and Viegas, 2004, 341-49) show the importance of these materials in the context of the trade with the Baetica.

Although the presence of these materials is closely related to army supply processes, we can equally report the spread of these products in the ambit of the Italic populations who moved to the Peninsula for political and economic reasons (publicani, mercatores and negotiatores), and of the indigenous populations who gradually acquired typically Mediterranean consumption habits. The trade elites of the city of Cádiz were, however, essential, because, whether for interest or obligation, they followed the different campaigns carried out by the Romans in these territories.

Despite the cultural homogeneity in these two stages for the coastal area between the Tejo valley region and the present Coruña, we are faced with two different pictures if we take into consideration the quantitative patterns of distribution: a larger concentration of finds in the south and on the coast than in the north and inland (Fabião, 1998, 176). This phenomenon is easily predictable and explainable, because the process of conquest was carried out from south to north and throughout the Atlantic façade.

The second stage Even though the second stage is a consequence of the process of conquest and pacification of the territory, between the mid Ist century BC and the Augustan period, it represents an important turning point in the economy and trade context of the ancient world. This territory is now integrated in a market economy, though at the level of the empire, which is the result of the intensification of sea traffic and the gradual romanization of the provincial territories.

As we can confirm from the available data, the materials associated with the first stage are concentrated in the central and southern areas of the present Portuguese territory, becoming scarcer as we move north (Fabião, 1989; 1998; Paiva, 1003; Arruda and Almeida, 1998, 201-31; 1999, 307-37; Luis, 2003, 263-82; Pimenta, 2003, 341-62; 2005), and towards the Gallaic coast (Naveiro, 1991).

This phenomenon, resulting from of an economic and political decentralization, reaches its peak in the Principate of Augustus with an administrative reorganization of the provinces, a decentralization of the financial system and the growth in provincial exports. At this time, the Northwest and the Atlantic coast enjoyed a strong economic impulse, which was the result of the exploitation of mining wealth. This exploitation started immediately after the Cantabrian wars and the increase in exports of other types of products (Blázquez, 1978, 82).

We can, for instance, compare the large concentration of wine amphorae from the Tirreno coast, of type GraecoItalic (18.36%) and Dressel 1 (41.71%) collected in Castelo de S. Jorge in Lisbon (Pimenta, 2005, 115) with the small amounts of these types found at hill fort (“castrejos) settlements in the northwest (Paiva, 1993) and Gallaic regions (Naveiro, 1991). We could say the same about the amounts of Baetic fish amphorae Maña C2b (= Type 7.4.3.2 and 7.4.3.3.3) documented in Castelo de S. Jorge (18.76%), Alcaçova de Santarém (Arruda, Almeida, 1998, p. 201-31), and in the excavations of Chões de Alpompé (Fabião, 1989; Diogo and Trindade, 1993-94) in relation to the same hill forts situated in the northern area of the present Portuguese territory and Gallicia (vd. Naveiro, 1991; Paiva, 1993).

The products from the Baetica, especially the Maña C2b amphorae (= Type 7.4.3.2 and 7.4.3.3) and Gaditanian ovoids (= “archaic” Dressel 10) for fish sauces, stand out among the group of these materials as well as the “early” Dressel 20 (= Dressel 25; Oberaden 83; Dressel 20 A; Haltern 71; Class 24; Type A), and Haltern 70 and Sala I/Lomba do Canho 67. These two latter amphorae transported different products, such as defrutum, preserved olives, and perhaps wine (Fabião, 1989; 1998; Naveiro, 1991; Paiva, 1993; Arruda, and Almeida, 1998; 2000; Morais, forthcoming).

In the second stage, between the mid Ist century BC and the beginning of the Imperial period, the quantitative patterns of material distribution tend to diminish because of the multiplication of the peninsular imports, especially products from the Baetica.

The study of the Culip VIII shipwreck has recently proved that there was a strong concentration of Haltern 70 amphorae throughout the Atlantic façade, especially in the NW Peninsular (Morais and Carreras, 2003, 93-115).

It is a period in which we witness the flourishing of trade throughout the empire and the inversion of export trends, since there is a gradual substitution of Italic products for the products produced in the provinces (Tchernia, 1986, 140-46 and 153-58).

As we have previously mentioned, together with these products, we could also identify two new different types of amphorae: the first corresponds to the flattened base 106

THE ATLANTIC ROUTE

One of the phenomena worthy of distinction in the area corresponding to the coastal areas of the present Portuguese territory and the Gallaic coast is the hegemony of Baetica products, in particular Haltern 70 amphorae. As we had the opportunity to report regarding

the distribution of these amphorae in the Roman Empire, they are particularly well represented in the region between the Douro River and the Gallaic Rias Baixas (Morais and Carreras, 93-112).

Figure 2.5.18 Flat Base amphorae, urceus type, from Gadalquivir

107

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 2.5.19 Lusitanian Ovoid amphorae

108

THE ATLANTIC ROUTE

2.6 ROMAN LIGHTHOUSES ON THE ATLANTIC COAST C. Fernández Ochoa and A. Morillo The constant discovery of new archaeological evidence, almost unknown until a few years ago in the Atlantic area, provides us with irrefutable proof of intense maritime trading activity in Roman times (Reddé, 1979; Chic, 1981: 223-249; Remesal, 1986: 77-79; Naveiro, 1991: 128-137; Fernández Ochoa & Morillo, 1994: 183194; Fernández Ochoa (ed.), 2003; Morillo, 2003). The evidence is very varied. Together with wrecks, or parts of ships, such as anchor bolts (very scarce in these waters), we have another important archaeological source: the materials spread throughout the western European coasts by means of maritime trade. But we should also not forget the remains of port facilities such as docks, banks (v. Fernández Ochoa (ed.), 2003) and other constructions aimed to help the navigation such as lighthouses. And to lighthouses we will pay special attention.

In these representations, lighthouses are always stairshaped. They can have one or several stages and some of combine several different phases. The lighthouse of Alexandria had three stages: the lower square section with a central core, a middle octagonal section, and, at the top, a circular section. By means of ramps and stairs wood was taken to the upper parts. Nevertheless, we do not know if this was a common pattern or if it was just an iconographic representation based on the Alexandrian model to identify easily the general concept of these buildings.

ROMAN LIGHTHOUSES IN THE NORTH ATLANTIC OCEAN AND THE CANTABRIAN SEA

The existence of a lighthouse at Flavium Brigantium (A Coruña) is mentioned by Orosius on the route from Rome towards Britannia (OR. Pag. I, 2, 71). Set next to the port entrance, the lighthouse of A Coruña, the so-called “Tower of Hercules” has an undoubted Roman origin. The plan of the tower is rectangular, built in opus vittatum with concrete filling in the interior and ashlars in the angles and openings. The inside is divided into three floors communicated through ramps (figure 2.6.1). By the beginning of Modern times the lighthouse had lost completely its perimeter walls, keeping only the central stage with the marks of the ramps. The outside of this central stage, as well as the upper part, was completely refurnished during the XVIIIth century. Thus, any reconstruction would be hypothetical (Hutter, 1973; Hauschild, 1977: passim; Bello, 1991b).

The presence of lighthouses on the European coasts of the Atlantic Ocean was as, or even more necessary than in the Mediterranean Sea. Remains of lighthouses have been found in Brigantium (A Coruña) and Dubris (Dover). In both cases they are not earlier than the IInd century AD.

Lighthouses are closely related to the Roman navigation system. Fires were known to have been used to send luminous night signals throughout all Antiquity; they were placed on cliffs or estuaries, indicating refuges for shipping. Thanks to literary testimonies, we know the great lighthouses of the past: Alexandria (whose construction on the island of “Pharos” in the year 280 BC by Sostratos from Cnido was used to name this type of building); Ostia – where Claudius built the lighthouse on the remains of a shipwreck and protected the entrance to the port; or the example at Puteoli. Other lighthouses, such as at Leptis Magna, Brigantium (A Coruña) or Dubris (Dover) have been archaeology recorded. Nevertheless, it is necessary to look at iconographic representations on coins, paintings, mosaic, or reliefs, to try and reconstruct these buildings, since the construction remains that have survived are very limited (Reddé, 1979b; Martínez Maganto, 1990; Giardina, 1995 & 1997).

The building has been dated to the IInd century AD, judging from the materials that L. Monteagudo found in the archaeological intervention in the surroundings of the tower (Bello, 1991: 167). An inscription on stone found near the building points to its possible builder, C. Sevius Lupus, from Aeminum (Coimbra, Portugal). P. Le Roux 109

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 2.6.1 Recreation of the lighthouse of Brigantium in Roman times (E. and P. Cabarcos, Factoría Gráfica)

relates the construction of the tower to the military garrison at A Cidadela (Sobrado dos Monxes, A Coruña), although, strictly speaking, Brigantium was not a military port. He also states the existence of a statio of the portorium, deduced from the dedications to Marcus Aurelius and Lucius Verus by Reginus, exactor, possibly a civil servant for the collection of the shipping tariffs at the port of A Coruña (Le Roux, 1977: 92; v. Fernández Ochoa & Morillo, 1994: 61). The location of the port of Coruña (San Claudio, 2003) at such a strategic point between the Atlantic and Cantabrian routes of navigation would justify the construction of such a building.

Figure 2.6.2 Eastern lighthouse of Dover, near St. Mary-in-Castro Church (A. Goy Diz)

The old Dubris or Portus Dubris (Dover, Kent) lies at the mouth of the Dour River (Dubras) on the southeast coast of Great Britain, close to the European mainland. On both cliffs around the estuary lighthouses were built to protect the entrance to the port, the main naval base of classis Britannica on the British side of the channel. The foundations of the western lighthouse, which was set next to the Roman fort at Dover and dismantled at the beginning of the XIXth century, are located at the socalled ‘Devil’s Drop’ or ‘Drop Redoubt’, where some brickwork with the mark of classis Britannica has been found. This fact points to a construction, or rebuild, at the beginning of the IVth century (Wheeler, 1929: 40-46). The remains of the eastern lighthouse are preserved within the bell tower of St. Mary-in-Castro Church, next to the old castle. The eastern lighthouse has an octagonal plan and is square inside. At ground level, it is 4 meters square. It is preserved to a height of 13 meters and Wheeler thinks that it may have between 20-25 meters high (Wheeler 1929: 33). But according to Booth, the original height was around 15 meters (2007: 17). Outside it has eight decreasing steps, each 25 cm narrower than the following one. Booth has suggested recently that it did not have a stepped-shaped profile, but that its height decreased following a conical shape (2007: 17) (figures 2.6.2 and 2.6.3).

Figure 2.6.3 Reconstruction of the Eastern lighthouse of Dover (G. de la Bédoyère, 1991)

It is built in masonry with a concrete core and two rows of bricks regularly disposed, which provided a polychromic effect. The door and windows are topped with round arches and the jambs also have alternate brick

courses with mortar joints. Only the first four steps of the Roman building are preserved, the rest was very much altered in medieval times. The first mention of the tower as the bell tower of the church of the castle dates back to 110

THE ATLANTIC ROUTE

1252 AD (Wheeler, 1929: 30-34; Collingwood, 1930: 66, fig. 25; Booth, 2007: 9-19). Judging from the absence of classis Britannica marks on the bricks, Wheeler pointed to the Ist century AD as the construction date (1929: 39). Nevertheless, this same criteria, as well as the similarity of the brick found in the second fort of the Britannica fleet, built between 130 and 150 AD next to the port, has suggested a parallel construction dates for both (Booth, 2007: 20). The oldest lighthouse in this Atlantic area would be that of old Gesoriacum (Boulogne-sur-Mer), although it has not survived into our times. Caius had it built in the year 39/40 AD within his project of the conquest of Britannia. It was set in the foreland that protected the port and estuary of the Liane River, near the place where the Roman fort of Bononia might have been built. The lighthouse was reconstructed or restored by Charlemagne in the year 810 AD. In 1544 the English integrated it within a brick fort defended by fours towers in the angles, the so called Tour d’Odre that collapsed in 1644.

Figure 2.6.4 Probable location of the Roman lighthouse of Gijón (FMC and UP Gijón)

We have a reinterpretation of this lighthouse thanks to the old descriptions and drawings from the XVth to XVIIIth centuries. Built on the cliff, the tower had several decreasing geometric steps, topped with a pyramidal shape that might have reached 46 m height. The materials may have been stone and brick, such as at Dover (D’Erce, 1966; Reddé, 1979b: 868). In the last few years we have proposed the existence of a Roman lighthouse in the Pre-Roman/Roman hill fort at Campa Torres (Fernández Ochoa et al., 2005), which protects the entrance to the Roman port of Gijón (Asturias, Spain) (figure 2.6.4).

many studies. It is a large yellow-grayish block of marble, 1.62 x 0.80 high and 0.50 thick, on which a dedication to the emperor Augustus was carved. The transcription reads as follows in figure 2.6.5. The fourth and part of the fifth line have been intentionally defaced. They contained the name of the devotee, Cn. Calpurnius Piso, erased because this person fell out of favour and was punished by Tiberius (with a damnation memoriae) in AD 20 Syme, 1969: 129-132).

This settlement, the oppidum Noega of the ancient sources (MELA III.13), was the place chosen by the conquerors to erect a great commemorative monument. The marble altar devoted to the Emperor Augustus by Cn. Calpurnius Piso in the year 9/19 AD was also part of it (Syme, 1969: 129-132). This inscription, now belonging to an antiquarian, was already known in the XVIth century (CIL II 2703; ERA nº 12) and it has been the subject of

The dedication to Augustus led scholars to consider the altar one of the three “Sestian altars” that were placed in the Hispanic finis terrae and described by Roman sources. The existence of one of these altars devoted to Augustus by L. Sestius Quirinalis in a peninsula on the northwest of Hispania is stated in the geographic descriptions of Mela (MELA, III, 13), Plinius (PLIN.

IMP.CAESARI.AVGVSTO.DIVI.F

COS.XIII.IMP.XX.PONT.MAX

PATR.PATRIAE.TRIB.POT.XXXIII

SACRVM Figure 2.6.5 Calpurnius’ plaque (9-10 BC) from the Campa Torres 111

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 2.6.6 Excavations by Manuel Reguerra in the Campa Torres in 1783

placing the plaque by Cn. Calpurnius Piso on that building and then transferred sometime in the XVIth century to a chapel in the vicinity.

Nat.IV, 112) and Ptolomeus (PTOL. Geog. II, 6, 6), and also in the itinerary guide know as Anonymous of Ravenna. These altars, set up between the years 22 and 19 BC, were very famous in ancient times and were to become symbols of Augustan control in the northwest of Hispania. All the authors, except Mela, place the “Sestian altar” on the Galician coast. This contradiction in the sources has caused much controversy in the Galician and Asturian historiography, both sides keen to possess such an important monument within their territories. Nowadays, following Syme’s opinion (1969: 129-132), Plinius’ work is more favoured; he places the altar on some headland of the Galician rugged coastline, between Finisterre and Arosa. Despite all the attempts made by Asturian scholars, as already pointed out by A. Schulten (1943: 192-193), there is no evidence for us to include this altar in relation to those set up by L. Sestius. Even the chronology of that monument is much later, almost 30 years before Sestius’s rule (22-19 BC). This difference has been barely taken into account when placing the “Sestian altars” in the Astur Noega (Fernández Ochoa & Morillo, 2002: 903).

In 1783 some excavations were undertaken by the architect M. Reguera, sponsored by Jovellanos (López Álvarez, 1997). Reguera’s plans show apparently two different constructions. The bigger building has a quadrangular plan and is delimited by a wall 1.05 m wide and 9.24 m long. Inside this building there is another one, also square in plan, 4.76 on a side, whose walls run parallel to the first one. Between them there is an ambulacrum, 4.76 m on a side1, which probably had an inner stair. While the perimeter wall seems to have been made with ashlars, with two faces, the construction is solid and thus it could have had a core of opus caementicium covered with ashlars (figure 2.6.6). Such strength in the base can only suggest a significant height, that is to say the construction of a tower. It is possible that there were two stages, without ruling out the possibility of the existence of a third stage crowning the building, this could have been some 15-20 m height. Inside, there could have been straight flight stairs and quarter-turn staircases that would occupy the space described as a corridor or ambulacrum (cf. Fernández Ochoa et al., 2005: 132-142).

This important commemorative religious inscription can be reinterpreted from the findings at Campa Torres. Despite the frequent, and incorrect, identification of the “Sestian altars” in modern sources, two facts about the remains at Campa Torres are beyond all doubt. First, the remains of the two buildings destroyed in the XVIth century (around 1575), whose ruins were somewhere in the Campa Torres, near the cliff and a fountain (Somoza, 1908: 299-322). Judging from the texts, the larger construction must have had a quadrangular plan, towerlike, and, perhaps, stairs. All the extant sources agree in

The building described in Reguera’s plan has been lately reinforced when N. Martinez found the ruins 1 The measurements have been taken in an ancient measurement, “Castilian feet”, from Reguera’s plan: 1 Castilian foot = 28 cm. Taking into account the different correspondences of these measures by the different authors, the proposed dimensions could vary slightly but this does not mean a change in the proportions.

112

THE ATLANTIC ROUTE

Figure 2.6.7 Campa Torres (Gijón). Drawing of the remains of the tower after the excavations undertaken in the year 1884 according to Nemesio Martinez

0.80 m high) was carved as a part of one of the building’s faces, and was, surely, in a prominent position. Taking into account that the plaque refers to a consecration, it is reasonable to think that it was actually consecrating the monument. There are examples of monuments consecrated to Augustus (Bará Arch, La Turbie trophy), or to a specific divinity like Mars (Adamclissi trophy, the Tower of Hercules). In places where it has been possible to prove the monument, the dedication is well visible and on the main façade. Thus, it is reasonable to think that the inscription at Campa Torres would have had a similar location.2

of that same construction in a lithograph dated to 1884 (Martínez Sienra, 1884, reed. 1986: 71). Martinez not only expresses a romantic view of the derelict building but he captures also the results of the 1884 archaeological interventions (Adan, 1997: 215). That same year, the Gijón newspapers reported the pillaging of the remains that had been brought to light after the excavations (Maya & Cuesta, 2001: 27, nota 16) (figure 2.6.7). Being cautious, it can be said that this building follows a very characteristic pattern of a Roman turris, with clear antecedents in the Hellenistic world. The application of this architectural pattern is very well known in funerary monuments from Roman time (Gros, 2001: passim), but there is no doubt about the existence of many similar constructions with different purposes: defensive, trophy/ commemorative buildings, and, of course, lighthouses.

On the other hand, the large size of the plaque makes it unlikely that the ashlars used for the building were of the same size. It seems that the inscription was carved on a marble ashlar with a prominent place within the building, both because of its size and the material chosen. Moreover, that carved ashlar must have been imported, as this material does not exist in Asturias3 and there were no active epigraphic officinae at that early date. No other constructive material has remained. After the dismantling of the tower during the second half of the XVIth century, the ashlars were reused in other buildings, perhaps in the construction of the embankment of the port of Gijón, as it appears in the well -known engraving by D. Fernando Valdés, dating back to 1635. But, as a hypothesis,

As for the second building from Reguera’s plans, it is difficult to ascribe any function to it. Its form and close relation to the other may point to an enclosure with a sacrificial altar recalling the ara pacis, although it could have also been just a platform for access to the turris. In fact, in the words of the people familiar with the excavation, the structures identified in the drawings correspond to the foundations of the buildings, a theory that makes it possible to consider the construction of the upper layers.

2

The oldest written references place the plaque in the only building visible at that time: the tower-like structure (v. supra). There are no data that let us put the inscription in relation to the second building, although such a consecration could also have been sited on a possible altar. (v. Alföldy et al., 2003: 258-260). 3 Taking into account the circumstances of the current location of the plaque, in a private collection, it has not been possible to make an analysis of the material, which could confirm its Hispanic, or foreign, origin.

As has already been said, the oldest descriptions associate the monumental inscription by Cn. Calpurnius Piso to Augustus with this building (De Morales, 1586; De Avilés, 1956: 211-212). The plaque (1.62 m long and 113

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 2.6.8 Roman lighthouses and maritime routes in the European Atlantic coast (Fernández Ochoa et al., 2005)

Although the interpretation of the monument still poses many questions, we have a tower-like structure, set on a rise on the Cantabrian coast, with a monumental inscription, the first and only one known in the north of the Peninsula, which alludes to a consecration of Augustus in the year 9-10 AD. All these elements together, in Late Augustan times and in a recently pacified and strongly militarized area, points to a situation in which the official initiative and possible political aim must be taken into account.

to have this building built during his rule in commemoration of some event that had happened in the previous years, possibly during the conquest. This event could have been the arrival of Roman troops at the edge of the Oceanus, providing an echo of the similar situation (leading to the “Sestian altars”) that had occurred some years before. Its role as a signalling lighthouse at the best Asturian port, in a militarized and recently conquered area and in one of the most important points of the Cantabrian coastline, makes complete sense. These altars, trophies, or towers of a symbolic or commemorative character, in spite of their ritual consecration, are of more political than religious intent (Fernández Ochoa & Morillo, 2000: 904-905).

Calpurnius’ plaque in the hill fort of Campa Torres must have formed part of a great building, dating from the end of Augustan times, on the Asturian coast. It must have been constructed thanks only to the power and control of the army in charge of the administration of the new territories. It is possible that Cn. Calpurnius Piso wanted

Despite the fact that we have no construction remains that can be positively related to a lighthouse, the topographical situation of the Jaizquibel mountain (Fuenterrabía), which protects the entrance to the estuary of the Bidasoa River, where the Roman port of Oiasso (Irún, Spain) was set, could indicate the existence of a

we prefer to consider the use of local stone for the construction of the Roman building, a lighter and easier material to obtain in the area.

114

THE ATLANTIC ROUTE

typical than we can so far understand from the classical sources (Fig. 2.6.8). They were essential features for navigation in this area. The great increase of archaeological data along the shores of the Atlantic Ocean, which reveal the existence of well-established maritime routes, makes us rethink this question, taking into account very different methodological premises. Nevertheless, the limitations of navigation in Roman times (navigating close to the shore, avoiding night journeys, avoiding cliffs and reefs, etc.) mean also that the lighthouses would be erected at strategic points related to auxiliary ports that could provide shelter at night and during bad weather. This conception is different to the main modern role of lighthouses – i.e. the indication of dangers along the coast.

similar structure here. Jaizquibel is a large rocky eminence rising over the coast and going deeper into the sea, forming Cape Higuer. It has been traditionally identified with the “Oiasso foreland”, mentioned by Ptolomeus as being in the surroundings of Irún (PTOL. Geog. II.6.10; II.7.2). Schulten placed the Veneris Iugum on that hill; this was the famous temple of Venus mentioned by Avienus (AVIE. O.M. 158) (Schulten, 1955: 100). At its top, 541 m. high and within the same area of the old fort of San Enrique, a coin from Nerva was documented among some construction remains of uncertain date, but in any event prior to the construction of the fort (Rodríguez Salis & Martín Bueno, 1981; Esteban, 1990: 172). The archaeological soundings carried out here did not yield any other testimony from Roman times (Fernández Ochoa & Morillo, 1994: 145).

The lack of archaeological data near important Atlantic and Cantabrian ports (Oiasso (Irún) and Portus Victoriae (Santander) on the Hispanic shoreline; Burdigala (Burdeaux); the Armorican ports or the mouth of the Rhine; Londinium and the British ports), represents much more a research problem than a historical reality.

FINAL CONSIDERATIONS Archaeological research has made it possible to identify two Roman lighthouses along the North Atlantic façade of Europe (Dover, A Coruña): a possible one in Gijón and one more known to us through depictions and descriptions at Boulogne-sur-Mer. These structures can be grouped into two different categories with regard to their dimensions. A Coruña and Boulogne have around 20 m on a basal side, whereas Gijón and Dover are much smaller, about 10 m. Of course, the height of these buildings must have been in relation to their bases. Nowadays it is difficult to establish whether this duality responds to any unbiased criteria as, for example, the actual importance of the port to the Roman state, or specific visibility difficulties, or whether the architect in charge of the project had other reasons.

Apart from their close relationship with ports of refuge, another characteristic of lighthouses in this maritime area is their connection with the Roman Army. Some of them are also military bases (classis Britannica in Dubris and Gesoriacum). Others were erected at militarized sites, i.e. Brigantium, a lighthouse that must have related to the nearby fort of cohors I Celtiberorum, A Cidadela (Sobrado dos Monxes, A Coruña), or Gijón itself, the port of the legionary camp of the legio VI victrix, and the legio VII gemina in León, more than 100 km away. Lighthouses were part of state support for ocean navigation. This support was highly necessary to keep the maritime routes safe and open for supplying the British and German frontiers.

It seems that the Roman lighthouses of the North Atlantic Ocean and the Cantabrian Sea must have been more

115

Part Three MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

3.1 THE STRUCTURE OF MILITARY SUPPLY: AN EXPLANATORY MODEL1 C. Carreras In the last years, there has been a new interest in the role played by the Roman army in the long-distance exchange between different territories and provinces. The Roman legions became one of the main stimuli for the longdistance exchange during the Principate, since they fostered commercial transactions between distant lands. This continuous flood of commodities from far-away centres was not a spontaneous activity, but a wellorganised network that involved either military or civil staff. The army always required a constant supply of staples for its survival. As Vegetius (VEG. Epit. 3.3) said: “For armies are more often destroyed by starvation than battle and hunger is more savage than the sword. Secondly, other misfortunes can in time be alleviated: fodder and grain-supply have no remedy in crisis except storage in advance.” 2

the army supply was always a problem in Republican times, since there was neither an institution nor a post responsible for it. For instance, Caesar’s victory was at stake twice during the civil war due to lack of corn and supplies amongst his soldiers. Such was, for instance, the case during the manoeuvring around Ilerda (BC 52) in the campaign against Afranius and in Northern Africa (CAES. Afr. 19.47) against king Juba. The Republican experience provided the basis for the development of a complex structure of military supply that would characterise the Principate and the Late Empire (Erdkamp, 1998). Unfortunately, there are no direct written sources that explain how this structure worked and evolved over the period. Hence, scholars have had to build up explanatory models based on a few indirect documents and a huge volume of archaeological data.

1

The network, it is commonly believed, was created during the Principate because of problems that the armies in the field had to face in Republican times. Despite Cato’s phrase ‘war feeds itself’ (LIV. 34.9.12), the Republican generals used to be very careful in supplying their soldiers in campaign and times of peace. Of course, Roman troops requisitioned supplies from their enemies or allies in time of war as many sources record.3 However, they could never solely rely on those potential resources.4

THE STRUCTURE OF ROMAN MILITARY SUPPLY: AN EXPLANATORY MODEL The Roman army, whether legions or auxiliaries, constituted an essential instrument for the survival of the Empire, as well as its distinctive way of life. Maintenance of this powerful war machine required a series of basic supplies for its subsistence, which normally were obtained from the territory where the troops were settled (Wierschowski, 1984; Remesal, 1986). Vegetius (VEG. Epit. 3.3) describes what the army’s main objectives were in the late fourth century AD concerning supply: “...there should be careful consideration given to supplies and their issue in order that fodder, grain and other army provisions customarily requisitioned from provincials may be exacted in good time, and quantities always more than sufficient be assembled at points well-placed for waging war and very well fortified.” This situation, common to all the armies in history, was probably particularly complex in the Roman world, where legions were settled far away from the capital of the Empire (Van Berchem, 1977; Remesal, 1986; Whittaker, 1989), which

Livy (LIV. 23.49.2) mentions that societas publicanorum obtained contracts for the military supply in Republican times, and, besides, that they were exempted from military service and liability in case of piracy or shipwreck. For instance, in 52 BC Caesar (CAE. Gal. 7.3.1) entrusted the supply of all the Roman legions in Gallia to the negotiator Gaius Fufius Cito. Nevertheless, 1 This paper was published in 2002 in Erdkamp (2002). The present article is an updated version. 2 Similar thoughts are expressed by Caesar (CAES. Civ. 1.72.1), Frontinus (FRON. Str. 4.7.1) and Ammianus (AMMIAN. 25.7.4). 3 Caesar (CAES. Gal. 4.32; 5.17), Josephus (J. B.J. 2.528). 4 Sun-Tzu in his Art of war (IV BC) voices the same idea “a good general manages to feed his troops from the enemy” (II.15)

119

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

posed great difficulties to transport and the flow of information (Duncan-Jones, 1990, 5-76; Carreras, 1994).

Britain or Germany could not supply enough corn for the army settled there, let alone such items as sour wine or olive oil, which were produced in more southerly latitudes.

Interestingly, the Romans maintained a common diet for all the soldiers in the Roman army. A few possible explanations may be given. Firstly, it was less risky at the physical level to maintain the same diet when the army was, for a short time, away from their own territory. In addition, on a psychological level, this local diet preserved soldiers’ emotional links with their homeland (smells, tastes) and their rejection of foreign foods (neofobia). Moreover, at the sociocultural level, a common diet for all the soldiers fighting in distant lands was a way to maintain a collective mind as well as preserve the same identity of the people they defended (Carreras, 2000, 176).

A second alternative for local military supply was billeting, which means that local civilians provided accommodation and food for soldiers following the practice known as hospitium. This was a common practice in the Eastern Roman Empire (Safrai, 1994; Kissel, 1995) in provinces such as Syria or Palaestina, where local people had to lodge soldiers as well as provide food and clothing. Only soldiers, veterans, teachers, doctors, orators and philosophers were exempted from compulsory billeting (Isaac, 1999, 298). Talmudic sources illustrate the fear of civilian Jews facing the hospitium burden and soldiers’ behaviour in their homes. For instance, Shimeon Hatemani in the second century AD (tos. Betzah 2.6) explains: “A patrol of gentiles came into town and they [the townspeople] were afraid that they [the soldiers] might harm them and therefore we prepared them a calf and we fed them and gave them to drink and rubbed them with oil so that they would not harm the townspeople.”

To guarantee the regular supply of the troops, Rome organised an extremely complex administrative system that provided what any unit required, making these goods available to the final destination. The army could obtain these supplies from: a) the local territory b) the own province

Both systems allowed the military units to obtain necessary foodstuffs without investing much in either transportation or infrastructure. Ideally, most resources should come from the local land and population, whereas small amounts of extra items may have been purchased from local private traders in the local markets, as the Vindolanda tablets seem to document (Bowman and Thomas, 1983). Every unit had its own finance administrator, sometimes this function was undertaken by a tabularius or a signifer (VEG. Epit. 2, 20; PSI 1603), who, on his officer’s orders, took note of all transactions made by the unit. Normally, the responsibility for supplies was under the control of an evocatus and signifer (CIL VIII.18224), assisted by a quaestor to make payments and an actarius, who registered transactions in books (Le Bohec, 1994, 51).6 At Vindolanda, the optiones were in charge of these books (Tab. Vindol. II, 127), as they show purchases made locally.7 Provincial procuratores provided the soldiers’ pay, as well as other expenditures in the camp. Accounts were then sent back to the procurator.

c) the other provinces of the Empire a) Supply from the local territory Generally, the military units obtained directly their supplies (i.e. food, cloth, skins, and metals) from the near environment. Therefore, they could sustain themselves due to their own resources (i.e. prata, figlinae, fabricae), or from local civilians through taxes, requisition or trade. If the army wished to maintain peaceful relations with their neighbourhood it was wise not to requisition or raise special taxes to get local supply. Most native revolts were brought about by soldiers confiscating goods or collecting taxes (Mócsy, 1966; Petrikovits, 1974; 1976; Remesal, 1986). The basic Roman military diet in peacetime consisted of corn (wheat or barley), sour wine (acetum), salt, cheese, bacon-fat (laridum) and olive-oil.5 Depending on the size of the units, as well as the production capacity of the neighbourhood, the army requirements may often have been fulfilled by the local resources. Nevertheless, this was not always the case.

Ideally, the local units should store in their forts or nearby horrea sufficient supplies to last for at least one year. Tacitus (TAC. Ag. 22.2-3) records that his father-in-law, Agricola, ordered one year’s supplies for each unit when he was governor in Britannia (AD 78-84). Calculations on the granary volumes in the British limes showed that they could contain enough corn for one year.8

Alternatively, the military units could first of all obtain their required foodstuffs from their own lands (prata), where they may have cultivated cereals, raised domestic animals (cattle, horses, sheep, pigs) or obtain wood. The soldiers themselves may have cultivated these lands or leased them to neighbouring civilians (Mócsy, 1967). However, the frontier regions were not specially suited to provide all the ingredients for their basic diet and not necessarily in suitable amounts. For instance, the limes of

6

The post of actarius is documented from Septimius Severus onwards (AUR.VIC., Cae. 33.13). 7 Tablets 4 (Tab. Vindol. II, 190) and 5 (Tab. Vindol. II, 191) (Bowman and Thomas, 1983, 88-95) include a series of purchases carried out by a person called Privatus, perhaps a slave. He bought wheat, barley, wine, sour wine, garum, lardum, pig and venison. 8 Davies (1989, 187) calculates that a Roman soldier consumed approximately one-third of a ton of corn annually (333 kg).

5 Appian (APP. Ib. 54) says: ‘...they had no [vintage] wine, salt, sour wine or oil, but fed on wheat and barley, and large quantities of meat and hare boiled without salt, which upset their digestion’.

120

THE MILITARY SUPPLY FROM THE LATE REPUBLIC TO THE EARLY EMPIRE PERIOD

Normally, the border regions of the Empire (Limes Africanus, Germanicus or Britannicus) were not suitable for many crops and also recorded low densities of population, so they did not have enough resources to feed a large, inactive group such as the army. Therefore, the units had to resort to remoter regions in the same province, as seems to have been the case in Britannia or Germania Inferior.9

mercatores or foreign traders in bulk, in order to distribute the supplies amongst the units by using the public or private transport network. Sometimes, due to special military requirements, the army was obliged to requisition foodstuffs from the provincials, with the compulsory permission (diploma) known as indictiones. There are numerous testimonies such as papyrus PSI 683, in which clothing, ropes and oils were requisitioned, as well as other papyri (BGU 266; P. Gen. 35; P. Flor. 278).13

b) Supply from own province

When the geographical conditions required a constant collaboration of the civilians in the supply and military transport, laws were issued to fix general norms. This is the case in the decree of S. Sotidius Strabo Libuscidianus (Mitchell, 1976; AE 1976, 653), legatus augusti in Galatia during the reign of Tiberius, who asked for wagons and pack animals from the community of Sagalassos in order to take care of military overland transport. The procurator had the right to use a large number of animals, while senators and soldiers were only permitted a more limited number. This special treatment for the procurator is related to his special function regarding the military supply in the province (Remesal, 1990, 60).

The situation became more complex when supplies came from distant places in the province and diverse administrative levels therefore had to intervene. Since the Augustan period, as Strabo recalls, the highest financial authority within each province was the procurator augusti (STR. 3.4.20) (Mócsy, 1966). He was responsible for military supply,10 as well as other duties such as gathering direct and indirect taxes in the province (XX hereditatium, XX libertatis, portoria), which constituted part of the imperial fiscus.11 According to the number of military troops, the procurator assigned a proportional amount of money for each unit’s officer in order to obtain supplies from the local market. However, as said before, depending on what they could get from their own local resources, each unit may have made different purchases. Sometimes supplies could easily be obtained from a network of local private traders, who carried the army’s requirements from distant areas in the province, as for example a wagon of hides coming from Catterick to Vindolanda.12

Overland transport was a general problem for Roman military strategy.14 With the only exception of Legio (Hispania), all legionary headquarters during the Principate were established in places where the geographical conditions allowed the easy and safe movement of supplies over water, which otherwise would have required cumbersome overland transportation.

Apart from Strabo’s passage, the relationship between procuratores and military detachments regarding supplies becomes evident from an inscription from a tabularius of the legio III Augusta dedicated to the procurator IIII publicorum Africae, Ti. Ci. Proculus Cornelianus (AE, 1956, 123). In addition, there is a letter from Pliny the Younger (PLIN. Ep. 27), acting as a governor of Bythinia, in which he mentions that the procurator Maximus required the service of soldiers assigned to him. Maximus needed those soldiers to collect grain in Paphlagonia, destined probably to supply the Roman army during the Parthian campaign.

When supplies had to come from a great distance, special detachments (milites, frumentarii15) were sent to fetch the required supplies and protect convoys.16 The rostrum from Sept. 105 AD from Cohors I Hispanorum (Fink, 1971, nº 63), which was quartered in the province of Moesia inferior, sheds some light, since it documents the absence of soldiers from the camp due to several missions. For instance some soldiers went to Gallia to collect grain and clothing, while others fetched horses on the river Erar, transported herds from Haemus, or protected food convoys in the province of Moesia. The post of frumentarii is of great interest because their inscriptions are usually found in the capital of provinces,

The procurator had to organise the supply of all the military detachments that could not obtain resources locally. As provincial finance administrator, he could purchase food and other supplies from provincial

13 There is an interesting inscription (AE 1958, 236) in Hama of Domitian times, in which the procurator of Syriae, Claudius Athenodorus, is exhorted to prevent abuse in asking for labour, pack animals and wagons from local civilians (Mouterde and Mondésert, 1957). 14 Tacitus (TAC. Ann. 15.8) refers to a senseless campaign of Caesinius Paetus in Cappadocia interrupted because of a lack of supplies which had to be obtained overland. 15 Originally frumentarii were responsible for corn supply as their name suggests. However they also acted as spies (Clauss, 1974; Paschoud, 1983) or security agents, as the rostrum of legio III Cirenaica reveals (Fink, 1971, nº 10). 16 Troop dispersion, it is said, was one of the causes of Varus’ defeat in AD 9 in the Teutoburg forest. Dio (56. 19) records that Varus: ‘...did not keep his legions together...but distributed many of the soldiers...for the alleged purpose of guarding various points, arresting robbers or escorting provision trains’.

21

Britannia: Anderson (1992); Germania: (Brandt and Slofstra, 1983; Willems, 1986). 10 The figure of procurator augusti is fully discussed in relation to military supply by Wierschowski (1984), Remesal (1990) and Dise (1991). 11 Some provinces probably raised fewer taxes than the cash they required to feed the army and all the administrative staff. Therefore, one may expect some kind of compensation between the provincial treasuries, either imperial fiscus or senatorial aerearius Saturni. This compensation favoured a balanced economic relationship between central and peripheral provinces. 12 The tablet Vindol. II, 343 details purchase and sale of corn and hides by Octavius (Bowman et al., 1990, 43).

121

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 3.1.1 Distribution map of beneficiarii information,20 and supervising traffic, markets or the collection of taxes (portoria) (Schallmayer, 1990). This last responsibility was also shared with centurions (P. Oxy. 1185). When beneficiarii were located in the main centres of communication (statios), they became, together with their assistants (stratores,21 exceptores, vilici), key figures in the military supply, second only to the procurators.22

probably acting in the governor’s officium, or even outside of the province, far away from their original units. For long-distance provisioning, a suitable communications network was required, along with adequate means of transport under civilian or military control. Furthermore, administrative control was also required to avoid fraud. This administrative inspection was carried out by the beneficiarii (consularis, procuratoris),17 amongst others, when supplies were in transit, and by the unit commander after reaching final destinations.18

The map of the distribution of inscriptions of beneficiarii in the Roman Empire (figure 3.1.1) (Schallmayer, 1990), illustrates their location on the main roads to military districts as well as in their limites. Provinces, whose supply was based on overland transport, such as Danubian and Germanic ones, seem to document a large number of beneficiarii because their logistics may have been more complex. In contrast, Eastern and African provinces, as well as Hispania, appear to have obtained most of their supplies locally (Carreras, 1997).

The first beneficiarius documented appears in the late Republic during the Civil War (CAES. Civ. 1.75), though the first epigraphic record dates from Tiberius’ reign (CIL IX.5711). Many of these beneficiarii are found in the provincial governor’s officium as assistants in the daily paperwork (Davies, 1989, 44). Nevertheless, they operated often in the main centres of communication, where they were assigned a wide variety of duties. We find them acting as police officers,19 obtaining

Amh. 77; P. Amh. 80). However, the responsibility may have also fallen on centuriones or stationarii (Davies, 1989b, 175-185). 20 Austin and Rankow (1995) show the function of beneficiarii as information suppliers to the Roman military intelligence. 21 There are some interesting works on stratores and their functions of control and military supplies related to statios, e.g. Santero (1981) and Perea (1998). 22 Some beneficiarii were specially adscript procuratores, beneficiarii procuratoris, chiefly as foremen for the collection of portoria.

17

Responsibilities of the beneficiarius are not clearly defined until the Flavian period, when they appeared in the main hubs of communication and in the governor’s officium (Domaszewski, 1902; Popovic, 1989; Schallmayer, 1989; Ott, 1995; Stoll, 1997). 18 The commanding officer had to test the quality of the rations and prevent quartermasters from cheating (Digest. 49.16.12; P. Dura 61). 19 There are a myriad of papyri in Egypt that record beneficiarii acting as crime investigators in a district (SB 9657; SB 9238; SP 9203; P.

122

THE MILITARY SUPPLY FROM THE LATE REPUBLIC TO THE EARLY EMPIRE PERIOD

The available documentation on beneficiarii seems to offer a relatively complete picture. They appear for instance together with stratores in the main entrance of Dura Europos (Baur and Rostovzeff, 1929; AE 1931, 116-7) collecting taxes, likewise in Zarai (Numidia) (CIL VIII, 4508), Mogontiacum (CIL XIII, 11816), Lambaesis23 (AE 1914, 234) or Aquincum (CIL III, 10429), where they controlled the commercial traffic between the colony and the military territory. In all these inscriptions, the beneficiarii were liable for controlling quality, quantity and, probably, price of commodities in transit, some of which were meant for military consumption. The digestae (39. 4, 4, 1, Paulus) include a law from the reign of Hadrian, which reproduces the norms for official transport. It says that official transportation required a document signed by the praesides (libellus) listing all the products and quantities destined for own consumption, to the province governor, procuratores or the army. Hence, the document had to be shown to the toll official (publicanus), so he could exempt the public cargo from the portorium, as far as it did not surpass the quantity and variety mentioned (Whittaker, 1989, 64).

German limes (Nierstein, Erpolzheim, Altrip, Germersheim, Strassburg). However, the region with the highest concentration in the whole Empire was the German limes, foremost in key strategic places such as Mogontiacum (17), Stockstadt (20), Osterburken (30), and others no less important, such as Oberubung, Stuttgart, Jagsthausen, Friedberg or Grosskrotzenburg.24 All the evidence points to the same conclusion: beneficiarii acted as links between the financial administration of the province under the procuratores and the commanding officers and administrative staffs of the military units. Their role was to control a regular flow of supplies obtained in the same province outside the military territory, and also from other provinces. When supplies came from other provinces,25 the whole structure became even more complex because it surpassed the procurator’s authority, and hence other posts or institutions were required to co-ordinate military demands, taxes, purchases and transportation from the production provinces (Schlippschuch, 1987). c) Supply from other provinces of the Empire

The importance of the role of beneficiarii in the supply of the Danubian provinces has been recently emphasised by Dise (1991, 78), who pointed out the high concentration at key strategic sites such as Celeia (Noricum), with 23 inscriptions of beneficiarii procuratoris between the late first century AD to AD 170. Celeia was midway on the overland route from the Mediterranean ports to the Danubian limes (Dise, 1996). The distribution of beneficiarii consularis is even more suggestive for the later period (AD 161-235), because their inscriptions are well represented in most centres of communication in Pannonia superior (Siscia: 10; Savaria: 10; Poetovio: 5) and inferior (Intercisa: 8; Mursa: 5).

It is obvious that not all the products consumed by the army could be obtained within the province, so they should be provided regularly at a reasonable price and in specific amounts by other provinces. Some scholars argue that a market exchange mechanism could have perfectly well fulfilled these requirements without direct state intervention (Hopkins, 1980; Fulford, 1989), or possibly with some incentives (Whittaker, 1989). Nevertheless, inadequate transport infrastructure and communication in Roman times makes this hypothesis extremely unlikely.26 First of all, the central authority in the province had to know the exact requirements, chiefly staples (i.e. corn, wine, olive-oil), of the units in the province as a whole. It is perfectly reasonable that the province’s procurator could play this role as far as finances were concerned. Then, he should contact either traders or local producers to order exactly the amount required at a fixed price. Since procuratores did not have any authority outside their province, the contact and negotiations with traders and producers had to be undertaken by other institutions.

Schallmayer’s work (1990) contains most inscriptions of beneficiarii in the Roman Empire, so it allows us to analyse their distribution in some detail (see figure 3.1.1). In some western provinces such as Belgica and Germania inferior, concentrations are documented in the main centres along the Rhine (Köln, Bonn, Remagen, Xanten, Neuss), in two Atlantic ports (Boulogne, Zierikzee) and in the main overland routes coming from Gallia (Trier, Nettersheur, Arlon, Namur, Aachen). As can be observed, many beneficiarii inscriptions are located on the frontiers between provinces and places where there is a shift in the means of transport. Consequently, they may have been responsible for controlling military supplies coming from other provinces in a place where the goods changed hands and where, as a consequence, there was more risk of illegal practices in the handling of public cargoes.

The only known existing institution that acted in this way was the annona, an institution responsible for a 24 The distribution pattern of beneficiarii in other provinces is similar, though it varies according to the logistics and geography of each territory (Schallmayer, 1990). 25 The supply from other provinces is well documented in the rostrum of the cohors XX Palmyrenorum from Dura Europos and the cohors I Hispanorum from Moesia Inferior. It is also evident from the presence of continental grains Britannia (York, Caerleon), quotes of export of Gaulish grain to Hispania or British corn to Germania, and even a wide variety of amphorae such as Dressel 20, Gauloise 4 or Dressel 6 (Carreras, 2000). 26 Duncan-Jones (1990, 7-58) shows clearly the difficulties of communication between Rome and Alexandria, where the death of the emperor could only be known a few months later in winter. With regards to transport costs and time consumption on travelling see Carreras (1994).

With reference to Germania superior and Gallia, there are examples from along the Rhone (Nimes, Vienne, Lyon), on the Alps routes (Geneve, Vevey) and along the main overland routes between the Rhone valley and the 23

The inscription of Lambaesis (AE 1914, 234) shows that signiferii, beneficiarii, pecuarii and conductores took part in collecting the portorium.

123

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

redistributive mechanism that supplied corn and other products to the populace of Rome at a reasonable price. The Roman army may have employed a similar mechanism or even the same institution to obtain the required supply from the production provinces. For instance, the transport of olive-oil coming from Italia (amphorae Dressel 6) or Hispania (amphorae Dressel 20) may have been arranged beforehand as taxes in kind or at a fixed price, and then carried by local traders to the destined provinces after paying vecturae. The archaeological distribution of some amphorae types with extremely high densities in military sites seems to confirm the existence of this redistributive public system (Remesal, 1990; Carreras and Funari, 1998).

It is quite interesting that these remarkable amphorae distributions in the frontier regions seem to disappear gradually in Severan times, as if this dynasty’s reforms affected not only transport control of public cargoes (Remesal, 1986, 105), but also inter-provincial military supply, at least of olive-oil. Summing up, the military supply network in peacetime was defined by 4 hierarchical levels of decision. The highest rank was probably represented by the praefectus annonae who co-ordinated the extra-provincial supply, whereas procuratores organised finances, transport and supplies within each individual province. A third level identified by beneficiarii were in charge of controlling supplies and regular transport, while commanding officers and clerks were responsible for local purchases. Of course, the general model for military supply in the Principate proposed here presented slight variations in each province according to their logistic and geographical features.

However, the question arises whether this redistributive public system was part of the existing annona or a different institution such as the late annona militaris.27 It appears feasible that a post such as praefectus annonae28 could co-ordinate the extra-provincial requirements of the procuratores, since he already had recourse to the infrastructure and information needed to convey the supplies to each province. Therefore, the state did not have to pay the whole stipendium to soldiers, since it could withdraw all the amounts due for imported supplies (in victum). Perhaps this cash never even reached the unit tabularius or the provincial procurator, but remained in the praefectura annonae, compensating the accounts between provincial treasures. As Remesal (1986, 104) proposes, a central authority was needed in order to coordinate and maintain a system of compensations between the provincial treasuries, fiscus and aerarius Saturni. Only the praefectura annonae could fulfil this role.

However, requirements in wartime may have modified this structure, since transportation became extremely dangerous and more supplies were necessary (Roth, 1999). Vegetius (VEG. Epit. 3.3) pays special attention to these times of crisis: “Therefore, before the war is commenced, there should be careful consideration given to supplies and their issue in order that fodder, grain and other army provisions (annonariae species) customarily requisitioned from provincials may exacted in good time, and quantities always more than sufficient be assembled at points well-placed for waging war and very well-fortified. But if tax revenue be insufficient, everything should be compulsory purchased from prior contributions in gold.”

Furthermore, the system may have involved the direct intervention of frumentarii helping in intra-provincial transport and provisioning together with the beneficiarii, under the supervision of the praefectus annonae or the provincial procuratores. The result of this complex system in the archaeological record is the distribution of olive-oil amphorae such as Baetican Dressel 20 in high numbers in military sites in Britannia, Germaniae, Raetia and Noricum and Italian Dressel 6, with high concentrations in the Danubian limes (Bezezcky, 1987).29

The peacetime generic model underwent changes when the army was on campaign, since some temporary posts were created. There is a series of testimonies of these posts such as praefectus vehiculorum (Eck, 1975; Bérard, 1984) for the Dacian campaigns of Domitian; curator copiarum expeditionis in the Marcomannic campaigns of Marcus Aurelius (CIL VI, 1589); and the Dacian campaigns of Commodus (AE 1934, 2).30 A visual image of all the requirements of the Roman army on campaign is provided by the representation of the Dacian wars on Trajan’s column (Richmond, 1935).

27 The creation of the annona militaris in Septimius Severus’ time was an hypothesis put forward by van Berchem (1937), based on the inscription of M. Rossius Vitulus (AE 1911,7; 1914, 248), who held a new post of procurator annonae, during the campaigns against Clodius Albinus (AD 196-7). This temporary post (and others) did not continue afterwards. Therefore it does not demonstrate the existence of either a new tax or institution. Guey (1937) suggested that the institution was created in the Trajan period, but in fact the only occurrence dates from the time of Severus Alexander (HA, Sev. Alex. 15.5). Moreover, Cerati (1975) demonstrated a long time ago that what is known as annona militaris from Diocletian onwards (P. Beatty Panopl. 2.245-249), was only a part of the taxes in kind destined to the army (Remesal, 1986, 104). 28 Praefectus annonae was a magistrature created between AD 8-14 by Augustus. 29 Another possible public supply was the wine carried by the Gauloise 4 amphorae (Laubenheimer, 1985), which appears in great amounts in military sites in Britain and Germany. For instance, they are the majority imports together with Dressel 20 containers in the fort of Walheim until the cohort moved to other fort, as was recorded also at other British forts (Carreras and Funari, 1998).

Soldier’s diet on campagin31 The study of military supply has become quite popular subject in the recent years since most scholars realize its importance to understand the economy of Roman Empire as a whole. It is needless to say that a military supply is a 30 From Nero onwards, a series of exceptional posts were created to manage the military supply in wartime. They are discussed in detail by Remesal (1986, 96-108) in relation to the historic moment when they appeared. 31 This paper was published in 2006 in Ñaco and Arrayas (2006). The present article is an updated version of such paper published in 2006

124

THE MILITARY SUPPLY FROM THE LATE REPUBLIC TO THE EARLY EMPIRE PERIOD

quite complex subject, difficult to cover in all its diverse aspects. One of such aspects is its evolution through time.

Moving in hostile territories meant that Romans could not rely on local food supply, but they required external supplies organized in collaboration with local and central authorities (Erdkamp, 1995; 1998, 84-102). Cato’s famous quote “bellum se ipsum alet” (war feeds itself) (LIV. 34.9.12) should be understood in its right context in 195 BC, when the Roman army commanded by this proconsul in the Hispania Citerioris was supplied by allied civitas. However, the Republican army always attempted to organize carefully all details about supply when it was in enemy territories and closed off from any maritime port or river from which external supplies could easily arrive. For instance, T. Quintus Flaminius in the year 198 BC avoided going inland because he “was afraid that if he went far away from the sea and got into regions that were poorly tilled and barren, while Philip avoided a battle, lack of provisions would compel him to come back with his task undone” (LIV. 32.27.2).

There has been a few excellent works analyzing the military supply in the Republican period (Labisch, 1975; Erdkamp, 1998), which was a bit different from the one recorded in the Augustan period (Wierschowski, 1984; Remesal, 1986, Herz, 1988). As was mentioned by some authors such as Kissel (1995) and Roth (1999), one relevant difference was the nature of the Republican army, always on the move, compared to a more static one from Augustus onwards. It seems that there was a major change in the late Ist century BC, which affected the logistics of the military supply from production areas, distribution and quality control. The present paper attempts to shed some light on the reasons behind such change, relating it to the military situation of NW of the Iberia Peninsula.

As Kissel (1995) and Roth (1999) have pointed out, any Roman commander should take into account troop requirements, i.e. logistics. An army in continuous movement rarely has fixed headquarters except in winter (hibernia), its supply lines and suppliers always changing along the routes to the front lines.

The Republican Roman army was, at least in the last two centuries BC, a conquest force always in foreign lands. This constant expansion of the Roman borders meant that legions were never in a quite and stable headquarters, but in eventual establishments ready to march again. Therefore, the Republican army was difficult to supply since it was in hostile territories, so traders run always a risk when reaching their temporary camps.

Sometimes winter camps (hibernia) were far from the hostile territory and close to accessible supply points such as the Mediterranean coast. On occasions, legionaries were hosted by local populations (hospitium militare) who provided food and accommodation instead of the State (Ñaco, 2003, 223-231). Despite the economic advantages for the Roman State, such institutions generally led to conflicts between soldiers and locals that could end in revolts. Logistics can be understood as the science that studies military supply of food, weapons, medicines and equipment (i.e. clothes, tents, etc.), including the intricacies of acquisition, transport, management and consumption. The concept of logistics did not have an equivalent in Roman times, but other terms such as res frumentaria (CAES, Gal. 1.23, 1.39, 2.2, 7.3, 7.10), or synonyms such as copia, annonaria or commeatus (SAL, Jug. 46.5; CAES. Gal. 1.49, 2.5, 3.2; Civ. 3.49, 3.53). It is believed that around 90% of army requirements were food for men and fodder for their animals. Therefore, food and fodder supplies were the main worries for any good commander. When the Roman armies were marching on campaign, soldiers used to carry equipment and food for survival for at least one month. Livy (LIV. 43.1.8) said that legions of Gaius Cassius on their march from Iliria to Macedonia in 171 BC carried food for 30 days, as did Quintus Marcius Philippus in the IIIrd Macedonian war (172-167 BC) (LIV. 44.2.4) and Scipio Aemilianus in Numantia (134-133 BC). According to Frontinus (FRON. Str. 4.1.1) the soldiers of Scipio carried rations for only a “few days”. Other sources document that Pompey’s men at Ilerda

Figure 3.1.2 Simulacra Haltern

125

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

held rations for 22 days (CAES. Civ. 1.78) or ‘half a month’ (CIC. Tusc. 2.37).

Aquitani,33 a Gaulish tribe settled south of the Garonne. The isthme gaulois route, crossing from the Mediterranean to the gulf of Biscay following the Aude and Garonne, was therefore not reliable in terms of military supply.

This meant that soldiers, together with pack animals and wagons, should carry such baggage (impedimenta) for a hostile environment, where weight could hinder their speed and fighting capacity. If average rations were for 16 days, it meant that they consumed rations for 8 days on the outward journey and had then 8 days’ worth for the return to headquarters. If a day’s march could cover a maximum of 40 km, Roman units could cover a distance of 320 km from their headquarters.

Nevertheless, Strabo (STR. 3.4.18) mentions that the classis Aquitanica supplying the Augustan troops due to scarcity of grain34: “This was so much the case for the Romans in Cantabria that, although a proclamation was made that mice-catchers would gain bounties graded in proportion to the number caught, the Romans could barely come through with their lives; and, besides the plague, there was a scarcity, not only of other stuffs, but of grain too; and only with difficulty could they get supplies out of Aquitania on account of the rough roads”. No signs of Roman constructions of Augustan age in the gulf of Biscay are documented, despite the fact that indirect information may suggest portus AmanumFlaviobriga (Castro Urdiales) and sites around Santander were possible eventual ports for the Roman navy. It took some time for Augustus to have full control of the “mare externum”. As Augustus says in Res Gestae (26): “I extended the boundaries of all the provinces which were bordered by races not yet subject to our Empire. The province of the Gauls, the Spains, and Germany, bounded by the Ocean from Gades to the mouth of the Elbe, I reduced to a state of peace”.

Figure 3.1.3 Legionaries on Trajan’s column

Augustan campaigns may therefore have employed the inland supply route along the Ebro valley and “ruta de la Plata”, and perhaps also the Atlantic route, taking advantage of infrastructures already known after of years of military campaigning in Hispania. Augustus’ military headquarters were at Segisama (Burgos), close to the Ebro valley route as well as Herrera de Pisuerga, while other headquarters such as Astorga and León were also accessible by other routes such as the “ruta de la Plata” or the Atlantic route.

In the case of the NW Iberian Peninsula, mountainous ranges would limit troop speed to at least half the average. Therefore, soldiers may have covered only 20 km a day and a total of 180 km from their headquarters, with 8 days of rations. Looking at the pattern of distribution of military camps in Asturian and Cantabri wars (26 AC – 19 AC) (Morillo, 2006, 38, fig. 1), all the territory was covered by marching 160 km from the different military settlements. Astorga and Rosinos de Vidriales were the headquarters of X legio Gemina, Herrera de Pisuerga was the base of the legio IIII Macedonica, while the VI Victrix seems to have been settled in León and, perhaps Lugo (Morillo, 2006, 38). There was also a series of temporary and auxiliary forts (i.e. La Loma, El Cincho, La Poza…) covering the distance from Herrera to the coast of the gulf of Biscay.

When a legionary marched during a military campaign he was used to carry his own weapons (between 18-22 kg) (Stolle, 1914; Fuentes, 1991), clothes (around 3 kg) (Junkelman, 1997, 94-102), cooking set (vas) and food (cibus castrensis). Vegetius (VEG. Epit. 1.19) mentioned that soldiers carried 60 librae (20 kg) apart from the weapons (18-22 kg), therefore they bore 40 kg as a whole. It seems that Marius was responsible (at the end of the IInd c. BC) for the reform that reduced common baggage for units by increasing weight carried by soldiers. The legionaries enrolled at that time were called Marius’s mules (PLU. Mar. 13.1; FRON. Str. 4.1.7) (Fuentes, 1991).

It is interesting that all military camps were established on the southern side of the Picos de Europa, instead of along the coastal strip where supply would have been easier.32 The reason appears quite simple as the Roman army did not fully control the gulf of Biscay until 27 BC when Marcus Valerius Messala Corvinus routed the

Employing pack-animals meant a major effort to obtain animal fodder, less mobility for the army, and a longer 33

J. Caesar conquered the coast of Brittany routing the Venetii in 50 BC, but it seems that the Gulf of Biscay was not under Roman control. 34 Florus (Epit. 2.33.46) and Orosius (Hist. 6.21.4) also refer to the same intervention of the classis Aquitanica to solve some difficulties in supply.

32

Fernández Ochoa and Morillo (1994) have surveyed all the archaeological evidence of the Romans in the Gulf of Biscay, but there are only a few remains of the Augustan period. Sites at Coaña and Campa Torres record some Augustan material but only in small amounts, which cannot suggest any military supply.

126

THE MILITARY SUPPLY FROM THE LATE REPUBLIC TO THE EARLY EMPIRE PERIOD

baggage train that was easier for enemy forces to attack (IOS Bell.Iud. 2.63) (Menéndez, 2000, 174-175). Roth (1997, 77-83) calculated that each contubernium (8 legionaries) shared a tent, a hand-mill, tools and baskets (145 kg) carried by 2 mules. Fodder for so many animals may well lead to the sort of difficulties experienced by Mark Anthony in 36 BC, when he was forced to abandon some of his mules (PLU. Ant. 45, 4). And the need to feed mules obliged the army to send units into enemy territory to find food for them and lead to further difficulties, such as occurred in Perseus’ war in 171 BC (LIV. 42.27-31).

Cato refused external supply (LIV. 34.9.12), the army’s own (or subcontracted civilian) transport took items to warehouses and military camps. However if supplies arrived from other provinces the commanders asked the Senate to contract societates publicanorum to forward them as required (Malmeinder, 2002). Livy makes a few references to food contractors for the Roman army, including a dubious story involving publicani in the supply of the army in Hispania during the Second Punic War (LIV. 23.48.4-49, 4). In 215 BC, the commanders in Hispania urged the Senate to send clothing and corn from Italy to feed the army. Since the Senate lacked the economic means for supply, they asked civilians to give them credit. Nineteen individuals grouped in three societates publicanorum agreed on two conditions: exemption of military services and the assumption of risks of transport and hostilities (Erdkamp, 1998, 114).

Of course, marching legions in distant territories required external supply by sea, river or land transport. Supply by river and sea was faster, easier and cheaper, so it was the most suitable means whenever possible. As Erdkamp (1998, 61) points out, there were normally peaks in the demand of shipping to supply armies on campaign and so civilian vessels were sometimes requisitioned for this. For instance, Cato in the year 195 BC in Hispania, “sent a proclamation along the coast to collect ships of every kind” (LIV. 29.24.9; 34.8.5). Similarly Pompeius requisitioned ships to transport their 30 cohorts from Brundisium according to Plutarch (PLU. Pomp. 62.2).

Later, one of these publicani (M. Postumius from Pyrgi) was charged with fraud, either because he was supposed to have sunk cargoes for insurance payouts or to have pocketed money. When two tribunes started persecution the publicani closed ranks and the Senate took a firm line against them (LIV. 25.3.8-5, 1).

In Republican times an important part of the army’s supply came from overseas, so keeping these supply lines open and moving was a key issue. There were three major factors to this:

These two quotes suggest that the publicanii (via public contracts with Senate) were responsible for mobilizing goods to supply Roman armies. Perhaps they were the same men who were defined as traders (negotiatores, mercatores) in places such as Noviodunum (Gallia) and Vaga (Numidia), close to military camps. In addition they helped soldiers invest and trade their battle spoils.

a) Obtaining resources from other territories b) Mobilizing such products to army headquarters c) Quality control of the process

Such traders also carried other private goods for legionaries, such wine (in Dressel 1 vessels) or olive-oil (Brindisi amphorae). Archaeological testimonies of such military supplies are found in the remains of such amphora sherds recovered along supply lines such as the Ebro valley (Dressel 1 and Brindisi amphorae), the coast of Gallaecia (Brutus’ campaigns, 138-136 BC) (Dressel 1 amphorae), Hispania and Aude in Gallia (Dressel 1 and Pascual 1 amphorae).

The Roman State sought to obtain resources (in kind or cash) from provinces such as Sardinia, Sicily, Numidia and Hispania to supply her legionaries fighting in Greece, Asia or Hispania. Publicanii were the government functionaries responsible for gathering taxes (poll and land taxes) from local populations, as well as tolls and other dues – portorium (port tax), decuma (tithe) and scriptura (agistment) – from 123 BC onwards35 (CIC. Pomp. 6.15) (Cimma, 1981; Malmeinder, 2002).

With regards to quality control, there were no mechanisms as such within the administration to check the flows of supplies or quality. Once goods arrived at their final destination, quaestors (with the help of scribae and apparitors) logged the incoming and distribution of supplies.

During the Second Punic War, corn from Sicily was shipped to Tarentum where the Roman army was amassed (LIV 27.8.18). Corn was also shipped to the Roman army in the Second Macedonian War (200-197 BC) from Sicily (LIV. 29.35.1; PLB. 15.1), Sardinia, Sicily, Spain (LIV. 30.3.2) and from Numidia, Sardinia and Sicily (LIV. 32.27.2). Notwithstanding the fact that such supplies were sometimes requisitioned, as in Sicily in 205 BC by Scipio (LIV. 29.1.14), normally they were acquired through regular taxation or purchases.

In the last years of the Republic, it seems that the system collapsed. Perhaps the constant civil wars complicated the situation for partisan publicanii and so commanders could not rely on them for external supplies and had to look to allies or third-parties. For instance Pompeius complained continuously to the Senate because he did not have enough resources to combat Sertorius in Hispania (SAL. II.98). Pompeius received supplies from the Gaditanian L. Cornelius Balbus when his army was at Cartago Nova; this seems to have been why Balbus was

If resources were obtained in the same territory or province, such as in the case of Hispania in 195 BC when 35

Ñaco (2003, 117) insists that evidence from publicanii farming taxes cannot be later than 123 BC with the lex Sempronia Asiae.

127

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

granted citizenship in 72 BC by the lex Gellia-Cornelia (CIC. Pro Balbo).

of the early empire, but it varied over time depending on the means of preparing and consuming it. In the military camps soldiers used to have enough time to grind cereal as well as bake bread, however legionaries when marching carried buccellatum (biscuits) instead, as it was lighter and pre-cooked. Poliby (PLB. 6.39.13) reveals that a Roman soldier of the IIIrd c. BC received one Athenian medinnus of grain a month (850gr, as grain or bread a month, or 650gr of biscuit) (Roth, 1999, 43). Sometimes a legionary ate puls (porridge), adding water to cereal, when there was no time for anything else (J. Caesar, CAES. Afr. 67.2). Barley was only common in Republican times (PLIN. Nat. 18.15.94). It was also used in emergencies, such as Caesar’s blockade at Dyrrachius, and when Augustus’ troops were on the Dalmatian campaign in 34 BC (D.C. 41.47-51).

Excessive dependence on private societates publicanorum made the system unreliable in terms of the obtaining, mobilizing and controlling of supplies. Augustus’ reforms appear to have been a consequence of these late Republic setbacks. A legionary’s diet on campaign The best descriptions of a Roman legionary’s diet come from Late Roman Empire sources, such as lives of the Roman emperors described in the Historia Augusta (Davies, 1989). For instance Hadrian (HIST.AUG. Adr. 10.2) used to have meals with his soldiers, sharing the contents of the cibus castrensis – lardum (lard), caseus (cheese) and acetum (vinegar). Similar items appear in the description by Avidius Cassius (HIST.AUG. Avid. Cas. 5.3) when he forbade soldiers to carry any food on the expedition except lardum (lard), buccellatum (biscuit) and acetum (vinegar). A similar description appears in another quote from Historia Augusta (HIST.AUG. Trig.Tyr. 18.6-9), where he mentions the consumption of vinum (wine), lardum (lard) and wheat.

A second staple was a fat of some kind, mostly lardum (lard), as a (poor, though highly nutritious) substitute for olive-oil, which was preferred of course in military headquarters (Remesal, 1986; Carreras and Funari, 1998). It was a practical food in that it was a solid fat and easy to transport, whereas olive-oil required a container (amphora, barrel or skin) and had to be transported on campaign. Only in the Near East, where there is a lack of pigs, did legionaries sometimes eat other fats on campaign. Herod (IOS. Bell. Iud. 1.299), for instance, collected cereals, wine and olive-oil for his army near Samaria to get them to Jericho. However Aelius Gallus (STR. 16.4.24; PLIN. Nat. 11.239), on his expedition to Arabia, forced his soldiers to consume butter instead olive-oil, something that was considered a terrible hardship.

Early in 115 BC Appian (APP. Hisp. 54.227) mentions a range of items from the legionary diet, including wine, salt, vinegar, olive-oil, wheat, barley, hare, venison and game. However, the consumption of hare led to gastric problems for soldiers who were not used to it. Another interesting quote comes from Plutarch (PLU. Crass. 19.6), where he describes the defeat of Crassus against the Parths in 54 BC. His army consumed lentils and biscuits before the battle, which was supposed to be a bad omen.36

Acetum (vinegar) was a liquid that legionaries consumed mixed with water, as an alcoholic beverage. Apart from its calorific properties, it boosted courage before fighting. It also had medical properties and could help blood coagulate and be used for cleaning wounds (CEL. 5.26.21-24); in large quantities it made a crude anesthetic. Roman doctors also knew its properties as an antiscorbutic (with high vitamin C content). Scurvy was a miserable disease and was widespread amongst legionaries during the IInd Punic War. Cato (PLU. Cato 1.7) recommended consuming acetum on campaign.

Some references of the Roman armies in the IVth century AD, also record staples of the legionary diet. Vegetius (VEG. Epit. 3.3.9-10) refers to “frumenti vero et aceti vel vini nec non etiam salis” (wheat, vinegar, wine, salt) without a reference to any sort of fat. Finally, the Codex Theodosianus (7.4.6) mentions that in AD 360 soldiers consumed bucellatum, lardum, veal, wine and vinegar (Perea, 2006, 347). Of course, there was much more dietary choice when legions were in a permanent camp, as shown in the Vindolanda tablets and from other archaeological evidence (Bowman, 1994; Carreras, 2000, 170-186). Other interesting evidence of military consumption in times of peace, or when settled in garrisons, comes from papyri such as SB 6970, which refers to lentils, salt and vinegar; or P. Dura. 54, which lists the military festivals in Dura Europos where beef was the food of choice (Perea, 2006, 347).

Salt was also essential for the soldiers’ diet (Perea, 2006). It was used to preserve meat (i.e. laridum), as well as to keep water in hot conditions. Appian (APP. Hisp. 54.227) refers to the battle of Intercatia, in which Luculus’ army consumed hare and venison without salt. The result of such consumption brought about diseases amongst legionaries, some of whom died of dysentery. Vegetius (VEG. Epit. 3.3.9-10) proclaims salt as a staple for the army, together with frumentum (cereal), acetum (vinegar) and vinum (wine).

There were three staples for marching soldiers. The main campaign staple was cereal, preferably wheat in the days

Caesus (sheep or cow cheese) was another item of the soldier’s diet because of its energy giving and nutritive powers, as well as being easy to transport and preserve. A ‘comfort’ food, it was the staple of Roman farmers (VERG. Ecl. 2.1-13; 28-30; 45-73).

36

Plutarch underlines that both items were “funerary food”, so a bad omen for an army on campaign of course (Perea, 2006, 346).

128

THE MILITARY SUPPLY FROM THE LATE REPUBLIC TO THE EARLY EMPIRE PERIOD

Supply of staples during the campaigns against Cantabri and Astures (26-25 BC) Supplying armies in mountainous areas, such as the territory occupied by the Cantabri and Astures, generated particular difficulties, because little food could be obtained from the territory. According to Strabo, northern mountaineers had a completely different diet from the legionary: “All the mountaineers lead a simple life, are waterdrinkers, sleep on the ground, and let their hair stream down in thick masses after the manner of women, though before going into battle they bind their hair about the forehead. They eat goat’s-meat mostly, and to Ares they sacrifice a he-goat and also the prisoners and horses; and they also offer hecatombs of each kind, after the Greek fashion — as Pindar himself says, "to sacrifice a hundred of every kind." They also hold contests, for light-armed and heavy-armed soldiers and cavalry, in boxing, in running, in skirmishing, and in fighting by squads. And the mountaineers, for two-thirds of the year, eat acorns, which they have first dried and crushed, and then ground up and made into a bread that may be stored away for a long time. They also drink beer; but they are scarce of wine, and what wine they have made they speedily drink up in merry feastings with their kinsfolk; and instead of olive-oil they use butter.” (STR. 3.3.7).

Figure 3.1.4 Castro Urdiales patera

Of course no signs of such cereal supply are registered in the archaeological record, but Strabo’s quote (STR. 3.4.18) reveals that it was insufficient and they required an extra supply from the Classis Aquitanis. The Augustan campaigns against the Cantabrii and Astures involved more than 30,000 legionaries, belonging to 7 legions, and 20,000 auxiliary troops (Santos Yaguas, 1991, 90). Calculations of the grain requirements needed to feed all these troops during the campaigns reveal a total amount of between 10,000 and 24,000 tons38 (Santos Yaguas, 1991, 90), which probably placed extraordinary pressure on a territory that lacked agricultural surpluses. Such extraordinary demand in a very short time would have probably required different sources, and a special supply such as the one recorded by the Classis Aquitanis.

Following Strabo’s description, the Cantabrii and Astures consumed acorns instead of cereals, beer instead of wine, and butter instead of olive-oil. It seems that the supply of meat was less problematic since there were cattle, sheep and game available. With regards to salt, there were local sources of rock-salt. Strabo again: “Their rock-salt is red, but when crushed it is white” (STR. 3.3.7). The present-day town of Cabezón de la Sal (Cantabria) was exploiting its salt production since Roman times. It is said that the term “Cabezón” comes from an old Latin measure for salt.

Again, Strabo (STR. 3.4.16) makes it clear that conditions along the northern coast were extraordinarily hard:

Roman legionaries therefore should have been able to acquire their main staples (cereals, wine-vinegar, oliveoil) from nearby regions or external territories. Obtaining cereals was relatively easy since the nearby territories of Vettones, Vacceii and Celtiberians, in general, yielded large amounts of grain, as ancient sources reveal in the two centuries of the Hispania conquest (Ñaco, 2003). Cereals probably arrived from the eastern and southern areas near the territories of the Cantrabii and Astrures, and obtained by gathering taxes in kind from the civitates stipendiarii. Publicanii were probably responsible for collecting such taxes, under the supervision of procuratores37, who mobilized the cereal supply and controlled its quality.

“Iberia also produces quantities of those roots that are useful for dyeing. As for olive-trees, grape-vines, fig-trees, and the similar plants, the Iberian coast on Our Sea is richly supplied with them all, as is also a great part of the outer coasts. But the ocean-coast on the north has none on account of the cold, and, for the most part, the rest of the ocean-coast has none on account of the slovenly character of the people and the fact that they live on a low moral plane – that is, they have regard, not for rational living, but rather for satisfying their physical needs and bestial instincts – unless some one thinks those men have regard for rational living who bathe with urine which they have aged in cisterns, and wash their teeth with it, both

37 This magistrate appears in the Augustan period in Imperial provinces as the equivalent of quaestor in senatorial provinces. The procurator was responsible for finances in Imperial provinces and for supplying the armies. Strabo (STR. 3.4.20): “Caesar also has procurators there, of the equestrian rank, who distributed among the soldiers everything that is necessary for the maintenance of their lives”.

38 Garnsey (1983) suggests an equivalent of 300 modii of wheat per year (200 kg), while Davies (1971) calculates a total of 494.4 kg per year (3 librae a day).

129

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

they and their wives, as the Cantabrians and the neighbouring peoples are said to do.”

north of the river Tagus, in present-day Extremadura, there were olive groves in 147 BC.

With regards to meat, local sources were probably used for the supply of lardum, caesus, and salted meats. It seems that local economies, and even the farming activities of the armies themselves in their pratae, may have provided enough resources for the legionaries during the Augustan campaigns.

“Viriathus overran the fruitful country of Carpetania without hindrance, and ravaged it until Gaius Plautius came from Rome bringing 10,000 foot and 1300 horses. Then Viriathus again feigned flight and Plautius sent 4,000 men to pursue him but Viriathus turned upon them and killed all except a few. Then he crossed the river Tagus and encamped on a mountain covered with olive trees, called Venus’ mountain. There Plautius overtook him, and eager to retrieve his misfortune, joined battle with him, but was defeated with great slaughter, and fled in disorder to the towns, and went into winter quarters in midsummer not daring to show himself anywhere”. Appian (APP. Hisp. 64, 146)

However, the most difficult staple to manage for the army was olive-oil. The NW Peninsula was not suitable for olive trees, so it was imported from outside the region. Normally amphorae recovered in local excavations ought to provide information on the source of imports such as olive-oil. Nevertheless the presence of olive-oil amphorae is scarce in most archaeological sites in the NW Peninsula, and therefore cannot represent total oil consumption (Carreras, 1997; Carreras and Berni, 2003). Most olive-oil amphorae recovered in the NW are Baetican Oberaden 83 and Dressel 20, but their percentages cannot be compared to other military sites in Gaul, Britain, or Germany.

Later on, Roman colonies in the Extremadura fostered the development of villa exploitation in the countryside (Cerillo, 1984). Some of these rural settlements, known from partial excavations and surveys, may have produced olive-oil as well as other crops. Villae such as at La Cocosa (Badajoz) and la Dehesa de Torre-Aguila (Montijo) seem also to have produced olive-oil. The study of this villa at Torre-Aguila has produced a date for olive-oil production in the Ist century AD, although most contexts are dated later, from II-III century AD (Rodríguez Martin and Gorges, 1999).

It was thought, therefore, that some olive-oil may have been obtained from closer areas and transported to the NW in skins (utrii) or barrels. An initial hypothesis was that olive-oil came from the river Ebro in the Logroño and Navarra regions, where some villae with probable olive-oil presses have been recorded (Carreras, 1997). Such regions were in the army’s control and near the route that led to Segisama and Herrera de Pisuerga, as well as the port of Oiasso on the Cantabrian Sea:

Even further north of the Extremadura, in the present-day province of Salamanca, there is production of olive-oil in Sierra de Gata. The oil of this region was traded in the late XVIIIth century to the main northern markets, such as León or Zamora (Ringrose, 1972, 112). There is no evidence that olive-oil production in Sierra de Gata and other areas of Salamanca could go back to Roman times, but at least such territory could have provided excellent growing conditions for olive trees. The same applies for other modern olive-oil producing regions in Portugal, such as Proença-A-Velha (Idaha-la-Velha) close to the river Tejo, or Mirandela very close to León. However, there is no evidence of Roman olive exploitation, which is said to have been introduced by Visigoths and Muslims.

“Through these districts runs the road from Tarraco to those outermost Vasconians on the Ocean who live about Pompaelo, and about the city of Oeaso, which is at the Ocean itself – a road of two thousand four hundred stadia, reaching to the very frontier of Aquitania and Iberia” Strabo (STR. 3.4.11) Of course there were other potential sources for this olive-oil coming from the south. Morillo (1999, 325) argues that Baetican olive-oil may have come to the NW following the Plata route,39 a pre-Roman road going from Mérida to Astorga. However most amphorae from Astorga in the Tiberian period come from Baetica (Dressel 7-11 and Haltern 70), but only a few are oliveoil containers. If there were imports of Baetican olive-oil, they should have arrived by maritime routes, especially suited for amphorae, and following the same pattern as other amphora vessels.

Nowadays there are two varieties of olives produced exclusively in the Extremadura and Salamanca: Manzanilla from Cáceres and Verdial from Badajoz. Amphorae imports during the Augustan campaigns Most amphorae assemblages studied in the NW Peninsula cannot be dated to Augustan campaigns, but over longer phases involving pre-Roman layers, or later. With regards to the pre-Roman layers, the best example is the Campa Torres excavation, at which a mixture of local phases and Roman conquest layers have been uncovered (Carreras, 2001). The amphora assemblage was very poor in terms of quantities and variety, with a few Dressel 1, Oberaden 83, fish-sauce amphorae and Haltern 70 sherds. Perhaps the most interesting factor was the predominance of Haltern 70 vessels amongst other amphorae types.

Although the Baetican origin for the olive-oil is not feasible, there are other potential sources along the Plata route. Appian (APP. Hisp. 64, 146) indicates that further 39 The Plata route (Silver route) was a pre-Roman itinerary basically employed for transporting metals, such as silver, from the northern mines to the southern communities.

130

THE MILITARY SUPPLY FROM THE LATE REPUBLIC TO THE EARLY EMPIRE PERIOD

A second good example of later phases was the excavation at Blanco Cela, 21 (González, 1996). Between 1993 and 1995 there was a series of rescue excavations in Blanco Cela, 21, a site in the city of Astorga. The excavations provided a group of structures related to a V-shaped ditch, which appeared to be part of an early Augustan military camp dated around 26 BC. The ditch was filled at the beginning of the Ist century AD, when the area was redeveloped into a civilian establishment (González, 1996). The whole excavation covered an extent of around 110 m2, with structures of fossae perfectly evident, and, once abandoned, they were filled with contemporaneous material and thus provided a dumping area that covered all the previous structures.

Figure 3.1.5 Picture of Blanco Cela 21 (Astorga)

With regards to dating, fine wares provide an approximate dating to the second quarter of the Ist century AD. There is the presence of Italian Samian ware from the change of era, in planta pedis with stamps of Svavius Titius, Ateius, Romanus Rufus, Avellis and Felix (10 BC-AD 15/20). Dating can be confirmed by two coins of the Augustan period (Celsa) and Tiberius (AD 22/23-30).

mind that Asturica Augusta (Astorga) was an inland town, so the land transport of heavy containers, such as amphorae, was even more difficult. Perhaps this is the reason behind the low amphora density recorded (56,300 cg/m2), which is logical considering the distance of Asturica from the sea and navigable rivers, since amphorae were containers especially suited to these two means of transport.

In relation to the composition of the amphorae assemblage, the following table shows the proportions of each typology in quantitative terms. It must borne in

Within the assemblage, fish-sauce amphorae from southern Spain are the most common, together with Eastern Mediterranean Dressel 2-4. Basically, the

Table 3.1.1 Quantified table of the amphorae from Blanco Cela, 21 (Astorga) Ámphorae

EVE

Fragments

Weight

21

2470

Handles

Spikes

Density

Baetica Dressel 20 Dressel 28 Haltern 70

87

2245,45

1

300

79

2891

1

272,73

30

4730

5

228

18150

6

2

400

2

363,64

2628,18

Campania Dressel 2-4

4300,00

Oriental Dressel 2-4

75

Dressel 5 P&W66 Rodia

18

Southern Spain

4

16500,00

3

110

35

4410

18

4009,09

100,00

105

17300

6

15727,27

Dressel 10

37

2

440

400,00

Dressel 7-11

62

3

460

418,18

22

50

3850

23

1690

4

1960

15

19

2290

8

480

316

613

61931

Lipari Richborough 527 Gallia

3

3500,00 1536,36

Tarraconense Dressel 2-4 Local ?? TOTAL

131

1 1

1781,82 2081,82 436,36

41

6

56300,91

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

The highest finance rank in the province was the procurator, in our model the second level below the praefectus annonae. The procuratores headquarters were established at an early stage at Londinium, as demonstrated by the presence of the tomb of the procurator C. Iulius Alpius Classicanus (CIL VII, 30), which has to be dated after AD 60/1, as well as by the find of a tablet signed by proc. brit. dederunt (‘proclaimed by the procuratores of Britannia’) (Frere, 1987, 187). The early role of Londinium in the military supply seems quite remarkable. It provides the reason for the foundation of the city in the first place, and explains the high concentrations on this site of terra sigillata gallica (TSG) (Marsh, 1981) and olive-oil amphorae (Funari and Carreras, 1998, 28) during the first years of the conquest. Although it appears feasible that procuratores continued to reside in London, there are later inscriptions of these posts in centres of the military area (see figure 4), sometimes dating to times of conflict. Examples can be found at Inveresk (CIL VII, 1082), Risingham (CIL VII, 1003) or Watercrook (RIB 752); or even civilian sites such as Brampton (CIL VII, 875) (Birley, 1981, 287-299).

amphorae Dressel 10 and Dressel 7-11 are the best represented in the excavation. Other amphorae, that occasionally may have contained muria, is represented by Haltern 70, which is not especially common in Blanco Cela. However other sites from Astorga show Haltern 70 as the most common amphorae of this Roman town (Carreras and Piero, 2003). In this sense it must be remembered that an important concentration of Haltern 70 is documented at Braccara Augusta over the same chronological period (Morais, 2004). Perhaps some contents were transferred to other vessels to complete the landward journey to inland settlements such as Asturica. The predominance of Haltern 70 becomes most apparent in coastal settlements of the NW Iberian Peninsula, whereas its importance inland is not so overwhelming. On the contrary, fish-sauce amphorae are better represented inland than at the coast. At first sight, the composition of the amphorae assemblage from Blanco Cela 21 coincides with other assemblages from the NW Iberian Peninsula, with the only difference in densities of Haltern 70 depending on inland or coastal sites. Therefore, this excavation may have been representative of the amphorae circulation in the Atlantic façade of the NW Iberian Peninsula in a period after the conquest of the Cantabric and Asturian zone. Therefore it represents a clear military distribution, before starting the exploitation of military resources and construction of cities from new.

At a lower level in the model, beneficiarii and stratores intervened, of which a series of 22 and 2 inscriptions are documented in Britain. As can be seen in figure 4, their location coincides with the main centres of overland communication, above all in the region close to Hadrian’s Wall,40 on the routes towards the northern border – the Antonine Wall – and to the temporary Welsh frontier. Only in the case of Housesteads (CIL VII, 645) does the inscription indicate that the beneficiarius acted as portorium collector, although the location of the remaining ones suggest that they may have also controlled the overland traffic of military supplies (Holder, 1982, 74). In the case of Risingham (CIL VII, 996) it says that the place was the first statio, a point for toll collection. Two tablets also record the presence of beneficiarii in Vindolanda. One of them mentions a loan of an amount of grain (Tab. Vindol. II, 180), while on the other a foreigner asks for his intervention in an internal affair (Tab. Vindol. II, 344).41

This chapter has described the minimum requirements of a legionary, in terms of the staples needed for a fighter’s diet. Such basic requirements have been related to Augustus’ campaign during the Cantabrii and Astures campaigns so as to define possible supplies and infrastructures. Most land transport routes were not available at that time, but later requirements of supplying armies and transporting local gold would modify the infrastructures. BRITANNIA: A CASE-STUDY OF THE EXPLANATORY MODEL

On the frontier, the evidence shows the purchase of particular provisions in the local markets, such as the

Britannia was chosen as a case-study for this particular supply model; any model should be compared to a real situation in order to know whether it fully represents the complexities of the real world. The province of Britannia is a very suitable case-study for the analysis of military supply due to its insular condition and because of the huge bibliography on the subject. Epigraphic and archaeological testimonies can be used to test the model. The latter include well-published excavations of military forts, granaries and monographs on ceramic distributions, amongst which amphorae stand out (Middleton, 1979; Selkirk, 1983; Anderson, 1992).

40 Beneficiarii are documented at Dorchester-on-Thames (CIL VII, 83), Winchester (CIL VII, 5), Wroxeter (RIB 293), Lancaster (CIL VII, 2712), Catterick (CIL VII, 424), Binchester (RIB 1030; CIL VII, 424), Greta Bridge (CIL VII, 280), Lanchester (CIL VII, 411), Chesterholm (RIB 1696), London (JRS 52, 1962), Housesteads (CIL VII, 645; CIL VII, 691), Aldborough (Britannia, 1987, 373), Birdowswald (CIL VII, 824), Brough-under-Stainmore (CIL VII, p.313), Chester (RIB 505, RIB 545, RIB 532), Piercebridge (Britannia, 1986, 438-9), York (Britannia, 1979, 307), Risingham (CIL VII, 996) and Brougham (RIB 783); whereas stratores are located at Irchester (RIB 233) and Dover (Britannia, 1977, 426-7). 41 The second text is extraordinarily interesting since a foreigner, hominem transmarinum, interpreted as possibly a trader, was beaten by a centurio. He turns for help to the commanding officer, the praefectus, who is ill at that moment, and then to a beneficiarius and finally to another centuriones. The trader addressed the beneficiarius as the second-in-line, which may suggest that he acted either as a policeman or as official who, being the farmer of portoria or responsible for supplies, was close to the group of traders.

Following our hierarchical model of administrative posts related to military supply, it was relevant to analyse the distribution of the administrative staff in the province. 132

THE MILITARY SUPPLY FROM THE LATE REPUBLIC TO THE EARLY EMPIRE PERIOD

ones recorded in the Vindolanda tablets (Tab. Vindol. II, 190; II, 191; II, 192; II, 302; II, 343) (Bowman and Thomas, 1983; Bowman, 1994). In one of the tablets (Tab. Vindol. II, 302), the slave of the praefectus Verecundus is ordered to buy food on the local market. It is said that he should buy 100 or 200 eggs if they are of good quality, or 8 sextarii of muria and a modius of olives if the price is reasonable. Therefore, this tablet refers to a direct purchase.

perhaps corresponded to the stationes destined to collect portoria.45 After this initial stage, the official cargoes destined to the legions continued their journeys towards the limes britannicus, either by land or sea, with the collaboration of the Classis (Cleere, 1977). Land transport was undertaken by the military itself or by civilians. The Classis Britannica played a determinant role not only in military supply but also in shipping metals coming from state-owned mines such as Charterhouse on Mendip, Shropshire, Halkyn mountain, Derbyshire, Alton Moor or Dolancothi Weald. Therefore, return trips from the northern border could be used to carry public cargoes, such as metals from Welsh or southern mines.

With regards to the archaeological testimonies, there is a wide variety of products consumed by the army that come either from the southern parts of Britannia or from other provinces (Anderson, 1992, 42-101). In view of the low volume, these transactions were probably undertaken by individual mercatores, who would have transported the goods to the camp entrances (canabae).42 However the supply of other products, such as olive-oil, only mentioned in a small quantity in one tablet (Tab. Vindol. II, 203), and corn (York, Caerleon), which were basic staples (Dickson, 1989), suggests direct state intervention. The extraordinary amounts necessary, and the distance from Britain, necessitated intervention in terms of both the purchase in other provinces, as well as their distribution through the redistributive mechanism. In the particular case of olive-oil, the location of the Baetican Dressel 20 amphora sherds and stamps favours a study in detail of the internal supply network, since it shows a significant concentration in the military zone.43

From the south, the military supplies were mainly directed to four or five reception points in the north such as Carlisle, Corbridge, South Shields, York or Chester. These military ports register the highest densities of Dressel 20 amphorae in the north, since they became “breaking points”, or store and distribution centres, before reaching each final military unit. It must be born in mind that Chester records three inscriptions of beneficiarii, perhaps because the legionary camp was close to a river port, where metals from local mines were also shipped. Therefore, beneficarii could supervise the arrival of military supplies as well as the export of metals. York was a similar case, as a river port with a legionary camp, and records the presence of at least one beneficiarius.

On the basis of Dressel 20 amphorae densities and stamps from different Romano-British sites, it becomes possible to reconstruct the transportation of the amphorae from the calling ports in southern Britannia to the northern frontier (Carreras and Funari, 1998, 21-30). The Baetican amphorae arrived at the main ports such as Richborough, which was the headquarters of the Classis Britannica, London, Exeter or Colchester, where the highest concentrations of Baetican amphorae in the south are documented.44 It is more than probable that these were the places where the administrative staff from the procurator’s office undertook an initial control and

The last stage of the journey from the ports to the camps was undertaken by the military administration through beneficiarii located in the main centres of communication. The Vindolanda texts are explicit and show the continuous movement of people and merchandise from this fort to other military centres, such as Carlisle, Corbridge, Ribchester, Catterick, Binchester, Aldborough or York. As a whole, the supply network in Britannia can be summarised by the general model that identifies the different hierarchical levels, with the only exception of the top one, the praefectura annonae.

42 Most amphora types documented in the military zone respond to a pattern of trade of modest volume, and probably reflect private trade (wine, sour-wine, garum, olives) through a market exchange system (Carreras, 2000, 117-188). Apart from the presence of a foreigner (Tab. Vindol. II, 344), two accounts of a man called Gauus (Tab. Vindol. II, 192; II, 207) appear in Vindolanda. He may have been a trader supplying textiles and foodstuff. Other people in the tablets seem to have supplied products in small quantities (Tab. Vindol. II, 343; II, 196). 43 Dressel 20 amphorae predominate in military sites during the Principate in Britannia, reaching percentages in weight between 6090% of the overall amphora assemblage (Carreras and Funari, 1998). The amphora densities, as well as stamp distribution, coincides with troop movements in each period, and reveals a close relationship between Baetican olive-oil producers and the army. This relationship is not reflected in any other amphora type, with perhaps the exception of the wine-carrying Gauloise 4. 44 Higher densities of amphorae are interpreted as indicators of change in containers. This was done to facilitate the transportation of olive-oil to the final destination. Actually, a change of means of transport does not always represent a change in container, since Dressel 20 amphorae have been found on all Romano-British sites.

However, the system seemed to have changed in the Severan period, when less Dressel 20 imports are recorded in Britain (Carreras and Funari, 1998, 63-64). Inter-provincial military provisioning was no longer a military priority. On the contrary, internal supplies for all necessities were preferred in order to reduce transport costs.46 Some scholars argue that Septimius Severus modified the military diet and replaced olive-oil with bacon-fat (laridum), which could be obtained locally in 45 There is no documentation on the location of stationes in Britannia. However, the mentioned ports enjoyed the best possibilities. Only one port, Dover, which was the headquarters of the Classis from the midsecond century AD onwards, records the presence of a strator, and therefore the possible existence of a local statio. 46 The  tituli on Dressel 20 amphorae with the names of the Imperial family suggest that public transportation in the Severan period was in Imperial hands, so that the state saved some considerable money in vecturae.

133

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

all the western frontier provinces.47 As a result of these Severan reforms, the flow of supplies between provinces was an exception to the normal rule. Far from being surprising, the subsequent mentions of annona militaris

may therefore be interpreted as a reinforcement of provincial ways of supply instead of inter-provincial exchanges.

47

This is an interpretation of a passage in Herodian (HDN. 3.8.5). Davies, 1989a, 188.

134

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

3.2 SUPPLYING ARMIES IN THE IBERIAN PENINSULA DURING THE REPUBLIC P. Erdkamp LIVING OFF THE LAND

INTRODUCTION

All armies ‘lived off the land’ to some extent. They needed water and fodder, which only exceptionally and for brief periods was imported into the war zone. In addition, corn and other field crops were either harvested when ripe, or collected from the field where they were being processed, or taken from the stores and granaries. In much the same way livestock was gathered from the countryside, mostly by cavalry troops, for consumption by the whole army. Food was usually foraged only within hostile territory. In contrast, grazing, and often also the collection of fodder and wood, were nearly daily activities, regardless of the friendly or hostile nature of the territory. The advantages of sustaining one’s army in hostile territory by living off the land were that it reduced the costs, or rather shifted the burden to the enemy, and that it limited the amount of transportation that was required. In fact, as we shall see, living off the land was the only means available to sustain one’s army in isolated regions. Its main disadvantages, however, were that it exposed armies to enemy actions and significantly reduced one’s operational freedom.

Many are apt to picture Spain in their imaginations as a soft southern region, decked out with the luxuriant charm of voluptuous Italy. On the contrary, though there are exceptions in some of the maritime provinces, yet, for the greater part, it is a stern, melancholy country, with rugged mountains, and long sweeping plains, destitute of trees, and indescribably silent and lonesome, partaking of the savage and solitary character of Africa (Irving, 1995, 4). In 1829 a British traveller described the landscape of the Iberian Peninsula in these austere and dark, yet also romantic colours. The harshness of the Iberian landscape is clearly revealed in the accounts of warfare in the Iberian Peninsula over the ages. It was a land dominated by uninhabitable mountains, barren plateaus, unnavigable rivers, and land routes that could only be traversed by the omnipresent asses and mules. Waging war under such circumstances forced armies to operate on the fringes of the logistical possibilities. The frequent mentioning of problems of food supply in our sources on Roman wars in Iberia is therefore not surprising. The way wars were fought was determined by the geography and climate, and by the ecological restraints on agriculture and transport that were imposed by the environment. In fact, Rome’s difficulties in conquering the interior of the Peninsula may at least partly be explained by logistical factors. Nothing much seems to have changed in this regard between Roman times and the Napoleonic Age, when Spain was the theatre of war for British, French and Spanish troops. In his analysis of the military geography of the Napoleonic Wars Sir Charles Oman quoted an old saying: ‘In Spain large armies starve, and small armies get beaten.’ (Oman, 1902, 85). This chapter will try to determine to what extent this was true in Roman times (Labisch, 1975; Erdkamp, 1998; Roth, 1999).

Foraging was executed in a manoeuvre during which large numbers of soldiers, servants and pack-animals gathered food stuffs while usually being guarded by other troops. Both during the Third Punic War and the Numantine War, for instance, Scipio Aemilianus cautiously guarded his foraging parties with cavalry and infantry (APP. Pun.100; Hisp. 88). Unfortunately for the Romans, their Iberian opponents were excellent horsemen, while their light-armed infantry was well suited to attacks on foragers, thereby severely limiting the extent to which Roman armies could live off the land. Owing to their mobility, horsemen and light-armed troops had an important role in attacking foraging troops as well as in their defence. Foraging parties were either vulnerable when dispersed over a wide area, or when 135

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

legionaries carried food for twenty-two days in their army train, the light-armed troops had nothing. “As a result, large numbers of them were deserting to Caesar every day” (CAES. Civ. 1.78). Afranius tried to reach the military depots in Ilerda, but his army never made it there. Caesar’s army restricted their opponents’ movements so much that lack of water compelled them to surrender (CAES. Civ. 1.81).

returning laden with supplies. During his operations against Numantia in 134 BC Scipio Aemilianus was well aware of the dangers the enemy posed to his foraging parties: “Our troops will return from their foraging laden and tired, bringing with them animals and wagons and burdens. For this reason the fighting will be severe and unequal” (APP. Hisp. 87). Tactical considerations compelled the frumentatio to be executed in as concentrated an area as possible. Therefore, as long as conditions allowed, foraging would be aimed primarily at stores in hamlets and villages, and at locations where soil and drainage offered favourable conditions for arable farming.

Apart from the tactical weaknesses inherent in living off the land, it exposed armies to changes of the landscape. Polybius (PLB. 35.1) wrote that only the approach of winter made an end to the fierce fighting in Spain. Bad weather made land and sea transport in the autumn and winter nearly impossible. Most importantly, the commencement of fighting was only made possible by the new harvest. Before harvest last year’s stores were nearly empty. Foraging missions therefore had to cover a wide area to sustain the army, while stores were easily moved into walled towns and thus out of reach of armies passing by. Caesar (CAES. Civ. 1.48) writes that Fabius’s troubles were increased because off the time of year: there was none of last year’s grain in the stores, while this year’s was not yet ripe. The Pompeians had carried most grain to Ilerda, and Fabius’s troops had eaten all that had been left.

In the Iberian landscape good opportunities to forage and graze the animals were often limited to certain areas. Frontinus (FRON. Strt. 2.5.31) describes such a situation regarding Pompey’s operations against Sertorius. “When Pompey was near the town of Lauron in Spain, there were only two tracts from which fodder could be gathered, one near by, the other farther off.” Sertorius uses these circumstances to his advantage when ambushing and destroying a large troop returning from foraging. The requirements of a large army would soon exhaust the immediate surroundings. Inevitably this meant longer foraging routes, and thus the increased threat of attack. In short, the collection of food and fodder was by no means a marginal affair.

Unlike Caesar, who was in a hurry to eliminate the Pompeian troops in Spain in order to deal with Pompeius himself in the East, most generals who expected their army to live off the land would wait for the crops to ripen. Cato the Elder’s campaign in 195 BC offers a famous example:

A good example of the tactical significance of foraging is provided by Caesar’s account of the campaign conducted in Spain by his legate Gaius Fabius in the spring of 49 BC: “He had built two bridges over the river Sicoris, two miles apart, and sent men across them to forage, since he had used up all the supplies of fodder on the near side of the river during the preceding days. The commanders of the Pompeian armies were doing much the same thing, and for the same reason, and there were frequent skirmishes between their cavalry forces. On one occasion, two of Fabius’s legions had gone out, as was the daily custom, to guard the foragers. They had crossed the river by the nearer bridge and the wagons and all the cavalry were following, when there was a sudden squall of wind and a rush of water that broke down the bridge and cut off a good part of the cavalry”. (CAES. Civ. 1.40)

“It happened to be the time of the year when the Spaniards had their corn stored in the barns. Cato forbade the army contractors to supply any corn to the troops, and sent them back to Rome with the remark, "War feeds itself." Then, advancing from Emporiae, he laid the enemy’s fields waste with fire and sword, and spread terror and flight in all directions”. (LIV. 34.9.12-13) The plundering of food was not limited to crops in the field. After harvest the soldiers had even better access to the crops as they were being processed on the farms. Cato had deliberately timed the start of his campaign to the time of year when conditions for living off the land were optimal. Also Q. Metellus Macedonicus started his tenure as provincial governor with a campaign at harvest time against the Vaccaei, who lived in the north-west of the Peninsula (APP. Hisp. 76). We may be sure that both commanders deliberately delayed the start of the campaigning season until after the crops had ripened because of the advantages this offered for the Roman army’s food supply.1

The Pompeian commanders promptly attacked the foraging party and the two legions, but Fabius’ forces were saved just in time by the arrival of reinforcements. Again, we see that the gathering of sufficient fodder, the transportation of which required the use of wagons, led to frequent fighting and potentially to full-scale battle. Avoiding battle under such conditions was simply impossible, and this is how Caesar eventually defeated the Pompeian army in Spain. Using the superior strength of his army, he made it impossible for his opponents to sustain their troops off the land. “Afranius’s men were finding it difficult to secure supplies of fodder and water”, Caesar (CAES. Civ. 1.78) writes, and while the

In short, living off the land seemed a straight-forward and cost-effective way of sustaining an army, but it brought 1 Glover (1963) 260 says that the time of year to live off the countryside during the Napoleonic Wars in Spain was at the time of harvest or just afterwards.

136

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

Caesar’s army draw provisions from various towns in 45 BC (CAES. Hisp. 26). Clearly ad hoc was the requisition of livestock by Caesar’s army in 49 BC to make up for the failing grain supply (CAES. Civ. 151-52).

with it its own set of interrelated disadvantages. In the first place, an unavoidable condition for successfully living off the land was tactical strength. Owing to their mobility, horsemen and light-armed infantry were equally important in attacking and defending troops that collected food, fodder and water. While Roman superiority on the battlefield largely relied on their heavy infantry, in Spain they were often confronted with native opponents who boasted good cavalry and light infantry. In the second place, the military calendar of soldiers who lived off the land was largely determined by growth cycles of the major crops. The circumstances for foraging were best at harvest-time or just afterwards, when crops were being processed on the farms. During winter and in the months before harvest, fields were bare, stores were depleted or taken into walled towns, and accessible resources were so thinly spread over the countryside that foraging parties had to cover impractically large areas.

In short, the military requirements of grain as well as money, which were at first fulfilled by direct demands, were gradually institutionalized into money-taxes and levies-in-kind, which were supplemented by ad hoc levies in times of war. As early as the final years of the Second Punic War, supplies from Hispaniae were occasionally even sent abroad, but it would be wrong to assume that the Iberian Peninsula sustained Rome or overseas armies in any regular way before the days of the Principate. To be sure, grain (and clothing) was sent from Sicily, Sardinia and Spain in order to support the Roman troops invading Africa at the end of the Second Punic War (LIV. 30.3.2). At about the same time grain was sent from Iberia to the city of Rome and distributed at 4 asses per modius to the population by the aediles. We may compare this to two shipments of grain that were sent from Africa in the years 201 and 200 BC and that were also sold cheaply to the city’s populace (LIV. 30.26.5f; 30.38.5; 31.4.6; 31.50.1). These shipments to Rome have to be seen in relation to the military circumstances in Spain and Africa at the time and were the result of windfalls arising from the end of fighting. Army supplies became available for overseas consumers as troops were discharged and tributes demanded from the vanquished peoples increased. Hence, such shipments should be seen as fortunate, but unplanned side-effects of the military food supply in Africa and on the Iberian Peninsula.

IMPORTS AND EXPORTS Roman commanders preferred the security and flexibility of organized provisioning. The strategical advantages and security offered by organized provisioning and longdistance supply urged the Romans to gain access to food as directly and strongly as possible. Hence, the Romans created governmental structures that siphoned off resources from the subjected communities to their armies. Whether acquisition involved ad hoc requisitions or regular taxation-in-kind depended on the geographical and military context. In his analysis of the development of the Iberian Peninsula from a theatre of war into regular provinces, John Richardson has shown that in Hispaniae the complex system of taxation emerged from the ad hoc levies demanded by the Roman commanders in order to pay and feed the Roman army during the Second Punic War and afterwards (Richardson, 1976; 1986). Tribute was demanded from defeated peoples, as when in 205 BC the Ilergetes were ordered to supply six months’ of grain to the Roman troops (LIV. 29.3.5). These ad hoc levies were supplemented with contributions from Rome’s local allies. In 171 BC we find the first mention of the levy of a regular half-tithe (vicensuma), when the local communities complained about fraud concerning the conversion of grain levies into payments at an unfavourable rate (LIV. 42.2.12). The regularisation of the grain levies may possibly date from the reorganisation by Ti. Sempronius Gracchus in c. 180 BC of the relations between Rome and the subordinate peoples in what by now were the two provinces of Hispania Citerior and Ulterior.

Roman wars on the Iberian Peninsula required imports from abroad from the start, and in fact, continued to do so until the first century BC. Not surprisingly, as soon as the first Roman army was sent to Spain, we also see the first mention of shipments of grain. In 218 BC, Livy (LIV. 22.11.6) tells us, grain ships that had left Ostia westwards were captured by a Punic fleet. We are also told that in subsequent years the Roman generals who operated in the Iberian Peninsula ardently demanded grain shipments (LIV. 25.37.7; 26.2.4). The historiographic context of these passages may not inspire much confidence, but this makes the importation of military supplies not less likely. The strongest evidence for structural reliance on imports pertains to cases claiming that outside supplies were not deemed necessary. As we have seen, Cato as army commander in Hispania Citerior in 195 BC sent back the suppliers of grain because he thought that at the time the circumstances were right for the war to sustain itself (LIV 34.9.12). Similarly in 181 BC the Senate was informed by the commanders that in that year the usual supplies of grain and money were not necessary, because the successes in the province ensured sufficient supply from local resources (LIV. 40.35.4). Just as in 203 BC victory in the Iberian war zone had allowed grain to be shipped to Africa and Rome, in 195 and 181 BC it allowed the provinces to be self-sustaining. Both passages clearly indicate that this was not normally the case. The need for

Although it is impossible always to distinguish between regular (annual) taxation, ad hoc levies, and enforced purchases, it is clear that in the first century BC local communities still contributed grain and other provisions to Roman armies that were garrisoned and campaigning in Spain. We see for instance that when some of the local communities defected to Sertorius, the governmental troops experienced problems with their supplies (SAL. Hist.2.47.5). The defecting communities had probably been supplying government troops with food. Also 137

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

offered them little booty. Therefore no sutler followed them, and no long train of pack-animals followed the army column.” One might suspect that the Iberian Peninsula may have solicited similar comments, the more so in view of the lack of Roman success in some of Rome’s Iberian wars, the poverty of some isolated regions, and the lack of booty in regions disturbed by continuous fighting. Nevertheless, Roman soldiers were apparently corrupted by the presence of sutlers (local merchants), prostitutes and other camp-followers even during the difficult and unpopular campaigns that were fought in inland Spain during the 150s and 130s BC. This may be deduced from the fact that, in preparation for his operations against Numantia in 134 BC, Scipio Aemilianus forbade the presence of traders, prostitutes and fortune-tellers. He made an exception though for those who sold plain soldiers’ food. His disciplining of the army included regulation of the kinds of food allowed at the various times of day, and the utensils the soldiers were allowed to cook with (APP. Hisp. 85; PLU. Mor. 201c; FRON. Strat. 4.1.1). The point of these measures was primarily to abolish unnecessary luxuries and thereby reduce the army train. Despite occassional disciplinary measures, however, sutlers and other camp-followers may be regarded as a normal component of a Roman army of these days. From a military viewpoint one may even argue that such traders offered a vital opportunity to the soldiers to get rid of bothersome masses of booty. However, beyond supplying the soldiers with wine and luxury food stuffs, sutlers and petty traders played no significant role in the food supply of the Roman armies.

imports at this stage is not surprising. From 187 BC onwards, four legions (i.e. approx. 40,000 men including allies and non-combatants) were permanently stationed in the provinces of Hispaniae. This was a significant part of Rome’s total armed forces at the time. The extent to which such a force, which lived off the land in hostile territory only during short and irregular periods, could live off the resources of the province was surely limited. Winter quarters and garrisons clearly exceeded the carrying capacity of their immediate hinterland. If Rome wanted to avoid overburdening its local allies, it had to shift part of the effort to regions that saw less fighting. Sicily and Sardinia were indeed annually contributing tax grain to Rome since the days of the Second Punic War. Both islands paid an annual tax of ten percent of the harvest, while an additional tenth was levied during Rome’s wars against Antiochus III and Macedon, part of which went to the armies there. Even though the sources keep silent on the destination of Sicilian and Sardinian tax grain in all other years, we may be sure that much, if not all of it went to the armies that were permanently stationed in the Hispaniae and in Gaul. In fact, even the war against Sertorius necessitated imports from the neighbouring province of Gallia Narbonensis (CIC. Font.13). It was only during the Principate that we have clear evidence of Iberian provisions, mainly olive oil and grain, reaching the armies abroad on a regular basis (D.C. 60.24.5; TAC. Ann. 1.71; Hist. 2.32). SUTLERS AND CONTRACTORS

Because of the inadequacy of the Roman bureaucracy, it is often supposed that they relied on contractors to supply the armies with essential provisions (Roth, 1999, 230231; Ñaco, 2003). The role of large-scale contractors in the military grain supply is based largely on a dubious story which involves publicani in the provisioning of the Spanish army during the Hannibalic War (LIV. 23.48,449,4 and 25.3,8-5,1). In the year 215 BC the commanders of the army in Spain informed the Senate of a shortage of money, clothing and corn. They would try to get money themselves, but clothing and corn had to be delivered from Italy, if the province were not to be lost. Lacking the necessary resources the Senate decided to appeal to those who had made profits before from state-contracts to deliver the required supplies to the Spanish army, and nineteen individuals in three societates were willing to supply provisions on credit. Some years later one of these publicani – M. Postumius from Pyrgi – was charged with fraud. Together with a colleague – T. Pomponius from Veii – he had deliberately sunk worthless ships or pocketed money for non-existent vessels and cargoes. The accusations and the ensuing trial led to a political row. At first the Senate was reluctant to take action, because it was afraid of offending the class of public contractors. When two of the tribunes started prosecution nonetheless, the publicani closed ranks and at first tried to obstruct their action using the veto of one of their colleagues; later they resorted to simple violence. As a result, the Senate took a firm line with the publicani (LIV. 25.3, 8-5,1).

In 178 BC a Roman army on its way to fight the Histrians struck camp on the coast near the colony of Aquileia. Livy (LIV. 41.1.2-5.3) describes in detail how people gathered from the surroundings to offer their wares for sale to the troops. A market was established on the beach, where individual soldiers bought food and other goods. Two brothers arrived from Aquileia with wagons loaded with supplies (commeatus). The reason for Livy’s detailed description is that the Histrians took the Roman troops by surprise, captured the camp with all its luxuries, but were defeated when intoxicated by wine and good food. Although the story is a typical late-annalistic tale – we are supposed to believe that the rumour of the capture caused an emergency levy throughout Italy, even though the Romans claimed victory before nightfall of the same day – the fiction pertains to the Histrian surprise attack, not the creation of a market by traders and farmers who were attracted by the buying power of a nearby army. In other words, the situation reflects what a first-century BC author deemed a likely occurrence. Although we have no such evidence for the Iberian provinces, there is no reason to assume that similar markets did not emerge near Roman armies in friendly territory there. At all times armies attracted multitudes of campfollowers, especially those fighting successfully in rich areas. Things were different in poor regions. Livy (LIV. 39.1.6-7) tells us about Liguria that it is “a poor region, which compelled the soldiers to a scanty way of life and

138

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

The details provided by Livy are extremely doubtful. The most serious criticism is that the story contains many anachronistic elements. The Senate, we are told, was reluctant to take action, because it was afraid of offending the class of public contractors. The relationship between the Senate and the publicani, as portrayed by Livy in this case, is in stark contrast to Polybius’ second-century BC description of public contractors as dependent on the Senate (Badian, 1972, 45; Brunt, 1976, 178). The author on whose account Livy (directly or indirectly) derives was clearly hostile to the class of the wealthy publicani, who had by the late second century BC acquired a powerful but disputed position in Roman politics. Moreover, Polybius (PLB. 6.17) illustrates the great importance of these societates with examples of contracts in several domains: “All over Italy an immense number of contracts, far too numerous to specify, are awarded by the censors for the construction and repair of public buildings, and besides this the collection of revenues from navigable rivers, harbours, gardens, mines, lands – in a word every transaction which comes under the control of the Roman government – is farmed out to contractors.” Contracts to supply corn to the Roman armies are not among them, a sate of affairs more easily understood if the food supply, which would have been a huge assignment, formed no part of the contracts.

fed away from the coast. If the troops were to profit from external supplies, either the troops had to move to the supplies, or these had to be moved to the armies. The extent to which the armies managed to live off external supplies is crucial for our understanding of Roman warfare on the Iberian Peninsula. This in turn requires us to have a deeper look at the conditions of land and river transport in this part of the Mediterranean and at the distribution system devised by the Roman army. Transportation in the Iberian Peninsula was complicated by the fact that it is a vast land mass, the centre of which is separated by hundreds of kilometres from the sea. More-over, the Iberian land mass is extremely mountainous, which not only hampers all communication over land, but, in combination with the climate, also makes the rivers quite useless for the purpose of transportation. The rivers are not only steep, they are also governed by extreme seasonal fluctuations. Rainfall in the summer months is very low. To make matters worse, owing to the extreme heat most precipitation evaporates, so that even the larger rivers carry little water at this time of year. There are virtually no navigable rivers, even for small boats, except for the lower reaches of the Guadalquivir and the Ebro. As Sir Charles Oman observed in his study of the Napoleonic wars, “it is a land where the rivers count for little, and the hills for almost everything, in settling military conditions.” (Oman, 1902, 75).

In a later instance, however, Livy is thought to have provided corroboration of the theory, when he mentions Cato’s measures regarding the food supply of his troops in Spain in 195 BC. As has been mentioned before, in 195 BC Cato sent back the supplies, because he wanted to emphasize that he would make the war sustain itself. Some scholars see this as the dismissal of the publicani, who had been contracted for the provisioning of food to Rome’s Iberian army.2 The text, however, does not allow only this interpretation, for ‘redemptor’ may just as well have the meaning of ‘supplier’ or ‘entrepreneur’ in general or of ‘transport contractor’. Hence, Livy (LIV. 34.9,12) provides no evidence for a supposed role of publicani in the military food supply.

The means of transport during the Peninsular War were hardly any different from those of the Roman wars. Mules and animal-drawn carts still managed most of the transportation over land; river boats were still towed upstream by man or animal; the capacity of carts or boats had not significantly altered, nor their speed. Because virtually all overland routes consisted of dirt tracks that were only passable when the ground was dry; prolonged and heavy rainfall such as normally occurred in the Iberian winters severely impeded overland travel and made the use of wagons virtually impossible. Moreover, many important communication routes ran at right angles to the rivers and across watersheds, which meant that one often had to cross rivers and difficult passes. In few places, a modern study notes, one can travel more than fifty miles without having to climb a mountain (Ringrose, 1970, 4). Due to these conditions, ninety per cent of all transportation in the eighteenth century was done by ass or mule (Ringrose, 1970, xxi). As an example of the seasonal problems of overland transportation we can point to Caesar’s springtime campaign in Spain in the year 49 BC, when

Summing up, we have to conclude that hard evidence for large-scale contracts to supply corn to the Roman armies is completely lacking. There is no mention of publicani as contractors of the military grain supply besides the dubious story in Livy’s books 23 and 25. The lack of any evidence for the much better documented period of the civil wars pointing in this direction is also more easily understood when assuming that there was no role of large-scale contractors in the acquisition of food supply on behalf of the Roman armies.

“a storm of such intensity sprang up that it was agreed that there had never been a greater rainfall in that district. On this occasion it washed down the snow from all the mountains, overtopped the banks of the river, and in one day broke down both the bridges which C. Fabius had made. This caused serious difficulties to Caesar’s army. [...] The states which had entered into friendly relations with Caesar could not supply provisions, nor could those who had

ROADS AND RIVERS While the Romans easily managed the shipment of supplies from Italy, Sardinia, or Sicily to the Iberian harbours, the real problems arose when armies had to be 2 Briscoe (1973) 70, even refers to Cato’s conflicts with publicani during his censorship. Cf. Badian (1972) 28; Brunt (1976) 210; Richardson (1976) 150f; (1986) 93.

139

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

travelled some distance for forage return, being cut off by the rivers, nor could the huge supplies which were on their way from Italy and Gaul reach the camp”. (CAES. Civ 1.48).

to reckon with a daily consumption of approximately 60 tons. In other words, the daily consumption of grain alone required 600 mules or 200 mule-carts or 100 ox-wagons, to which – under favourable circumstances – must be added 120 mules or 32 carts or 16 wagons for the amphorae of wine and olive oil. While the demand for operational flexibility called for the use of pack-animals in the army train, which had to move across all kinds of terrain, the supplying of the depots over land allowed the use of ox- or mule-drawn wagons. Whenever possible, the shuttle system between depot and army also made use of river boats. During his campaign in 49 BC Caesar shifted the burden of transporting supplies to the civilian population. When local peoples surrendered, Caesar demanded that they assisted his army with grain. Hence, he tells us, they collected pack animals from their territories and moved grain supplies to his army’s camp. We may compare this to a situation during the First Punic War, when nearby communities supplied the army depots at Herbessos, which the Romans had established on behalf of their army during the siege of Agrigentum.4

It was not so much the costs involved in land transport that hampered the Roman war effort as the sheer impossibility of hauling large amounts of provisions over long distances. However, warfare could not be limited to those regions that lay within easy reach of sea and river transports. In Spain, for instance, this would have limited warfare to the coastal zone and the hinterland of the lower courses of the Guadalquivir and the Ebro. The limitations of river transportation in the Iberian Peninsula inevitably increased the importance of local resources.

ARMY DEPOTS AND SUPPLY LINES Because the provisions that arrived from overseas or that were collected locally could not be hauled along by the army, the supply system required three phases: transportation to the supply depots, transportation from the depots to the armies, and the army train itself. To get an idea of the quantities involved, let us take as example an army of 40,000 soldiers and non-combatants, 4,000 horses and 3,500 mules, which would consume more than 1000 tons of wheat and 800 tons of barley monthly, not counting green or dry fodder, and wine, olive oil, or other food items. Despite the seeming impressiveness of these requirements, limited numbers of freighters could easily handle such amounts. The biggest ships from midRepublican times reached a size of over 1000 metric tons. However, the regular freighters in operation during the Republic were more in the range of 100 to 200 tons. Ten or fifteen of such freighters were sufficient to hold the monthly provisions for an army of 40,000 men. Supplies had to be shipped during the Mediterranean sailing season, since the increased risks at sea during autumn and winter provided a serious obstacle to the overseas shipments. The fourth-century AD military handbook by Vegetius gives the period of May 27 to September 14 as safe, and from the latter date to November 11 as “doubtful and more exposed to danger”. From November 11 to March 10 the sea is closed: mare clausum. Assuming that ships sailing from Sicily, Sardinia or Italy managed four return trips during the sailing season (but also taking into account some spoilage and losses), 30 to 50 freighters were sufficient to ship all the food supplies that were annually needed by the entire army.

Supply depots and winter quarters were mostly located on the coast, where they could be supplied by seafaring vessels, or on navigable rivers, where the resources of the hinterland could also be gathered with ease. During their operations on the Iberian Peninsula in the early years of the Second Punic War, the Scipio brothers are said to have regularly collected grain stores in Castrum Album, which Livy describes as a fortified place (LIV. 24.41.34). We may assume that Carthago Nova functioned as the Barcids’ depot, where supplies from Carthage were stored. Hence, the future Scipio Africanus dealt them a significant blow when he captured the place with all its stores. Supply depots were not only used in a situation of long-distance provisioning: they could also be stocked from the surrounding area. In 153 BC, the second year of renewed fighting in the region, the Roman general Nobilior first made a fruitless attempt to capture the town of Axinium, where the Celtiberians had collected stores of food, and then laid siege to the town of Numantia. The town of Ocilis meanwhile harboured the Roman depot of provisions and money. A string of failures caused Ocilis to defect to the opponents, thus compelling the Roman army to abort the siege and go into winter-quarters. Since it was rather late in the year and Nobilior had lost his provisions, his troops suffered from dearth during the winter (APP. Hisp. 47). When preparing the siege of Pallantia in 136 BC Aemilius Paulus “collected provisions in a place that he fortified”.5 When the siege turned out to be protracted the supplies of food and fodder, which the Romans must have gathered locally, soon failed (APP. Hisp. 81-82).

Logistically much more challenging was the transportation over land. A mule could carry about 100 kg; a light mule-drawn cart managed about 300 kg and an oxdrawn vehicle 600 kg.3 Continuing our example, we have

the Peninsular War the maximum allowed load of an ox-cart of the commissariat was some 275 kg., while the maximum capacity is estimated at 1000 lb. (= ca. 455 kg). 4 PLB 1.18. Cf. APP. Pun. 18: the magazine in Tholon, serving the army besieging Utica in 203 BC, was captured by Syphax of Numidia. 5 As Roth (1999, 51) notes, lack of a reliable walled town forced Aemilius Paulus to built a fortified place as safe depot of his army’s supplies.

3

The estimated capacity of mules in early modern times varies somewhat. During the Peninsular War some 100 kg. (200-230 lb. = 91-104 kg.) seems to have been usual. Ward (1957) 86; Glover (1977) 110. This is probably a fairly close comparison to the ancient world. Ringrose (1970, 39), on the other hand, is much lower with roughly 60 kg. Ringrose (1970, 39), estimates the capacity of a Spanish ox-cart as approx. 250 kg. During

140

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

Convoys from the army depots, or otherwise regular visits to the depots, supplied the army with the necessary supplies. We have seen in the passage quoted above about the problems arising in his army due to springtime flooding that Caesar expected provisions from Gaul and Italy to reach his troops. However, it will be clear that logistical considerations imposed a maximum distance over which supplies could reach the army. Apart from the size of the army, the distance between army and supply depot was the most important factor in determining required transport capacity. The distance that could be managed in a day varies considerably for mule or ox-cart. The early modern data indicate that oxen could not manage more than 25 km a day. Mules, on the other hand, could be relied upon to travel up to 70 or 80 km a day. On continuous duty in the transportation of heavy loads, their average, however, must more realistically be lowered to some 40 km a day. Each animal would need fodder in addition to its grazing during the trip. In other words, they would eat part of their cargo (in the case of pack-mules, during 5 days some 10 % of its load). At a distance of 100 km between army and depot, the shuttle system required some 930 ox-carts, drawn by approx. 3,700 oxen, continuously making an 8-day round trip. Alternatively, about 4,000 pack-mules would be necessary, making a 5-day round trip. Along routes that were used regularly, stores of green fodder, hay and straw might be prepared either from military supplies or from stocks that were requisitioned or plundered from the civilian population. If all green and dry fodder had to be carried by the animals for the duration of the trip, the range would be much smaller, since the animals would consume their load within a few days. We may conclude that the shuttle transport over a distance of 100 km between a sizeable army and its depots required 3,000 or 4,000 mules. Above 100 km the required number of animals and their consumption of fodder would soon reach impractical levels.6 In other words, the number of animals and their feeding constituted the major limitation of the range between supply depots and army.

We may conclude that logistical restrictions severely limited the extent to which armies operating in Spain could profit from supplies shipped from abroad. Under the best of circumstances the range of these supplies went little beyond 100 km from the coast or the lower stretches of the rivers. Supply depots and winter quarters were located at the logistically most convenient places. However, campaigns further inland required armies to go beyond this range.

ARMY TRAIN Because of reasons of flexibility, the army train made use almost exclusively of pack-animals. Wagons and carts were primarily used for the movement of goods to and from the supply depots. In addition, material too heavy for one mule, like siege equipment, had to be moved in wagons. The troop-train, however, which must be distinguished from the wagon train, had to be able to follow the army in rough terrain. For that reason Scipio Aemilianus forbade the use of wagons by his troops during his campaign against Numantia in 134 BC (APP. Hisp. 85). Was it possible to carry enough food in the army train for a month? At 30 days the effective load of a pack-animal would be reduced to approx. 40 kg., since it would need to carry 60 kg for its own consumption. It would therefore require the impractical number of 40,000 mules – consuming more than the rest of the army – to carry the entire supplies of wheat and barley for a month. Even in ten days the men and horses would consume some 545 tons of grain, while the effective load of a pack-animal would be 80 kg. The required number of mules would still be 7,000. Only at 5 days would the number of mules be reduced to 3,000. It will be clear that it would have been impossible to carry supplies of straw or hay in the army train, and that in order to take along food for more than a few days, the men themselves had to carry part of it. The auxiliaries of the Pompeians, Caesar (CAES. Civ. 78) notes, were physically unaccustomed to carrying large amounts of food, unlike the Roman legionaries, it is implied. Assuming that Roman soldiers and servants carried their own food for 8 days, this would reduce the rations to be carried by pack-animals to approx. 265 tons. Carrying food for ten days in such a case would nevertheless require 3,500 pack-animals, part of them carrying the barley consumed by the horses and the mules for non-comestible goods.

In the presence of hostile troops it was necessary to have soldiers accompany the supply convoys to the storehouses and the armies. For instance during the siege of Numantia in 141 BC, supplies were brought by a detachment under the command of a military tribune that lost 700 men when the enemy attacked (Appian, Iber. 77). In 45 BC “our [Caesar’s] cavalry went out rather far in the direction of Corduba in pursuit of those who were conveying supplies from the town to Pompeius’s camp” (CAES. Hisp. 11). Alternatively, detachments of troops were located in towns or fortifications along the supply routes. Whenever supply routes were stretched, securing them could take up many troops. The longer the supply lines, the more convoys of wagons and muletrains would be necessary to supply the army and the more troops were required for their protection. Only a numerous and strong army could manage to protect these supply lines.

Hence, organized provisioning called for the frequent supply from the army depots to the army. Occasionally our sources show that in preparation of campaigns commanders told the troops to take with them food for a specified period. For instance, when Scipio during the Second Punic War was confronted with a mutiny in Iberia, he devised a ruse to install a false feeling of confidence among the mutineers. He ordered loyal troops “to furnish themselves with provisions for a considerable time on the pretext that they were marching under Marcus against Andobales” (PLB. 11.26.6). In other words, this

6

Likewise Ward (1957, 84), points out that during the Peninsular War lack of transport limited the range of the magazine system.

141

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

The accomplishment of the Romans in military food supply surpassed that of their opponents, especially in the organization of acquisition and distribution. These two aspects were inextricably connected: there was little point in acquiring food stuffs, if these could not reach the armies. It would require such enormous numbers of packor draught-animals and, therefore, such vast amounts of fodder to move food over large distances over land that it would simply be inefficient to undertake the effort, even if it were possible to find the means that would be required. Hence, considerations of transportation determined that in inland war zones armies had to be sustained from internal resources. No tithes from Sicily, no shipments from Africa sustained the Roman troops moving into the interior of the Iberian Peninsula.

would have been the normal preparation for such a campaign. A similar ruse was employed by Cato, who, while being quartered near Emporion, was approached by Iberians requesting the assistance of at least part his army. Cato did not want to split his forces. In order to give not only the envoys, but also the enemies the false impression that help was indeed under way, he ordered one third of his troops to prepare food and bring it into the ships. We may add that when in 79 BC Metellus expected to be able to conquer the town of Langobritae swiftly, he ordered his troops to take only five days of provisions. However, when the siege proved not to be easy, he had to send out a foraging party (PLU. Sert. 13.6). As we have seen, in 49 BC the Pompeian legionaries carried provisions for 22 days in their train, which may have been an exceptionally large number of days and only made possible by the fact that part of the troops, as we have seen, had nothing (CAES. Civ. 78).

When external resources were not available, armies had to be on the move continuously. The element of duration was related to size – large forces obviously depleting local resources faster than small ones. In winter local resources were rapidly exhausted, which forced armies even sooner than in summer to move on to locations not yet visited by plundering troops. Both numbers and duration determined the military pressure that could be maintained during a particular campaign.

Carrying provisions for a long period would give an army operational independence from the supply base, but would involve the disadvantages of a large train. The train had to be protected in case of attack; for this reason the pack-animals were interspersed between the army contingents, when marching in combat formation. Furthermore, the army train was vulnerable and troublesome on rough terrain; a large number of packanimals consumed the army’s stores of barley and not only required the often arduous gathering of large amounts of fodder, but also bound the army to terrain offering that much fodder. In short, commanders had good reasons for trying to keep the army train down to the absolute essential (FRON. Str. 4.1,1; 4.1,7; PLU. Mar. 13,1).

We have seen that food supply was a vital element in all major wars that Rome waged in the Iberian Peninsula. The campaigns of the Scipio’s (father, uncle and son) and early second-century generals were only made possible by the shipments of overseas grain. Difficulties increased as soon as the wars shifted inland. The limitations of logistics that exposed the Romans to the weaknesses of their food supply and to the superiority of their opponents in guerrilla-tactics caused Roman campaigns in the 150s and 130s BC to fail. In 153 BC the defection of the town in which all his supplies were stored forced Nobilior to abort the siege and go into winter-quarters (APP. Hisp. 47). In 151 BC the same problem emerged, when Lucullus tried to subject the town of Intercatia, located in the valley of the Duero in the territory of the Vaccaei. Food shortages in the army, which were probably due to the exhaustion of the surrounding area, induced Lucullus to conclude a treaty with the people of Intercatia. Lucullus now tried his luck at Pallantia, a town of the Vaccaei in the same region, but here the Pallantian horsemen successfully prevented his foraging parties from gathering supplies, with the result that Lucullus also withdrew into winter-quarters (APP. Hisp. 54f.). Finally, in 136 BC Aemilius Lepidus Porcina besieged the same town of Pallantia. Foraging parties collected supplies in the surrounding countryside, but the siege of Pallantia being protracted the food supply of the Romans failed. According to Appian, the general withdrew after a long time, but not before many of his men had died of starvation (APP. Hisp. 81f.). In all cases the weaknesses of the Roman army’s food supply compelled the Romans to abort their operations. Problems of food supply also plagued the Roman armies during the civil wars of the first century BC, requiring dangerous foraging missions and shifting much of the burden to the local population. The vital importance of the army’s food supply is

FOOD SUPPLY AND STRATEGY The aim of armies was not solely to stay alive by arranging adequate supply of food and drink, but also to win wars. The supply of armies on campaign had to be realized in a way that least hampered military operations, which meant that the supply system had to ensure optimal logistical security while strengthening flexibility in time and space – i.e. providing the opportunity for an army to operate where and when one wanted. The aims of security and flexibility were often contradictory, leaving it to the commander to decide which he regarded as most important in each individual case. ‘Living off the land’, in its strict sense of plundering, has often been regarded as offering the most flexibility, because armies seemed to be able to go wherever they wanted, not being restrained by supply lines and bases. However, as we have seen, this brought with it its own limitations, not to mention the basic insecurity of provisioning. The difficulties inherent in feeding armies compelled governments and commanders to make optimal use all available means, i.e. living off the land when the tactical and environmental circumstances allowed, shifting the burden as much as one could to local allies, and relying on external supplies as far as possible.

142

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

being in want of provisions, and his army much reduced, burnt his camp in the night and returned to Lusitania”. (APP. Hisp. 67-68).

indicated by Caesar’s lament that disrupted supply as a result of an unexpected flooding in the spring of 49 BC caused a complete reversal of his situation: “In a mere handful of days, circumstances had totally changed and there had been a reversal of fortunes: our men were now labouring under shortages of all the necessities of life, while their opponents had plenty of everything and were considered the stronger force”. (CAES. Civ. 1.52).

CONCLUSIONS Considerations of supply had a much more direct impact on campaigns in the Iberian Peninsula than in, for instance, Greece or Asia Minor. This is not to say that food supply was less crucial in the latter war zones. However, stockpiling depots on the coast or navigable rivers full of provisions that were shipped from taxpaying provinces like Sardinia and Sicily or from largescale suppliers such as Carthage or Numidia freed the Roman commanders from many of the restrictions that governed warfare in the Iberian heartlands. While the Romans often employed large armies in Greece or Syria, they rarely concentrated more than a few legions against the Iberians. The circumstances of military food supply therefore determined that in isolated regions the Romans could not employ their military potential as fully as in the coastal regions of the Mediterranean. The long-drawn-out nature of Roman warfare in the Iberian and their troubles in enforcing Roman rule among the inland peoples may have been partly due to the geographical circumstances which made these tribes less susceptible to the might of Rome.

In short, the element of pressure on the enemy, requiring permanent security of supply, was essential for the besieging of cities. From the viewpoint of military provision, sieges were among the most difficult operations to undertake. Local supplies would soon be exhausted by large armies, the more so when the population had gathered in fortified towns and had evacuated their stores behind the walls. Ironically, this could also work in the Romans’ favour, as witnessed by Appian in his account of Viriathus’s operations in 142 BC: “Viriathus continued to make frequent incursions by night or in the heat of the day, appearing at every unexpected time with his light-armed troops and his swift horses to annoy the enemy, until he forced Servilianus back to Itucca. Then at length Viriathus,

143

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

3.3 GARRISONS, MILITARY LOGISTICS AND CIVIL POPULATION IN THE LATE REPUBLIC: AFRICA AND HISPANIA1 T. Ñaco already broken out (Morstein-Marx, 2001; Parker, 2004). Almost immediately, a quaestor of L. Calpurnius Bestia (cos.111) left for the city of Vaga, which at that time was an independent commercial enclave under the Numidian sphere, to ensure payment of the negotiated amounts (SAL. Jug.29.5). In 109, the consul Q. Caecilius Metellus Numidicus decided to establish a military garrison in that same city with the intention of supervising the provision of supplies and war equipment, realising that the presence of numerous merchants of Italic origin could benefit Rome strategically by guaranteeing supplies for the emplacement situated at his own rearguard (Sal. Iug. 47.1-2) (Berthier, 1981).

T. TURPILIUS SILANUS, PRAEFECTUS OPPIDI IN THE CITY OF VAGA (AFRICA), 109/108 BC1 In the West, in the midst of the war against Jugurtha, T. Turpilius Silanus, praefectus fabrum of the consul Q. Caecilius Metellus (APP.Num.2.3; PLU.Marc.8.2), was made commander in chief of the African city of Vaga between 109-108 BC. Despite the discussion concerning the status of this individual, which has gone on for more than a century,2 this paper’s interest in the events occurring at Vaga is restricted to the consequences derived from the presence of a Roman garrison in the foreign city, and to the attributes of Turpilius as praefectus oppidi, an expression coined by Sallust, who describes him as a genuine military governor.3 Years later, shortly before the Italic slaughter at Cirta, which led to the conflict between Rome and Jugurtha, the latter tried to gain time in event of a Roman reprisal by bribing influential members of the Senate and negotiating a truce when hostilities had

During the autumn and winter of that same year, Metellus ordered his troops to withdraw towards the rearguard areas where he could demobilise his soldiers without great danger. They wintered in the urban centres that previously had become part of the Roman faction, which was damaging to the cities, particularly to the civilians, who were forced to billet idle legionaries in their own houses (SAL. Iug. 61.1-2). The hospitium militare was a common practice, not only in the Roman world; many armies throughout history have roomed their soldiers, mobilised or not, with civilians. The Roman military authorities demanded all types of services of foreign cities, specially in the provinces, but occasionally also in Italy (maintenance, housing and even the soldiers’ pay), getting to the point where, in certain cases, these charges were not able to be undertaken either by public funds or private citizens. All this undermined the autonomy of communities that did not enjoy privileged legal statutes that could protect them from these measures (Ñaco, 2001; Roth, 1999). Actually in Vaga there took place a revolt against the Roman garrison that kept the military emplacement, instigated by Jugurtha and aided actively by members of the city’s élite.4 A local festivity was

1 This project was financed in 2007 by a research grant from the Harry Frank Guggenheim Foundation (New York, USA). The research was done with the support of the Programa “Ramón y Cajal” (MEC-UAB), Project Hum-2004-04213/Hist- of the Ministry of Education and Science and the Grupo de Investigación Consolidado AREA (SGR200500991) of the Catalonian Government, and with the concession by the Catalonian Government of the scholarship Beca per a estades per a la recerca fora de Catalunya (2006 BE-2 00112) at the University of Oxford (Classics Faculty). I would also like to thank César Carreras Monfort for his invitation to take part in this book, and his patience as well, and to Fernando López Sánchez and Joaquín Muñiz Coello for their help. Any errors in this paper, of course, remain the responsibility of the author. 2 On the status of T. Turpilius Silanus (RE, VIIA, 1948, 1430-1431) and his position as praefectus fabrum in the Late Republic, see the bibliography that accompanies the ample scientific discussion concerning this individual in two relatively recent articles (Welch, 1995, 131-145; Badian, 1997). At a more general level there are the works by (Verzar-Bass, 2000; Ñaco, 2006). 3 SAL. Jug. 66.3; 69.4; PLN. Nat. 5.49.2, and in similar, though not identical terms: CAT. Orat. 173.9; FRO. Aur. 1.2.11; GELL. 15.1.6; LIV. 23.15.2; 28.15.15, in this last case narrating the defection to the Romans in 206 BC of the Carthaginian commanders of the Carthaginian garrisons previously located in two cities: inde duo munita oppida cum praesidiis tradita a praefectis Romano.

4

In his eagerness for victory, Jugurtha attempted to bribe with his gold not only the slaves of the Romans, but also even the soldiers that were part of Roman garrisons in Numidian cities: Sal. Jug. 66.1: servitia

145

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

chosen for the day of the revolt, inviting to the celebrations a majority of the soldiers, centurions, military tribunes, and even the commander in chief, Titus Turpilius Silanus, praefectus oppidi, a term certainly descriptive of the responsibilities attached by the consul to an official of his army, quite surely a member of his general staff (Sal. Jug. 66.2-3).5

validity of which could be revised. An example of this could be the spread of a complex bureaucratic Roman administration and a permanent fiscal policy (the extent of which depends on the historiographical interpretation). The binomial “military conquest = tribute charges” is not, at least for Republican Hispania, as easy to apply as might seem at first sight; some of these traditional approaches have generated less unanimity and more critical positions in the last years.8 In fact, in 1977, R. Knapp suggested the study of the establishment of Roman garrisons in provincial cities, both historically and archaeologically, as an essential part of a specific policy of implementation of Roman rule in the Iberian Peninsula, at least during the first century of military intervention, and even later. The deployment of enclaves, formed by legionaries or auxiliary troops of different origin, and situated near or inside the urban centre of foreign cities, became a mechanism of military and political domination, not only over the dediticii, but also over those populations directly or indirectly menaced by the inexorable advance of the Republican army in Hispania. From the perspective of the North-American scholar, the political and strategic control of the territory constituted two coinciding objectives, or even two faces of the same coin (Knapp, 1977, 15-35).

When narrating that Turpilius was the only one amongst the Romans and Italics in the dinner that survived the killing, Sallust refers to him simply as praefectus, without epithet, which in turn can be interpreted as an unmistakable reference to his military affiliation, not specifying that during the previous months his foremost responsibility had been the governing of the city (Sal. Jug. 67.3).6 Once the rebellion was snuffed out and the members of the local senate executed, numerous suspicions of treason fell on Turpilius, who is once again called praefectus oppidi by the Latin historian, and was later judged and condemned to be flogged and decapitated. This punishment would have been inflicted nam is civis ex Latio erat (Sal. Jug. 69.4), a controversial phrase that to some take to mean that Turpilius was a civis Latinus, while for others it is the product of an adulterated text.7 Beyond the exact title and origo of Turpilius, his mission as commander-in-chief of the Roman garrison in Vaga must be considered an integral part of his functions as an officer of the army, thus his power as the supreme Roman authority in the Numidian city. Likewise, the abuses committed on the civilians and civic authorities must be interpreted in a strictly military context, comparable to the actions of the famous praefectus at Gades (206 BC) and other governatori romani in città provinciali, according to the terms established in his day by Tibiletti (1953).

On the other hand, the interest of authors such as Erdkamp (1998), Roth (1999), Carreras (1998), and others in delving deeper into the study of logistics in Roman armies during the last centuries of the Republic, the transition towards the Empire, and even further, by analysing the relationship between military strategy and the need for supplies both during the months of military campaign and winter, has been crucial in bringing to our attention a modern reconsideration of the phenomenon of military garrisons, and in particular, their intrinsic relation to the indigenous populations with which they coexisted, or better, said, of which they served themselves (Groeman-Van Waateringe, 1989; Goldsworthy, 1996; Erdkamp, 1998; 2007; Roth, 1999; Carreras, 2004; 2006). In the case of the Iberian Peninsula, the recent studies of Cadiou (2003) have allowed us to contrast the already classic position of Knapp in respect to the role of garrisons in the history of Roman intervention in Hispania, highlighting the need for a renovated approach of this issue in general. More specifically, Cadiou (2003) points out that, according to the available evidence, the garrisons that were always part of the treaties of deditio formalised with the Hispanics, in many cases do not seem to respond to a simple strategic control of the defeated, but rather to the deployment of logistical support points, which would guarantee the continuous supply of the

GARRISONS, CASTELLA, ROMAN ARMIES, AND THE HISPANI (133-72 BC) The material consequences of a war of conquest lasting two centuries have conditioned historical interpretations towards prioritising a cause-effect relationship between military activity and other phenomena, such as assimilation, acculturation and romanisation, although it is also true that all this has sometimes been established without questioning several fixed and out-dated topoi, the Romanorum allicere et eos ipsos qui in praesidiis erant, pecunia temptare. 5 et ipsum praefectum oppidi T. Turpilium Silanum. 6 Turpilius praefectus unus ex omnibus Italicis intactus profugit. (Bullo, 2002). 7 quem praefectum oppidi unum. Appian (App. Num. 2.3) seems to confirm the title of phrourarchos, therefore equivalent to the general meaning of “commander-in-chief” contained in praefectus oppidi. On the other hand, Plutarch (Plu. Marc. 8.2) following Greek nomenclature, prefers an identification with praefectus fabrum, which has generated a lengthy discussion in respect to whether Latin citizens could have occupied this or other positions. In the case of accepting the literal interpretation, appearing in a majority of moderns editions (civis ex Latio), see Fontanella (2002). Nevertheless, Badian (1997) is sceptical towards this literal interpretation, preferring civis ex Collatia (in reference to a dark Latin city), making Turpilius a Roman citizen, perhaps a member of the élite in Collatia, but a Roman after all.

8 The influence of the so-called culture militaire in the administration of Hispanic affairs is noteworthy according to Le Roux (1995). Critiques of the traditional model may be seen in Ñaco (2003; 2007); France (2005; 2006); Chaves (2005). The possible use by the Romans of the socalled “Iberian coinage” (with different motivations depending on the historiographic traditions) has also caused discussion amongst specialists Ripollés (2005) and López Sánchez (2007) that relates punctual issues from some of the mints in Celtiberia, between the late second and the beginning of the first centuries BC, with a possible use as payment to auxiliary troops under the service of Rome, during the “post-Numidian” and “pre-Sertorian” time of complex historical interpretation.

146

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

Republican army, both during advance and retreat, on the front lines and in the rearguard areas (Cadiou, 2003; Ñaco, 2006).

beginning of the Empire, due, amongst other reasons, to the gradual incorporation of auxiliary troops (Whightman, 1977a; 1977b; Roymans, 1996; 2004).

It is archaeology, nevertheless, that returns substantial data on the use of Roman garrisons with logistical motivations in Republican Hispania. In the last decades, due to the amazing advance of the so-called “military archaeology” in the Iberian Peninsula (Morillo and García-Marcos, 2002; Garcia-Bellido, 2006), a whole series of enclaves has been identified of notably smaller dimensions than the regular camps or castra, and in the case of those that will be mentioned here, their beginnings are dated to the last third of the second century and the first decades of the first century BC, or even further into this last one. According to the specialists, they would correspond to what the literary sources call castella (Jimenez, 1995; Aounallah, 1996). They are enclosures, normally fortified, constructed ex novo, and independent from non-Roman settlements, eventually situating themselves on adjacent territory, and housing a relatively reduced number of troops. The truth is that their functions could be related just as much to the control of a certain territory in a particular moment of a determined military conflict as to the logistics of the advance of the Republican army in Hispania.9

The main tasks entrusted to certain strategic Roman enclaves included the construction of infrastructures for communication, necessary for the movement of troops (roads, bridges, aqueducts, etc.), as well as their maintenance and security after the war. These enclaves were not always fortified like castella, but were logically also in charge of the general security of the surrounding territory, while exploiting the material resources coming from the neighbouring populations (Bishop, 1999; Ñaco, 2003). In Italy, from very early periods in Roman history, the construction of roads was entrusted explicitly to the legions, due to the priority given to their military use, allowing for a more efficient movement of armies, and facilitating the gradual extension of the conquest throughout the peninsula. The treaties of castrametatio explicitly mention the presence of castrorum metatores and of castrorum mensores, which have been identified as military engineers in charge of making certain operations of land measurement at a large scale, with the object of finally emplacing the castra and the rest of the legion’s quarters, later built by the own soldiers. The more than likely participation of the legions in these kinds of activities probably coincided with the winter months, when the soldiers were quartered for long periods of time (Conticello, 1991; Witcher, 1997; Uggeri, 2004; Prag, 2007). The discovery of some of the most ancient Republican milestones in Hispania, dated to the last quarter of the second century BC, has been linked to the hypothetical Roman presence in El Camp de les Lloses (Tona, Osona), a site only occupied during the last decades of the same century and the early ones of the first century BC. Although for the moment, no relevant defensive structures have been documented, it has been thought that at least some of the occupants were part of a minor unit of the Roman army, quartered very near to one of the oldest sections of Roman road in the Iberian Peninsula, which they may have helped construct or defend and maintain (Mayer and Roda, 1986; Molas, 1993; Beltrán and Pina, 1994; Molas et al., 2000; Hermon, 2000; Salomon, 1996). Although there is no definite evidence to prove it, Roman presence in this and other similar enclaves may have also been used to centralise the recruitment of auxiliary troops coming from the immediate surroundings, due principally to the characteristics of this region – populated mainly by the Ausetani – which remained aside from the important revolts until the Sertorian period (Ñaco, 2006; Pina, 2003).

As has been shown for the Roman conquest of Britain, the Roman army could take advantage of pre-existing urban structures, in some cases previously abandoned, with the intention of temporarily quartering smaller legionary units in enemy territory, mostly for strategic reasons and because of circumstances in the development of the military conflict in the region (Frere, 1986; Breeze, 1986-7; James, 2001; Manning, 2006). A structure like the castellum could keep a modest garrison, formed by legionaries and auxiliary troops, and remain, if necessary, in an occasionally hostile territory, surviving on its own for some time. This high capacity of adaptability, especially in a militarily insecure land, was quite useful when the instability provoked by the constant changes in war also affected previously pacified regions, such as during invasions and internal revolts.10 Incidentally, during Ceasar’s conquest of Gaul, and specially for the post-Caesar control of Gallia Belgica, archaeology has demonstrated that in these types of contexts, in which the Roman legionaries were temporarily quartered in isolated garrisons, not only were indigenous metalworking techniques easily incorporated, but the daily use of essentially alien material culture is also perceived, like in the case of fibulae. All this could have influenced decisively in a certain “militarisation” of the romanisation process amongst the populations that inhabited those regions, between the end of the Republic and the

During the Sertorian War, the relative ease for the Sertorian faction to supply itself, strike alliances or get the necessary auxiliary troops from the Hispanics in each moment, as well as being able to organise the production of weapons at a large scale (Quesada, 2006), has, as a whole, been related to the known strategic ability of Sertorius throughout the whole campaign. A good example of this would be the taking and establishment of military garrisons in towns, such as Contrebia Leucade,

9 Amongst one of the most recent examples (Palmada, 2003; Mataloto, 2004; Diosono, 2005; Mercado et al., 2006; Brotons and Murcia, 2006; Olesti, 2006). 10 In the province of Corsica-Sardinia, theoretically under Roman control since 237 BC, there were incessant, although intermittent revolts until well into in the first century AD. Led by nomadic communities of the mountainous interior of both islands, the attacks fell on the agricultural cities of the coast (Zuoca, 2001; Rowland, 2001; Ñaco, 2003; Mastino, 2005).

147

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

where the Republican faction had previously accumulated important provisions, and where Sertorius left a legate, L. Insteius, as commander-in-chief of the garrison (LIV.91, frg. 22W) (Broughton, 1952; Balil, 1963; Garcia Mora, 1995; Pina and Pérez Casas, 1998). At the same time, the biography that Plutarch dedicated to the insurgent leader, insists that on his arrival in Hispania, he renounced any kind of exaction, as had been common practice till then, freeing all Hispanics from their duties of housing active Roman garrisons and demobilized soldiers in the winter (hospitium militare), while at the same time denying to his own troops what was surely a very deeply rooted custom in the Roman army (PLU. Sert. 6.4) (Muñiz, 1982; Ñaco and Prieto, 1999).11

to pay the legions, auxiliary troops, weapons, and supplies) contrasted with the precariousness that the treasury underwent at the time, due to simultaneous military expenditures on different fronts and a certain decrease in state revenues, from which may be deduced a growing pressure on Hispania to ease at least a part of those expenses (SAL. Hist. 2.47.6-7).14 In fact, amongst the urgent measures undertaken by the respective commanders, we know that Pompey received supplies from his colleague in the Narbonense, M. Fonteius (CIC. Font. 13) in 76 BC, while also leading various incursions in Gaul (LIV. Per. 93) (Beltrán and Pina, 1994; Garcia Mora, 1991; Payen, 2002; Scardigli, 2002). It would seem logical to a certain point to establish a parallel between Roman actions in the Narbonense and in the Citerior during the whole period, taking into account not only Rome’s decision to act on both territories almost at the same time, but also the tight link existing between the two provinciae in terms of logistics, and according to some researchers, even politically.15 Stationed near the Vascones for the winter, Pompey embarked on the foundation of Pompaelo or Pompelo (STR. 3.4.10), probably because that was the principal route through which supplies arrived from the Narbonense, although other motivations should not be discarded, either ideological, like the imitatio Alexandri pointed out by L. Amela, or the existence of a firm policy of veteran settlement in the Iberian Peninsula, similar to what had happened in Gaul (CIC. Font. 6.13-14; CAES. Civ. 1.35.4) (Salinas, 2006). Incidentally, a passage in Plutarch, referring to the activities of Lucullus against Mithridates, could aid in locating the destination of some of Pompey’s veterans, when mentioning the soldiers of Pompey, citizens now, were snugly ensconced with wives and children in the possession of fertile lands and prosperous cities (PLU. Luc. 34.4), together with the territorial implications of this type of settlement, or the extension of the Roman dynast’s cliental sphere of influence in Hispania (Olesti, 1995; Amela, 2000).16

In a letter to the Senate preserved in a fragmentary text of Sallust and dated to the winter of 75/74 BC, Pompey exposes before the patres the precariousness with which he had to face his campaigns against Sertorius, indicating, perhaps with an excess of zeal, that Sertorius had spent winters camped amongst the enemies, instead of having withdrawn, literally “into the cities”, as was usual (SAL. Hist. 2.98.5) (Roth, 1999, 143).12 For this period, Cáceres el Viejo is one of the few military camps well documented archaeologically in Hispania, located in a region marked by the traditional resistance of the Lusitani and other groups contrary to Roman domination. Iberian brooches have been documented inside the camp, indicating perhaps cohabitation between legionaries and auxiliaries, or at least the use by the Romans of indigenous material culture (Pamment, 1995; 1996). As of that moment, nevertheless, a totally different model of military occupation seems to have been consolidated. There was a deduction in colonies, and above all, a certain proliferation of castella or fortifications aimed at the defence of chosen strategic points, and especially amongst them, communication routes (Knapp, 1977; Alonso and Fernández, 2000). The future leader of the optimates described in his letter the difficulties in getting supplies for his army in Hispania Citerior; he could not compromise his personal wealth, nor obtain supplies from Gaul like the year before, so he solicited the dispatch of urgent help directly from Italy (SAL. Hist. 2.98.9).13 A few years before, in Gaius Cotta’s speech to the plebs, also narrated by the same Roman historian, the consul warned that the demands of the commanders serving in Hispania (money

PREFECTS AND CITIES ACCORDING TO DE BELLO AFRICO (46 BC) A large number of garrison commanders, regardless of their title, who were stationed in foreign cities and identified by Tibiletti (1953), correspond to the central period of the civil wars (53-42 BC), although they do not always originate from the Eastern Mediterranean, the principal focus of interest in his famous article of 1953. Incidentally, in a study on procuratores and praefecti

11 Sertorius was able to take advantage of his own experience as a military tribune at the service of the proconsul T. Didius (98 BC) when the legionaries that composed the Roman garrison in the city of Castulo, received hospitality inside the urban centre and were killed in a bloody revolt, later suppressed by Sertorius (PLU. Sert. 3.6-10) (Konrad, 1994; García Mora, 1998; Ñaco, 2001). 12 The case of Cyprian cities is very well known. According to Cicero in a letter to Atticus of 50 BC, they always preferred to bribe their predecessors in Cilicia and avoid having the legions camp in their cities’ territories: CIC.Att.5.21.6-7 (Muñiz, 2004). 13 Hispaniam citeriorem, quae non ab hostibus tenetur, nos aut Sertorius ad internecionem vastavimus praeter maritumas civitates, ultro nobis sumptui onerique; Gallia superiore anno Metelli exercitu stipendio frumentoque alui et nunc malis fructibus ipsa vix agitat; ego non rem familiarem modo, verum etiam fidem consumpsi (Amela, 2003, 83).

14 Namque imperatores Hispaniae stipendium, milites, arma, frumentum poscunt. 15 According to CIC. Font. 6.13, Fonteius dedicated a part of the wealth obtained from Pompey’s confiscations of the lands of several Gallic tribes that rebelled in 77 BC to buy cereals for Pompey’s armies fighting Sertorius in Hispania (Ebel, 1976 74-102; Hermon, 1993, 145, 262-273; Amela, 2003, 82-83). 16 In the later period of the Civil Wars, the participation of Hispanics in conflicts “between Romans”, siding with one faction or the other, is a well known phenomenon. A good example of this would be the military help received by Sextus Pompeius from some of the cities of Ulterior (Love, 2002).

148

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

during the Republic, Nicolet (1966) inferred a few years later the lack of evidence on the legal status of the men who occupied the position of prefect, which in his opinion “ne sont pas à proprement parler militaires, qui ne s’intègrent pas dans les grands commandement légionnaires” (Harmand, 1967). Nicolet (1966) centred his attention especially in the last century of the Republic, when some of the praefecti named each year by the provincial governors did not have a specifically designated function, or in any case, took care of responsibilities that were different from those theoretically imposed by their title. In contrast to the situation in previous centuries, during the last decades of the Republic, a majority of officers of the Roman army, amongst whom the magistrates could choose their leading deputies, were knights and noble Italics, all of them loyal to leaders who had promised them support in their business and private interests, and not to the State (De Blois, 2000; 2007). In general, these officers were part of the cohors amicorum of the magistrate in his provincia, aiding him in administrative, legal or military tasks, all in return for certain corollary benefits, related mainly with the continuity of their own private businesses in the region (FRONT. Str. 2.7.3; LIV. Per.81; CIC. Att. 5.21.10.1).17

display of his tactical abilities, together with another Roman knight, the latter originally from Utica (CAES. Afric. 68.2-4) (Broughton, 1952, 303). Publius Cornelius, who was probably a prefect, should also be mentioned, although his rank and the position he held are unknown, because the anonymous author of this work only adds that he was a reservist (evocatus) of Metellus Scipio, in charge of the defence of the Pompeian garrison emplaced in the city of Sarsura, finally paying for it with his life when Caesar conquered the stronghold (CAES. Afr. 76.1) (Broughton, 1952, 302). Furthermore, a delegation composed by legates of the city of Vaga presented itself before the future dictator to plead for the urgent dispatch of a garrison in order to defend the Numidian city from the Pompeian enemy, naturally in exchange for a whole series of compensations, mainly dealing with logistical support. Nevertheless, the petition arrived late, because precisely at that moment, Caesar’s camp received news of the city’s destruction by king Juba, so that the possible treaty became void of any interest (CAES. Afr. 74) (Bullo, 2002, 105).19 Likewise, after conquering the city of Sarsura, the same Caesar distributed amongst his troops the cereal that the previous prefect of Metellus Scipio, P. Cornelius, had guarded till that very moment, later advancing towards the next city to besiege, Thysdra, also under Pompeian defence (B. Afr. 76.2). This last episode, together with other events that may only be found in this anonymous work, enable us to confirm the establishment of warehouses for weapons and supplies inside the urban precinct of foreign cities under Roman military control, logically provided with garrisons that would keep them, constituting an essential element of Republican military praxis. In fact, the almost exponential increase of notices concerning this phenomenon during the Civil Wars is a clear indicator of the direct connection between this practice and the use of a series of Roman officers of ambiguous denomination as responsible for the defence of supplies vital to carry on with war.20 Precisely in relation with Roman military administration in some of the African communities is an inscription from Curubis (Korba), referring to the fortification works of the city, which was partial to the Pompeian faction during the central period of the Civil War between 48 and 46 BC. According to the inscription, the responsibility of this construction work befell two legati and one praefectus, T. Tettius, who would probably correspond with the “military governor” of this Pompeian stronghold in African territory.21

In principal, prefects, legates, military tribunes, exmagistrates, and even officers of lesser rank could be in charge of any mission under the orders of a magistrate. Despite this, an ambiguousness persists in the terminology used in the sources of the first century BC, to the point that praefectus – with no epithet – ends in denoting a military context without describing any specific function. This practice should be related both to the conquest and occupation of cities by the Republican dynasts, with the ensuing expansion of the imperium romanum and the legion of dependent cities and communities, and to the management of their own armies, using for this the resources available from allied and, when necessary, of conquered cities.18 Some of the best examples come from the African campaigns of the Civil War (46 BC), preserved in this case in the De Bello Africo. Such is the case of Gaius Minucius Reginus, commander – and possibly prefect – of Pompey’s garrison in Zeta, member of the equestrian order and part of the close circle of friends of Metellus Scipio, supporter of Pompey; he was eventually captured by Caesar in a 17 As highlighted by Muñiz (1998, 135-151), M. Scaptius, a member of the equestrian order with important connections in the Senate, enjoyed the position of praefectus with Ap. Claudius, proconsul in Cilicia in 5351 BC, although he actually concerned himself with the promotion of his own business affairs in the region, under the protection of his powerful “friends” in Rome. In fact, Cicero’s letters indicate that he accepted Scaptius’ “application” to renew his position unwillingly, a responsibility which fell entirely in the sphere of competence of Ap. Claudius’ successor in the provincial government of Cilicia, i.e., the same Cicero (CIC. Att. 5.21.10). 18 Sometimes the numismatic evidence, despite certain controversial interpretations, seems to indicate the presence of individuals with the title of PRAEF(ectus) who, almost at the end of the Civil Wars, or already in the Augustan period, could have been in charge of city garrisons and even of the government of wider territories, such as the prefect AMBATUS, from Campestris Babba (Mauritania Tingitana): (AMBATUS PRAE F. ITER IVLIA CAM P.) (Guadan, 1969; Amandry, 1984; Alexandropoulos, 2000).

19 Likewise, SAL. Iug. 77 mentions the temporary presence of a Roman garrison in Lepcis Magna (Bullo, 2002, 166). 20 Of great relevance is the information concerning the logistical determinants in Caesar’s African campaigns that appear in this work (although certainly under the bias of the future dictator), relating garrison policy with the supply needs of the Roman armies, both for Caesar and Pompey, and specially with the benefits resulting from the taking of cities that held warehouses full of supplies and weapons, despite the resistance put up by the enemy garrisons: CAES. Afr.3.1; 9.1-2; 20.4; 21.1; 24.1-3; 33.1-5; 36.2; 37.1-2, documents parties of soldiers on missions with the object of finding supplies in the surroundings, with no restriction on the methods used (commeatus); 67.1-3 (Roth, 1999, 130); 74.1; 76.1; 79.1; 85.1 21 CIL 24099 = ILS 5319 (Breal, 1895; Degrassi, 1965, no.394; Broughton, 1952, 291; Zucca, 1996; Bullo, 2002, 103).

149

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

3.4 OCTAVIAN’S REFORMS C. Carreras under them who had had some difficulty with the leaders, or with their fellow-officers”.

The army in the Late Republic was based in the personal power of commanders who promised rewards and lands to the discharged soldiers. Such personal loyalties raised numerous internal problems in the armies which brought about continuous political crisis as civil wars (D.C. 52.27). The civil conflicts that involved the key political figures of Late Republic also affected systems of army supply since they no longer trusted in suppliers and their mechanisms.

In the year 40 BC, Sextus Pompeius’s fleet blockaded Italy creating great problems of supply for the city of Rome; L. Cornelius Balbus was also consul in this year. The old friend of J. Caesar became advisor of Octavius, who took advantage of all the social networks and influence of the Gaditan and the other Caesar’s collaborators. For instance, the new tax on bequests (i.e. vicesima hereditatium) was issued during this year in order to obtain resources for feeding the plebs of Rome and against the blockade.

The last years of the Republic were difficult ones for publicani, since they were subject to proscriptions from both sides. Appian says that 2000 knights were killed as a result of such proscriptions (APP. BC 4.5):

It is difficult to know whether the advice of L. Cornelius Balbus may have affected some of the reforms undertaken by Octavian and, also Agripa. However, some changes seem to have attempted to solve long lasting hardships such as supplies for the plebs and military.

“As soon as the triumvirs were by themselves they joined in making a list of those who were to be put to death. They put on the list those whom they suspected because of their power, and also their personal enemies, and they exchanged their own relatives and friends with each other for death, both then and later. For they made additions to the catalogue from time to time, in some cases on the ground of enmity, in others for a grudge merely, or because their victims were friends of their enemies or enemies of their friends, or on account of their wealth, for the triumvirs needed a great deal of money to carry on the war, since the revenue from Asia had been paid to Brutus and Cassius, who were still collecting it, and the kings and satraps were contributing. So the triumvirs were short of money because Europe, and especially Italy, was exhausted by wars and exactions; for which reason they levied very heavy contributions from the plebeians and finally even from women, and contemplated taxes on sales and rents. By now, too, some were proscribed because they had handsome villas or city residences. The number of senators who were sentenced to death and confiscation was about 300, and of the knights about 2000. There were brothers and uncles of the triumvirs in the list of the proscribed, and also some of the officers serving

After the battle of Actium (31 BC), Octavian dispersed an army of 60 legions, around 300,000 armed men, giving lands and cash rewards (Res Gestae, 3). Therefore, he retained only 28 legions, the number that remained to protectthe Principate. Discharging half of the army meant an extraordinary expenditure in purchasing land for new veteran colonies, Augustus even paying from his own purse – 860,000,000 sesterces according to the Res Gestae (28) from 30 to 14 BC. Cash rewards on discharge from 7 to 2 BC reached an amount of 400,000,000 sesterces (Res Gestae, 16). Additionally, Augustus established a minimum length of service (25 years) and regular payment of 225 denarii1 and fixed discharge cash reward. Such regulations of soldier service attempted to break any past loyalty between legionaries and commanders. From now on, soldiers’ 1 In fact this measure was taken by Caesar (SUET. Jul. 26), but Octavian confirmed it later.

151

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 3.4.1 Map of republican military camps (Morillo) from the camp of Segisama. Both fronts revealed some degree of continuity with regards to military fronts in the Republican period following the Ebro valley and the Lusitanian coast. In fact the tradition of two Spanish provinces, Ulterior and Citerior, as well as their respective military commanders seemed still in place in the early stages of the Astures and Cantabrii wars.

loyalties were with the State and the Princeps (public oath). In order to raise funds for the discharge cash reward, he created in AD 6 the aerarium militare with his initial donation of 170,000,000 sesterces (Res Gestae, 17). The income of this new treasury came from two new taxes: 1% of sales at auction and 5% of heritance tax. The next Octavian responsibility was to command all the non-peaceful provinces, in other words, the ones that still had legionary or auxiliary troops in its territory, such as Hispania, Gaul, Syria, Cilicia, Cyprus or Egypt. Those were the imperial provinces whose taxes were gathered by Octavian, and become part of the province’s own treasury called “fiscus” (Eck, 2003).

With regards to supply, this initial division on two fronts suggested also two possible sources of supply according to different products and their accessibility to military camps. Following this logic, it does not appear strange that the initial division of Roman Spain in three provinces (Tarraconensis, Baetica and Lusitania) in 27 BC included conquered territories of Gallaecia and Asturia as the new created province of Lusitania. Lusitania covered the whole region from the river Anas (Guadiana) to the Cantabric Sea, and had Emerita Augusta as capital. This was a colonial foundation of Augustus in 25 BC.

Since 27 BC there were two types of provinces, senatorial and imperial, each of them collecting taxes for an independent treasury: fiscus for imperial provinces and aerarium Saturnii for senatorial ones. Despite this division of treasuries, Octavian contributed at least 4 times to the aerarium Saturnii with his own money a total amount of 150,000,000 sesterces (Res Gestae, 17).

The provinces of Tarraconensis and Lusitania became imperial provinces while Baetica became the only senatorial province in Hispania. Therefore, taxes gathered in Tarraconensis and Lusitania were employed to pay troops and were part of the imperial fiscus. Whereas, the Baetican taxes became part of the senatorial aerarium

Meanwhile the war against the Astures and Cantabrii began in Hispania in 29 BC on two initial fronts: one facing the Astures from the military camps of Lucus, Asturica and Legio, and the other facing the Cantabrii 152

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

Figure 3.4.2 Map of the early division of Spanish provinces (27 BC)

eos agros sine controvertia possi dere iubeo …/… Actum Narbone Mario XVI et XV K. martias M.Druso Li bone Lucio Calpurnio Pisone cos (Grau and Hoyas, 2001; Sánchez Palencia et al., 2007)

saturnii. Despite these two different treasuries, army supplies could be bought from any province in the Roman Empire. In the case of the Spanish provincial division, one interesting fact is that Augustus modified the provincial limits only a few years later (around 16-13 BC).2 The second division included the territory of Gallaecia and Asturia as part of the province of Tarraconensis, defining the northern border of Lusitania up to the river Duero. Many scholars have discussed the reasons behind such change in the borders of these new provinces, employing different arguments such as strategic – having a sole commander or administration in relation to the exploitation of the gold mines.

Perhaps, one of the “brains” behind such a division was M. Vipasius Agripa, who was in the NW Peninsula in 1918 BC crushing some indigenous rebellions. Despite his short time in the Iberian Peninsula, he seems to have been responsible for organizing the road network, mines exploitation (ingots with Agrippa inscription) (Rodà, 2007) and being patron of different cities such as Emporiae, Ulia and Gades. His face appears represented on Gaditan coins with the legend municipi parens and municipi patronus parens. Again, the link between Aggripa and Gades becomes quite suggestive in relation to the Atlantic supply route to the NW.

The Bierzo edict, dated to 15 BC (Alfoldy, 2000), which refers to the province of Transduriana, appears to confirm the same dating for the provincial division: Imp. Caesar divi fil. Aug. trib. pot. VIII{I} et procos. Dicit Castellanos Paemeiobrigenses ex gente Susarrorum desciscentibus ceteris permansisse in officio cog novi ex omnibus legatis meis qui Transdurianae provincia prae ferunt itaque eos universos im munitate perpetua dono quosq agros et quibus finibus possede runt Lucio Sestio Quirinale leg meo eam provinciam optinente{m}

From a supply viewpoint, border changes did not affect tax gathering since both provinces were under Augustus’ control. Taking into account that the second division took place after a period of peace (the war finished in 19 BC), a single command of the whole army in the NW was more sensible that having two different fronts and commands. On the other hand, the province Citerior Tarraconensis was the territory with more civitates stipendiorum in Hispania, in other words cities that paid taxes (PLIN. Nat. 4.117; 3.7; 3.18). Therefore, the provincial fiscus had more resources to pay troops deployed in the territory.

2 Santos Yaguas (1991, 124-133) summarizes different views of scholars about the date and reasons behind this second division. Nowadays, most people agree with the proposal by Lomas (1975, 142) quoting Dio Casio (D.C. 54. 25.1) of a second visit of Augustus to Hispania in 16 BC or 13 BC.

Pliny the Elder includes a complete account of the cities in Hispania using information collected by M. Vipsanius Agripa before his death in 12 BC.

153

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 3.4.3 Map of the late division of Spanish provinces (16-13 BC)

Table 3.4.1 Categories of the cities in Roman Spain from Pliny the Elder Lusitania

Baetica

Citerior-Tarraconensis

Total

Roman colonies

5

9

12

26

Roman municipia

1

10

13

24

Latin municipia

3

27

18

48

Civitas foederata

0

3

1

4

Friend Civitas

0

6

0

6

Civitas stipendiaria

36

120

135

291

Total

45

175

179

399

offices. The next Julio-Claudian emperors also reduced the number of public leases, so by the end of the Ist century AD, large societas publicanii disappear whereas some taxes were collected by minor contractors (conductores).

At that time, Lusitania had probably a low density of population and a low number of communities, including only 36 civitas stipendiaria, from which taxes could be farmed. On the contrary, Baetica and Tarraconensis registered a large number of civitas stipendiaria (120 and 135 respectively) from which the State could obtain good revenues. Thereby, the province of Tarraconensis was in a better situation to put up with all the costs of paying the Roman army in the NW. Despite the contribution of each province, army supplies were always obtained from the production areas by purchase through normal commercial networks or by public networks.

The procuratores Augustii collected taxes in the imperial provinces, some of which were allocated to the army acting in their respective territories. In this case, an individual army could transport goods to their camps in the province. However, problems arose when supplies came from other provinces. Remesal (1986) suggested that the figure of praefectus annonae was responsible for keeping a balance between treasuries, aerarium Saturnii and fiscus, as well as contributions from different provinces. His duties were basically to supply grain to the plebs of Rome, but perhaps he may have had an active role in army provision. This last question is still open to debate.

On the other hand, legal reforms limited the involvement of publicanii in gathering taxes and dues (Cimma, 1981, 99; Malmendier, 2002), while imperial agents took those responsibilities. In 27 BC, Octavian distinguished between “imperial” and “senatorial” provinces and appointed procuratores Augustii to lease out tax collections at a provincial level. Therefore, the way of obtaining resources for the plebs of Rome and the army became more centralized through these new equestrian

Anyway, there should have been a way that legati augusti may have asked for resources to the Roman State through 154

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

from 18 BC onwards (Gacía Bellido, 2007). With the foundation of Caesaraugusta in 15 BC, a new military mint provided currency to the military, but in this case it had a special relation only with Herrera de Pisuerga.

the provincial procuratores, who demanded from a central authority some resources in money to buy supply locally, or in kind, from other provinces (i.e. corn, wine, olive-oil). Products in kind were shipped from the provinces by public contracts between traders (i.e. mercatores, navicularii), who were paid an amount for this service (vecturae). Those traders enjoyed exemptions during the Principate and may have carried their products together with the public cargoes. Once they reached the destination province, they were sometimes helped by the army there to the final distribution point (i.e. Classis).

It is hard to differentiate Haltern 70 amphorae from Augustus’ contexts in this region from other JulioClaudian contexts, but they appear to have been imported in great numbers in this period in places such as Porto, Vigo or Braccara Augusta (Morais and Carreras, 2005). This amphora type was predominant in this region with percentages of around 60-50% of the overall amphorae assemblages. It came together with Gaditan fish-sauce amphorae (ovoids, Dressel 7 and Dressel 9). Other sites in the Cantabric Sea with Augustan horizons, such as Campa Torres, also reveal an overwhelming presence of Haltern 70 (Carreras, 2001).

Another Augustan reform was the creation of a kind of mail service (cursus publicus), a group of public messengers transporting letters and other kinds of public consignments from one side of the Empire with their official diplomata (Eck, 1999; Kolb, 2000). The cursus publicus required a minimal infrastructure for land transport with a series of relay points (i.e. mutationes – 12-18 km; mansios – 25 km) and the construction of a complex road system. Augutus and the army were the main road constructors of the Empire, and the emperor took over the responsibility of road construction in Italy in 20 BC.

It must be borne in mind that Haltern 70 distribution in the whole Roman Empire was concentrated in the NW Iberian Peninsula in a very short period, chiefly from Augustans to the Flavians (Morais and Carreras, 2004). Despite the lack of conclusive evidence, it is believed that such numbers of Baetican imports should be related to the army, either during the Cantabric campaigns and later, when army units were supervising gold exploitation from the NW mines (i.e. Las Médulas).

Such land routes were also the main communication networks for a flow of information from different points of Empire, but also the transport routes for staples. Testimonies of such movements of goods directed to the army are found in the extraordinary amount of amphorae recovered in the main military camps. In Octavian’s period, the new system started to be put in place, but it was a relatively simple one. For instance, amphorae recovered in the NW Iberian Peninsula showed a clear pattern of distribution related to the movements of the army during the Cantabrian wars (29-19 BC).

As mentioned above, Haltern 70 was a rather special amphora in that it could carry a variety of products, according to the inscriptions we have (Aguilera, 2004). This means that the same container transported different foodstuff from the Guadalquivir valley (i.e. defructum, olives, muria and perhaps wine). An uncorroborated suggestion is that a multi-purpose container might have been used carry the products of a unique consumer, the army. The suggestion remains a tentative one.

Related to the movement of staples towards the legions settled in the NW, there was also the supply of coins. Bellido (2007) pointed out that the first military coins supplying the Cantabrian campaigns were struck by Augustus in a mint of Lucus Augustus. (26-24 BC). They are the so-called “caetra” coins. They appear with the face of Augustus on one side and the representation of the Lusitanian-Gallaecian shield on the other, and the legend “Imp Aug divi f”. Distribution of these coins is concentrated in the NW sector of the Peninsula, the military sector. There is more excellent evidence of the Atlantic military supply of coins from the treasure of Alvarelhos (Porto) dated to 26 BC and consisting of 5000 denarii and silver flans (argentum infectum) with Gallaecian metrology (366 g). Most denarii are Italian and 6 are caetra-type (García Bellido, 2007, 161).

With regards to the control of military supply, no information is available for this period, so it may have still relied on the procuratores and controls by the army at the main reception points. As has been underlined, the Octavian period supposes a complete change in military supply, with more public involvement in collecting taxes through procuratores and a widespread collaboration of private trades in mobilizing goods. Publicanii lost control of this key supply area, perhaps because they were not completely reliable in the past. In addition, the Republican structure gave too much economic power to a few societates publicanorum, which could generate excessive tensions for the new leaders. One may wonder whether the Astures and Cantabri campaigns (29-19 BC) became special grounds for experimenting with some of the reforms that Augustus and Agrippa had in mind. Discharging part of the army, the existence of two treasuries, provincial division, tax gathering, military supply, and coinage issues all had an excellent test case in Hispania. All these Octavian reforms were evident quite early in the NW Iberian Peninsula.

A second period of coin supply (24-18 BC) in the NW coincided with the creation of 3 new mints: Emerita (24 BC), Colonia Patricia (18 BC) and Celsa (BC). After the foundation of Emerita Augusta (25 BC), the new colony also became the main supplier of currency to the troops in the province. However the mints of Colonia PatriciaCorduba and Colonia Celsa complemented such supply

155

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

amphora, the trader’s name and controller’s name, and when it was carried (Remesal, 1986).

Perhaps the changes observed in the NW were part of a more ambitious idea of Augustus, as he himself states in the Res Gestae (26):

Additionally there were officers such as stratores and beneficiarii who controlled army supplies in transit (Schallmayer, 1990; Carreras, 1997; Nélis-Clement, 2000). They were posted in the main hubs of communication, close to military garrisons, where they could check the quality and quantity of supplies carried to the army. Finally, the individual military detachments had their own officers (optio, signifier, primus pilum, praefectus castrorum) responsible for controlling supplies, as well as assigning them to the relevant units and warehouses.

“I extended the boundaries of all the provinces which were bordered by races not yet subject to our empire. The provinces of Gauls, the Spains, and Germany, bounded by the ocean from Gades to the mouth of the Elbe, I reduced to a state of peace. The Alps, from the region which lies nearest to the Adriatic as far as the Tuscan sea, I brought to a state of peace without waging on any tribe an unjust war. My fleet sailed from the mouth of the Rhine eastward as far as the land of Cimbri to which, up to that time, no Roman had ever penetrated either by land or by sea, and the Cimbri and Charydes and Semnones and other peoples of the Germans of that same region through their envoys sought my friendship and that of the Roman people”

Nevertheless, some army campaigns in inland regions where no maritime or river transports were available, required special treatment. Conveying large amounts of foodstuffs by land transport needed a special infrastructure of pack animals and fodder, not always easy to find in frontier regions. That is why a new office appeared in Nero’s time, the praefectus vehiculorum, a functionary who seems to have been directly involved with the land transport of official cargoes when there were peaks in demand (Eck, 1999; Carreras, 2003).

EPILOGUE Changes in military supply continued during the whole Julio-Claudian period and even later, when the new system was adjusted to new requirements and challenges. For instance, the control mechanisms were not efficient enough in the Augustan period, so it was improved over time. Testimony of such control systems are the painted inscriptions (tituli picti) on Dressel 20 olive-oil amphorae directed for the Roman army. They include sophisticated inscriptions with the weight of the empty and full

Sometimes the special requirements of military supply forced the creation of extraordinary posts, such as curator copiarum, during Marcus Aurelius’ involvement in the Marcoman wars (Remesal, 1986), but the system designed in the Octavian period appears to have remained relatively unchanged until the Late Roman Empire.

156

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

3.5 DEMAND AND MILITARY SUPPLY IN THE NORTHWEST OF HISPANIA THROUGHOUT THE EARLY EMPIRE1 A. Morillo spot, both from wild animals and from domestic ones, which were sometimes abandoned or confiscated. This is the hypothesis drawn up by Groenman-van Waateringe (1997, 261-262), who points out that the references in the Classical sources to the problems of the service corps to obtain some products points to a shortage at a particular moment or campaign, but not to its importance on the “standard”3 military diet.

DEMAND AND MILITARY SUPPLY: A STRATEGY MATTER The supply of troops, either in the field or camped, was a major challenge for the Roman military administration. The army organized a wide range of basic supplies for the different troops, from essential items to manufactured products. The value of these goods was taken from the soldiers’ pay. Broadly speaking, this system continued until the IIIrd century AD, the time when changes were made to the systems (Breeze, 1993, 536). Throughout the Early Empire, the service corps supplied troops and officers with clothing and articles for personal use, household furnishings, currency and, of course, food. 1

The diet of the regular army set up by Augustus seems to be much more varied and rich. Apart from corn, some other foods were available: fruit, vegetables, olives, meat, salt pork, salted meat or fish, cheese and, of course, wine and oil, not to forget salt, vinegar and honey. The paleobotanic and archaeozoologic research carried out on the Roman camps along the northern frontiers of the Empire gives evidence of the presence of some species of animal and vegetables. The amphorae tell us, partially at least, about the arrival of wine, oil and salted meat and fish. However, we hardly have any data on the rest of products and we can only know about their arrival through texts (Davies, 1971).

Food was of course the most important requirement. Minor problems with supply could cause a defeat or a mutiny, as Vegecio2 quotes. Judging by the references from classical sources (Labish, 1975, passim), the diet of a Roman soldier seems to have been based mainly on corn (Davies, 1971: 138). This appears to be true, at least, for armies in the field, where this supply was a real logistical problem. There can be no doubt that corn (wheat, barley) has undeniable advantages in terms of its preservation and delivery, which was made in sacks, whereas troops on campaign cannot carry vegetables, fruit or fresh meat. But, together with corn, some researchers claim the complementary role of some other food, such as meat, which is always possible to get on the

The service corps not only dealt with supplying food to the military camps. Some other goods and manufactured products were also necessary: horses and beasts of burden, timber, wooden and leather objects, woollen and linen fabrics, plant fibres such as esparto-grass for cordages, weapons, armour, and some objects made of bronze and iron, herbs and medicines, bone items, glass vases and containers and pottery of all kinds (terra

1 This work is a part of the following research project: Campamentos romanos en Hispania: análisis diacrónico de las estructuras defensivas. It has been given by the Ministerio de Educación y Ciencia (HUM200600534), which is being undertaken from Ist October 2006 under my direction. A first version of this same article has been published: A. Morillo, “Abastecimiento y producción local en los campamentos romanos de la región septentrional de la Península Ibérica”, A. Morillo (ed.), Arqueología Militar Romana en Hispania. Producción y abastecimiento en el ámbito militar, León, 2006, 33-74. 2 De rei militari, III, 3.

3

This same author points out that the frequent references in classical texts to the problems of corn supply during the conquest of the Iberian Peninsula are due, on the one hand, to the fact that the campaigns took place in summer, when corn was not ripe enough to reap and it was necessary to bring it from the stores where the previous harvest was kept; on the other hand, due to the few inhabitants in some zones, such as the two Mesetas, which is reflected on the few farming resources, it was not enough for the Roman army to be able to maintain position (Groenman-van Waateringe, 1997: 262).

157

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

sigillata, lamps, common ware and thin-walled pottery, construction material, etc.). All form parts of the essential supplies the army must get to maintain optimum operating capacity and, also, to raise troop morale (Morillo, 1992, 61; 2000, 623).

considered that the system was developed especially by Claudius. But, for the moment, we are not in a position to say to what extent the supply of the army that was settled in Hispania during the Augustan and Julius Claudian period was organized, compared with what we know for later times and that I have outlined previously (Morillo, 2000, 624; Morillo, 2006, 36). Epigraphic evidence concerning supply to the exercitus Hispanicus, recently collected, refers, on the whole, to a later moment than Vespasian’s coming to power, when the legio VII gemina was the only garrisoned legion in the Peninsula (Carreras, 1997, 165-167). Undoubtedly, the procurators Augusti, who appears in Hispania as early as Augustus’s reign, must have played an essential role within this embryonic organization, which had to supply the army with staples regularly.

Undoubtedly, the most important part of troop supply must have consisted of staples, although we have little information. Except for some ceramic containers holding oil, wine, olives or salted meat or fish, only manufactured products provide some clues about imports of provisions and the places they came from. But we have to deal with hypothesises about military supply, when we do not have enough information. One of the best known examples about this lack of information concerning supply can be found in the campaign of the Cantabrian wars.

The Julius Claudian period is the least known from the point of view of the administrative organization and yet it offers the great promise from an archaeological point of view. I am going to leave out the Republican period, as there was not, at that moment, a state institution in charge of army supply (Carreras, 2002, 70). On the contrary, the administration was consistent on this matter all through the Early Empire, finding solutions which depended on various elements: the general political and military situation; the step forward in the Romanizing process in Hispania; the specific needs of the State and the determined military strategy (in terms of the amount and arrangement of military units throughout Hispania) (Morillo, 2006, 36-37).

The supplies could be imported or locally produced in the camps, in the military factories4 or in the surroundings. In view of the high cost of long distance transport it was preferred to exploit of the surrounding area, which was sometimes under army control (prata).5 But selfsupplying only happens when agriculture and craft production in the civil hinterland is not developed enough to supply the military with everything they need by means of trade, taxes, or ordinary requisition (Groenmanvan Waateringe, 1997: 263). The supply policy of the state to the Roman army and the problems this involved are matters that have attracted the attention of various previous researchers, such as the work of Van Berchem (1937 and 1977) and Pavis d’Escurac (1976). However, it has been over the last decades that progress in the archaeological knowledge has made it possible to make more significant advances, resulting in some publications either of a general nature (Birley, 1981; Middleton, 1983; Wierschowski, 1984; Remesal, 1986; Herz, 1988; Breeze, 1993 and 2000; Adams, 1999), or more detailed archaeological analyses on specific areas (Middleton, 1979; Davies, 1997; Birley, 1997; Taylor, 2002), or about items such as pottery (Biegert & Lauber, 1977; Gassner & Jilek, 1997; Swan, 1997), or currency (Wigg, 1977; García-Bellido, 1996: 104-106 and 2004: 47-68).6

THE SUPPLY OF THE ROMAN ARMY THROUGH THE CANTABRIAN WARS (29-19 BC) It is highly likely that the new provisioning system started by Augustus had been put to the test for the first time during the Cantabrian Wars, which were a general trial for many aspects of the military strategy, later tested and regularized during the conquest of the northern frontiers (Morillo, 1996, 81; 2002, 69). Undoubtedly, preparing for such campaigns must have been extremely thorough with regard to military supply, as it involved transferring an army made up of some tens of thousands of men (seven legions together with their auxiliary corps) towards an isolated mountainous area and with no possibility of being supplied by sea or waterway (figure 3.5.1). On the other hand, the shortage of agricultural resources available in such areas made it very difficult to find supplies on the spot by means of requisition.

I am not going to deal here with some matters already analyzed in previous works, nor yet will deal in detail with the working of the administrative system, which has been studied fully by some authors such as Remesal (1986) and Carreras (1997; 2002). The main features of this organization could possibly have been set up by Augustus, who established the praefectura anonnae, in the last years of his reign. However, it is usually

From some texts by Cesar and other authors we know of the absolute tactical necessity to have the army constantly provisioned.7 From this point of view the campaign in the 26/25 BC against the Cantabrian and Astures must have

4

Vegetius gives a list of the craftsmen the legions had. This shows to what extent the unit worked in the field as an almost self-sufficient production centre (VEG. Epist. 2, 7). Archaeology gives us more and more evidence (essential bibliography, Petrikovits, 1974; 1976 and 1976b; Sander, 1962; Driel-Murray & Gechter, 1984; Hanel, 2006). 5 For the exploitation of military territory: Schulten, 1894; Mócsy, 1953, 1966 and 1967; Mac Mullen, 1963; Vittinghoff, 1974. 6 For the Iberian Peninsula: the recent monograph (Morillo, 2006).

7 CAES. Civ. 1, 72, 1; CAES.Gal. 4, 32; J., BJ 2, 528. Vegetius graphically explains that the troops were defeated more frequently by hunger than by battle (VEG. Epi. 3). Chapter three of book III by this author is wholly devoted to the tactical needs of the army to take care of food supply and forage.

158

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

Figure 3.5.1 Distribution of the Roman camps in the north-western area of the Iberian Peninsula during the Cantabrian Wars (according to A. Morillo)

Considering the archaeological evidence that is little by little being released, it seems much more likely to think that provisioning either during the campaigns of 26/25, or 19 BC commanded by Agrippa, were carried out, mainly by means of supply centres on the coast of Tarraconensis, where the goods coming, possibly, from Italy and other Mediterranean regions would be stored. It would be through the valley of the Ebro, by waterway or land, that military equipment would reach the northern battlefields, where harassment and punishment operations took place. It appears less probable that the Baetica played any direct role in this respect, but we should not forget that the legions stationed on the western front that attacked the Astures from the Galician area, must have been supplied with all they needed. The abundant Betic agricultural resources might have been directed to this theatre of operations by sea up to the Rias Baixas in Galicia and the low course of the river Miño, where they would have followed on by land. In any event there is no archaeological evidence beyond pure hypothesis. As I have already remarked, the written sources seem to give the northern region of Gaul (the Aquitania province) some role in the supply of the Roman army during the Cantabrian and Astures campaigns. Such role has not yet been traced.

been very hard since, despite more and more archaeological evidence, there is a lot of popular mythology about this topic (Morillo, 2006, 37-40). One of the few references to the problems of military supply during the Cantabrian Wars we have is the mention to the classis Aquitanica. According to Strabo (STR.3.4.18), Floro (FLOR. Epit.2.33.46) and Orosius (ORO. Hist.6.21.4), this squadron, coming from the ports in Aquitania, could have landed somewhere on the Cantabrian coast during the war against the Cantabri and Astures. Floro and Orosius give this controversial intervention a purely military role, but Strabo points out that the intervention was made due to the fact that the army had a need of wheat and other food supplies. The archaeological remains cannot back this early Roman presence in any of the known Cantabrian settlements, although an occasional Roman landing need not have left clear material or constructive proof (Fernández Ochoa & Morillo, 1994, 39). In this same work I have already shown some indirect data which pointed to the former Portus(S)Amanum-Flaviobriga, the current Castro Urdiales, as the detachment of a possible Roman intervention by sea (Fernández Ochoa & Morillo, 1994, 182-183), which could also have taken place at other coastal sites, such as the bay of Santander (Morillo & Fernández Ochoa, 2003, 447).

Being an army in the field, most of the equipment needed would have been food and animals, whereas manufactu159

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

red products would hardly have any importance. This can be proved thanks to the few ceramic remains found at the first field camps in Cantabria and Asturias (Peralta, 1999; 2000, 273-282; 2001; Camino Mayor et al., 2002).

Two different kinds of minting have been identified: first the sestertii, dupondii and asses (RPC 1-3), probably made in north Galicia about 27-23 BC (Villaronga, 1969, 593; García-Bellido & Blázquez Cerrato, 2001, II, 287). The finding of coins belonging to this first issue, together with no minted tokens in Lugo, has made it possible to mark this place as a centre of production (Ferrer, 1996, 283-288). There were also the asses of a rougher kind (RPC 4). Some years ago I referred to the great concentration of coins belonging to this second group on the southern slopes of the Cantabrian Range and to the east of the León mountains; that is to say: in the Meseta, specially in the camp of Herrera de Pisuerga (Morillo et al., 1996, 201-204). Their typological features, as well as the different areas they reached, and their stratigraphy in the camps in the Meseta, reassured the hypothesis that we have two different mintings from two different dates (Morillo et al., 1996, 202). Recently, García-Bellido has described the coins of type RPC 4 as copies, extending their time of minting, at least, until the change of Era (1996, 107).

Research into coin supply to the troops during the Cantabrian Wars and the Claudian period as a whole has made significant progress (García-Bellido, 2006). Beyond all doubts it seems nowadays that the bronze issues characterized by a caetra or Galician buckler on the reverse, sometimes together with falcata and dagger, are the official coinage, whose minting was boosted by the military authority responsible for meeting the supply needs of the troops transferred to the northern region during the Cantabrian Wars. The amount of such finds together in a particular place, as well as the implied meaning of the iconography on the reverse, indicates their military nature (Morillo & Gómez Barreiro, 2006, 357) (figure 3.5.2).

The preliminary data provided by the legio IIII in Herrera, released some years ago, have been endorsed by the whole analysis of the monetary circulation in such camps. There, most of the coins with a caetra back of the type RPC 4, are found in archaeological areas dated to the time of Augustus, that is to say, between 10 BC – 10 AD, judging by the association of materials (Morillo et al., 1996, 202). It is interesting to link the minting of the second group of caetra to the moment of the settlement of the legio IIII Macedonica in its camp in Herrera de Pisuerga, which took place at the time of the campaign of M. Agrippa against the Cantabrians in the year 19 BC. This finished with the complete subjection of this population (Morillo, 2000b: 619). In this situation, it makes sense that Agrippa released a new coin issue with a caetra reverse, as happened during the first campaigns of the Cantabrian Wars, so as to supply the troops stationed on the western front. This would prove the concentration of the group RPC 4 in the camp of Herrera. I would date this group to an early time, but later than the first issue (19 BC), and they would no longer be considered as copies due to the fact that they had been issued by the main military authority (Morillo & Gómez Barreiro, 2006, 258). Trying to identify, from an archaeological point of view, the currency used by the troops during the Cantabrian Wars raises several problems. Archaeological contexts from this period clearly identified are few in number. However some sites, such as La Espina del Gallego (Peralta, 1999) and El Cincho (García Alonso, 2002), are starting to give some information about the currency used in the field by the Augustan army. Together with this data, we have the installation level of the troops in settlements like Herrera de Pisuerga, and, above all, we deal with coins lost in the first years of the fixed Augustan camps. These coins are found in later archaeological horizons. From a numismatic point of

Figure 3.5.2 Moneta militaris with reverse of caetra found at Astorga (courtesy: Imagen MAS) 160

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

Figure 3.5.3 Distribution of the Roman camps in the north-western area of the Iberian Peninsula during the Armed Peace (19/15 BC-10/20 AD) (according to A. Morillo)

León mountains. The legionary camps of the legio III Macedonica in Herrera de Pisuerga, the legio VI victrix in León and the legio X gemina in Astorga and Rosinos de Vidriales were the main centres of operations inside a strip of military area, like a limes, which anticipated the strategy applied some years later during the campaigns that took place at the southern frontiers of the Empire (Morillo, 1996, 80-81; 2002, 77-78). The deployment of a considerable military contingent in the north of the Peninsula continued throughout the Claudian period (figure 3.5.3).

view, the circulating monetary group during the Cantabrian Wars is defined by the presence of: late Republican and Augustan denarii of the first issues; P. Carisio’s denarii minted in Emerita, issued with a reverse of caetra, denarii; residuary Iberian unities and Roman bronze issues from Lugdunum, Nemausus and some mints in the Tarraconensis. These coins arrived at the different camps in the pockets of the troops placed here to finish off the conquest (Morillo and Gómez Barreiro, 2006, 343). The army must have been re-supplied with money by the authority during the development of the campaigns. In this sense we must interpret the presence of Augustan denarii issued in Rome or in Emerita as well as the bronze issues with caetra reverse minted at two different times; perhaps those coinciding with the two campaigns against the northern people.

We have much more archaeological evidence about this period (19-15 BC – 10/20 AD) than that of the Cantabrian Wars. The excavations on the permanent legionary camps from Augustan time also give some important information about products and the army supply lines.

THE SUPPLY OF FIXED CAMPS IN AUGUSTAN TIMES: IMPORTS AND SELF-SUPPLY DURING THE ARMED PEACE

The border military settlements, set up in areas away from the key places of the Empire and from the main roads, either waterway or sea, had to survive, at first, by themselves; they relied on the resources that the military management left and their administration of the surrounding areas. Undoubtedly, major items, especially food, must have been imported from other sites in the Peninsular or other Mediterranean areas, as happened during the Cantabrian Wars. A little evidence available at

As a consequence of the end of the campaigns against the Cantabrians and Astures, most of the military units that took part departed for new conflict zones and the others placed on permanent stations arranged as a defensive cordon along the northern range of the Cantabrian mountains, the frontiers, and the eastern slopes of the 161

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

examples have been found of wine amphorae of an Italic origin such as the Dressel 2-4 from Campania (Carreras, 1996, 206).

this early stage of the northern camps that sheds light on the origin of some products (corn, wine and oil) is given by the manufactured products, which came, mainly, from central and southern Italy and found their way to the Hispanic settlements.

However, the high costs of goods transport by land towards the Hispanic camps that settled in the peripheral areas recently conquered and militarized, and far away from the main communication routes, resulted a few years later in the development of a complex craft system used to fulfil the basic needs for manufactured products (Morillo, 1992, 167). Judging by the data that are being released, we understand that the military workshops from Augustan times were a part of a policy undertaken by the Roman army so as to respond to supply problems.

The study of the importation of manufactured products, especially ceramics, found in military contexts, is based on stratigraphy at such sites and they make a valuable contribution to the reconstruction of historical processes. The early contexts in Astorga, León and Herrera de Pisuerga camps show a great amount of Italic terra sigillata containers (ITS), with or without decoration, with shapes belonging to 10-15 BC and 10-15 AD. The study of potter’s marks has made it possible to know the origin of these ceramic manufactures, as well as to establish the possible distribution routes. Most of the items came from the production centre in Arezzo, followed at a distance by Pisa and some other centres, such as Puzzoli, the Po Valley and Lugdunum (cf. Morillo and García Marcos, 2003b, 296, charts 1-3; see also Morillo et al., 2005, 142-144). Although we do not know how it was organised from an administrative and commercial point of view, the close relationship that it has been possible to establish between producers such as the Ateii and the anonna militaris (Menchelli, 1997, 193), which, judging by Hispanic finds, could possibly be extrapolated to some other producers like P. Cornelius, highly present in the castra of the legio VI victrix in León, suggest some interesting deductions about the provisioning in the military centres of ITS, which was possibly commissioned from some particular workshops or potters.

The archaeological excavations undertaken in camps from Augustan and Tiberian times show the existence of local productions of all kinds of pottery, from local terra sigillata and lamps, to common ware and thin-walled pottery (Morillo & García Marcos, 2001; Morillo, 2006b, 182-185). The best known examples are finds of Italic terra sigillata made by L. Terentius in the camp of the legio III Macedonica in Herrera de Pisuerga (García y Bellido. 1960; 1956/61; Pérez González, 1989, 199-240), which are an exception in the Roman world due to the link of a figlinarius with a specific military unit. Today more than one hundred marks have been found on cups and dishes that have been recovered, especially in Herrera de Pisuerga. Recently, two finds from the level of the camp of the legio X in Astorga and one more mark found in Caesaraugusta have been released (Morillo & García Marcos, 2001, 151; Morillo, 2006, 43, note 18). Terentius’ officina must have been in operation between 10 BC and 10/15 AD (Pérez González, 1989, 214).

Together with the ITS of central Italic origin, at this early stage there are also lamps of types Dressel 4 and Vogelkopflampe, Loeschcke IA and Loeschcke III (Morillo, 1999, 65-66 and 73). Among the items of this last type, two fragments of glazed lamp, green-bluish in colour, likely to be from Latium or Campania have been recovered (Morillo, 1999, 86-87).

But, well known as they are, finds in local terra sigillata in imitation of Italic ones and signed by L. Terentius are not the only production made in the camp of Herrera. The recovered materials show the existence of a real industrial complex, possibly set outside the walls of the camp enclosure, in operation during the time of Augustus. Within this productive complex I have identified workshops dedicated to the manufacture of different ceramic products besides ITS as lamps (Morillo, 1992, 167-168; Morillo, 1993b; Morillo, 1999: passim), common ware and thin-walled pottery (Pérez González et al., 1991; Reinoso, 2002). These items indicate a real fabricae militaris (Morillo, 2000b, 618), very similar to the ones in border areas of the Empire (figure 3.5.4). The military ceramic complex from the time of Augustus in Herrera de Pisuerga had, at that moment, no parallels in the Peninsula. Nevertheless, it has recently been detected in the Augustan levels of the camp of the legio VI victrix in León a new production of local terra sigillata of an Italic tradition signed, in this case, by the potters C. Licinus Maximus, L.M. Gen ( ) and the “potter of the caliga”, so known by the shape of his un-epigraphic mark. However, unlike the productions by L. Terentius in Herrera, the name of the potter never appears with that of the military unit occupying the enclosure (Morillo & García Marcos, 2001, 151-154; García Marcos, 2005).

Within this initial period, there is some thin-walled Italic pottery, dated to the last years of the Ist century BC. I should place in Etruria, Latium, Campania, and in the Padua area, the possible production centre of common ware and thin-walled pottery found in the camps of northern Hispania. This can be proved by the presence of some fragments of pottery, signed by Aco and Sarius. There is also some thin-walled Gallic pottery, such as the bowls of the type Mayet X, ascribed to the workshops of Loyasse and La Muette (Lyon) (cf. Morillo & García Marcos, 2003b, 298-300). The common ware from this period shows a great typological variety too. The origin of most of these containers, like mortars, seems to be Latium and Campania (see Pérez González & Fernández Ibáñez, 1989; Alcorta, 2002, 148-152). Concerning the amphorae, together with some Baetic items for salted meat or fish (Dressel 7-11), some

162

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

documented through an important complex of Augustan milestones, land boundary stones or prata of the legion in relation to the land of the nearest eastern towns, to be exact Iuliobriga and Segisamo (cf. García y Bellido, 1981, 118); (González Echegaray & Solana, 1975, 157175); (Le Roux, 1982, 114-118; Morillo, 2000b, 613614). The epigraphic formula is always the same although the abbreviations and links vary slightly: terminus augustalis dividit prata legionis IIII et agrum Iuliobrigensium (or Segisamunensium). As for the rest of the legionary camps, it is only known an Augustan milestone assigned to the legio X gemina (Descosido, 1982), even though it is dated at a time later than the reign of Augustus, when such unit quarters on Rosinos de Vidriales. Also from a Claudian time, seems to be the deposit of Augustan milestones found in Castrocalbón and Soto de la Vega, where it is demarcated the territory of an auxiliary unit, the cohors III Gallorum in relation to the ones of the ciuitates Baeduniensium, and Luggorum (García y Bellido, 1961, 155-159; Roldán, 1974, 439440). However, neither the camp of the legio X gemina in Astorga nor the one of the legio VI victrix in León have given, for the moment, any news about the existence of a military territory round such quarters. Perhaps in these cases, and following the hypothesis drawn up by Groenman van Waateringe (1997, 263), the civilian frame could have been developed enough from an economic point of view to provide supplies to the service corps, which made it unnecessary to segregate an area and place it under the military authority of the Army for its direct exploitation, as it could have been the case of the camp of the legio III in Herrera de Pisuerga. Anyway, the study of the characteristics of the military territories still raises many mysteries, either inside or outside our frontiers. We do not know exactly their measures, ways of setting in relation to the nearby population centres, their economic role or how their exploitation was. Le Roux (1982, 114118) gives a complete picture about the matter.

Figure 3.5.4 Lamp of the type Dressel 4 (Vogelkopflampe) from the sector San Millan (Herrera de Pisuerga), manufactured in the military workshop of legio IIII Macedonica in this town (Photograph: A. Morillo)

Although it has been possible to confirm its manufacture in C. Licinius Maximus’ workshop, the find in the same archaeological levels of a lamp fragment on which it might be possible to determine the mark of the legio VI victrix, L. V. I, in embossed capital letters (Morillo, 1999, 296-297 nº 40, fig. 170), could be, in some way, related to the output of a potter from León. Although we have little information at the moment, in the Augustan camp of Herrera de Pisuerga it has also been possible to detect the manufacture of bone (Pérez González & Illarregui, 1994; Matía, 1999) and metal tools (Fernández Ibáñez, 1999; 1999b; 2002), which could possibly have been together with some textile and leather items.

Frequently, the local agricultural produce was not enough or it did not supply the necessary products. In this case, they had to turn to the supply from other regions. The provision of corn, wine or oil was directly controlled by the State, which carried out an economic policy of directed market, such as Remesal (1986, 111) has well defined, directed towards very specific targets: the provision to the large city and the forces deployed in the border provinces. The State must take on the economic load of such transport regardless of its cost, as a public service intended for the support of its defensive army but it, logically, tried to reduce the costs of transport.

Dealing again with the matter of basic commodities supply and despite the difficulty its archaeological and epigraphic verification involves, the accounts endorse the fact that as the reign of Augustus goes by, it is also verified in this field the same process of regularization and rationalization of the provisioning. As in the case of manufactures, this rationalization and decreasing of costs leans on the development of local resources as far as possible. In some camps the surrounding area is put under the direct rule of the army (territorium legionis), which will farm the land so as to get the agricultural and livestock farming produce set aside for the direct sustenance of the legion. The best-known example in the Iberian Peninsula is the one of Herrera de Pisuerga,

Within the area of cost reduction, another feature that confirms the rationalization process and the optimization in the provision of key resources developed by the service corps throughout the reign of Augustus is found in the archaeological evidence of amphorae, that shows the increasing participation of other provinces and Hispanic territories in food supply, as the reliance on foreign imports loses importance quickly. In the Hispanic 163

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

military archaeological registries, the few documented amphorae of Italian origin make way for ceramic containers for salted meat or fish and oil from the Baetica and wine from the eastern Tarraconensis.

north of the Peninsula, taking into account the insecurity that the historical and archaeological analysis of this kind involves. Yet, an increasing presence of salted meat or fish containers of Betic origin is stated, as well as wine amphorae from the Layetanian zone. Imports from the coastal zone of the eastern Tarraconensis, following the valley of the Ebro river, where the newly-founded Caesaraugusta colony starts to become a first-rate trade centre and the Baetic centres through the capital of the Lusitania, would turn these into the main routes of military supply.

Some years ago I dealt with the problems that the provision of oil to these camps posed, where, in spite of having few remains of oil amphorae Oberaden 83 and Dressel 20, the abundance of lamps confirms the existence of considerable provision of oil coming from the production centres in the Baetica. The answer to this apparent contradiction would be the existence of alternative containers for the transport of oil, about which we can only speculate, more than on interpretations built from a supposed weak romanization, or from the rejection of oil for cultural reasons (Mañanes & Balil, 1974-75).8 Perhaps, amphorae or ceramic containers, whose typologies do not correspond to that of the containers normally employed for oil transport, and about which also have little evidence, were used. But it is more likely, in our opinion, that non ceramic containers, whose low weight would facilitate its transport by land on carts or on baggage mules, were used. Rather than barrels, which have the added difficulty of permeability, I am inclined to the use of wineskins (cullae), the usual containers for oil packaging in bulk until a few decades ago, well evident in the sources, the epigraphy and the iconography (cf. Marlière, 2002). The difference between the military settlements on the northern border of the Empire, where oil amphorae from the Baetica are much more abundant (Remesal, 1986; Carreras & Funari, 1998, 41), and the Hispanic bases, lies on the urgent necessity to provide these by land, whereas the trade by water and sea is perfectly feasible in the first case (Morilllo, 2000, 629).9

I have already pointed out previously that the traffic of provisions of all types to the army from the producing or commercial centres in the east and the south of the Peninsula was an important stimulus for the development of an incipient trade towards the peripheral regions in the north of the Peninsula, as is proved through some finds of objects such as the lamps (Morillo, 1999, 324-325). But, undoubtedly, the demand caused by the several tens of thousands of soldiers for products like grain, wine, oil and salted meat or fish, a need that had to be filled by the annona militaris, that had to meet the costs of supply, must have had an effect of an enormous importance within the producing agricultural zones, either in the Tarraconensis or in the other two Hispanic provinces. The Hispanic producers and negotiatores must early on have seen the possibility of high benefits if they obtained a sole contract of provision with the military administration to supply certain agricultural and farming products. And, compared to their Italian, Gallic or African competitors they had the advantage of geographic nearness which reduced significantly the cost of transport, as well as large expanses of fallow land which could be cultivated or reoriented from a productive point of view.

This hypothesis raises the difficulty of direct archaeological verification, which, obviously, is not easy to obtain due to the perishable character of a material like leather. Nevertheless, in our opinion, the facts I present here on the high levels of oil consumption in the main military and civil settlements in the northern region of the Peninsula, in spite of its indirect character, are an unanswerable proof that the transport of such substance in wineskins, destined either for culinary use or as fuel, reached considerable levels in the inner areas of Hispania (Morillo, 2000, 629).

Whenever the study of the economic history of the Roman Hispania is approached we have to face a problem springing from what Balil called “geographic restriction” of the economic documentation, since both the literary references and the archaeology focus especially on the Baetica and, to a lesser extent, on the Mediterranean coast. On the contrary, large regions like the two Mesetas, the valley of the Ebro river and the Lusitania are blacked out by the lack of evidence (Balil, 1972, 59). This fact means that we always think about economic production in the Mediterranean Hispanic areas in terms of exporting to foreign markets, but never towards the inner part of the Hispanic provinces. It is essential to break this centrifugal tendency in the analysis of the economic history of Hispania, appraising the whole inner market as it deserves to be. Undoubtedly, the supply to the Hispanic army became an essential factor for the Hispanic market from the reign of Augustus. It is likely that the needs for supply of the northern army were an invaluable incentive or driving force for the economic activity of the Hispanic provinces, both for manufactured or agricultural products (Morillo, 2006, 48).

The case of oil supply could give us a perfect example of the casuistry of other nutritional products for which our information is even less, like wine, salted meat or fish and, particularly, cereals. Only the typology of amphorae shaped containers lets us establish, a priori, producing regions and routes of supply to the Augustan camps in the 8

Those hypotheses are only based on the blank areas that the maps of amphorae finds in northern Hispania show, but other matters, such as the presence of the army or the abundance of lamps for lighting have not been taken into account. 9 For the majority of researchers, transport by land of oil stored in amphorae meant an increase in the cost almost impossible to bear. Sillières has calculated that over 100 miles, the cost of a product transported by land will be doubled.

Taking again oil as the departure point for our analysis, it is really significant that the massive planting of olive 164

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

the Ebro were encouraged by the State (Morillo and Pérez González, 1990, 459).

trees in the Baetica took place about the 20 BC, that is to say, at near the time the Cantabrian wars finish. Such chronology has been established by Chic from the verification of the first massive exports of Baetic oil towards the German limes, which happened about 10 BC (1997: 98). At this time, there must have appeared the first Baetic oil amphorae (Oberaden 83), which made way for the characteristic Dressel 20 (Berni, 1998, 33) at the beginning of Claudian times. But it also draws our attention to the temporal coincidence of the foundation of two new Augustan colonies in the eastern part of the Baetica.

The recent analysis of the circulation of money we have undertaken in the camps of Leon and Herrera de Pisuerga (Morillo & Gómez Barreiro, 2006; 2006b)10 confirms that the real monetization of these camps took place in midAugustan times. Such activity mainly depends on the bounteous municipal and colonial flotations in the valley of Ebro minted in the reign of Augustus. The mint houses of Caesaraugusta, Calagurris and Celsa provide the majority of the supply under Augustus, especially with the mintings from post 15/13 and ante 2 AC.11 These cities seem to assume the main role in supplying the camps, chiefly Caesaraugusta12, and their issues become the main part of the money supply in these centres.

I am talking about Acci (Guadix, Granada) and Tucci (Martos, Jaén), whose foundation in a senatorial province far away from the conflicts has always caught attention (Roldán, 1974, 188-194). Do we have proof of a process of agricultural colonization impelled and encouraged by the Roman State using veterans of the Cantabrian campaigns, whose final objective would be to increase the production of certain products, such as oil, to meet the needs of troops on active service? (Morillo, 2006, 48).

As for the supply of bronze coins, a position similar to the issues of the mints in the valley of Ebro, led to the silver coins devoted to Augustus’ grandchildren Caius and Lucius (RIC I2: 205-212). These are widely represented across the Peninsula, especially in the northwestern area (Centeno, 1987, 214-216).

This same process of encouragement of agricultural development, focused on products like wine or oil, could be verified in other regions, such as the Catalonian coast, the valley of the Ebro river, or Lusitania. Even the development of the salting industries set along the Betic and Lusitanian coasts must have been driven, somehow, by the demand of military supply (Morillo, 2006, 49-50). It is much more difficult to search for the origin of cereal supply, whose marketing in sacks makes it remarkably difficult for archaeologists. I am not in a position to show the origin of the cereal sent to the military centres in the north, since it could come either from the ager publicus of the Hispanic provinces or from the privatus. It is possible that the Baetica, the most romanized and intensely exploited province, which, judging by what we know about Claudian times, had a wheat (D.C. 9.24.5) surplus, should have borne, at first, the main part of cereal supply to the Hispanic troops, while the farming of other regions did not allow the diversification or reorganization of the sources of supply (Morillo, 2006, 50).

Centeno (1987) has related this abundance to the payment of troops. Such issues have usually been attributed to Lugdunum, but the great amount and dispersion of the silver series have made it possible to admit the working of several auxiliary mints. One of them, at least, would have been placed within the Iberian Peninsula. This is a proposal supported by the discovery of two impressions from the issue RIC I2 207 in Calahorra (Giard, 1976, 52 and 226; RIC I, 2: 55). The creation of an “official” auxiliary mint of Lugdunum in Hispania meant the silver mint would have had the same beneficiaries in mind as the chief mint had. If the denarii from the Gallic city were intended for the military in Germania, the auxiliary Hispanic mint would be used for the payment of the stipendium militare to the three legions detached in Hispania (IIII Macedonica, VI victrix and X gemina) and to the auxiliary troops (García-Bellido, 1996, 104). The available archaeological data about the supply of other items is almost non-existent. There has been work on the relation of the army quartered in the north, and on M. Agrippa in particular, to the exploitation of lead in the Serena (García-Bellido, 1994-95) and the zone of Linares (Rodá, 2004). We cannot overlook that lead has both civilian and military applications, so some of the

In spite of all the indirect evidence, it is not possible to measure yet the direct incidence of the supply to the Hispanic army on the economic history of the Augustan and Claudian period as a whole. The money supply to the troops stationed in the northern region seems to be the cornerstone of the military policy applied to Hispania. The great majority of the circulating currency between 27 BC and 41 AD in the Iberian Peninsula belongs to the abundant Hispano-Roman series minted from the reign of Augustus in the workshops of the Hispanic provinces, especially in the valley of the Ebro river. As I have already pointed out, it seems to be the need to provide the necessary money for their minor transactions to the several thousand soldiers quartered on the region, that explains why some mints in the valley of

10

This analysis was a part of the research project by the MCYT under Mª P. García-Bellido from the CSIC (BHA2002-00047). 11 So as to define the horizons of Hispano-Roman coin minting we follow García-Bellido’s hypothesis (1996b; 2000). 12 The data we used until some time ago did not make it possible to hold this importance for Caesaragusta in terms of the military supply, Calagurris being the mint best represented in the military contexts (Morillo and Pérez González, 1990; García-Bellido, 1996, 104). The new data from the northern camps within the quoted research project by MªP. García-Bellido has confirmed the outstanding role of the colony in the supply of the troops (Morillo and Gómez Barreiro, 2006, 373).

165

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

production might have been destined for the northern camps to make either slingshots or other objects of military use (Morillo, 2006, 51-52).13

phase is characterized by the development of applied decoration and the appearance of marks in planta pedis from the 15 AD on (Morillo and García Marcos, 2003, 301; Morillo et al., 2005, 144-146).

TIBERIUS AND THE DIVERSIFICATION OF THE SOURCES OF SUPPLY (10/20-40 AD) At the beginning of the reign of Tiberius, despite the basic continuity of the policy instigated by Augustus, a thorough military reorganization was undertaken (Morillo, 2002, 83-86; 2006, 52). Such reorganization can be seen in the internal restructuring experienced at that moment by the camps of the legio III Macedonica in Herrera (Pérez González, 1996, 91; Morillo et al., 2006b, 317) and of the legio VI victrix in León (García Marcos, 2002, 178; Morillo and García Marcos, 2004, 275). One of the clearest consequences of the changes between permanent and less permanent military bases was the abandonment of Astorga in about 15 AD by the legio X gemina, which moved to Rosinos de Vidriales (Morillo, 2002, 84). In place of the former castra the city of Asturica Augusta was founded, the capital of the conventus Asturum and the first important civilian centre in the northern region of the Peninsula (Morillo and García Marcos, 2000: 598). The redefinition of military strategy as applied to Hispania was corroborated through the reports of Strabo14 and coinciding with the stabilization of Roman troops along the Rhein-Danube frontier subsequent to the defeat of Q. Varus in 9 AD. The gradual development of the process of Roman settlement in the northern region, organized mainly by the army deployed in the north of Hispania, required a better strategy for adapting to the new needs, among which the role of troops in large-scale auriferous exploitation stands out (Morillo, 2002, 83).

Figure 3.5.5 Base of container of local terra sigillata of Italic tradition signed by C. Licinius Maximus, potter linked to the Augustan camp of the legio VI victrix (Photograph: V. García Marcos)

At the beginning of Tiberius’ rule it is also possible to notice the arrival of the first containers of Gallic terra sigillata, which, at the start of the time of Claudius superseded almost completely the Italic imports. A surprising fact that I cannot leave out is that the incipient production of home terra sigillata of Italic origin stopped sharply at the beginning of the Tiberian era, coinciding with the army reorganization of quarters and forces. The closing of these Hispanic military workshops, at a moment when Gallic terra sigillata had still not begun to arrive in quantity, and, therefore, was not a great competitor for the markets, may be due to a decision in this respect by the military administration. Although it is not possible to understand the reasons at this time, it might have occurred from pressure by certain economic groups who decided to open a new market previously closed by the presence of local manufacturers.15 On the other hand, the suspension of the activity of the military ceramic officinae was not total and it seems to have only impacted on the production of local terra sigillata of Italic origin (Morillo and García Marcos, 2001, 155; 2003, 300).

From this moment on there is a gradual increase in the archaeological evidence, which makes it possible to make a clearer reconstruction of the composition and origin of military supply, at least in terms of manufactured products. Several signs tell us that the military supply system strengthened and was reorganized on firmer and more diversified bases, perhaps at the same time as the development of the praefectura of the annona were taking place. During the last years of the reign of Augustus and all through that of his successor, Tiberius, the Italic manufactures continued monopolizing the imports. It is during Claudius’ time that they were superseded for good by products from the south of Gaul. As for the ITS this

Thin-walled pottery is represented, in Tiberius’ time, mainly by hemispherical bowls of shape Mayet XXXV. Nevertheless there is still an important presence of shape Mayet XXXIII at the site of Herrera de Pisuerga because of the maintenance of the local production of thin-walled pottery (Reinoso, 2002, 377). The great variety of production centres of thin-walled containers from that moment on (Italy, southern Gaul, southern Hispania)

13

Within this same context, we should perhaps consider the sub-aquatic deposits of Comacchio and the barge loaded with lead ingots sunk in the Po River near Ferrara. The site of the shipwreck, as well as the legionary marks on the ingots, together with Aggripa’s name, may suggest that it was destined for some centre of metal processing in northern Italy where some legions were garrisoned during the time of Augustus (Morillo, 2005b, 24). We should bear in mind the need for lead before the imminent Alpine campaigns. 14 Strabo (STR. 3.4.20; 3.3.8) and later Tacitus (TAC. Ann. 4.5.1) show the new situation of the Roman Peninsular troops, reduced to three legions with their corresponding auxiliary corps.

15

166

We will come back to this point later.

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

makes it very difficult to pronounce on the definite origin of the items (Martín Hernández, 2007), yet there is no doubt some vessels continued to arrive from the centres of central and northern Italy. Other supplies possibly came from the workshop in Lyon (Reinoso, 2002, 375). On the other hand, at about the end of this period, there must have been deliveries of the first Betic containers of shape Mayet XXXIV and vases from workshops in the Ebro river valley, properly verified in Herrera and León (Reinoso, 2002, 375- 377; Martín Hernández, 2006, 406- 408). Common ware shows a great variety of shapes (jugs, pots, lids). Increasing productive decentralization makes it very difficult to determine the factories or the origin of the containers. Nevertheless, there are some exceptions, such as the dishes with Pompeian inner red slip (Morillo and García Marcos, 2001, 301). Also of central Italic origin seem to be the containers of glazed pottery (skyphoi, lamps, of the type Loeschcke III) verified in León and Herrera de Pisuerga (Astures, 1995, 281; Morillo, 1996b; Morillo, 1999, 85-87). The origin of most of the lamps is still Italy, although the military lamp workshop in Herrera de Pisuerga kept running, making items of types Dressel 4, Loeschcke IA and Loeschcke III (Morillo, 1992, 67-68; Morillo, 1999, 76-77; Morillo and García Marcos, 2003, 301). One of the most striking aspects regarding lamps is, together with the Italic imports, the presence of lamps of type Dressel 3 or the “Andujar type” (figure 3.5.6). The production of this type starts with Tiberius and continues through Claudius’ government. The dispersion of the Dressel 3 lamps confirms the location of their producing centres in southern Hispania, concentrated mainly in the valley of the Guadalquivir and its tributaries. Sotomayor identified a lamp workshop of such a type in Andújar, to which some other workshops in Córdoba and Mérida have been added in recent times (cf. Morillo, 1999: 105106 with bibliography). In the north of the Peninsula the lamps of Andújar type are concentrated mostly in the important cities and camps, mainly Astorga and León, or they appear isolated in sites next to the main north-south road links, principally the so called “Vía de la Plata”, which connects Emerita Augusta with Asturica and the Asturias Transmontana. In our opinion this special diffusion of the “Andújar type” artefacts arises from a sui generic trade, in which lamps are not the main objects, but secondary items that went with the loads of Betic oil towards the main northern settlements, placed especially in the conventus Asturum. The high cost of lamp transport by land would be covered by means of the oil delivery. The nearness to the “Vía de la Plata” of most of the finds may allude to this natural thoroughfare as the main route for the marketing of both products (Morillo, 1999: 325).

Figure 3.5.6 Lamp derived from Dressel 3 (“Andujar type”) found in the excavation in Casa Pallarés (León) (Photograph: A. Morillo)

This half-way site could easily be Emerita Augusta, the capital of Lusitania and the redistribution and marketing regional centre for consumer goods and manufactured products. The negotiatores or mercatores, whether on their own initiative to sell their wares in Asturica, or asked to do so by the people in charge of supplying the military troops still stationed in Herrera de Pisuerga, León and Rosinos de Vidriales, could have gone to Emerita so as to provide northern consumers with Baetic and Lusitanian oil without having the need to turn to the production centres. The lamps of type Dressel 3 could have been a highly profitable adjunct to this northern trade (Morillo, 2000: 630-631). The items found in the northern deposits were possibly made in the Lusitanian capital too, as has been proved by archaeometric analysis carried out on items from León, Astorga and Herrera de Pisuerga (García Giménez et al., 1999). The lamps of “Andújar type” captured almost completely the northern markets until the time of Vespasian. From a chronological point of view, the arrival of “Andújar type” lamps to this region coincides, as well, with the time when the few remains of Baetic amphorae of type Dressel 20 start to increase (Morillo and García Marcos, 2003, 301).16

Such lamps must surely be related to the Baetican oil trade which moved oil from Baetica, or from some halfway location, to the northern region of the Peninsula.

16

A first approach to the archaeological material found in the area of San Pedro (León) shows more presence of amphorae for oil Dressel 20 than in the deposits of Astorga and Herrera de Pisuerga. Perhaps, it is

167

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Judging by either the presence of “Andújar” lamps and amphorae of Dressel 20, it is highly likely that Baetica undertook the supply of oil to the northern centres. The analysis of other kinds of amphorae shows a significant presence of wine, most of it coming from Italy (amphorae of the type Dressel 2-4). Some other wine containers also start appearing: from the Catalonian coast (Pascual 1, the Baetica, the Eastern Mediterranean, Rhodes and the Gaul types (Gauloise 4, P&W 9, Rhodian). There are also some salted meat or fish amphorae of Baetic origin (type Dressel 7-11).17 The situation is very similar to the one depicted by Carreras and Berni for the neighbouring settlement in Astorga (2003), which will quickly become a regional centre for the redistribution of goods (Morillo et al., 2005, 146), and we can speculate similarly about the different supply of staples to the Hispanic army during this period. Together with the route from the Mediterranean ports through the valley of the Ebro there is another one of increasing importance: the one that goes from Emerita towards the north, following the line of the so called “Vía de la Plata”, or the former roads which communicated the Lusitanian capital with the Asturian zone through northern Lusitania (Morillo, 1999, 325).

areas of the Empire (RPC: 52). In the same way, some coin workshops that took the lead in the supply of copper coins provided to the troops, such as Celsa and Bilbilis, reduced their issues drastically. The main supply source during Tiberius’ time was centred on the workshop at Turiaso, followed by the mints of Calagurris, Caesaraugusta, Clunia and Cascantum. This is the situation that camps such as at Herrera and León demonstrate (Morillo and Gómez Barreiro, 2006 and 2006b), but also some cities, e.g. Asturica (Blázquez Cerrato, 2006). As for silver coinage, the abundance of Tiberian denarii with a reverse featuring Livia seated, found in northern areas of the Peninsula, often together with Gaius and Lucius’ Augustan denarii within the same contexts, has lead us to pose the possibility that, like these last examples, some Tiberian issues destined as pay for the troops, came from some auxiliary Hispanic mints (Morillo and Gómez Barreriro, 2006b, 356). Coinage supply to the northern camps stops suddenly at the end of Tiberius’ reign, or during the short rule of Caligula, coinciding with the departure of the legio IIII Macedonica towards Mogontiacum (Mainz) in 39 AD.

On the other hand, as I pointed out above, it is just during Tiberius’ rule that the land was exploited under military jurisdiction, at least in the case of the prata militaris arranged round the camp of the legio IIII Macedonica in Herrera de Pisuerga. This fact must have been farreaching in terms of the supply of certain products, such as meat, milk, vegetables and perhaps some kind of cereals. I have already pointed out that the Asturian camps behaved differently. Perhaps this was because the native population, in the process of acculturation, could provide such products directly, and, in this way, the military administration did not have to undertake directly the exploitation of cattle and farming. To this effect, we should not forget the major population reorganization that took place in the neighbouring Asturian mining districts, which resulted in the functional specialization of certain settlements to farming (Sánchez-Palencia, 1995, 148149). It is possible that farming surpluses were not only used for the support of the mining population but for military troops as well.

CLAUDIUS, NERO AND THE ROLE OF GAUL IN MILITARY SUPPLY Like Tiberius’ reign, this period is characterized, in the field of manufactured products, by the overwhelming presence of imported items, with local production in the minority, at least until a later time in Nero’s administration. As for the origin of the manufactured products, there is the almost total disappearance of Italic imports to be totally superseded within the Hispanic trade routes by Gallic manufacture of terra sigillata and the early Hispanic production of terra sigillata, as well as by lamps, thin-walled pottery and common ware from Gallic or Hispanic workshops. The relaxing of restrictions on workshops in Tiberian/Claudian times will have some consequences, such as the spread of small craft installations and consequent regional productive decentralization with the aim of reducing production and transport costs (Morillo and García Marcos, 2003, 302).

From the point of view of coinage supply, the Tiberian era meant a continuation of Augustan policy. Nevertheless during this period it is possible to notice a decrease in numismatic contribution in military settlements, which must already have been pretty well monetized at the end of Augustus’ rule. This decrease is also stated in other areas of the Peninsula (Blázquez Cerrato, 2002, 273-274; Ripollés, 1982, 324-330) and it was undoubtedly partially motivated by this Emperor’s own decision to centre in Rome the supply to the western

In the archaeological contexts of this period, which after the departure of the legio IIII Macedonica in the 39 AD relate just the legionary settlements of the legio VI victrix in León and of the legio X gemina in Rosinos de Vidriales,18 the containers of southern Gallic terra sigillata completely replace the imports of ITS in the stratigraphy belonging to the central decades of the first century AD (Astures, 1995, 275-279). Among the vases of STS the so-called marmorate stand out, with evidence of shapes Drag. 27, Drag. 18 and Drag. 19. Together with the imports from La Graufesenque, coming through the

the military nature of León which distinguishes this settlement from that in Astorga (Carreras and Berni, 2003). As for Herrera, the departure of the legio IIII Macedonia in 39 AD reduces the possibilities of finding such material in the archaeological record. 17 These remarks come from direct analysis of the materials in the area of San Pedro (León), dated between Tiberius’ rule and 70 AD.

18 The departure of this military unit to Pannonia took place in 63 AD; it returned to the Peninsula between 68 AD and the winter of the years 69/70 AD.

168

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

Mediterranean ports, is the first production from Montans. The abundance of containers from this pottery in the north of the Cantabrian range corresponds to its essentially Cantabrian spread (Fernández Ochoa and Morillo, 1994, 184-185) and lets us pose the possibility that, for the first time, containers from the Astures Transmontani territory reached the military centres (Morillo et al., 2005, 146-147). All through this period, the first production of Hispanic terra sigillata from the valley of the Ebro is documented, especially that of the so-called early potters, such as M.C.R., Asiaticus, Maternus and Ullo, who worked with shapes taken from the formal repertoire of the GTS. As for the lamps, the central decades of the first century AD are dominated by shapes Loeschcke IB, Loeschcke III and Loeschcke IV, as well as the Dressel 3 shape. If, as I have already pointed out, this last type is of Hispanic manufacture then it is more difficult to comment on the origin of other kinds of lamps; among those there are some Italic examples of great quality, but they seem to be, in the great majority, Hispanic manufactures, especially from the south but also from the valley of the Ebro river (Morillo, 1999). This is also the case with the thin-walled pottery, among which the Hispanic production, especially Baetic (Mayet XXXIV) and from the valley of the Ebro river, far and away outnumbers the foreign examples (Morillo and García Marcos, 2003, 302-303; Martín Hernández, 2006).

Figure 3.5.7 Local production of thin-walled pottery of the “Melgar de Tera type” (courtesy Imagen MAS)

With regard to the period between the last years of Tiberius and the central years of Nero’s rule, we have first-rate archaeological Hispanic contexts, such as the monopoly of the southern Gallic terra sigillata, which totally displaced the ITS in a few years. This is a perfectly documented phenomenon from an archaeological point of view, which, of course, is reflected at the Hispanic camps of Claudian time. There is no doubt that the significant presence of STS in such quarters implies that the south Gallic potters took charge of supplying this ceramic to the Hispanic troops. But if we take into account, with the exception of some highly- priced quality products, the high cost of long-distance transport of ceramic containers, the abundance of STS would appear confirm that such containers travelled as seconddary loads, or added value, together with basic necessities, on their journey from the Gallic ports (Narbo, Burdigala, etc.) towards military and civil Hispanic centres.

Like the thin-walled pottery, the common ware shows a clear tendency to the commercial and productive reorganization (figure 3.5.7). Only some dishes with a high-quality Pompeian red inner slip allows us talk of Italian imports. The amphorae found in the excavations in some military contexts make it possible to record, for the moment, a continuity of the tendency started during Tiberian times with regard to the import of staples. There is also a presence of wine Italic imports, from the eastern Mediterranean and the Catalonian coast, salted meat or fish Baetic amphorae, and some oil amphorae also of Baetic origin. Taking into account the information coming from the analysis of the amphoral containers, such regions as Italy and the Laietania coast (wine), and Baetica (salted meat or fish and oil) seem to have been in control of supply to the Hispanic castra.

What these products were we can only speculate on, but ruling out the oil, for obvious reasons, we must focus our research on wine and cereal.

Moreover, through this period, significant progress on the exploitation of Baetic olive trees took place (Chic, 1997b, 78). However, as I have already pointed out (Morillo, 2000, 627), it is a methodological mistake to deduce the presence or the origin of certain substances only from the archaeological verification of certain amphoral containers. It is necessary to have other alternative and indirect sources to solve this difficult matter.

Only a few amphorae archaeologically support Gallic imports of wine to Hispania. But, as has already been pointed out, the scarcity of amphoral containers may also hide, in this case, important exports of perishable containers such as wineskins and barrels. It is more difficult, if possible, to provide evidence for cereal 169

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

exports in sacks, an activity which, no doubt, must have existed.

2006b, 185). These would include the thin-walled containers made in the workshop of Melgar de Tera, placed near the camp of Rosinos de Vidriales, whose production started at this time (Gimeno, 1990; Lion, 1997; Carretero, 2000), or in the headquarters of the legio VI victrix (Martín Hernández, 2006, 410-411; 2007).

Nevertheless, despite the scarcity of evidence for the possible import from Gaul of some agricultural products and livestock to the Hispanic army, I cannot find another admissible explanation for the presence of such a great quantity of ceramics from Gallic potteries within military Hispanic contexts (Morillo, 2006: 59-60).19 A similar scheme, although it was not driven by government authority but by private merchants, must have existed in the early civil settlements, which started to appear in the northern region (Fernández Ochoa and Morillo, 1999; 2005). In both of them the predominance of the STS made in La Graufesenque shows us that the Mediterranean route was much more important than the Atlantic one which, however, has been perfectly well verified (Fernández Ochoa et al., 2005; Martin, 2005; López Pérez, 2005).

However, despite the close geographic relationship of such industries with the military troops, which clearly point to them as providing the main market, in both cases they seem to be civilian workshops placed in the cannabae (or surrounding areas) that spotted an opportunity to make a profitable deal for the manufacture of ceramic containers for a military market always hungry for consumer goods. At this time the production of brick in the legionary or auxiliary quarters also starts. In this case there is no doubt about the military nature of the production thanks to the legionary marks. In fact, in Herrera de Pisuerga at the time of Nero and Flavius (Pérez González, 1996, 95; Morillo, 2002, 88) some bricks have been found with the mark of the two new military units recently there, the ala Parthorum and the cohors I Gallica (Morillo et al., 2006, 320-322). In Rosinos de Vidriales some marks on bricks are also documented. In the case of the legio X gemina, these date possibly from 68 and 69/70 AD. (Wahl, 1984, 73-74), although other researchers do not rule out that the legion might have branded its production previously to its first departure in 63 AD (Martín Valls et al., 2002, 150). On the contrary, in the camp of the legio VI victrix in León, no legionary marks on this kind of material have been documented, which would confirm the introduction of the custom for military stamping at a later date in Nero’s administration, through the units coming from the Germanic limes transferred towards Hispania.

We must not forget that, through Claudius’ reign, the praefectura of the annona, responsible for military supply, shows a great development, boosted, no doubt, by the need for provisioning troops during the conquest of Britannia (Morillo, 2003, 25), which must have required much larger quantities of essential products, as well as improved organization. This import of manufactured products and, perhaps, of essential goods from Gaul to Hispania raises some difficult questions. Taking into account the high cost of long-distance transport, and that it is cheaper to use products from closer regions, was farming in the neighbouring regions, such as the high Ebro valley and the Meseta, not developed enough to supply the needs of the military centres? I rather tend to an ‘economic policy’ solution, whereby some interest groups – in this case the Gallic oligarchy – take precedence over others. The economic promotion of some regions can be obtained not only by means of a correct fiscal policy, but much more easily by purchases organized by the State, such as those which must have been destined for the military camps. To this effect, we must not forget Tiberius’ policy, contrary to that of the Italic aristocrats, or the promotion of Gallic oligarchs to the Senate by Claudius, which would confirm the significant economic role which, as happened later in the case of the Hispanics, some Gallic families would have played (TAC. Ann. 11.23-25). Perhaps the closing of the military manufactures at the beginning of Tiberius’ reign might have happened because of similar pressure by certain groups to open up a market, such as thin-walled pottery, previously protected in the military market.

Recently some workshops for the manufacture of militaria have been identified in the camp of the legio VI victrix in León: one of them on the outskirts and the other one inside the camp (Campomanes, 1998/99; Aurrecoechea, 2006). THE REGULARIZATION OF SUPPLY FROM FLAVIAN TIMES Some decades after the conquest, the departure of most of the troops in 69/70 AD, and the cutback in the Hispanic army to a single legion (the legio VII gemina, which had settled in León from 74 AD) and several auxiliary units, must have simplified the problem of military supply, which then continued without important changes until the mid IIIrd century. We also witness the definite provision of an administrative system of military supply with the appearance of the procurator per Asturia et Gallaecia (79 AD), the highest authority responsible for military supply, and the beneficiarii, who acted as go-betweens for the procurators and officers in charge of the military units, keeping control of the trade (Carreras; 2002, 79). Within the same organization there were the frumentarii, linked to the legio VII gemina, who appear in other

From late in Nero’s rule (ca. 60 AD), new local productions are documented for the military (Morillo, 19 The presence of the main pottery centres – La Graufesenque and Montans – in peripheral but perfectly integrated areas inside the main agricultural districts in southern Gaul would confirm the close link of ceramic containers to the commercialization of wine and cereal. They were commercialized towards the Mediterranean and Atlantic areas through the ports of Narbo and Burdigala.

170

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

provinces. Such civil servants are linked to the transport of official convoys, either to Hispania or from there to other provinces (Carreras, 1997, 167-168). The epigraphic evidence concerning the annonaria organization increases from Flavian times, making it more difficult to reconstruct certain elements of it.

meet at the conventus’ capital. One of them comes from Tarraco and, going through the important city of Caesaraugusta and up the Ebro river, passes through La Rioja and crosses the northern Meseta at the foot of the Cantabrian mountain range. The other major route, the so-called “Via de la Plata”, starts from the capital of Lusitania before crossing, south-north, the west of the Meseta, Moreover, Asturica is the start and finish point for routes which use the Cantabrian ports towards the central part of the Asturian coast, as well as for the three routes which join Galician lands to the Meseta. Through this complex road framework some Italic or African manufactures arrived in regional deposits. (Morillo, 1999, 325-326). On the contrary, the Cantabrian trade does not seem to have more than a complementary role.21 The lack of epigraphical documents cannot be a definitive argument since there is some indirect archaeological proof, such as the presence of HTS, which backs up some traffic of staples. Such traffic is confirmed by the existence itself of a Statio Segisamonensium in Amaya (Burgos) (CIL II.2915), perhaps a place to collect the portorium and for the control of goods (Carreras, 1997, 167).

On the other hand, the agriculture and livestock exploitation in regions such as the Meseta and the valley of the Ebro river reach their highest yield and are in a good condition to meet the needs of the troops deployed in the Tarraconensis. The most convincing proof of this exporting role taken by of some northern regions stems from the abundance of Hispanic terra sigillata (HTS), which, made in several workshops in the valley of the Ebro river (and mainly in the area of La Rioja), spreads increasingly through northern Hispania. In the same way as I suggested for STS, the HTS was used, no doubt, for staples (wine, cereal, and perhaps oil), which confirms that agricultural production in the valley of the Ebro was working at full capacity, despite the scarcity of direct archaeological documents (Solovera, 1987; Mezquíriz, 1999).20 Taking into account the ‘crossroads’ status of the high Ebro valley, the HTS made in the pottery workshops of the region cannot have been restricted to local products, but was employed for other goods from other regions (Meseta, the Catalonian coast, the middle and lower Ebro), which travelled either by waterway or land. All these northern regions with high agricultural production, together with the Baetica region, which had to take on the supply of oil, seem to be responsible for military supply from the last years of Nero’s rule. Recently, Carreras has pointed out the concentration of epigraphical evidence on the beneficiarii in the administrative centres in Emerita and Tarraco, as well as the northwest – where some inscriptions have been documented in Astorga, Luyego, Villalís, La Rua and Amaya (1997, 165-167). Only the inscriptions in Estepa, Figueras and Nimes escape this distribution. The absence of military staff along the main land routes has led this researcher to suggest that there must have been little road traffic of supply to the army, and that it would have been mainly by sea transport, especially the routes that went to the Atlantic ports from Asturica. Although I myself have upheld the importance of the Cantabrian seaway for the coastal strip (Fernández Ochoa and Morillo, 1994, 186187), and, obviously, sea transport is cheaper, the material found at sites on the southern flanks of the Cantabrian range, where there were, of course, military camps, suggests the unquestionable predominance of Mediterranean, Baetic, and Lusitanian products, as well as material from other inland areas, such as the high Ebro river valley (Morillo, 2006, 62).

Figure 3.5.8 Military stamp of legio VII gemina found in León (courtesy Museo de León) 21 On the other hand, as we have already pointed out in another place, we cannot accept as an argument to emphasize the importance of the Cantabrian commerce in this region that the roads between Astrurica and the Cantabrian and Atlantic coasts were used to drain the gold extracted from the Astures mines towards the ports to its transport to Rome waterway (Carreras 1997, 167). Not taking into account that there is no archaeological data to base this hypothesis, it is logical to think that the Roman administration would always choose to transport the merchandise by land, in the form of a well watched convoy with military troops so as to avoid the chance of losing the treasured metal and the subsequent monetary and economic upsets. Without any doubt they kept away from the long and dangerous crossing around the Peninsula and, although it has been traditionally held that the gold, transferred from Asturica to Tarraco, would be loaded in this last port to Rome, it is also possible to think about the transport by land through Gaul, a slower but safer route (Fernández Ochoa and Morillo, 1999; Morillo, 2000, 633).

From halfway through the Ist century AD, Astorga takes shape as the great centre of redistribution for northwestern trade. Two of the main routes of Roman Hispania 20 A quite important local production of oil has also been pointed out (Fernández Castro, 1984; González Blanco & Hernández Vera, 1984; Hanel, 1989; Carreras, 1997, 171). But it is possible that some installations to make wine have made out as oil mills.

171

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Among theses, weapons and military equipment (militaria), manufactured in army factories that are better and better known, stand out, as well as coinage turned out by official mints or workshops, temporarily supervised by the State. The situation is very different for goods requiring producers (and middlemen) outside the State: both essential products (cereal, wine, oil) and manufactured items. This group presents more questions about supply. The answers vary for supplies from Italic and Gallic producing centres during the Cantabrian Wars and the first stages of the settled Hispanic army, or when encouraging local production of essential products in particular Hispanic regions (Baetica, the eastern coast of the Tarraconensis and the Ebro river valley). And we should not forget the local supply of certain manufactured products, especially ceramics, witnessed in Augustan camps such as León and Herrera de Pisuerga. From Claudian time on, weapons and military equipment factories (even extending to brick material too) met the needs of some units. The system was definitely regularized from Flavian times, coinciding with the decrease of officers of the Hispanic army and the consolidation of farming exploitation in areas such as Meseta and Ebro valley, which reached their full potential.

Being the logistical operation perfectly integrated into an absolutely developed economic and commercial system, kept by a dense road network, the production of manufactured products is unnecessary. Except for some particular materials, like bricks, made in León (García and Bellido, 1970, 588-599); Campomanes, 2006) and A Cidadela (Caamaño, 1989) (figure 3.5.8), or militaria, whose production has also recently been verified in the camp of the legio VII gemina (Aurrecoechea, 2006), the local productions present in military environments came from civilian workshops established on or around the cannabae. SOME LAST CONSIDERATIONS The supply to troops, either if they were in the field or settled at permanent camps, was a major problem for the Roman military administration. Solving it by the army depended on the type of products and the actual potential of the province (and neighbouring provinces) where the troops settled. Both the real needs of each unit and the potential of each province with regard to the supply changed significantly through the High Empire. However the archaeological evidence known so far shows that the military supply policy on the part of the appropriate administrations can adapt easily to the different situations, always following a logical policy and always aimed at cost reduction.

The main difficulty is the identification of the sources of supply of essential products, because they do not yield archaeological proof, especially in the case of cereals. In this area we must deal with indirect testimony, such as the origin of manufactured products that might have accompanied other goods on the routes towards the consuming military centres.

Generally, the goods produced directly by the State were supplied straight to the army from the production centres.

172

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

3.6 HISPANIA: A PARTICULAR SUPPLY NETWORK1 C. Carreras During the Principate, once the NW Hispania was already conquered by the Roman legions of Augustus, the whole region changed. Gold mines were started being exploited by the Roman State under the supervision of procuratores metallorum (Santos Yaguas, 2004), a network of Roman roads was built and a series of colonies were founded in the following years. However, the Roman legions were still in the NW, perhaps in less numbers but with a mission of keeping an eye on Astures and Cantrabian tribes and mainly, controlling the gold mines and its metal transportation. 1

other lands. Therefore, some territories have tended to achieve self-production of all staples required. Likewise, the Roman army settled in the NW Peninsula had to organize itself to obtain all the required staples from the nearby military camps. The best known example is the case of legio III Macedonica established from the Augustan period (AD 39) in the town of Herrera de Pisuerga, which had its own figlinarius, L. Terentius, who was responsible for coarse ware and samian ware production for his own unit. On the other hand, 15 stone inscriptions have been recorded in the nearby area that seem to limit the legion prata with regards to the urban territories of Segisama and Iuliobriga. Such territories may have been use to feed all the animals kept by the army.

Military headquarters were more stable from the JulioClaudian period onwards, and from Flavian times only one legion (VII Gemina) remained in Hispania together with some auxiliary troops. Despite not being a frontier territory, the Roman army in the NW received its corresponding supply of staples as any other province.

Sometimes local productions were not enough to fulfill all the unit requirements, so they required supply from other regions or provinces. According to the general model (see chapter 3.1), procuratores were mainly responsible for supplying the army in each province. The first procuratores of Hispania is from Augustus’ time in which he appointed a magistrate for each of the new provinces: Tarraconense, Lusitania and Baetica2. Procuratores normally lived in the capital of the province, as monumental epigraphy records also in Hispania.

Regarding other provinces such as Germania or Britannia, the Hispania province shows a completely different supply organization due to the geographical conditions of these territories. It has been pointed out that the Peninsula relief made communications difficult between coastal and inland regions from antiquity. Even in the XVIII-XIXth century there was no thriving trade between complementary economies of coastal regions in the Mediterranean or Atlantic and Castile (Ringrose, 1972). Until the XIXth century with the development of the railway, there was a lack of communication between the periphery and the central Peninsula, except some waterways along the Guadalquivir, Ebro, Guadiana, Tajo or Duero valleys. Water transport allowed cheaper costs and less time but it was only suitable for some privileged regions. Difficult transport conditions may have clearly affected the types of agricultural production of each territory depending on better or worse communication facilities to

However, there was a specific procurator for the gold mines of the NW, procurator metallorum, who used to stay at Asturica Augusta (Astorga). One of this first procurator metallorum was Pliny the Elder after AD 73. The distance from the capital Tarraco, made it necessary to have a specific procurator close to the mines. He was responsible for the administration of the whole mining district, which refers to the territory of Asturica et Callaecia, and probably from either mining and military interests.

1

2

This paper was published in 1997 in Gerión, 151, pp.151-176. The present article is an updated version of the paper published in 1997.

See Pflaum (1960-1) or Balil (1965) for a relation of all procuratores of each province.

173

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 3.6.1 Distribution of beneficiarii et al.

Tarraco4, whereas frumentarii are documented at Emerita, Corduba and Astigi5. The presence of frumentarii in Emerita, Corduba and Astigi may suggest that these regions were corn granaries for the NW army.

Therefore, from AD 79 onwards, there is the first procurator of the NW, per Asturica et Callaecia, who was responsible for the mines in the conventi bracarensis, lucensis and asturicensis (Lomas, 1975, 190). Such division in three common conventii suggests a kind of coherence in the regional administration with similar problems and complementary economies that had little contact with the Meseta and Mediterranean territories.

However, the most remarkable distribution comes from the beneficiarii since 35 inscriptions have been documented so far in Hispania6. A total of 14 of those inscriptions are recorded at Tarraco in the Tarraconense governor’s officium7. Only in three cases of those inscriptions did the inscriptions indicate the beneficiarii origins such as Toletum (CIL II.04164), Asturica (CIL II.04144) and Braccara (RIT 00905). It is interesting that the NW origin of two of them match the concentration of other beneficiarii inscriptions in this NW sector.

With regards to military supply, the presence of procuratores metallorum in Asturica may have facilitated its administration with the military outposts all over the NW region3. It is difficult to know the relation between those procuratores and the army settled there, but it was supposed to be very close since they were keeping routes where metal was sent towards the city of Rome.

4

T. Claudius Candidus (CIL II.04114) and L. Valerius Flavus (RIT 00043) are the only two stratores registered so far in a maritime city (e.g. Tarraco), probably farming taxes (i.e. portoria) from the local maritime traffic. 5 C. Titius Similus (CIL II.00484) as centurio frumentariorum and M. Valerius Secundus (AE 1905.00025) as frumentarius are recorded at Emerita, Q.Antonius Granius Erasinus (AE.2003) at Corduba and Magnius Donatus (CIL II.01167) at Astigi). 6 The number has increased quite a lot from the early study in 1997 (Carreras, 1997). 7 They have the title of beneficarii consularis as CIL II.04144; CIL II.04145; CIL II.04148; CIL II.04149; CIL II.04152; CIL II.04153; CIL II.04154; CIL II.04160; CIL II.04164; CIL II.04167; RIT 00189; RIT 00192; RIT 00194; RIT 00198; RIT 00905.

The distribution of inscriptions of beneficiarii, stratores, frumentarii and procuratores in Hispania may help us to understand supply infrastructures in the province (see figure 3.6.1). Procuratores inscriptions are concentrated in the capitals of provinces (Tarraco, Emerita and Corduba) and Asturica for the specific procurator metallorum (Carreras, 1997, 166, fig. 4). Likewise, the two inscriptions of stratores recorded so far appear at 3

For an overall analysis of the military distribution in the NW and its archeological evidence see Carreterro (1993) and Morillo (2006).

174

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

Additionally there are three inscriptions in the Lusitania governor’s officium8, which demonstrates that military supply in Hispania was closely related to both Roman provinces, Tarraconense and Lusitania. Most of the remaining inscriptions of beneficiarii appear in the NW where the armies were settled. For instance, there is an inscription of a beneficiarius procuratoris at Astorga (IRP Leon 00126) whose origin is braccarensis, another in Luyego (AE 1967.229) and five at Villalís of beneficiarii procuratorii (CIL II, 2552; CIL II, 2553; CIL II, 2555; CIL II, 2556; AE 1910, 1). Another inscription was found in the mining district of Castro Corporales (EAstorga 00148), referring also to a beneficiarius procuratoris, and another at Rua de Valderroes, near Aquae Flaviae (CIL II.2610) on the route towards the Atlantic ports. Such high concentrations of beneficiarii inscriptions in the NW region seem to be related to both transport of stables to the army and also the metals mined there. All those beneficiarii appear to be supervised by the procuratores metallorum of Astorga, since they are meant as beneficiarii procuratoris.

Atlantic ports, which apart from supplying the army requirements may have been excellent alternatives to land transport over the whole Tarraconense province. Perhaps there is an inscription from Villalís (CIL II, 02552) dated AD 163 that may show the relationship between soldiers involved in keeping the gold mines, evacuating gold, transporting supplies for the army and those responsible for controlling such processes, beneficarius procuratoris: I(ovi) O(ptimo) M(aximo) [p]ro salute M(arci) Aureli An [t]onini et L(uci) Aureli) Veri [A]ugustor(um) ob natale(m) Aqu[i] lae vexillatio leg(ionis) VII G(emina) F(elicis) sub cura Licini Patern[i] c(enturionis) leg(ionis) eiusd(em) et Hermetis Augustor(um) lib(erti) proc(uratoris) et Lu creti Paterni dec(urionis) coh(ortis) I Celt(iberorum) et Fabi Martian[i] b(ene)f(iciarii) proc(uratoris) Augustorum et Iuli Iuliani sign(iferi) leg(ionis) eiusd(em) IIIII Id(us) Iunias Laeliano et Pastore co(n)s(ulibus)

In the eastern NW region there is another inscription at Segisama-Amaya (Sesamón) (CIL II, 2915) near the camp of the Legio IIII Macedonica on the route towards the Upper Ebro valley.

CIL II, 02552

Finally there is a series of inscriptions in other places such as Estepa (CIL II, 4144), Cartago Nova (HEp. 11.00330), Saetabis (AE.1994.01068), Dénia (CIL.II.03587), Cambrils (AE.1955.246) or Figueras (CIL II, 04624). Figure 3.6.1 shows the distribution of all those inscriptions of procuratores, stratores, frumentariores and beneficiarii in Hispania with a high concentration in the NW Peninsula as well as the capitals of the provinces, Tarraconense and Lusitania.

With reference to the land route towards the Ebro valley, there are only a few military inscriptions along the two main roads in towns of the Cluniense conventus such as Clunia, Uxama, Vareia, Tritium or Calagurris, apart from the beneficiarius and miles recorded at Segisama. The inscription from Segisama (CIL II.2915) has special value since it records together with the beneficiarius, the presence of a stationis Segisamonensium, where portoria (taxes) of transport merchandises were gathered:

Regarding beneficiarii distribution, there are some blank areas between the capital of the provinces and the army settlements in the NW. Such lack of inscriptions may be due to difficulties in land transport that may have fostered another means of transport towards the military settlements of the NW region. The absence of other military inscriptions9 on the land routes towards Asturica Augusta from Tarraco and Emerita point to the high costs of land transport.

[3] A[1] [3 Genio(?)] sta [t]ionis Segi samonensium [1] Aelius Mari timus b(ene)f(iciarius) co(n)s(ularis) exedram cum basi d(e) s(uo) f(aciendum) c(uravit)

On the contrary, high concentrations of military inscriptions are documented along the routes that communicate Asturica with different ports of the Atlantic coast. Some of those inscriptions belong to vexillationes that kept control of the whole mine district (Jones, 1976, 60). It seems feasible to think that the NW gold may have been sent from time to time through the Atlantic ports such as Brigantium or Portus Cale. There was a complex infrastructure of Roman roads leading towards those

CIL II.03915 If the epigraphic documentation appears to be confusing and somehow particular to some extent, the archaeological evidence is even more complex. Of course, the NW sector records a good sample of imports from other parts of Hispania of Samian ware (TSH), fine wares, mortaria or lamps. Nevertheless, the real amount of such imports should be evaluated by quantifying them and comparing them to other regions. At first sight, it looks as if those imports were never in large amounts. Local productions of Samian wares and other fine wares demonstrate difficulties in military supply at least in the early stages.

8 The three inscriptions are CIL II.491; AE.1983, 0487 and AE.1976, 0275 (Schallmayer, 1990). 9 The Plata route (Emerita-Asturica) only documents an inscription at Salmantica and Capera, whereas there is no inscriptions from Tarraco to Calagurris in the route from the Ebro valley (see Roldan, 1974; Le Roux, 1982, fig. 3)

175

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Tabla 3.6.1 Densities of amphorae in different sites in Hispania Site

Dressel 20

%

Baetican Fish-sauce

%

1. Legio (León)

1460

10,2

4606

32,3

2. Petavonium

52

7,8

78

11,7

3. Campa Torres (Gijón)

6

5,0

4

3,3

4. Uxama

7

0,3

248

12,0

NW

5. Corona de Corporales

0

0

0

0

6. Corona del Cerco

0

0

0

0

7. Castrelín

0

0

0

0

299

7,1

1532

36,6

8. Asturica Augusta (Astorga) NE 9. Can Fatjó

63

0,1

32

2,6

10. Iluro (Mataró)

2175

2,2

5220

5,4

11. La Salut (Sabadell)

189

2,3

127

1,5

0

0

14

5,0

12. Setmenat 13. El Bosquet

0

0

3378

8,2

106

0,1

148

0,2

15. Tona

0

0

4892

14,3

16. Sant Pau del Camp

0

0

0

0

17. L’Alzinar gran

0

0

0

0

14. Can Feu

18. Vinya del Ticó

0

0

0

0

19. Iesso (Guissona)

280

0,7

2465

6,7

20. Ausa (Vic)

192

2,7

1368

19,9

21. Can Ventura de l’Oller

48

0,2

59

0,2

22. Vil·la de l’Estadi

0

0

170

7,1

23. Aguacuit

1651

19,0

3336

38,4

24. Can Jofresa

1709

13,2

2609

20,2

Table 3.6.1 illustrates that point and shows amphora densities and percentages in different sites in Hispania. It also reveals that other Baetican amphorae such as fishsauce vessels were arriving in the NW in higher percentages, so transport conditions did not hinder the import of Dressel 20s. Therefore, it must be concluded that no redistributive system involving Dressel 20s was in place in the NW of the Iberian Peninsula.

PUBLIC DISTRIBUTION NETWORKS: THE LACK OF BAETICAN DRESSEL 20’S AMPHORAE Some staples coming from other provinces were transported by a combination of public and private networks. The model explained before argued the direct intervention of procuratores and beneficiarii in mobilizing staples such as grain or olive-oil through a complex network. In the case of olive-oil, amphorae assemblages of Baetican Dressel 20 amphorae recorded in the military sites of Britannia and Germania10, were a distinctive sign.

This statement can be also confirmed by the following interpolation map that includes Dressel 20 amphora densities from 104 British sites, 24 Spanish sites and other from Gallia and Germania. Amongst the 24 sites from Hispania, some of the key northwestern settlements such as Asturica, Legio, Petavonium, Campa Torres and Uxama are included11.

Densities of Dressel 20 amphorae in the NW of the Iberian Peninsula are quite low compared to other places in Hispania (Mediterranean coast) or other western provinces such as Gaul, Britain or Germania.

The difference in Dressel 20 amphora densities appear to demonstrate that the importance of State intervention in

10 In both provinces, the Dressel 20 amphorae stand for more than 50% of any local amphora assemblages, so their volume is incredible high compared to closer productions, which seems to demonstrate a State intervention in their distribution (Remesal, 1986; Carreras, 1998).

11

Amphora densities from Britannia come from Carreras and Funari (1998) whereas densities from Hispania come from table 1, together with percentages and Baetican fish-sauce amphorae.

176

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

Figure 3.6.2 Map of densities of Dressel 20

compared to other vessels (60-90%) and the number of Dressel 20 stamps (around 1800) (Carreras and Funari, 1998) appear to confirm the existence of redistributive system of the Baetican olive-oil in this province. However, the NW volumes and percentages (less than 10%) as well as the number of stamps (only one in Asturica Augusta) challenge the idea of a redistributive system in place in the NW12. The unique alternative is a local supply of olive-oil difficult to document. In fact, Dressel 20 amphorae are scarce in the whole NW Atlantic and Cantabric coast (Naveiro, 1991; Urteaga, 2005), with some presence in Lusitanian sites, including some amphora stamps (Fabiao, 1994).

distant places such as Britannia or Germania, which justifies the large volume of Baetican olive-oil amphorae. However, there are other regions with low densities of Dressel 20s that may have produced their own local olive-oil (i.e. Tarraconense coast) or may have imported other olive-oil from nearby regions. This seems to be the case in the NW of the Iberian Peninsula where a good number of lamps are documented, while the olive-oil amphorae are almost non-existent (Morillo, 2003; Carreras and Berni, 2003). As already said, NW troops could receive local oils from two close production areas such as Navarra-Logroño or Extremadura-Salamanca. Nevertheless, no signs of a possible container for carrying this liquid have been documented so far in the NW, so the answer remains open.

It is evident that military units stationed in the NW not always remained close to the camps, but they used to perform different missions in the rest of the Peninsula (i.e. police, vigilance), so supply could come from other parts of the Iberian Peninsula.

In order to compare the supply requirements of two provinces such as Britannia and Hispania, it must be pointed out that the Spanish army from the Flavian period onwards included roughly 10,000 soldiers whereas the British army was 4 times larger, 40,000 soldiers. Taking into account that each soldier received 48 librae of oliveoil a year according to the P. Beatty Panop. (2, 245-9), this meant that the British army required 9,142 Dressel 20 amphorae a year, while the Spanish army only needed 2,151 (Remesal, 1977-8; 1986, 76-77). In the case of Britannia, the volume of Dressel 20 and its percentage

Notwithstanding the negative evidence of olive-oil amphorae, the consumption of olive-oil in the NW is well known because of the abundance of lamps (i.e. lucernae) (Morillo, 1992; 2003) as well as its local imitation in 12 The case of Hispania is an exception to other provinces in the western Roman Empire in which there is an important presence of Dressel 20 amphorae. In eastern provinces such as Pannonia, the State supplied the armies with Adriatic-Istrian olive-oils (Brindisi and Dressel 6B amphorae).

177

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Asturica, Braccara and Tritium (Amaré and García Marcos, 1995). Since olive-trees could not endure the climate conditions of the NW of Iberian Peninsula, the olive-oil consumed in these lucernae would have come other latitudes. Apart from the Dressel 20 no other oliveoil amphorae appear in the amphorae assemblages, so olive-oil would have travelled to the NW regions in perishable containers such as skins (cullae) or barrels (tonnae).

villae are Moncalvillo, Medrano, La Morlaca, Camino de Pago, Turrios, Murillo de río Leza and Eros de San Martín; whereas Navarra documents the villae of Liédena and Cerrao (Fernández Castro, 1983; González and Hernández, 1983). Of course, there are some doubts with some of them about whether they were olive-oil or wine presses, but they demonstrate that there was a minimal olive-oil production. Such number of supposed olive-oil presses can be only compared to the ones recorded in Baetica or the Catalan coast in the Tarraconense (Hanel, 1989).

The iconography of the Castro Urdiales patera, with a wagon carrying a barrel to which amphorae contents are transferred demonstrates that it was a well known transport container in the region.

None of these rural sites in the Upper Ebro records any amphora or ceramic workshop, so there is no evidence of how the local production of olive-oil was traded. Perhaps skins or barrels were the alternative containers for such production, but there are no archaeological remains. The olive-oil production in the Ebro valley is not an exception since other Northern regions such as Tarraconense or Narbonense also record a thriving production of which hardly any containers are known14. With regards to the relation between the Upper Ebro and the military settlements in the NW, the presence of so many Roman villae producing olive-oil suggest a possible provincial supply of olive-oil for the army. Therefore, the overall redistributive model for supplying olive-oil may have had a local supplier in the case of the Iberian Peninsula. There are military inscriptions in the Upper Ebro valley, such as the ones documented in Calagurris, Vareia and Tritium, where soldiers may have travelled to guarantee army supplies; there is also the statio of Segisamo where the beneficiarius Aelius Maritimus (CIL II.02915) dedicated an inscription.

Figure 3.6.3 Castro Urdiales patera

On the other hand, the supply from the south through the Plata route is another possible alternative for this provincial supply of olive-oil. Production of olive-oil to the north of the Tagus river is recorded from mid II century BC (APP., Iber.64.146). With the foundation of Emerita, Augusta (circa 25 BC) started villae exploitation in the Extremadura countryside (Cerillo, 1984). Although only a few villae have been excavated so far, two of them have recorded olive-oil presses, Cocosa (Badajoz) and Dehesa de Torre-Aguila (Montijo), revealing intensive olive-oil production.

There are two potential close areas to the NW for oliveoil production. The Upper Ebro valley was one of the nearest places to the military camps. Olive-oil production in Navarra and La Rioja continues to the day, though other crops are replacing the olive-oil groves (Solovera, 1987, 15). In fact, the whole Ebro valley used to be covered by olive groves, as mentioned by Avienus (AVIE. O.M. 5.505), who describes the river as oleum flumen (the olive-oil river). Olive-oil production in Aragon is documented by Prudentius (Passio. XVIII. marit), who describes Caesaraugusta surrounded by meanders covered with olive groves. These same trees were used by Saint Braulius (Ep. X) when he supplied the presbiter Iactantius. This literary evidence, as well as the lack of olive-oil amphorae in the Ebro valley (Beltrán, 1983), confirms the existence of a local olive-oil production13.

Again, this olive-oil production has not been related to any kind of container, so it is difficult to prove any consumption in the NW military markets. Even closer to the NW military sites, there was olive-oil production in Sierra de Gata (Salamanca), which was 14

The olive-oil production in Narbonense is discussed by Bellet (1976), Brun (1986) and Garcia (1992). With regards to the olive-oil production in the Tarraconense the number of possible sites is increasing every day. So far 16 sites are documented in Catalonia: Santa Maria de Sales, Vilauba, Sentromà, Hort d’en Bareres, Partida de Fosses, Can Sans, Bovalà, Cantaperdius, Castellot, Torà, Corbins, Els Ametllers, Tolegassos, Puig Rodon, Santa Cristina and Turó de Remolencs (Mari et al., 1982; Mari and Mascort, 1988; Hanel, 1989; Casas et al., 1994).

At the archeological level, the region of Navarra and La Rioja concentrates an important number of Roman villae, with supposed olive-oil presses. In La Rioja, the recorded 13

On the contrary, the Baetican fish-sauce amphorae are widely distributed in the Ebro valley as well the NW.

178

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

Figure 3.6.4 Mapa Hispania amb els llocs de producció d’oli

challenge the overall model of public redistributions to the army proposed by Remesal (1986). In the case of the Iberian Peninsula, it seems that olive-oil was provided by provincial sources, from either Tarraconense or Lusitania. The State was still directly responsible for gathering taxes in kind, or purchasing oil and transferring it to the final military destination. Unfortunately, such a possible means of olive-oil supply has left no archaeological trace, as they appear to have used an organic container to transport it. The advantages of consuming olive-oil from Tarraconense or Lusitania were obvious: cutting transport costs, reducing complexity, and limiting dependence on distanct provinces.

traded in northern markets, such as León and Zamora, in the XVIIIth century (Ringrose, 1972, 112). However, it is unknown whether such production began in Roman times. Nowadays, there are two varieties of olives produced exclusively in Extremadura and Salamanca: Manzanilla from Cáceres and Verdial from Badajoz. Finally, Beltran (2008, 300, fig. 26) records a couple of Roman villae in the Duero valley that are supposed to have produced olive-oil during the Principate and Late Roman periods. One is known as villa de Fontao (Lavra) (Gorges, 1989; Brun, 1997, 51) and the other is Fonte do Milho (Porto) (Russell, 1951). This second source for the olive-oil consumed at the military sites of the NW is an attractive hypothesis, but more data is still required to confirm it.

To some extent, both epigraphic and archaeological testimonies identified by olive-oil distributions, confirm that the supply of the Hispanic army was regional. In other words it depended on staples from the NW and closer regions such as Upper Ebro valley and Lusitania. Moreover, external contacts were normally by sea, as distributions of military epigraphy show; these are mainly beneficiarii, located on the Roman roads towards the Atlantic ports. Therefore, Hispania provided a unique supply system, quite different from most western provinces of the Roman Empire.

On the contrary, Baetican olive-oil traded to the NW in other containers apart the Dressel 20 amphora lacks support, due to the fact that other amphorae (i.e. fishsauce amphorae, Haltern 70) from the same province are recorded in good numbers. Summing up, the lack of olive-oil Dressel 20 amphorae at the military sites of the NW Iberian Peninsula does not 179

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

theless, supply organization outside the Empire was substantially modified according to the requirements of military campaigns (expeditio).

PRAEFECTUS VEHICULORUM: UNRECORDED FIGURE IN HISPANIA When Roman troops were on campaign they carried all the equipment they needed to fight and survive in the military camps for at least one month15. This means that either soldiers or animals had to transport all the baggage (impedimenta) through hostile territory; the weigtht had not to limit their fighting capability and speed. Thereby the sources reveal that in 49 BC, Pompey’s troops at Ilerda disposed of 22 days’ rations (CAES. Civ 1.78), whereas Cicero asserts that soldiers carried rations for only half a month (CIC. Tusc. 2.37). Severus Alexander’s troops carried rations for 17 days (HIST.AUG. Alex. Sev. 47.1) and Julian’s for 17 or 20 days (AMM. 17.8.2; 17.9.2).

Those campaigns were not always conducted close to the coast or riverbanks. When troops had to rely exclusively on land transport, then problems started to arise. As example can be recalled in the campaigns of Corbulos in Armenia (AD 57-58), who had Trapezus as his headquarters on the coast of the Black Sea. The whole campaign of T. Quinctius Flaminius (BC 198) against Philipus depended on the availability of troop supplies, since the Roman general never dared go inland fearing a lack of staples. He only relied on supply near the coast that was guaranteed (PLU. Flam. 4.1; LIV. 32.27.2).

As can be observed, ancient sources record that Roman soldiers transported rations for 15 to 20 days, which gave them a limited autonomy. After using half the rations, soldiers had to return to their military camp, be supplied on the field, or feed themselves from the territory16. Supposing that they consumed 15 days’ rations, and had to return to their military camp, they should consume half rations each way (i.e. circa 8 days). If a legionary can cover 40 km in one day’s march, it means that Roman units could fight as far as 320 km from their base point.

As can be observed above, mobility inland in the Roman army depended to a great extent on the control of supply as well as availability of safe land transport. If the military unit wanted to become independent then the soldiers’ carrying capacity conditioned it. According to Vegetius (VEG. Epit. 1.19) each legionary could carry 60 librae (20 kg) without weapons, apart from pack animals and wagons that were included in the unit17.

Of course, when legions were marching in distant territories they depended on parallel supply (by river, maritime or by land) in order to avoid returning to their headquarters. These were the logistics employed on the German campaigns by Tiberius (5 BC) when the Roman fleet supplied staples along the river Elbe and the coast of the North Sea (VEL.PAT. 2.106.3). Likewise, Trajan was supplied by a Danubian fleet in the Dadic war (AD 101105), as illustrated on his famous arch in Rome (Richmond, 1935, 1-40). A detailed account of expeditiones with logistical support by sea is documented by Tacitus (TAC. Agricola) in his biography of his father-in-law, Cneo Iulius Agricola. Agricola was governor of Britannia (AD 77-84) and fought a series of campaigns in the north of the British Isles.

When a soldier started a military expedition he carried his own weapons (between 18-22 kg), clothes and underwear18, cooking equipment (vas) as well as the food rations (cibus castrensis). If one takes into account the 20 kg of carrying capacity of a soldier recorded by Vegetius (Epit. 1.19), plus weapons of around 18-22 kg, a legionary transported around 40 kg, without doubt a very heavy weight19. It appears that C. Marius was responsible at the end of IInd century AD for a military reform that reduced the common baggage for troops. As an alternative every soldier should carry more weight. That is why people joining the army were known as “Marius’s mules” (PLUT. Mario, 13.1) (Fuentes, 1991). Frontinus’ says (FRONTIN. Strat. 4.1.7): “For the purpose of limiting the number of pack animals, by which the march of the army was especially hampered, Gaius Marius had his soldiers fasten their utensils and food up in bundles and hang these on forked poles, to make the burden easy and to facilitate rest; whence the expression ‘Marius’s mules’.”

A SOLDIER’S EQUIPMENT (SARCINA)

During the Principate, as opposed to the Republican period, supply of staples to headquarters or military depots (i.e. granaries) was completely organized, with its corresponding administrative spheres (praefectus annonae, procu-ratortes, beneficiarii), to reach the borders (limes), where troops were garrisoned. Never-

Of the total weight, 14-18 kg corresponded to cibus, according to the number of days’ rations. The previous weight provided a range of 15 days. Any increase in the

15 Only T. Livius reveals that legionaries carried food, frumentum, for 30 days as is the case of Gaius Cassius in his campaign from Ilirium to Macedonia in 171 BC (LIV. 43.1.8), or Quintus Marcius Philippus in the Third Macedonian War (172-167 BC) (LIV, 44.2.4). According to Frontinus (FRONTIN. Strat. 4.1.1), Scipio Aemilianus’ soldiers used to carry rations for “many days” at Numantia (134-133 BC) (LIV. Per. 57). 16 Despite Cato’s famous sentence “war feeds itself” (LIV. 34.9.12), or the idea of Sun-Tzu (IV century BC) that “a good general manages to feed his troops thanks to the enemy” (SUN-TZU 2.15), relying on enemy resources increases risks. Thereby, it is the less recommended solution, except if there is no other.

17 The decision was not easy. The more pack animals a unit required the more independent it could be, but it need more fodder to feed them and produced a longer convoy that was more difficult to defend (Erdkamp, 1998, 23). 18 There was no standard uniform, but equipment was quite similar for everyone with a tunic and cloak, weighing around 3 kg (Junkelman, 1997, 94-102). 19 Junkelmann (1986, 200) tested a group of people carrying Roman equipment that weighed between 43-46 kg; he took them trekking in the Alps at a daily average of 25 km.

180

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

number of rations would have represented an increase in weight for the soldiers – 1.2 – 1.3 kg per day according to Stolle (1914) and Roth (1991).

(Kolb, 2000). Agrippa, lieutenant of Augustus, established a system of couriers who undertook relays (iuvenes) covering a certain distance between stationes located on military roads, (Eck, 1999, 94; Herzig, 1999). Sometimes, apart from riders, wagons transported the mail of the governors and other civil servants, as Suetonius reports (SUET. Aug. 49, 3-50): “To enable what was going on in each of the provinces to be reported and known more speedily and promptly, he at first stationed young men at short intervals along the military roads, and afterwards post-chaises. The latter has seemed the more convenient arrangement, since the same men who bring the dispatches from any place can, if occasion demands, be questioned as well”.

The alternative of transporting baggage was to employ a pack animals20 and wagons, which meant a major effort to obtain fodder21. This alternative also limited the mobility of the unit and generated a longer convoy and a weak point for enemies (IOS. Bell.Iud. 2.63) (Menéndez, 2000, 174-175). Military columns then needed to employ an escort to repel possible attacks, as was the case with Gelbuda’s convoy going to Neuss during the revolt of Civilis (TAC. Hist. 4.35). The need to obtain fodder also involved sending a detachment for this task and thereby dispersing troops in enemy territory. Roman fodder parties were constantly at risk, e.g. Perseus’ war in 171 BC22.

Initially, no magistrate was directly responsible for the post system. However, as soon as the office of praefectura vehiculorum appeared23, such function was attached to it. Of course, at the end of IInd century AD when the praefectus vehiculorum became a sexagenarian magistrate, he seems the only one responsible for post and official transport on particular roads (i.e. per Flaminiam – see appendix). However, it appears that this magistrature had more duties at an earlier period.

The logic behind Marius’ reform was based on the more weight a soldier could carry the more autonomy and mobility for the unit. The unit required less extra supplies for its pack animals. Within a soldier’s sarcina, the variable weight was basic food (cibus castrensis). In general, legionaries carried cooked food that required minimum preparation. Often they had no fire in order so as to avoid being discovered and therefore food for legions inside camp differed rather from campaign rations.

At least twice the office was combined with the cura copiarum exercitus, i.e. army supply on the first German campaign (L. Aurelius Nicomedes, CIL VI. 1598: AD 168-175) and the second (M. Aurelius Papirius Dionysios, CIL X.6662: AD 178-180) in the time of Marcus Aurelius. Sometimes this post coincided with administrative duties in ports, conductoris publici portorii (C. Antonius Rufus – CIL III.13283 – and T. Iulius Saturninus – AE 1940, 0101: AD 138-161) in the time of Antoninus Pius. Furthermore, some careers (cursus honorum) of praefecti vehiculorum suggest that they had other offices before, related to maritime transport (praefectus classis), and even later to the praefectura annona.

LAND SUPPLY ON CAMPAIGN: THE PRAEFECTUS VEHICULORUM When Augustus ruled the Roman Empire it was at the end of a period of continuous expansion. Roman troops were well settled on the limes that divided Roman civilization from other communities. The old hibernia, or winter camps, became permanent camps for the legions and auxiliary troops. They were responsible from now on for keeping the frontiers. Of course, minor conflicts were dealt with by responsible of troops and corresponding legati legiones, although any major conflict required the Emperor’s consultation.

Eck (1999, 100) already hinted at the function of vehiculatio as including activities related to the supply of food and clothes to the troops. What were such requirements if the supply structure relied on praefectus annonae, procuratores and beneficiarii? (Carreras, 1997). The answer is simple: his function must have been to obtain the transport needed to mobilize legionaries and their supply to unfamiliar places. Therefore they supplied the army in places where there was a lack of infrastructure. Thereby, every new campaign on the frontier brought about an increase in military resources and supply (food, equipment). If the conflict area was close to a river or coast, movements of troops and equipment did not require a substantial infrastructure. On

Such was the flow of information required between the city of Rome and her frontiers, the Emperor set up a communications network known later as cursus publicum 20

Roth (Leiden, 1999, 77-83) suggests that each contubernium (8 soldiers) had 2 mules assigned in order to carry the common equipment such as hand-mill, tent, tools and baskets (circa 145 kg). Perhaps, some of these tools such as mola were not required on short campaigns, on which they would consume biscuits (bucellatum) rather than mill grain. 21 In the Partic campaign, Mark Anthony (36 BC) had to get rid of some of his dying mules and thereby some of his army’s equipment (PLUT. Antonio 45,4). 22 Erdkamp (1998, 141-155) describes numerous examples of conflicts in Republican period because of difficulties in obtaining fodder. Probably, J. Cesar was one of the generals who took more risks on his campaigns since he preferred speed rather than stores of fodder and minimal supplies (CAES. Gall 5.17.2-5; 7.16.3; 20.9; AUL. Hist.BGall 8.10.3; 16.4).

23

Initially, it was thought that praefectura started in Hadrian’s reign based on the interpretation of texts from Historia Augusta (HIST.AUG. Adr. 7.5) and Aurelius Victor (AUR.VICT. Caes. 13,5). However, Eck (1975; 1999, 95-99) has demonstrated that this was introduced earlier, at least in the Claudian-Nero period (see Inscriptions 1 and 2 in the appendix).

181

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

the other hand, when the theatre of operations was inland the logistics of land transport make movements quite complicated.

Domitian period according to the inscription of Auximum (CIL IX.5835) (Bérard, 1984): ..., priusquam te Germanicus arbitrum sequenti annonae dedit ommiumque late praefecit stationibus viarum,…

The first evidence in difficulties regarding logistics by land transport appears in the edict of Sextus Sotidius Strabo Libusciduanus, legatus pro praetore. In the time of Tiberius he was obliged to exact in Sagalassos (Pisidia) any mean of transport (mules, wagons) to move troops and supplies by paying24. It must be born in mind that Sagalassos is located inland, in a very mountainous area (heights around 1000 m). This region was at that time a border area in the Roman Empire. The province of Galatia was conquered in AD 25, but Capadocia and Comagene were annexed in AD 17. In this sense, the edict of Sextus Sotidius should be understood as a requirement for the means of transport to mobilize goods from the Lician coast (p.e. Attaleia, Side) to places inland were the troops were settled.

Stationes viarum are interpreted here as relay places of vehiculatio, which were required for land transport. Therefore supplies could reach theatres of operations and follow troops on campaign. In times of peace, supplies could be organized through the structure of praefectus annonae, procuratores provinciales and beneficiarii; in war the post of praefectus vehiculorum was especially important in terms of military supply. Going back to the origin of this office, according to Eck (1975; 1999) the first account was an anonymous inscription from the Claudian era (AE 1950, 0170; see inscription 1), located in Epirus. At that time, the only military campaigns were the ones related to the conquest of Britannia, some conflicts by Corbulus against Caucos and Frisons (AD 48-49), and also the peaceful annexation of Thracia. Of course, it is well known that Claudius had a keen interest in improving communications in Germania with the construction of roads and a channel between the river Rhin and Mousel. Lack of documentation makes it difficult to relate the beginning of this office to a particular moment. It seems that supply was a general worry for emperors and was solved with the creation of this new post, the praefectus vehiculorum.

In this particular case, all the effort of providing the means of transport was put on local communities, such as the one from Sagalassos. Although it is not said how long it was in place, this kind of measure imposed a terrible burden for local communities, which probably required wagons and animals for other daily activities. This kind of edict may have been exceptional, but provincial administrations were perfectly aware that it was not a resource to be abused. Sometimes local communities did not have the means of transport for the army’s requirements and “confiscation” would generate extreme economic pressure on them. Within this context of land transport requirements for military logistics, the praefectura vehiculorum appears to have an important role. It seems that it appeared during the rules of Claudius and Nero (Eck, 1999, 197), or that of Vespasian (Eck, 1975). It was an office that managed postal services, including transport and infrastructures, but, above all it was aimed at facilitating army supplies in any situation. That is the interpretation of the passage by Statius (STAT. Silvae 5.9) that refers to Plotius Grypus (AD 92-93), who was praefectus vehiculorum in the

During Nero’s time, another praefectus vehiculorum is recorded thanks to an inscription from Messina (CIL X.6976; see inscription 2), L. Baebius Iucundis. The career of this eques appears to be concentrated in the province of Egypt, where he enjoyed different military posts as well as the office of iuridicus. Although it is not known where this praefectus vehiculorum25 carried out his duties, it must be borne in mind that the Eastern provinces suffered continuous conflicts in Nero’s period, such as the Jewish war (AD 67-69). The Roman troops sent to crush Jewish rebellions were led by Vespasian and Titus, required special supplies, and perhaps the praefectus. L. Baebius played an active role. He was previously tribunus militum in the legio XXII Deiotarianae and settled in the military camp of Nicopolis in Alexandria from AD 25. Therefore, he knew well the supply logistics of the annona of Rome.

24

An extract of the inscription (AE 1976, 0653; 1978, 0789; 1989, 0727) reads: ...praestare autem debebunt vehicula usque Cormasa et Conamam neque tamen omnibu/s huius rei ius erit sed procuratori principis optimi filioque eius suu da[to us]que ad carra decem aut/pro singulis carris mulorum trium aut pro singulis mulis asinorum binorum quibus eodem te/mpore utentur soluturi pretium a me constitutum praeterea militantibus et iis qui diplomum hab/ebunt et iis qui ex alis provincis militantes commeabunt ita ut senatori populi Romani non plus quam/ decem carra aut pro singulis carri muli terni aut pro singulis mulis asini bini praestentur soluturis id quod/ praescripsi equiti Romano cuius officio princeps optimus utitur ter carra aut in singula terni mula aut/ in singulos [mu]los bini asini dari debebunt eadem condicione si quo amplius desiderabit conducet/ arbitrio locantis centurioni carrum aut tres muli aut asini si quo amplius desiderabit conducet/ arbitrio locantis centurioni carrum aut tres muli aut asini sexs eadem condicione iis qui frumen/tum aut aliudq(u)id tale vel quaestus sui cau{s}sa vel usus portant praestari nihil volo neque cuiquam p/ro suo aut suorum libertorum aut servorum iumentu mansionem omnibus qui erunt ex/ comitatu nostro et militantibus ex omnibus provincis et principis optimi libertis et servis et iumentis/ eorum gratuitam praestari oportet ita ut reliqua ab invitis gratuita non e(x)sigant (Mitchell, 1976, 106-131).

Of course each period in which there is a mention of a praefectus vehiculorum there were changes in the Roman 25

During the first two centuries, the praefectus vehiculorum was not assigned to any province, nor to a particular road as will happen in Severus’ time. Therefore it is believed that there was a central office in charge of the cursus publicum for the whole Empire. Nélis-Clement (Les beneficiarii: militaires et administrateurs au service de l’empire (Ier s. AC – VI s. PC), 2000, Burdeos, 176) suggests that the provincial administration of the cursus was in the hands of governors and other provincial posts; while praefectus vehiculorum was responsible for special requirements and general ones that affected the whole system.

182

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

From Marcus Aurelius’ time there are the first inscriptions to link the praefecti vehiculorum with military supply on campaign (cura copiarum)26. It is L. Aurelius Nicomedes (CIL VI.1598; see inscription 11) with the expeditio Germanica prima, and M. Aurelius Papirius Dionysios (CIL X.6662; see inscription 12) with the expeditio Germanica secunda. This is the first time that the relation between both functions, military supply and transportation, becomes explicit. L. Aurelius Nicomedes was a freeman close to the Imperial family; he had been already teacher of Lucius Verus, and therefore a trustee for a key post.

frontiers or military campaigns that generated new requirements in transport infrastructures and supply. So, the anonymous character of Apri in Vespasian’s reign (AE, 1973, 0485; see inscription 3) coincides with a turbulent period that started with the Civilis rebellion (AD 69) and a resumption of the Jewish wars (AD 69; Masada AD 73). In addition there were the Neckar occupation (AD 73-74), the new organization of Asia Minor after the annexation of Comagene (AD 74-77), the various campaigns in Britannia (Agricola AD 77-82), Germania (Veleda AD 77-78), and activity along the Danube (AD 78). The detailed study of this character appears in the work of Eck (1975). All those movements and shifts in frontiers required important logistical changes, in which the office of praefectus vehiculorum played a singular role, together with coordination with other praefectures such the annona.

In the distribution of offices in the Marcoman war, expeditio Germanica secunda, with an inscription of praefectus vehiculorum, M. Aurelius Papirius Dionysios, he was assigned a specific responsibility on the via Flaminia, in other words the road that goes from Rome to the north of Italy, probably the centre of operations. Since Rome was the convergence point for products obtained from the annona (wheat, olive-oil...), this was the possible origin of some supplies destined for troops on the German frontier.

Unfortunately, our textual sources do not provide many details on this post, and epigraphy only includes the cursus honorum of each particular person. Such cursus may infer a career of a typical profile of an equites following a military career related to transport and supply. These are the profiles of an outstanding figure such as Q. Plotius in Domitian’s time (CIL IX.5835; see inscription 4a and 4b), studied by Bérard (1984), who appears in a complex period involved in the Dacia (AD 85-86 and 88-89 AD) and Germania (organization of Decumates fields, and wars against Yapigians, Cuadii and Marcomans – AD 90-92). Likewise, two inscriptions are documented in the same period in Hama (Syria), in which there is a requisition of transport, similar to Sagalassos, by the procurator Claudius Athenodorus (Mouterde and Mondesert, 1957).

Close to him, there was another army supply official: praepositus copiarum expeditionis Germanicae secundae, in charge of T. Claudius Candidus (CIL II.4114)27, helped by an imperial slave, a dispensator rationum copiarum (CIL V.2155). To some extent, these three coetaneous characters reveal different administrative levels in the army supply chain. The praepositus lives near the operations’ hub and knows the logistic requirements; he reports to a dispensator to order supplies by requisition, or purchase, from places nearby. The praefectus vehiculorum manages supplies from distant locations, as well as their distribution to the different units on the front lines.

From Trajan’s period, there is a Q. Marcius Fronto (AE 1955.225; see inscription 5), who may have taken part in the logistics of the numerous campaigns of this emperor, such as in Germania (AD 96-97), the Dacian wars of AD 101-102 and AD 105-107, or the conquest of Arabia (AD 105-106). On the contrary, the following anonymous character (CIL XIII.01857; see inscription 6) coincides with the last years of Trajan and the peaceful reign of Hadrian, in which there was only the Parthian campaign (AD 114) and the conquest of Asiria and Mesopotamia (AD 116), and also the reorganization of frontiers (Britannia, Germania and Raetia).

The last praefectus documented in the Antonine dynasty was M. Aurelius Mindius (AE 1928, 97; see inscription 13), who took this post in Commodus’ reign. However, there was a change of duties during the time of Septimius Severus, which meant decentralization (Eck, 1999, 105) and the creation of posts of praefecti for different roads and provinces (see index: praefecti vehiculorum in provinces). Why did the new Emperor make these changes? Many of the Septimius Severus’ decisions appear to be related to difficulties suffered by him and his troops during the civil wars against Pescennius Niger and Clodius Albinus. During these campaigns, despite the fact that Septimius disposed of more loyal troops, he was

There is a series of inscriptions of 4 praefecti from Antonius Pius. The first is the one to L. Volusius Maecius (AE 1955, 0179; see inscription 7), who perhaps took part in the reorganization of the Dacian limes (AD 140). Afterwards there is a wide chronology that goes from AD 138 to 161, which records T. Appalius Alfidius (CIL IX.5357; see inscription 8), C. Antonius Rufus (CIL III.13283; see inscription 9) and T. Iulius Saturninus (AE 1940.0101 and AE.1934, 0107; see inscriptions 10a and 10b). These last-named may have taken part in other later military actions to AD 145, such as the Mauritan revolts (AD 149-152), the widening of the Neckar limes (AD 160), or the Armenian war (AD 161-163).

26 The expression cura copiarum appears for first time in Trajan’s reign. It involves the responsibility for obtaining staples for the army by purchasing or expropriating it under the direct orders of the Emperor. Remesal (1986, 98) includes a list of praefecti cohorti or tribunes fulfilling such task. 27 The figure of T. Claudius Candidus, one of the loyal generals of Severus during the Civil war against Pescennius Niger and Clodius Albinus, perhaps had a direct implication in changes produced in the praefectura vehiculorum from AD 193 when it becomes a sexagenarian magistrate and provincial responsibilities.

183

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

always in trouble supplying them as well as the city of Rome. After the death of Pertinax at the hands of the Praetorian Guard and the arrival of Didianus Julianus, Septimius was greeted as Emperor in AD 193 by his troops at Carnutum, basically the legions from Pannonia and Germania. From that moment on, Didianus Julianus sent messages from the Senate to the troops asking for their desertion from Septimius, and even demanding they kill him (HIST.AUG. Sev. 5.1-8). However Septimius escaped from Didianus’ envoys and he sent messages to the Praetorians to kill him, which at last they did. In these early skirmishes of the Civil War, the power of the cursus publicum was a key element in the final result and therefore the post of praefectus vehiculorum was without doubt particularly relevant.

After Septimius’ victory, he took terrible revenge on the supporters of his rival, including Italian, Gaulish and Spanish senators. He took their lives and confiscated their estates (HIST.AUG. Sev. 11.8-14.3). Some stamps of Dressel 20 amphorae have been interpreted as the result of such confiscations amongst the Baetican olive-oil producers (Manacorda, 1977; Remesal, 1986), as well as changes in the tituli picti  of the same amphora following the formula Dominorum Nostrorum (Rodríguez Almeida, 1989, 35-37), which suggests the control of the annona transport by the same emperor. Such measures reveal the difficulties that Septimius faced in the Civil Wars from staple producers (i.e. Baetican olive-oil), private transporters (i.e.  of the olive-oil amphorae) and civil servants.

But conflict did not end here. In AD 193, the governor of Syria, Pescennius Niger, was proclaimed Emperor in the east, with the support of 3 legions, 2 from Palaestina and the IInd Trajana from Egypt, together with 50,000 auxiliaries. Pescennius could have cut grain supplies to the city of Rome from the African provinces, whose granaries were quite low at that time (HIST.AUG. Sev. 8.2-4). Interestingly, Septimius had to send his troops by road in three columns led by his lieutenants, Lucius Marius Maximus Perpetuus Aurelianus to Thrace, Lucius Fabio Cilo to Perinthus, and Tiberius Claudius Candidus, with most troops, crossing the Danubian region. Perhaps the election of such land routes, longer and more difficult, was due to the lack of support by the fleet and private traders28. In the end Septimius managed to beat Pescennius Niger in 194 and become sole Emperor again until Clodius Albinus29 challenged his power. While Septimius was in the East, the governor of Britannia, the afore-mentioned Clodius Albinus, maintained correspondence with senators (HDN. 3.5.2) asking for support to access the Imperial throne. Again the official mail had an important role.

In this context, it seems easy to understand the reform of the post of praefectura vehiculorum. Much of the power of this office was removed when other praefecti posts were created for each road and province (Eck, 1999, 105). Although there are other late inscriptions in which no relationships to a particular road or province (see inscriptions 14 to 21) are mentioned, most documented inscriptions of late praefecti vehiculorum indicate a geographical location.

When Septimius heard of the aims of his rival, he obtained the support of the Senate against Clodius as the Senate and Roman plebs were tired of war (HIST.AUG. Sev. 10.2; D.C. 75.4.2-6). Clodius Albinus obtained only the support of the Gallic and Hispanic aristocracy, and beat Severus’ troops in the first battles in Gallia, until the battle of Tournus (AD 197). He was routed at this cruel battle in which more than 100,000 soldiers lost their lives (D.C. 75.6.1).

Since the Iberian Peninsula was a territory where land transport to and from military settlements was essential, the lack of inscriptions of praefecti vehiculorum may reveal alternative supplies. As was initially justified, it appears that the military in the NW depended basically on regional supply, so merchandise did not travel widely, only those products that needed to arrive by a combination of means from the Atlantic coasts (maritime, river and land transport).

FINAL COMMENTS This paper attempted to analyze the difficulties in supplying Roman armies. Within this context, equally common to any army in history, the office of the praefectus vehiculorum should be especially linked, even though his role is partially obscured from lack of information. However, it seems that this was a key post in the supply of troops on campaign. The partial evidence has been presented again here and the hope is that new documents will shed new light on this subject.

28

The sources do not give reasons. Septimius offered Clodius Albinus the opportunity to become Caesar if he agreed to give him support against Pescennius, and this did it (D.C. 73.15.1-2; HDN. 2.15.3-4). 29

184

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

PRAEFECTUS VEHICULORUM 1) …]co Po[3] (v. 41 dC – Eck, 1999) – AE 1950, 0170 (Pflaum, 1961, 1029) – Buthrotum (Epirus, Albania) …]co Po[3]/ [3I]Ivir(o) II quinq(ennali) [3]/ [3 l]eg(ionis) V Macedonic(ae) [3]/ [aug]uri patrono col(oniae) [3]/ praef(ecto) v]ehic(u)lor(um) et [3]/ [praef(ecto)] castr(orum) [3]/ [3] Orestarum [3]/ [3] suo [ 2) L. Baebius Iucundis (54-68 dC – Eck, 1999) – CIL X.6976 – Messina L(ucius) Baebius L(ucius) f(ilius)/ Gal(eria) Iuncinus/ praef(ectus) fabr(um) praef(ectus)/ coh(ortis) IIII Raetorum/ trib(unus) milit(um) leg(ionis) XXII/ Deiotarianae/ praef(ectus) alae Astyrum/ praef(ectus) vehiculorum/ iuridicus Aegypti 3) ... (69-79 dC – Eck, 1975) – AE 1973, 0485 – Apri (Thracia, Turquia) trib(uno) cohortis M(milliariae)/ equ[itatae Chalci]/ denoru[m praef(ecto) alae]/ I Panno[n]io[rum praef(ecto)]/ vehiculorum [Imp(eratoris) Cae]/saris Vespasian[I Aug(usti)] e[t] aedili castren[si]/[p]roc(uratori) XX hereditatium/ [p]roc(uratori) provinc(iae) Lusita/[n]iae proc(uratori) provinciae/ [T]hraciae censori e[i]us/ [de]m provinciae patrono/ [colo]niae pub(lice) d(ecreto) d(ecurionum) 4)a Q. Plotius Maximus Trebellius Pelidianus (v. 92 dC, Bérard, 1984) – CIL IX.5835 – Auximum (Lacio) Q(uinto) Plotio Maximo/Collin( )/Trebellio Pelidiano/eq(uite) pub(blico)/ trib(uno) leg(ionis) II Traian(a) Fort(is)/trib(uno) coh(ortis) XXXII Volunt( )/trib(uno) leg(ionis) VI victric/ proc(uratori) aug(usti)/ pro magis( ) XX hered( )/ praef(ecti) vehiculor(um)/ qq( ) patr(oni) col( ) et suo pont( )/ coll( ) fabr( )auxium ob/ eximium inter municip( )/ suos amorem/ l( ) d( ) d( ) d( ) 4)b Idem – Stacio, Silv. 5.9.10-23 ..., priusquam te Germanicus arbitrum sequenti annonae dedit ommiumque late praefecit stationibus viarum,… 5) Q. Marcius Fronto Turbo Publicius Severus (107-108 dC) – AE 1955, 225 (Pflaum, 1961, 1029) – Cyrrhus (Siria) [Q(uinto) Marcio]/ C(ai) f(ilio) Tro(mentina) Fron/toni Turboni/ Publicio Severo/ domo Epidauro/ p(rimo) p(ilo) bis praef(ecto) vehic(ulorum) trib(uno)/ coh(ortis) VII vigil(um) trib(uno) eq(uitum) sin[g(ularium)]/Aug(usti) trib(uno) pr[ae]t(orianorum) proc(uratori)/ ludi magni praef(ecto)/ Qui[r(ina) Va]lens/ o[b m]eritis (!) 6) …]latin P[3] (110-130 dC) – CIL XII.01857 (Pflaum, 1961, 1029) – Vienne ...]latin P[3]/ [leg]ato [//leg(ato) Aug(usti) pro pr(aetore) [prov(inciae) Lugu]/unens(is) adlecto in[ter praetor(ios) ab]/ Imp(eratori) Caes(ari) T(ito) [Aelio Hadriano Antonino Aug(usto)]/ Pio [p(atri) p(atriae)3]/ P[//praef(ecto) class(is) Mi[senat(i) proc(uratori) provinc(iae) Lug]udu(ensis)/ [proc(curatori) provi]nc(ciae) [R]aet[iae et regni] Nor(ici) pro[c(uratori)]/ [provinc(iae)] Pon[ti subpraef(ecto) veh]iculor(um) [ 7) L. Volusius Maecianus (v. 145 dC) – CIL XIV.5347 (Pflaum, 1961, 1029) – Ostia L(ucio) V[olus]io L(uci) f(ilio)/ Ma[e]cian[o]/ co(n)s(uli) desig(nato) praef(ecto) aer(arii) Satur[n(i) pr(aefecto) Aeg(ypti)]/ pr(aefecto) ann(onae) pontif(ici) m(inori) a libell(is) et [cens(ibus) Imp(eratoris)]/ Antonini a studiis et proc(uratori) [biblioth(ecarum)]/ pr(aefecto) vehicul(orum) a libell(is) Antoni[ni Aug(usti) pr(aefecto)]/ coho(rtis) I Aeliae class(icae) pr(aefecto) fabr[um p(atrono) c(oloniae)/ Lucius V[olusi]us Mar[ 8) T. Appalius Alfidius Secundus (138-161 dC) – CIL IX.5357 (Pflaum, 1961, 1029) – Firmum (Picenum) T(ito) Appalio T(iti) f(ilio) Vel(ina)/ Alfino Secundo/ proc(curatori) August(i) XX hered(itatium)/ proc(uratori) Alpi(um) Atrectianarum/ pr(aefecto) vehicul(orum) sub praef(ecto)/ class(is) praet(oriae) Ravenn(atis) pr(aefecto)/ alae I Aug(ustae) Thrac(um) trib(uno) coh(ortis)/ I Aeliae Britton(um) praef(ecto)/ coh(ortis) IIII Gallor(um) patron(o)/ colon(iae) flamini divor(um)/ ommium augur(i)/ II vir(o) quinq(uennali) bis/ ob merita eius/ d(ecreto) d(ecurionum) 9) C. Antonius Rufus (138-161 dC) – CIL III.13283 Pflaum, 1961, 1029 – Senia (Dalmatia, Croacia) I(nvicto) M(ithrae)/ spelaeum cum/ omne impen/ sa Hermes C(ai)/ Antoni Rufi/ praef(ecti) veh(iculorum) et/ cond(uctoris) p(ublici) p(ortorii)/ ser(vus) vilic(us) Fortu/nat(ianus) fecit 10)a T. Iulius Saturninus (138-161 dC) – AE 1940, 0101 (Pflaum, 1961, 1029) – Senia (Dalmatia, Croacia) S(oli) I(nvicto)M(ithrae)/Faustus/T(iti) Iuli Saturni/ni praef(ecti) vehi/culor(um) et conduc/toris p(ublici) p(ortorii) servus/pro se et suis/ v(otum) s(olvit) l(ibens) m(erito) 185

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

10)b Idem – AE 1934, 0107 – Capidava (Moesia Inf, Rumania) T(itus) Iul(ius)T(iti) f(ilius) Fab(ia) Sat[ur]/ninus prae(fectus) veh(i)/ cul(orum) trib(unus) m(ilitum) leg(ionis) I Ital(icae) p[r]/aef(ectus) coh(ortis) I Cl(audiae) equitatae/ scriba tribunic(ius) ap/paritor Imp(eratoris) Caes(aris) Ti/ti Aelii Hadriani/ [A]ntonini Aug(usti) Pii/ [p(atris) p(atriae)] cond(uctor) Illyrici/ [u]triusq(ue) et Ripae/ Thracicae 11) L. Aurelius Nicomedes (168-175 dC, Bérard, 1984) – CIL VI.1598 (Pflaum, 1961, 1029) – Roma – exp. Germanica prima (cura copiarum) qui et] Ceionius et Aelius vocitatus est L(uci) Caesaris fuit a cubiculo et divi Veri Imp(eratoris) nutr[itor]/[a divo Antonino Pio equo publico et sac]erdotio Caeniniense item pontif(icatu) min(ore) exornatus ab eodem proc(urator) ad silic(es) et praef(ectus) vehicul(orum) factus et ab Imp(eratore) Antonino/ [Aug(usto) et divo Vero cura copiarum exercit]us ei iniunct(a) hasta pura et vexillo et corona murali donatus proc(urator) summarum rat(ionum) cum Ceionia Laena uxore sua hic situs 12) M. Aurelius Papirius Dionysios (178-180 dC, Bérard, 1984) – CIL X.6662 (Pflaum, 1961, 1029) – (Antium) – exp. Germanica sequnda (cura copiarum) Pii felicis Aug( ) Ducenario/praef(ectus) vehicul(orum) a copis aug( )/per viam Flaminiam/centenario consiliario/aug( ) sacerdoti confrarreati/onum et diffarreat onum/adsumpto in consilium ad IS LX M N/iurisperito antiates publ( ) 13) M. Aurelius Mindius Matidianus Pollio (180-192 dC) – AE 1928, 97 (Pflaum, 1961, 1029) – Efeso  [   ] /  [    ] /     /   / .   / /     -/     ]/  , /   , -/   , -/  %    /    -/  %    -/      ,    -/    , -/    ,  -/ %, % , % -/       -/  ,    /      /  , , / []  -/   [.....]/  [.....]  14) L. Baebius Aurelius Juncinus (v. 195) – CIL X.7580 – (Pflaum, 1961, 1029) – (Cerdeña) L(ucio) Baebio L(ucio) f(ilio) Gal(eria) Aurelio/ Iuncino proc(utarore) heredit(atum)/proc(uratore) aug(usti) praef(ecto) prov(inciae) Sard(inia)/praef(ecto) vehicul(orum) as HS CC/ praef(ecto) vehicul(orum) as HS C/ proc(uratore) ad annonam Ostis/ ad HS IX proc(uratore) Bybliothec( )/ad HS IX Laurentium/ Lavitatim/ Q(uintus) Montanius Pollio/ dec( ) eq( ) strator eius/ v( ) e( ) praesidi rarisima 15) [Fulvius] (198-209 dC) – AE 1931, 0002 (Pflaum, 1961, 1029) – Lepcis Magna (Africa, Libia) [[[6]]]/ [[[6]]]/ [[[6]]]/ [[[6]]]/ [[[6]]]/ [3 pra]ef(ecto)/ vehiculorum proc(curatori)/ XX her(editatium) fratri pio/ Fulvia Nepotilla so/ror cum Q(uinto) Ful/vio Didabibuliano(?)/ marito et Q(uinto) Ful/vio Didabibu[l]ia/no(?) et Q(uinto) Fulvio Seve[r]o/ Iuniore/ sacerdotio Caeni/nense ornato/ liberis posuit 16) ... (v. 205) – CIL XIV.2110 (Pflaum, 1961, 1029) – Lanuvium prae]fec(to) vehiculor[um]/ [donis] donati bello Br[ittannico]/ [aedilis] municipi(i?)] Lanivini et dic[tatoris]/ [sua pe]cunia refe[cit 3]/ [3]bus et mulier[um? 17) P.Sallustius Sempronius Victor (v. 227) – CIG II.2509 (Pflaum, 1961, 1029) A. / / / / / / / / / / / / / / / / / / /  B. ./ ./  18) ... (v. 248) – CIL VI.1645 (Pflaum, 1961, 1029) – Roma praef(ecto)/ vehicul(orum) pro(curator)/lud(i) ma[fni proc(curatori)] Lusit(aniae)/ trib(uno) p[raet(orianorum)] Philipporum A[ug(ustorum)] p(rimi) p(ilari)/ duci leg(ionum) Dac(iae)/ (centurioni)/ corn(icularii) praef(ectum) pr(aetori) 186

THE MILITARY SUPPLY FROM THE LATE REPUBLICAN TO THE EARLY EMPIRE PERIOD

19) Ziadinnio (s. III dC) – CIL III.14412.4 (Pflaum, 1961, 1029) – Vcumakovvci (Moesia Inf, Bulgaria) D(is) M(anibus)// Ziadinnio/ pr(a)ef(ecto) vehi(culorum) vix(it)/ an(nos) LX mil(itavit) an(nos)/ XXXVI [I e]t Aurel(iae)/ Salviae vix(it) an(nos)/ LXX hic sepulti sunt/ quorum memoria(m)/ et tit(ulum) bene meri(tis) patri/ et matri [Zi]abitianus/ prot(ector) domes(ticus) fecit 20) … – IDR.03.01,281 – Dacia ]/[cond(uctoris) p(ublici) p(ortorii)]/ [et praef(ecti)]/ [ve]hicul[o]/rum/Mercato[r]/ vili[c(us)]/ [v(otum) s(olvit) l(ibens) m(erito)] 21) ... – IK 13,00855ª – Ephesus (Asia) ]/ [proc(uratori) provin]c(iae) Na[rbo]/[nensis item i]nsul[ae]/ [Cyrni accepto]rem [vehi]/[culorum adiut(ori) pr]oc(uratoris) [ab ornamentis Praefecti vehiculorum en provincias (sexagenaria) ITALIA Q.Gabinius Barbarus (193-198 dC) – CIL X.7585 (per Flaminiam) M. Opellius Macrinus (v. 205) – Dio Casio 78.11.3 (per Flaminiam) Modius Terventinus (v. 214) – CIL VI.31338 (viarum Appiae, Traianea item Anniae) Claudius Severianus (v. 214) – CIL VI.36899 (viarum Appiae, Traianea item Anniae) Mamilius Superstes (v. 214) CIL VI.36899 (viarum Appiae, Traianea item Anniae) ...I(.)nianus (v. 219) – CIL VI.31370 (viarum Appiae, Traianea item Anniae) …c)elerianus (v. 219) – CIL VI.31370 (viarum Appiae, Traianea item Anniae) Ulbius Gaianus (v. 250) – CIL V.5797 (per Transpadanam et partem Norici) GALLIA (LUGDUNENSIS, NARBONENSE ET AQUITANIAE) C. Claudius Firmus (198-209 dC) – IGR III.181 L. Mussius Aemilianus (v. 247 dC) – CIL VI.1624 … (v. 268) CIL VI.1641 BELGICA ET DUAS GERMANIAE Q. Iulius Maximus Demetrianus (III dC) – CIL VIII.12020 PANNONIAE, MOESIA SUP ET NORICI … (v. 200) – CIL III.6075

187

Part Four THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

4.1 THE INCORPORATION OF THE BAETICAN HINTERLAND INTO THE WESTERN SUPPLY DURING THE LATE REPUBLIC – A READING BASED ON THE DISTRIBUTION OF THE GUADALQUIVIR’S MINORITY AMPHORA TYPES R.R. de Almeida content, preferred lines of distribution, backgrounds and means of production. In this way, recent discoveries in some consumption settlements of the Atlantic façade contributed new data that can help to clarify some of these aspects.

INTRODUCTION The history of research in the Guadalquivir valley generated a focus centred on certain types, specifically on Dressel 20 amphorae, which, by the attractiveness of the epigraphical information and its ubiquity in the Empire`s capital, received enormous attention and dominated amphoral studies:1 the majority of studies and papers have been dedicated to the epigraphy of Dressel 20 amphorae (Chic Garcia, 2001) and this has resulted in a lack of excavations in its kilns, as well as uneven knowledge of the production centres and other amphorae types.

Without doubt, the main remaining aspect that we can highlight, concerning the framework of the amphorae production and the few centres excavated in the Guadalquivir and Genil’s rivers is the persistent lack of data indicating a late Republican or even Augustan production stage, as well as human occupation clearly allocated to these periods. These facts require appropriate answers in terms of the geography of distribution and consumption.

In the last third of the XXth century, the recognized importance of Haltern 70 and Dressel 28 amphora types, as well as the identification of the late Republican types, the identification of other types and their detection in consumption centres has revolutionized this panorama. Despite the remarkable progress achieved especially in the last two decades, there are still many open questions about certain types: their origin and morphological development, its chronological framework, specific 1 From the large number of kilns identified in the Guadalquivir valley, including the course of the River Genil (Bonsor, 1989; Ponsich, 1974; Ponsich, 1979; Ponsich, 1991; Chic Garcia, 1985; Chic Garcia, 1990 Chic Garcia, 2001) only a small number was excavated – highlighting up El Tejarillo (Remesal Rodríguez, 1983), Las Delícias (Saéz Fernandez et al., 2001), the city of Arva (Remesal Rodríguez et al., 1997), Azanaque (Romo Salas, 1993; Romo Salas and Vargas Jiménez, 2001). In the majority of these excavated sites, and in other surveyed on the middle Guadalquivir valley, the predominance of Dressel 20 oil amphorae compared to other types is significant. However, this changes substantially on the lower course of the river, approaching the mouth of the river, as can be concluded from the excavations at the Hospital de Las Cinco Llagas, Seville (Garcia Vargas, 2000; Garcia Vargas, 2003). The image seems to be characterized by the existence of a production activity with permanent structures from the middle of the first century AD. This production focuses almost exclusively on amphorae Dressel type 20 in the case of the middle/higher Guadalquivir, and shared with other types dedicated to wine and fish products, in the case of the area next to the end of the river.

191

The current picture results therefore in a paradox when it can be seen that amphorae produced in this region during the late Republic reached consumption contexts not only on the Atlantic façade, but also at other Peninsular areas or even the central Mediterranean region. As a consequence, the total ignorance of kilns in the Ist century BC, or even the first quarter of the Ist century AD, are contradicted by some assemblages from various consumption centres on the Atlantic façade – particularly the colony of Scallabis, at the time one of the best documented and studied (Almeida, 2006) – in which considerable numbers are found of certain amphora types precedent from the Guadalquivir. These types reveal a markedly Roman typology and seem to have a minority character within the Guadalquivir production contexts. Accordingly, the fact that the reception of imported goods in Guadalquivir amphorae in the consumption centres of the Atlantic façade occurs within the late Republican chronology, converts its production/exportation into a tangible and irrefutable reality. This reality has to be understood within specific late Republican supply contexts, particularly from the second third of the Ist

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

The remaining 90% is composed of coastal Baetican material – with type Mañá C2b amphorae are the overwhelming majority2 – and Italic imports.

century BC, that are dominated by other productions from Italy and the Baetican coast, also the subject of study in this volume, but in which those of the Guadalquivir have to be considered as another component.

The extraordinary volume of Maña C2b containers is quite revealing of the role that they played in troop supplies, especially in western contexts. The massive and continuous importation of peninsular production Maña C2b amphorae can be placed from the second quarter of the Ist century BC at least to the time of Augustus’s principate, and directly related to the distribution/ reception of Italic goods. However, taking into consideration the volume of imports in Atlantic contexts, the Mañá C2b was not merely one vessel among the many that constituted the container traffic required to transport the supplies generated by the conquest process and the military needs of the Atlantic façade campaigns. Rather, it seems to correspond to a preferred container which was directly involved in supplying the west peninsular in the final moments of the Republic.

THE GUADALQUIVIR’S LATE REPUBLIC MINORITY TYPES IN THE CONTEXT OF ATLANTIC AMPHORAE IMPORTS Comparing the sites of the Atlantic façade (especially those of the current Portuguese territory) with other peninsular sites, the considerable number of late Republican Guadalquivir types stand out. The latest investigations, that have defined, characterized and revised the ancient Guadalquivir types (Fabiao, 1988; Fabiao1998a; Fabiao, 2001; Almeida, 2006), show that the recently named Ovóide 1 (=Class 67) and Ovóide 6 (=Classe 24) are predominant, followed by the Ovóide 4 (=“Haltern 70 unusually small variant”?). Particularly important is the amount of the first two, which highlights the pattern of most Peninsular sites – with the exception of the peninsular southeast (Molina Vidal, 1997) – characterized by their low frequency, giving a false image of what should have been a container with a quite considerable volume of production/distribution. This is what we can deduce from sites such as Lomba do Canho (Fabião, 1989), Mesas do Castelinho (Guerra and Fabião, 1994), Santarém (Almeida, 2006) or Castro Marim (Arruda et al., 2006).

Regarding Italic imports, the Tirrenic Dressel 1 wine amphorae are almost exclusive (Fabião, 1998; Arruda and Almeida, 1999). This seems logical considering that the diffusion/reception of Italic wines in the west peninsular arose from the Roman conquest process and the preferences of Roman soldiers (Fabião, 1998a: 175). Paradoxically, Brindisian oil container types, and even Lamboglia 2 wine amphora, are only represented in very small percentages. However, the predominance of Dr. 1 wine over the Lamboglia 2 and the Italian Southern oil types is not an exclusive situation for some Atlantic sites (Fabião, 1989; Fabião, 1998a; Arruda and Almeida, 1999; Pimenta, 2005). From the large number of documented Dr. 1 remains, it is not surprising that there is a low percentage of Baetican containers that were hypothetically used for the wine. Baetican Dressel 1 and Ovóide 4 amphorae appear moderately from the time of the first Roman presence, such as at Santarém (Arruda and Almeida, 1999; Almeida, 2006) or Mesas do Castelinho (Fabião and Guerra, 1994), growing progressively more frequent in the ceramic repertoire throughout the second half of the Ist century BC, in parallel with the decrease of Italian types (Fabião, 1989; Arruda and Almeida, 1999; Fabião, 1999; Fabião, 2001; Arruda and Almeida, 2001).

As well as the mentioned types, the end of the Republican period saw the reception of other minority types, such as the Guadalquivir’s Maña C2b and Dressel 1, and also types Ovóide 2, Ovóide 3 and Ovóide 5, documenting unequivocally its association with the previous types in the sets of imported goods. These numbers have a significant meaning because they allow the recognition in the late Republican Guadalquivir productions a percentage volume far more significant than might have been inferred a few years ago and what must now be seen from the viewpoint of its integration within the range of products intended for military deployments, perhaps since the time of the Civil War (Fabião, 2001: 666). This export trend was accentuated while Caesar was praetor and especially during the second half of the Ist century BC, culminating with the supplies at the limes of Germania, from whose military sites several of the amphorae majority types (Oberaden 83, Haltern 70) have identified and catalogued.

Some specimens of “ancient tripolitan” types are also found, presumably intended for oil transport, in small but significant percentages, and they also suggest a visible trend in the Tagus valley – specifically in Santarém (Almeida e Arruda, 2005), Chões de Alpompé (Diogo, 1982; Diogo and Trindade, 1993-1994) and Lisbon (Pimenta, 2005) – as well as in the SE peninsular (Molina Vidal, 1997), Valencia (Pascual Berlanga and Ribera I Lacomba, 2001), Tarragona (Diaz and Otiña, 2003) and other sites of the current Catalan area (Asensio, Devenat and Sanmartí, 1999). This attested peninsular distribution

As for the products carried in these minority containers, the assessment is more problematic. Ovóide 4 amphorae are associated with wines and Ovóide 6 with olives/oliveoil, but the possible contents of the other forms are more problematic. Amphoral imports from the other provinces of the Empire can possibly throw some light in this direction. For an understanding of Guadalquivir imports it should be noted that in the period between the second third and the beginning of the last quarter of the Ist century BC, they represented only 10% of the imports.

2 Fish sauces, and/or perhaps even meat in brine were established as key items of the Roman diet and therefore major elements in supplies to military deployments (Molina Vidal, 1997).

192

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

Figure 4.1.1 Guadalquivir’s late Republic minority types 193

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

is the apparent synchrony in the various regions of the province and, simultaneously, their obvious differences, as a result of the different degrees of romanization of the various territories (Garcia Vargas, 2001: 102).

in coetaneous chronological contexts leads us to consider its dissemination as a result of institutional army supply (Pascual Berlanga and Ribera I Lacomba, 2001). The large-scale view of general reception/consumption that can be traced from the late Republican period is maintained, in its general outline, until the first years of the Augustan principate, the period when some significant changes began to occur.

The initial chronology for this process can be placed in the first quarter of the Ist century BC on the Mediterranean coast (Cerro del Mar) and the Strait of Gibraltar (El Rinconcillo in Algeciras and Bolonia in Tarifa), and with all probability in the Bay of Cadiz. Despite the maintenance of some archaic elements, the introduction of amphoral types of clear Roman affiliation originated a remarkable typological split, confirming the rise of the romanization phenomenon and announcing a notable exchange in the structure of amphorae production and the distribution of items they contained (Garcia Vargas, 1998).

According to what has just been said and characterized for the amphoral import profile of the late Republican period in the Atlantic façade, it can be seen that two of the three basic containers of army foodstuffs are evident: salted foods (transported in Maña C2b amphorae) and wine (transported in Dr. 1 amphorae). Considering the third component of the Mediterranean food triad, olive oil, in relation to Guadalquivir containers, it is now accepted that these containers, the so-called “ancient” form of ovoid morphology, could be polyvalent (Fabião, 2001; Garcia Vargas, 2001), or at least bivalent. Therefore, as well as the Ovóide 1(=Class 67) and Ovóide 6 (=Class 24), perhaps types Ovóide 2, Ovóide 3 and Ovóide 5 may also have transported this product. It remains tempting to relate these morphologies with those of Italic containers intended for oil exports, with which they have remarkable morphological affinities.

Regarding what may be defined as the first step in this process, the imitation and/or reproduction of amphorae, we can see that it was a common phenomenon in the western provinces during the Roman period. The desire to imitate amphorae types reflects a successful pre-existing business, along with the desire to introduce these imitations into consolidated markets. Over time, these new containers may have acquired a certain prestige within those consolidated markets and their own morphological evolution can differentiate them from the original amphora. Furthermore, we should not forget that in its origin the reason for imitation was the requirement for the local distribution of regional agricultural products (Garcia Vargas, 2001). In the case of the Guadalquivir, the documented reproductions of Roman types are limited so far to Dressel 1 – several examples can be recognized in the Iberian peninsula, particularly in the present Portuguese territory (Fabiao, 1998; Almeida, 2006) – and now also we can include the Maña C2b type, the container par excellence of the Gaditan territory.

THE START OF GUADALQUIVIR EXPORTS IN THE FRAMEWORK OF BAETICAN PRODUCTION The current scenario allows us to recognize that for a large number of the territories integrated within the province of Baetica there already existed, in different scales, a previous tradition of amphorae manufacture, some of which had diverse repertoires of their own. These productions were inherited from the Semitic tradition, which developed mainly in the area of Cádiz and Málaga, and proved to be of great variety and successful within the areas they were intended for, so much so that they continued for almost a century after the process of conquest and pacification of the province (Garcia Vargas, 2001: 102).

The second step in this process, the creation and production of clearly romanized forms, is characterized by a framework of local generalized “interpretations” of Italic amphorae, more or less contemporary. In the case of the coastal territory, the most well known are the Dressel 12 and the Gaditan ovoid. In the Guadalquivir valley, to the already known Ovoid 1 (= Class 67) and Ovóide 6 (= Class 24) can be added the Ovóide 2, 3, 4 and 5 amphora types. These forms result from the creative process of Roman-inspired amphorae with typical western characteristics, occurring in parallel, or immediately after the “reproduction” process of Italic models (Fabião, 1999).

It was not until at least the first quarter of the Ist century BC that these typologies began to be replaced by form of a clear Italic affiliation, testifying to the romanization process. By the final years of the first half of that century there was systematic production of formally Roman amphorae, morphologically separated from the Semitic and “Ibero-Punic/Turdetan” types (Garcia Vargas, 1996; Garcia Vargas, 1998; Fabião, 1998; Fabião, 1998a; Fabião, 2001). This development, which can be seen directly through the ceramic evidence, is the most easily noticeable one. Other issues related the appearance of certain types in this process, with changes occurring to the structures of production and the mechanisms of distribution are the more difficult to define.

As a result, it is now obvious that this juncture opens a new perspective on the Baetican hinterland production and provides a family of amphorae with ovoid bodies that clearly have morphological similarities with Apulian and Adriatic types. Regardless of motivation and similarities with other forms, the question is that the development looks like another example of cultural assimilation, whereby the southern Italic influx seems to have been dominant and the main factor responsible for the

A first point to stress in this replacing process of “late Punic” and “Ibero-Turdetan” containers by Roman ones 194

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

Garcia, 1999: 260). This situation is based on the direction of an effective presence of Italic elements in the Gaditan Bay since at least the time of Caesar, although other evidence suggests the possibility of a spontaneous colonization some time before the first half of the century. That would explain the oldest dates of Gaditan ovoid amphorae in shipwrecks and in some consumption areas (Garcia Vargas, 1998: 57; Garcia Vargas, 2001: 196).

romanization of the amphorae (Fabião, 2001: 666; Bernal Casasola and Jiménez-Camino Álvarez, 2004: 601). But how to consider the start of this whole process? It is not quite yet certain whether the production that began in the Guadalquivir Valley was from the same centres that supported pottery production in Turdetan times. Analysis of the lower Guadalquivir region, currently the best characterized and well known, shows that in the few investigated cases the ceramic repertoires from the end of the IInd century and the early Ist century BC are mostly Turdetan and that the few known production centres are related to an urban nucleus of some importance. This may be seen from Cerro Macareno (Fernández Gómez et al., 1989), Palacio Arcebispal of Seville, Itálica (Luzón Nogué, 1973) and from the Plaza Argollón de Carmona (Rodríguez Rodríguez, 2001).

This scenario suits quite well what we know from Caesar’s western policy and his interest in Atlantic expansion dating back to the expedition to Brigantium on the Atlantic coast in 61 BC; at this time Julius Caesar was the Ulterior governor and it does not seem a mere coincidence that he was accompanied by the Gaditan L. Cornélio Balbo as his praefectus fabrum (Chic Garcia, 1995a).

At none of these last three sites, whose terminal stages reach Republican times, are attested productions of Roman types. Therefore, such absence of any signs of romanization in the amphoral repertoires from Turdetan centres does not allow us to understand how this “industrial conversion” occurred. In spite of this, the lack of Turdetan types on the circuits of supply, which coincides with the start of exports of romanized types, can suggest that the indigenous structures of production were inadequate for that purpose (Chic Garcia and Garcia Vargas, 2004: 311). At the same time, the absence of urban kilns after the IInd century BC could be related to the creation of rural kilns responsible for amphoral production from the beginning of the Ist century BC, for which there is so little evidence (Escacena Carrasco and Padilla Monge, 1992; Garcia Vargas, 1996); this possibility can be supported, based on the possible urban (or not) profile of the first “settlers” (Fabião, 2001: 671).

This is precisely the chronological time span to which are dated the Guadalquivir ovoid amphorae appearing at Lomba do Canho (Fabião, 1989: 41 seq.), in Santarém (Almeida, 2006), whereas other Republican forms on the Portuguese coast could also correspond to this episode, or to other events related to movements of military contingents in the region in the context of the civil war (Fabião, 1989; Fabião, 1998a; Garcia Vargas, 2001: 106). Equally appropriate in this context seems to be the direct relation of the Guadalquivir early types and Gaditan ones. The data from Santarém (Arruda and Almeida, 1998; Almeida e Arruda, 2005; Almeida, 2006) and from the SE peninsular (Molina Vidal, 1997) show a close harmony and that the marketing and redistribution of the Baetican or even Italic products would be distributed by Gaditan agents. Taking into account all the above, the great effort required for start-up production of containers for distributing local products, in which context the interior region of Baetica took part, can be related to the interests and needs associated with army supplies for the military campaigns in Lusitanian territory and episodes related to the civil war across the peninsula (Garcia Vargas, 2001: 107).

In other areas of the province, specifically the Bay of Cadiz, the evidence of the production of “Roman” types in the Ist century BC is demonstrated by the location of production establishments in the city’s territory. Small kiln centres do not require the organized colonization of the territory, but only the presence of Roman or Latin settlers organized in conventus civium Romanorum, or set by their own in the territory of indigenous towns (Chic Garcia and Garcia Vargas, 2004: 311).

The systematic associations of some minority types at several sites in the Portuguese territory – namely the Lomba do Canho (Fabião, 1989), Scallabis (Almeida, 2006) and Mesas do Castelinho (Fabião and Guerra, 1994) – the first, a military site, the second an indigenous village where a military site was established, and the third an indigenous village in which the presence of military is assumed – confirm the joint presence of such types as part of the military supplies. The same association may also be found in some wrecks of the western Mediterranean, such as Grand Conglué 3 (Liou, 2001), but the context of its distribution cannot be determined. Nevertheless, it is not certain that all the movement of foodstuffs was produced for military needs. It must also be considered that there was an important “civil market” that began with the expansion of Mediterranean food uses

It seems that production may have been gradually controlled by Roman or romanized elements, and manifests itself in a typological repertoire clearly dominated by Roman morphologies, where, with the exception of Maña C2b, the other Guadalquivir amphoral Republican types are imitations of Italic containers, or forms inspired by them. Its production area would have probably exceeded the city territories, and included the dependent nuclei, which agrees with the existence of a probable conventus civium Romanorum within many of them (Chic Garcia and Garcia Vargas, 2004). In the case of the Bay of Cadiz, this process seems to be directly related with those territories dependent on colonies or with municipalities founded by Caesar (Chic 195

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 4.1.2 Map – Spread of minority types around the Atlantic façade

its currently assumed diversified production – are already known in a way that can be regarded as significant in archaeological contexts throughout the Empire, especially in its western half, it remains necessary to further this study. Many questions remain unresolved in terms of the origin of the morphologies, their typological definition, formal development, production and distribution, chronological aspects, the specific content of each type, not to mention the production structures and contexts – always more difficult to uncover. However, these types provide a better understanding of Roman Baetica’s complex productive beginning, particularly in relation to its interior region.

on the Atlantic façade (2001), which probably happened from the beginning of the Empire. Despite the paucity of visible evidence of late Republican production in the Guadalquivir valley, which may be due to the lack of systematic excavations of the kilns, recent research has managed to achieve remarkable advances, particularly in relation to our knowledge of the containers from this region. This allows a relatively well characterized framework of form repertoires, basic distribution patterns and exports from the final years of the Republic. Even though these “ancient” amphoral types from Guadalquivir – minorities in the contexts of

196

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

4.2 WEST AFRICAN AMPHORAE (II BC – I AD) C. Aranegui

Moroccan archaeology reveals the importance of fishing activities (Ponsich, Tarradell 1965; Lagóstena et al. 2007) and the commercial traffic deriving from them (Aranegui et al. 2005, 107-134; Cerri 2007, 195-204). There is some documentation related to both the cetariae, testimony of the predominant production activity in its wide northern territory, and the figlinae, generally associated with them (Aranegui et al., 2004, 366-378; Gliozzo et al., 2005, 203-213).

its stamps, but also to suggest a correction in favour of shape Dr. 12 of some supposed imitations of Dr. 1 (Hesnard 1998, 291-293). This seems justifiable because only Haltern 70 amphorae (defrutum, mulsum, olives…) can, in part (Carreras 2000, 419-426), escape consideration as containers for fish products (García Vargas cit.). Certainly, this container has firing failures, as found not only in Sala but also in the workshop AC-7 of Ain Mesbah (Asilah) and in Khedis, together with Salé (Arharbi, Najbi 2004, 169-188), where there is also production of T-7.4.3.3 and Dr. 7-11º forms.

The period between the VI and I centuries BC is best represented in the ceramics by the production of amphorae from Kouass (Asilah) (Ponsich 1968; Kbiri Alaoui 2007) and Banasa (Arharbi, Lenoir 2002, 1-45), but as we move towards the end of the Republic and the beginnings of the Roman Empire, the periods contemplated in this volume, the local pottery information decreases to the point whereby it is easier to evaluate the amphorae of western Mauritania (and then Tingitana), for the outside findings rather than their internal context. Nonetheless, all the salted fish facilities, big or small, are dated at the present from those times.

The intensity of the investigation differs between north and south of the Strait. This has influenced the thesis that considers the south as a subsidiary of the north (Ponsich 1993, 49-62), a thesis that can be clarified as further studies in Morocco are made (Cheddad 2004, 989-1010). This is the reason here for considering the Strait as an integrated space (Chaves et al., 1997, 1307-1320), and for this reason I will use this term with regard to amphorae used for fishing derivatives, leaving the possibility open for Morocco also to have produced those types documented in the Bay of Cadiz, the main trade centre of the area.

Perhaps in comparison to the Baetica, there were some similarities to oil-use vessels in the shape Sala 1/Lomba do Canho 67 (LC 67) (Boube 1987-1988, 183-207) – the first one with Roman typology and firing failures found in Morocco – discarded both by the most recent investigation (Fabião 1989) and by their tituli (lumpha, lympha, lumpa) (García Vargas 1998, 199-206), because it is an area notoriously involved in the distribution of salsamenta, which is closely related to the Guadalquivir valley, an area more diversified in the management of its own resources and particularly related to the northern coast of the Strait Circle, where Cadiz held the role of principle settlement. In these terms, not only is it acceptable to interpret the local shape Dr. 1C from the Baelo subsoil (Bolonia, Cadiz) as salsarian amphora (Étienne, Mayet 1994, 131-138), according to a series of

A third point to be taken into account affects the typological evolution of the amphorae under consideration here. In our area of study, towards the date of the fall of Carthage (146 BC), there is a considerable increase in shape T-7.4.3.3 (Ramon 1995), which reaches an unprecedented level of distribution through the Mediterranean and the Atlantic, both in terms of numbers and geographical extension. This occurs at roughly the same time as T-12.1.1.1, and similar forms, decrease as they are eventually substituted (Aranegui et al. 2005, cit.). It is at that time that we also find a substantial increase in the minting of coins with Punic symbols and tuna fish on the reverse (Gadir, Tingis, Lixus, Sexi, Abdera…). Soon after there was an increase in the marks stamped on salsarian amphorae (Lagóstena 2001, 289), which began to have Latin epigraphs.

CURRENT POSITION

197

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

Figure 4.2.1 T-7.4.3.3. Amphorae; Haltern 70; Sala 1/LC67; with almond-shaped rim; Dr. 8 y Dr. 9 present in Lixus between 50 BC and 10 AD

contents of the amphorae, as well as Italic anthroponyms related partially to families settled in the Baetica region. The dispersion area of products from Tingis and Lixus, superimposable on other Baetican products, follows the routes that go from Rome to Gaul and to the limes germanicus (Cerri, cit.), and also from Gades, through the Atlantic, to the south of Brittany (Morais 2005, 6-40). Despite some preliminary findings from Ain Mesbah and Thamusida (Gliozzo et al., cit) relating to the production of Haltern 70 and B IIA, we have no documentation for this period that allows evaluation and comparison to finds from the Baetic or Lusitania.

All this led to important changes in the economy of the Strait, well before the development of Mauritania (25 BC) or it becomes Tingitana (42-43 AD), changes brought about by the naval power of Cadiz in the Atlantic, a Roman city from 206 BC. In that area, between 70-50 BC, appear oval amphorae with shapes Haltern 70, Sala 1/LC 67, old Tripolitania and Dr. 12, the latter being soon replaced by group Dr. 711 (Aranegui et al.), a general phenomenon noticed in Andalusia, the south of Portugal, Morocco and Tripolitania. It is a typology of amphorae with a developed neck and with handles longer than those found in the Punic tradition, possibly justified by new practices in the usage of the containers. It is not easy to identify each of the production centres of the Strait in the absence of detailed investigations. This group, understood as plural, must have had complementary main and secondary workshops (Fabião 2000, 717-730). The Mauritanian model of Haltern 70 (from the workshops of Ain Mesbah and Kheddis, at least) presents clear, yellowish or orange clays, with chalky oil remover, and with almost no apparent edges. There is no monographic research that specifies it, but it seems that the fabrication of Dr. 12 was repeated in the same workshops.

The stratigraphic evolution of Lixus indicates a phase between 50 BC and 10 AD known as the medium Mauritanian horizon. Another phase is superimposed on this, from 10 to 50 AD, identified as the late Mauritanian horizon (Aranegui et al. 2005, cit., 4 and 115-133). Medium Mauritanian By the middle of the first century BC the typology and amphorae clays in the Strait, are diversified in Lixus according to a traffic that involved a greater number of production centres. More than half of the total number of pieces belong to amphora T-7.4.3.3, which, without proven evidence, could be partially assigned to the

The reign of Juba II saw the evolution of shapes Beltrán IIA and Dr. 7-11 which present tituli alluding to the

198

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

Figure 4.2.2 Lixus. Trade of amphorae between 10 and 50 AD it appears exclusively with pastes from Guadalquivir. The group of amphorae with different origins to those from the Strait entail, on one hand, an Italic assemblage: Dr. 1, Lamboglia 2, Brindis and Dr. 2-4 amphorae, and on the other, amphorae from Lepcis, Cos and a piece of G3 shape that must be reconsidered regarding the date of the context, which is early for the distribution of these amphorae from the workshops of the Rhone.

surroundings of Lixus, considering their abundance and their ceramic paste. Prospections made by INSAP (Rabat) in the coastal lagoon of the estuary of Loukkos, identified pottery activities which remain unknown. It is at this time that the shape T-12.1.0.0 practically disappears and the Latin inscriptions on T-7.4.3.3 are affirmed. One of these pieces bears the nominal inscription AMIS.E, also found in stratum IV of a workshop in Cadiz (Muñoz 1993, 328, fig.17,9). The amphora Dr. 12 appears in the mid Mauritanian, with lips of rectangular section or slightly bell-shaped (figure 4.2.1), with the same variety of pastes observed in the group T-7.4.3.3. The ceramic character of some ovoid amphorae with pastes assignable to the Guadalquivir valley, because of the little pieces of sand of their oil remover, (Martin-Kilcher 1993, 269-320) is different to other cleaner and clearer pastes which are possibly assigned to the coast of Cadiz or the north of Morocco. The shape Sala 1/LC 67 is first to appear (Fabião 2000b, 65-682) and it represents 13% of the total number of amphorae. Not so abundant is the shape Haltern 70 (4% of the amphorae), mostly with pastes from the Guadalquivir in this chronological stage. The group of amphorae with almond-shaped rims which we considered in the tradition of Dr. 20A, represents 2% and

Shapes T-12.1.0.0., T-7.4.3.3., Dr. 12, Sala 1/LC67 and Haltern 70 have production centres in Morocco. Recent Mauritanian (figure 4.2.2) The representation of this phase in excavations of the southern slope it is not high, due to the removal of the stratigraphy in Lixus’ medieval times. With a sample of 247 NMI can be noticed an increase of Dr. 7-11, Dr. 20 and B II amphorae parallel to the decline of LC67 and the relative increase of Haltern 70, while imports of Lamboglia 2 and Brindis amphorae are maintained, Dr. 24 of Italic origin increase and some pieces of G4 from Narbonne appear (figure 4.2.3). Shapes Haltern 70, Dr. 7-11 and B IIA have production centres in Morocco.

199

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

Figure 4.2.3 Lixus. Dr. 20A; Dr. 2-4; G4; cylindrical and Cos amphora. 10-50 AD

Figure 4.2.4 Finds of the Khedis workshop, near Salé (courtesy of H. Najdi)

200

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

4.3 A CASE STUDY OF HALTERN 70 AMPHORAE C. Carreras conditions allowed the preservation of the cargo to the present day. It appears that the ship carried around 60 Haltern 70 amphorae, some of which have disappeared by natural circumstances and human intervention.

Haltern 70 is an amphora type associated with the Atlantic façade. In fact it is a minority vessel in other regions, but it becomes common on the Lusitanian coast and inland as well as the NW Iberian Peninsula. Naveiro (1991) documented in his survey of the main coastal sites in Galicia that Haltern 70 was the majority type. Likewise, Carreras and Berni (2003) recorded high percentages of Haltern 70 in military sites in the NW such as Astorga, Campa Torres and León. On the other hand, Lusitanian sites also provided high percentages of Haltern 70 in places like Arganil (Fabiao, 1989) or Braccara Augusta (Morais, 1998; 2004).

An initial study of the archaeological material dated the shipwreck to around the change of the millennium. The first proposal was between 20 BC and AD 20 according to a single sherd of Italian Samian ware (Form 14.2; date -15/15), some local pottery, the presence of three Pascual 1 amphorae, and the Haltern 70 and Baetican olive-oil amphorae (Dressel 20-Oberaden 83) (Vivar, 2005; Nieto, Vivar and Carreras, 2005). The working hypothesis was that the Culip VIII ship was following a coastal route between the ports of Narbonne and Emporion. Probably one of these ports was the end destination of a larger ship with an important cargo of Haltern 70 amphorae which were later transferred to smaller ships, such as Culip VIII, for coastal distribution through the small ports of Narbonense and north of Tarraconense.

Actually, it was expected to find high percentages of Dressel 20 amphorae to demonstrate the existence of a redistributive system in place to supply the NW Roman army with Baetican olive-oil, and it turned out that Haltern 70 amphorae were the key Baetican vessels in the Atlantic region. This chapter attempts to shed some light on this mysterious typology, whose distribution may explain the importance of the Atlantic route from Augustus onwards. Most information on this typology was collected in 2005 with the publication of the Cala Culip VII shipwreck (Carreras et al., 2005), but there have been some new updates.

After the underwater excavation the archaeologist, Xavier Nieto, decided to ask for the collaboration of CEIPAC1, in order to study the amphorae cargo (Carreras et al., 2005).

Cala Culip is found in the northern part of Cap de Creus, a coastline of NE Spain (Girona) that tends to suffer heavy storms and strong northern winds. Such weather conditions forced Roman merchant ships to take refuge in inlets such as Cala Culip. In the case of Cala Culip, the rocks near the surface use to take sailors by surprise and their ships broke upon them.

HALTERN 70 AND THE CONTENT OF CULIP VIII Culip VIII is the first excavated Roman shipwreck in which Haltern 70 amphorae comprise the main cargo, although it is supposed that similar cargos are not unique in other Roman wrecks. One of the interesting details of the Haltern 70 assemblage is a certain morphological variability, which

So far this small inlet has recorded 8 Roman shipwrecks and the ship discovered in 2001 is known as Culip VIII. It was a ship of small dimensions, around 10 meters in length, of which hardly any remains from the wooden structure have been preserved. When Culip VIII crashed onto the coastal rocks, it sank into a longitudinal space that later was covered by posidonia. These special

1 The research group CEIPAC from Universidad de Barcelona (http://ceipac.ub.edu) studies inter-provincial economic relations in Classical Antiquity, mainly Roman amphorae and their epigraphy. Since 1989 they have excavated the Mte. Testaccio of Rome, the main amphorae dumping area in the whole Roman world.

201

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 4.3.1 Haltern 70 amphorae from Cala Culip VIII

minimum number of three samples, and one stamp D, and one Oberaden 83-Dressel 20, which may have belonged to the crew. As well as one sherd of Italian Samian ware (Form 14.2), there were coarse wares, with a kantharos as the outstanding pot, as well as weights and fishing objects.

is startling because all the vessels are contemporary. Therefore an attempt was made to classify the amphorae according to common features. Three groups were obtained according to rim classification (subgroups A, B and C) and two according to the spikes (1 and 2). Because they were contemporaneous vessels with the same dating, it was thought that the variability could be explained by different amphora workshops with diverse fabrics. Therefore the same analyses by X-ray diffraction were carried out on 20 samples in the laboratory of the faculty of Chemistry in the Universidad de Sevilla, where there is a databank of Guadalquivir river clays (González et al., 2005). However, the result of such analysis and the correspondent data clustering did not provide any matches related to morphological classifications.2

THE WORLD OF HALTERN 70 AMPHORAE Up to now Haltern 70 were considered minority containers in most amphorae assemblages from excavations and little attention was paid to them. The typology was known from excavations in military camps from the German and Raetian limes, such as Haltern, Oberaden and Vindonissa, but under different names. Moreover, they were documented in the first Roman colonies of Britannia, such as Colchester, though their role was obscure.

Another important aspect is that the Culip VIII revealed two complete vessels: one weighed 13 kg and held 28.8 litres and the other weighed 14.8 kg and held 31.5 litres. Again, the variability between the vessels is remarkable.

The 1977 publication of the Port-Vendres II shipwreck brought to light this typology. The presence of some painted inscriptions (tituli picti) that mention a derivative from wine, defructum, on Haltern 70 shapes brought about a major change. From that moment new finds in the limes provinces, as well as new shipwrecks, began to add the final pieces to the vessel’s identity. In the 1990s a renewed interest was shown in the production centres of Haltern 70 amphorae, and the discovery of imitations modified its rather primitive image.

Taking advantage of the presence of residues in some Haltern 70 samples, such remains were analyzed by electronic microscope and gas chromatography (Treserras and Matamala, 2005). Six samples studied presented remains of pitch (resin of Pinus sp.), as well as characteristic grape fitolites (4 samples), so it seems that Haltern 70 amphorae of Culip VIII contained wine or derivatives. Apart from the amphorae Haltern 70, the Culip VIII transported other types such as Pascual 1, with a

Similarly, the systematic study of amphorae assemblages in the Atlantic in the same period, where this vessel is predominant, has modified knowledge. The excavation of Culip VIII has allowed us to reevaluate the current

2

The authors detected two groups, one of 17 samples and one of three. Neither could be associated to the groups of rims and spikes distinguished by form.

202

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

Figure 4.3.2 Haltern 70 typology from Cala Culip VIII

picture of the Haltern 70 and new questions to be answered (Carreras et al., 2005). Every amphora type has its own problems, and for Haltern 70 the following questions arose:

a. Chronology For a long time, Haltern 70 finds were dated between 60/50 BC, however new finds have questioned these time lines. 203

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

tes some problems in the span from 60/50 to 30 BC because there are still only a few well-dated examples.

b. Typology Some confusion still exists in assigning amphorae to this typology, since there are some other types similar in shape, and imitations, which makes classification complicated.

With regards to the end of production, it was originally accepted that the date was around 50 BC, however the continuous finds in later Flavian contexts, such as in Britannia, Culip IV, Lyon, Nijmegen, Pompeya, Herculano and Rome, has modified the end of production date towards the end of the Ist century AD. Some finds in Braccara, Mainz and Rome (Rizzo, 2003) appear in Antonine contexts, but it is still undetermined whether they are contemporary productions or residual.

c. Origin and imitations The detailed study of different workshops in Baetica has allowed us to have a better knowledge of diverse production areas. In the same way, identification of similar vessels (Verulamium 1908), Gaulish imitations (Haltern 70 similes) and later evolutions (London 555-Augst 21) permit a better understanding of imitation phenomena related to this amphora type.

As a conclusion, one can state that Haltern 70 amphorae were in circulation from 30 BC up to AD 90, with the possibility for widening this date range to 60/50 BC in its beginning, and early II century AD for its final production stages.

d. Content Perhaps one of the most controversial issues related to Haltern 70 is content, as painted inscriptions record at least three different substances. Most scholars agree that this vessel was a multiuse container. e. Distribution One of the least-known aspects that can still provide new information is distribution – both temporal and geographical. So far, the most outstanding feature is the important presence of Haltern 70 in Lusitania and the NW of the Iberian Peninsula (conventus bracarense, lucense and asturicense) Haltern 70 have traditionally been studied in consumption and transit locations, as distinct from production areas, but in recent years there has been a balance in study. Today, most new information comes from production centres and nearby areas. CHRONOLOGY OF AMPHORAE HALTERN 70 The chronology of amphorae Haltern 70 was initially defined according to its presence in the Madrague de Giens shipwreck (Tchernia et al., 1978), dated between 60-50 BC, in which an individual was documented.3 Some authors have doubts about such identification, and consider the individual as a Baetican ovoid (Étienne and Mayet, 2000, 190). Therefore, this unique evidence is still under evaluation.

Figure 4.3.3 Workshop production of Haltern 70 – “Simpósio Rota Altântica (Peniche, Portugal)”

Nevertheless, the beginning of such production was not far from this date, since three examples appear in preAugustan contexts (50-25 BC) at the Plaça de la Font in Tarragona (Gebelli and Díaz, 2000). Confirmation of such an early date also comes from Gergovia (Gallia), in which excavations in the Porte Ouest have recorded Haltern 70 amphorae in the phase 2 contexts, dated between 55-30 BC: it is also documented between 30/20 BC in contexts of Saint-Romain-en-Gal and Lyon (dépotoir du Loyasse). However, the initial period still genera-

TYPOLOGY A vessel that is produced for more than 100 years may undergo different changes in shape over the years. Of course, there are common traces all over this period such as ovoid body and solid spike of different dimensions as Culip VIII illustrates and a ball of clay inside it. Its rounded handles with a central groove, which was one of its most distinctive features; it also has an elongated rim that finishes with an everted rim in form of collar with a little step.

3 Vivar (2001) documents a Haltern 70 find in the Cala Galladera shipwreck (Port de la Selva), where the provisional date is late IInd century BC and early Ist century BC. In fact he provides a date range between 125-75 BC. Again, it is an exceptional find that requires further study.

204

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

Figure 4.3.4 Imitations: 3 and 4, Verulamium 1908; 5, Haltern 70 similis; 6, London 555 Over time the body and rim became longer and such evolution was a typological distinction for its diverse variants.

and Pascual 1. Other contemporary vessels linked formally to Haltern 70, and produced also nearby were Dressel 7-11 and Dressel 25.

First phase (from the beginning to 20 BC) Due to the lack of well-dated examples from the beginning to 20 BC, the initial variant has not been completely defined morphologically. Probably it was a period of experimentation with high variability and forms that linked Haltern 70 with other types such as the Baetican Lomba do Canho 67, as well as distant forms such as Lamboglia 2 or Apuliotas, or Layetana 1 related again to other wine amphorae – Dressel 1

Augustus-Tiberius variant It is characterized by small neck and everted rim quite strongly marked outside. Also evident is a clear step at the start of the handles. Culip VIII examples would be prototypes of Haltern 70 amphorae of this period. Claudian variant A stylization of this amphora is observed approximately from the Claudian period in both the neck and 205

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

in Emerita Augusta (Alba and Méndez, 2002) that produced this typology in two different sizes. So far it is difficult to determine the type of fabric of this production and its pattern of distribution. In fact, there are some formal differences, since the solid spike of this production does not include a clay ball, which was typical in the Guadalquivir valley. Other production of this form appears in the Lusitania province located in the Tagus estuary (Raposo et al., forthcoming).

body shape. It is evident above all in the rim with a less marked step, longer rim and trumpet-shape of its mouth. The examples that better represent this variant appear in the Port-Vendres II shipwreck. Flavian variant It maintains a continuous stylization appearing in the neck, body and rim. For instance the rim is longer, with hardly any step and a rounded border. Actually, some Flavian examples in Britannia show a little step in this rim extremity which is distinctive at this time.

Associated to the classical Haltern 70 of the Guadalquivir valley there is another typology known as Verulamium 1908, which was initially identified as a late variant.

Martin-Kilcher (1994) distinguished a later variant dated between AD 130 and 190, but it is coetaneous with Haltern 70 and another amphora type known as Verulamium 1908.

Verulamium 1908 It can be distinguished from Haltern 70 by its stylized body and an extraordinarily long rim. There is a wide variety because only a few vessels have been identified so far, which probably reflects an evolution over time. First datings were established to around the II century AD (Fenchurch St. – London and Augst) and that is why this type was considered a late version of Haltern 70. However new finds of Verulamium 1908 at Usk, Strasbourg, Roma and León in Julio-Claudian contexts modified this chronology, since both typologies were coetaneous at least for 50 years.

Finally, it must be added that Haltern 70 amphorae were occasionally stamped, although there are only a few individual finds with stamps. Additionally, Haltern 70 include an important collection of tituli picti that refer to contents and mercatores names. ORIGIN AND IMITATIONS From early times Haltern 70 amphorae were associated according to its fabric with other productions in the Guadalquivir valley, such as olive-oil Dressel 20 amphorae. Furthermore, the presence of some stamps are common to both typologies, Dressel 20 and Haltern 70, such as CFVFAVITI from Alcalá del Río. A common origin was also suggested by coincidence in names of mercatores from tituli picti of Port-Vendres II (i.e. QVRITTIREVOCATI) in Haltern 70 and Dressel 20. However the first workshops in the Guadalquivir valley were not identified until the revision of the Torre de los Herberos (Orippo) material (Carreras, 2000b).

All the examples documented so far present the characteristic Guadalquivir valley fabric; thereby it seems the only production area. El Cortijillo (Peñaflor) has been proposed as a possible workshop but no details are known. With regards to contents, neither tituli nor remains inside vessels have been recovered therefore it cannot be associated with any particular content. Researchers tend to think that it contained the same contents as Haltern 70. Finally, its distribution is rather misleading because most examples are recorded in Britannia, Germania, Gallia, Rome and Hispania. It is likely that as soon as this amphora type is better identified then its pattern of distribution will become clearer.

Nowadays the number of possible workshops in the Guadalquivir valley has increased with finds at Hospital de las Cinco Llagas, Ilipa, Las Pilas, Los Morales, Cortijo del Maestro, La Gabriela, Cerro de los Pesebres, Cortijo del Mohino, El Remolino and Las Delicias. Moreover another workshop has been recorded at Pinguele (Huelva) which seems to be associated to the production of the Marismas region.

Anther associated form with a distinctive typology is the so-called Haltern 70 similis of Gaulish origin (Desbat and Dángreaux, 1997). This vessel, very similar to the Baetican Haltern 70, was often confused with other forms in early researches.

Nevertheless the Haltern 70 type was also produced in other workshops in the Baetican province, on the coastal Mediterranean and Atlantic, in centres such as El Olivar, Javier Burgos, Puente Melchor and Cerro de los Mártires around the Cádiz bay, as well as Venta del Carmen, El Rinconcillo and Villa Victoria in the Straits region. It is interesting to note that such production from the coastal regions of Baetica, which shared fabrics with fish-sauce amphorae, has hardly been documented outside the production areas.

Haltern 70 similis This form has a very similar typology to the classical Baetican prototype and is easy to misidentify. At a morphological level it can be distinguished by its flat handles, a continuous rim without step, and the flat internal part of the spike. This variant was produced in the Narbonense province, with workshops known in Fréjus, and also other centres in Lyon. The Fréjus fabric is a fine, creamy colour with few inclusions, mainly limestone, similar to other Gaulish types. On

But the panorama of Haltern 70 production is even more complex, since a workshop has been excavated recently 206

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

Figure 4.3.5 Production areas of Baetican Haltern 70

the contrary, fabrics from the Lyon area brown with numerous sedimentary inclusions, mainly in the external coat. This vessel was also contemporary to Baetican Haltern 70, with a date that extends from AD 20 to the end of the Ist century AD.

The last typology associated with the Baetican Haltern 70 is the so-called London 555 (Sealey and Tyers, 1989), or Augst 21 (Martín-Kilcher, 1994), with evident formal differences but with some links to the Baetican prototype.

With reference to contents, only one titulus pictus is known from Mainz (Ehmig, 2003, 10) in a very bad condition, which the author reads as [O]LIV(AE), in other words olives. The distribution of this vessel in concentrated in the military markets of the German and British limes, through the Rhone axis.

London 555 (Augst 21) It is a container of elongated ovoid body, with a solid spike without any ball of clay inside, rounded handles with groove and long vertical rim with a horizontal groove. Recently, Monsieur (2001) has distinguished two phases in terms of formal changes. A phase A

207

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

dated between AD 50 and 100, in which the body is a long ovoid and the horizontal groove is very small. In the following phase B, AD 100-125, the body is still more stylized and the groove greater. With regards to origin, all examples known so far show a distinctive fabric from the south of Gaul.

added. In the example from Mainz the inscription reading is quite clear and refers to MVR(ia).

In relation to contents, the excavation at Pand Sand (London) documented a London 555 filled with oliveoil stones (Sealey and Tyers, 1989). Such content is confirmed by numerous tituli picti registered in diverse finds in Nyon (Oliva dulce), Augst (Olivae ex dulce), London-Walrook (Oliva), London-Cornhill (Oliva alba), Vechten (Oliva alba, Oliva alba), Chesterholm (Oliva alba) and Mainz (Oliva alba). Its distribution, like Haltern 70 similis, is concentrated in the German and British limes as well as the Rhône axis.

With regards to the remains found inside examples of Haltern 70, there is evidence of pitch (pine resin) used as a waterproof coat to contain liquids such as wine. This is the case of examples recovered at Braccara and from the Culip VIII shipwreck amongst others. Furthermore, fitolite analyses of other examples have provided remains of derivatives of wine together with olive remains, thereby both products were associated. Finally, an example from Broch of Gurness (Orkneys, Scotland) included fish products. Again, residue analyses reveal the function of Haltern 70 as a multiuse vessel including derivatives from grape, preserved olives and fish products.

Painted inscriptions in relation to Haltern 70, therefore, suggest at least three types of contents: derivatives from grape (defructum, sapa), olives, and muria.

Apart from grape derivatives, it is feasible that wine was another main content of this amphora type, though no titulus refer to it. However, the large amount of Haltern 70 documented at Atlantic sites cannot only be explained by products such as olives, must syrup and muria. It seems more feasible that such large quantities were due to the transport of staples such as wine.

CONTENT The issue that has probably generated more discussions about Haltern 70 amphorae is their contents, and the need to find a suitable container for the famous Baetican wine. The first painted inscriptions discovered from Haltern 70 in Weisenau (Mainz) referred to olivae nigrae ex defructum, in other words, olives preserved in a derivative from must (defructum), so it appears that the contents had been identified. Nevertheless the finds from the Port-Vendres II shipwreck (Tchernia et al., 1978) revealed a series of tituli picti reading only defructum excellens. This discovery generated a polemic over whether defructum was a boiled wine (vin cuit) that was the main content of this vessel.

DISTRIBUTION A last interesting aspect of this amphora is its pattern of distribution with a high concentration in the Atlantic coast of Lusitania and the NW Tarraconense. Research studies over the last 15 years in these Atlantic regions have shown the trading pattern of this amphora type. Initially its presence was known in the military market of the German and Raetian limes, as well as the Rhone axis (i.e. Lyon, Vienne). From this particular region there are a series of amphora assemblages from the Augustan to Flavian periods, in which Haltern 70 are thought to represent only 1.6% (Lyon – Bas-de-Loyasse) to 8% (Saint-Romain-en-Gal) of the overall amphorae. It appears that Haltern 70 were especially important along the German limes in Augustan contexts, as suggested by the quantities found in the excavations of Ost-Mauer in Xanten.

Actually there is a wide variety of painted inscriptions and remains inside these amphorae that allow us to confirm that Haltern 70 were multiuse containers. In other words they contained different products, although it the proportion of use for each product is unknown (Aguilera, 2005a). Amongst products identified by tituli there appear two synonyms of must syrup, sapa (titulus from Amiens) and defructum (at Pompey, Weisenau, Port-Vendres II, Mainz…). The must syrup is a boiled must, and not boiled wine as some authors indicate. It is a well-known product in Mediterranean countries and was used as a sweetener and an additive to improve some wines. Aguilera (2005b) undertakes a thorough study of different Latin terms that define must syrup and its applications by Roman agronomists, such as for olive preservation. In fact there are groups of de tituli that suggest an association of olives preserved in must syrup, such as oliva ex defructum in Weisenau, Soissons, Mainz etc…

Anther region with Haltern 70 imports was Britannia, although the number is relatively small as the island was conquered in the Claudian period (Carreras, 2000a). Despite this the type was documented over a wide range of sites, but in low percentages, never reaching 1% of any amphora assemblage. On the other hand the Levantine coast of Hispania records a good representation of this vessel, mainly from the mining regions near Cartago Nova. However they are never in the majority in any of these assemblages, but well represented in Augustan contexts.

The final controversial product linked to Haltern 70 is muria – a type of fish-sauce. The reason for the controversy came about because the tituli identified were in bad condition and the readings prone to error. This was the case with the tituli from Celsa and Zaragoza, to which two other examples from Pisa and Mainz should be

Nevertheless the publication of Naviero’s work (1991) presented a turning point in the study of Haltern 70

208

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

Figure 4.3.6 Quantified (cg/m2) distribution of Haltern 70 (Carreras et al., 2005)

In addition, the distribution of Mediterranean shipwrecks seems to respond to a coastal route following the Levant of Hispania towards the ports of Gallia (Narbona, Arles or Frejús), as well as a route going through the Balearic Islands and Bonifacio straits towards the Italian Peninsula.

distribution, revealing that the form was predominant at most Galician sites in percentages that reached 75% of the overall amphora assemblage. Such percentages only appeared at that time with Republican wine amphorae (Dressel 1) and olive-oil vessels (Dressel 20), always linked to military distributions. This special distribution pattern in the NW Iberian Peninsula was later confirmed by studies of amphora assemblages at Campa Torres, Astorga and León. Again Haltern 70 appeared in very important percentages, reaching 60-75% of the overall assemblage. Therefore, this was a regional phenomenon that affected at that time only the NW of the Iberian Peninsula.

However the most remarkable finds of Haltern 70 distribution are the concentrations in the NW Iberian Peninsula, including northern Lusitania. In order to distinguish such phenomenon it was necessary to carry out a quantitative study of amphorae assemblages from sites of the western Roman Empire. The first results appear in figure 4.3.6 (Carreras et al., 2005).

On the other hand, an important presence of Haltern 70 was documented in some Lusitanian archaeological sites such as Arganil (Fabião, 1989) and Mesas de Castelinho (Fabião and Guerra, 1994), so the regional phenomenon appears to affect to the whole Atlantic coast. It was the publication of Morais’ work (2005) on Braccara Augusta and other nearby zones which revealed great densities of Haltern 70 documented in northern Lusitanian sites.

A total of 188 archaeological sites have provided quantified amphora assemblages in the provinces of Britannia, Germania, Gallia, Italia, Hispania and Lusitania. All amphora assemblages have been quantified with different measures, so the same equivalents were established to obtain amphora weights, which were divided into each excavation area in order to get a standardized density. Densities were employed because only standardized measurements enable us to compare diverse sites.4

The land distribution of this vessel is always confirmed by the distribution of shipwrecks with cargoes of Haltern 70, either as primary or secondary cargo. Most shipwrecks are located along the Atlantic and Mediterranean coasts, above all the Levantine coast of Hispania, the Balearic Islands and Bonifacio straits. Such distribution seems to indicate clear maritime itineraries towards the Atlantic Finisterre, and even some finds in Britannia, while others suggest Mediterranean routes.

As a result of this quantified exercise, a quantified interpolation map was created from these values with the help of Idrisi software (see figure 4.3.6). It clearly shows 4

Unfortunately, this methodology does not allow us to evaluate a temporal dimension and it is difficult to compare the same period in diverse places. Therefore complementary measures are reequired.

209

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 4.3.7 Quantified (cg/m2) distribution of Haltern 70 (2007)

from military sites along the German and Raetican limes (such as Dressel 20), but a high proportion of Haltern 70.

a high concentration of Haltern 70 in the production area (Baetica) and the Atlantic façade of Lusitania, mainly the NW of the Peninsula. More quantified data is required from provinces such as Gallia, Germania, Italia and the SE of Hispania to complement this eventual image. Nevertheless, the interpolation map appears to confirm proportions of Haltern 70 recorded in each region.

Nevertheless, absolute amphora quantities inland in the NW military sites are not as high as the coastal areas. That is why it was suggested that some of the contents of the coastal Haltern 70 were transferred to other containers to be transported and consumed by the NW troops. Therefore high densities on the coast, such as Braccara Augusta and Vicus, may identify “breaking points” in the routes towards the NW military settlements.

In 2005 it was realized that the sample did not represent all the western Roman provinces in the same way. That is why it was decided to complement the sample in 2007 with 250 sites, including more published examples from Gallia, Germania, Raetia and Italy. The results can be observed in figure 4.3.7, which changes little, except for the high density in central Gaul produced by sites such as Gergovia, Bribacte and Lyon. Again, the military nature of these three sites may explain the high concentrations of Haltern 70 that do not appear in other settlements in Gaul.

However all those settlements record high proportions of Haltern 70 instead of the well-known olive-oil Dressel 20 amphorae, which are predominant in German and British military sites from the middle of the first century AD (reaching percentages of 60-70% in many assemblage). In this sense, differences in amphorae may be due to chronological reasons since most Haltern 70 appear to have been imported in the Augustan or Julio-Claudian periods.

How can such high concentrations in the NW of the Iberian Peninsula be explained? In recent years attempts have been made to relate such high proportions of Haltern 70 in the NW with the presence of Roman legions on campaigns against the Cantabrii and Astures (27-24 BC), as well as later military garrisons and mining exploitation in the region. In this sense the study of amphorae from Astorga and León allows us to recognize that military sites in the NW do not document the typical amphorae

Taking into account that Haltern 70 were multiuse amphorae, and that no olive-oil containers appear in the NW, would it be possible that this vessel also occasionally carried olive-oil? Another possibility is that Haltern 70 vessels were multiuse containers for providing the different kinds of staples for the legionaries, and 210

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

Flavian periods. In the case of imitations, recent discoveries in the Tajo valley and Mérida show that these phenomena were more complex than initially expected.

therefore a dedicated military amphora. In any event there is no evidence to support such proposals. On the way back from the inland military settlements in the NW, gold and other metals could be transported to the Atlantic ports. Such a theory contradicts the accepted one that gold was transferred to the Mediterranean ports (i.e. Tarraco) through the Ebro valley.

With regards to contents, the multiuse function is expected to be confirmed with new examples. Perhaps, a titulus of vinum is needed in order to include wine as another potential content, but caution is required at present. Finally, the lack of Dressel 20 amphorae and the possibility that olive-oil may have been another possible content is only a hypothesis with no physical evidence.

QUESTIONS TO BE ANSWERED

With regards to distribution, the image of an Atlantic concentration of this amphora type should be confirmed by new data, and interpreted in relation to other archeological and historical data. If there were staging points on the routes towards the military camps or consumption centres, such patterns should be fully explained. In this sense, the important dissemination of Baetican products (coastal salsamenta and other products from the Guadalquivir valley) across the overall Atlantic horizon from Gibraltar to Germania, crossing through Britannia and Gallia, ought to make us think about the reasons of such trade and their protagonists.

There are still many questions to be answered regarding Haltern 70 amphorae. The great number of these amphorae along the Atlantic façade of Lusitania suggests that probably some of these questions may be answered by new finds in this region, a key territory for the study of Haltern 70. An initial question is the starting point of the form’s production, around 60 BC according to Madrague de Giens, and its general consumption from 30 BC onwards. There are not many examples recognized of this first variant and it is difficult to establish morphological relations with other similar vessels (i.e. Lomba do Canho 67, gaditan ovoids…), thereby any new evidence is to be welcomed.

From the Punic period onwards, the Gaditan populations were the main traders along the whole Atlantic façade, and with the Roman conquest their role was even more important. Support of some Gaditan characters and leading players, such as the Balbo family, for the Romans at the end of the Republic (Pompey, Julius Cesar, Augustus) may explain such a predominant role in Atlantic trade, and, as far as we are aware, amphoraeborne commodities.

Another aspect that should be contrasted is the predominance of Haltern 70 production in the Augustan period. Further well dated and quantified contexts are required to fully understand the amount of amphora production in each period, as well as imitations, and, above all, the form’s evolution in the Julio-Claudian and

211

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

4.4 BRACARA AUGUSTA R. Morais to the one found in Braga (vid. table 4.4.1). We do not know, however, if this difference is due to the fact that this study is based on a temporary calculation from a limited number of excavations (that according to the authors only corresponds to 4.19% of the surface occupied by the intramural city) or if it is due to a real supply difference between these two cities of the NW Peninsular.

While analysing comparatively the rhythms and patterns of consumption regarding the goods imported during the early empire period, we decided to highlight the amphorae since these transported essential consumer goods fundamental for the diet of its population. The analysis of the rhythms and patterns of consumption of these products in the city should, however, take into consideration the comparison with other samples equally representative of materials gathered from other sites in the Peninsula.

The explanation for this difference, could, however, lie in two truly significant aspects: on the one hand – as we can see from the chart of transportation costs from Baetica presented by C. Carreras Monfort (Carreras Monfort, 1996: 210, Fig. 3) – the supply of Asturica would be more expensive than that of Bracara (figure 4.4.2); on the other hand, that difference could be due to a larger degree of self-sufficiency with Asturica and its region, taking into account its administrative importance and the presence of civil servants, documented by the epigraphy, with responsibilities in both territories, such as a legatus augusti per Asturiam et Gallaeciam (CIL, II, 2634), a praeses provinciae Gallaeciae (CIL, II, 2635) and a procurator Asturiae et Gallaeciae (CIL, II, 2643).

In the case of amphorae we are particularly interested in the data provided by J. Naveiro López for the Gallaecia area (Naveiro López 1991a: 63-73) and by C. Carreras Monfort for the Asturian territory (Carreras Monfort, 1996: 205-210). In both cases, and at first sight, the studied sites reveal an identical consumption profile to that of Braga, which was directly involved in the Atlantic trade route. However, as C. Carreras Monfort points out (id. ibidem, 1996: 205-207), there are some differences in the types of supply between the different sites, whether we are dealing with coastal places (for example Campa Torres and the NW coastal centres) or inland centres, which present particular supply dynamics (for example Legio and Asturica Augusta).

In spite of the fact that we are confronted with two major NW cities that benefited directly from the supply of Atlantic products, the analysis of the types of amphorae and their quantitative expressions reveals a consumption profile that is truly differentiated.

Regardless of these differences, if we compare the quantity and types of amphorae existing at these Asturian sites (figures 4.4.1 a and b) with the data obtained in Braga (table 4.4.1), we verify a larger flux of imports in this city.

We must point out right away the greater Romanization and economic importance of Asturica during the early years of the city. In fact, besides the above-mentioned inscriptions that compel us to stress the main administrative role of Lucus and Bracara, we also have a significant group of wine amphorae from Italy (Dressel 24) and from the Aegean region (Dressel 2-4 and Rodia Type). These amphorae prove the existence of indigenous elites and of a population of Mediterranean origin in Asturica, who valued and could afford these luxurious products (vid. Carreras Monfort, 1996: 206). This is not a

In fact, a monographic study recently published by C. Carreras Monfort and P. Berni (Carreras Monfort and Berni, 2003: 633-673) on the amphorae of Astorga seems to corroborate that analysis. As a matter of fact, as we can see in table 4.4.2, the quantity and diversity of the types of amphorae collected in Astorga is significantly inferior 213

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

a

b Figures 4.4.1a & 4.4.1b Graphs of amphorae finds (in Carreras Monfort, 1996)

surprising situation if we take into consideration that the city was chosen to be the residence of the legati iuridici and the procuratores augusti and metallorum, who were involved with mining exploitation (Tranoy, 1981).

in turn, was due to the relative nearness of the coast and the population density of its territory. These two factors might have naturally attracted traders, as the presence of negotiatores during the Claudian period seems to document.

The significant presence of amphorae in Braga, in particular of Haltern 70, makes us believe that the importance of the city lay in its market potential, which,

The supply of the city must have also benefited from this situation by getting other products, most of which were complementary to the transport of amphorae. 214

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

Table 4.4.1 Amphora found in Braga in the Early Empire period Amphorae

NF

NMI

%

Rodian Type

58

24

1.65

Dressel 2/4 (italic)

57

28

1.92

Dressel 2/4 (gaulish)

6

6

0.41

Dressel 2/4 (baetica)

20

9

0.62

Dressel 2/4 (tarraconensis)

5

4

0.27

Dressel 2/4 (eastern)

3

2

0.14

Dressel 2/4 (african)

3

3

0.21

Dressel 28

2

2

0.14

Gauloise 4

18

11

0.75

Matagallares I

1

1

0.07

Beltrán 68

1

1

0.07

Urceus type

52

41

2.81

3614

933

64.04

2

2

0.14

Dressel 20

40

27

1.85

Dressel 7/11 (baetica)

127

69

4.74

Dressel 7-11 (lusitanian - Peniche)

13

11

0.75

Haltern 70 "Arcaic Dressel 20"/Tipo B (Augustan-Tiberian)

Dressel 7/11? (lusitanian)

1

1

0.07

Beltrán II A

2

2

0.14

Beltrán II B

8

8

0.55

Puerto Real 1

2

2

0.14

Dressel 14 (baetica)

17

17

1.17

Dressel 14, A (lusitanian)

50

50

3.43

Dressel 14, B (lusitanian)

73

70

4.80

Dressel 14, C (lusitanian)

7

7

0.48

Dressel 14 Tardia

2

2

0.14

Dressel 14, ind. (lusitanian)

20

18

1.24

Almagro 51 C (baetica)

4

4

0.27

Almagro 51 C (lusitanian)

12

10

0.69

"Almagro 50" / Keay XVI, A

2

2

0.14

"Almagro 50" / Keay XVI, C

1

1

0.07

Almagro 50 / Keay XXII

9

7

0.48

Beltrán 72, B

2

2

0.14

Richborough 527

10

6

0.41

PE 18

2

2

0.14

Dressel 30

1

1

0.07

Majuelo II / Almagro 51 C, A

1

1

0.07

Africana Grande B

1

1

0.07

Regional Form I

1

1

0.07

Regional Form II (Beltrán 72, B)

1

1

0.07

Regional Form II ("Almagro 50" / Keay XVI, C)

86

49

3.36

Undeterminate Local (Late ?)

6

5

0.34

Module 1 (local)

3

3

0.21

Module 2 (local)

8

8

0.55

Module 3 a (local)

1

1

0.07

Moduloe3 b (local) Total

1

1

0.07

4356

1457

100.00

215

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Table 4.4.2 Amphorae from Astorga (in Carreras Monfort e Piero Berni, 2003)

We have come to realize that Bracara Augusta, just like other nuclear cities of the Roman world, played an important role in the inter-provincial trade as an importer and redistribution centre of foodstuffs and manufactured goods.

We have also seen that the city had a major role as a productive centre, directly manifested in several documented handmade activities, related to the needs of the urban centre and the rural world that were under its control.

This role carried out by the city, which can be considered as a basic unity in the exchange process of the region, based itself necessarily on an intermediary network; right from the beginning it is possible to find large-scale traders (negotiatores), specifically documented in the Claudian period, which make us assume the existence of small traders (mercatores), whether they were congregated or not in specific collegia or corpora.

Just like in other cities (vid. Foraboschi, 1990: 820), the city markets (mundinae) must have been organized according to the available farmed goods and these were arranged taking into account their perishability, their weight, and the routes and costs of transportation. As E. Gabba (Gabba, 1988: 144-149) and J.M. Frayn (Frayn, 1993) have demonstrated, these markets were of the utmost importance for exchange of surpluses. They took place every eight days in the same city and were sometimes situated on private lands (Frayn, 1993: 4-5), after a weekly rotation with other centres of the region (vid. Columella, Rust. I, 18; Pliny, N. H. XVIII, 3, 13; cfr. De Martino, 1979: 163).

Together with these mercatores and collegia, not yet epigraphically documented, we can also infer the presence of administrative workers in charge of financial matters (tabularii), similar to what is documented in Lugo.

216

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

Figures 4.4.2 Hispania – Costs of transportation from Baetica (in Carreras Monfort, 1996)

concentrate preferentially in its outskirts, or following the networks of routes, and obeying a system of occupation of the radial type, directly dependent on the city and essentially connected to its supply (id. ibidem; id. 1995).

As well as these markets, there were nearly permanent shops (tabernae) and buildings for periodic sales of goods (macellum) (De Ruyt, 1983; Frayn, 1993: 1-10), characteristic of each community, Bracara Augusta included.

However important these data may be for the understanding of the economic role of the city, they need to be revaluated taking into account the wider perspective of the territory. To accomplish this we must take into consideration the models of distribution, the creation of infrastructures and the organisation of the landscape (through the analysis of possible centres for example) so that we can make maps of the settlement distribution. In fact the value of these maps for the NW Peninsular region is not absolute because of the difficulty one has to define the organization of the nuclei, their chronology and mutual relationships, which is the result of the lack of excavations in most cases.

In this way, we can assume that the power of farm production in the economic organization of the region might have influenced the different manufacture of nonfood items that depended on food production, and, consequently, the character and volume of trade which was mainly made up of agricultural products. In fact, just like other Roman cities, Bracara Augusta enjoyed a mixed economy based on agriculture, on trade, and on industry. This assumption is developed not only from the analysis of the materials and the remains of the workshops documented in the city but also from the analysis of the proto-historic and Roman settlements in the region where the city is situated. Among many other studies, the one carried out by Manuela Martins in the basin of the Cavado River (Martins, 1990) gives us important elements for a better understanding of the Roman occupation in the rural space near the city.

Leaving aside these unresolved issues, we can, however, demonstrate that within the nucleus Bracara Augusta, or in its territory, a large range of instrumenta necessary for farming activity or domestic life were produced: oillamps, common wares, fine wares, amphorae, dolia, building materials and several other objects (such as pondera and terracottas). As far as the foreign trade between provinces is concerned, we should emphasize that Bracara Augusta

The new types of sites, small villages or uillae seem to appear at the first stage, together with the city. They 217

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 4.4.3 Location of Bracara Augusta in the context of the road and sea route networks

transported exclusively, and in large amounts, these byproducts, especially defructum. This latter had to possess special characteristics to be able to be commercialized as an alcoholic drink (or if we prefer as low quality wine). In this context, and taking this last hypothesis into consideration, we can only conclude that the result of that process is a low quality and very sweet wine that was drunk at meaner tables or sold at the tabernae, while consumption of the better appreciated and finer quality wines (with a higher degree of alcohol) was only reserved for the most privileged classes.

benefited from a real economy of scale. Such a conclusion is the result of a group of imported materials found so far in the city, which inclusively leads us to believe that the site was selected in part for its market potential. As we have several times stressed, this potential was also enhanced by the relative proximity of the coast, which made it possible to access to privileged areas in order to carry out an economic exchange of several goods from other regions. In this way, we can also accept that the city might have shared the role of redistribution centre (figure 4.4.3) with the two other capital cities of the NW Peninsular, Lucus Augusti and Asturica Augusta.

Still to be solved is the presence of only small amount of olive-oil amphorae. To try to explain this we have formulated two hypotheses, which, despite not being contradictory, might help explain the scarcity of these amphorae: firstly we can accept that this product could have been produced in other regions, namely in Trás-osMontes, and secondly people probably continued using animal fat that was used as a substitute for olive-oil, which could explain the small numbers of these types of amphorae.

In fact, as we can see from the amount and types of amphorae so far collected, we know that the city was in contact – directly or indirectly – with the larger production centres, and thus well integrated into the Empire, reflecting the rhythms and fluxes of regular exchanges which were in the meantime discovered in the context of the NW Peninsular (Morais, 1998a: 81).

We can also highlight the presence of amphorae that are not normally found outside their places of production. We refer to the wine amphorae Dressel 7-11 of Tarroconensis production and to the wine amphorae Galoise 5 and 7 from the south of Gallia, amongst others. As far as the first amphorae are concerned, we admit that there is a weak representation outside their area of production. However that is not the case for the second type of amphorae. We believe that the small amounts of these amphorae are due to the absence of researches. In fact, at

We realized that these centres had a privileged relation with the province of Baetica through the massive importats of Haltern 70 amphorae and the considerable presence of Dressel 7-11 fishing amphorae. As far as the massive presence of Haltern 70 amphorae is concerned, and in an attempt to reveal the question related to their content, we put forward two alternatives: these amphorae both transported wine and additionally the by-products referred to in the tituli picti, or they 218

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

Baetica, has definitely proved the importance of the Atlantic route as a privileged circulation axis in the regular supply of the NW Peninsular. This route was also important for the supply of Britannia and some northern sites of the German limes, which could hardly have been supplied by other maritime trade routes as these would have turned out to be much more expensive than the Atlantic one.

least as far as the NW Peninsular is concerned, we were able to report that the Galoise 5 and 7 amphorae were present at different sites in what is today northern Portugal and Galicia. We have also reported the moderate appearance of the wine amphorae Dressel 2-4, probably from Tunisia, and of the Matagallares I and Béltran 68 amphorae equally documented in the city. The residual presence of these amphorae in the city should, nonetheless, be borne in mind so that in the near future we can map of their diffusion and register their distribution and supply routes. For the same reasons we can highlight the presence of fish amphorae Puerto Real I, because, with exception of the production centre of Puente Melchor and of specimens found underwater, these amphorae have only been documented so far in the region of the German limes, in Augst and on the coast of Mauritania, in the Strait of Gibraltar.

On the other hand, we should not find that privileged situation strange, especially if we think that the Baetic province was one of the most fertile areas of the west in this period, a region where everything was produced and in large quantities. However, according to the quantitative and chronological data obtained, we should distinguish two different phases in this supply route: a first phase, between the time of Augustus and the mid Ist century, dominated by largescale imports of Haltern 70 amphorae, and a second phase related to the conquest of Britannia, in the year 43, which allowed the consolidation of the Atlantic circulation, especially of fish-sauces, still associated with the Haltern 70 amphorae.

Throughout this review of amphorae gathered in the city, we have also highlighted the residual presence of amphorae of undefined content, represented by the amphorae Type 8.1.3.3 (PE – 18) from Ibiza and by the African amphorae Dressel 30 and “Africana Grande” B from Mauritania Caesariensis and Byzacena respectively. The Richborough 527 amphorae were found in a different situation; a considerable number of these amphorae were documented and they were probably meant to transport alum, a product which, as we have seen, was originally from the island of Lipari (Italy) and was used to fix paint. We are of the opinion that the absence of specimens of this type of amphorae in the present Portuguese territory is due to lack of investigations rather than to a restricted trade or to a specific trade route.

A significant number of other ceramics is also documented in the city, such as the mortaria from Italy and Baetica, and the fine table wares, represented by terra sigillata and by thin-walled wares, originally from different areas and production centres. After all that has been said we can easily conclude that the city attracted a complex exchange network with several different provinces of the Empire (figure 4.4.4). This fact does not mean, however, that the city got its supplies directly from the production centres or from the distribution ports connected to those centres. In fact, as the several shipwrecks so far documented show, the ships transported a variety of goods. These goods had different origins since they had been bought in the main ports that were responsible for the storage of various products.

Regarding the large concentration of Haltern 70 amphorae and the need for a more detailed observation of the formal characteristics of this amphora, it was possible to recognize a group of amphorae that presents an unquestionable Guadalquivir fabric, but whose formal characteristics lead us to suggest its integration within the known Dressel 7-11 and Dressel 14 forms.

Several shipwrecks confirm this interpretation, such as Cala Culip IV, a ship that transported Dressel 20 amphorae, thin-walled wares from Baetica, terra sigillata from the south of Gallia (originally from the production centre of La Graufesenque) and Italic oil-lamps and mortaria; the Cabrera 3 wreck, with Baetican and Tripolitanian materials aboard, and the shipwreck of Port Vendres I, carrying amphorae to be used in the transport of oil-lamps and fish-sauces from Lusitania, African terra sigillata.

The identification of a group of materials, whose regional and local origin was duly supported by archeometric study, was extremely important for the history of the economy and the trade of the NW Peninsular in the middle and late empire period. We are obviously referring to two different sets of amphorae that we have differentiated according to their morphological and fabric characteristics: the first, which we have temporarily named Regional Form I (Galoise 4 imitation) and Regional Form II (“Almagro 50” Keay XVI, var. C and Béltran 72 imitation), has a strong resemblance to the fish amphorae recognized from the Gallician kiln of San Martiño de Bueu; the second set, which corresponds to a group of forms of local fabric, has similarities to the flatbottomed amphorae used for the transport of wine.

We might thus assume that the majority of the goods collected in the city might have come, directly or indirectly, from products stored in Cádiz, especially if we think about the privileged commercial relationship this city had with the province of Baetica.

This appreciation of the group of amphorae so far gathered in the city, and of its privileged relation with

We cannot, however, assume that the city only enjoyed an indirect supply from the port of Cádiz. As a matter of 219

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 4.4.4 Origin by areas and centres of production of the Early Empire products imported by the Roman city of Bracara Augusta The same connection to the port of Brigantium can be accepted for the arrival of Hispanic terra sigillata products from Tricio (La Rioja) even though it is much wiser to think that these might have been transported overland through the Douro Valley resorting to draught and pack animals used to cross the mountainous routes.

fact, it is perfectly possible that the city (and other important nuclei from the NW) might have been supplied with shipments of boats mainly dedicated to high-seas navigation that set sail from other Hispanic ports, such as from Hispalis, Tarraco and Nova Carthago or even from ports outside the Peninsula situated in Narbonne, Arles, Pozzuoli, Ostia and Carthage.

However, as we have highlighted in the chapter concerning sea-transport networks, the city did not benefit directly from the ports of call along the Atlantic route, the itinerary of which included obligatory stops at Brigantium, Burdigala and Gesoriacum. It is thus possible that the city benefited from a supply from intermediate ports of call, such as that of Cale and of the River Minho, or even directly from the secondary ports of call connected to the traffic of small and large cabotage situated in the mouths of the Leça, Ave and Cávado rivers.

This might be proved through the presence in the city of amphorae of Italic and Aegean origin and of amphorae from the south of Gallia, which naturally came together with other ceramic products equally documented in Bracara Augusta, as is the case of the terra sigillata, the mortaria and the thin-walled wares. We should equally accept the hypothesis that some of the products collected in the city might have come from other routes of circulation. We are referring to some products from the south of Gallia that might have followed other exchange circuits by using the well-known AudeGaronne route, also known as the Gallaic Isthmus route. Such a connection might imply that the city was supplied from the port of Brigantium, which was in turn connected to that route through the Cantabrian coast.

With respect to the analysis and diffusion of the Haltern 70 amphorae in the present Portuguese territory and in the region of Gallicia, we reported the existence of many places that indirectly could have been related to the supply of the city and its region. In the present Portuguese territory we have distinguished, among 220

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

the “ports of call” of the Douro and Minho rivers, might have unloaded part of their cargoes in that river, in the socalled Cala – a deep area with exterior reefs that worked as a natural breakwater, just like the leixões of the mouth of the Leça River. As we have equally highlighted this situation does not overrule that there might have existed a river port on the Cávado River, probably situated near the city in Areal de Caída or a little further downstream in the area of Barca do Lago.

others, the role of the Cale, an important anchorage, which possessed port structures on its banks, and the leixões on the mouth of the Leça River, whose exterior reefs might have worked as a natural breakwater. The group of materials recovered so far is a witness of this privileged position, especially the amphorae recovered from the excavations of the old prison (Aljube), situated in Rua de S. Sebastião (Porto), and the amphorae collected in the old and recent excavations in Castro de Guifões, near the mouth of the Leça River.

It is in this context that we can also accept as correct Ausonius’ reference (Ordo, XIV) when he sites the city of Braga near the sea beaches. We might say that he anticipated by fifteen centuries the “garden situated by the sea”.

As far as the direct supply of the Roman city of Bracara Augusta and nearby territory is concerned, we have pointed out the importance of the Cávado River. We have admitted that the larger vessels, besides heading towards

221

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

4.5 THE MINING AREA OF QUINTA DA IVANTA R. Morais south of Gallia and Hispanic sigillatas with a chronology of mid to late Ist century; no fragments of amphorae have however been registered (vd. Pinto, 1994, 5-19; Figs. 112).

The testimonies of mining settlements from mining places so far identified in the Portuguese territory are numerous, such as the ones at Serras de Santa Justa, Jales, Três Minas, Sª do Castelo (Urros), Macedinho, Covas de Seixo de Ansiães, Selores and Poço das Freitas. The same can be said about settlements situated in proximity to them, such as the mines of Mouros (Cerdeira), S. Domingos and Aljustrel. Nevertheless, apart from the latter one, the study of the amphorae from these settlements has not greatly aroused curiosity.

The case of Aljustrel is different given the existence of two burial grounds: the Valdoca and the Monte de Farrobo. The first was the result of systematic excavations carried out by Abel Viana and Veiga Ferreira, recording a large group of 496 incineration graves (65%) and inhumations (35%), the latter practically without goods (Ferreira and Andrade, 1966, 3). The vast spoils there recovered, studied by Jorge and Adília Alarcão are made up of common wares, together with early and late imperial sigillatas, thin-walled wares, glasses and oil-lamps datable to the Augustan period up to the second half of the IInd century (Alarcão and Alarcão, 1966, 1-98, Ests IXXXVIII). Once again the amphorae are practically absent; we only know references to amphorae fragments collected among the graves or on the gravestones (as in the case of numbers 304 and 305) (vd. Ferreira and Andrade, 1966, 3, 6), but without allusion to shapes and respective quantities. In fact they only published two amphorae, without context (Alarcão and Alarcão, 1966, 100-102; Est. XXXVIII), that fit Lusitanian Dressel 14, whose pearled rims, similar to rim numbers 4 to 6 found in Quinta da Ivanta, can be attributed to the IInd century (vd. Mayet and Silva, 2000, 99-100).

In fact, even the amphorae of Aljustrel, previously published by Laura Trindade and Dias Diogo in 1995 (Trindade and Diogo, 1995, 11-14), and recently presented by Carla Maria Braz Martins in the catalogue of her PhD thesis (Martins, 2005), correspond to fragments without context, from the old gatherings and presently deposited in Lisbon museums. The same can be said about the necropolises associated with mining settlements. As a matter of fact, besides some funerary steles without context found in the mines, or nearby, only the necropolises of Corredoura (Campo, Valongo) and Aljustrel (cemeteries of Valdoca and Farrobo) have been so far documented. There is also one reference to a necropolis associated with Três Minas, situated in Veiga (Lago das Covas). Unfortunately, we only know that glasses and “clay vessels” had been collected there, but they have in the meantime disappeared. However, from that place six funerary tombstones, five of which make reference to people from the Clunia conventus, are still known (vd. Almeida, 1973, 554-555).

The necropolis of Monte do Farrobo, excavated by Ruy Freire de Andrade, O. da Veiga Ferreira and Priest Serralheiro in 1959, is very probably part of the necropolis of a villa, with a spoil datable to 50 and 350 AD, with special emphasis on the period between 50 and 150 AD (vd. Alarcão, 1974, 31). Among the recovered spoil, there are fibulas, glasses, oil-lamps and common and fine wares (thin-walled wares, glazes, African sigillata), but unfortunately no amphorae fragments.

The necropolis of Corredoura (Campo, Valongo), in which we are particularly interested because of its location, is related to the mining settlements of Santa Justa and Pias, given its location below Pias hill (Pinto, 1994, 7). The scarce materials dug there that are truly datable are essentially represented by sigillatas from the 223

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 4.5.1 View of the excavation (northeast/southeast area)

It is part of a nucleus or mining site of the Douro mining region (figure 4.5.3), especially rich in transversal gold veins (with high percentage of gold per ton2), thin and short, from quartzites and quartzose schists, of irregular mineralization, defined in the Geological Chart of Portugal as belonging to the “graywacke-schist anteordovícian complex and derivative metamorphic series” (Carvalho, 1969) (figure 4.5.4). As far as the administrative and political unity of the Roman era is concerned, this region was incorporated within the territory of the conventus bracaraaugustano, and was part of an autonomous mining district, situated between the civitas of Cale and of Tongobriga (Lemos, Meireles, forthcoming). It comprises an area about 30 km long by 10 km wide, oriented from northwest to southeast, from Valongo to Castelo de Paiva (Carvalho, 1969, 101) and possesses some mining sites with Roman remains, or what we assume were exploited by them. Among these, known as Fojos, we highlight the following: Fojo das Talhadas, Fojo do Escritório, Fojo da Pirâmide, Fojo das Escadas, Fojo do Alcantilado, Fojo das Pombas and Fojo da Valéria (vd. Martins, 2005, 265293). To these we should add the small site of Quinta da Ivanta, probably a “mining settlement”, which was examined as a rescue excavation in 1999, under the coordination of José Marcelo Mendes Pinto, and more recently excavated by Lídia Baptista and Liliana Barbosa, archaeologists working for the Ricardo Teixeira & Vítor Fonseca Archaeology Company,3 in the ambit of a plan designed to minimize the impact of a future real estate enterprise predicted for that area.

Figure 4.5.2 Underground gallery The spoil from Quinta de Ivanta comes from a small habitat associated with the mining exploitation of a primary Roman quarry (figure 4.5.1 and 4.5.2), situated on the north slope of Serra de Santa Justa, between Ribeira de Valongo and the entrance of Fojo das Pombas.1

2 According to Allan (1965) the gold contents vary from 20 and 40 g/t. 3 We thank all the archaeologists involved for the opportunity of studying the amphorae collected at this site.

1

Quinta da Ivanta is situated in the municipality of Valongo, a district of Oporto, and is referred to in the military chart 1/25.000, sheet 123: 41º 11’ 15’’ Latitude North; 8º 29’ 4’’ Longitude West (Greenwich).

224

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

Figure 4.5.3 Military chart 1:25.000, Page 123 (41º11’15’’ Latitude North; 8º 29’4’’ Longitude West, Greenwich)

Figure 4.5.4 The mines in the mountains of Santa Justa and Pias in the context of the NW Iberian Peninsula

The amphorae stand out among the remains collected. They are represented by a set of more than 70 specimens, exclusively represented by Baetic and Lusitanian productions.

corresponding to 74.2%; the remaining 25.8% correspond to Lusitanian productions (exclusively represented by the type Dressel 14). The Lusitanian productions, except for the fragments numbers 4 to 6, which present a pearled rim datable to the IInd century, can be dated to the Ist century (nº 1 to 3).

As we can see in the chart and histogram (figure 4.5.5), the imported amphorae from the Baetic predominate, 225

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 4.5.5 Chart of the imported amphorae from Quinta da Ivanta

The Baetican productions, except for four fish amphorae belonging to three Dressel 7-11 specimens (4.8%) (nº 9 to 10) and a Béltran II B specimen (1.6%) (nº 11), correspond to productions from the Guadalquivir. Among these, as one would expect, the Haltern 70 amphorae stand out, with about 30 examples (48.4%) (nº 13 to 19), followed by about eleven Dressel 20 finds (17,7%) (nº 20 to 26) and one Urceus type specimen (1.6%) (nº 27-28).

In fact, as we can see in the chart and histogram, Haltern 70 amphorae are the most abundant in the whole set, with about 30 specimens.

The typological review of the amphorae and their stratigraphic contextualisation allow us to accept a diachronic occupation of the place from the period of Augustus to the beginning of the IInd century, with special focus on the first half of the Ist century. The presence in the same stratigraphic unities of Italic terra sigillata fragments, and from the south of Gallia, and, in less quantities, of Hispanic terra sigillata, confirms the previous dating. Despite the state of fragmentation of the majority of these fragments, which make it impossible to fit them into a specific typology, we can highlight two fragments of Italic terra sigillata dishes Consp. Types 18.1 and 18.2 (nº 29), datable to 15 BC and 37 AD, two fragments from the south of Gallia of the form Drag. 30, decorated with a motive of St. Andrew from the mid Ist century AD, and a fragment of a dish with a stamp of the potter Albinus (i), a potter working in La Graufesenque between 40 and 80 AD (Oswald, 1964, 10; Polak, 2000, 162) (nº 30). According to the stamp [OFALBINI] it is a dish, probably Drag. 15/17 or 18, datable to the time of Nero, the period in which this potter was most productive (vd. Polak, 2000, 162).

Such a quantity by far exceeds the relation percentages of these amphorae in the NW Peninsular, which, as we know, and despite being representative (vd. Fabião, 199394, 219-245), do not reach the percentage values documented here.

But the presence of Haltern 70 amphorae is not the only interest of this sample; the proportional significance of the presence of Dressel 20 amphorae, with about eleven specimens, is very curious.

The proportionally significant difference of Dressel 20 amphorae found in this mining settlement of Quinta da Ivanta, relative to other sets known in the NW Peninsular, can only be explained in the context of supply of an institutional and state nature to mining sites, and by the unquestionable importance of olive oil in these contexts. Curiously (or not), we find that the materials of the old excavations in Aljustrel present a consumption profile similar to that of Quinta da Ivanta, even though they reveal a wider spread of occupation. Among the published materials (Trindade and Diogo, 1995, 11-14; Martins, 2005, 500-507)4 there are – with the exception of a basal fragment attributed to a Dressel 2-4 amphora from Campania and a fragment of type Almagro 50/ Keay XXII (Lusitanian 4) – Haltern 70 Baetic amphorae, Dressel 20, Dressel 7-11, and the Lusitanian Dressel 14

This sample is of great interest because it proves that there is a significant coherence between the strong concentration of amphorae in the NW with the good sea and river routes, following the itineraries that led to the gold workings and military camps, and, as such, related to supply of the armies in the first decades of the imperial period (Morais and Carreras Monfort, 2003, 112).

4 From the description of the fabric and the drawings, two of the pieces classified by the author (Martins, 2005, 418-423; figs 49-56, 500-507) as Dressel 20 amphorae, in fact resemble more a fragment of Dressel 14 Lusitanian (number of register 9.2.7) and a fragment of a mortaria from Baetica (number of register 9.2.8). With the exception, perhaps, of the basal fragment with the register number 9.2.15, the bases classified as Dressel 20 should be equally attributed to other shapes.

226

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

Serra de Santa Justa, in Valongo, benefited from the proximity of the Douro hydrographical basin. This river, as Strabo reports (III, 3, 4), was navigable for large boats for a distance of about eight hundred stadia (corresponding to about 25 leagues, a little less than 150 km). The 16 km that separate Serra de Santa Justa from Oporto are actually negligible if we take into consideration that the products could circulate in vessels of small gauge in one of the Douro’s tributaries by Serra de Santa Justa, the Sousa River. Complementing or overlapping the transport of products by waterways (and in the opposite direction the flow of ore), we can also identify two fundamental road axes: the first in a northsoutheast direction, connecting this area to Bracara Augusta through via XVI, and the second connecting the coast to the interior.

amphorae (“Lusitanian 2”). Once again, just like the sample from Quinta da Ivanta, the shapes that morphologically fit the Ist century and beginnings of the IInd century (in particular the first half of the Ist century) are predominant, as is suggested by the rim fragments from the Haltern 70 and Dressel 20 amphorae, equally well represented in this sample. Apart from the formal aspects of the amphorae we have just analysed, the set recovered in Quinta da Ivanta poses other questions that we consider interesting. One of the questions has to do with the consumption profile of this place, which, being a mining area, registers some specificities. Among them we should again stress the significant presence of Dressel 20 amphorae, which, as containers for oil, were fundamental both for diet and as fuel for lighting. The remaining amphorae are present in similar conditions to other Roman places of consumption situated on the coast or served by navigable waters, and thus do not correspond to a habit or specific consumption profile. We can notice a more monotonous type of supply that is only represented by peninsular Baetic and Lusitanian products. In this situation are the Baetic Haltern 70 amphorae, which, as the tituli picti show, were used to transport a diversity of goods, among which the derivative of wine products (defrutum, sapa), oil products (olivae), and curing products (muria) (vd. Aguilera, 2003, 119-120; Carreras, 2003, 117), and the amphorae of Baetic and Lusitanian origin that transported fish products, which were very important in the diet because of the large amount of salt they contained.

But, as one can imagine, the proximity of Oporto was one of the most important factors in the supply of this mining area of Quinta da Ivanta. As we had assumed and reported in different studies (Morais and Carreras Monfort, 2003, 93-112; Morais, 2005; Morais, forthcoming a; Morais, forthcoming b), the Cale, situated on the knoll of Oporto’s Cathedral, also called Penaventosa, was one of the most important anchorages in the NW Peninsular. As proof we have a vast assemblage of amphorae found in a large landfill from the mid XVIIIth century, from the time of the building of the old Aljube of Oporto, and which has turned out to be one of the most important deposits of amphorae so far documented in Portuguese territory (Morais, forthcoming b).

Another question directly related to the first has to do with the transport and trade circuits of these products in the supply context of this mining environment. In this case the question does not pose many difficulties. As a matter of fact, if we pay close attention to the analysis of a map of the road network of the NW, we can easily conclude that this place, situated on the northern slope of

All in all, we would say that the identification and systematic excavation of mining settlements could corroborate the issue we have raised here: the existence of a hybrid supply market, in the control of private interests, together with a true annona system of state control exemplified by the trade of Dressel 20 amphorae.

227

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 4.5.6 Amphorae and sigillata from Quinta da Ivanta

228

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

4.6 RÍAS BAIXAS AND VIGO (VICUS ELENI) A. Fernández route – known and used from the second millennium BC – by Gaditan merchants in the VIth century BC. It does not seem that on this occasion the Gaditan merchant’s aims were merely to obtain tin and other metals, but to integrate this area in the Phoenician Mediterranean trade circles based on Cádiz and Carthage (Morais, 2005, 50).

INTRODUCTION The numerous rescue archaeology interventions carried out in the Rias Baixas, especially in the city of Vigo, over the past few years has produced an enormous amount of finds, mainly fine wares and late imported amphorae. However, the number of publications dealing with this material – especially amphorae and coarse wares – has barely reached double figures since J. Naveiro’s comprehensive work on the north-western trade in 1991 (Naveiro, 1991). Therefore, the publication of several papers, such as those on the materials found on recently excavated castros, such as O Neixón or Montealegre (Aboal, Castro (coords.), 2006; Ayán Vila (coord.), 2006; Ayán Vila (coord.), 2008; González Ruibal, 2006-2007; González Ruibal et al. 2007), and on the imported fine wares found in Vigo and the surrounding villae (Fernández, 2007; Fernández et al. 2007; Fernández et al. 2008; Fernández, Lago, 2007; Fernández, Soto Arias, 2007), may be counted as exceptional.

LATE REPUBLICAN PERIOD The evidence showing late republican trade links prior to the actual Roman presence, specifically amphorae, fine wares, imported coarse wares and prestige and luxury goods, is restricted to the pre-roman settlements in castrexos and to their surrounding areas – natural harbours – especially on both shores of the Ría of Vigo – Castro de Vigo, Punta do Muíño do Vento, Castro de Toralla, Montealegre, etc. – on the mouth and hinterland of other Rías – Castro de O Neixón, Baroña, A Lanzada, Alobre, Torres de Oeste – and around the mouth of the river Miño – Santa Trega, A Forca, Fozara or Troña (Naveiro, 1991).

Our aim in the present paper is to approach trade on a specific coastal area, the Rías Baixas, from the IIIrd century BC to the VIIth century AD, on the basis of amphoric evidence (provenance, chronologies, consumption patterns, etc.). The suggested geographical frame (the Rías Baixas) is located on the SE of modern Galicia (figure 4.6.1). This region’s coastline, which extends from the mouth of the river Miño to the cape of Finisterre, Naveiro’s zones VII and VIII (Naveiro, 1991, Map 2), is “broken” by four major bays (Rías) – Vigo, Pontevedra, Arousa and Muros/Noia – and substantially penetrates inland along river valleys, especially the Miño and the Ulla valleys.

Italy Late republican amphorae from both of the Italian coasts, the Tyrrhenian and the Adriatic, are present in the Rías Baixas. The Tyrrhenian Dressel 1 type is attested in its three suggested variants (A-C), although the B and C types are predominant (Naveiro, 1991, 66). Dressel 1A type production begins in the mid IInd century BC, as a direct evolution of the so called Greco-italic amphorae, spanning until the mid Ist century BC. In the northwest, they usually appear in Ist century BC contexts (Ibid., 66), in association with Late Phoenician amphorae and Campanian wares. Lamboglia’s type A is only represented in Castro de Troña (Ibid., Fig. 14), Montealegre (González Ruibal et al. 2007, Fig. 14) and Torres de Oeste (Naveiro, 2004, 76), whereas the examples of Dressel 1 of the B and C types are much more common, being present in Castro de Montealegre,

Due to this coastal morphology, the castros located on the Rías Baixas enjoy excellent conditions – natural harbours – for maritime trade, allowing for their long-lived implication in commercial relationships and explaining their wealth in imported goods. Many of these pre-roman settlements function as true commercial emporia form an early date, especially since the re-opening of the Atlantic 229

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 4.6.1 Location of the Rias Baixas and Vigo (Vicus Eleni) in the Iberian Peninsula although some of these pieces could be of Baetican provenance (González Ruibal et al., 2007, Fig. 14), O Facho de Donón (Ibid., 55), Santa Trega (Peña Santos, 1986), Torres do Oeste (Naveiro, 2004), O Neixón (Ayán Vila, coord., 2008, 129), Castro de Vigo (Hidalgo, Viñas Cue, 1992-93, Fig. 9) and, also in Vigo, the beach sand levels in O Area l site, specifically in the number 5 Hospital Street (Acuña Piñeiro, 1996, 35) and the Parcela 14 excavations (Castro Carrera, 2007, 357), as well as in underwater finds from the Ría (figure 4.6.2. 1).

unreservedly, in the Santa Trega and Alobre castro sites (Naveiro, 1991, Fig. 14) and, in Vigo, on the excavations at number 5 Hospital Street (Acuña Piñeiro, 1996, 35) and at the Beiramar tunnel – of the Lamboglia 2 type (Cortegoso, 2004, 31). The most common type, Lamboglia 2, is produced from the mid IInd century to the mid Ist century BC, being replaced by the Dressel 6A type (García Vargas, 2007, 323).

The presence of Dressel 2-4 amphorae types, of Italian provenance and late republican date, is almost negligible, for the majority of these containers found in the northwest are of local production of early imperial date. The known examples from the Rías Baixas are an individual Dressel 2-4 type amphora found in the 2003 excavation at Castro de Neixón (Rodríguez Martínez, 2005, 155) and the castro sites of Santa Trega and Troña examples mentioned by Naveiro (Naveiro, 1991, 66). Dressel 2-4 type examples are also documented in Iria (Pérez Losada, 2002, 95, based on evidence provided by J. Naveiro), although the associated material – Haltern 70, Dressel 2-4 from Tarraco, and Lusitan Dressel 20 and 14 types – leads us to date them to the early imperial period.

Amphora manufacture in the Baetica during the late republican period focused on two major areas, the Guadalquivir valley and the Atlantic coast, as far as the Gibraltar strait, including an important centre in the bay of Cádiz (ibid., 2007). The amphorae of “Turdetan” tradition of the Pellicer D type coming from the Guadalquivir valley are first documented in the northwest in Castro de Montealegre (González-Ruibal et al., 2007, Fig. 9). The production of these containers begins in the IIIrd century BC and, although we have labelled them as coming from the Guadalquivir valley, they are also produced on the Gaditan coast.

The amphorae originating from the Adriatic coast are much less common. Although we are aware that some of the known examples could have a northern Adriatic provenance, we have decided to label these containers, similar to the Dressel 1 type but for the oval belly, as originating in Apuleya or Brindisi – including the Lamboglia 2 types and others. Adriatic amphorae of late republican date have been found, although not

The main amphora type originating in this area is the Haltern 70 type, traditionally a wine container. It seems that these amphorae are produced from the second quarter of the Ist century BC, although most of the examples found in the northwest are dated on the early decades of the Ist century AD (Naveiro, 1991, 66). Of the older Haltern 70 type, examples have been recovered in Castro de Montealegre (González-Ruibal et al., 2007, Fig. 15),

Baetica

230

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

Figure 4.6.2 1: Dressel 1C, underwater find (Photo: Archive of the Municipal Museum Quiñones de León, Vigo); 2, 4 and 5: Lomba do Canho 67, Beiramar tunnel, Vigo (Cortegoso, 2004); 3: LC 67 Castro de Montealegre (González Ruibal et al., 2007)

Santa Trega (Peña Santos, 1986) and in the excavation at the Beiramar tunnel in Vigo (Cortegoso, 2004, 31), in all three cases in association with another wine container also manufactured in the Guadalquivir valley: the Lomba do Canho 67 amphora type (figure 4.6.2, 2-5).

type examples found on the Rías Baixas should all be dated in the third quarter of the Ist century BC, as shown by the stratigraphic evidence from Castro de Montealegre (González-Ruibal et al., 2007, 56). The containers for salted fish products of the late Phoenician Mañá-Pascual A4/T-12.1.1.0 type – direct descendants of the old MPA4/T-11.2.1.0 type (Saez, 2008, 641) and probably present at some coastal castros

This container, also produced in the Baetican coast and Morocco, can be dated between 75 BC and the Augustean period (García Vargas, 2007, 325), although the LC 67

231

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 4.6.3 1 and 2: T-7 of Beiramar tunnel, Vigo (Cortegoso, 2004); 3 and 4: T-7 of Castro de Montealegre (González-Ruibal et al., 2007)

number 5 Hospital Street (Acuña Piñeiro, 1996, 35) excavations.

on the Rías Baixas such as A Lanzada, Montealegre, or Punta do Muíño do Vento since the Vth century BC (González-Ruibal et al., 2007, 50) – originate in the Atlantic coast and the Gibraltar strait. The chronology of the Ramón-S12 group, of Gaditan production, concentrates in the IIIrd and IInd centuries BC, although the shape seems to have developed in the IVth century BC (Sáez, 2008, 640). The T-12.1.1.1. and the hybrid .½ and .2 types have been found on the castro sites of A Lanzada, A Forca and Troña (Naveiro, 1991, 69), and during the 2003, 2005 and 2006 excavation campaigns at O Neixón (Ayán Vila, coord., 2008).

Some mention should be made of the long-standing tradition for imitation of Italian containers in the Gibraltar strait area. Only two examples of this sort of Baetican ware can be attested in the Rías Baixas: an example of the Dressel 7-11 type found at O Neixón during the 2004 campaign that can be dated in the mid Ist century BC (González-Ruibal, 2005, 222) and one of the Dressel 1C type, from Montealegre (supra), which could be an early imitation originating in the Rinconcillo pottery workshop, in Algeciras (García Vargas, 2008, 326).

The other late Phoenician container originating in the Gibraltar strait area found in the Rías Baixas is the amphora traditionally known as Mañá C2a (T-7.4.3.2/ T-7.4.3.3) type. It is the last of the shapes of Phoenician tradition manufactured in the Gibraltar strait area, with a chronology spanning from the mid IInd century BC to, in principle, the Augustean period (Sáez, 2008, 647). A double use for this container – salted products and wine – may not be ruled out (Ibid., 647). It is the most common Gaditan late republican amphora type in the Rías Baixas, having been found in castro sites such as A Lanzada, Santa Trega (Naveiro, 1991, 69), O Neixón, during the 2004 campaign, Alobre, Punta do Muíño do Vento (González-Ruibal, 2005, 222), Montealegre (González-Ruibal et al. 2007, 53) (figure 4.6.3. 3-4) and in O Areal site in Vigo, on the Beiramar tunnel (Cortegoso, 2004, 31) (figure 4.6.3. 1-2) and the

Early Empire The pacification of the northwest by Augustus completed the conquest of Hispania and confirmed the Roman domination of these regions, although the Roman influence had been increasing here since Croesus’ (96-94 BC) and Caesar’s (61-60 BC) military expeditions. The new political situation – creation of a road system and foundation of administrative centres – helped to alleviate the clear dichotomy between the coast and the hinterland so evident in the previous period, in which imported goods were limited to coastal and river-side castro settlements (Naveiro, Pérez Losada, 1992, 69). Regardless of these new conditions the coast, especially the Rías Baixas, remained instrumental for the reception 232

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

Figure 4.6.4 1: Oberaden 83 of Beiramar tunnel, Vigo; 2: Haltern 70 of Beiramar tunnel, Vigo (Cortegoso, 2004) conventus and an urban centre for consumption and redistribution, the Dressel 20 type only amounts to approximately 1.99% of the total of amphorae dated to the early Empire (Morais, 2005, Fig. 26). In the Rías Baixas its presence is attested in the castro sites of Troña, Santa Trega and Berga (Naveiro, 1991, 69), on the excavations carried out on the preserves factory found in Marqués de Valladares Street (Torres et al., 2007, 482), on the Beiramar tunnel (Cortegoso, 2004) in Vigo, and in Iria (Pérez Losada, 2002, based on information provided by J. Naveiro). This last excavation also produced one Oberaden 83 type (figure 4.6.4 1) olive oil amphora, direct predecessor of the Dressel 20 type, chronologically falling within the Augustean period (García Vargas, Bernal, 2008).

of imported goods in the northwest during the early Empire. The castro settlements on the Rías were now joined by “semi-urban” settlements with a secondary agglomerated plan – Vicus Eleni (Vigo), Tude (Tui), Forum Iriensis (Iria), Aquis Celenis (Caldas de Reis) or Turoqua (Pontevedra) – (Pérez Losada, 2002). Baetica Once again, the Baetica can be divided into two major productive sub-areas: the Guadalquivir valley and the Atlantic coast, as far as the Gibraltar strait, including the major production centre in the bay of Cádiz. During the early decades of the Ist century AD the Haltern 70 type amphorae, produced in the Guadalquivir valley from the late Republic, reached the north-western coast in massive numbers. This type of container is present in almost every castro and new settlement on the coast and in the hinterland from approximately the late Ist century BC to early Ist century AD (figure 4.6.4 2). Furthermore, numbering in the thousands, it is the most common amphora type in the northwest, and the concentration of its finds (NMI) in this area is greater than in any other in the Roman world (Morais, Carreras, 2005). Its exclusive use as a container for Baetican wine is currently under discussion, for the tituli picti mention grape-derived products, such as the mulsum, the sapa or the defrutum (García Vargas, 2007, 329).

The containers for salted products of the Dressel 7-11 (Ist century AD), Beltrán IIA (Ist century AD) and Beltrán IIB (Ist-IInd centuries AD) types, along with those produced on the pottery workshop of Venta del Carmen, in Algeciras (VC1), all originate in the Atlantic coast and the Gibraltar strait area (García Vargas, Bernal, 2008). Of these types, only the first three are attested in the Rías Baixas. The oldest of these, the Dressel 7-11 type (Beltrán I), which amalgamates the Dressel 7, 8, 9, 10 and 11 types, is present in Castro de Montealegre, Santa Trega, A Lanzada, on the 2004 campaign in O Neixón – although the excavators considered thus finds as late republican in date – (González Ruibal, 2005, 222), in Torres de Oeste (Naveiro, 1995), Iria (Pérez Losada, 2002, 95, based on information provided by J. Naveiro) and in Vigo, both in the castro and in the preserves factory in Marqués de Valladares Street (Torres et al., 2007, 482). A complete example of an amphora belonging to this group, specifically of the Dressel 10

The olive oil amphorae of the Dressel 20 type are still very rarely found, reinforcing the traditional notion, based on textual evidence (Strabo, III, 3, 7), that animal fat was used in the northwest as a substitute (Naveiro, 1991, 67). Even in Bracara Augusta, the head of a 233

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Ría of Vigo (Naveiro, 1991, Fig. 15; Viñas Cue, 2004, 57).

type, could also be recovered from the sea on the area around A Guarda (Viñas Cue, 2004, 58). The Beltrán IIA type, derived from the former types during the Augustean period (García Vargas, 2007, 331), is only documented in the castro site of Santa Trega (Naveiro, 1991, Fig. 15). The later Beltrán IIB amphora type offers a similar case, for it is only attested in the Rías Baixas in Santa Trega (Naveiro, 1991, Fig. 15) and possibly in Castro de O Neixón (González-Ruibal, 2005, 222). The scarce presence of these types of containers in the northwest is confirmed by analysis of the amphorae found on Bracara, from which barely one fragment/ individual of the Beltrán IIA and eight of the Beltrán IIB types have been recovered (Morais, 2005, 109).

LATE ROMAN PERIOD Since J. Naveiro’s work on ancient north-western trade in 1991 (Naveiro, 1991), nothing, other than the occasional brief “mention” in some reports, has been published about amphorae in late contexts in Galicia. The scarce evidence that we use here is based on Naveiro and on our own personal impressions, gathered during our study of fine wares in the Rías Baixas during the late Empire. During the late Empire, the castro settlements seem to gradually loose importance – many of them are abandoned – against a new sort of settlement of the villa type, which spread across the coast, especially in the Ría of Vigo (Pérez Losada, Fernández y Vieito, 2008, 481506). These settlements, many of which acquired an industrial nature, were to be the new recipients of goods imported through maritime trade, along with the semiurban settlements of the secondary agglomerate type.

Tarraconense The amphora types known as Tarraconenses or Layetan, of the Pascual 1 and Dressel 2-3 types, also called Dressel 2-4, originate in the area around Tarraco (López Mullor, Martin, 2008). The Pascual 1 amphora type was manufactured in many pottery workshops around Llanfranc and Benifallet from 40 BC to 60/70 AD, peaking between 20 BC and 20 AD (ibid., 698). In the Rías Baixas, this type is represented in Castro de Montalegre (González Ruibal et al, 2007, 63) and in the excavation at the Beiramar tunnel (Cortegoso, 2004, 31). None of these two examples are supported by any stratigraphic evidence to inform us as to the moment in which the Layetan wine arrived in the northwest.

IVTH CENTURY In the villa of Toralla (Vigo), where we have conducted a study dealing with fine wares from the early IVth to the early Vth century AD (Fernández et al., 2008, 575-585), the few recovered amphorae fall into de Lusitan Almagro 50C/Keay 16 (figure 4.6.5. 1) type and the Cylindrical/ African II type. Apart from this imported material is one white pasted Almagro 50C, which we have identified as a IVth century AD. Regional II type from the Bueu pottery workshop, was also found.

Another individual Layetan wine amphora of the Dressel 2-3 or 2-4 type groups (González-Ruibal et al. 2007, Fig. 27), probably of the Dressel 2A type which peaks in the mid Ist century AD (López Mullor, Martin, 2008, 705), was found in Castro de Montealegre. This amphora type has also been documented in the secondary agglomerate in Iria, in association with examples of the Italian Dressel 2-4 type and other Ist century AD containers (Pérez Losada, 2002, 95, based on information provided by J. Naveiro).

From the beginning of the Vth century AD, the villae seem to have gone through a process of collapse around the Rías Baixas, being abandoned in favour of the raised settlements, which became occupied once again, and especially of the commercial centres, such as Vigo which seems to have acquired a new category as a harbour for the reception and redistribution of African and eastern goods.

Lusitania Traditionally, it was accepted that Lusitan pottery workshops became active during Claudius reign, in the prime of the imperial Atlantic policy (Fabiao, 2005, 84). Recent studies on Lusitan amphorae have, however, brought to light the existence – in many cases in castro sites in the Rías Baixas – of oval containers similar to the Haltern 70 and Dressel 7 types, of Augustean dates (Morais, Fabiao, 2007). These containers for salted products have been found in Castro de Santa Trega, Monte Castro (Panxón) and Castro de Vigo (ibid., Fig. 1, 5 and 6). The manufacture of the Dressel 14/Beltrán IV/Lusitana 2 type, also a preserves container, became generalized in the Tajo and Sado valleys in the second half of the Ist century AD (Fabiao, 2008). The finds of this kind of amphora once again focus on the Rías Baixas area, for example in the cape of Mar and in underwater finds around the Island of Torralla, both in the

The east With eastern amphorae we refer to all those containers manufactured in the eastern Mediterranean in Late Antiquity. Known as Late Roman Amphorae (LRA), we will use Riley’s standard international typology. In Vigo, the LRA 1, 2, 3 and 4 types have been found so far. The LRA 1 type, and it’s variants, is easily the most common amphora type in O Areal site in Vigo. This wine container was manufactured from the IVth to the VIIth centuries AD in many workshops in Cyprus, the north of Syria and the southern coast of Turkey. It’s presence in Vigo – from the mid Vth century AD (Fernández, 2009, forthcoming) – must be associated with that of fine wares – LRC and LRD. Therefore, the earliest variant (LRA 234

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

Figure 4.6.5 1: Almagro 50C of the villa of Toralla; 2: LRA 1B of Vigo (Photos: Archive of the Municipal Museum Quiñones de Leon, Vigo)

1A) is absent whereas the LRA 1B, characteristic of the VIth and VIIth centuries AD, is the most common, for example in the excavations at UARC II, Parcela 14, Marqués de Valladares or number 5 Hospital Street (figure 4.6.5. 2). Specifically, in the UARC II excavation, a complete 1B type has been recovered in association with a Bonifay’s C5 type lamp (Bonifay, 2005) dated to the second half of the VIth century.

Africa Although much less common than the eastern, the African amphorae are present in Vigo in all late periods, from the IVth century AD – villa of Torrala – to the VIIth century AD. In some excavations, such as those carried out in number 5 Hospital Street, Parcela 13 and Parcela 19 – all dealing with the O Areal necropolis site – these African cylindrical amphorae were found to be used as tombs, making their identification more difficult. The late fine wares found in the number 5 Hospital Street excavation suggest a date between the Vth and the VIth centuries AD, and therefore the amphorae must belong to the Keay 35, 57, 55 or 62 types, typical of this period. The later levels in UARC II, dated to the VIth and the first half of the VIIth centuries AD on the basis of the fine wares, are rich in examples of the LRA 1B type, in association with the African Keay 62 and Keay 61 types.

The wine amphorae of the LRA 2 type and its variants, produced in the Argolid and in some areas between Rhodes and Cyprus from the mid Vth to the VIIth centuries AD, are less common in Vigo. Variants of the LRA 2A and 2B types are documented in UARC II, and only the former in Marqués de Valladares and number 5 Hospital Street. The LRA 3 is a fusiform amphora originating in Lydia, in the region between Ephesus and Sardis. Its fabric is rather micacious, brown in colour and very soft. In Vigo, its single-handled version – characteristic of the IVth century AD – has been documented in Marqués de Valladares, but its pointed base and small capacity indicate a late Vth century date. The LRA 3 type has also been found on an excavation carried out in Colon Street in 2000.

SUMMING UP According to the available evidence, imports in the Rías Baixas area during the late republican period are limited to wine and salted products. From the IInd century BC, the wine consumed in the pre-roman settlements is of Italian origin, mainly from the Tyrrhenian coast, although some wine produced in the Adriatic coast was also imported. The presence of Baetican wine containers from the mid Ist century BC – early Lomba do Canho 67 and Haltern 70 types – reveals how the Baetican wine began gradually displacing the Italian wines from the Rías Baixas’ markets from an early date, until becoming

The last amphora type attested so far in Vigo is the LRA 4 type, or “Gaza Amphora” due to its provenance (Palestine), production of which spans from the IVth (LRA 4A) to the late VIIth centuries AD. In Vigo, fragments from the rim have been identified in the UARC II, in Marqués de Valladares, in association with late contexts, and in the necropolis at number 5 Hospital Street. 235

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Amphorae belonging to the straight bottomed types Dressel 28, Gauloise 3, 4 and 5 and Urceus – containing Gaul and Baetican wine – the Dressel 2-4 type – also Gaul and Baetican – or the eastern types (Rhodes, Crete), are remarkably absent during the post-Augustean period, but all of them are documented in the capital, Bracara Augusta (Morais, 2005). The reason for this absence may lie in the high quality of the wine that they typically contained, causing its demand, similarly to that for olive oil, to be restricted to high rank social elites inhabiting administrative centres.

virtually the only one consumed, apart from some Tarraconense produce, during the early Empire. The finding of Vth century BC. Phoenician amphorae proves that preserves were being consumed in the northwest from an early period. During the Republic, the fish products consumed came exclusively from the Gaditan coast, as shown by the T-12 amphora type in the IIIrd and IInd centuries BC, and the T-7 type for the IInd and Ist centuries BC. The absence hitherto of other Gaditan salted products containers, such as the T-8 and T-9 types, mass production of which lasted until the last stages of manufacture of this family of containers of Phoenician tradition, is therefore remarkable.

Unfortunately, our evidence for the IIIrd and IVth centuries AD is very limited, but these centuries should be dominated by containers for salted products from the Lusitania– Almagro 51A, B and C and Almagro 50 types – the Baetica – Beltrán 72 type – and Africa – the large African I-II and Tripolitanian III types – along with the local/regional Bueu produces – Regional I and II types – as shown by the finds in Braga (Morais, 2005). The only available evidence for the IIIrd and the IVth centuries AD in the Rías Baixas comes from the villa of Torralla, and indicates that the Baetican preserves had been substituted by fish preserves from Lusitania, Africa and the local factories (Bueu). We don’t know about the sort of wine consumed during this period, although we may assume that, as is the case in other places such as Hispalis (García Vargas, 2008) the Spanish wines disappear from the markets from the late IInd century AD, being partially replaced by wine from Gaul in the IIIrd century and African wine in the IVth and part of the Vth centuries, when the latter is joined by eastern wines until the cessation of imports during the VIIth century AD. Olive oil presents a similar case, with no known imports into the Rías Baixas after the IInd century AD until the arrival, in the early IVth century AD, of the first African amphorae which, in some cases, could transport olive oil.

The deep changes introduced in the northwest since the Roman conquest (vid supra) also affected the consumption of imported goods. During the early imperial period, wine is the most widely imported product, well above salted products and olive oil. During the Empire, the Italian wine is replaced – the Dressel 2-4 amphora type is very rarely found – with Baetican wine and its derivates, transported in Haltern 70 type amphorae. We have already pointed out the enormous number of individual finds in the northwest as mentioned by Morais, especially in castro sites in the Rías Baixas (Morais, Carreras, 2005). Layetan wine, transported in Pascual 1 and Dressel 2-3 types amphorae has also been documented. This wine would have been of a better quality, probably being brought on demand by the newly arrived Latin population or the local elites. During the Ist century AD, substantial amounts of salted products from the Bay of Cádiz were still arriving at the Rías Baixas, although now they are transported in containers of Italian tradition, as the Dressel 7-11 type. Even though we lack any precise quantitative study, we can tell that the number of Gaditan containers for salted products decreases progressively during the Ist century AD, becoming almost negligible in the IInd century AD, as shown by the few individuals of the Beltrán IIB type, typical of this century, that have been recovered. This situation may be due to two factors:

From the late IIIrd century, the Rías Baixas begin receiving imports of sigillata and amphorae from north and central Tunisia, and from the mid Vth century, oriental products such as LRC and LRD fine wares, LRU perfume containers, and amphorae of the LRA 1, 2, 3 and 4 types. Although we still lack any comprehensive study dealing with the amphorae found in Vigo we can propose that the number of eastern containers is very high, significantly surpassing the number of African examples, as is the case with fine wares, for the first time in the Roman west, the LRC ranging above the ARS in NMI (Fernández, 2009, forthcoming). Most of these amphorae of oriental origins were wine containers, although we cannot rule out the possibility that the smaller ones, such as some of the LRA 1 and especially some of the LRA3 types, transported wines for specific purposes – liturgical wine – or perfumes. While wine arriving at the Rías Baixas came from the east from the Vth to the VIIth centuries AD, imports from Africa included salted products and olive oil. For the moment, we ignore the role played by other manufacturing areas – Baetica, Lusitania, the Aegean, etc. – in the supply of the Rías Baixas during this later period.

The progressive introduction of Lusitan preserves – Lusitan oval shapes – from the early Augustean period and, more clearly so, from the mid Ist century AD. The emergence in the Rías Baixas of a number of fish preserves factories – seven in the Rías Baixas region, four of which are located in the Ría of Vigo (Currás, 2007, Fig. 1) – which would supply the local/regional markets during this period and the following centuries. The small number of olive oil amphorae found in the Rías Baixas proves that olive oil was not a product on demand by the majority of the population. Nevertheless, the Baetican olive oil, transported in Oberaden 83 type amphorae, is represented since the Augustean period. The olive oil was, similarly to the Layetan wine, demanded by the army, the recently arrived population in charge of organizing the new territory and the local aristocratic elites in an early trend towards Romanisation.

236

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

Figure 4.6.6 Amphorae from the Rias Baixas. 1. Dress. 1A; 2. Dress. 1B-C; 3. Dress. 2-4; 4. Lamb. 2; 5. Amphora of Apuleya or Brindisi; 6. Pellicer D; 7. Haltern 70; 8. LC 67; 9. T-12; 10. T-7; 11 y 12: Dress. 7-11; 13. Haltern 70. 14. Oberaden 83; 15. Dress. 20; 16 y 17. Dress. 7-11; 18. Pascual 1; 19. Dress. 2-3; 20. Dress. 14/Beltrán I; 21. Beltrán IIA; 22. Beltrán IIB; 23. Almagro 50C; 24. Regional II; 25. Africana II; 26. Keay 35; 27. Keay 62; 28. Keay 61; 29. LRA 3; 30. LRA 4; 31. LRA 1B; 32. LRA 2

237

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

4.7 ASTORGA AND LEÓN C. Carreras Padre Blanco 7-11 (AA/PB 7-11/86), Oscar Grau (AA/OGR/92-3), P. Castro (AA/PC7/91), La Cruz (AA/LC 20-24/92), Cosamai (AA/OCC/82), Calvo Sotelo (AA/CS10/90), Modesto Lafuente (AA/ML3/91), Porteria 12 (AA/POR12/89), SantoCildes 5 (AA/PS5/88), Rodríguez Cela (AA/RC5/86), El Mesón (AA/M9/91), José de Maya 3 (AA/JM3/90), Plaza Romana (AA/PR15/91), Macias 7 (AA/MM7/93), San Javier (AA/SJ), E. Castro (AA/IEC/92), Blanco Cela 21, and Santiago Crespo.1 Although the whole settlement covers an area of 16 ha, this small sample of more than 6,000 m2 is representative enough to allow us to draw some conclusions.

In 1995 and 1996 we started the study of amphorae assemblages in the NW of Spain. Our first contact with this region was an initial analysis of the amphorae from Campa Torres (Gijón), where the assemblage included vessels from the Republican period, basically Dressel 1, and Augustan period, such as Oberaden 83 – early Dressel 20 and Haltern 70 (Carreras, 2001). Of interest was the important percentage of Haltern 70s found at Campa Torres, which confirmed some patterns already pointed out by Naveiro (1991) in his survey of the Galician coast. Naveiro (1991) identified Haltern 70 as the most remarkable amphora import at the maritime Roman sites of Gallaecia, reaching high percentages in any assemblage. Such trends have been confirmed since then, and any site in Galicia records Haltern 70 as its main amphora import, as the recent work of Pérez Losada (2002) demonstrates.

The results from this representative sample were later compared to nearby sites such as Campa Torres, Legio (León), Petavonium (Rosinos de Vidriales), Tiermes, Arxa and Uxama Argaela (Osma). At first sight, all the assemblages from NW sites present common traits in both amphora typologies and quantities. There are few amphorae found inland, since such containers were better suited for water and maritime transport. Legio and Asturica were the sites with a wider variety of amphorae as well as larger amounts. Of course, Asturica was a referent point in NW Spain with more than 12 different fabrics documented and 25 diverse forms (see figure below) and amphora density 5,327 cg/m2.

However, it was thought at that time that the most interesting amphorae assemblages should have been in the military areas of the NW. That is why we decided to study the amphorae from Asturica Augusta (Astorga) and Legio (León). Initially it was believed that the assemblage from these two military sites from the NW Spain would be similar to other military camps from the German and British limes. On the contrary, the amphora assemblage from both sites was completely different from other military settlements and revealed a more regional pattern that was similar to the findings of Naveiro (1991) in Galicia.

Amphorae from Asturica were quantified by number of sherds, weight, estimated vessel equivalent (EVE) and rim percentage, handles and spikes (Carreras, 2000a, 45-70). The results of these quantifications appear in the table below, which includes 12 fabrics, and the distinctive typologies within these fabrics.

The amphorae studied in 1995 and 1996 corresponded to more than 29 rescue excavations from the city of Astorga, which unearthed an approximate area of 6,704 m2, and were published in 2003 (Carreras and Piero, 2003). The excavation studied includes Manuel Gullón 3-5 (AA/MG 3-5), General Mola 8-10 (AA/GM 8-10), Cepedana (AA/80/ESC), Bastión (AA/BA/85), L. Pelaez (AA/LPS/ 90), Garrote (AA/AG4/91), Autobuses (AA/EA/86), Altamira (AA/CA 8-10/90), Peñicas (AA/PI/90), Puerta Obispo (AA/POB/89), General Mola 3-5 (AA/GM 3-5), Puerta Romana (AA/PVR), Padre Blanco 5 (AA/PB5/85),

Only the Duero valley fabric may be a particular regional ceramic unknown from other sites. It is a series of sherds red-orange, dark red or grey-orange in colour, which 1 It is supposed that most late dumping areas were located outside the city walls, such as the excavations off C/Bastión and “Las Lolas” demonstrate (García Marcos and Vidal, 1996).

239

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Table 4.7.1 Amphora quantities from Asturica Augusta (Carreras and Berni, 2003) Typology

Sherds

Weight (gr.)

EVE

Density

AFRICA

27

2080

32

34

Tripolitana 1

2

220

35

3

GUADALQUIVIR

6

990

-

16

378

37330

343

622

2

400

23

6

Haltern 70 Oberaden 83 (Dr20) Dressel 20

266

22550

12

375

Dressel 20 parva

4

860

100

14

Dressel 23

1

150

-

2

Dressel 2-4

26

6790

33

113

Dressel 7-11

1

750

100

12

Astorga 44

1

310

25

5

BAETICAN COAST

975

115360

178

1922

Dressel 7-11

25

3640

293

60

Beltrán II-A

1

440

7

7

Beltrán II-B

4

900

8

15

CAMPANIA

15

2080

-

34

Dressel 2-4

225

28730

220

478

Anfora “a bastone”

12

2260

-

37

GALLIA

30

2460

-

41

RHODIA

253

21330

108

355

RICHBOROUGH 527

18

1588

45

26

P&W 66

1

30

-

1

EASTERN MED.

73

4950

-

82

Agora G-199

9

600

-

10

Dressel 2-4

240

20870

220

347

Pseudokoan

2

370

25

6

LATE ROMAN

4

470

-

7

Late Roman 1

1

250

-

4

TARRACONENSE

11

4170

-

69

Pascual 1

83

12430

176

207

Dressel 2-4

46

6470

38

107

DUERO VALLEY

9

1960

20

2

Unknown

220

15855

124

264

TOTAL

2971

319643

2093

4770

appear to be sometimes over-fired. The fabric breaks easily and the microscope reveals quartz, limestone and mica. Parallels turn up at other sites in the region such as Léon (Legio) and Rosinos de Vidriales (Petavonium). One sample (500403) studied by Aureli Alvarez (Autonoma University of Barcelona) reveals the presence of cyenit, a mineral typical of the Duero valley (Carreras and Berni, 2003).

regarding other sites of similar size is the amount and density of each particular type. The most common amphorae in the Asturica assemblages come from the coast of Baetica (Cádiz), a situation also found at Legio and even other sites such as Barcino or Xanten in Augustus’ time (Carreras, 2007). It must be borne in mind that most assemblages of Asturica are dated to the early periods of the Ist century AD (Julio-Claudian periods).

Apart from this particular regional typology, most amphorae recorded at Asturica are common to most western Roman settlements. Perhaps the main difference

However, the best represented typology in Asturica is Haltern 70 amphora; again it is quite an exceptional case and only comparable to some sites in Lusitania and 240

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

Table 4.7.2 Amphora quantities from Legio – León (unpublished report)

Gallaecia. Therefore the high percentage of Haltern 70 in Asturica appears to reflect an Atlantic circulation, since no other site in the Ebro valley records such high numbers of this Baetican amphora.

Typology

Sherds

Weight (gr.)

Density

129

17070

2574

Oberaden 83 (Dr20)

2

610

92

Dressel 20

58

9680

1460

Verulamium 1908

3

340

51

Dressel 2-4

1

210

31

Dressel 7-11

7

1080

162

Dressel 28

1

100

15

BAETICAN COAST

169

26840

4048

Dressel 7-11

11

2540

383

Dressel 10

5

1160

174

GUADALQUIVIR

The almost complete absence of Gaulish vessels reinforces the idea that supply from Gulf of Biscay was not really significant, due to difficulties in transport infrastructure. As far as amphorae assemblages are concerned, amphorae commodities consumed in Asturica were probably coming from the Atlantic ports of Gallaecia and Lusitania, where Haltern 70s were also predominant. The low densities of amphora from Asturica demonstrate that the city was far away from the main Mediterranean commercial circuits, and also highlights the difficulties in land transport (see De Soto in this volume). However both Legio and Asturica document a wide variety of types that, despite all obstacles, managed to reach such distant markets in relatively good proportions.

Haltern 70

With regards to Dressel 20 amphora, an important percentage is also recorded in Asturica, although it never enjoyed the pre-eminent role of other western military sites from Germania, Britannia and Raetia. Similar is the case of Legio, but not so acute. It must be borne in mind that the presence of oil-lamps in these both sites is remarkable.

Dressel 2-4

45

8650

1304

GALLIA

2

180

27

CAMPANIA

Dressel 2-4

4

1330

200

Gauloise-4

18

2250

339

RHODIA

53

2350

354

Dressel 2-4

176

15740

2374

Pseudokoan

13

1190

179

5

680

102

EASTERN MED.

With regards to Legio (León), the number of excavations studied was relatively low, since there were only a few rescue excavations at that time. Amphora assemblages were studied from only four sites (San Pedro, Damaso Merino, Corral de San Guisán and Pallarés), covering an area of 663 m2 out of 19 ha, that it is to say the whole extent of the Roman site. Therefore the sample is quite small to be generally representative of the trade patterns at this military site, providing a total density of 14,227 cg/m2.

TARRACONENSE Dressel 2-4

Again the Baetican coastal fabric is the best represented in the sample, as happened at Asturica, although Haltern 70 is still the most common amphora type. Similarly to Asturica, there was a wide variety of amphorae from all over the Roman Empire. The most far-distant vessels came from the Eastern Mediterranean (Rhodia, Dressel 2-4, Pseudokoan), containing probably highly-valued wines from those latitudes.

DUERO VALLEY

2

540

81

Unknown

22

1790

269

TOTAL

721

94330

14227

to the Ebro valley, such as Tiermes, Atxa and Uxama Argaela, where Tarraconense imports are dominant. To some extent, all appear to suggest that there were two independent circuits with different suppliers: one related to the Ebro valley and another to the Atlantic coasts.

The main supplier of amphorae to Asturica and Legio was the province of Baetica, something that comes as no surprise if it is considered that such imports are also common in all the provinces around the Atlantic Ocean, such as Aquitania, Lusitania, Belgica, Britannia or Germania. Likewise, the Roman settlements in the NW appear to have consumed mainly olive-oils, wines, fishsauces and olives from Baetica and the coast of Gallaecia (Naveiro, 1991, 66-67) to the main Asturian bases – Campa Torres (Carreras, 2001), conventus Braccarense (Morais, 1998), and even inland (Legio, Petavonium). From the whole assemblage of Astorga 60% in weight and 11 out of 25 amphora types documented are Baetican in origin, which confirms the importance of this supplying province. Such predominance of Baetican imports is not present in sites in the Meseta closer

Amongst the Baetican vessels, the best represented are the fish-sauce amphorae (liquamen, muria, garum) from the coastal strip. Fish-sauces were products of some luxury and obtained from the complex processes of drying, salting and maceration of fish species, such as tuna. The most common types documented are Dressel 7-11, Beltrán II A and II B, produced during the Principate. Dressel 7-11 amphorae from Guadalquivir, and occasionally some Haltern 70, may have comprises all the potential fish-sauce vessels arriving in the NW. The presence of fish-sauces is generalized in the NW, reaching high densities in centres such as Asturica and Legio, which is direct testimony to the purchasing power of these two communities. Although the production of fish-sauces is well known from other sites along the Atlantic and Cantabric coastlines (Lusitania, 241

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

ceramic assemblages in Asturia, which reinforces the idea of difficulties in land transport in the Conventus (Zarzalejos, 2005).

Gallaecia, Asturia, Aquitania), there is no other physical evidence of their consumption in Legio and Asturica. With regards to typologies, the Haltern 70 amphorae produced in the Guadalquivir valley is the best represented, not only in Asturica but also in Legio, apart from Campa Torres and most sites of the Gallaecian coast (Naveiro, 1991, 66) and Bracara Augusta (Morais, 1998, 44-45, 70). It is the only region in the Roman Empire where this amphora type is the best represented. The important percentages of Haltern 70 were initially related to the military campaigns of Augustus, the period of peak imports, although it continued being imported in the NW until the Flavian period. It seems feasible that the Asturian and Cantabrian wars were the cause of the arrival of such imports to the NW sector of the Peninsula. Later, the Mediterranean populations who settled in the region (i.e. soldiers, civil servants, veterans) were probably the main consumers of those products, which probably fetched high prices when they reached Asturica and Legio.

Figure 4.7.1 View of Gijon bay including the site of Campa Torres

Other NW sites produced similar quantities and percentages (Carreras, 2001). For instance Campa Torres (Gijón) also registers a high percentage of Haltern 70 (73%), although the densities are much lower than those recorded in Asturica, Gallaecia and Lusitania.

Another military site whose amphorae have been studied is Petavonium (Rosino de Vidriales), also in the NW of the Peninsula and near Asturica and Legio. Although Haltern 70 amphorae are present in the assemblage, the number is very low, only one sherd, compared to the eastern and fishsauce amphorae that predominate.

Table 4.7.3 Amphora quantities from Campa Torres (Carreras, 2001) Typology

Sherds

Weight (gr.)

Density

Haltern 70

447

10650

208

Oberaden 83 (Dr20)

10

1160

22

Dressel 28

1

90

1

Table 4.7.4 Amphora quantities from Petavonium (umpublished report)

GUADALQUIVIR

Typology

3

300

1

Dressel 2-4

10

775

15

Greco-Italica

1

85

1

RHODIA

1

80

1

Pascual 1

5

710

13

LOCAL

3

410

8

Unknown

8

325

6

TOTAL

489

14585

284

Sherds

Density

Haltern 70

430

1

43

Dressel 20

90

1

9

Dressel 20 picola

430

3

43

BAETICAN COAST

780

4

78

Beltran II-A

70

1

7

CARROT

80

1

8

EASTERN

3220

7

322

60

1

6

RHODIA

100

1

10

LOCAL

1260

4

126

Unknown

110

2

11

TOTAL

6630

26

663

GUADALQUIVIR

BAETICAN COAST Dressel 7-11

Weight (gr.)

CAMPANIA

TARRACONENSE

GAULOISE Gauloise 4

The assemblage from Campa Torres seems to come from two periods: Republican and Augustan. According to the number and densities of amphora recorded at Campa, it cannot be considered a Roman port-of-trade for the supply of inland armies during the Asturian and Cantabric wars. The mountain range of Picos de Europa was an impressive barrier for land transport, and still remains so in present times. Similar conclusions come from the study of other

What can we say of Haltern 70 presence in the sites closer to the Ebro valley? Fortunately some Roman settlements have provided a good sample of its amphora circulation. Tiermes (Soria) is a Celtiberian site, later occupied by Romans, that has recorded a few amphora sherds. The amphora studied from the Casa del Acue242

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

Table 4.7.6 Amphora quantities from Uxama Argaela (umpublished report)

ducto showed a predominance of Eastern Dressel 2-4 amphorae and a small presence of Haltern 70. The following figure documents the amphora assemblage from Tiermes.

Typology

Table 4.7.5 Amphora quantities from Tiermes (umpublished report) Typology

Sherds

Weight (gr.)

Density

1

BAETICAN COAST

1

170

Density

GUADALQUIVIR

1

200

14

Dressel 20

1

100

7

BAETICAN COAST

9

3060

218

Beltran II-A

2

430

30

160

Dressel 2-4

42 40

CAMPANIA Dressel 2-4

Weight (gr.)

CAMPANIA

GUADALQUIVIR Haltern 70

Sherds

1

280

70

EASTERN Dressel 2-4

3

940

235

TARRACONENSE

2

160

40

Unknown

1

40

10

TOTAL

9

1750

437

2

410

29

TARRACONENSE

165

19640

1402

Dressel 2-4

20

3190

227

Gauloise-4

3

250

17

Local

4

560

40

Unknown

5

1050

75

TOTAL

115

28890

2063

The important presence of Tarraconense amphorae instead of Eastern wine containers, and the absence of Haltern 70, reveals a completely different pattern of amphora consumption in this region compared to the NW. Two different circuits of amphorae circulation could therefore be anticipated according to the percentages and the presence and absence of particular amphora types. The following histogram shows the percentages of Haltern 70 and Baetican fish-sauce amphorae from all the abobementioned sites and makes clearly shows the two distinctive circuits.

As can be seen, the amphora densities are much lower eastwards, towards Legio and Asturica, and Tarraconense amphorae dominate all assemblages. This is also the case at another site called Atxa (near Vitoria), with only two sherds of Dressel 2-4 (Tarraconense). Our final example is the site of Uxama Argaela (Soria), which records no Haltern 70 amphorae and the Tarraconense vessels are the best represented in the assemblage.

80 70 60

Haltern 70 Baetican fish-sauce

50 40 30 20 10 0

Figure 4.7.2 Percentages of Haltern 70 and Baetican fish-sauce amphorae in the NW

243

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Figure 4.7.3 Interpolation of Haltern 70 densities in the Iberian Peninsula

vessels with no Haltern 70 (Dressel 7-11, Dressel 12, Dressel 22, Dressel 24 and Pascual 1), as happens at C/La Cadena 23 with 20 amphorae (Dressel 2-4, Dressel 28…) and C/Reconquista with 18 vessels (Dressel 2-4, Dressel 5, Dressel 7-11, Pascual 1). Therefore it appears that Haltern 70 amphorae did not travel much along the Ebro valley.

With regards to the Ebro valley there is an exceptional lack of Haltern 70 evidence. Initially Beltran (1990) identified Celsa (Velilla del Ebro) as the only site with some examples (density of 1 cg/m2), but other examples were found later at El Palao (Alcañiz), in relatively high densities (4405 cg/m2), and Zaragoza. The case of the colony of Caesaraugusta (Zaragoza) is remarkable, since a series of deposits of complete amphorae have been documented that included hardly any Haltern 70. For instance the recent excavation at Plaza de las Tenerías, with a chronology from Augustus to the II century AD (Cebolla et al., 2004), records a deposit of 814 complete amphora but no Haltern 70. The amphorae present are Dressel 1A, Dressel 1B, Lamboglia 2, Dressel 2-4, Dressel 7-11, Dressel 12, Dressel 24, Oberaden 83, Dressel 20 and Pascual 1. This is not the only occurrence: the deposit of Santo Sepulcro contained 70 complete

The previous figure X is the result of an interpolation of amphora densities of Haltern 70 (Idrisi Kilimanjaro), showing its distribution in the Iberian Peninsula. Apart from the high densities in Baetica and along the Lusitanian and Gallaecian coasts, it clearly reveals low percentages in the Ebro valley and most sites in the Meseta. Therefore Haltern 70 supply in Asturica and Legio, as well as other Baetican amphorae is likely to have arrived through the Atlantic ports.

244

THE ROMAN TRADE ON THE ATLANTIC FAÇADE: THE AMPHORAE

4.8 THE HYBRID MARKET SYSTEM AS SUGGESTED BY HALTERN 70 AMPHORAE R. Morais As a result of the analysis of the density map calculated for Haltern 70 amphorae in the Roman Empire (program IDRISI 4.11) – as presented in a recent study (Morais and Carreras Monfort, 2003, 111) – one realises that there is a strong concentration of these amphorae in the Atlantic façade and, in particular, in the NW Peninsular.

indicate, even if indirectly, a close relationship between the ways of production of these amphorae and the characteristics of farming production and trade. This is to say, as with Dressel 20 amphorae, the production of Haltern 70 amphorae would have followed a clear export strategy and would have implied the creation of a patterned shape that was produced in large quantities and in different regions of Baetica, especially in the Gaudalquivir valley.

In that study we suggest that there was a great coherence between those densities and good sea and river routes, which followed the itineraries that extended to the gold exploitations of the NW and the military camps. These routes were, as such, preferably used for the supply of the armies in the first decades of the imperial period (Morais and Carreras, 2003, 112).

The complementary relationship of Haltern 70 and Dressel 20 amphorae seems, however, to precede the annonary system. In fact Carreras’ (2003, 118) observation about the heavy concentration of Haltern 70 amphorae in Xanten in the period of Augustus is extremely interesting, suggesting that these amphorae were gradually substituted by the appearance of the first “archaic” oil Dressel 20 amphorae.

We should equally take into consideration the assumption of a direct and low-cost system of supply in the trade of Haltern 70 amphorae, in accordance with the principle of market economy maximization (vd. Dicken and Lloyd, 1990, 181-184; Carreras, 1999, 94), only comparable to the annonary system of Dressel 20 amphorae.

In these circumstances we can say that, at least from Augustus onwards, the necessary conditions had been created for the maintenance of a hybrid supply system, in the hands of private individuals, together with a true annonary system of state nature that began with the trade of Dressel 20 amphorae.

A plausible explanation for this situation may be related to the existence of a hybrid market system, in close relationship with the variables cost of transportation/ ethnicity of the native populations.2

The people in charge of the sale and commercialisation of Baetic oil products, who now benefited from a supply system to the armies and the city of Rome, certainly created suitable conditions for the maintenance of a hybrid system of supply, keeping exports of other types of products (i.e. fish, wine and its derivatives), whether they were integrated within the state system or not (as, for example, other types of Baetic amphorae found in the Testaccio seem to prove) (Carreras, 1999, 91-98).3

An argument in favour of a hybrid market system is suggested by the names gathered in the tituli picti and in stamps of Haltern 70 and Dressel 20 amphorae, i.e. the tituli with the names Q(uinti) VRITTI REVOCATI and M. VALERI EVPHEMI (Colls et al, 1977; Manacorda, 1977, 131, Fig 1 [CIL IV 9611]; Beltrán-Lloris, 2000a, 324; 331, note 86), and the stamps CFVFAVITTI present in both typologies (Carreras, 2001, 425-442). These seem to

3

The study of the oldest amphorae of the Testaccio is yet to be done because the levels of the Ist century and early IInd century have so far not been excavated (Carreras, 2001, 425). It is, however, possible that we may find Baetic amphorae and amphorae from other places in these layers, just like other contemporary registers known in the Roman Empire.

1

The quantitative calculation of amphorae density divided by the extension of the excavated area. For a further discussion on this method, see Carreras (2000a, 45-62). 2 On wine consumption and its derivatives in the NW, see Morais and Carreras Monfort, 2003, 104-106.

245

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

The provinces kept on receiving large amounts of Baetic products4, as the countless remains recovered from land and sea excavations prove.

of a diversity of products from Baetica, as the tituli substantiate. These tituli mention derivatives from wine products (defrutum, sapa), oil products (olivae) and curing products (muria) (vd. Aguilera, 2003, 119-120; Carreras, 2003, 117). These contents are corroborated by the presence in Haltern 70 amphorae of resin residues (Colls et al., 1977; Chic García, 1980; Juan and Matamala, 2003, 165-166; Morais, 2004, 545-565), and the remains of grape pips (Chic, 1980; Juan and Matamala, 2003, 165- 166), and olive stones (Liou and Domergue, 1990, 29; Nieto et al., 1989; Liou, 2000, 1063) – as verified by recent photolitho analyses of an example of this type (Juan, 1998, 88) – and also fish remains (Fitzpatrick, 1989).

This situation changes for the supply of Dressel 20 amphorae along the Atlantic façade and, in particular, in the NW peninsular. Even though their numbers are significant (vd. Fabião, 1993-94, 219-245), they do not reach the volumes documented in other places. This issue, so far considered insoluble (Morais, 2005), may have a very simple explanation, one exclusively related to the annonary supply system: the Atlantic route – certainly intensified after the conquest of Britannia and strategically used to supply the armies and populations – seems to have been less well used for the supply of Baetic olive oil. We are not obviously depreciating the importance of this route, but only trying to understand the reason why the presence of these amphorae in this region is not very significant in comparison with other regions that benefited from the traditional routes of the annona supply. A study carried out by Carreras (2001, 425-442), regarding the distribution of the stamps C. SEMPRONI PLOYCLITI and III ENNIORVM IVLIORVM, corroborates the opinion that the supply of Baetic oil to the legions and the city of Rome was structured with a series of previously established routes. Among these routes we should highlight the traditional routes of distribution which followed the Rhine-Rodano axis and the Gaulish isthmus and towards Rome, the largest consumption market in ancient times.

These data are enough to demonstrate that Haltern 70 amphorae were used to transport a diversity of products, including marine products. A good example is muria, a salty concoction, used for the preparation of sauces and fish curing, made from the fermented viscera and gills of tuna, mixed with the blood and other fluids of the same fish. The proximity of Marismas, one of the production regions of Haltern 70, situated in Baixo Gadalquivir – a region where there was an important lagoon with a diversified economy known as Lacus Lagustinus (vd. Carreras, 2000 b, 419-426) – in the Gaditanian area, is on its own suggestive of the possible use of Haltern 70 to transport diversified products, some of which were of Gaditanian origin. This might explain the highlighted presence of lids of Gaditanian fabric collected in Aljube do Porto, which are only quantitatively justifiable by the equally highlighted presence of Haltern 70 amphorae.

As we had the opportunity of suggesting about the context of Haltern 70 from Aljube, we can equally assume that in this hybrid market system there would have been a complementary position between some of these pottery production sites of the Guadalquivir and certain production areas of the Andalusian coast that resorted to the empty amphorae of the Guadalquivir for the sale of their products.

This might equally explain the production of Haltern 70 amphorae in the Cádiz Bay in Puente Melchor (García Vargas and Lavado, 1995; Lagóstena, 1996; García Vargas, 1998), and in the Algeciras bay in Venta del Carmen (Bernal and Lorenzo Martínez, 1998a; 1998b; 2000), and, as we have already seen, their early imitation in Lusitanian production sites associated with the production of fish sauces (Morais and Fabião, forthcoming). In fact, it will be easier to think that these productions imitated a type of amphora that was known to transport a diversity of products, a type that was well known in the distribution and consumption markets, especially along the Atlantic façade.

In other words we could be facing a particularly interesting situation – that some Haltern 70 might have been used for the transport of Gaditanian products in the context of a relative autonomy of certain pottery production sites, whose independent productive activities would lead them to benefit from the sale of their products to the Andalusian coast. This possibility gains some consistency if we think that we are faced with atypical amphora used for the transport

4 The vitality of this trade is also documented along all the coastal route of the Atlantic façade, with special emphasis on the maintenance of importats of large amounts of Haltern 70 amphorae up to the Flavian period (Morais and Carreras, 2003, 93-112; Morais 2004, 545-565; Morais 2005).

246

Part Five THE ATLANTIC TRADE ELITES

THE ATLANTIC TRADE ELITES

5.1 THE GADITAN ELITES AND THE FIGURE OF L. CORNELIUS BALBUS C. Carreras Seven years later, 61 BC, J. Caesar returned to Hispania Ulterioris as propraetor with a hunger for real military success and revenues that could help pay his debts in Rome. He waged war on the northern Lusitanian tribes who lived in the Mons Herminius (PLU. Caes. 12):

THE FIGURE OF L. CORNELIUS BALBUS The figure of L. Cornelius Balbus deserves special attention, because he may have been responsible to some extent for the administrative changes that took place in the Augustan period with regards to military supply. Lucius Balbus was born at Gades in 97 BC inside the Roman province of Hispania Ulterior. He was very young when he enrolled in the Q. Metellus’ army (CIC. Bal. 5) and collaborated with Cn. Pompeius and his commander C. Memmius in their fight against Sertorius (79-71 BC) (Rodríguez Neila, 1992, 29).

“As soon as he came into Spain he was very active, and in a few days had got together ten new cohorts of foot in addition to the twenty which were there before. With these he marched against the Calaici and Lusitani and conquered them, and advancing as far as the ocean, subdued the tribes which never before had been subject to the Romans. Having managed his military affairs with good success, he was equally happy, in the course of his civil government. He took pains to establish a good understanding amongst the several states, and no less care to heal the differences between debtors and creditors. He ordered that the creditor should receive two parts of the debtor’s yearly income, and that the other part should be managed by the debtor himself, till by this method the whole debt was at last discharged. This conduct made him leave his province with a fair reputation; being rich himself, and having enriched his soldiers, and having received from them the honorable name of Imperator.”

“He admits that he served in Spain, in a most severe war, with Quintus Metellus, with Caius Memmius; that he served both in the fleet and in the army; and, when Pompeius came into Spain and began to have Memmius for his quaestor, that he never left Memmius; that he went to take possession of Carthage; that he was present at those two hardly contested and most important battles of Sucro and the Durius; that he remained with Pompeius to the end of the war.” (CIC. Bal. 5) His long experience with the Roman legions made him aware of supply requirements and weaknesses in the whole supply system. As a result of his commitment to Pompey’s army during Sertorian’s war, he was granted Roman citizenship in 72 BC thanks to the Lex Gellia Cornelia. As a result he took the nomen Cornelius and became a member of the equestrian order, in other words a Roman knight.

L. Cornelius Balbus seems to have taken part in these campaigns, providing maritime support and supplies, and he may have traded part of the booty obtained (Rodríguez Neila, 1992, 59). It is believed that a Gaditan fleet of 80 or 90 ships transported soldiers as far as Brigantium (La Coruña) to complete the conquest of the NW Iberian Peninsula (D.C. 37, 52-53; ZON. 10.6). As a result of the first campaign, J. Caesar appointed L. Cornelius Balbus praefectus fabrum, chief engineer.

In 69 BC J. Caesar was appointed quaestor in the Hispania Ulterioris under the command of C. Aristius Vetus, when he met L. Cornelius Balbus and became an intimate friend. It is interesting to see the relationship between a person becoming an economic magistrature in the province and a member of the Gades elite involved in trading ventures and also, probably, food production (i.e. fish-sauces).

In terms of archaeological evidence, there are only a few Republican military camps confirmed in Lusitania (Cáceres el Viejo, Alpiarça, Lomba do Canho) that may have taken active part in those campaigns. In terms of 249

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

happens with Mamurra and Labienus (CIC. Att. 2.3.3). The statutes of the Urso colony specified a special seat in the local theater for the praefectus fabrum, which means that it was a key post in the Baetican province (Lex Urso 127).

amphora assemblages, all those Republican military camps record an important percentage of Baetican vessels, basically fish-sauce containers. At the end of the Republic, the praefectus fabrum was responsible for all the units employed in engineering work, such as building bridges, artillery, siege and defense works (Dobson, 1966; Deehene, 1981). It is unknown whether the appointment of L. Cornelius Balbus added further duties to this office. It should be borne in mind that L. Cornelius Balbus helped Pompeius and later J. Caesar with ships and food, so he could have provided army supplies directly.

However, the office of praefectus fabrum involved different tasks during the Principate, which were more related to production and engineering inside the Roman troops, and became part of the early stages of an equester’s career (Saddington, 1985; Welch, 1995; González, 2004). The office disappeared in the IInd century AD.

Portuguese archaeologists have documented in recent years an important increase in Baetican amphora imports from around mid Ist century BC onwards, approximately the same date as the Lusitanian campaigns of J. Caesar. Most common amphora imports from this southern region are the Gaditan ovoids (early Dressel 10) carrying fish-sauce products, Oberaden 83 carrying olive-oil, Haltern 70 and Lomba do Canho 67 that held a variety of products (i.e. defructum, olives, perhaps wine) (Fabiao, 1989; Morais, 1997-8; 1998).

During the Gaulish wars, J. Caesar was concerned about corn supplies for his legions (CAES. Gal. 5.17.2-5; 7.16.3; 20.9). Perhaps this was why he created the special office responsible for corn supply that was held by the Roman knight C. Fufius Cita, “qui rei frumentariae iussu Caesaris praeerat” (CAES. Gal. 7.3.1). Later he went to Rome as a representative of J. Caesar and became artifex of the Lucca agreement (56 BC) between Pompey and Caesar, which was the basis of the first triumvirate. Therefore he was a key figure in the Late Republic and his active role earned him many enemies. Some of them sued him because they argued that he could have not receive Roman citizenship under the law of Lucius Gellius Publicola and Cnaeus Cornelius Lentulus Clodianus. He was defended by Pompey, Crassus and Cicero, from whom we got the testimony of his trial (CIC. Bal.).

It is still difficult to make a clear distinction between the arrivals of imports in the NW of the Iberian Peninsula for each chronological phases, however there is a shift between an Italian supply of the late IInd c. BC – first half Ist BC (Dressel 1 and Brindisi) – to a Baetican one from the middle of the Ist c. BC onwards. Probably L. Cornelius Balbus fostered the provincial involvement in supplying Roman armies, either directly or through contacts with other Gaditans. In this context, the praefectus fabrum’ office may have played a key role.

The prosecution of Balbo was a direct attack on Pompey, who granted citizenship to the Gaditan. As Cicero pointed out:

Syme (1939, 355) already linked this office of praefectus fabrum to intendance and supply, pointing out the category of people appointed who were confidents of J. Caesar:

“Therefore it is the status of Cornelius as a citizen, and the action of Pompeius that are now on their trial before this court. For you admit that this man was born of a most honourable rank in the state to which he belongs, and that from his earliest manhood disregarding all his private affairs, he was passed his whole time in our wars in the company of our own generals, and that he has been absent from no labour, from no siege, and from no battle. All these things are full of glory, and are the peculiar glory of Cornelius; nor is there any crime in any part of such conduct.” (CIC. Bal. 6).

“…a proconsul chose for his agent and chief officer of intendance and supply a knight of no small consequence, the praefectus fabrum. The names alone of these officers are sufficient testimony2” (2) Balbus under Caesar in Spain, Mamurra in Gaul. It might also be conjectured that men like Ventidius, Salvidienus and Cornellius Gallus have been praefectus fabrum. Under the Principate, however, the position soon declined in importance.

Conflict between Pompey and Caesar caused him some disquiet as he was good friend of both. Despite being on Caesar’s side, the civil war did not produce for him the results he might have expected from his political efforts in Rome. Caesar with the support of Gades managed to beat Pompey’s lieutenant Titus Labienus, and his sons, in the battle of Munda (Osuna) in 45 BC. As a result, the Gaditans received Roman citizenship.

In 60 BC, Balbo goes with his friend J. Caesar to Rome, where he met Cicero and renewed his old friendship with Pompey. Caesar offered the same office, praefectus fabrum, to L.Cornelius Balbus during the campaigns in Gaul (CIC. Bal. 63). It seems that the office was a type of assistant for the proconsul or praetor in charge of the province; therefore, it was an office very close to commander. Rodríguez Neila (1992, 60) argues that a praefectus fabrum was responsible for negotiating the share of booty that corresponded to the commander, as

After the death of Caesar in 44 BC, Balbo organized support for Octavian before Mark Anthony, which brought about a second Triumvirate that included 250

THE ATLANTIC TRADE ELITES

Lepidus. Probably as a reward for all these tasks he received the praetorship in 42 BC and consulship in 40 BC, becoming the first non-Italian to reach this post. Then he disappears from the political scene.

such families appear in the local monumental epigraphy as well as painted inscriptions from fish-sauce amphorae (Dressel 7-11), which started production around 30 BC to the Flavian period. The following table includes painted inscriptions on Baetican Dressel 7-11 fish-sauce amphorae (Etienne, 2002; Lagóstena, 2002-3).

Since most Haltern 70 and fish-sauce vessels found in the NW Iberian Peninsula came from Baetica, it seems that Gaditan traders, perhaps the same gens Balbo, were involved in transporting supplies to the armies stationed in this region. Therefore the monopoly of army supply by Italian publicanii had changed and now provincials were fully involved in trade matters.

Some present cognomina appear to be of Greek origin (i.e. Hymnus, Poliddori, Hermetis, Nymphodorus, Anthus), which suggest that they were freedmen of the main local families involved in production and trade of salsamenta. Since most freemen adopted the praenomen and nomen of their former owners, these two elements can help us link painted amphora inscriptions to the main families in Gades featured on the monumental epigraphy.

Balbo the Elder acquired one of the great fortunes of Roman Principate; when he died (32 BC) he bequeathed to the plebs of Rome a sum of 25 denarii a head, even more than J. Caesar (D.C. 48.32.2). It is believed that a part of his fortune was obtained in trading ventures by himself and his family (Curchin, 1983; Chic, 1997).

To date a series of nomen have been identified in both, painted inscriptions and local monumental epigraphy: – Annii, Atii, Baebi, Caecilii, Clodii, Hermetii, Firmii, Manlii, Rusticii, Herculanii, Valerii and Svavis. Although a few of those nomen were universally popular, such as Valerii, others seem to be common only in Baetica and near Gades, such as Anni or Baebii. There are a good number of inscriptions from the gens Annii, Baebi, Clodii or Valerii in Gades, but also the same nomen is recorded in the Guadalquivir valley and other regions of Baetica (Lagóstena, 2002-3; 2004).

Gades was at that time at its height. Not only was Balbo an outstanding citizen playing politics in Rome, but also his nephew, Lucius Cornelius Balbus “the Younger”, was enjoying a promising military and political career. He fought on Caesar’s side in Egypt, the East, Africa and Hispania. Balbo the Younger, nephew of the elder, held the office of quaestor in Hispania Ulterior (44 BC) when he started numerous building projects in the city of Gades (CIC. Farm. 10.32). Later he was responsible for enlarging the port and its infrastructures according to Strabo (STR. 2.5.14; 3.5.3). Balbus “the Younger” continued his career as quattorvir, propraetor and member of Senate, becoming eventually proconsul in Africa, where he fought against the Garamantes (a Sahara tribe) in 19 BC. He obtained a fortune from his victory, explaining why he obtained an ovatio at Rome.

Chic (1997, 19) argues that some members of this nobilitas from Gades invested some of their trade income purchasing states in the territory of Jerez and the mouth of the Guadalete and Guadalquivir. Most Baetican products and amphora workshops come from these regions, so it seems that the second generation of rich Gaditan merchants became involved in agricultural production. Remembering that most Haltern 70 amphorae were produced along the Gaditan coast (García Vargas, 1998), i.e. the areas of Lacus Ligustinus and the Guadalquivir valley, it is reasonable to think that some Gaditan families were involved in its production and distribution around the Atlantic coast.

By the end of the Republic and Augustan period, Gades enjoyed huge commercial success. Despite being only a municipium civium romanorum (PLIN. Nat. 4.119), the city recorded more than 500 knights in the Augustan period (STR. 3.5.3). The mint issued large amounts of coins until its end in 19 BC, when M. Vipsanius Agripa modified mine administration in Hispania (Alfaro, 1988).

One interesting feature is that the common nomina from Gades, appear also in a series of monumental inscriptions in Lusitania. There are testimonies of Annii, Attii, Baebii, Caecilii, Clodii, Firmii, Manlii, Rusticii and Valerii in the Roman province of Lusitania. Nowadays, it is difficult to establish any personal relations between those people, although some family connections cannot be ruled out.

It seems that wealth in Gades was widely distributed amongst a series of families – nobilitas – directly involved in trade and food production. Testimonies of

251

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Table 5.1.1 List of tria nomina present on Baetican Dressel 7-11 (Etienne and Mayet, 1998; Etienne, 2002; Lagóstena, 2002-2003) Nomen

Cognomen

Praenomen

Tipologia

A

H

L

Dr.8

A

P

L

A

Victoris

P

Aemilius

Pudentis

L

Albucius

Actus

D

Dr.7

Annius

Hymnus

L

Dr.8

Nuevas n.101; 164; 176; 230; 276; 294 – Cádiz

Annius

Senecio

C

Dr.8

Carteia, Ilurco – C.Annius Senecio

Aquilius

Evocatus

M

Dr.8

Archivus

Munigua, Ecija – L.Aemlius Pudentis (flavi)

M

Aspasia

Poliddori

Atilius

Macrius

A

Attinii

AA

Attinii

CC

Attius

Insc. Nomen

Dr.10-11

Severus

P

Baebi

N

M

Baebi

Anthus

L

Dr.8

Baebi

Claricus

M

Dr.10-11

C

A

A

C

H

QQ

Caecilius

M Hermetis

Clodius

Herm

Prado del Rey (Cádiz) – C.Attius Severus Nuevas 146; 169; 203

C

Caecilius Clodius

Dr.8

Dr.9

M

Nuevas 265; 283 Nuevas 78; 228; 269

Hermet Cordius

Gratius

P

Cornelius

Hermerotis

C

Domitius

Felicio

Cn

Dr.7

Dorionis

Quietus

L

Dr.7

Firm

A

Grani

Iuni

Q

Grati

Rufi

P

Inseqventi

Nuevas 293

Velinorum

Nuevas …]pseqvens 261

Iossipi

Co

Iunius

Festus

L

L

N

M

L

N

N

Licinius

Lun

Licinius

Mahetis

Livicis Manlius

M

Dènia CIL II.4962

Dr.7

L Severus

L

Mari

Ponici

L

Mevi

Rufi

L

Minici

Rustici

C

O

M

N

Dr.7

Nuevas 165

252

THE ATLANTIC TRADE ELITES

Nomen

Cognomen

Nicandri

Dorothei

Non

Domillius

C

Numisi

Nicerotis

M

Numisi

Silonis

L

Porcius

Nymphodorus

M

Purelli

Gemelli

L

Quintilus

Hermae

M

Thori

Praenomen

Tipologia

Insc. Nomen

T

Titini

Herculani

Valerius

Catullus

A

Valerius

Caldonius

Cl

Valerius

Svavis

M

Verri

Apri

L

Vritti

Verecundi

T

Nuevas 4; 16; 70; 73; 76; 119; 161; 176; 177; 181; 231; 259; 278; 279; 286; 289

Aemilianus Barb Mam Dorionis Proviniorum Saturni

253

THE ATLANTIC TRADE ELITES

5.2 THE ATLANTIC TRADERS: THE GADITAN AND LUSITANIAN ELITES R. Morais says that “added to that, too, is the present peace, because all piracy has been broken up, and hence the sailors feel wholly at ease.”; by the archaeological testimonies of port infrastructures (anchorage and coastal places) and by the volume of goods found, among which we have to highlight the amphorae.

THE GADITAN ELITES The image of the Romans as a fighting people led them to minimize the importance of trade and production. The Gaditanians, who controlled the largest trade and navigation centre in the extreme west of the Mediterranean and one of the largest of the empire, were aware of this situation. Strabo himself (STR. III, 5, 3) says that the majority of the five hundred men assessed as Gaditanianian Knights (equites) did not live in Cádiz, but lived on the sea or in Rome, some of whom occupied high equestrian posts, including those of Egypt’s and the praetorium’s prefectures.

Pliny’s reference to the Gaditanians (PLIN. Nat. II, 167168), as the ones who sailed the Atlantic freely and in both directions, confirms the former references. Thus the commercial relationship with the Baetic Province turned out to be of the utmost importance to the NW Peninsular. The importance of the city of Cádiz and its port is highlighted by Mela when he mentions that the growth of the commercial traffic in the city led to the building of a new portus (MELA III, 4) and by Strabo, when he says that the same growth led to the foundation of a new Neapolis and of an anchorage (=επινειον) (MELA III, 5, 3), stressing the fact that the city possessed the largest commercial fleet and had the biggest ships sailing the Mediterranean and the Atlantic (MELA III, 5, 3). The logistical importance of the harbour of Cádiz in the context of the Baetica is equally present in Strabo’s words when he says that this region produces all things and in great quantities and that its surplus products are bartered off with ease thanks to a large number of commercial relationships with other provinces (STR. III, 2, 4). He also makes reference to the large quantities of grain, wine and olive oil of best quality that are exported from this region (STR. III, 2, 6).

In this process the role of the rich Gaditanian oligarchies stands out because they controlled the economic and commercial activities of the “Gibraltar Strait Route” and the majority of the sea circuit, Mediterranean and Atlantic. The maritime connection of the Northwest to the rest of the Empire’s commercial world was done through the harbour of Cádiz, which served as a link to this distant part of the Empire. This port was also directly connected to the supply of Britannia, through the Atlantic route, and to the Mediterranean basin through the port of Puteolos and later of Ostia. The vitality of this trade is equally documented by Strabo himself (STR. III, 5, 11) when he says that before the Romans’ arrival the Gaditanian traded with the famous Tin Isles, the Cassiterides, exchanging pottery, salt and copper utensils for animal hides, tin and lead.

But the literary sources are not the only ones that indicate the importance of the traded goods in the area of influence of the Gaditanian circuit. As a matter of fact, as we have been presenting throughout this work, the data so far gathered in places inside and outside the Peninsula strongly allude to the existence of specific networks of trade. These are associated both to the

The consolidation of this route is indirectly witnessed by Pliny (PLIN. Nat, 2, 167) when he says that, with Augustus, all the Atlantic was navigable; by Horatio (HOR. Ars. I, 31-33) when he says that one could do several journeys a year in the Atlantic without suffering any misadventures; by Strabo (STR. III, 2, 5), when he 255

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

long distance exportation of these products and to their short distance sale through cabotage systems of navigation.

Augustan legion (Diogo and Trindade, 1999). There is another missing inscription dedicated to this character of the second generation (CIL II 35, IRCP, 1852), which probably belonged to the base of a statue erected by the Scallabis colony in Alcácer where he lived as a result of the services he did for the city.

In the present Portuguese territory one of the best examples of the existence of secondary distribution networks through the cabotage system and subsequent land transport is the archaeological site of Castelo da Lousa, situated in Mourão (Évora). The study of the amphorae collected in this place shows that they are the type of products that were usually traded within the ambit of the military campaigns. They are represented by wine amphorae from the Tirreno coast, type Dressel 1, and by Baetic products that can be attributed to the Cádiz Bay (Types 9.1.1.1., 7.4.3.2., 7.4.3.3., Gaditanian ovoids, Dressel 9, Dressel 10 C and type “urceus”) and to the Gadalquivir (“early” Dressel 20, Haltern 70, Sala I/Lomba do Canho 67 and Dressel 12). Just like in other places situated on the Atlantic coast of the present Portuguese territory and Galicia we could also find ovoid amphorae of Lusitanian production (Morais, 2003, p. 3640) and amphorae of “regional” production similar to the Haltern 70 and Sala I / Lomba do Canho 67).

As far as the third generation is concerned, the one of L. Cornelius Bocchus, we know of two important inscriptions: one (FE, 60, 1999, nº 2753), found a few years ago in the baths of Cassius in Olisipo (vd Diogo and Trindade), is very probably part of a statue erected in his homage by decree of the decurions of that city; the other (CIL II 2479 and 5617, IRCP 1894), again found in Alcácer do Sal, belongs to a monumental plaque of a building that he is supposed to have had constructed in the city. Following the presented order, one inscription completes the other. The cursus honorum of this man is thus well documented: belonging to the equestrian order he might have began his public activity as a military tribune of the 7th Augustan Legion, followed by the praefectus fabrum post, which he occupied for five consecutive times; then his activity as an officer ends and he returns to Salacia, where he is twice elected as edis curuis (duumvirus) of the municipium, obtaining later the perpetual priesthoods as flamen and pontifex. At the top of his advancement, he occupied two other posts that surpass the municipal nature of his social mobility: he was elected, first, flamen of the Lusitania province and then twice praefectus Caesarum, substituting the members of the imperial family in the euergetism of the city (González Herrero, 2002, 33-57). Despite the fragmentation of this plaque, which does not let us see the reference to the type of prefecture this character might have assumed as praefectus Caesarum, this is, either duumuir or quattuoruir, it is possible to say that he was the former, taking into account another equally fragmented plaque5 found in Alcácer. There we can read, besides the praenome Lucius and the flaminate, the reference to the duumvirate.

If Castelo da Lousa is a good example of a short distance supply through the cabotage system of navigation and subsequent land transport, the data collected in other places outside the Iberian Peninsula are a good proof of the long distance exportation of these products. Among other places we could mention the data provided by the archaeological register of the city of Lyon (vd. Desbat, Lemaître, 2000, p. 793-815) and the Titán, Cap Sicié and Grand Conglué 3 shipwrecks found in the region of Provence (vd. among others, Parker, 1992, p. 106-107, 201-202 and 424-425) because they are a solid evidence of the trade success of these products through an open sea navigation and the use of large navigable river arteries.

THE LUSITANIAN ELITES The Cornelli Bocchi, one of the most enigmatic families of the Roman Lusitania, are a distinguished gens of the equestrian class referred to in documented inscriptions in Tróia, Setúbal, in Alcácer do Sal (Salacia) and in Lisbon (Olisipo).

If we analyse the cursus honorum of the Cornelli Bocchi of the second and third generation we realize that they had similar careers: both followed the first equestrian militia, belonging to the IIIrd and VIIth Augusta Legion respectively and ended up overcoming the municipal character of their social promotion through the provincial flaminate, being the highest authority after the governor.

In all, we know of six inscriptions that mention three different generations: Caius, Lucius son of Caius and Lucius son of Lucius.

2 [L(ucio)] C[ornelio C(aii) [ f(ílio)] Boccho / [fl]am(ini) pro[u]inc(iae) tr(ibuno) mil(itum) /[Co]lonia Scallabitana / [ob e(ius)] merita in cólon(iam). 3 L(ucio) Cornelio / L(ucii) f(ílio) Gal(eria) Boccho /Salaciensi / flamini proui[n]ciae Lusitania[e] / praef(ecto) fabrum V(quinquies) / trib(uno) mili(tum) leg(ionis) VII (septimae) / Aug(ustae) / d(ecreto) d(ecurionum). 4 [L(ucius) Cornelius L(ucii) f(ílius) Boc]chus pr(aefectus) Caesarum bis / [flam(en) prouinc(iae)? pon]t(ifex) perp(etuus) flamen / perp(etuus) / [duumuir aedilis?] II(bis) pr(aefectus) fabr(um) V(quinquies) tr(ibunus) mil(itum) / de s(ua) p(ecunia) f(ecit). 5 IRCP 188.

The reference to the first generation was marked in a fragment of an honorific plaque gathered in Tróia, Setúbal, but now gone (CIL II 5184, IRCP 2071). This inscription was dedicated to a Cornelius Bocchus of the second generation (Lucius son of Caius), whose most important activity might have occurred during the period of Augustus taking into account the reference to the IIIrd 1

[L(ucio) C]ornelio C(aii) f(ílio) / [B]occho / [flami]ni prouinc(iae) / [tr(ibuno)] mil(itum) leg(ionis) III (tertia) Aug(ustae).

256

THE ATLANTIC TRADE ELITES

We do not naturally hold the same importance for L. Cornelius Bocchus, nor do we support the idea that the post that he carried out five times in a row of praefectus fabrum was exclusively related to the supply of the armies; we are simply of the opinion that this post might involve the same type of responsibility as that of a proconsul assistant or a praetor in charge of a province.

As Marta González Herrero (2002, 54-55) has pointed out, they are in fact the first provincial Hispanic-Roman flamines to have equestrian military careers in the JuliusClaudius period. The reasons for such a promotion are varied. Among the possible ones we would like to highlight the importance of their services in the equestrian career as members of the first militia of IIIrd and VIIth Augustan Legion. In this position – as personal assistants of a cum imperium magistrate – they would certainly have carried out other non-military services (Welch, 1995, 144), which leads us to think that there might have existed a mixed career at a military and municipal level (Dobson, 1966, 68, footnote 41).

As a matter of fact, as we know, one of the measures taken by Augustus in 27 BC was the centralisation of imperial power through the new equestrian posts; this measure meant the gradual loss of power of the societates publicanorum and of the role played by the praefecti fabri to the imperial agents, the procuratori augusti, responsible for the levying of taxes in the imperial provinces, some of them related to the armies. According to Remesal (1986), the balance between the fiscus of the senatorial and imperial provinces was now under the responsibility of the praefectus annonae, who was in charge of the transactions between the traders (mercatores, naviculari) and the sate, through a payment for services provided (vecturae) (Remesal, 1986).

However, we feel we should highlight the cursus honorum of L. Cornelius Bocchus as prefect of the craftsmen, a post he carried out five times in a row. In fact, if we compare him to other characters who, at the end of the first republic and in the first decades of the empire occupied the same type of posts, especially the post of praefectus fabrum, we assume that L. Cornelius Bocchus had a close connection to economic and trade activities. We would thus be faced with a post that had some resemblance to the role played by the praefecti fabri of the late Republican period, who were, among other tasks of economic nature, responsible for the supply of armies on campaign.

We are thus facing a moment of change that was, as with any other type of far-reaching reforms, slowly being evolved so as to gain a larger concentration of imperial power. As proof of these changes, and at the same time of the continuity of ‘cronyism’ on the part of the powerful families related to trade activities and their social promotion, we have the figure of Balbus, the Younger, nephew of Cornelius Balbus. And this character can be used to integrate the importance of our L. Cornelius Bacchus, both in terms of relative contemporaneity and the type of posts that he carried out in the euergetism context that both may have undertaken. As far as Balbus is concerned, we know that he took up a post in Cadiz (44 BC) and that he may have started countless building projects (CIC. Fam. 10.32, apud. Carreras, 2006). He was also later responsible for the harbour enlargement and its infrastructures (STR. 2.5.14; 3.5.3, Carreras, 2006).

One of the best examples we know for the late Republican period corresponds to the role played by the Gaditanian L. Cornelius Balbus. This character, as Carreras has often pointed out (vd 2006, 172-173), was unquestionably associated with the conquest of our territory; as a very young man he cooperated with Cn. Pompeu and his captain Memmius in his fight against Sertorius (79-71 BC), supplying by sea the armies on campaign, which may have granted him citizenship in 72 BC. Later, in 68 BC, it would be Julius Caesar, who, as Quaestor of Hispania Ulterior, under the command of C. Aristius Vetus, would become his close friend. In 61 BC, this time as propraetor, Julius Caesar granted him the post of praefectus fabrum, with the responsibility of supplying the armies and giving them logistical support. We can thus understand the presence of a Gaditanian fleet with about 80 to 90 vessels responsible for the transportation of Julius Caesar’s troops up to Brigantium (La Coruña) to complete the conquest of the NW Peninsular (D.C. 37.52-53; ZONAR. 10.6, apud. Carreras, 2006). The success of this intervention might have secured him in the keeping of this post during Julius Caesar’s campaigns in Gallia (CIC. Balb. 63, apud. Carreras, 2006).

However, we must not forget that in both cases we were confronting one of the most flourishing periods of growth, both economic and in terms of social promotion – the end of the late Republican period and the beginning of the first decades of the imperial period. As Manuel Salinas de Frias (1995, 162) has well pointed out in his study entitled El Gobierno de las Provincias Hispanas durante la República Romana (218-27 BC), fast social advancement was due to different factors, such as:  the duration of the wars and the importance of military activity and of the regular soldiers there detached;  the profitability of his domain through the economic overexploitation of his resources, especially mining interests;

Later, in 40 BC, a date that coincides with the cereals blockade of Sicily to Rome ordered by Sextus Pompeius, Cornelius Balbus was already part of the Roman Consulate and a faithful counsellor of the future princeps. At the time of his death (32 BC), he had acquired one of the major fortunes of the Roman principality.

 the possibilities of personal enrichment and political influence given to those who carried out their duties successfully. 257

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Lying nearby was one of the most important urban agglomerations of the Roman period, Alcácer do Sal, the ancient Roman Salacia and residence of the Cornelli Bocchi. It has a privileged geographical situation on the right bank of the Sado River, which has attracted traders since earliest times, as the presence of remains datable to the VIIth century BC onwards proofs (Etienne, Makaroun and Mayet, 1999). Amphorae production centres were gradually created, especially along the right bank of the Sado River, in order to supply this huge industrial complex. Among these centres we may highlight, from upstream to down, the following: Barrosinha, Bugio, Enchurrasque-ira, Abul, Pinheiro, Quinta da Alegria and Setúbal (Mayet, Schmitt and Tavares da Silva, 1996; Mayet and Tavares da Silva, 1998; Mayet and Tavares da Silva, 2002; Mayet and Tavares, 2005). The abundance of both fish and salt here has, at least since proto-historical times, made this region one of the main suppliers of fish products in the Roman world.

In this way, L. Cornelius Bocchus – just like his father – might have been responsible for the supply of the native populations situated on the coastal settlements of the Atlantic façade. On a larger scale, he may have also been directly involved in the trade circuits of the Atlantic route and in the supply of the armies during the first decades of the imperial period. The appearance of inscriptions in Tróia, Setúbal, in Alcácer do Sal and in Olisipo poses extremely interesting questions concerning the study of the economy in Roman Lusitania in the beginning of the imperial period. In fact, if we relate the Cornelii Bocchi to the trade activities and, in particular, to the world of the amphorae, as Encarnação has recently suggested (2004), we might also have to associate them with the fish-sauce industry and its exportation. This hypothesis, already put forward by Dias Diogo and Laura Trindade, regarding the inscription of Olisipo (FE 1999, nº 275), seems to be, in fact, very suggestive.

The association of the Cornelli Bocchi to the beginnings and development of the largest industrial complex situated in Tróia is obviously still under discussion. In fact, according to the data so far available, this industrial complex seems to have begun output from the mid Ist century onwards (Etienne, Makaroun and Mayet, 1999, p. 30); if so, we should put aside the connection of the Cornelli Bocchi with such an industrial complex. However, we are of a different opinion. The lack of information so far registered is not an immediate indicator of the beginnings of these activities along the Sado, and in particular in Tróia. In fact, the complexity and importance that this industrial centre will gain in the Roman Empire indicates that there must have existed an initial stage, at least from the kingdom of Augustus onwards, not only regarding those industrial processes involved with salting and fish-sauces production, but also with amphorae production.

Particularly interesting is the finding of the inscription in Tróia, Setúbal, not only because it is near Alcácer do Sal, where the Cornelli Bocchi lived, but also because we may be in the presence of the first owners and promoters of the major industrial complex of fish-sauce production in the Roman Empire, whose vastness and importance is yet to be fully known. About forty kilometres south of Lisbon and the estuary of the Tejo lies Tróia, which was an island in Roman times. It stretches along the estuary of the Sado, over two kilometres wide, as a peninsula surrounded by the sea and the left bank of the river, making it an excellent place for the development of the salting and fish-sauce industries (vd. Etienne, Makaroun and Mayet, 1999).

258

THE ATLANTIC TRADE ELITES

5.3 WAS THERE A TRUE ATLANTIC VOCATION AMONG THE FIRST EMPERORS? R. Morais Even though we can consider the first emperors’ “Atlantic vocation” an “artificial” issue, in the context of a more complex analysis of the romanization of the Peninsula, we know that the Romans were aware of the difference between coastal and inland cities. Such an awareness is, for example, implicit in Plinys’ work, who, based on the information collected in Varro, says “the cities worthy of mention on the coast, beginning from the Tagus, are that of Olisipo, famous for its mares, which conceive from the west wind; Salacia, which is surnamed the Imperial City; Merobrica; and then the Sacred Promontory, with the other known by the name of Cuneus, and the towns of Ossonoba, Balsa, and Myrtili” (PLIN. Nat. 4.35.21).

A comparative analysis of the inhabited nuclei along the Atlantic façade (Morais and Carreras, 2003, 93-112; Morais, forthcoming) suggests that the first emperors focussed special attention on circulation in the Atlantic. Such attention is reflected in the previous campaigns carried out by the Roman generals during the Republican period, which occurred from the last third of the IInd century BC onwards. These campaigns, which benefited from the special assistance of the city of Cádiz and her elites, thus gradually gained control over the Atlantic route and obtained essential support in terms of material and logistical resources. Among the episodes of the conquest process and occupation of the territory, we should highlight the activities Julius Caesar’s action, in particular his alliance with the Gaditanians (i.e. L. Cornelius Balbus). Such an alliance allowed him to reach the city of Brigantium, which led to the subjugation of the indigenous populations, favouring sea communications along the north-south axis of the Empire and, in the short term, permanent contacts with the Baetica province.

This idea is equally present in a later source, datable to the VIIth century, known as the Anonymous of Ravenna’s (or Ravenate) Cosmography. Even though this is not a source datable to the Roman period we know that this work is based on older sources. Taking this aspect into consideration, the enumeration of eight groups of cities spread through the eight territorial-political divisions of Roman Hispania is a curious fact. Furthermore, we can distinguish two sets in that enumeration: the first regarding the cities situated on the coast, grouping them along the coastal façade of both seas, the Atlantic (Ocean) and the Mediterranean (Great Balearian Gallic Sea), and the places situated inland, grouped around certain road networks and coinciding frequently with the already known mansiones from the Antonine Itinerary.

However, it is only by the time of Augustus that the Atlantic world is fully integrated in a market economy, as a consequence of the intensification of the sea traffic and of the gradual romanization of the provincial territories. At that moment, the Northwest and the Atlantic coast enjoyed a major economic impulse as a consequence of the exploitation of mining wealth, which started being exploited immediately after the end of the Cantabrian wars, and of the increase in commercial flows with the exploitation of other products. Caligula himself paid special attention to the Roman traders’ private interests and was even responsible for having a lighthouse built in Gesoriacum to help commercial navigation. Claudius, however, was responsible for the impetus by giving major encouragement to trade, as well as conquering Britannia and proclaiming himself ruler of the ocean.

However, perhaps the issue should not be presented in this way, especially if we think that we are confronting two different realities that, nevertheless, can be at the same time complementary. What we can certainly accept is that the foundation or continuation of cities established before Roman arrival along the Mediterranean and Atlantic coasts not only facilitated the logistics associated with the conquest process but were also essential to the 259

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

out from points of reception on coastal areas, or which could be reached by navigable watercourses” (Fabião, 1998, 176), as was the case with the old Roman cities of Bracara Augusta and Aeminium.

inter-provincial and district-based trade (Morais, forthcoming, a). On the other hand, the foundation of inland cities, like Augusta Emerita, Caesaraugusta and Clunia, besides being directly associated with the conquest of strategic positions and situated by the junction of secular roads, was the result of a decisive political strategy to control the vast territories that were to be distributed among the veteran soldiers. The foundation of inland cities should adhere to a series of prerequisites previously established, which included the creation of other minor but indispensable centres to complete the administrative network.

It is from the interconnection and complementarity of these different realities that we should question the relevance of such an “Atlantic vision” of the first emperors. What cannot be denied is that the privileged situation and strategic importance of the different nuclei situated by the coast, or easily reachable by river routes, would have certainly influenced the policies of the emperors. These policies were not only based on the administrative control of the territories but also on the economic advantages of waterways over land routes. This reality, and not the issue of the emperors’ “Atlantic vision”, was what truly influenced the politics and visions adopted across all the Roman Empire.

In the centre of these two realities we have other urban nuclei apparently situated in inland positions but which benefited from “secondary circuits of distribution, carried

260

CONCLUSIONS: THE NEW ATLANTIC FRONTIER C. Carreras & R. Morais Another interesting issue is the chronological horizon, since most commercial contacts started at a very early date (Ist BC) with very rich archaeological materials that were prototypes of later types (i.e. amphorae). The evidence from this regional trade may shed some light on the evolution of some pottery types that are not found so readily in other parts of the Empire. At the end of the Ist century BC, the NW of the Iberian Peninsula endured the series of military campaigns led by Augustus against Cantabrii and Astures which probably fostered Atlantic maritime routes in order to supply Roman troops.

When the Atlantic route in the Roman period is discussed, people tend to deal with it in a restricted way, only covering a limited coastal territory that belongs to one or two modern States. Nevertheless, Roman traders did not recognize such modern political boundaries and they understood the Atlantic – mar exterior – as a unique environment connecting different territories of the Roman Empire – even barbarian ones. For a long time, Atlantic navigation in Roman times has been questioned due to the evident difficulties of this fierce ocean, but new material evidence (i.e. shipwrecks, archaeological distribution, etc.) question every day such a view. The presence of numerous shipwrecks along the Atlantic coast, the large volume of amphorae recovered from coastal and inland sites, the distribution of pottery and other products (i.e. stone-mills) demonstrate a constant commercial exchange amongst the diverse Atlantic coastal populations.

It is believed that the complexity of supplying the distant wars may have provided Augustus with a steep learning curve, but one he might have benefited from a couple of decades later in terms of the development of supply lines for the Germanic limes. Of course, there are still many questions to be answered, such as what ship itineraries were chosen to move around the Atlantic and how these trading ventures were financed. However, there is rich archaeological evidence that can be explained now in some detail.

Nowadays, it is well accepted that the mar exterior was another navigation route in Roman times and there were even longer voyages towards distance places, such as the British Isles. Nevertheless, scholars debate the volume and scale of such maritime traffic along the Atlantic coast. They raise questions about whether there was regular commercial traffic or simply occasional trade ventures.

Meanwhile, navigation in the Central Atlantic and English Channel is well documented by ancient sources, such as Julius Caesar in his Bellum Gallicum and the archaeological record (i.e. pottery, amphorae, and coins from both sides of the Channel). Some of these initial accounts of Roman commercial exploitation from the Central Atlantic coincide with the distribution of Italian Dressel 1 amphorae during the IInd-Ist century BC. Such vessels were widely distributed in Gallia as a whole (Fitzpatrick, 1987), reaching the south and south-east British Isles, where the redistribution port of Hengisbury Head stood out. This port, located in Christchurch Bay opposite the Isle of Wight, received in the Late Iron Age materials from Armorica (i.e. ceramics) and CornwallDevon (i.e. coarse wares), as well as amphorae imports – all demonstrating the existence of lively trade in the Ist century BC.

We believe that this present volume addresses all these questions and opens new ground concerning the strategic importance of the Atlantic route in the Roman Empire. So far, little was known about the thriving Roman trade on the Lusitanian coasts and NW of the Iberian Peninsula, a key part of the Atlantic route that reinforced contacts between the Mediterranean and the Atlantic world. In this context, the new evidence from the western Iberian Peninsula opens new possibilities for understanding the Roman economy in a more integrated way, one where southern provinces directly supplied northern ones with raw materials and added-value products, and vice versa. 261

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Itineraries chosen by those traders bringing Italian amphorae appear to be the main Gaulish rivers. In 60 BC, the Roman frontier was in the province of Gallia Narbonense, whereas the whole of northern Gaul was controlled by Celtic tribes. One of the routes chosen was that known as “isthme gauloise”, a route starting from Narbonne that followed the river Aude upstream, later following a land route to reach Toulouse. This town had a statio in the middle of the Ist century BC (Cic., Pro. Front. IX.19), therefore representing a commercial border with the rest of Gallia.

Roman fleet controlled the English Channel as a whole, and was the sole maritime power in the Atlantic, between the Strait of Gibraltar and the mouth of Rhine. As Roman legionaries wielded power along the Atlantic shores, Italian Dressel 1B amphorae appeared regularly all over those conquered lands. As can be seen by the description of historical events, Roman policy at the end of the Ist century BC attempted to integrate the whole Atlantic – Mar exterior – under Roman rule. There was a series of decisions and military campaigns that attempt to create a new Roman Sea.

Goods continued their way down the river Garonne up to Burdigala (Bourdaux), where they could continue their navigation through the Atlantic. The whole route is documented by the large amount of amphorae Dressel 1 recorded, chiefly in the area of Toulouse, where a terminus may have been located that required a change of container (Tchernia, 1986)1. In the second half of the Ist century BC another wine amphora from NE Spain turned up, the Tarraconense Pascual 1, following the same itinerary as its Italian predecessor (Carreras, 2009). This type also recorded high densities along the whole “isthme gauloise” route but is also found at many Gaulish sites in Britanny and Normandy.

Augustus continued J. Caesar’s work, completing pacification of the Gaulish tribes living in the Atlantic by campaigns against the Aquitanii (French Basque country) or Morini (Pas de Calais). With regards to the Low Countries, the campaigns emptied the territory, which was afterwards continuously threatened by German incursions. Therefore, Agrippa, as a governor of Gaul, fought against Suebii and Chatii in 39-38 BC and resettled the Roman allies, the Ubii, at Köln (Ara Ubiorum – later Colonia Agrippina). With the new organization of Gaul by Agripa, the capital of Gallia was founded at Lyon (Lugdunum) in 43 BC, a key point that linked the Mediterranean with the Rhine, along the river Rhone following Chàlon-sur-Saône, and later the land route up to Metz, where the Moselle was taken as a riverine route up to Trier (Treveris) and Köln (Ara Ubiorum). Finally, the Romans could also reach the Rhine along the Atlantic coast from the Strait of Gibraltar and along the main rivers of Gallia and Germania.

An interesting detail is that Tarraconense amphorae are hardly recorded in the NW Peninsula, which means that once they reached the port of Burdigala they were traded north instead of south. There is a feeling that the whole military market of the NW Peninsula was under the control of Gaditan traders, who were not keen to introduce other Tarraconense products. Perhaps, the distribution of the Baetican Haltern 70 amphorae (that may have contained wine) in the NW could explain the lack of a presence of Tarraconense vessels.

However the region was far from pacified. In 17 BC, the governor of Belgium, Marcus Lollis, lost the legion V Alaudae against the Sugrambrii, another German tribe from the eastern Rhine. Augustus was obliged to send his own stepsons, Drusus and Tiberius, to fight against these German tribes and to keep peace in the Lower Germania.

The Venetii were the Gaulish tribe controlling the Atlantic trade from Burdigala up to the British Isles. They settled the Breton coast, and, according to J. Caesar, were the main sea power of the western Atlantic. This explains why J. Caesar started his campaigns in Gaul in 58 BC, taking into account the importance of these maritime supply routes. He was well aware of the potential of the Venetii as he knew, as a legatus of the Ulterior from his Gaditan friend Lucius Cornelius Balbo, that they exerted maritime control over the Central Atlantic.

Between 16 and 13 BC, the two generals organized the Roman legions around the Rhine, settling the main military garrisons at Nijmegen, Xanten (Vetera I) and Neuss. In addition, Drusus decided to build a series of channels and dams to improve navigation of the main rivers towards the North Sea and Eastern Germania since supplies were normally carried through such river routes. Therefore, the Romans took full advantage of the Atlantic and North Sea navigation routes to reach their military outposts.

The Roman campaigns continued (57 BC) all over the Atlantic coast region (British Channel), beating the Nervii and the Atuatucii in the Meuse valley in the Netherlands. Therefore, J. Caesar faced the Venetii in 56 BC so as to control all the Gulf of Biscay and part of the British Channel.

Between 12 and 9 BC, Drusus started his campaigns to conquer Germania from two legionary camps, Vetera (Xanten) and Mainz, and following the rivers Lippe and Main. The main aim of such campaigns was the control of the Vlatava valley (between the Elbe and Danube), because of its mining resources, which were especially attractive for the Romans. His first campaigns were against the the Sugambrii, who were routed, and later he returned to the Rhine and moved his legions form Nijmegen towards the North Sea. From the coast he

Between 55-54 BC, the Romans reached the British Isles for first time and even fought some of the local chieftains, Cassillevanus, in Kent. As a result, in the year 50 BC, the 1

There was an International conference on this route, and others in the Gallia, held at Lattes with the title Itinéraires des vins romains en Gaule. IIIer-Ier siècles avant J.C. (2007).

262

CONCLUSIONS: THE NEW ATLANTIC FRONTIER

Map of the main sites mentioned in the text

263

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

the early stages of the Cugernii site at Xanten again included an important number of Tarraconense, Italian, Greek and Baetican vessels, probably destined for the military from Vetera I. The most common types of these Augustean-Tiberian contexts at Xanten were Southern Spanish Dressel 7-11 and the “Atlantic” Haltern 70.

attacked the Frisii and Chaucii. In the following year Drusus again fought the Sugambrii, following the river Lippe to reach almost as far as the River Wesser. Drusus established a line of military camps along the river Lippe, starting from Oberaden, whose foundation was dated to 11 BC by dendrochronology. He probably founded in the same years the military camp at Haltern along the Lippe, well-known by pottery specialists for its Italian Terra Sigillata (Samian Ware) and amphorae such as Haltern 70. The wooden military camp was fully excavated in XIXth century and was occupied until the defeat of Varus (AD 9), providing an interesting amphorae assemblage dated from this short period (circa 11 BC to AD 9) with amphorae Dressel 1B, Dressel 9 similis, early Dressel 20 (Haltern 71-Oberaden 83), Rhodian and, of course, our Atlantic amphora – the Haltern 70.

This present volume has attempted to collect evidence of the lively trade in the Atlantic from the Ist century BC up to Ist century AD, when the Romans decided to conquer the territories of the Atlantic littoral. The papers here cover the commercial phenomena detected from the Strait of Gibraltar up to the Galician coasts of the NW Iberian Peninsula, which was probably determined by the military campaigns in the NW during the Augustan campaigns against Cantabrii and Astures, and later the NW exploitation of the rich mines there. However, our feeling is that the phenomena was not limited to the Iberian Peninsula but affected the whole Mar Exterior (Atlantic), from the coast of Armorica, Brittany, Normandy, Belgium and Germania Inferior. Despite obvious differences between all these territories, there were some common traits in material culture, information, traders and logistics which cannot be explained in a fragmentary way. The present volume attempts to fill a gap from the western side of the Atlantic, but we are aware that some questions raised here can be only answered from other territories of the Mar Exterior.

Drusus’ campaigns continued until AD 9, when he reached the Elbe on his second campaign against the Chatti. Incidentally, finds of a possible temporal camp of his have been recorded in Hedemünden near the Elbe. On his way back, Drusus fell from his horse and died. From this moment (AD 9-8), Tiberius becomes the commander-in-chief in Germania and carried on his offensive against the Sugambrii, who were routed and later deported to the western Rhine, where they were settled at a site near Vetera I known as Xanten, and changed their name to Cugernii. Amphorae finds from

264

BIBLIOGRAPHY ADAN, G. 1997: “Intervenciones arqueológicas de la Comisión Provincial de Monumentos HistóricoArtísticos de Oviedo (1844-1978)” Lancia 2, 207233.

AA.VV. 2005: El món romà a les Illes Balears, Barcelona. AA.VV. 2006: Historia de la pesca en el ámbito del Estrecho. I Conferencia Internacional (El Puerto de Santa María, 2004), Sevilla.

AGUILERA, A. 2005ª: “Los tituli picti”. In Carreras, C.; Aguilera, A.; Berni, P.; Garrote, E.; Marimon, P.; Morais, R.; Moros, J.; Nieto, X.; Puig, A.; Remesal, J.; Rovira, R.; Vivar, G. (2005) Culip VIII i les àmfores Haltern 70. Monografies CASC 5, Girona, 58-69.

ABAD, L. 2003: “El tránsito funerario. De las formas y los ritos ibéricos a los modelos romanos”. In L. Abad (ed.) De Iberia in Hispaniam, Fundación Duques de Soria, Madrid, 75-100. ABAD, L.; SALA, F. 1997: “Sobre el posible uso cúltico de algunos edificios de la Contestania ibérica”, QPAC 18, 91-102.

AGUILERA, A. 2005b: “Defructum, sapa y caroenum. Tres nombres y un producto: arrope”. In Carreras, C.; Aguilera, A.; Berni, P.; Garrote, E.; Marimon, P.; Morais, R.; Moros, J.; Nieto, X.; Puig, A.; Remesal, J.; Rovira, R.; Vivar, G. (2005) Culip VIII i les àmfores Haltern 70. Monografies CASC 5, Girona, 120-132.

ABAD, L.; SALA, F. (eds.) 2001: Poblamiento ibérico en el Bajo Segura. El Oral (II) y La Escuera, Real Academia de la Historia, Madrid. ABAD, L.; SALA, F.; GRAU, I. (eds.) 2005: La Contestania 30 años después, Alicante.

ALARCÃO, J. 1991: O domínio romano em Portugal. Mem Martins.

ABASCAL, J.M.; ABAD, L. (coords.) 2003: La ciudad y los campos de Alicante en época romana, Canelobre 48, Alicante.

ALARCÃO, J. 1994: “Lisboa romana e visigótica”. In Lisboa subterrânea. Lisboa, 58-63.

ABOAL FERNÁNDEZ, R.; CASTRO HIERRO, V. (coords.) 2006: O castro de Montealegre (Moaña, Pontevedra), Noia.

ALBA, M.; MÉNDEZ, G. 2002: “Evidencias de industria paleolítica y de un alfar altoimperial en Augusta Emérita: intervención arqueológica realizada en el solar de esquina entre la prolongación de la calle Anas y el final de la avenida de Lusitania”. Mérida, excavaciones arqueológicas 8, 375-410.

ACUÑA, F. 1971: “Los Lares Viales en la Galicia Romana” II C.N.A. Coimbra, 1970, 353-357. ACUÑA PIÑEIRO, A. 1995: Informe de valoración. Excavación arqueológica del solar nº 5 de la calle Hospital, Vigo (Pontevedra), Informe inédito depositado en el Servicio de Arqueología de la Xunta de Galicia, Santiago de Compostela [unpublished].

ALCOCK, J.A. 2001: Food in Roman Britain, Wiltshire. ALCORTA, E. 2002: Lucus Augusti II. Cerámica común romana de cocina y mesa hallada en las excavaciones de la ciudad, Lugo.

ACUÑA PIÑEIRO, A. Informe de valoración II. Ampliación de la excavación arqueológica del solar nº 5 de la calle Hospital, Vigo (Pontevedra), Informe inédito depositado en el Servicio de Arqueología de la Xunta de Galicia, Santiago de Compostela [unpublished].

ALEXANDROPOULOS, J. 2000: Les monnaies de l’Afrique Antique 400 av. J.-C. – 40 ap. J.-C. Toulouse. ALFARO, C. 1988: Las monedas de Gadir/Gades. Madrid. ALFOLDY, G. 2000: “Das neue Edikt des Augustus aus El Bierzo in Hispanien. Einfürung.” Zeitschrift für Papyrologie und Epigraphik 131, 177-205.

ADAMS, C.E.P. 1999: “Supplying the Roman army: bureaucracy in Roman Egypt” In A. Goldsworthy and I. Haynes (eds.) The Roman Army as a community, JRA Suppl. 34, Porthsmouth, 119-121.

ALFÖLDY, G., ABASCAL, J.M.; CEBRIÁN, R. 2003: “Nuevos monumentos epigráficos del foro de 265

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

AMARO, C. 1993: “Vestígios materiais orientalizantes do claustro da Sé de Lisboa”. In Estudos Orientais. Lisboa: Instituto Oriental da Universidade Nova de Lisboa. 4, 183-192. AMARO, C. 1995: “Núcleo Arqueológico da Rua dos Correeiros”. Lisboa: Fundação Banco Comercial Português. AMELA, L. 2000: “La turma salluitana y su relación con la clientela pompeyana”. Veleia 17, 79-92. AMELA, L. 2003: Cneo Pompeyo Magno. El defensor de la República romana, Madrid. ANDERSON, J.D. 1992: Roman military supply in North-East England. BAR Brit.Ser. 224, Oxford. ANDRÉ, J.M. (ed.) 2004: Hispanité et Romanité, Casa de Velázquez, Madrid. ANDREAU, J. 2001: “Rome capitale de l’Empire: la vie économique”. Pallas 55, 303-317. ANGELUCCI, D.E. 2003: “A partir da terra a contribuição da Geoarqueologia. In Mateus, J.E. e Moreno, M. (eds.) – Paleoecologia Humana e Arqueociências – Um Programa Multidisciplinar para a Arqueologia sob a Tutela da Cultura. Lisboa: Instituto Português de Arqueologia. Trabalhos de Arqueologia, 29, 37-103. AOUNALLAH, S. 1996: “Castella et civitates dans le pays de la Carthage romaine”. L’Africa Romana. Atti del XI Convegno di Studio (Cartagine, 15-18 dicembre, 1994), Orieni 1996, 1505-1512. ARANEGUI, C. 1996: “Los platos de peces y el más allá”. Hom. a M. Fernández Miranda, Complutum 6,1, 401-414. ARANEGUI, C. 1997: en AA. VV., Los Iberos, príncipes de Occidente, Barcelona. ARANEGUI, C. 1999: “Personaje con arado en la cerámica ibérica (ss. II-I AC). Del mito al rito”. Mélanges C. Domergue 2, Pallas 50, 109-120.

Segobriga”, Zeitschrift für Papyrologie und Epigraphik 143, 255-274. ALMAGRO, M. 1980: “El monumento ibérico de Pino Hermoso (Orihuela, Alicante)”. TP 37, 345-362. ALMEIDA, R.R. 2006: As Producoes Anfóricas do Guadalquivir no Quadro das Importacoes de Scallabis. Contibuto para o conhecimento dos tipos minoritários no occidente peninsular, Master Thesis. Universidade de Lisboa. Lisboa. ALMEIDA, R.R.; ARRUDA, A.M. 2005: “As ânforas de tipo Maña C em Portugal”, V Congresso Internacional de Estudos Fenício-Púnicos, Sardenha. ALONSO, A.; FERNÁNDEZ CORRALES, J.Mª 2000: “El proceso de romanización de la Lusitania Oriental: la creación de asentamientos militares”. In J.-G. Gorges; T. Nogales, (coords.), Sociedad y cultura en Lusitania romana. IV Mesa redonda internacional, Mérida 2000, 85-100. ALONSO, C., MENANTEAU, L., GRACIA, F.J.; OJEDA, R. 2007: “Geoarqueología y paleomorfología litoral de la ensenada de Bolonia. Primeros resultados y nuevas propuestas”, in A. Arévalo & D. Bernal eds., Las cetariae de Baelo Claudia. Avance de las investigaciones arqueológicas en el barrio meridional (2000-2004), Arqueología Monografías, Sevilla, pp. 521-538. ALONSO, D.; PINEDO, J. 1999: Metamorfosis. El puerto de Cartagena antes del tercer milenio, Cartagena. ALVAREZ SAENZ DE BURUAGA, J. 1976: “La fundación de Mérida”, Emerita Augusta. Actas del Symposium Internacional Conmemorativo del Bimilenario de Mérida, Madrid, 19-30. ÁLVAREZ SANTOS, J.A. 1999: La terra sigillata en Cantabria: los fondos del Museo Provincial de Prehistoria y Arqueología y del Museo Arqueológico Nacional, Trabajo de investigación inédito, Santander [unpublished].

ARANEGUI, C. 2004a: “A propósito del vaso de los guerreros del Castellar de Oliva (Valencia)”, Soliferreum. Studia archaeologica Emeterio Cuadrado Díaz ab amicis, collegis et discipulis dicata, AnMurcia 17-18, 2004, 229-238.

ALVES, F.; DIOGO, A.M.; CARDOSO, J.P. 2001: “Considerações sobre os dois grandes cepos de âncora em chumbo com alma de madeira do séc. V-IV AC, provenientes do ancoradouro natural da Ilha Berlenga (Peniche) e sobre os achados de ânforas de “tipo púnico” em águas portuguesas”. In Os Púnicos no Extremo Ocidente. Actas do Colóquio internacional, (Lisboa, 27-28 de Outubro de 2000). Lisboa: Universidade Aberta, 239-260.

ARANEGUI, C. 2004b: Sagunto. Oppidum, emporio y municipio romano, Barcelona. ARANEGUI, C. 2006a: “Imaginario ibérico”. D. Vaquerizo, J.F. Murillo, eds., El concepto de lo provincial en el mundo antiguo. Homenaje a la profesora Pilar León Alonso, Córdoba, 113-124. ARANEGUI, C. 2006b: “Dar, intercambiar, comprar y vender en el Mediterráneo antiguo”. I. Aguilar, ed., El comercio y el Mediterráneo. Valencia y la cultura del mar, Valencia, 73-87. ARANEGUI, C. “Arte ibérico en la Edetania”. In L. Abad, M. Hernández, eds., Alicante. [forthcoming].

ALVES, F.; REINER, F.; ALMEIDA, M.J.; VERISSIMO, L. 1988/89: “Os cepos de âncora de chumbo descobertos em águas portuguesas – contribuição para uma reflexão sobre a navegação ao longo da costa atlântica da Península Ibérica na Antiguidade”. O Arqueólogo Português. Série 4. N.º 6/7, 109-185. AMANDRY, M. 1984: “Notes de numismatique africaine I”. Revue Numismatique 26, 85-94.

ARANEGUI, C.; DE JUAN, C.; FERNÁNDEZ, A. 2004: “Sagunto como puerto principal. Una aproximación náutica”. Méditerranée occidentale antique: les échanges, ANSER III, Soveria Mannelli, 75-100.

AMARE, M.T.; GARCÍA MARCOS, V. 1995: “Una producción de lucernas en Asturica Augusta”. Zephyrus XLVII, 273-285.

266

BIBLIOGRAPHY

Povoamento Proto-Histórico do Litoral Atlântico intitulado Ora Maritima. Das Colunas de Hércules a Finisterra. Outubro de 1997. Póvoa de Varzim. [unpublished].

ARANEGUI, C.; MARTÍN BUENO, M. 1995: “L. Planius Rusinus en el puerto de Denia”. Saguntum 28, 261-263. ARANEGUI, C.; MATA, C.; PÉREZ BALLESTER, J. 1997: Damas y caballeros en la ciudad ibérica, Madrid.

ARRUDA, A.M. 1999-2000: “Los Fenicios en Portugal. Fenicios y mundo indígena en el centro y sur de Portugal (siglos VIII-VI AC)”. Cuadernos de Arqueología Mediterránea, 5-6. Barcelona.

ARANEGUI, C., KBIRI ALAOUI, M., VIVESFERRÁNDIZ, J., 2004: “Alfares y producciones cerámicas en Mauritania Occidental. Balance y perspectivas”. D. Bernal, L. Lagóstena, eds., Figlinae Baeticae, BAR Int. Ser. 1266, Oxford, 366-378.

ARRUDA, A.M. 2001: “Importações “púnicas” no Algarve: cronologia e significado”. In Os Púnicos no Extremo Ocidente. Actas do colóquio internacional (Lisboa, 27 e 28 de Outubro de 2000). Lisboa: Universidade Aberta, 69-98.

ARANEGUI, C., ed., et al., 2005: Lixus-2 Ladera Sur. Excavaciones marroco-españolas en la colonia feniciACampañas 2000-2003, Saguntum-extra 6, Valencia.

ARRUDA, A.M.; ALMEIDA, R.R. 1998: “As ânforas da Classe 32 da Alcáçova de Santarém”. Conimbriga 37, 201-231.

ARANEGUI, C., HASSINI, H., VIVES-FERRÁNDIZ, J. “Las ánforas de época republicanas en Lixus (Larache, Marruecos)”, VI CISFP (Lisboa 2004) [forthcoming].

ARRUDA, A.M.; ALMEIDA, R.R. 1999: “Importações de vinho itálico para o território português: contextos, cronologias e significado”. In Économie et territoire en Lusitanie romaine. Actas da IIIe Table ronde sur la Lusitanie romaine. Madrid. Collection de la Casa de Velázquez. 307-337.

ARASA, F.; IZQUIERDO, I. 1998: “Estela antropomorfa con inscripción ibérica del Mas de Barberán (Nogueruelas, Teruel)”. AEspA 71, 79-102. ARCE, J. 1979: Zaragoza.

Caesaraugusta, ciudad

ARRUDA, A.M.; ALMEIDA, R.R. 2001: “Importação e consumo de vinho bético na colónia romana de Scallabis (Santarém, Portugal)”. In Actas del Congreso Internacional Ex Baetica Amphorae. Conservas, Aceite y Vino de la Bética en el Império Romano (Écija-Sevilha 1998), Écija, 703-715.

romana,

ARCELIN P.; TUFFREAU-LIBRE M. (ed.) 1998: La quantification des céramiques. Conditions et protocoles. Glux-en-Glenne: Centre Archéologique Européen du Mont Beuvray.

ARRUDA, A.M.; SOUSA, E. 2003: “Cerâmica de paredes finas da Alcáçova de Santarém”. Revista Portuguesa de Arqueologia. Lisboa, 6: 1, 235-286.

ARÉVALO, A.; BERNAL, D. (eds.) 2007: Las cetariae de Baelo Claudia. Avance de las investigaciones arqueológicas en el barrio industrial (2000-2004), Sevilla.

ARRUDA, A.M.; VIEGAS, C. 1999: “The roman Temple of Scallabis (Santarém-Portugal)”. Journal of Iberian Archaeology. Porto. I, 185-224.

ARHARBI, R.; LENOIR, É. 2002: “Les niveaux préromains de Banasa”. Ières Journées Marocaines d’archéologie et du patrimoine, Rabat, 1-45.

ARRUDA, A.M.; VIEGAS, C. 2004: “Les mortiers de l`Alcáçova de Santarém (Portugal)”. Actes du Congrès de Vallauris, 341-349.

ARHARBI, R.; NAJBI, H. 2004: “Les amphores de Khedis à l’époque maurétanienne, Méditerranée occidentale antique: les échanges”. III ANSER, Soveria Mannelli, 169-188.

ARRUDA, A.M.; VIEGAS, C.; BARGÃO, P. 2005: “As ânforas da bética costeira na Alcáçova de Santarém”. Revista Portuguesa de Arqueologia, vol. 8, nº1, Instituto Português de Arqueologia, Lisboa, p.279297.

ARRUDA, A.M. 1993: “Ocupação da Idade do Ferro da Alcáçova de Santarém no contexto da expansão fenícia para a fachada atlântica peninsular”. Estudos Orientais. (Actas do Encontro “Os fenícios no território português”). Lisboa. 4, 193-214.

ARRUDA, A.M.; VIEGAS, C.; BARGÃO, P. 2007: “Ânforas lusitanas da Alcáçova de Santarém”. In Actas do Colóquio de Homenagem a Françoise Mayet (Setúbal, Junho de 2004).

ARRUDA, A.M. 1994: “A Península de Lisboa entre o Norte atlântico e o Oriente mediterrânico”. In Lisboa subterrânea. Lisboa, 52-57.

ASENSIO, D.; MIRÓ, M.; SANMARTÍ, J. 2005: “Darreres intervencions arqueològiques al Castellet de Banyoles (Tivissa, Ribera d’Ebre): una ciutat ibèrica en el segle III AC”. Món ibèric als Països Catalans, XIII Col. Int. D’Arq. de Puigcerdà, 1, 615-627.

ARRUDA, A.M. 1997a: As cerâmicas áticas do Castelo de Castro Marim. Lisboa. ARRUDA, A.M. 1997b: “Os núcleos urbanos litorais da Idade do Ferro no Algarve”. In Noventa séculos entre a serra e o mar. Lisboa: Instituto Português do Património Arquitectónico e Arqueológico, 243255.

ASTURES 1995: Astures. Pueblos y culturas en la frontera del Imperio Romano, Gijón. AUBET, Mª E. 1997: Tiro y las colonias fenicias de occidente. Barcelona.

ARRUDA, A.M. 1997c: “O Castelo de Castro Marim: Um porto comercial na foz do Guadiana”. Resumo de comunicação apresentada no Colóquio sobre o

AURRECOECHEA, J. 2006: “Talleres dedicados a la producción de equipo militar en los campamentos 267

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

romanos de León, con especial referencia a los restos de lorica segmentata”. In A. Morillo (ed.), Arqueología Militar Romana en Hispania. Producción y abastecimiento en el ámbito militar, León, 309334. AUSTIN, N.; RANKOV, B. 1995: Exploratio. Military and political intelligence in the Roman world from the Second Punic war to the battle of Andrianopolis. London.

BELLIDO, MªP. 2007: “El abastecimiento de moneda al ejército romano en el NO. durante las guerras cántabras”. Larouco 4, 45-66.

AYAN VILA, X.M. (coord.) 2006: Os castros de Neixón, Noia. AYAN VILA, X.M. (coord.) 2008: Os castros de Neixón II, Noia.

BELTRÁN, F.; PINA, F. 1994: “Roma y los Pirineos: la formación de una frontera”. Chiron 24, 103-133.

BELLO, J.M. 1991a: “La investigación sobre la Torre de Hércules”, Ciudad y Torre. Roma y la Ilustración en La Coruña, La Coruña, 165-170. BELLO, J.M. 1991b: “La torre romana”, Ciudad y Torre. Roma y la Ilustración en La Coruña, La Coruña, 171178.

BELTRÁN, J. 2002: “La arquitectura funeraria en la Hispania meridional durante los siglos II AC y I DC”. D. Vaquerizo, ed., Espacios y usos funerarios en el Occidente romano, Córdoba, 293-328.

BADIAN, E. 1997: “Notes on a Recent List of Praefecti Fabrum under the Republic”. Chiron 27, 1-19. BADIE, A.; GAILLEDRAT, E.; MORET, P.; ROUILLARD, P.; SÁNCHEZ, MªJ.; SILLIÈRES, P. 2000: Le site antique de La Picola à Santa Pola (Alicante), Casa de Velázquez, Madrid.

BELTRÁN, M. 1970: Las ánforas romanas en España, Zaragoza. BELTRÁN, M. 1983: “El aceite en Hispania a través de las ánforas: la concurrencia del aceite italiano y africano”. In Producción y Comercio de Aceite en la Antigüedad, II Congreso, Madrid, 515-550.

BALIL, A. 1965: “Funcionarios subalternos en Hispania durante el Imperio Romano”. Emerita 33, 297-319. BALIL, A. 1968: “Economia de la Hispania Romana”. In Estudos de Economia Antigua de la Península Ibérica: Ponencias Presentadas a la 1ª Reunion de Historia de la Economia Antigua de la Península Ibérica. Barcelona, 289-370.

BELTRÁN, M. 1990: Guía de cerámica romana. Zaragoza. BELTRÁN, M. (dir.) 1995: Azaila, Institución Fernando el Católico, Zaragoza. BELTRÁN, M. 2008: “La producció i el comerç de les àmfores de la “Provincia Hispania Tarraconensis”: homenatge a Ricard Pascual i Guasch: Actes de les jornades d’estudi celebrades al Palau Marc de la Generalitat de Catalunya els dies 17 i 18 de novembre de 2005 coord. por A. López Mullor y J.Aquilué, Barcelona, pp. 271-318.

BALIL, A. 1972: “Economía de la Hispania romana (s. IIII DC)”. Studia Archaeologica 15, Santiago de Compostela. BALIL, A. 1974: “De nuevo sobre Galicia y sus relaciones marítimas durante la época imperial romana”. III Congresso Nacional de Arqueología. Porto, 1973, 211-221.

BELTRÁN, M.; FATÁS, G. 1998: Caesar Augusta, ciudad romana, Historia de Zaragoza, II, Zaragoza.

BALIL, A. 1980: “¿Restos de un puerto romano en La Coruña?”, Brigantium 1, 167-171.

BENCIVENGA, C. 1984: “La ceramica iberica da VeliAContributo allo studio della diffusione ceramica iberica in Italia”. DAINST(M) 25, 20-33.

BALIL, A. 1989: “De la escultura romano-ibérica a la escultura romano-republicana”. In J. González, (ed.) Estudios sobre Urso Colonia Iulia Genetiva, Sevilla, 223-231. BALIL, A. 1990: “Comunicaciones marítimas”, El comercio marítimo romano en el Mediterráneo occidental (Coloquio Int., Barcelona, 1988), PACT 27, 63-72. BARROS, L.; CARDOSO, J.L.; SABROSA, A. 1993: “Fenícios na Margem Sul do Tejo. Economia e Integração Cultural do Povoado do Almaraz – Almada”. Estudos Orientais. Lisboa: Instituto de Estudos Orientais da Universidade Nova de Lisboa, IV, p. 143-181.

BENDALA, M., 1990: “El plan urbanístico de Augusto en Hispania. Precedentes y pautas macroterritoriales”. In W. Trillmich, P. Zanker, coord., Stadtbild und Ideologie. Die Monumentalisierung hispanischer Städte zwischen Republik und Kaiserzeit, München, 25-42. BÉRARD, F. 1984: “La carrière de Plotius Grypus et le ravitaillement de l’armée impériale en campagne”. MEFR 96, 259-324. BERCHEM, D. Van 1937: L’annone militaire dans l’Empire romain au IIIème siècle, Paris.

BECKER, C.; CONSTANTIN, C.; VILLEDIEU, F. 1989: “Types d’amphores en usage à Lugdunum du Ier au VIè siècle”. CEFR 114, 656-659.

BERCHEM, D. Van 1939: Les distributions de blé et d’argent à la plebe romaine sous l’Empire. Geneve.

BEKKER-NIELSEN, T. (ed.) 2005: Ancient Fishing and Fish Processing in the Black Sea region, Black Sea Studies 2, Aarhus. BELLET, M. 1976: “Les huilieres gallo-romaine de Provence”. Archéologia 92, 53-59.

BERCHEM, D. Van 1977: “L’annone militaire est-elle un mythe?” In Armée et fiscalité dans le monde antique. Paris, 331-339. BERMEJO, J.C. 1978: La sociedad en la Galicia casteña. Santiago.

268

BIBLIOGRAPHY

intervención arqueológica preventiva en la Plaza de África nº 3, Consejería de Educación y Cultura de la Ciudad Autónoma de Ceuta, Ceuta [unpublished].

BERNAL, D. 2006: “La industria pesquero-conservera en el Círculo del Estrecho. Consideraciones sobre la geografía de la producción”, Internacional Conference L’Africa Romana XVI (Rabat, diciembre de 2004), II, pp. 1351-1394.

BERNAL, D.; PÉREZ, J.M. 1999: Un viaje diacrónico por la Historia de Ceuta. Resultados de las excavaciones arqueológicas en el Paseo de las Palmeras, Madrid. BERNAL, D., RAISSOUNI, B., EL KHAYARI, A., ES SADRA, L., DÍAZ RODRÍGUEZ, J.J., SÁEZ, A.M., BUSTAMANTE, M., VILLADA, F., LAGÓSTENA, J., DOMÍNGUEZ PÉREZ, J.C.; PARODI, M.J. 2008: “El valle del río Martil en época preislámica e islámica. Primeros resultados de la Carta Arqueológica (2008)”, in D. Bernal, B. Raissouni, J. Ramos, M. Zouak & M. Parodi eds., En la orilla africana del Círculo del Estrecho. Historiografía y proyectos actuales, Colección de Monografías del Museo Arqueológico de Tetuán 2, Madrid, pp. 313-349.

BERNAL, D. 2007: “Algo más que garum. Nuevas perspectivas sobre la producción de las cetariae hispanas al hilo de las excavaciones en c/ San Nicolás (Algeciras, Cádiz)”, Congreso Internacional CETARIAE. Salsas y salazones de pescado en Occidente en la Antigüedad, BAR, i.s. 1686, Oxford, pp. 93-107. BERNAL, D. 2008: “El final de la industria pesqueroconservera en Hispania (ss. V-VII). Entre obispos, Bizancio y la evidencia arqueológica”, Ressources et activités maritimes des peuples de L’Antiquité, Les Cahiers du Littoral 2, nº 6, Université du Littoral Côte d’Opale, Boulogne-sur-Mer, pp. 31- 55.

BERNAL, D; RIBERA, A. (eds.) 2008: Cerámicas hispanorromanas. Un estado de la cuestión, Cádiz.

BERNAL, D. (ed.) 2009: Las factorías de salazones de Traducta, Universidad de Cádiz and Fundación Municipal de Cultura José Luis Cano de Algeciras, Algeciras, in press.

BERNAL, D., ROLDÁN, L., BLÁNQUEZ, J., DÍAZ, J.J.; PRADOS, F. 2008: “Un taller de púrpura tardorromano en Carteia (Baetica, Hispania). Avance de las excavaciones preventivas en el conchero de Villa Victoria (2005)”, in C. Alfaro y L. Karali eds., Purpureae Vestes, II Symposium Internacional sobre textiles y tintes del Mediterráneo en el Mundo Antiguo, Valencia, pp. 209-226.

BERNAL, D. & ARÉVALO, A. 2008: “Baelo Claudia y sus industrias haliéuticas. Síntesis de las últimas actuaciones arqueológicas (2000-2004)”, Ressources et activités maritimes des peuples de L’Antiquité, Les Cahiers du Littoral 2, nº 6, Université du Littoral Côte d’Opale, Boulogne-sur-Mer, pp. 9-30.

BERNAL, D; RIBERA, A. (eds.) 2008: Cerámicas hispanorromanas. Un estado de la cuestión, Cádiz.

BERNAL, D., ARÉVALO, A., AGUILERA, L., LORENZO, L., DÍAZ, J.J.; EXPÓSITO, J.A. 2007: “La topografía del barrio industrial. Baelo Claudia, paradigma de la industria conservera urbana hispanorromana”, in Arévalo, A. & Bernal, D., eds., Las cetariae de Baelo Claudia. Avance de las investigaciones arqueológicas en el barrio meridional (2000-2004), Sevilla, pp. 91-221.

BERNI, P. 1998: Las ánforas de aceite de la Bética y su presencia en la Cataluña romana, Col.leció Instrumenta 4, Barcelona. BERTHIER, A. 1981: La Numidie. Rome et le Maghreb. Paris. BEZECZKY, T. 1987: Roman amphorae from the Amber route in Western Pannonia. BAR Int.Ser. 386, Oxford. BIEGERT, S.; LAUBERT, J. 1997: “Zur sigillatabelieferung am vorderen limes”. In W. Groenman-van Waateringe, B.L. van Beek, W.J.H. Willems & S.L. Wynia (eds.), Roman Frontier Studies 1995. Proceedings of the XVIth International Congress of Roman Frontier Studies, Oxbow Monograph 91, Oxford, 295-299.

BERNAL, D., IGLESIAS, L.; LORENZO, L. 2009: “La zona de acceso a los embarcaderos de Traducta. Intervención arqueológica en c/ Méndez Núñez nº 4 (Algeciras)”, in D. Bernal, ed., Las factorías de salazón de Traducta. Primeros resultados de las excavaciones arqueológicas en la c/ San Nicolás (Algeciras, Cádiz), Universidad de Cádiz and Ayuntamiento de Algeciras, Algeciras, in press. BERNAL, D.; JIMÉNEZ-CAMINO, R. 2004: “El taller de El Rinconcillo en la Bahía de Algeciras. El factor itálico y la economía de exportación (s. I AC – s. I DC)”, Internacional Conference Figlinae Baeticae. Talleres alfareros y producciones cerámicas en la Bética romana (ss. II AC – VII DC), BAR International Series 1266, II, Oxford, pp. 589-606.

BIGAGLI, C. 2000: “Ancore”, in S. Bruni, Le navi antiche di Pisa. Ad un anno dall’inizio delle ricerche, Florencia. BILL, J.; BIRTHE, L.C. (eds.) 1999: “Maritime Topography and the Medieval Town”. Papers from the 5th International Conference on Waterfront Archaeology in Copenhagen, 14-16 May 1998. Copenhage: PNM Publications from The National Museum. Studies in Archaeology & History, vol. 4.

BERNAL, D., JIMÉNEZ, R., LORENZO, L., TORREMOCHA, A.; EXPÓSITO, J.A. 2003: “Las industrias de salazón de Traducta. Espectaculares hallazgos arqueológicos en la c/ San Nicolás 3-5 de Algeciras”, VII Jornadas de Historia del Campo de Gibraltar, Almoraima 29, pp. 163-183.

BIRLEY, A.R. 1981a: The Fasti of Roman Britain, Oxford. BIRLEY, A.R. 1981b: “The economic effects of Roman frontier policy”. In A. King & M. Henig (eds.), The

BERNAL, D., LORENZO, L., SÁEZ, A.M.; BUSTAMANTE, M. 2007 b: Informe preliminar de la 269

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

BLOT, J.-Y. 2003: De “Terra Nullius” a Parque de Reserva: a partilha do espaço no Fundeadouro da Ilha Berlenga. Lisboa: Academia de Marinha.

Roman west in the Third century, BAR Int. Series 109, Oxford, 39-53. BIRLEY, A.R. 1997: “Supplying the Batavians at Vindolanda”. In W. Groenman-van Waateringe, B.L. van Beek, W.J.H. Willems & S.L. Wynia (eds.), Roman Frontier Studies 1995. Proceedings of the XVIth International Congress of Roman Frontier Studies, Oxbow Monograph 91, Oxford, 273-280.

BLOT, J.-Y. 2004: “Le paléoestuaire du centre de Lisbonne (Portugal): sondages récents dans le cadre des travaux d’aménagement urbain”. AestuariACordemais: Estuarium, 2004, 193-216. BLOT, J.-Y. 2005: “A diacronia do Fundeadouro da Berlenga”. In Actas das I Jornadas de Arqueologia e Património da Região de Peniche (…), 886137.

BISHOP, M.C. 1999: “Praesidium: social, military, and logistical aspects of the Roman army’s provincial distribution during the early Principate”. In A. Goldsworthy; I. Haynes, (eds.), The Roman Army as a Community, Portsmouth, Rhode Island 1999, 111118. BLANCO, A. 1992: “Monumentos romanos de la conquista de Galicia”. La Romanizacion de GaliciACorunha, Cuadernos del Seminário de Estudios Cerámicos de Sargadelos, 16, (2ª edição), 95-104. BLÁZQUEZ CERRATO, C. 1999: “Notas sobre la contramarca cabeza de águila y su distribución geográfica en el territorio peninsular”, Anejos AEspA XIX (Actas II Encuentro Penínsular de Numismática Antigua, Oporto, 1998), Madrid, 91-100.

BLOT, J.-Y.; BLOT, M.L.P. 1990-1992a: “De la Glaciation de Würm aux derniers temps de la marine à voile: Éléments pour une carte archéologique du patrimoine immergé au Portugal”. O Arqueólogo Português, Série IV, 8/10, 1990-1992, 425-454. (Coautoria com J.-Y. Blot). BLOT, J.-Y.; BLOT, M.L.P. 1990-1992b: “Problématique d’une recherche entreprise pour le compte du Musée National d’Archéologie (M.N.A. de Lisbonne) dans le cadre du Programme de Recensement du Patrimoine Culturel Immergé”. O Arqueólogo Português, Série IV, 8/10, 1990-1992, 469-485. (Co-autoria com J-Y. Blot) BLOT, J.-Y.; PATA, A.J.; SALMÍNEN, M.; CALEJA, P.; ALELUIA, M.; GONÇALVES, P.; BETTENCOURT, J.; ROBB, G.; SWANSON, C.; JONES, C.; DELAUZE, H.G.; PACHOUD, G.; NERHOT, P.; REAL, F., BLOT, M.L.P.; ALVES, F.J.S. 2005: “O fundeadouro da Berlenga”. In Actas do Congresso A Presença Romana na Região Oeste. 31-55. BLOT, M.L.P. 2003a: Os portos na origem dos centros urbanos. Contributo para a arqueologia das cidades marítimas e fluvio-marítimas em Portugal. Lisboa: Instituto Português de Arqueologia. Trabalhos Arqueológicos, nº 28, Lisboa. http://www.ipa.mincultura.pt/pubs/TA/folder/28.

BLÁZQUEZ CERRATO, C. 2002: Circulación monetaria en el área occidental de la península ibérica, Archéologie et Histoire Romaine 6, Montagnac. BLÁZQUEZ CERRATO, C. 2006: “Circulación monetaria en Asturica Augusta”, MªP. García-Bellido (coord.), Los campamentos romanos en Hispania (27 AC-192 DC). El abastecimiento de moneda, Anejos de Gladius 9, Madrid, 118-137. BLÁZQUEZ, J.M. 1975: “Economía y sociedad”. In Tovar, A.; Blázquez, J.M. (eds.) Historia de la Hispania Romana: La Península Ibérica Desde 218 AC Hasta el Siglo V. Madrid: Alianza Editorial, 193348. BLÁZQUEZ, J.M. 1997: “Vías e itinerarios, de la Antigüedad al mundo romano”. In Gabinete de Antigüedades de la Real Academia de Historia.

BLOT, M.L.P. 2003b: “O Rio Tejo e a circulação aquática. Materiais submersos e breve história de um complexo portuário”. (Colaboração: S. Rodrigues). In Vila Franca de Xira, Tempos do Rio, Ecos da TerrACatálogo da Exposição. Vila Franca de Xira: Câmara Municipal. 2003. 81-94.

BLÁZQUEZ, J.M.; REMESAL, J. (eds.) 1999: Estudios sobre el Monte Testaccio (Roma) I. Barcelona. BLÁZQUEZ, J.M.; REMESAL, J. (eds.) 2001: Estudios sobre el Monte Testaccio (Roma) II. Barcelona.

BLOT, M.L.P. 2003c: “Problemáticas da Arqueologia Náutica e Portuária no quadro do estudo de portos antigos e medievais em Portugal”. Actas do Colóquio Portos medievais do Mediterrâneo. Portos antigos do Mediterrâneo (Mértola, 19 de Outubro de 2001). Arqueologia Medieval. Porto: Edições Afrontamento, nº 9, 207-220.

BLÁZQUEZ, J.M.; REMESAL, J. (eds.) 2003: Estudios sobre el Monte Testaccio (Roma) III. Barcelona. BLOT, J.-Y. 2002: “Elementos para a tonelagem dos navios na costa ibero-atlântica na Antiguidade: o testemunho dos vestígios de âncoras (cepos de chumbo)”. In Terrenos da Arqueologia da Península Ibérica, sessão 35, Actas do III Congresso da Arqueologia Peninsular. Vila Real: UTAD, 8, 571594.

BLOT, M.L.P. 2004a: Estuaires et histoire urbaine. Problématique portuaire et présence phénicienne en territoire portugais. AestuariACordemais: Estuarium. 2004, 171-192.

BLOT, J.-Y. 2002b “New Courses in Maritime Archaeology in Portugal”. In Ruppé, C.V., Barstad, J. F. (eds.) – International Handbook of Underwater Archaeology. New York: Kluwer Academic/Plenum Publishers, 465-495.

BLOT, M.L.P. 2004b: Paléogéographie côtière et activités portuaires sur le littoral portugais: une archéologie de contacts méditerranéens sur la façade atlantique. In De Maria, L. e Turchetti, R. (Eds.) (2004) – Evolución paleoambiental de los puertos y 270

BIBLIOGRAPHY

BOUBE, E. 1996: Le trophée augustéen, Collection du Musée Départamental de Saint-Bertrand-de-Comminges 4, Saint-Bertrand-de-Comminges.

fondeaderos antiguos en el Mediterraneo Occidental. I Seminario International ANSER, intitulado “El patrimonio Arqueológico Submarino y los Puertos Antiguos”. Alicante, 14-15 de Novembro de 2003. Roma: Rubbettino, 37-60.

BOUNEGRU, O. 1997: “Considerations concernat le transport des marchandises et le ravitaillement de l’armee sur le limes du Bas-Danube”. In W. Groenman-van Waateringe, B.L. van Beek, W.J.H. Willems & S.L. Wynia (eds.), Roman Frontier Studies 1995. Proceedings of the XVIth International Congress of Roman Frontier Studies, Oxbow Monograph 91, Oxford, 311-313.

BLOT, M.L.P. 2004d: Présences méditerranéennes antiques sur les côtes du Portugal. La problématique portuaire et les vestiges d’importations. In Zevi, A.G. e Turchetti, R. (Coord.) (2004) – Méditerranée Occidentale Antique: Les Échanges. III Séminaire International ANSER. Marseille, 14-15 de Maio de 2004. Roma: Rubbettino, 203-212.

BOWMAN, A.K. 1994: Life and letters on the Roman frontier. London.

BLOT, M.L.P. 2004e: “Circulação aquática e o papel dos portos fluvio-estuarinos nos contactos da Lusitânia Romana. O caso do litoral e dos rios de Portugal”. In V Mesa Redonda Internacional sobre Lusitania Romana: Las Comunicaciones. Cáceres. 465480.

BOWMAN, A.K.; THOMAS, J.D. 1983: Vindolanda: The Latin writing-tablets, London. BOWMAN, A.K.; THOMAS, J.D.; ADAMS, J.N. 1990: “Two letters from Vindolanda”. Britannia 21, 33-52.

BLOT, M.L.P. Arqueologia Urbana e Arqueologia do Meio Aquático. A problemática portuária como “ponte” entre dois territórios de investigação. IV Congresso de Arqueologia Peninsular. Faro, 15-17 de Setembro 2004 [forthcoming].

BRANDT, R.; SLOFSTRA, J. (eds.) 1983: Roman and native in the low countries: spheres of interaction. BAR Int.Ser. 184, Oxford.

BLOT, M.L.P.; ALVES J.G. 2004: “Représentations cartographiques et iconographiques des principaux ports de l’Algarve post-médiéval. Les premières expériences archéologiques de sites portuaires dans le territoire portugais”. In De Maria e Turchetti, R. (Coord.) (2004) – Rotte e Porti del Mediterraneo dopo la Caduta dell’Imperio Romano d’Occidente. IV Seminario ANSER. Genova, 18-19 de Junho de 2004. Roma: Rubbettino, 189-216.

BREEZE, D.J. 1993: “Demand and supply on the Northern Frontier”. In D. Breeze & B. Dobson (eds.), Roman Officers and Frontiers, Mavors Army Roman Researchs X, Sttutgart, 526-552.

BRÉAL, M. 1895: “Inscriptions from Curubis” CRAI, IV, 23, 31-34.

BREEZE, D.J. 2000: “Supplying the Army”. In G. Alföldy, B. Dobson y W. Eck (eds.), Kaiser, Heer und Gesellschaft in der Rómischen Kaiserzeit. Gedenkschrift für Eric Birley, Stuttgart, 59-64. BREEZE, D.J. 1986-7: “The logistics of Agricola’s final campaign”. Talanta XVIII-XIX, 1986-1987, 7-28.

BOLUFER, J. 1999: “L’arqueologia subacuàtica a Xàbia”. Revista de festes patronals.

BROTONS, J.F.; MURCIA, A.J. 2006: “El castellum tardorrepublicano del Cerro de las fuentes de Archivel (Caravaca de la Cruz, Murcia). Estudio preliminar”. In A. Morillo (ed.), Arqueología militar romana en Hispania. Producción y abastecimiento en el ámbito militar, León 2006, 639-653.

BONET, H.; MATA, C. 2002: El Puntal dels Llops, un fortín edetano, T.V. del SIP 99, Valencia. BONIFAY, M. 2005: Etudes sur la céramique romaine tardive d’Afrique, BAR International Series 1301, 2005, Oxford. BONIFAY, M.; CONGRES, C.; LEGUILLOUX, M. 1989: “Amphores tardives (V-VII siècle) à Arles et à Marseile. CEFR 114, 660-663.

BROUGHTON, T.R.S. 1952: The Magistrates of the Roman republic., vol.2. New York. BROWN, A.G. 1997: Alluvial geoarcheology: floodplain archaeology and environmental change. Cambridge.

BONSOR, G. 1935: The Archaeological Expedition Along the Guadalquivir 1889-1901, New York. Traducción española de Chic García, G. e Padilla Monge, A. (1989): Expedición arqueológica a lo largo del Guadalquivir, Écija.

BRUN, J.P. 1986: L’oleiculture antique en Provence. Les huilieres de department de Var. RAN, Suppl.15. Paris. BRUN, J.P. 1997: “Production de l’huile et du vin en Lusitanie romaine”. Conimbriga 36, pp.45-72.

BOOTH, K. 2007: “The Roman Pharos at Dover Castle”, English Heritage Historical Review 2, 7-21.

BUGALHÃO, J.; LOURENÇO, S. 2005: “Vestígios romanos na Ilha da Berlenga”. In Actas do Congresso “A Presença Romana na Região Oeste, 57-63.

BOTTE, E. 2007: “Les salaisons de poissons de Cumes (Italie) au Ier siècle de notre ère: une nouvelle inscription peinte sur amphore Dressel 21/22”, Ressources et activités maritimes des peuples de L’Antiquité, Les Cahiers du Littoral 2, nº 6, Université du Littoral Côte d’Opale, Boulogne-surMer, pp. 443-446.

BULLO, S. 2002: Provincia Africa. Le città e il territorio dalla caduta di Cartagine à Nerone, Roma.

BOUBE, J., 1987-1988: “Les amphores de Sala à l’époque maurétanienne”, BAM XVII, 183-207.

BURÓN, M. 1997: El trazado urbano en las proximidades del Foro en Asturica Augusta. La casa del

BURILLO, F. 1997: “Textos, cerámica y ritual celtibérico”. Kalathos 16, 223-242.

271

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

CARDOSO, G.; MARQUES, J.L.; RODRIGUES, S. 1998: “Forno romano de cerâmica descoberto em Peniche”. Al-madan, II série: 7, 178-179. CARDOSO, G.; RODRIGUES, S. 1998: “4ª Campanha de Sondagens Arqueológicas na Olaria Romana do Morraçal da Ajuda (Peniche)”. Al-madan, II série: 11, p.6. CARDOSO, G.; RODRIGUES, S. 2005: “Olaria Romana do Morraçal da Ajuda (Peniche – Portugal)”. In Actas do Congresso “A Presença Romana na Região Oeste, 83-102. CARMONA, P. 2002: “Geomorfología de la llanura valenciana. El río Turia y la ciudad”. Historia de la ciudad II, 7-28. CARRERAS, C. 1994: Una reconstrucción del comercio de cerámicas: la red de transporte en Britannia. Barcelona.

pavimiento de opus signinum, Arqueología en Castilla y León 2, Valladolid. BUSTAMANTE ÁLVAREZ, M.; MARTÍN-ARROYO SÁNCHEZ, D. 2004: “La producción de ánforas greco-itálicas de imitación y su imitación en la bahía gaditana durante el siglo II A. C.: los contextos de la avenida de Pery Junquera en San Fernando (Cádiz).”, Figlinae Baeticae. Talleres alfareros y producciones cerámicas (siglos II AC – VII DC), eds. D. Bernal Casasola & L. Lagóstena Barrios, BAR Int. Ser., 1266, Oxford, 441-446. CAAMAÑO, J.M. 1979: “Posible reutilización de caminos prerromanos en época romana”. Gallaecia, 3/4, 281-285. CAAMAÑO, J.M. 1989: “Estampillas de la cohors I Celtiberorum halladas en el campamento romano de Cudadela”, Gallaecia 11, 209-229. CÁCERES, I. 2007: “La ganadería en el desarrollo económico de la factoría de salazones”, in Arévalo, A. & Bernal, D., Las cetariae de Baelo Claudia. Avance de las investigaciones arqueológicas en el barrio industrial (2000-2004), Sevilla, pp. 499-511.

CARRERAS, C. 1996: “El comercio en Asturia a través de las ánforas”. In C. Fernández Ochoa (coord.) Los Finisterres Atlánticos en la Antigüedad. Gijón, 1995, 205-210. CARRERAS, C. 1997: “Los beneficiarii y la red de distribución militar en Britannia e Hispania”. Gerión 15, 151-176.

CADIOU, F. 2003: “Garnisons et camps permanents: un réseau défensif des territoires provinciaux dans l’Hispanie républicaine?”. In A. Morillo, F. Cadiou, D. Hourcade (eds.), Defensa y territorio en Hispania de los Escipiones a Augusto. Espacios urbanos y rurales, municipales y provinciales. Coloquio celebrado en la Casa de Velázquez, 19 y 20 de marzo de 2001, Madrid-León 2003, 81-100.

CARRERAS, C. 2000a: Economía de la Britannia romana. La importación de alimentos. Barcelona. CARRERAS, C. 2000b: “Producciones de Haltern 70 y Dressel 7-11 en las inmediaciones del Lacus Ligustinus (Las Marismas, bajo Guadalquivir)”, Ex Baetica Amphorae. Conservas, aceite y vino de la Bética en el Imperio Romano, vol. I Écija, 419-426.

CALVO DELCÁN, C. 1990: Opiano. De la Caza. De la Pesca. Anónimo. Lapidario órfico, Biblioteca Clásica Gredos, 134, Madrid.

CARRERAS, C. 2001: “Ánforas de la Campa Torres (Campañas 1985-1997)”. In J.L. Maya and F. Cuesta (eds.) El castro de la Campa Torres. Período prerromano. Gijón, 389-392.

CAMINO MAYOR, J., ESTRADA GARCÍA, R.; VINIEGRA PACHECO, Y. 2002: “El campamentu romanu del Castiillu La Carisa, una avanzada militar n’Asturia Transmontana”, Asturies 12, 24-29.

CARRERAS, C. 2003: “Aprovisionamiento del soldado romano en campaña: la figura del praefectus vehiculorum”. Habis 25, 291-311.

CAMPOMANES, E. 1998/99: “Hallazgo de un complejo metalúrgico romano en la ciudad de León. Excavaciones en la calle Plegarias con vuelta a la calle Ramiro III en la ciudad de León, Lancia 3, 269280.

CARRERAS, C. 2006a: “Transition of the military supply from the Late Republic to Augustan period: a case study of the NW of the Iberia Peninsula”. In A. Ñaco i I. Arrayás (eds.) War and territory in the Roman world. Guerra y territorio en el mundo romano. BAR Int.Ser. 1530, Oxford, 169-178.

CAMPOMANES, E. 2006 “El acueducto de la legio VII gemina y sus producciones latericias”, A. Morillo (ed.), Arqueología Militar Romana en Hispania. Producción y abastecimiento en el ámbito militar, León, 439-454. CANTÓ, A. Mª 1990: “Las tres fundaciones de Augusta Emerita”, Stadtbild und Ideologie (Die Monumentalisierung hispanischer Städte zwischen Republik una Kaiserzeit. Kolloquium Madrid, 1987), München, 289-297.

CARRERAS, C. 2006b: “A quantitative approach to the amphorae from Xanten: a more”. In Römische Amphoren der Rheinprovinzen unter Besonderer Berücksichtigung des Xantener Materials. Regionalmuseum Xanten, 13-15 Januar 2004. Xantener Berichte, band 14, Landschaftsverband Rheinland, Archäologischer Parx, Regionalmuseum Xanten, Mainz am Rhein, 25-39.

CARDOSO, J.L. 2001: “Achados subaquáticos de defesas de elefante, prováveis indicadores do comércio púnico no litoral português”. In Os Púnicos no Extremo Ocidente (Actas do colóquio internacional). Lisboa: Universidade Aberta, p. 261282.

CARRERAS, C. 2007: “Consumo de salazones béticos desde época de Augusto a los Julio-Claudios: mercados emergentes en Asturica Augusta (Astorga), Barcino (Barcelona) y Oppidum Cugernorum (Xanten)”. In Actas del Congreso Internacional CETARIAE. Salsas y salazones en Occidente durante 272

BIBLIOGRAPHY

y Salazones de Pescado en Occidente durante la Antigüedad, BAR International Series 1686, 2007, 355-365.

la Antigüedad. Universidad de Cádiz (noviembre 2005). BAR Int. Ser. 1686, Oxford, 215-220. CARRERAS, C. 2009: “Del Mujal a Xanten: noves visions del comerç del vi a la Tarraconense”. In M. Prevosti and A. Martin (ed.) El vi tarraconense i laietà: ahir i avui. Tarragona, pp.166-177.

CAZEILS, N. 2000: Dix siècles de pêche à la baleine, Rennes. CEBOLLA, J.L. et al. 2004: “La excavación arqueológica de las Tenerías, nº 3-5 (Zaragoza)”. Salduie 4, 463-472.

CARRERAS, C.; AGUILERA, A.; BERNI, P.; GARROTE, E.; MARIMON, P.; MORAIS, R.; MOROS, J.; NIETO, X.; PUIG, A.; REMESAL, J.; ROVIRA, R.; VIVAR, G. 2005: Culip VIII i les àmfores Haltern 70. Monografies CASC 5, Girona.

CENTENO, R.M.S. 1987: Circulação monetária no noroeste de Hspânia até 192, Anexos Nummus 1, Porto.

CARRERAS, C.; BERNI, P. 2003: “Ánforas”. In T. Amaré (ed.) Lucernas y ánforas. Axtorga IV. Universidad de León. 653-673.

CERATI, A. 1975: Caractère annonaire et assiette de l’import foncier au Bas-Empire, Paris. CERILLO, E. 1984: La vida rural romana en ExtremadurACáceres.

CARRERAS, C.; FUNARI, P.P.A. 1998: Britannia y el Mediterráneo: Estudios sobre el abastecimiento de aceite bético y africano en Britannia. Barcelona.

CERRI, L., 2007: “La produzione e il commercio di salsamenta della Mauretania Tingitana (I secolo DC)”. In Lagóstena (L.), Bernal (D), Arévalo (A) (eds.), Cetariae 2005. Salsas y Salazones de Pescado en Occidente durante la Antigüedad, BAR International Series 1686, 2007, 195-204.

CARRETERO, S. 1993: “El cuadrante noroeste peninsular en época romana: los efectivos militares y sus establecminetos”. Brigencio 3, 47-73. CARRETERO, S. 2000: El campamento romano del ala I Flavia en Rosinos de Vidriales (Zamora). La cerámica, Madrid.

CERVERA POZO, L., DOMÍNGUEZ BERENGENO, E.; GARCÍA VARGAS, E. 2007: “Estructuras de época romana en c/ Santa Verania 22”, Ilipa Antiqua. De la Prehistoria a la época romana., eds. E. Ferrer Albelda, A. Fernández Flores, J.L. Escacena Carrasco & A. Rodríguez Azogue, Alcalá del Río, 295-310.

CARRETERO, S. 2001: “El abastecimiento de productos cerámicos a la guarnición de la legio X gemina en Petavonium (Rosinos de Vidriales, Zamora, España)”, Rei Cretariae Romanae Fautorum. Acta 37 (Lyon, 2000), Abbingdon, 157-162.

CHAVES, F. 2005: “Guerra y moneda en la Hispania del Bellum Civile”. In E. Melchor et al. (eds.), Julio César y Corduba: tiempo y espacio en ela campaña de Munda. 49-45 AC, Córdoba. 207-245.

CARVALHO, T.P.; GUIMARÃES, C.; BARROCA, M.J. 1996: Bairro da Sé do Porto: contributo para a sua caracterização histórica. Porto. CASAS, J. et al. 1994: El món rural d’època romana a Catalunya. Girona.

CHAVES, F., GARCÍA VARGAS, F., FERRER, E., 1997: “Datos relativos a la pervivencia del denominado ‘Círculo del Estrecho’ en época republicana”, L’Africa Romana XII, 3, Sassari, 13071320. CHEDDAD, A., 2004: “Cohésion et desagrégations dans le circuit du détroit de Gibraltar”, L’Africa Romana XV, 989-1010. CHIC, G. 1981: “Rutas comerciales de las ánforas olearias hispanas en el occidente romano”, Habis 12, 223-350. CHIC, G. 1985: “Aspectos económicos de la Política de Augusto en la Betica”, Habis, 16, Universidad de Sevilla, p. 277-293. CHIC, G. 1988: La Epigrafía Anfórica de la Bética, Vol. II, Sevilla. CHIC, G. 1990 La Navegación por el Guadalquivir Entre Córdoba y Sevilla en Época Romana, Écija.

CASTANYER, P. 2006: “Les arts de pesca a Empuries”, Pescadors de l’antigua Empuries, Gerona, pp. 20-22. CASTELO-BRANCO, F. 1957: “Alguns Aspectos da Evolução do Litoral Português”. Boletim da Sociedade de Cartografia de Lisboa. Lisboa. Jul- Set. 1957, 337-353. CASTELO-BRANCO, F. 1989: “Evolução do litoral português”. In Serrão, J, (dir.) – Dicionário de História de Portugal. Porto: Livraria Figueirinhas, IV, 39-43. CASTRO, J.C. 1998: Sondaxes arqueolóxicas na parcela nº 14 da Unidade de Execución Rosalía de Castro I (Vigo), Informe Valorativo, Anta de Moura S.L. Informe inédito depositado en el Servicio de Arqueología de la Xunta de Galicia, Santiago de Compostela [unpublished]. CASTRO, J.C. 1999: Excavación arqueológica en área na parcela nº 14 da U.E. I-05 Rosalía de Castro 1, Vigo, Anta de Moura S.L. Informe inédito depositado en el Servicio de Arqueología de la Xunta de Galicia, Santiago de Compostela [unpublished].

CHIC, G. 1993 “La navegación fluvial en época romana” Revista de Arqueologia, 142, 28-39. CHIC, G. 1994 “Economía y Sociedad en la Bética Altoimperial. El Testimonio de la Epigrafía Anforica. Algunas Notas”, in González Roman, C. (ed.) La Sociedad de la BéticAContribuciones Para su Estudio, Universidad de Granada, Granada, p. 75122.

CASTRO, J.C. 2007: “La salina romana del yacimiento de “O Areal”, Vigo (Galicia)”. In Lagóstena (L.), Bernal (D), Arévalo (A) (eds.), Cetariae 2005. Salsas

273

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

culturali, Roma Castel Sant’Angelo, 11-25 Aprile, 1991, Roma 1991, 23-31; CORTEGOSO, M. 2004: Informe Memoria Técnica Control arqueológico de prolongación del Túnel del Arenal, Vigo (Pontevedra), Informe inédito depositado en el Servicio de Arqueología de la Xunta de Galicia, Santiago de Compostela [unpublished].

CHIC, G. 1995a: “Un Factor Importante en la Economía de la Bética: el Aceite”. Hispania Antiqua, XIX, Universidad de Valladolid, p. 95-128. CHIC, G. 1995b: “Roma y el mar: del Mediterráneo al Atlántico”, Guerra, exploraciones y navegación: del mundo antiguo a la Edad Moderna, A Coruña, 55-90. CHIC, G. 1997a: Historia económica de la Bética durante la época de Augusto, Sevilla.

COSTA, P.; CASTELLÓ, J.C. 1999: “La Cultura Ibérica: poblamiento y hábitat”. Historia de la Marina Alta. Alicante, 97-108. CURCHIN, I. 1983: “Personal wealth in Roman Spain”. Historia XXXII.2, 227-244. CURRAS, B. 2007: “Aportación al conocimiento de la industria de salazón en las Rías Baixas Gallegas”. In Lagóstena (L.), Bernal (D), Arévalo (A) (eds.), Cetariae 2005. Salsas y Salazones de Pescado en Occidente durante la Antigüedad, BAR International Series 1686, 2007, 135-149.

CHIC, G. 1997b: ”Olivo y vid en la Andalucía romana: perspectivas de una evolución”, J. Morilla, J. GómezPantoja y P. Cressier (eds.), Impactos exteriores sobre el mundo rural mediterráneo, Madrid, 63-86. CHIC, G. 2001: Datos para un estudio socioeconómico de la Bética. Marcas de alfar sobre las ânforas olearias, Vol. I e II, Écija. CHIC, G.; GARCÍA VARGAS, E. 2004: “Alfares y producciones cerámicas en la provincia de Sevilla. Balance y perspectivas”, Figlinae Baeticae. talleres alfereros y producciones cerámicas en la Bética romana (II AC – VII DC), vol. 2, BAR, Int. Ser. 1266, 279-348.

CURTIS, R. I. 1991: Garum and salsamenta. Leiden. CURTIS, R.I. 2001: Ancient Food Technology, Leiden. D’ERCE, I. 1966: “La Tour de Caligula à Boulogne sur mer”, Rev. Archéologie, 89-96. DAVEAU, S. 1995: Portugal geográfico. Lisboa: Edições J. Sá da Costa. DAVIES, J.L. 1997: “Native producers and Roman consumers: the mechanisms of military supply in Wales from Claudius to Theodosius”. In W. Groenman-van Waateringe, B.L. van Beek, W.J.H. Willems & S.L. Wynia (eds.), Roman Frontier Studies 1995. Proceedings of the XVIth International Congress of Roman Frontier Studies, Oxbow Monograph 91, Oxford, 267-272.

CHINER, P. 1990: La decoración arquitectónica en Sagunto (Valencia), Generalitat Valenciana, Valencia. CHRISTOPHERSEN, A. 1999: “The waterfront and beyond. Commercial activity and the making of townscapes”. In Bill, J. & Birthe, L.C. (eds.) – Maritime Topography and the Medieval Town. Papers from the 5th International Conference on Waterfront Archaeology in Copenhagen, 14-16 May 1998. Copenhage: PNM Publications from The National Museum. Studies in Archaeology & History, vol. 4, p. 161-168. CIMMA, M.R. 1981: Richerche sulle Società di publicani. Mailand.

DAVIES, R. 1971: “The Roman military diet”. Britannia 2, 122-142. DAVIES, R. 1989a: “The daily life of the Roman soldier under the principate”. In R. Davies (ed) Service in the Roman army. London, 33-70. DAVIES, R. 1989b: The investigation of some crimes in Roman Egypt. In R. Davies (ed) Service in the Roman army, London, 175-185. DAVIES, R. 1989c: Service in the Roman army. Edinburgh. DE AVILÉS, T. 1965: Armas y linajes de Asturias y Antigüedades del Principado, ed. y notas de M.G. Martínez, Oviedo. DE BLOIS, L. 2000: “Army and Society in the Late Roman Republic: Professionalism and the Role of the Military Middle Cadre”. In G. Alföldy, B. Dobson, W. Eck (Hg.), Kaiser, Heer und Gesellschaft in der Römischen Kaiserzeit. Gedenkschrift für Eric Birley. Stuttgart, 11-31. DE BLOIS, L. 2007: “Army and general in the Late Roman Republic”. In P. Erdkamp (ed), Companion to the Roman Army (Blackwell Companion to the Ancient World). Oxford, 164-179. DE GUADÁN, A.M. 1969: “Una nueva moneda de Tingis”. Numisma 19, 9-23.

CIOTOLA, A.; PICCIOLA, S.; SANTAGELI, R.; VOLPE, R. 1989: “Roma: tre contesti. 1. Via Nova – Clivo Palatino. 2. Crypta Balbi. 3. Via Sacra – Via Nova”. CEFR 114, 604-609. CLAUSS, M. 1974: Untersuchungen zu den principales des römischen Heeres von Augustus bis Diokletian, Bochum. CLEERE, H. 1977: “The Classis Britannica”. In D.E. Johnson (ed.) The Saxon shore. CBA Res. Rep. 18. London, 16-19. CLETO, J.; VARELA, J. M. 2000: “O gabinete municipal de arqueologia e história de Matosinhos”. Al-madan, II série: 9, 141-144. COLLINGWOOD, R.G. 1930: The Archeology of Roman Britain, London. CONLIN, E., GARCÍA VARGAS, E.; GARCÍA FERNÁNDEZ, F.J. “Alfarería y romanización en la Turdetania antigua: el caso de Carmona”, 26th Congress of the Rei Cretariae Romanae Fautores. [unpublished] CONTICELLO, B. 1991: “Le vie romane e le guerre di conquista”. Viae publicae romanae. X Mostra europea del turismo, artigianato e delle tradizioni 274

BIBLIOGRAPHY

DIOGO, A.M.D.; ALVES, F. 1988-1989: “Ânforas provenientes do meio fluvial nas imediações de Vila Franca de Xira e de Alcácer do Sal”. O Arqueólogo Português. Lisboa. Série IV. 6-7, 227-240.

DE LAET, S. 1949: Portorium, Brugghe. DE MORALES, A. 1586: Crónica General de España, III (ed. Gabriel Ramos Bejarano), Córdoba. DE VINGO, P.; FOSSATI, A. 2001: “Gli utensili da pesca”, in T. Mannoni y G. Murialdo, S. Antonino. Un insediamento fortificato nella Liguria bizantina, Bordighera, pp. 657-660. DEGRASSI, A. 1965: Publicae. Firenze.

Inscriptiones

Liberae

DIOGO, A.M.D.; CARDOSO, J.L. 1992: “Cerâmica campaniense proveniente da Foz do Arade (Portimão)”. Artefacto. Lisboa. I, 9-11. DIOGO, A.M.D.; CARDOSO, J.L. 2000 “Ânforas provenientes de um achado marítimo ao largo de Tavira, Algarve”. Revista Portuguesa de Arqueologia. Lisboa. 3:2, 67-79.

Rei

DESBAT, A.; DÁNGREAUX, B. 1997: “La production d’amphores à Lyon”. Gallia 54, 73-104.

DIOGO, A.M.D.; CARDOSO, J.L.; REINER, F. 2000: “Um conjunto de ânforas recuperadas nos dragados da foz do rio Arade, Algarve”. Revista Portuguesa de Arqueologia. Lisboa. 3:2, 81-118.

DESBAT, A.; LEMAÎTRE, S. 2000: “Les premières importations d’amphores de Bétique à Lyon”. In Congreso Internacional “Ex Baetica amphorae” vol. I. Sevilla-Ecija, 17-20 Desembre 1998, 793-815.

DIOGO, A.M.D.; TRINDADE, L. 1993-94a: “Ânforas provenientes de achados subaquáticos ao largo da Berlenga”. In Actas do Congresso “A Presença Romana na Região Oeste”, 109-117.

DESBAT, A.; MARTIN KILCHER, S. 1989: “Les amphores sur l’axe Rhône-Rhin à l’époque d’Auguste”, Amphores romaines et Histoire économique: dix ans de recherches (Siena, 1986), Roma, 339-365.

DIOGO, A.M.D.; TRINDADE, L. 1993-94b “Materiais provenientes de Chões de Alpompé (Santarém)”. Conimbriga 32-33, 263-281.

DESCOSIDO, M. 1982: “Término augustal de la Legio X”, Tierras de León 48, 91-96.

DION, R. 1954: “Itinérarires maritimes occidentaux dans l’antiquité”, Bulletin de l’Association des Géographes Français 243-244, 128-135.

DHEEDENE, H. 1981: “De praefectus fabrum: overzicht en studie aan de hand van epigrafisch ronnenmateriaal, onuitgegeven licentiaatsverhandeling”. Gent.

DIOSONO, F. 2005: “El castellum romano del Cerro del Trigo (Puebla de Don Fadrique, Granada) y el control del territorio en época republicana”. AespA 78, 119128.

DIAS, J.M.A.; RODRIGUES, A.; MAGALHÃES, F. 1997: “Evolução da linha de costa, em Portugal, desde o último máximo glaciário até à actualidade: síntese dos conhecimentos”. In Estudos do Quaternário, I. Lisboa: Edições Colibri. 53-66.

DISE, R.L. 1991: Cultural change and Imperial Administration, New York.

DÍAZ-REGAÑÓN, J.M. 1984: Claudio Eliano. Historia de los animales, Biblioteca Clásica Gredos, 66 y 67 (libros I-VIII y IX-XVII), Madrid.

DISE, R.L. 1996: “The beneficiarii procuratoris of Celeia and the development of the station network”. ZPE 113, 286-292.

DICKEN, P.; LLOYD, P.E. 1990: Location in space. London.

DOBSON, B. 1966: “The praefectus fabrum in the early principate, in: Britain and Rome”. Essays presented to E. Birley on his sixtieth birthday, s.l., 61-84.

DICKS, T.R.B. (1972) “Network analysis and historical geography” Area 4, 4-9.

DOMASZEWSKI, A. von 1902: “Die Beneficiarerposten und die römischen Straßennetze”. Westdt. Zeitschr. 21, 158-211.

DICKSON, C. 1989: “The Roman diet in Britain and Germany”. Archäobotanik 133, 135-154.

DOPICO, M.D. 1988: La tabula lougeiorum: estudios sobre la implantación romana en Hispania. Vitoria.

DIOGO, A.M.D. 1982: “A propósito de “Moron”. Estudo de alguns documentos provenientes dos Chões de Alpompé (Santarém)”. Clio. 4, 147-154.

DORE, J.N.; KEAY, N. 1989: Excavations at Sabratha 1948-51. Vol. II, pars i. Soc. For Lybian Studies. London.

DIOGO, A.M.D. 1987: “Ânforas provenientes do rio Tejo (Salvaterra de Magos), no Museu do Mar. Arqueología”. Porto: GEAP. 16, 112-114.

DREXHAGE, H.J. 1991: Preise, Mieten/Pachten, Kosten und Löhne im römischen Ägypten bis zum Regierungsantritt Diokletians. St. Katharinen.

DIOGO, A.M.D. 1993: “Ânforas pré-romanas de Chões de Alpompé (Santarém)”. In Estudos Orientais IV – Os Fenícios no território Português. Lisboa, 215-227.

DRIEL-MURRAY, C.V.; GECHTER, M. 1984: “Funde aus der fabrica del legio I Minervia am Bonner Berg”, Beiträge zur Archäologie des römischen Rheinlands 4. Rheinische Ausgrabungen 23, Köln, 1-83.

DIOGO, A.M.D. 1999: “Ânforas provenientes de achados marítimos na costa portuguesa”. Revista Portuguesa de Arqueologia. Lisboa. 2:1, 235-248.

DUNCAN-JONES, R. 1990: Structure and scale in the Roman economy. Cambridge.

DIOGO, A.M.D. 2005: “Vestígios de um possível naufrágio romano ao largo da Ilha do Farilhão”. In Actas do Congresso “A Presença Romana na Região Oeste, 103-106.

DUNCAN-JONES, R. 1994: Movement and goverment in the Roman economy. Cambridge.

275

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

Amphorae. Conservas, aceite y vino de la Bética en el Imperio Romano, Écija, 665-682.

DYCZEK, P. (2008): “On the origins of amphora Zeest 90/Dyczek 25 – ultimate solution?”, Rei Cretariae Romanae Fautorum Acta 40, Bonn, pp. 515-521. EBEL, CH. 1976: Transalpine Gaul. The Emergence of a Roman Province. Leiden.

FABIÃO, C. 2005: “Caminhos do Atlantico Romano: Evidências e Perplexidades”. In Fernández Ochoa (C.), García Días (P.) (eds.), Unidad y diversidad en el Arco Atlántico en época romana, BAR International Series 1371, 2005, 83-85.

ECK, W. 1975: “Die Laufbahn eines Ritters aus Apri in Thrakien. Ein Beitrag zum Ausbau der kaiserlichen Administration in Italien”. Chiron 5, 365-392

FABIÃO, C. 2008: “Las ánforas de la Lusitania”. In Bernal Casasola (D.), Ribera i Lacomba (A.) (eds.), Cerámicas hispanorromanas. Un estado de la cuestión, Cádiz, 2008, 725-745.

ECK, W. 1999: L’Italia nell’impero romano. Stato e amministrazione in epoca imperiale, Roma. ECK, W. 2003: The age of Augustus. Oxford.

FABIÃO, C.; GUERRA, A. 1994: “As ocupaçoes antigas das Mesas do Castelinho (Almodóvar). Resultados preliminares das campanhas de 1990-92”. Actas das V Jornadas Arqueológicas da Associação dos Arqueólogos Portugueses (Lisboa, 1993), vol.II, 275289.

EDMONDSON, J.C. 1987: Two industries in Roman Lusitania. Mining and Garum Production. Oxford: British Archaeological Reports (International Series; 362). EHMIG, U. 2003: Die römischen Amphoren aus Mainz. Frankfurter Archaeölogischen Schriften 4. Möhnesee. ERA= DIEGO SANTOS, F. 1985: Epigrafía Romana de Asturias, Oviedo. ERDKAMP, P. 1995: “The corn supply of Roman armies during the third and second centuries BC”. Historia 44, 166-191. ERDKAMP, P. 1998: Hunger and the sword. Warfare and food supply in Roman Republican wars. Amsterdam. ERDKAMP, P. 2007: “War and State Formation in the Roman Republic”. In P. Erdkamp, A Companion to the Roman Army (Blackwell Companion to the Ancient World). Oxford, 96-113. ESPINOSA, A.; SÁEZ, F.; CASTILLO, R. (2003) “Puertos y navegación”. In J.M. Abascal, L. Abad, (coords.) (2003), 161-182. ETIENNE, R. 2002: Salaisons et salses de poisson hispaniques. Paris. ÉTIENNE, R., MAYET, F., 1994: “À propos de l’amphore Dr. 1C de Belo (Cadix)”, MCV XXX, 131138. ÉTIENNE, R., MAYET, F. 1998: “Les mercatores de saumure Hispanique”. MEFRA 100, 147-164. ÉTIENNE, R., MAYET, F. 2000: Le vin hispanique. Paris. FABIÃO, C. 1989: Sobre as ânforas do acampamento romano da Lomba do Canho (Arganil), Cadernos da UNIARQ, 1, Lisboa, Instituto Nacional de Investigação Científica. FABIÃO, C. 1994: “O Azeite da Baetica na Lusitania”, Conimbriga, 32-33, Instituto de Arqueologia, Coimbra, p. 219-245. FABIÃO, C. 1998: “O vinho na lusitânia: Reflexões em torno de um problema arqueológico”. Revista Portuguesa de Arqueologia, 1: 1, Lisboa, 169-198.

FELGUEIRAS, G. 1958: Monografia de matosinhos. Lisboa. FERNÁNDEZ CACHO, S. 1995: “Las industrias derivadas de la pesca en la provincia romana de la Bética: la alfarería de El Rinconcillo (Ageciras, Cádiz)”, Spal, vol. 4, 173-214. FERNÁNDEZ CASTRO, MªC. 1983: “Fábricas de aceite en el campo hispano-romano”. In Producción y Comercio de Aceite en la Antigüedad, II Congreso, Madrid, 569-601. FERNÁNDEZ FREILE, B.E. 1998/99: “Cerámica engobada y de paredes finas del alfar de Melgar de Tera en la ciudad de León”, Lancia 3, 103-126. FERNÁNDEZ FREILE, B.E. 2003: La época romana en León. Aspectos arqueológicos. Estudio arqueológico de un vertedero romano situado en la calle Maestro Copín c/v San Salvador del Nido en la ciudad de León, León. FERNÁNDEZ FERNÁNDEZ, A. 2007: “Aportaciones al estudio de la Terra Sigillata Gálica Tardía del grupo atlántico de Vigo (Galicia, España)”. SFECAG, Actes du Congrès de Langres, 2007, 331-340. FERNÁNDEZ FERNÁNDEZ, A. “Resultados preliminares del estudio de la TS Focense (LRC) aparecida en Vigo (Galiza, España)”, in RCRF Acta. [forthcoming]. FERNÁNDEZ FERNÁNDEZ, A.; PÉREZ LOSADA, F.; VIEITO, S. 2007: Influencias mediterráneas sobre producciones tardías regionales en el noroeste peninsular: la cerámica de engobe rojo de la villa de Toralla (Vigo), en LRCW 2. Late Roman Coarse Wares, Cooking Wares and Amplhorae in the Mediterranean. Archaeology and Archaeometry, BAR International Series 1662 (I), 2007, 99-108. FERNÁNDEZ FERNÁNDEZ, A.; PÉREZ LOSADA, F.; VIEITO, S. 2008: “Cerámica fina de importación en Toralla (Vigo): abastecimiento y consumo en una villa costera atlántica tardorromana”. In Fernández Ochoa (C.), García-Entero (V.) y Gil Sendino (F.) (Eds.): Las villae tardorromanas en el occidente del Imperio: arquitectura y función. IV Coloquio

FABIÃO, C. 2001a: “O sul de Lusitãnia (Portuguese Algarve) e a Betica: concorrência o complementariedade?”, Ex Baetica Amphorae II, Écija, 717-730. FABIÃO, C. 2001b: “Sobre as mais antigas anforas de Baetica no ocidente peninsular”, Ex Baetica 276

BIBLIOGRAPHY

noroeste de Hispania en época julio-claudia”, Colloque L’Aquitaine et l’Hispania septentrionale à l’èpoque julio-claudienne. Organisation et explotation des espaces provinciaux (Saintes, 2003), Aquitania. Supplement 13, Bordeaux, 157-167.

Internacional de Arqueología en Gijón, Gijón, 575585. FERNÁNDEZ FERNÁNDEZ, A.; LAGO, M. 2007: “Dos piezas de Terra Sigillata Aricana de engobe ocre amarillo aparecidas en Vigo (Galicia, España)”. SFECAG, Actes du Congrès de Langres, 2007, 507510. FERNÁNDEZ FERNÁNDEZ, A.; SOTO, P. 2007: “Cerámicas finas tardorromanas orixinarias de África e de Medio Oriente atopadas na intervención da rúa Rosalía de Castro 1992 (Vigo)”. Castrelos 13 (Revista do Museo Municipal “Quiñones de León”, Vigo), 2007, 26-51.

FERNÁNDEZ OCHOA, C.; MORILLO, A.; VILLA VALDÉS, A. 2005: “La torre de Augusto en la Campa Torres (Gijón, Asturias). Las antiguas excavaciones y el epígrafe de Calpurnio Pisón”, AEspA 78, 129-146. FERNÁNDEZ OCHOA, C, ZARZALEJOS, M., GARCIA ENTERO, V., GARCIA MARCOS, V., MENENDEZ GRANDA, A., SANCHEZ HIDALGO, E. FOUCHER, V. 2005: “La difusión de los talleres de la Graufesenque y Montans en el cuadrante noroccidental de Hispania: materiales para un corpus de marcas de alfarero”, X. Nieto et al. (eds.), La difusió de la terra Sigillata Sudgàl-lica al nord d’Hispania, Monografies Museu d’Arqueologia de Catalunya 6, Barcelona, 79-102.

FERNÁNDEZ GÓMEZ, F.; GUERRERO MISA, L.J.; VENTURA MARTÍNEZ, J.J.; HOZ GÁNDARA, A.; SIERRA FERNÁNDEZ, J.A. de la; ALCÁZAR GODOY, J. e SUÁREZ LÓPEZ, A. 1996: – Oripo en la Antigüedad. Las Excavaciones Arqueológicas de 1979 a 1983, Dos Hermanas. FERNÁNDEZ OCHOA, C. 1994: Una industria de salazones de época romana en la Plaza del Marqués, Excavaciones arqueológicas en la ciudad de Gijón, Gijón. FERNÁNDEZ OCHOA, C. (coord.) 1996: Los Finisterres atlánticos en la Antigüedad. Época prerromana y romanAColoquio Internacional. Madrid, 201204.

FERNÁNDEZ RODRÍGUEZ, C. 2003: Ganadería, caza y animales de compañía en la Galicia romana: estudio arqueozoológico, Brigantium 15, La Coruña. FERREIRO, M. 1988: “La campaña militar de Cesar en el año 61”. In Menaut G.P. (ed.) – Actas 1er. Congreso Peninsular de Historia Antigua. Santiago de Compostela, II, 363-372.

FERNÁNDEZ OCHOA, C. (ed.) 2003: Gijón, puerto romano. Navegación y comercio en el Cantábrico durante la Antigüedad, Gijón.

FERRER SIERRA, S. 1996: “El posible origen campamental de Lucus Augusti a la luz de las monedas de la caetra y su problemática”, A. Rodríguez Colmenero (ed.), Lucus Augusti I. El amanecer de una ciudad, La Coruña, 425-446.

FERNÁNDEZ OCHOA, C.; GARCÍA DÍAS, P. (eds.) 2005: Unidad y diversidad en el Arco Atlántico en época romana, BAR International Series 1371, Oxford. FERNÁNDEZ OCHOA, C.; GARCÍA DÍAZ, P.; GIL SENDINO, F. 2003: “Gijón, enclave marítimo en la ruta comercial cantábrica. Evidencias arqueológicas e hipótesis sobre el puerto romano y los embarcaderos antiguos”, C. Fernández Ochoa (ed.), Gijón, puerto romano. Navegación y comercio en el Cantábrico durante la Antigüedad, Gijón, 97-116.

FINK, R.O. 1971: Roman military records on papyrus, Princeton. FINLEY, M.I. 1973: The ancient economy. London (1985, 2nd ed.) FONTANELLA, F. 1992: “Metello Numidico: una tradizione ostile (un confronto fra A Num. 2-3 e Sallustio)”. Atene e Roma 38.4, Oct.-Dec., 177-182. FRANCE, J. 2005: “La mise en place de l’impôt provincial sous le règne d’Auguste en Aquitaine et dans le Nord de l’Hispanie: un bilan”. In L’Aquitaine et l’Hispanie septentrionale à l’époque julioclaudienne. Organisation et exploitation des espaces provinciaux. IV Colloque Aquitania (Saintes, Septembre 2003), Bordeaux 2005, 65-93.

FERNÁNDEZ OCHOA, C.; MORILLO, A. 1994a: De Brigantium a Oiasso. Una aproximación al estudio de los enclaves marítimos cantábricos en época romana. Madrid. FERNÁNDEZ OCHOA, C.; MORILLO, A. 1994b: “La ruta marítima del Cantábrico en época romana”, Zephyrus XLVI, 225-231.

FRANCE, J. 2006: “Tributum et stipendium. La politique fiscale de l’empereur romain”. Revue Historique de Droit Français et Étranger 84.1, 2006, 1-16.

FERNÁNDEZ OCHOA, C.; MORILLO, A. 1999: La Tierra de los Astures. Implantación romana en la antigua Asturia, Gijón.

FREEMAN, L.C. 1979: “Centrality in social networks. Conceptual Clasification”. Social Network, nº 1.

FERNÁNDEZ OCHOA, C.; MORILLO, A. 2002: “Entre el prestigio y la defensa: la problemática estratégicodefensiva de las murallas tardorromanas en Hispania”, Morillo, A. (coord.), Arqueología Militar Romana en Hispania, Anejos de Gladius 5, Madrid, 577-589.

FREITAS, C.; ANDRADE, C. 1998: “Evolução do litoral português nos últimos 5000 anos: alguns exemplos. Al.madan. Almada: Centro de Arqueologia, II Série, 7, 64-70. FRERE, S. 1986: “The use of Iron Age hill forts by the Roman army in Britain”. Studien zu den Militär-

FERNÁNDEZ OCHOA, C.; MORILLO, A. 2005: “Ciudades y aglomeraciones secundarias en el norte y 277

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

GARCÍA BELLIDO, MªP.; BLÁZQUEZ CERRATO, C. 2001: Diccionario de cecas y pueblos hispánicos, I y II, Textos Universitarios 35 y 36, Madrid. GARCÍA CANO, J.M.; HERNÁNDEZ, E.; INIESTA, A.; PAGE, V. 1997: “El santuario de Coimbra del Barranco Ancho (Jumilla) a la luz de los nuevos hallazgos”. QPAC 18, 239-256. GARCÍA CANO, J.M.; PAGE, V. 2004: Terracotas y vasos plásticos de la necrópolis del Cabecico del Tesosro, Verdolay, Murcia, Monografías del Museo de Arte Ibérico del Cigarralejo 1, Murcia. GARCÍA DE FIGUEROLA, M.; GONZÁLEZ ALONSO, E. 1998/99: “Las contramarcas monetarias de la legio VI en Hispania”, Lancia 3, 127-140. GARCÍA DE LOS RÍOS, A.; OCAÑA, O. 2006: Cetáceos de Ceuta y áreas próximas. Estudio faunístico, ecológico y veterinario de las especies de cetáceos en la región de Ceuta, Estrecho de Gibraltar y Mar de Alborán, Ceuta. GARCÍA FERNÁNDEZ, M. 2004: Memoria Intervencion Arqueológica de Urgencia. Saneamiento integral del Aljarafe. Colector margen derecha, [unpublished]. GARCÍA GIMÉNEZ, R.; BERNAL, D.; MORILLO, A. 1999: “Consideraciones sobre los centros productores de lucernas tipo Andújar: análisis arqueométrico de materiales procedentes de Los Villares de Andújar (Jaén) y de la submeseta norte”, Arqueometría y Arqueología. II Reunión de Arqueometría. I Congreso Nacional (1995), Granada, 187-195. GARCÍA MARCOS, V. 1989/1990: “Marcas de alfarero en Sigillata Hispánica halladas en la ciudad de León”, Tierras de León 77-78, 89-114. GARCÍA MARCOS, V. 2005: “Importación de terra sigillata itálica y producciones locales de tradición itálica en la Meseta norte y el Noroeste peninsular”, Unidad y diversidad en el Arco Atlántico. III Coloquio Internacional de Arqueología en Gijón, BAR Int. Series 1371, Oxford, 87-108. GARCÍA MARCOS, V.; VIDAL ENCINAS, J.M. 1996: “Recent archaeological research at Asturica Augusta”. Proceedings of the British Academy 86. London. GARCÍA MORÁ, F. 1991: Un episodio de la Hispania republicana: la guerra de Sertorio. Granada.

grenzen Roms III – 13 Internationalen Limes Kongress (Aalan, 1983), Stuttgart 1986, 42-46. FRERE, S. 1987: Britannia, London. FUENTES, N. 1991: “The mule of a soldier”. Journal of Roman Equipment Studies 2, 65-99. FULFORD, M.G. 1989: “To East and West: the Mediterranean trade of Cyrenaica and Tripolitanian in Antiquity”. Lybian Studies 20, 169-191. FULFORD, M.G. 1992: “Territorial expansion and the Roman Empire”. World Archaeology 23, 294-303. FUMADÓ, I. 2006: “El caballero de Tamuda”, L’Africa Romana 16 (Rabat, 2004), Roma, III, pp. 20092018. GARCÍA ALONSO, M. 2002: “El campamento romano de “El Cincho” (La Población de Yuso). Un nuevo yacimiento de las guerras cántabras”, Sautuola VIII, 99-106. GARCÍA Y BELLIDO, A. 1960: “L. Terentius, figlinarius en Hispania de la Legión IIII Macedónica”, Hommages a León Herrmann, Col. Latomus XLIV, Bruxelles, 374-382. GARCÍA Y BELLIDO, A. 1961: “El “Exercitus Hispanicus” desde Augusto a Vespasiano”, AEspA 34, 114160. GARCÍA Y BELLIDO, A. 1970: “Estudios sobre la Legio VII Gemina y su campamento en León”, Legio VII Gemina, León, 569-599. GARCÍA BELLIDO, MªP. 1994-95: “Las torres-recinto y la explotación militar del plomo en Extremadura: los lingotes del pecio de Comacchio”, Anas 7-8, 187-218. GARCÍA BELLIDO, MªP. 1996A: “La moneda y los campamentos militares”, C. Fernández Ochoa (coord.), Coloquio Internacional de Arqueología: los finisterres atlánticos en la antigüedad (época prerromana y romana), Gijón, 103-112. GARCÍA BELLIDO, MªP. 1996b: “Las monedas hispánicas de los campamentos del Lippe: Legio Prima (Antes Augusta) en Oberaden”. Boreas 19, 247-260. GARCÍA BELLIDO, MªP. 2000: “The historical relevance of secondary material. The case of Augustan Spanish coins in German castra”, R. Wiegels (ed.), Die Fundmünzen von Kalkriese und die frühkaiserzeitliche Münzprägung (Osnabrück, 1998), Möhnesee, 121-137.

GARCÍA MORÁ, F. 1995: “Castra Aelia”. In F. Burillo (ed.), III Simposio sobre los Celtíberos. Poblamiento celtibérico, Zaragoza, 281-288. GARCÍA MORÁ, F. 1998: “La primera estancia de Quinto Sertorio en Hispania: Cástulo”. In J. Alvar (ed.), Homenaje a José Mª. Blázquez, Hispania romana. Vol.V, Madrid 1998, 151-159. GARCÍA MORENO, L.A. 1995: “Las navegaciones romanas por el Atlántico norte: imperialismo y geografía fantástica”, Guerra, exploraciones y navegación: del mundo antiguo a la Edad Moderna, A Coruña, 101-110. GARCÍA VARGAS, E. 1996: “La producción anfórica en la Bahía de Cádiz durante la República como

GARCÍA BELLIDO, MªP. 2004: Las legiones hispánicas en Germania. Moneda y ejército, Anejos de Gladius 6, Madrid. GARCÍA BELLIDO, MªP. (coord.) 2006: Los campamentos romanos en Hispania (27 AC – 192 DC). El abastecimiento de moneda (2 vols.), (Anejos de Gladius). Madrid. GARCÍA BELLIDO, MªP. 2007: “El abastecimiento monetario al ejército durante el periodo augusteo y tiberiano”. In A. Morillo (eds.) El ejército romano en Hispania. León, 159-174.

278

BIBLIOGRAPHY

GARCÍA VARGAS, E. “Las ánforas republicanas de Hispalis (Sevilla) y la “cristalización” del repertorio anfórico provincial” in Homenaje a la profa. Pilar Acosta, eds. E. Ferrer Albelda & R. Cruz-Auñón Briones [forthcoming].

índice de romanización”. Habis, 27, Universidad de Sevilha, Sevilha, p. 49-62. GARCÍA VARGAS, E. 1998a: La producción de ánforas en la Bahía de Cádiz en época romana (siglos II AC – IV DC). Écija.

GARCÍA VARGAS, E., BERNAL, D. 2008: “Ánforas de la Bética”. In Bernal Casasola (D.), Ribera i Lacomba (A.) (eds.), Cerámicas hispanorromanas. Un estado de la cuestión, Cádiz, 2008, 662-687.

GARCÍA VARGAS, E. 1998b: “Centros productores de ánforas olearias del valle del Genil: Nuevas Aportaciones Arqueológicas”, Boletín de la Real Academia de Ciencias, Bellas Artes y Buenas Letras “Vélez de Guevara”. N. 2, 1998, 105-133.

GARCIA, D. 1992: “Les elements de pressoirs de Lattes et l’oléiculture antique en Languedoc méditerranéen”. Lattara 5, 237-258.

GARCÍA VARGAS, E. 2000: “Ánforas romanas producidas en Hispalis. Primeras evidencias arqueológicas”. Habis, 31, 235-260.

GARNSEY, P.D.A. 1983: “Grain for Rome”. In Trade in the Ancient economy, London, 118-130.

GARCÍA VARGAS, E. 2001: “La producción de ánforas “romanas” en el sur de Hispania. República y Alto Império”. Actas del Congreso Internacional Ex Baetica Amphorae. Conservas, Aceite y Vino de la Bética en el Império Romano (Écija-Sevilha 1998), Écija, p. 57-174.

GARNSEY, P.D.A. 1994: “L’aprovisionament des armies et la ville de Rome”, Le ravitaillement en blé de Rome et des débuts de la République jusqu’ au Aut. Empire, (Naples, 1991), Naples-Roma, 31-34. GASSNER, V.; JILEK, S. 1997: “Zum problem der “Legionswäre” und ihrer verbretung von Germanien bis Moesien”. In W. Groenman-van Waateringe, B.L. van Beek, W.J.H. Willems & S.L. Wynia (eds.), Roman Frontier Studies 1995. Proceedings of the XVIth International Congress of Roman Frontier Studies, Oxbow Monograph 91, Oxford, 301-309.

GARCÍA VARGAS, E. 2003a: “La Industria Alfarera en el Bajo Guadalquivir en Época Romana”. In Vázquez Labourdette, A. (ed.), Arqueología y Rehabilitación en el Parlamento de Andalucía. Investigaciones Arqueológicas en el Antiguo Hospital de las Cinco Llagas de Sevilla, Sevilha, p. 124-138.

GEBELLI, P.; DÍAZ, M. 2000: “Importaciones béticas en Tarraco en contextos pre-augústeos”. Congreso Internacional Ex Baetica Amphorae. Conservas, aceite y vino de la Bética en el Imperio Romano. Sevilla-Ecija 1998, vol. IV, Écija, 1349-1358.

GARCÍA VARGAS, E. 2003b: “Las producciones de la figlina. Ánforas”. In Vázquez Labourdette, A. (ed.), Arqueología y Rehabilitación en el Parlamento de Andalucía. Investigaciones Arqueológicas en el Antiguo Hospital de las Cinco Llagas de Sevilla, Sevilha, p. 200-219.

GECHTER, M.; DRIEL-MURRAY, C.V. 1984: “Funde aus der fabrica der Legio I Minervia am Bonner Berg”, Beitrage zur Archäologie des römisches Rheinlades 4, Rhein. Ausgrabungen 23, 90.

GARCÍA VARGAS, E. 2004a: “El vino de la Bética altoimperial y las ánforas. A propósito de algunas novedades epigráficas”. Gallaecia, 24. GARCÍA VARGAS, E. 2004b: “La romanización de la “industria” púnica de las salazones en el sur de Hispania”. Actas del XVI Encuentros de Historia y Arqueología “Las Industrias Alfareras y Conserveras Fenicio Púnicas de la Bahía de Cádiz”, Ayuntamiento de San Fernando, Publicaciones CajaSur, Córdoba, 101-130.

GIAMMELLARO, A. SPANÒ 2004: “Pappe, vino e pesce salato. Appunti per uno studio della cultura alimentare fenicia e punica”, Kokalos. Studi pubbicati dall’Istituto di Storia Antica dell’Università di Palermo XLVI, I, pp. 417-464.

GARCÍA VARGAS, E. 2004c: “Las ánforas del vino bético altoimperial: formas, contenidos y alfares a la luz de algunas novedades arqueológicas”. Actas del Congreso Internacional Figlinae Baeticae. Talleres Alfareros y Producciones Cerámicas en la Bética Romana (ss. II AC – VII DC), BAR, International Series, Oxford, p. 507-514.

GIARDINA, A. (ed.) 1986: Le merci gli insediamenti. Rome.

GIARD, J.-B. 1976: Catalogue des Monnaies de l’Empire Romaine I: Auguste, Paris.

GIARDINA, B. 2005: “Il faro nel mondo antico: aggiornamenti e nouvi dati”, Orizzonti 6, 137-152. GIARDINA, B. 2007: “La reppresentazione del faro nell emisión numismatiche del mondo antico”, RIN 108, 145-168.

GARCÍA VARGAS, E. 2004d: “Las pesquerías de la Bética durante el Imperio romano y la producción de púrpura”, in Purpureae Vestes. I Symposium Internacional sobre Textiles y Tintes del Mediterráneo en época romana (C. Alfaro, J.P. Wild & B. Costa, eds.), Valencia, pp. 219-235.

GIMENO, R. 1990: “El alfar romano de Melgar de Tera”, I Congreso de Historia de Zamora II. Prehistoria e Historia Antigua (1988), Zamora, 587-610. GIRAO, A.; OLEIRO, J.M.B. 1953: “Geografia e Campos Fortificados Romanos”. Boletim do Centro de Estudos Geográficos. 6-7, 77-80.

GARCÍA VARGAS, E. 2007: “Hispalis como centro de consumo desde época tardorrepublicana a la Antigüedad Tardía. El testimonio de las ánforas”, Anales de Arqueología Cordobesa, vol. 18, 317-360.

GISBERT, J.A. 1985: “Hallazgos arqueológicos submarinos en la costa de Denia. Las ánforas de cronología romana republicana”. VI CIAS, Madrid, 411-424.

279

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

GRAU, L.; HOYAS, J.L. (eds.) 2001: El bronce de Bembibre. Un edicto del emperador Augusto. León.

GLIOZZO, E., CERRI, L., DAMIANI, D., MEMMI TURBANTI, I. 2005: “Amphora production and salsamenta trade. The case of Thamusida (Rabat, Morocco)”. III Congr. Int., AIAR, L’Archeologia in Italia: la scienza per i Beni Culturali, Bolonia, 203213.

GROENMAN-VAN WAATERINGE, W. 1989: “Food for soldiers, food for thought”, J.C. Barrett, A.P. Fitzpatrick & L. Macinnes (eds.), Barbarians and Romans in North-West Europe, BAR Int. Series 471, Oxford, 96-107.

GOLDSWORTHY, A.K. 1996: “Appendix: logistics”. In: The Roman Army at War. 100 BC – AD 200, Oxford 1996, 287-296.

GROENMAN-VAN WAATERINGE, W. 1997: “Classical authors and the diet of Roman soldiers: true or false”, W. Groenman-van Waateringe, B.L. van Beek, W.J.H. Willems & S.L. Wynia (eds.), Roman Frontier Studies 1995. Proceedings of the XVIth International Congress of Roman Frontier Studies, Oxbow Monograph 91, Oxford, 261-265.

GÓMEZ-PANTOJA, J. 1994: “Germánico y Caesaraugusta”, Polis 6, 169-202. GONZÁLEZ BLANCO, A.; HERNÁNDEZ VERA, J.A. 1984: “Más restos de industria oleícola romana en La Rioja”, Producción y Comercio del Aceite en la Antigüedad. II Congreso Internacional, Madrid, 611616.

GUERRA, A. 1996: “Os nomes do Rio Lima. Um problema de toponímia e de geografia histórica”. Hispania prerromana. Actas del VI Coloquio sobre Lenguas y Culturas Prerromanas de la Península Ibérica. Salamanca, 147-161.

GONZÁLEZ ECHEGARAY, J.; SOLANA, J.M. 1975: “La Legión IV Macedónica en España”, Hispania Antiqua V, 151-203.

GUERRA, A. 2005: “O Promontório Magno. Perspectivas da geografia antiga sobre o extremo ocidental da Hispânia”. Actas do Congresso “A presença romana na Região Oeste”, Bombarral, 119129.

GONZÁLEZ FERNÁNDEZ, MªL. 1996: “Consideraciones sobre el origen militar de Asturica Augusta”. In C. Fernández Ochoa (coord.) Los Finisterres atlánticos en la antigüedad (Gijón-1993). Gijón, 85-90. GONZALEZ RUIBAL, A. 2004: “Facing two seas: Mediterranean and Atlantic contacts in the north-west of Iberia in the first millennium BCP”. Oxford Journal of Archaeology. 23: 3, 287-317.

GUEY, J. 1937: Essai sur la guerre parthique de Trajan. Bucarest. HAGGETT, P. 1965: Locational analysis in Human geography. New York.

GONZALEZ RUIBAL, A. 2005: Avance de resultados da escavación arqueolóxica no Castro Grande de O Neixón. Campaña de 2004. Análise do material cerámico exhumado, en Ayan Vila (X.M.) (coord.), Os castros de Neixón, Noia, 2005, 204-229.

HAGGETT, P. 1969: Network analysis in geography, London. HANEL, N. 1989: “Römische Öl- und Weinproduktion auf der Iberischen Halbinsel au Beispiel von Munigua und Milreu”. Madrider Mitteilungen 30, 204-238.

GONZALEZ RUIBAL, A. 2006-7: Galaicos. Poder y comunidad en el Noroeste de la Península Ibérica (1200 AC – 50 DC), A Coruña, 2006-07.

HANEL, N. 2006: “Fabricae, Werkstätten und handwerkliche Tätigkeiten des Militärs in den Nordprovinzen des Römischen Reichs”, A. Morillo (ed.), Arqueología Militar Romana en Hispania. Producción y abastecimiento en el ámbito militar, León, 19-32.

GONZÁLEZ RUIBAL, A.; RODRÍGUEZ, R.; ABOAL, R.; CASTRO, V. 2007: “Comercio mediterráneo en el castro de Montealegre (Pontevedra, Galicia). Siglo II AC – inicios del s. I DC” Archivo Español de Arqueología 80, 43-74.

HARMAND, J. 1967: “Les préfets“. In L’armée et le soldat à Rome de 107 à 50 avant notre ère. Paris, 358-366.

GONZÁLEZ VILLAESCUSA, R. 2001: El mundo funerario romano en el País Valenciano, Casa de Velázquez, Instituto Juan Gil Albert, Madrid.

HAUSCHILD, TH. 1977: “El faro romano de La Coruña (Torre de Hércules). Problemas de su reconstrucción”, Actas del Bimilenario de Lugo (1976), Lugo, 131156.

GONZÁLEZ, J.; GONZÁLEZ, C.; GONZÁLEZ, M. 2005: “Annex 1: Análisis arqueométrico de las ánforas Haltern 70 originarias de la Bética y procedentes del pecio Culip VIII”. In Carreras, C. et al., 159-165. GONZALEZ, M. 2004: “Prosopografía del praefecti fabrum originarios de Lusitania”. Revista Portuguesa de Arqueología vol.7, n.1, 365-384.

HÄUSSLER, R. 2002: “Writing Latin-from resistence to assimilation: language, culture and society”. In N. Italy and S. Gaul, A.E. Cooley, ed., Becoming Roman writing latin? Literacy and Epigraphy in Roman West, Portsmooth, 61-76.

GORGES, G.J. (1979) Les villas hispano-romaines, inventaire et problématique archéologiques. Paris.

HAYES, J.W. 1983: “The villa Dionysios excavations. Knossos: the pottery”. ABSA 78, 98-169.

GRANADOS, O.; RODA, I. 1994: “La Barcelona de l’època romana”, III Congrés d’història de Barcelona. La ciutat i el seu territori, dos mil anys d’història (1993), Barcelona, 11-24.

HERMET, F. 1934: La Graufesenque (Condatomago), Paris. HERMON, E. 1993: Rome et la Gaule Transalpine avant César, Napoli-Laval.

280

BIBLIOGRAPHY

HERMON, E. 2000: “L’impérialisme romain et la formation de la province de Gaule transalpine”. Archéologie en Languedoc 24, 99-109.

eds. E. Ferrer Albelda, A. Fernández Flores, J.L. Escacena Carrasco & A. Rodríguez Azogue, 2007, 193-209.

HERZ, P. 1988: Studien zur römischen Wirtschaftsgesetzgebung. Die Lebensmittelversorgung. Historia Einzelschr 55. Stuttgart.

JACOMB, C. 1998: “Shore Whaling Sites of Banls Peninsula: An Archaeological Survey”, in Lawrence, S. & Staniforth, M., The Archaeology of Whaling in Southern Australia and New Zealand, pp. 68-78. Gianfrota, P.A.; Nieto, X.; Pomey, P.; Tchernia, A. 1997: La navigation dans l’Antiquité, Aix-enProvence.

HERZIG, H.E. 1999: “Die Gründung des Cursus publicus unter Augustus”. Bulletin IVS 99/1, 12-17. HESNARD, A., 1998a: en V. Blanc-Bijon, M.-B. Carre, A. Hesnard, A. Tchernia, Recueil de timbres sur amphores romaines II (1989-1990 et compléments 1987-1988), Université de Provence, Aix-enProvence, 291-293.

JAMES, S. 2001: “Soldiers and civilians: identity and interaction in Roman Britain”. In S. James; M. Millet (eds.), Britons and Romans: advancing an achaeological agenda, York 2001, 77-89.

HESNARD, A., 1998b: “Le sel des plages (Cotta et Tahadart, Maroc)”, MEFRA 110, Roma, pp. 167-192.

JIMÉNEZ-CAMINO, R.; BERNAL CASASOLA, D. 2007: “Redescubriendo a Traducta. Reflexiones sobre su topografía urbana y su secuencia ocupacional (ss. I-VII)”, Anuario de Arqueología Cordobesa 18, Córdoba, pp. 157-200.

HIDALGO, J.M.; VIÑAS CUE, R. 1992-3:, Nuevas cerámicas romanas de importación del Castro de Vigo (Campaña de 1987), en Castrelos 5-6 (Revista do Museo Municipal “Quiñones de León”, Vigo), 100293, 41-70.

JIMÉNEZ DE FURUNDARENA, A. 1995: “Castellum en la Hispania romana: su significado military”. Hispania Antiqua 19, 1995, 129-150.

HOBSBAWM, E.; RANGER, T. (eds.) 1983: The Invention of Tradition, Cambridge.

JIMÉNEZ SANCHO, A.; GARCÍA VARGAS, E.; GARCÍA FERNÁNDEZ, F.J.; FERRER ALBELDA, E. 2006: “Aportación al estudio de la Sevilla prerromana y romano-republicana. Repertorios cerámicos y secuencia edilicia en la estratigrafía de la calle Abades 41-43”, Spal, vol. 15, 281-312.

HODDER, I.R.; ORTON, C. 1976: Spatial analyses in Archaeology. Cambridge. HOLDER, P.A. 1982: The Roman army in Britain, London. HONRAD, C.F. 1994: Plutarch’s Sertorius. A Historical Commentary. Chapel Hill and London.

JIMENO, A. 2001: “Numancia”. In M. Almagro, M. Mariné, J.R. Álvarez, Celtas y Vettones, Ávila, 234247.

HOPKINS, K. 1980: “Taxes and trade in the Roman Empire (200 BC-AD 400)”. Journal of Roman Studies LXX, 101-125.

JONES, R.F.J. 1976: “The Roman military occupation of North-West Spain”. JRS lxvi, 45-66.

HURST, H. 1994: The Circular Harbour. North side. The Site and Finds other than Pottery, Excavations at Carthage. The British Mission, II, 1, Oxford.

JONGMAN, W. 1991: The economy and society of Pompeii. Amsterdam. JUNKELMANN, M. 1986: Die Legionen des Augustus. Der römische Soldat im archäologischen Experiment. Mainz.

HUTTER, S. 1973: Der Römische Leuchtturm von La Coruña, Madrider Beiträger 3, Mainz. IGLESIAS, J.M. 2003: “Establecimientos portuarios en el sector central del Mar Cantábrico”. Urteaga, M.M., Noain M.J. (eds). Mar Exterior, actas del Congreso Internacional. Pisa.

JUNKELMANN, M. 1997: Panis Militaris: Die Ernährung des römischen Soldaten oder der Grundstoff der Macht. Mainz. JUSTINO, D. 1988: A formação do espaço económico nacional. Portugal 1810 – 1913. Lisboa: Vega, Gabinete de Edições (Colecção Documenta Historica), vol. I.

IGLESIAS, Mª.J. 2007: Fase II. Evaluación patrimonial de la Unidad de Actuación I 06. Rosalía de Castro II, Vigo, Pontevedra, Informe inédito depositado en el Servicio de Arqueología de la Xunta de Galicia, Santiago de Compostela [unpublished]. ISARD, W. 1956: Location and space economy. Cambridge.

KBIRI ALAOUI, M. 2007: Revisando Kuass (Asilah, Marruecos). Talleres cerámicos en un enclave fenicio, púnico y mauritano, Saguntum-extra 7, Valencia.

ISSAC, B. 1999: The limits of the Empire. The Roman army in the East, Oxford.

KEAY, S.J. 1984: Late Roman amphorae in the Western Mediterranean. BAR Int.Ser. 196, Oxford.

IZQUIERDO, I. 1998: “Iberian anthropomorphic steles: La Serrada (Ares del Maestre, Castellón) and Mas de Barberán (Nogueruelas, Teruel) examples”. Journal of Iberial Archaeology 0, 115-131.

KEAY, S.J. 1992: “The Siena amphorae conference. Part 1: Amphorae and the Roman Economy”. Journal of the Roman Archaeology 5, 353-360. KISSEL, T.K. 1995: Untersuchungen zur Logistik des römischen Heeres in den Provinzen des griechischen Ostens (27 v.Ch. – 235 n.Ch.). Pharos studies 6. St. Katharinen.

IZQUIERDO, R. 2007: “Fortissimum Oppidum. Investigaciones en la muralla romana de Alcalá del Río”, Ilipa Antiqua. De la Prehistoria a la época romana,

281

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

KNAPPP, R. 1977: Aspects of the Roman Experience in Iberia 206-100 BC. Valladolid 1977.

LINEROS, R. 2005: “Urbanismo romano de Carmona I”, Carel, vol. 3, 987-1033.

KOLB, A. 2000: Transport und Nachrichten transfer im römischen Reich. Akademie-Verlag, Berlin.

LION, M.C. 1997: “El alfar de cerámica de paredes finas de Melgar de Tera” (Zamora, España), Rei Cretariae Romanae Fautores. Acta 35, 173-178.

LABISCH, A. 1975: Frumentum Commenatusque: Die Nahrungs-mittelversorgung der Heere Caesars. Meisenheim an Glau.

LIOU, B. 2001: “Las ánforas béticas en el mar. Les épaver en Méditerranée à cargaison d’amphores de Bétique”. Actas del Congreso Internacional Ex Baetica Amphorae. Conservas, Aceite y Vino de la Bética en el Império Romano (Écija-Sevilha 1998), Vol. III, Écija, p. 1061-1111.

LAFUENTE, A. 1992: “La producción cerámica ibérica en el taller de Fontscaldes (Valls, Alt Camp)”. In La cerámica de técnica ibèrica a la Catalunya romana (segles II-I AC), Barcelona, 47-77.

LIOU, B.; SCIALLANO, M. 1989: “Le trafic du port de Fos dans l’antiquité: Essai d’évaluation à partir des amphores”. SFECAG, Actes du Congrés de Lezoux, 153-167. LLUÍS, L. 2003: “Ânforas republicanas de Mata-Filhos (Mértola)”. Revista Portuguesa de Arqueologia, 6: 2, 363-382. LOMAS, F.J. 1975: Asturia prerromana y altoimperial. Sevilla. LÓPEZ ÁLVAREZ, X. 1997: “El artículo “Gijón” de Jovellanos para el Diccionario Geográfico-Histórico de Asturias” en Fernández Ochoa, C.: La muralla romana de Gijón, Gijón. LÓPEZ CASTRO, J.L. 1995: Hispania poena: los fenicios en la hispania romana (206 AC – 96 DC). Barcelona. LÓPEZ MULLOR, A.; MARTIN MENÉNDEZ, A. 2008: “Las ánforas de la Tarraconense”. In Bernal Casasola (D.), Ribera i Lacomba (A.) (eds.), Cerámicas hispanorromanas. Un estado de la cuestión, Cádiz, 2008, 690-724.

LAGÓSTENA BARRIOS, L. 1996: Alfarería romana en la Bahía de Cádiz. Cádiz. LAGÓSTENA BARRIOS, L. 2001a: “Aportación al conocimiento de la sociedad de la costa de la Ulterior en época republicana y julio-claudia. El registro β en los tituli picti de las ánforas salsarias de Castra Pretoria”. Lucentum XXI-XXII, 227-236. LAGÓSTENA BARRIOS, L. 2001b: La producción de salsas y conservas de pescado en la Hispania romana, Col. Instrumenta nº 11. Barcelona. LAGÓSTENA BARRIOS, L. 2004: “Las ánforas salsarias de BaeticAConsideraciones sobre sus elementos epigráficos”. In J. Remesal (ed.) Epigrafía anfórica. Barcelona, 197-220. LAGÓSTENA, L.; BERNAL, D; ARÉVALO, A. 2007: Cetariae 2005. Salsas y Salazones de Pescado en Occidente durante la Antigüedad, BAR International Series 1686, 2007. LAMBOGLIA, N. 1983: Il trofeo di Augusto alla Turbia, Bordighera (4ª ed.). LAUBENHEIMER, F. 1985: La production des amphores en Gaule Narbonnaise. CNRS 55. Paris.

LÓPEZ PÉREZ, M.C. 2005: “Galicia y los contactos comerciales con el sur de la Galia: la terra sigillata sudgálica”, X. Nieto et al. (eds.), La difusió de la terra Sigillata Sudgàl-lica al nord d’Hispania, Monografies Museu d’Arqueologia de Catalunya 6, Barcelona, 63-78.

LAWRENCE, S.; STANIFORTH, M. 1998: The Archaeology of Whaling in Southern Australia and New Zealand, The Australian Society for Historical Archaeology and The Australian Institute for Maritime Archaeology, Special Publication nb 10.

LÓPEZ SANCHEZ, F. 2007: “Los auxiliares de Roma en el valle del Ebro y su paga en denarios ibéricos (11090 AC)”. Athenaeum 95.1, 2007, 287-320.

LAWTON, B. 2004: Various and Ingenious Machines, Vol.6.1, Power generation and transport. Brill, Lieden/Boston.

LORENZO SANZ, E. 1979: Comercio de España con América en la época de Felipe II. Valladolid.

LE BOHEC, Y. 1994: The Imperial Roman army, London.

LOVE, B.J. 2002: “Sextus Pompeius and Spain, 46-44 BC”. In A. Powell, K. Welch, (eds.), Sextus Pompeius. Swansea, 65-102.

LE ROUX, P. 1976: “Lucus Augusti, capitale administrative au Haut-Empire”, Actas del Bimilenario de Lugo, Lugo, 83-105.

LUIK, M. 2002: Die Funde aus den Römischen Lagern um Numantia im Römisch-Germanisch Zentralmuseum, Mainz. MAC MULLEN, R. 1963: Soldier and Civilian in the Later Roman Empire, Cambridge.

LE ROUX, P. 1982: L’Armee romaine et l’organisation des provinces iberiques d’Auguste a l’invasion de 409. Paris. LE ROUX, P. 1995: “La contribution des armées: la culture militaire”. In Romains d’Espagne. Cités & politique dans les provinces. IIe. Siècle av.J.-C. – IIIe. Siècle ap. J.-C., Paris 1995, 21-34.

MACIEL, J; PEIXOTO CABRAL, J.M.; NUNES, D. 2002: “Os sarcófagos tardo-romanos do Museu Nacional de Arqueologia. Novos dados para a sua interpretação”. O Arqueólogo Português. Lisboa, Série IV, 20, 161-175.

LEÓN, P. 1979: “Plástica ibérica e ibero-romana”. In La baja época en la Cultura Ibérica, Madrid, 183199.

MALMENDIER, U. 2002: Societas publicanorum. Köln. 282

BIBLIOGRAPHY

Prehistoria y Arqueología de la Universidad Autónoma de Madrid 17, 67-89. MARTÍNEZ MERA, J. 2001: “Expedicións militares a gallaecia na época republicana”. Gallaecia 20, 297316. MARTIN-KILCHER, S. 1987: Die römischen amphoren aus Augst und Kaiseraugst. Vol. 1. Bern. MARTIN-KILCHER, S. 1993: “Amphoren den späten Republik und der frühe Kaiserzeit in Karthago”. Zu den Lebensmittelimporten der Colonia Iulia Concordia, Mitteilunge DAInst Rome 100, 269-320. MARTIN-KILCHER, S. 1994: Die römischen amphoren aus Augst und Kaiseraugst. Vol. 2. Bern.

MANACORDA, D. 1977: “Il Kalendarium vegetianum e le anfore della Betica”. MEFRA 89, 313-332. MANNING, W. 2006: “The Roman fort at Newstead: the weapons and the garrisons”. In R.J.A. Wilson, (ed.), Romanitas. Essays on Roman Archaeology in honour of Sheppard Frere on the occasion of his ninetieth birthday, Oxford 2006, 73-94. MANTAS, V. 1994: “Olisiponenses: epigrafia e sociedade na Lisboa romana”. In Lisboa subterrânea. Lisboa, 70-75. MARCO, F. 1996: “Romanización y aculturación religiosa: los santuarios rurales”. In S. Rebordea, P. López Barja, A cidade e o mundo: romanización e cambio social, Xinzo de Limia, 81-100. MARI, L. et al. 1982: “La vil·la romana de Cantaperdius (Bellvís – La Noguera)”. Ilerda XLIII, 95-117. MARI, L.; MASCORT, M. 1988: “Una instal·lació industrial oleícola d’època romana al municipi de Corbins (Segrià)”. 7è Col·loqui de Puigcerdà (1986), 267-274. MARÍN, MªC.; PADILLA, A. 1997: “Los relieves del domador de caballos y su significado en el conjunto religioso ibérico”. QPAC 18, 461-494. MARLIERE, E. 2002: L’outre et le tonneau dans l’Occident romain, Monographies d’Instrumentum 22, Montagnac. MARSH, G. 1981: “London’s samian supply and its relationships to the development of the Gallic samian inustry”. In A.C. Anderson and A.S. Anderson (eds.) Roman pottery research in Britain and North-west Europe: papers presented to Graham Webster. BAR Int.Ser. 123. Oxford, 173-238.

MARTINS, M. 1990: “O povoamento proto-histórico e a romanização da bacia do curso médio do Cávado”. Cadernos de Arqueologia: Monografias 5, Braga. MASTINO, A. 2005: Storia della Sardegna Antica (La Sardegna e la sua storia, II). Nutro, 65-144. MATALOTO, R. 2002: “Fortins e recintos-torre do Alto Alentejo: antecámara da “romanização” dos campos”, Revista Portuguesa de Arqueología, vol. 5.1., 161220. MATALOTO, R. 2004: “Fortins romanos do alto Alentejo: fortificaçao e povoamento na segunda metade do séc. I AC”. In P. Moret, T. Chapa (eds.), Torres, atalayas y casas fortificadas. Explotación y control del territorio en Hispania (s.III a. de C. – s. I DC). Jaén, 31-54. MATÍA MERINO, M.J. 1999: “Aspectos de la industria ósea procedentes de Herrera de Pisuerga”, Papeles Herrerenses II, Herrera de Pisuerga, 53-71. MATTINGLY, D.J. 1988: “Oil for export? A comparison of Lybian, Spanish and Tunisian olive oil production in Roman Empire”. Journal of Roman Archaeology 1, 35-56. MAYA, J.L.; CUESTA, F. (eds.) 2001: El castro de la Campa Torres. Período prerromano. Gijón. MAYER, M.; RODÀ, I. 1986: “La epigrafía republicana en Cataluña. Su reflejo en la red viaria”. Epigrafía hispana de época republicana, Zaragoza 1986, 157170. MAYET, F.; SILVA, C.T. 1993: “A presença fenícia no Baixo Sado”. Estudos Orientais. Lisboa: I.O.U.N.L., IV, p. 127-143. MAYET, F.; SILVA, C.T. 1994 “L’ Etablissement Phénicien d’Abul (Alcácer do Sal) ”. Les Dossiers de l’Archéologie. Dijon: s.n., 198, p. 22-25. MAYET, F.; SILVA, C.T. 2001: “O santuário de Abul b, uma presença púnica no baixo Sado?” In Os Púnicos no Extremo Ocidente. Actas do colóquio internacional (Lisboa, 27 e 28 de Outubro de 2000). Lisboa: Universidade Aberta, p. 173-195. MEDAS, S. 2000: La Marineria Cartaginese, Sassari. MEDEROS, A.; ESCRIBANO, G. 2000: “El periplo norteafricano de Hanón y la rivalidad gaditanocartaginesa de los siglos IV-III AC” Gerión 18, 77107.

MARTÍN HERNÁNDEZ, E. 2006: “Cerámica romana de paredes finas de época julioclaudia en el campamento de la legio VI victrix. Estudio preliminar de los materiales procedentes del Polígono de La Palomera”, A. Morillo (ed.), Arqueología Militar Romana en Hispania. Producción y abastecimiento en el ámbito militar, León, 399-418. MARTÍN HERNÁNDEZ, E. 2007: Cerámica romana de paredes finas de época julioclaudia en el campamento de la legio VI victrix en León. Los materiales procedentes del polígono de la Palomera, León [forthcoming]. MARTÍN VALLS, R., ROMERO CARNICERO, Mª V.; CARRETERO, S. 2002: “Marcas militares en material de construcción de Petavonium”, A. Morillo (coord.), Arqueología Militar Romana en Hispania, Anejos de Gladius 5, Madrid, 137-154. MARTIN, T. 2005: “Périple aquitain, comerse transpyrénéen et difusión atlantique des céramiques sigillées de Montans en direction des marchés du nord et du nord-ouest de la péninsule ibérique”, X. Nieto et alli (eds.), La difusió de la terra Sigillata Sudgàl-lica al nord d’Hispania, Monografies Museu d’Arqueologia de Catalunya 6, Barcelona, 21-62. MARTÍNEZ MAGANTO, J. 1990: “Faros y luces de señalización en la navegación antigua”, Cuadernos de

283

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

MONSIEUR, P. 2001: “L’amphore Gauloise Augst 21: une grande inconnue”. Acta Archaeologische Lovaniensia 13, 171-184. MOLINA, J. 1995: “Las ânforas ‘Lomba do Canho 67’. Aportaciones al estudio de un nuevo tipo: difusión y valoración económica”, In Actas del XIII Congreso Nacional de Arqueología, Vigo (1993), p. 419-424.

MENCHELLI, S. 1997: “Terra sigillata pisana: forniture militari e “libero mercato”, Rei Cretariae Romanae Fautores. Acta 35, Abingdon, 191-198. MENÉNDEZ, A.R. 2000: Las legiones del siglo III dC en el campo de batalla. Écija. MENÉNDEZ, A.R. 2006: “Organización administrativa de los abastecimientos al ejército romano altoimperial”. In A. Morillo (ed.) Arqueología militar romana en Hispania. Producción y abastecimiento en el ámbito military. León, 335-341.

MOLINA, J. 1997: La Dinámica Comercial Romana entre Italia e Hispania Citerior, (siglos II AC – II DC). Universidad de Alicante, Instituto de Cultura Juan Gil-Albert, Alicante. MOLINA, J. 2001: “Las primeras importaciones béticas en el Mediterráneo Occidental”, Ex Baetica Amphorae. Conservas, aceite y vino de la Bética en el Imperio Romano, Écija, 637-646. MOLINA, J. 2002: “La irrupción de Hispania en los movimientos socioeconómicos del Mediterrâneo occidental durante las Guerras Civiles”, Gérion, 20, p.281-306. MORAIS, R. 1997-1998: “Sobre a hegemonia do vinho e a escassez do azeite no Noroeste Peninsular nos inícios da romanização”, Cadernos de Arqueologia, série II, 14-15, Braga, 175-182.

MERCADO, M.; PALET, J.Mª; RODRIGO, E.; GUITART, J. 2006: “El castellum de Can Tacó/Turó d’en Roina (Montmeló-Montornès) i la romanització de la Laietània Interior. Cap a un estudi arqueològic del jaciment i del territori”. Notes. El patrimoni arqueològic del Baix Vallès 21, 2006, 241-266. MEZQUÍRIZ, MªA. 1999: “La producción de vino en territorio navarro durante la época romana”, El vino en la Antigüedad romana, Serie Varia 4, Madrid, 241252. MIDDLETON, P. 1979: “Army supply in Roman Gaul: an hypothesis for Roman Britain”. In B.C. Burnham and H.B. Johnson (eds) Invasion and response: the case of Roman Britain. BAR Brit.Ser. 73, (Oxford), 81-97. MIDDLETON, P. 1983: “The Roman army and long distance trade”, P. Garnsey and C.R. Whittaker, Trade and famine in classical antiquity, Cambridge Philological Society su Vol. 8, Cambridge, 75-83.

MORAIS, R. 1998: As ânforas da zona das Carvalheiras. Contribuição para o estudo das ânforas romanas de Braccara AugustACadernos de Arqueología 8. Braga. MORAIS, R. 2000: “Póster”. In Congreso Internacional “Ex Baetica amphorae” vol. I. Sevilla-Ecija, 17-20 Desembre 1998, 697-698.

MILLÁN, J. 1998: Gades y las Navegaciones Oceánicas en la Antiguedad (1000 AC – 500 DC). Écija.

MORAIS, R. 2004: “Braccara Augusta: um pequeno Testaccio de ânforas Haltern 70 – Considerações e problemáticas de estudo”. In Figlinae Baeticae. Talleres alfareros y producciones cerámicas en la Bética romana (s. II AC-VII DC), Cádiz, 2003. BAR Int.Ser. Oxford.

MÍNGUEZ, J.A. 1991: La cerámica romana de paredes finas, Zaragoza. MIRÓ, J. 1988: La producción de ánforas romanas en Catalunya. Un estudio sobre el comercio de vino de la Tarraconense (s. I AC-I DC), BAR Int. Series 473, Oxford. MITCHELL, S. 1976: “Requisitioned transport in the Roman Empire. A new inscription from Pisidia”. JRS 66, 87-105. MITCHELL, S. 1993: Anatolia (vol. I). Oxford.

MORAIS, R. 2005a: Bracara Augusta, Excavaçoes Arqueológicas. Autarcia e Comércio em Bracara AugustAContributo para o estudio economico da cidade no período Alto-Imperial. Vol. II. Unidade de Arqueología da Universidade do Minho/Núcleo de Arqueología da Universidade do Minho. Braga. MORAIS, R. 2005b: “Problemàtiques i noves perspectives sobre les àmphores ovoides tardorepublicanes. Les àmphores ovoides de producció Lusitana”. In C. Carreras et al. (eds) Culip VIII i les àmfores Haltern 70. Girona, 36-40.

MÓCSY, A. 1953: “Das territorium legionis und die cannabae in Pannonien”, Acta Archaeologica Acaemiae Scientiarum Hungaricae III, Budapest, 179-200. MÓCSY, A. 1966: “Das Lustrum Primipili und die Annona Militaris”. Germania 44, 312-326. MÓCSY, A. 1967: “Zu den prata legionis”, Studien zu den militärgrenzen Roms. Vorträge des 6. Internationalen Limeskongresses in Süddeutschland, Köln, 211-214. MOLAS, D. 1993: “Les recerques sobre les societats ausetana i lacetana. Estat de la question”. Laietania 8, 131-143. MOLAS, D.; PRINCIPAL, J.: MESTRES, M.; DURAN, M.; ALVAREZ, R. 2000: “El jaciment del Camp de les Lloses (Tona, Osona) i el seu taller de metals”. Saguntum extra-3, 271-282.

MORAIS, R. 2007: “Ânforas tipo urceus de produção Bética e produções regionais e locais do NW Peninsular”. In Actas del Congreso Internacional CETARIAE. Salsas y salazones en Occidente durante la Antigüedad. Universidad de Cádiz (noviembre 2005). BAR Int. Ser. 1686, Oxford, 401-415. MORAIS, R.; CARRERAS, C. 2005: “Geografia del consum de les Haltern 70”. In C. Carreras et al. (eds) Culip VIII i les àmfores Haltern 70. Girona, 93-116. MORALES, A.; ROSELLÓ, E. 2007: “Los atunes de Baelo Claudia y Punta Camarinal (s. II AC). Apuntes preliminares”, in Arévalo, A. & Bernal, D., Las 284

BIBLIOGRAPHY

MORILLO, A. 2002: “Conquista y estrategia: el ejército romano durante el periodo augusteo y julio-claudio en la región septentrional de la península ibérica”, A. Morillo (coord.), Arqueología Militar Romana en Hispania, Anejos de Gladius 5, Madrid, 67-94.

cetariae de Baelo Claudia. Avance de las investigaciones arqueológicas en el barrio industrial (2000-2004), Sevilla, pp. 489-498. MORALES, A.; ROSELLÓ, E. 2009: “Vertebrados de las factorías de la c/ San Nicolás y reflexiones zoológicas sobre las factorías romanas de salazones”, in D. Bernal, ed., Las factorías de salazón de Traducta. Primeros resultados de las excavaciones arqueológicas en la c/ San Nicolás (Algeciras, Cádiz), Universidad de Cádiz and Ayuntamiento de Algeciras, Algeciras, in press.

MORILLO, A. 2003a: “Lucernas”. In M.T. Amaré (ed.) Astorga IV. León, 15-632. MORILLO, A. 2003b: “La navegación oceánica durante la época romana: de la imagen legendaria a la vertebración de un espacio marítimo atlántico”, C. Fernández Ochoa (ed.), Gijón, puerto romano. Navegación y comercio en el Cantábrico durante la Antigüedad, Gijón, 14-41.

MORILLO, A. 1992a: Las lucernas. Cerámica romana de Herrera de Pisuerga. Santiago de Chile. MORILLO, A. 1992b: Cerámica romana de Herrera de Pisuerga (Palencia, España): las lucernas, Santiago de Chile.

MORILLO, A. 2005a: “La Arqueología Militar Romana en Hispania: nuevas perspectivas”, La Arqueología Clásica peninsular ante el Tercer Milenio. In el Centenario de A. García y Bellido (1903-1972), Anejos de AEspA XXXIV, Madrid, 161-186.

MORILLO, A. 1992c: “La producción de Vogelkopflampen de Herrera de Pisuerga (Palencia, España). Un testimonio revelador de la política militar augustea en la Península Ibérica”, Opvs XI, Siena, 115-134.

MORILLO, A. 2005b: “Hispania en la estrategia militar del Alto Imperio: movimentos de tropas en el arco atlántico a través de los testimonios arqueológicos”, Unidad y diversidad en el Arco Atlántico. III Coloquio Internacional de Arqueología en Gijón, BAR Int. Series 1371, Oxford, 19-33.

MORILLO, A. 1993: “Una nueva producción de lucernas en la Península Ibérica: el taller militar de Herrera de Pisuerga (Palencia, España)”, I Congreso de Arqueología Peninsular, I, Porto (Trabalhos de Antropologia e Etnologia 33 (1-2), 351-364.

MORILLO, A. 2006a: (ed.) Arqueología militar romana en Hispania. León.

MORILLO, A. 1996a: “Los campamentos romanos de la Meseta Norte y el Noroeste: ¿un limes sin frontera?”, C. Fernández Ochoa (coord.), Coloquio Internacional de Arqueología: los finisterres atlánticos en la antigüedad (época prerromana y romana), Madrid, 80-81.

MORILLO, A. 2006b: “Abastecimiento y producción local en los campamentos romanos de la región septentrional de la península ibérica”, A. Morillo (ed.), Arqueología Militar Romana en Hispania. Producción y abastecimiento en el ámbito militar, León, 33-74.

MORILLO, A. 1996b: “Lucernas vidriadas de época romana en la península ibérica”, Anuario Universidad Internacional SEK 2, 29-44.

MORILLO, A. 2006c: “Roman military productions in Spain”, A. Morillo & J. Aurrecoechea (eds.), The Roman Army in Hispania. An archeological guide, León, 181-188.

MORILLO, A. 1999a: Lucernas romanas en la región septentrional de la Península IbéricAContribución al conocimiento de la implantación romana en Hispania. Monografías Instrumentum 8, Montagnac.

MORILLO, A.; AMARÉ, MªT.; GARCÍA MARCOS, V. 2005: “Asturica Augusta como centro de producción y consumo cerámico”, Unidad y diversidad en el Arco Atlántico. III Coloquio Internacional de Arqueología en Gijón, BAR Int. Series 1371, Oxford, 139-161.

MORILLO, A. 1999b: “Contramarcas militares en monedas de la Submeseta Norte. Algunas consideraciones generales”, Anejos AEspA XIX (Actas II Encuentro Penínsular de Numismática Antigua, Oporto, 1998), Madrid, 71-90.

MORILLO, A.; FERNÁNDEZ OCHOA, C. 2003: “La bahía de Santander en el marco de la conquista romana de Cantabria”, C. Fernández Ibáñez y J. Ruiz Cobo (eds.), La Arqueología de la bahía de Santander, Santander, 439-450.

MORILLO, A. 2000a: “Anforas y envases perecederos. Nuevas aportaciones sobre la comercialización de aceite bético durante la época romana en la región septentrional de la península ibérica”, Congreso Internacional Ex Baetica Amphoras. Conservas, aceite y vino de la Bética en el Imperio Romano II, (Écija-Sevilla, 1998), Écija, 621-635.

MORILLO, A.; GARCÍA MARCOS, V. 2000: “Nuevos testimonios acerca de las legiones VI victrix y X gemina en la región septentrional de la Península Ibérica”, Actas Deuxième congrès de Lyon sur l’armée romaine: Les légions de Rome sous le HautEmpire, II, Lyon, 589-607.

MORILLO, A. 2000b: “La legio IIII Macedonica en la Península Ibérica. El campamento de Herrera de Pisuerga (Palencia)”, IIe Congrés de Lyon sur l’armée romaine. Les légions de Rome sous le Haut Empire (1998), Lyon, 609-624.

MORILLO, A.; GARCÍA MARCOS, V. 2001: “Producciones cerámicas militares de época augusteotiberiana en Hispania”, Rei Cretariae Romanae Fautores. Acta 37, Abingdon, 147-155.

MORILLO, A. 2000c: “Neue Forschungen zu römischen Lagern der iulisch-claudischen Zeit in Nordspanien”, Bonner Jahrbücher 200, 1-24.

MORILLO, A.; GARCÍA MARCOS, V. 2002: “Twenty years of Roman military archaeology in Spain”. In:

285

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

LIMES XVIII. Proceedings of the XVIIIth International Congress of Roman Frontier Studies held in Amman, Jordan (September, 2000), (BAR Int.Ser.1084), Oxford 2002, vol.2, 779-789.

Iugurthinum 26-27)”. Classical Philology, 2001, 468476. MOUTERDE, R.; MONDÉSERT, C. 1957: “Deux inscriptions grecques de Hama”. Syria 34, 278.

MORILLO, A.; GARCÍA MARCOS, V. 2003a: “Legio VII gemina and its Flavian fortress at León”, Journal of Roman Archaeology 16, 275-286.

MULA, M.J.; ROSSER, P. 1993: “El poblado ibérico amurallado del Cerro de las Balsas (Albufereta, Alicante). Resultados preliminares de los sondeos practicados por el COPHAM (1990-1991)”. Lqnt 1, 105-117.

MORILLO, A.; GARCÍA MARCOS, V. 2003b: “Importaciones itálicas en los campamentos romanos del norte de Hispania durante el periodo augusteo y julioclaudio”, Rei Cretariae Romanae Fautores. Acta 38, Abingdon, 295-304.

MUÑIZ, J. 1982: El sistema fiscal en la España romana. República y Alto Imperio. Zaragoza. MUÑIZ, J. 1998: Cicerón y Cilicia. Diario de un gobernador romano del siglo I AC Huelva.

MORILLO, A.; GARCÍA MARCOS, V. 2004: “Arqueología romana en la ciudad de León: balance de dos décadas de investigaciones”, J. Blánquez Pérez y M. Pérez Ruiz (eds.), Antonio García y Bellido. Miscelánea, Serie Varia 5, Madrid, 263-291.

MUÑIZ, J. 2004: Moral e imperio. La tradición romana sobre el estado (BAR International Series 1254). Oxford.

MORILLO, A.; GARCÍA MARCOS, V.; FERNÁNDEZ OCHOA, C. 2002: Imágenes de arqueología leonesa. Antonio García y Bellido y el Noroeste peninsular en la Antigüedad, Valladolid.

MUÑOZ, A. 1993: “Las cerámicas fenicio-púnicas de origen submarino del área de La Caleta (Cádiz)”, Cuadernos de Prehistoria y Arqueología de Castellón 15, 287-333.

MORILLO, A.; GÓMEZ BARREIRO, M. 2006a: “Circulación monetaria en Herrera de Pisuerga (Palencia)”, MªP. García-Bellido (coord.), Los campamentos romanos en Hispania (27 AC-192 DC). El abastecimiento de moneda, Anejos de Gladius 9, Madrid, 338-421.

NAVEIRO LÓPEZ, J.L. 1981: “As ánforas romanas de A Corunha (I)”. Brigantium 2, 117-126. NAVEIRO LÓPEZ, J.L. 1982: “As ánforas romanas de A Corunha (II)”. Brigantium 3, 63-74. NAVEIRO LÓPEZ, J.L. 1991: El Comercio Antiguo en el Noroeste Peninsular. Museu Arqueolóxico e Histórico, nº 5, La Coruña.

MORILLO, A.; GÓMEZ BARREIRO, M. 2006b: “Circulación monetaria en los campamentos romanos de León”, MªP. García-Bellido (coord.), Los campamentos romanos en Hispania (27 AC-192 DC). El abastecimiento de moneda, Anejos de Gladius 9, Madrid, 258-298.

NAVEIRO LÓPEZ, J.L. 1996: “Registro cerámico e intercambio en el Noroeste en la época romana”. NAVEIRO LÓPEZ, J.L. 2005: Torres de Oeste. Monumento Histórico e Xacemento Arqueolóxico, Pontevedra.

MORILLO, A.; GÓMEZ BARREIRO, M. 2006c: “Las acuñaciones de Tiberio en Turiaso: novedades arqueológicas en los campamentos de la legio VI victrix en León”, XII Congreso Nacional de Numismática (Madrid-Segovia, 2004), Madrid, 375-392.

NAVEIRO LÓPEZ, J.L.; PÉREZ LOSADA, F. 1992: “Un Finisterre Atántico en época Romana: la costa Galaica (N.W.) de la Península Ibérica” BAR International Series 577, 1992, 63-93.

MORILLO, A.; PÉREZ GONZÁLEZ, C. 1990: “Hallazgos monetarios de Herrera de Pisuerga en colecciones privadas”, Actas II Congreso de Historia de Palencia I. Prehistoria, Arqueología e Historia Antigua (1989) Palencia, 443-461.

NÉLIS-CLEMENT, J. 2000: Les beneficiarii: militaires et administrateurs au service de l’empire (Ier s. AC – VI s. PC), Burdeos. NICOLET, C. 1966: “Procurateurs et préfets à l’époque républicaine”. Mélanges d’archéologie, d’épigraphie et d’histoire offerts à Jerôme Carcopino, vol.III, Paris 1966, 691-709.

MORILLO, A.; PÉREZ GONZÁLEZ, C.; ILLARREGUI, E. 1996: “Acerca de las monedas de la caetra procedentes de Herrera de Pisuerga (Palencia) y algunas reflexiones sobre estas acuñaciones del Noroeste peninsular”, Anejos AEspA XIV (I Encuentro Peninsular de Numismática Antigua, 1994), Madrid, 199-206.

NIETO, X.; VIVAR, G; CARRERAS, C. 2005: “La data del naufragi”. In Carreras, C. et al., 156. NOGUERA, J.M. 2002: Ibers al Ebre, col. Daliner 3, Móra d’Ebre. NOGUERA, J.M. 2003: “La escultura hispanorromana en piedra de época republicana”. In L. Abad, ed., cit., 151-208.

MORILLO, A.; PÉREZ GONZÁLEZ, C.; ILLARREGUI, E. 2006: “Asentamientos militares de Herrera de Pisuerga: introducción histórica y arqueológica”, MªP. García-Bellido (coord.), Los campamentos romanos en Hispania (27 AC-192 DC). El abastecimiento de moneda, Anejos de Gladius 9, Madrid, 305-323.

NORDSTRÖM, S. 1968: “Los ojos de los oenochoes ibéricos en el levante español”. Stockholm Studies in Classical Archaeology V.

MORSTEIN-MARX, R. 2001: “The alleged ‘massacre’ at Cirta and its consequences (Sallust, Bellum

ÑACO, A. 2001: “Milites in oppidis hibernabant. El hospitium militare invernal en ciudades peregrinas y 286

BIBLIOGRAPHY

problemas e potencialidades”. Incontro de arqueologia urbana, sep. Bracara Augusta. Braga, XLV / 94, 87-98.

los abusos de la hospitalidad sub tectis durante la República”. DHA 27.2, 2001, 63-90. ÑACO, A. 2003a: Vectigal Incertum. Economía de guerra y fiscalidad republicana en el occidente romano: su impacto histórico en el territorio (218133 AC). BAR Int. S. 1158, Oxford.

OTT, J. 1995: “Die Mechanismen bei der Befördenung von Beneficiariern der Statthalter”. In Y. Le Bohec (ed) La hiérarchie (rangordnung) de l’armée romaine sous le Haute-Empire, Lyon.

ÑACO, A. 2003b: “Roman Realpolitik in taxing Sardinian Rebels (177-175 BC)”. Athenaeum (Athenaeum. Studi periodici di letteratura e storia dell’antichità – Pavia)91.2, 2003, 531-540.

PAÇO, A.; JALHAY, E. 1955: “Tesouro monetário da Citânia de Sanfins”. Anais da Academia Portuguesa da História. 2ª série, 189-275.

ÑACO, A. 2006: “Rearguard strategies of Roman Republican Warfare in the Far West”. In T. Ñaco, I. Arrayás, (eds.), 149-167.

PAIVA, M.B.C. 1993: Ânforas romanas de castros da fachada atlântica do norte de Portugal. Dissertação de Mestrado em Arqueologia apresentada à Faculdade de Letras da Universidade do Porto, Porto, policopiado. [unpublished].

ÑACO, A. 2007: “The Late Republican West: imperial taxation in the making?”. In O. Hekster, G. De Kleijn, D. Slootjes (eds.), Crises and the Roman Empire. Proceedings of the Seventh Workshop of the International Network Impact of Empire. Nijmegen, June 20-24, 2006, Leiden-Boston 2007, 219-231.

PAIXÃO, A.C. 2001: “Alcácer do Sal proto-histórica no contexto mediterrânico”. In Os Púnicos no Extremo Ocidente. Actas do Colóquio Internacional (Lisboa, 27 e 28 de Outubro de 2000). Lisboa: Universidade Aberta, p. 149-172.

ÑACO, A.; ARRAYÁS, I. (eds.) 2006: War and territory in the Roman world. Guerra y territorio en el mundo romano. BAR Int.Ser. 1530, Oxford.

PALMADA, G. 2003: “La fortificació republicana d’Olèrdola (Sant Miquel d’Olèrdola, Alt Penedès)”. Revista d’Arqueologia de Ponent 13, 2003, 257-288.

ÑACO, A.; PRIETO, A. 1999: “Moneda e historia monetaria en la Hispania republicana: ¿economía, política, fiscalidad?”. Studia Historica. Historia Antigua 17, 193-241.

PAMMENT, J. 1996: Roman Republican Castrametation. A reappraisal of historical and archaeological sources (BAR Int.Ser. 630), Oxford.

OLCINA, M.H. 2005: “La Illeta dels Banyets, El Tossal de Manises y La Serreta”. L. Abad, F. Sala, I. Grau, cit., 147-178.

PAMMENT, J. 1997: “A possible function for the location of the Roman Republican fortress at Cáceres el Viejo in Extremadura, Western Spain”. In W. Groenman-Van Waateringe, et al. (eds.), Roman Frontiers Studies-1995. Proceedings of the XVIth International Congress of Roman Frontier Studies. Oxford, 53-58.

OLCINA, M.H. 2006: “Lucentum: origin and evolution of a Roman municipium in the sinus Ilicitanus”. In L. Abad, S. Keay, S. Ramallo, eds., Early Roman towns in Hispania Tarraconensis, JRA supplementary series 62, Portsmouth, Rhode Island, 105-117.

PANELLA, C. 1983: “Le anfore di Cartagine: nuovi elementi per la ricostruzione dei flussi commerciali del Mediterraneo in età imperiale romana”. Opus II, 53-75.

OLESTI, O. 1995: El territori del Maresme en època republicana (s.III-I AC). Estudi d’arqueomorfologia i història. Mataró.Orton, C.; Tyers, P.; Vince, A. (1993) Pottery in archaeology. Cambridge.

PANELLA, C. 1989: “Le anfore italiche del II secolo DC”. CEFR 114, 139-178.

OLESTI, O. 2006: “El control de los territorios del Nordeste peninsular (218-100 BC): un modelo a debate”. In T. Ñaco; I. Arrayás (eds.) 119-148.

PARKER, A.J. 1992: Ancient Shipwrecks of the Mediterranean & the Roman Provinces. BAR International Series, 580, Oxford.

OLMOS, R. 2000: “El vaso del ‘ciclo de la vida’ de Valencia: una reflexión sobre la imagen metafórica en época iberohelenística”. AEspA 73, 59-85.

PARKER, V.L. 2004: “Romae omnia venalia esse. Sallust’s Development of a Thesis and the Prehistory of the Jugurthine War”. Historia 53.4, 2004, 408-423.

OLMOS, R. 2003: “Combates singulares: lenguaje de afirmación de Iberia frente a Roma”. T. Tortosa, J. Santos, eds., Arqueología e Iconografía. Indagar en las imágenes, Roma, 79-97.

PARODI, M.J. 2001: Ríos y lagunas de hispania como vías de comunicación. La navegación interior en la Hispania romana. Ecija: Editorial Graficas Sol.

OLMOS, R. 2005: “Iconografía celtibérica”. A. Jimeno, ed., Celtíberos, Soria, 253-260.

PASCHOUD, F. 1983: “Frumentarii, Agentes in rebus, Magistrani, Curiosi, Veredarii: problèmes de terminologie”. Bonner Historia-Augusta Colloquium 1979/81, 215-243.

ORTEGA, J.R.; ESQUEMBRE, M.A.; CASTELLÓ, J.S.; MOLINA, F.A. 2003: “Una pieza singular: la terracota de una birreme del poblado ibérico del Cerro de las Balsas (La Albufereta, Alicante)”. Saguntum 35, 147-158.

PASCUAL BERLANGA, G.; RIBERA i LACOMBA, A. 2001: “El consumo de productos béticos en Valentia y su entorno: la continuidad de una larga tradición”. Actas del Congreso Internacional Ex Baetica Amphorae. Conservas, Aceite y Vino de la Bética en

OSÓRIO, MªI.N.P.; SILVA, A.M.S.P. 1994: “Intervenção arqueológica municipal no Porto: resultados, 287

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

el Império Romano (Écija-Sevilha 1998), Écija. Vol. II, p. 565-576. PAVIS D’ESCURAC, H. 1976: La préfecture de l’Annone, service administratif impérial d’Auguste à Constantin. Rome. PAYEN, P. 2002: “Sertorius et l’Occident”. Pallas 60, 93-115. PEACOCK, D.P.S.; WILLIAMS, D.F. 1986: Amphorae and the Roman economy. London. PEÑA SANTOS, A. 1986: Yacimiento galaico-romano de Santa Trega (A Guarda, Pontevedra). Campaña de 1983, Arqueoloxía / Memorias 5. Xunta de Galicia, Santiago de Compostela, 1986. PERALTA, E. 1999: “Los castros cántabros y los campamentos romanos de Toranzo y de Iguña. Prospecciones y sondeos (1996-97)”, Las Guerras Cántabras, Santander, 201-276. PERALTA, E. 2000: Los cántabros antes de Roma, RAH 5, Madrid. PERALTA, E. 2001: “Los castra aestiva del bellum cantabricum: novedades arqueológicas”, I Congreso Internacional de Historia Antigua (2000), Valladolid, 173-182. PEREA, S. 1998: Los stratores en el ejército imperial (funciones y rangos). Madrid. PEREA, S. 2006: “El uso de la sal en el ejército romano y su abastecimiento”. In A. Morillo (ed.) Arqueología militar romana en Hispania. León, 345-359.

Congreso de Arqueología Peninsular, IV, Porto, 1993), 259-267. PÉREZ LOSADA, F. 2002: Entre a cidade e a aldea. Brigantium 13. Coruña. PÉREZ LOSADA, F.; FERNÁNDEZ FERNÁNDEZ, A.; VIEITO COVELA, S. 2008: “Toralla y las villas de la Gallaecia atlàntica”. In Fernández Ochoa (C.), GarcíaEntero (V.) y Gil Sendino (F.) (Eds.): Las villae tardorromanas en el occidente del Imperio: arquitectura y función. IV Coloquio Internacional de Arqueología en Gijón, Gijón, 479-504. PÉREZ MACÍAS, J.A.; DELGADO DOMÍNGUEZ, A. 2007: “Los metalla de Riotinto en época julioclaudia” in Las minas de Riotinto en época julioclaudia, eds. J.A. Pérez Macías & A. Delgado Domínguez, Huelva, 37-183. PETRIKOVITS, H. von 1974: “Militärische Fabricae der Römer”. In Actes du IX Congrès International d’Etudes sur les frontières romaines (Marmara, 1972). (Bucarest) 399-407. PETRIKOVITS, H. von 1976a: “Römisches Militärhandwerk”. In Beiträge zur römischen Geschichte und Archäologie von 1931-1970. (Bonn), 598-611. PETRIKOVITS, H. von 1976b: “Militärische Fabricae der Römer”, Beiträge zur Römischen Geschichte und Archäologie. 1931 bis 1974, Bonn, 612-619. PFLAUM, H.G. 1960: Les carrières procuratoriennes équestres sous le Haut-Empire romain, Paris. PICARD, G.CH. 1947: Thophées d’Auguste a SaintBertrand-de-Commingues, en Memoires de la Société Archéologique du Midi de la France 21, 5-52.

PEREIRA, I. 1993: “Figueira da Foz. Santa Olaia”. Estudos Orientais. Lisboa: I.O.U.N.L., IV, p. 285-304. PEREIRA, I. 1997: “Santa Olaia et le Commerce Atlantique (Itinéraire Industriel et Maritime) ”. In Itinéraires Lusitaniens. Paris: s.n., p. 209-254. PÉREZ BALLESTER, J. 2003: “El comercio: rutas comerciales y puertos”. In H. Bonet, R. Albiach, M. Gozalbes, coords., Romanos y Visigodos en tierras valencianas, Diputación de Valencia, Valencia, 115130. PÉREZ GONZÁLEZ, C. 1989: Cerámica romana de Herrera de Pisuerga (Palencia, España): la terra sigillata, Santiago de Chile. PÉREZ GONZÁLEZ, C. 1996: “Asentamientos militares de Herrera de Pisuerga”, C. Fernández Ochoa (coord..), Coloquio Internacional de Arqueología: los finisterres atlánticos en la antigüedad (época prerromana y romana), Gijón, 91-102.

PIMENTA, J. 2003: “Contribuição para o estudo das ânforas do Castelo da São Jorge (Lisboa)”. Revista Portuguesa de Arqueologia. Lisboa, 6: 2, 341-362. PIMENTA, J. 2005: As Ânforas Romanas do Castelo de São Jorge (Lisboa), Trabalhos de Arqueologia, 26, Instituto Português de Arqueologia, Lisboa. PINA, F. 2003: “¿Por qué fue reclutada la turma Salluitana en Salduie?”. Gerión 21.1, 197-204. PINA, F.; PÉREZ CASAS, J.A. 1998: “El oppidum Castra Aelia y las campañas de Sertorius (77-76 AC)”. JRA 11, 245-264. PINTO, I.V. 2003: “A cerâmica comum das villae romanas de São Cucufate (Beja)”. Colecção Teses. Lisboa: Universidade Lusíada. PINTO, I.V.; MORAIS, R. 2005: “Complemento de comércio das ânforas: cerâmica comum bética no território português”. In Actas del Congreso Internacional CETARIAE. Salsas y salazones en Occidente durante la Antigüedad. Universidad de Cádiz (noviembre 2005). BAR Int. Ser. 1686, Oxford, 235-254. POMEY, P. 1979: La Navigation dans l’Antiquité. Aixen-Provence. PONSICH, M. 1968: Alfarerías de época fenicia y púnico-mauritana en Kuass (Arcila, Marruecos), Papeles del Laboratorio de Arqueología de Valencia 4, Valencia.

PÉREZ GONZÁLEZ, C.; FERNÁNDEZ IBÁÑEZ, C. 1989: “Sellos de alfarero sobre “mortaria” en la península ibérica”, Publicaciones Instituto Tello Téllez de Meneses 60, 67-98. PÉREZ GONZÁLEZ, C.; ILLARREGUI GOMEZ, E.; FERNÁNDEZ IBÁNEZ, C. 1991: “Pisoraca. Un interesante conjunto de yacimientos arqueológicos”, Revista de Arqueología 120, 18-26. PÉREZ GONZÁLEZ, C.; ILLARREGUI, E. 1994: “Un taller de útiles óseos de la Legión IIII Macedónica”, Trabalhos de Antropologia e Etnologia 34 (3-4) (I 288

BIBLIOGRAPHY

jecto de investigação arqueológica e arqueométrica”. In Simposio A Costa portuguesa no panorama da rota Atlântica durante a época romana. Peniche. [unpublished].

PONSICH, M. 1974: Implantation Rurale Antique Sur le Bas Guadalquivir I, Paris. PONSICH, M. 1979: Implantation Rurale Antique Sur le Bas Guadalquivir II, Paris. PONSICH, M. 1987: Implantation Rurale Antique Sur le Bas Guadalquivir III, Paris. PONSICH, M. 1988: Aceite de oliva y salazones de pescado. Factores geo-económicos de Baetica y Tingitania, Madrid. PONSICH, M. 1991: Implantation Rurale Antique Sur le Bas Guadalquivir IV, Paris. PONSICH, M. 1993: “Le circuit du détroit de Gibraltar dans l’antiquité”, Homenatge a Miquel Tarradell, Curial, Barcelona, 49-62. PONSICH, M., TARRADELL, M. 1965: Garum et industries antiques de salaisons dans la Méditerranée occidentale, Bibl. École de hauts études hispaniques 36, París. POPOVIC, M. 1989: “Une station de bénéficiares à Sirmium”. Comptes Rendus Paris, 116-122.

REAL, M.L.; TÁVORA, M.J.; OSÓRIO, M.I.P.; TEIXEIRA, F.F. 1985-86: “Escavações arqueológicas no Morro da Sé. sep.” Boletim Cultural da Câmara do Porto. 2ª série: 3 / 4, 7-42. REDDÉ, M. 1979a: “La navigation au large des côtes atlantiques de la Gaule à l’époque romaine”, Mélanges Ecoce Française de Rome. Antiquité 91, 481-489. REDDÉ, M. 1979b: “La représentation des phares à l’époque romaine”, Mélanges Ecoce Française de Rome Antiquité 91, 845-872. REINOSO DEL RIO, M.C. 2002: “Cerámica de paredes finas en el asentamiento militar de Herrera de Pisuerga (Palencia)”, A. Morillo (coord.), Arqueología Militar Romana en Hispania, Anejos de Gladius 5, Madrid, 369-380. REMESAL RODRÍGUEZ, J. 1977-8: “Economía oleícola bética: nuevas formas de análisis”. AEspA 1977-8, 87-142.

PRAG, J. 2007: “Auxilia and Gymnasia: a Sicilian Model of Roman Imperialism”. JRS 97, 68-100. PRIEGO, M.; RUIZ-GÁLVEZ, M. 1998: La Europa Atlántica en la Edad del Bronce.Un viaje a las raíces de la Europa occidental. Barcelona.

REMESAL RODRÍGUEZ, J. 1983: “Transformaciones en la exportación del Aceite Bético a Mediados del Siglo III DC”, in: J.Mª Blázquez y J. Remesal (eds.), p. 115-131.

PURPURA, G. 1989: “Pesca e stabilimenti antichi per la lavorazione del pesce in Sicilia, III. Torre Vindicari (Noto), Capo Ognina (Siracusa)”, Sicilia Archeologica XXII, pp. 25-37. QUESADA, F. 2006: “Armamento indígena y romano republicano en Iberia (siglos III-I AC): compatibilidad y abastecimiento de las legiones republicanas en campaña”. In A. Morillo (ed.), Arqueología militar romana en Hispania. Producción y abastecimiento en el ámbito militar. León, 75-96.

REMESAL RODRÍGUEZ, J. 1986: La annona militaris y la exportación de aceite bético a Germania. Madrid. REMESAL RODRÍGUEZ, J. 1990: “Die Procuratores Augusti und die Versorgung des römischen Heeres”. In 14. Limeskongress, (Viena, 1986), 55-65. REMESAL RODRÍGUEZ, J. 1992: “Instrumentum domesticum e storia economica: le anfore Dressel 20”. Opus XI, 105-113. REMESAL RODRÍGUEZ, J. 1996: “Mummius Secundinus. El Kalendarium Vegetianum y las confiscaciones de Severo en la Bética (HA Severus 12-13).” Gerión 14, 195-221.

RABANAL, M.; GARCÍA, S. 1996: “La red viaria en el noroeste hispanorromano su importancia y significado”. Hispania Antiqua 20, 273-296.

REMESAL RODRÍGUEZ, J. 1997: Heeresversorgung und die wirtschaftlichen Beziehungen zwischen der Baetica und Germanien. (Stuttgart).

RAMALLO, S.; BROTONS, F. 1997: “El santuario ibérico de La Encarnación (Caravaca de la Cruz, Murcia)”. QPAC 18, 257-268. RAMALLO, S.; RUIZ, E. 2002: “Cartago NovACapital de Hispania Citerior”. In J.L. Jiménez, A. Ribera, coords., Valencia y las primeras ciudades romanas de Hispania, Ayuntamiento de Valencia, Valencia, 113122. RAMON, J., 1995: Las ánforas fenicio-púnicas del Mediterráneo central y occidental, col. Instrumenta 2, Barcelona. RAMON, J. 2004: “La ciutat romana d’Ebussus”. In M. Orfila, M.A. Cau, coords., Les ciutats romanes del llevant peninsular i les illes Balears, Barcelona, 261313. RAPOSO, J.; FABIAÕ, C.; GUERRA, A.; BUGALHÃO, J.; DUARTE, A.L.; SABROSA, A.; DIAS, M.I. PRUDENCIO, M.I. “Ãnforas romanas do Estuário do Tejo: balanço e perspectivas de um pro-

REMESAL RODRÍGUEZ, J. 2004: “Alfares y producciones cerámicas en la provincia de Córdoba. Balance y perspectivas”, Actas del Congreso Internacional Figlinae Baeticae.Talleres Alfareros y Producciones Cerámicas en la Bética Romana (ss. II AC – VII DC), BAR, International Series, Oxford, p. 349-361 REMESAL, J.; REVILLA, V.; CARRERAS, C.; BERNI, P.1997: “Arva: prospecciones en un centro productor de ânforas Dressel 20 (Alcolea del Río, Sevilla)”, Pyrenae, 28, p. 151-178. REVILLA, J. 1993: El alfar romano de l’Aumedina (Tivissa, Tarragona). Producción cerámica y economía rural en el Bajo Ebro. Barcelona. REVILLA, V. 1995: Producción cerámica, viticultura y propiedad rural en Hispania Tarraconense (siglos I AC-III DC), Barcelona.

289

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

RODRÍGUEZ MARTIN, F.G.; GORGES, J.G. 1999: “Prensas de aceite y de vino en una villa romana de la cuenca media del Guadiana: Torre-Aguila, Bardaño (Badajoz)”. In J.G. Gorges and F.G. Rodríguez Martin (eds.) Économie et territoire en Lusitanie romaine. Paris, 403-426.

RIBEIRO, O.; LAUTENSACH, H.; DAVEAU, S. 1987: Geografia de Portugal. Lisboa: Edições J. Sá da Costa. RIC I2 = SUTHERLAND, C.H.V. 1984: The Roman Imperial Coinage I. From 31 BC to AD 69, London. Mócsy, A. (1967) “Zu den Prata Legiones”. In H. Schönberger (ed.) Studien zu den Militärgrenzen Roms, 211-214.

RODRÍGUEZ MARTÍNEZ, R.Mª 2005: “Avance de resultados da escavación arqueolóxica no Castro Grande de O Neixón. Campaña de 2003. Análise do material recuperado”. en Ayan Vila (X.M.) (coord.), Os castros de Neixón, Noia, 2005, 150-157.

RICHARDSON, J.S. 1976: “The Spanish mines and the development of provincial taxation in the second century BC”. JRS 66, 147ff.

RODRÍGUEZ NEILA, J.F. 1992: Confidentes de César. Los Balbos de Cádiz. Cádiz.

RICHARDSON, J.S. 1986: Hispaniae. Spain and the development of Roman imperialism, 218-82 BC. Cambridge.

RODRÍGUEZ RODRÍGUEZ, I. 2001: “Las áreas artesanales. los alfares”, Carmona Romana. Actas del II Congreso de Historia de Carmona (29 sept.-2 octubre 2001), ed. A. Caballos Rufino, Carmona, 311-320.

RICHMOND, I.A. 1935: “Trajan’s army on Trajan’s column”. PBSR 13, 1-40. RILEY, J.A. 1979: “Coarse pottery”. In J.A. Lloyd (ed.) Excavations at Sidi Khrebish/Benghazi (Berenice). Suppl. Lybia Antiqua V.II, 91-446.

ROLDÁN, J.M. 1974: Hispania y el ejército romano. Salamanca. ROMO SALAS, A.S.; VARGAS JIMÉNEZ, J.M. 2001: “Azanaque. Evidencias Arqueológicas de un centro de producción anfórica”. In Actas del Congreso Internacional Ex Baetica Amphorae. Conservas, Aceite y Vino de la Bética en el Império Romano, Vol. I, Écija, p. 405-417.

RINGROSE, D.R. 1972: Los transportes y el estancamiento económico de España (1750-1850). Madrid. RIPOLLÉS, P.P. 1982: Circulación Monetaria en la Tarraconense Mediterránea. Valencia. RIPOLLÉS, P.P. 2005: “Coinage and Identity in the Roman Provinces: Spain”. In Ch. Howgego, V. Heuchert, A. Burnett (eds.), Coinage and Identity in the Roman Provinces, Oxford, 79-93.

ROSELLÓ, E.; CAÑAS, J.M. 1994: “Análisis de la ictiofauna de la Plaza del Marqués (Gijón)”, in Fernández Ochoa, C., ed., Una industria de salazones de época romana en la Plaza del Marqués, Gijón, pp. 189-197.

RIPOLLÈS, P.P.; LLORENS, MªM. (eds.) 2002: ArseSaguntum. Historia monetaria de la ciudad y de su territorio, Bancaja, Sagunto.

ROTH, J. 1999: The logistics of the Roman army at war. Leiden.

RIZZO, G. 2003: Instrumenta Urbis I. Ceramiche fini da mesa, lucerne ed anfore a Roma nei primi due secoli dell’Impero. CEFR 207, Roma.

ROUGÉ, J. 1975: La Marine dans l’Antiquité. Paris. ROWLAND, R.J. 2001: The Periphery in the Center. Sardinia in the Ancient and Medieval Worlds, (BAR Int.Series, 970), Oxford.

ROBINSON, H.S. 1959: The Athenian Agora. Vol. V: Pottery of the Roman period. Princeton. RODÀ, I. 1993: “Escultura republicana en la Tarraconense: el monumento funerario de Malla”. In T. Nogales, coord., I RERH, Mérida, 207-219.

ROYMANS, N. 1996: “The sword or the plough. Regional dynamics in the romanisation of Belgic Gaul and the Rhineland area”. In N. Roymans (ed.), From the Sword to the Plough. Three studies on the earliest romanisation of Northern Gaul, Amsterdam 1996, 9126.

RODÀ, I. 1998: “El papel de Agripa en la trama urbana de la Hispania augustea” en Los orígenes de la ciudad en el noroeste hispánico. Actas del congreso internacional, Lugo 15 – 18 de mayo 1996. (Lugo 1998), 275-293.

ROYMANS, N. 2004: Ethnic Identity and Imperial Power. The Batavians in the Early Roman Empire, Amsterdam 2004.

RODÀ, I. 2004: “Agripa y el comercio del plomo”, Mastia 3, 183-194.

RPC = BURNETT, A, AMANDRY, M.; RIPOLLES, P.P. 1992: Roman Provincial Coinage. I. From the death of Caesar to the death of Vitellius, LondonParis.

RODÀ, I. 2007: “La figura de Agripa, el ejército y la promoción de las obras públicas en Hispania”. Larouco 4, 27-44. RODDAZ, J.M. 1993: “Agripa y la Península Ibérica”, Anas 6, 117-122.

RUIZ-GALVEZ, M. 1986: “Navegación y comercio entre el Atlántico y el Mediterraneo a fines de la Edad del Bronce” T.P. vol 43, 9-42.

RODRIGUEZ ALMEIDA, E. 1989: Los tituli picti de las ánforas olearias de la Bética, 1989, Madrid, 35-37.

RUSSELL, F. 1951: “As escavaçoes arqueologiques do castellum de Fonte do Milho”. Anais do Instituto de vinho de Porto 12, pp.17-88.

RODRÍGUEZ COLMENERO, A. 1996: “Lucus Augusti, ¿una ciudad sin ambiente?”. In Rodríguez, A. (ed.), Lucus Augusti I. El amanecer de una ciudad. Coruña 1996, 419-424.

SADDINGTON, D.B. 1985: “Praefecti fabrum on the Julian-Claudian period”. In E. Webber and G. 290

BIBLIOGRAPHY

SANTERO, J.M. 1981: “Collegium stratorum”. Habis 12, 261-272. SANTOS YANGUAS, N. 1991: La romanización de Asturias. Madrid.

Dobesch (eds.) Römische Geschichte, Altertumskunde und Epigraphik: Festchrift Artur Betz zur Vollendug seines 80. Lebensjahres. Wien, 529-546. SÁEZ FERNÁNDEZ, P. 2002: “Algunas consideraciones sobre el territorio de las ciudades de la Bética”, in González Román, C. e Padilla Arroba, A. (eds.), Estudios sobre ciudades de la Bética, Granada, p. 389-445.

SANTOS YANGUAS, N. 2001: “Vías y comercios en las sociedades del norte peninsular en la antigüedad: el territorio de Aurigotes, caristios y várdulos”. In F. Wulff; G. Cruz and C. Martínez Maza, (eds.), Comercio y Comerciantes en la Historia Antigua de Málaga. (Siglo VIII AC – año 711 DC). Actas del IIdo. Congreso de Historia Antigua de Málaga. 1998. Málaga, 2001. II Congreso de Historia Antigua de Málaga. Málaga.

SÁEZ FERNÁNDEZ et al. 2001: “Excavación arqueológica de urgencia en el alfar romano Las Delícias (Écija, Sevilla)”, Anuario Arqueológico de Andalucía 1997, Vol. III, Sevilha p. 562-575. SÁEZ FERNÁNDEZ, P.; TINOCO MUÑOZ, J.; GARCÍA VARGAS, E.; GARCÍA-DILS DE LA VEGA, S. 2001: “Excavación arqueológica de urgencia en el alfar romano de Las Delicias (Écija, Sevilla), 1997”, Anuario Arqueológico de Andalucía, vol. III, 562-575.

SANTOS YANGUAS, N. 2004: “Una inscripción en Villalís (León): Los procuratores metallorum y la administración minera romana del oro en territorio de los Astures”. Astórica 23, 9-34. SANTOS, M.J. 2006: “O Largo Vitorino Damásio (Santos-o-Velho, Lisboa: contributo para a história da zona ribeirinha de Lisboa”. Revista Portuguesa de Arqueologia. Lisboa: Instituto Português de Arqueologia, Vol. 9, nº 2, 369-399.

SÁEZ ROMERO, A.M. 2008: “La producción de ánforas en el área del Estrecho en época tardopúnica (siglos III-I AC)”. In Bernal Casasola (D.), Ribera i Lacomba (A.) (eds.), Cerámicas hispanorromanas. Un estado de la cuestión, Cádiz, 2008, 635-659.

SANZ, R. 1997: Cultura ibérica y romanización en tierras de Albacete: los siglos de transición, Inst. de Estudios Albacetenses, Albacete.

SÁEZ, A., BUSTAMANTE, M., BERNAL, D. & LORENZO, L. “Excavando en la orilla africana del Estrecho. Síntesis estratigráfica de la actuación arqueológica preventiva en la Plaza de África nº 3 (Ceuta)”, Tabona, Homenaje a Pilar Acosta, Tenerife, [forthcoming].

SCARDIGLI, B. 2002: “Trent’anni di studi sertoriani”. In G. Urso (a cura di), Hispania Terris Omnibus Felicior. Premesse ed esiti di un processo di integrazione. Atti del convegno internazionale, Cividale del Friuli 27-29 settembre 2001. Pisa, 143161. SCHALLMAYER, E. 1989: “Zur Herkuntf und Funktion der Beneficarier”. Roman Frontier Studies, 400-406.

SAFRAI, Z. 1994: The economy of the Roman Palestine, London. SALINAS DE FRÍAS, M. 2006: “Geografía real y ficticia de la epopeya sertoriana”. In G. Cruz Andreotti; P. Le Roux; P. Moret (eds.), La invención de una geografía de la Península Ibérica. I. La época republicana, Madrid, 153-174. SALOMON, M. 1996: “De la via Heraclea à la via Domitia” Archéologie en Languedoc 20.2, 99-108. SÁNCHEZ-PALENCIA, F.J. 1995: “Minería y metalurgia en la región astur en la antigüedad”, Astures. Pueblos y culturas en la frontera del Imperio Romano, Gijón, 141-158. SÁNCHEZ-PALENCIA, F.; FERNÁNDEZ, M.D.; OREJAS, A.; SASTRE, I.; RUIZ, M. 2007: “Minería romana de oro del Noroeste de Hispania”. In A. Morillo (eds.) El ejército romano en Hispania. León, 135-158. SÁNCHEZ PINTO, L. (2004): “Antiguas producciones naturales de Canarias”, Fortunatae Insulae. Canarias y el Mediterráneo, Tenerife, pp. 209-219.

SCHALLMAYER, E. 1990: Der römische Weihebezirck von Osterburkern I. Forschungen und berichte zur vor-und frühgeschichte in Baden-Wüttemberg. Band 40. (Stuttgart). SCHLIPPSCHUH, O. 1987: Die Händler im römischen Kaiserreich in Gallien, Germanien und den Donauprovinzen, Rätien, Noricum und Pannonien, Amsterdam. SCHULTEN, A. 1894: “Das territorium legionis”, Hermes 29, Wiesbaden, 481-516. SCHULTEN, A. 1922: Fontes Hispaniae antiquae, I. Avieno, Ora Marítima. Barcelona-Berlin. SCHULTEN, A. 1940: Fontes Hispaniae antiquae, V. Las guerras de 72-19 a. de J. C.. Barcelona. SEALEY, P.R.; TYERS, P.A. 1989: “Olives from Spain: a unique amphora find in British waters”. Antiquaries Journal 69, 53-72. SELKIRK, R. 1983: The Piercebridge formula: a dramatic new view of Roman history, Cambridge.

SANDER, E. 1962: “Der praefectus fabrum und die Legionsfabriken”, Bonner Jahrbücher 162, 139-161. SANMARTÍ, J.; UGOLINI, D.; RAMON, J.; ASENSIO, D. (eds.) 2004: La circulació d’àmfores al Mediterrani occidental durant la Protohistoria (segles VIII-III AC): aspectes quantitatius i anàlisi de continguts, II Reunió Internacional d’Arqueologia de Calafell, Arqueomediterrània 8, Barcelona.

SENNEQUIER, G.; COLONNA, C. 2003: L’Algérie au temps des royaumes numides. Ve siècle av. J.-C. – Ier siècle ap. J.-C., París. SEVILLANO, Mª A.; VIDAL ENCINAS, J.M. 2001: Urbs magnifica. Una aproximación a la arqueología 291

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

SWAN, V.G. 1997: “Vexillations and the garrisons of Britannia in the second and Early Third centuries: a ceramic view-point”, W. Groenman-van Waateringe, B.L. van Beek, W.J.H. Willems & S.L. Wynia (eds.), Roman Frontier Studies 1995. Proceedings of the XVIth International Congress of Roman Frontier Studies, Oxbow Monograph 91, Oxford, 289-294.

de Asturica Augusta (Astorga, León). Museo romano (Guía-catálogo), Astorga. SHENNAN, S. 1988: Quantifying archaeology. Edinburgh. SIDEBOTHAM, S.E. 1986: Roman economy policy in the Erithrean Thalassa: 30 BC – AD 217 Leiden. SILLIÉRES, P. 1997: “Producción, transporte y comercialización del aceite y el vino hispánicos”, J. Morilla, J. Gómez-Pantoja y P. Cressier (eds.), Impactos exteriores sobre el mundo rural mediterráneo, Madrid, 87-109.

SYME, R. 1969: “A Governor of Tarraconensis”, Epigraphische Studien 8, 125-133. TARRADELL, M. 1950: “Sobre unos discos de cerámica procedentes de Tamuda y sus paralelos”, I Congreso Nacional de Arqueología (Almería, 1949), Cartagena, pp. 326-330.

SILLIÉRES, P. 2005: “Les voies de communiation dans les provinces occidentales de l’Empire romaine” En M.M. Urteaga and MªJ. Noain (coord.) Mar exterior. El occidente atlántico en época romana. Actas del Congreso Internacional, Pisa (2003). Roma, 2132. SILVA, A.C.F. 2000: “Proto-história e romanização do Porto”. Al-madan, II série: 9 (Outubro), 94-103.

TAYLOR, A. 2002: “Supplying the frontier zones: the rôle of the East Anglia Fens”, F. Freeman, J. Bennett, Z.T. Fiema & B. Hoffmann (eds.), Limes XVIII. Proceedings of the XVIIIth International Congress of Roman Frontier Studies (Amman, 2000), Bar Int. Series 1084 (II), Oxford, 843-850. TCHERNIA, A. 1971: “Les amphores vinaires de Tarraconnaise et leur exportation au debut de l’Empire”, AEspA 44, 38-85.

SIMPLÍCIO, M.C.; TEIXEIRA, S.B.; BARROS, P. 2000: “Arqueologia e Geodinâmica do Litoral – O caso de Quarteira (Algarve – Portugal)”. “Terrenos” da Arqueologia da Península Ibérica. Actas do 3º Congresso de Arqueologia Peninsular. Setembro de 1999. UTAD. Vila Real. Porto: ADECAP, 609-622.

TCHERNIA, A. 1986: Le vin de l’italie romaine. Roma. TCHERNIA, A.; POMEY, P.; HESNARD, A. 1978: L’épave romaine de La Madrague de Giens (Var). Gallia suppl. 34, Paris.

SIRKS, B. 1991: Food for Rome. Amsterdam. SOARES, A.M. 1887: “O teor em radiocarbono das conchas marinhas: um indicador paleo-oceanográfico”. Setúbal Arqueológica. Setúbal, 11-12, 1725. SOLANA, J.M. 2001: “Los caminos de Hispania hace dos mil años recogidos en las fuentes escritas”. Solana (ed.) La Península Ibérica hace 2000 años. Actas del I Congreso Internacional de Historia. Valladolid.

TEICHNER, F. 1997: “Note sur le fonds numismatique romain de Foz do Rio Arade (Portimão, Portugal)”. ConimbrigACoimbra. 36, 123-160.

SOLOVERA, M.E. 1987: Estudios sobre la historia económica de la Rioja romana. Logroño.

TIBILETTI, G. 1953: “Governatori romani in città provinciali”. Istituto Lombardo di Scienze e Lettere. Rendiconti. Classe di Lettere e Scienze Morali e storiche 86, 64-100.

TEIGELAKE, U. 2003: “Tracing Ship Trafic Without Ships – Alternative Methods of Finding Evidence for Pre- and Early Historical Inland Water Transport”. In Beltrame, C. (ed.) (2003) – Boats, Ships and Shipyards. Oxford: Oxbow Books, 154-159.

SOMOZA, J. 1908: Gijón en la Historia General de Asturias, Vol. I, Gijón, (ed. facsimil, 1971).

TIERNEY, J. 1963: “The map of Agrippa”, Proceeding of the Royal Irish Academy 63, 151-166.

SORIGUER, M., ZABALA, C.; HERNANDO, J. 2007: Malacofauna e ictiofauna del yacimiento de la Plaza de África de Ceuta, Unpublished Report, Instituto de Estudios Ceutíes (Ayudas a la Investigación, 2006), Ceuta. SOTOMAYOR, M. 1969: “Hornos romanos de ánforas en Algeciras”, X Congreso Nacional de Arqueología, Madrid, 389-399. STEFANI, G. 1991: Uomo e ambiente nel territorio vesubiano. Guida all’Antiquarium di Boscoreale, Herculano. STOLL, O. 1997: “Die Beneficiarier-Randorgung und Funktion. Einige Bemerkungen zue neueren Forschung”. Laverna 8, 93-112. STOLLE, F. 1914: Der römische Legionär un sein Gepäck (Mulus Marianus): Eine Abhandlung über den Mundvorrat die Gepäcklagt und Tornister des römischen Legionärs und im Anhang Erklarung der Apokolypse 6.6. Strasbourg.

TOMAS, V.; RODRÍGUEZ COLMENERO, A. 2007: “El nuevo exsactor de Brigantium”, Larouco 4, 221222. TOMBER, R. 1992: “Early Roman pottery from Mons Claudianus”. Cahiers de la Cerámique Egyptienne 3, 137-142. TORRES, J. 1978-79: “Tesouro monetário do castro de Alvarelhos: estudo numismático: seriação cronológica e histórica”. Santo Tirso Boletim Cultural, 1: 2-3, 15250. TORRES, V.C; CASTRO, J.C.; PRIETO, S. 2007: “La factoría romana de salzón de “O Areal”, Vigo (Galicia): un complejo industrial salazones altoimperial, en Lagóstena (L.), Bernal (D), Arévalo (A) (eds.), Cetariae 2005. Salsas y Salazones de Pescado en Occidente durante la Antigüedad, BAR International Series 1686, 2007, 475-485.

292

BIBLIOGRAPHY

VAQUERIZO, D. 1999: La Cultura Ibérica en Córdoba. Un ensayo de síntesis, Córdoba.

TORTOSA, T. 2003: El yacimiento de la Alcudia (Elche). Pasado y presente de un enclave ibérico, Anejos de AEspA, Madrid.

VARGAS, M. 2005: Cazadores de ballenas en el Golfo de Cádiz, Algeciras. VERZAR-BASS, M. 2000: “Il praefectus fabrum e il problema dell’edilizia pubblica”. In M. CébéillacGervasoni (dir.) Lés élites municipales de l’Italie Péninsulare de la mort de César à la mort de Domitien entre continuité et rupture. Classes sociales dirigeantes et pouvoir central, Rome 2000, 197224. VIEGAS, C. 2003: “A terra sigillata da Alcaçova de Santarém: cerâmica, economia e comércio”. Trabalhos de Arqueologia. Lisboa, 26.

TOVAR, A. 1975: “La conquista romana”. In Tovar, A.; Blázquez J.M. (eds.) Historia de la Hispania Romana: la Península Ibérica desde 218 AC hasta el siglo V. Madrid 9-121. TRANOY, A. 1981: La galice romaine: recherches sur les NW de la péninsule ibérique dans l’antiquité. Paris. TRANOY, A. 1990: “L’organisation urbaine dans le Conventus Scallabitanus”. In Les Villes de la Lusitanie Romaine, Hiérarchies et territoires. Table ronde internationale du Centre Nacional de Recherche Scientifique (Talence 1988). Paris: CNRS (Coll. Maison des Pays Ibériques; 42), 11-20.

VILLARONGA, L. 1969: “Emisión monetaria augustea con escudo atribuible a P. Carisio y a la zona norte de Hispania”, IX CAN (Mérida, 1968), 591-600.

TRANOY, A. 1995: “La rive nord du douro à l’époque Romaine: contribution épigraphique à l’étude d’une zone frontière”. Gaya, 6, 125-136.

VILLEDIEU, F. 1984: Turris Libisonis. Fouille d’un site romain tardif à Porto Torres, Sardaigne. BAR Int.Ser. 224. Oxford. VIÑAS CUE, R. 2004: Ánfora romana, in Até o confín do mundo: diálogos entre Santiago e o mar, Museo do Mar de Galicia, 2004, Vigo, p. 57-58.

TRILLMICH, W. 1990: “Colonia Augusta Emerita. Die Hauptstadt von Lusitanien”, Stadtbild und Ideologie (Die Monumentalisierung hispanischer Städte zwischen Republik una Kaiserzeit. Kolloquium, Madrid, 1987), München, 299-318.

VITTINGHOFF, F. 1974: “Das problem des ‘militärterritoriums’ in der vorseverischen Kaiserzeit”, I diritti locali nelle province romane con particolare riguardo alle condizioni guiridiche del suolo (1971), Roma, 109-124. VIVAR, G. 2001: “El yacimiento arqueológico submarino de cala Galladura (Port de la Selva)” Revista d’Arqueologia de Ponent 12, 237-246.

UGGERI, G. 1998: “Portolani romani e carte nautiche. Problemi e incognite” Studi di filologia e Leteratura, IV. Porti, approdri e linee di rotta nel Mediterraneo Antico, 31-78. UGGERI, G. 2004: La viabilità della Sicilia in età romana. Journal of Ancient Topography / Rivista di Topografia Antica, Suppl. II, Galantina, 21-25. URIOL, J.L. 1977: “Apuntes para una historia del transporte en España. Los caminos de rueda del siglo XVIII”, en Revista de Obras Públicas, 3143. 145168. URIOL, J.L. 1985: “Las calzadas romanas y los caminos del siglo XVI” en Revista de Obras Públicas, 3237. 553-563.

VIVAR, G. 2005: “Els objectes de la tripulació”. In Carreras, C. et al., 147-153. WACHSMANN, S.; DUNN, R.; HALE, J.; HOHLFELDER, R.; CONYERS, L.; ERNENWEIN, E.; SHEETS, P. BLOT, M.L. “Some Aspects of Phoenician Maritime Involvment in Portugal”.– I.J.N.A. [forthcoming].

URIOL, J.L. 1986: “Viajes y viajeros por España a principios del siglo XVI” en Revista de Obras Públicas, 3251, 805-811.

WAHL, J. 1984: “Ein Ziegelstempel der Legio X Gemina aus dem Alenkastell bei Rosinos de Vidriales (Prov. Zamora)”, MDAI(M) 25, 72-78.

URTEAGA, M. 2003: “El puerto romano de Oiasso (Irún) y la desembocadura del río Bidasoa”, C. Fernández Ochoa (ed.), Gijón, puerto romano. Navegación y comercio en el Cantábrico durante la Antigüedad, Gijón, 192-211.

WASHINGTON, I. 1995: The Alhambra. Barcelona. WELCH, K.E. 1995: “The Office of Praefectus Fabrum in the Late Republic”. Chiron 25, 131-145. WESTERDHAL, C. 1991: “The maritime cultural landscape”. The International Journal of Nautical Archaeology (IJNA). 5-14. London.

URTEAGA, M. 2005: “El puerto romano de Irún (Guipúzcoa)”. In M.M. Urteaga and MªJ. Noain (coord.) Mar exterior. El occidente atlántico en época romana. Actas del Congreso Internacional, Pisa (2003). Roma, 85-106.

WESTERDHAL, C. 1993: “Links between sea and land”. In Coles, J.; Fenwick, V.; Hutchinson, G. (Eds) – A Spirit of Enquiry. Essays for Ted Wright. Exeter. 9195.

URTEAGA, M.; OTERO, X. 2002: Erromatar garaia, Bertan 17, San Sebastián-Donosti.

WESTERDHAL, C. 2000: “From land to sea, from sea to land. On transport zones, borders and human space”. In Litwin, Jerzy (ed.) (2000) – Down the river to the sea. Eighth International Symposium on Boat and Ship Archaeology, Gdánsk, 1997. Gdánsk: Polish Maritime Museum, 11-20.

VAN DOMMELEN, P. 1998: “Punic persistence: colonialism and cultural identities in Roman Sardinia”. In R. Lawrence, J. Berry, eds., Cultural identity in the Roman Empire, Londres, 25-45. 293

THE WESTERN ROMAN ATLANTIC FAÇADE: A STUDY OF ECONOMY AND TRADE IN THE MAR EXTERIOR...

WOODING, J.M. 1996: Communication and Commerce along the Western Sealanes AD 400-800. Oxford: BAR International Series 654.

WHEELER, R.E.M. 1929: “The Roman Lighthouses at Dover”, Archaeological Journal 36, 28-46. WHITTAKER, C.R. 1989: Les frontières de l’Empire romaine. Besançon.

WOOD, M.; QUEIROGA, F. 1992: Current Research on the Romanization of the Western Provinces, BAR International Series 577, 1992.

WIGG, D.G. 1997: “Coin supply and the Roman army”, W. Groenman-van Waateringe, B.L. van Beek, W.J.H. Willems & S.L. Wynia (eds.), Roman Frontier Studies 1995. Proceedings of the XVIth International Congress of Roman Frontier Studies, Oxbow Monograph 91, Oxford, 281-288.

ZABALETA, M.M. 2000: “Hallazgos numismáticos de los comienzon de Bracara Augusta”. In 3º Congresso de Arqueologia Peninsular (“Arqueologia da Antiguidade na Península Ibérica”). Porto. VI, 395-99.

WIGHTMAN, E.M. 1977a: “Soldier and Civilian in Early Roman Gaul”. Limes. Akten des XI Internationalen LimesKongress. Budapest, 75-86.

ZARZALEJOS, M. 2005: “Comercio y distribución de cerámicas romanas en Asturias”. In C. Fernández Ochoa and P. García (eds) (2005) Unidad y diversidad en el Arco atlántico en época romana. BAR Int.Ser. 1371, Oxford.

WIGHTMAN, E.M. 1977b: “Military Arrangements, Native Settlements and Related Developments in Early Roman Gaul”. Helinium 17, 105-126. Wierschowski, L. (1984) Heer und Wirtschaft. Das römische Heer der Prinzipatszeit als Wirtschaftsfaktor. Bonn.

ZBYSZWESKI, G.; FERREIRA, O.V.; SANTOS, C. 1968: “Acerca do Campo Fortificado de ‘Chões’ de Alpompé (Santarém)”. O Arqueólogo Português 2, 49-57.

WILLEMS, W.J.H. 1986: Romans and batavians: a regional study on the Dutch Eastern river area, Amersfoort.

ZUCCA, R. 1996a: La Corsica romana. Oristano. ZUCCA, R. 1996b: “Inscriptiones latinae liberae rei publicae Africae, Sardiniae et Corsicae”. L’Africa Romana XI. Cartago 1994, Ozierei 1996, 14251489.

WITCHER, R. 1998: “Roman roads: phenomenological perspectives on roads in the landscape”. TRAC97. Proceedings of the Seventh Annual Theoretical Roman Archaeology Conference, Oxbow Books, Oxford 1998, 60-70.

ZUCCA, R. 1998: Insulae Baliares. Le issole Baleari sotto il dominio romano, Roma.

294