Life along Communication Routes from the Roman Period to the Middle Ages: Roads and Rivers 2 1803275979, 9781803275970

Table of contents :
Cover
Title Page
Copyright Page
Contents Page
List of contributors
Introduction
Šibenik’s Donje Polje
Toni Brajković, Andrija Nakić and Željko Krnčević
Discovering Roman Roads of the Trilj Area
Domagoj Bužanić, Jelena Bužanić, Blaž Glavinić, Karla Ivak, Marko Jukić, Lucija Prusac, Domagoj Tončinić and Ivan Vidović
A Fort on the Road
Andrej Janeš and Nikolina Vrančić
An Island of Connectedness: Intra-Insular and Inter-Regional Communications of Rab Island (North-East Adriatic, Kvarner Gulf), Croatia
Ana Konestra, Fabian Welc and Paula Androić Gračanin
The Roadside Station at Žuta Lokva
Ivana Ožanić Roguljić
Petrijanec-Aqua Viva
Lovorka Štimac-Dedić
The unknown about the known: the Možđenec-Sudovec Roman road
Jere Drpić
New Data, Old Town: The Case of Aquae Balissae
Andrej Janeš, Marina Matković Vrban and Ivana Hirschler Marić
The Road from Mursa to the North: Segments of the Road in Remote Sensing Data
Mislav Fileš and Miroslav Vuković
On the Road from Ad Herculem to Rittium (Segment Čortanovci-Surduk)
Biljana Lučić, Alessandro De Rosa, Sara Zanni, Gorana Lemajić, Sonja Štefanski-Zorić, Radoslav Muždeka and Uroš Nikolić
Lifeline of the Frontier: The Road and Port Network and the Concept of Supply and Distribution in Moesia Superior
Nemanja Mrđić and Angelina Raičković Savić
Signposts of Settlement: Roman Communications and Avar-Age Sites in Continental Croatia: Overlap Pattern Analysis
Lorena Jurakić and Pia Šmalcelj Novaković
Small Items, Big Results
Anita Rapan Papeša

Citation preview

Life along Communication Routes from the Roman Period to the Middle Ages Roads and Rivers 2 Edited by Ivana Ožanić Roguljić, Jere Drpić, Angelina Raičković Savić

Life along Communication Routes from the Roman Period to the Middle Ages Roads and Rivers 2

Edited by

Ivana Ožanić Roguljić Jere Drpić Angelina Raičković Savić

Archaeopress Archaeology

Archaeopress Publishing Ltd Summertown Pavilion 18-24 Middle Way Summertown Oxford OX2 7LG www.archaeopress.com

ISBN 978-1-80327-597-0 ISBN 978-1-80327-598-7 (e-Pdf) © the individual authors and Archaeopress 2023 Cover: Map of the River network, Roman road network and Roman provinces in the S-I Europe (author: J. Drpić, base maps: WMS_ Esri Shaded Relief, WMS_ MapTiler-topo; http://server.arcgisonline.com/arcgis/rest/services/ World_Shaded_Relief/MapServer/tile/{z}/{y}/{x}; https://api.maptiler.com/tiles/topo/{z}/{x}/{y}.webp?key=mBJY9x5P2uYxCRKXnlsF; vector layers: Roman road network; McCormick et. al. 2013, “Roman Road Network (version 2008)”, https://doi.org/10.7910/DVN/TI0KAU, Harvard Dataverse, V1; Roman provinces: https://github.com/ klokantech/roman-empire; COPERNICUS Land Monitoring Service, 2019: EU-Hydro; https://land.copernicus.eu/ imagery-in-situ/eu-hydro)

All rights reserved. No part of this book may be reproduced, or transmitted, in any form or by any means, electronic, mechanical, photocopying or otherwise, without the prior written permission of the copyright owners. This book is available direct from Archaeopress or from our website www.archaeopress.com

Contents

List of contributors�������������������������������������������������������������������������������������������������������������������������������������������������������������������� ii Introduction���������������������������������������������������������������������������������������������������������������������������������������������������������������������������������v Šibenik’s Donje Polje������������������������������������������������������������������������������������������������������������������������������������������������������������������1 Toni Brajković, Andrija Nakić and Željko Krnčević Discovering Roman Roads of the Trilj Area������������������������������������������������������������������������������������������������������������������������10 Domagoj Bužanić, Jelena Bužanić, Blaž Glavinić, Karla Ivak, Marko Jukić, Lucija Prusac, Domagoj Tončinić and Ivan Vidović A Fort on the Road��������������������������������������������������������������������������������������������������������������������������������������������������������������������21 Andrej Janeš and Nikolina Vrančić An Island of Connectedness: Intra-Insular and Inter-Regional Communications of Rab Island (North-East Adriatic, Kvarner Gulf), Croatia���������������������������������������������������������������������������������������������������������������������29 Ana Konestra, Paula Androić Gračanin and Fabian Welc The Roadside Station at Žuta Lokva�������������������������������������������������������������������������������������������������������������������������������������43 Ivana Ožanić Roguljić Petrijanec-Aqua Viva����������������������������������������������������������������������������������������������������������������������������������������������������������������47 Lovorka Štimac-Dedić The Unknown about the Known: The Možđenec-Sudovec Roman Road��������������������������������������������������������������������55 Jere Drpić New Data, Old Town: The Case of Aquae Balissae���������������������������������������������������������������������������������������������������������������62 Andrej Janeš, Marina Matković Vrban and Ivana Hirschler Marić The Road from Mursa to the North: Segments of the Road in Remote Sensing Data����������������������������������������������69 Mislav Fileš and Miroslav Vuković On the Road from Ad Herculem to Rittium (Segment Čortanovci-Surduk)������������������������������������������������������������������77 Biljana Lučić, Alessandro De Rosa, Sara Zanni, Gorana Lemajić, Sonja Štefanski-Zorić, Radoslav Muždeka and Uroš Nikolić Lifeline of the Frontier: The Road and Port Network and the Concept of Supply and Distribution in Moesia Superior�������������������������������������������������������������������������������������������������������������������������������������������������������������������������108 Nemanja Mrđić and Angelina Raičković Savić Signposts of Settlement: Roman Communications and Avar-Age Sites in Continental Croatia: Overlap Pattern Analysis������������������������������������������������������������������������������������������������������������������������������������������������������117 Lorena Jurakić and Pia Šmalcelj Novaković Small Items, Big Results��������������������������������������������������������������������������������������������������������������������������������������������������������130 Anita Rapan Papeša

i

List of contributors Paula Androić Gračanin [email protected] Cardinal Stefan Wyszynski University in Warsaw Wóycickiego 1/3 (23) 01-938 Warsaw, Poland

Karla Ivak [email protected] Kolodvorska 102b 48246 Marija Bistrica, Croatia Andrej Janeš [email protected] Hrvatski restauratorski zavod Odjel za kopnenu arheologiju Kožarska 5 10000 Zagreb, Croatia

Toni Brajković [email protected] Muzej grada Šibenika Gradska vrata 3 22000 Šibenik, Croatia Domagoj Bužanić [email protected] Odsjek za arheologiju Filozofski fakultet Sveučilišta u Zagrebu Ivana Lučića 3 10000 Zagreb, Croatia

Marko Jukić [email protected] Jarebinjačka 24 22203 Rogoznica, Croatia Lorena Jurakić [email protected] Pirovačka 9 10040 Zagreb, Croatia

Jelena Bužanić [email protected] Muzej seljačkih buna Muzeji Hrvatskog zagorja Samci 64 49245 Gornja Stubica, Croatia

Ana Konestra [email protected] Institut za arheologiju Jurjevska ulica 15 10000 Zagreb, Croatia

Alessandro De Rosa [email protected], Italia

Željko Krnčević [email protected] Muzej grada Šibenika Gradska vrata 3 22000 Šibenik, Croatia

Jere Drpić [email protected] Institut za arheologiju Jurjevska ulica 15 10000 Zagreb, Croatia

Gorana Lemajić [email protected] Institute for the Protection of Cultural Monuments Sveti Dimitrija 10, Sremska Mitrovica, Serbia

Mislav Fileš [email protected] Institut za arheologiju Jurjevska ulica 15 10000 Zagreb, Croatia

Biljana Lučić [email protected] Institute for the Protection of Cultural Monuments Sveti Dimitrija 10, Sremska Mitrovica, Serbia

Blaž Glavinić [email protected] Vrapče Donje 43 10090 Zagreb, Croatia

Marina Matković Vrban [email protected] Ulica Hrvatskog proljeća 47 Ribnica, Velika Gorica, Croatia

Ivana Hirschler Marić [email protected] Hrvatski restauratorski zavod Odjel za kopnenu arheologiju Kožarska 5 10000 Zagreb, Croatia

Nemanja Mrđić [email protected] Institute of Archaeology Kneza Mihaila 35/IV 11000 Belgrade, Serbia ii

Radoslav Muždeka [email protected] Serbia

Sonja Štefanski Zorić [email protected] Serbia

Andrija Nakić [email protected] Javna ustanova u kulturi tvrđava kulture Šibenik Vladimira Nazora 1 22000 Šibenik, Croatia

Lovorka Štimac-Dedić [email protected] Gradski muzej Varaždin Šetalište J. J. Strossmayera 3 40000 Varaždin, Croatia

Uroš Nikolić [email protected] Ruma County Museum Glavna 192 22400 Ruma, Serbia

Domagoj Tončinić [email protected] Odsjek za arheologiju Filozofski fakultet Sveučilišta u Zagrebu Ivana Lučića 3 10000 Zagreb, Croatia

Ivana Ožanić Roguljić [email protected] Institut za arheologiju Jurjevska ulica 15 10000 Zagreb, Croatia

Ivan Vidović [email protected] Side Košutić 10 10090 Zagreb, Croatia

Lucija Prusac [email protected] Stjepana Gradića 3 10010 Zagreb, Croatia

Nikolina Vrančić [email protected] Orehovečki brijeg 73a 10000 Zagreb, Croatia

Angelina Raičković Savić [email protected] Institute of Archaeology Kneza Mihaila 35/IV 11000 Belgrade, Serbia

Miroslav Vuković [email protected] Odsjek za arheologiju Filozofski fakultet Sveučilišta u Zagrebu Ivana Lučića 3 10000 Zagreb, Croatia

Anita Rapan Papeša [email protected] Gradski muzej Vinkovci Trg bana Josipa Šokčevića 16 32100 Vinkovci, Croatia

Fabian Welc [email protected] Cardinal Stefan Wyszynski University in Warsaw Wóycickiego 1/3 (23) 01-938 Warsaw Poland

Pia Šmalcelj Novaković [email protected] Institut za arheologiju Jurjevska ulica 15 10000 Zagreb, Croatia

Sara Zanni [email protected] University of Pisa Italia

iii

iv

Introduction The research presented in the book was carried out within and was partly financed by the project of the Croatian Science Foundation, LRR - Life on the Roman road: communications, trade and identities on Roman roads in Croatia from 1st – 8th CE (UIP-05-2017-9768). The project aimed to explore various aspects of Roman life brought by the construction of the roads. It is often said that roads in Roman times were the arteries of the Empire itself. They connected provinces and cities; without them, the Romans could not have conquered and held onto the territories they ruled. They were means of moving military officials and civilians and transporting goods. Roman roads gave excellent links for organizing life and establishing various settlements in periods after the collapse of the Empire. The engineering and surveying skills of the Romans provided the basis for many of today’s routes. Life on some Roman roads continued after the fall of the Empire and has not stopped until today. The project’s main objective is an interpretation of everyday life on the Roman road from the beginning of Roman rule to the period of decline, acculturation and integration of new identities. This project tried to explore a chance for a fresh perspective: how does life develop around Roman roads and a path during the primary usage of the roads and after it’s over? With an interdisciplinary approach to archaeological research and analysis of small archaeological materials, the project tried to present how life on the Roman road was organized. Activities on the project utilize landscape surveys and studies of material assemblages to track the development and decline of the Roman way of life. The broad perspective of the project by the end of the research offers a starting point/foundation for a significant number of future comparisons, analyses and projects. During the Life on the Roman Road project, Roman roads are studied as the essential artery of the Roman world but also as a metaphor for introducing, accepting and living the Roman way of life. After the decline of the Roman world, Roman roads represent the metaphors of heritage and tradition of what was once a vast Empire. The publication in front of us consists of thirteen papers dedicated to the regions of Roman provinces Dalmatia, Pannonia and Upper Moesia. Contributors to the publication gave their brand new data about research on selected study areas. The first five articles are dedicated to the Roman province of Dalmatia.  Šibenik’s Donje Polje  (review article) by Toni Brajković, Andrija Nakić and Željko Krnčević and  Discovering Roman Roads of the Trilj Area (preliminary communication) by a group of authors Domagoj Bužanić, Jelena Bužanić, Blaž Glavinić, Karla Ivak, Marko Jukić, Lucija Prusac, Domagoj Tončinić and Ivan Vidović present a new data collected by a field survey of road sections. A Fort on the Road (review article) by Andrej Janeš and Nikolina Vrančić brings us a report about the fort situated on the coast in present-day Novi Vinodolski; fort Lopar represents a rare example of an early 4th-century AD fortification in the north Adriatic. Ana Konestra, Paula Androić Gračanin and Fabian Welc present (original scientific paper) An Island of Connectedness. Intra-Insular and Inter-Regional Communications of Rab Island (North-East Adriatic, Kvarner gulf) bring us essential insight into insular communications. The Roadside station at Žuta Lokva (preliminary communication) by the editor of the publication and PI of the project, Ivana Ožanić Roguljić, presents the analogies of the only fully excavated roadside station in Croatia. The following five articles are dedicated to Roman Pannonia. Petrijanec - Aqua Viva (preliminary communication) by Lovorka Štimac Dedić provides information about this important Roman settlement. A brand new approach to studying roads that in ethnological records are called “Roman Roads” is seen in the article The Unknown about the Known; The Možđenec – Sudovec Roman Road (preliminary communication) by Jere Drpić. In the article by Andrej Janeš, Marina Matković Vrban and Ivana Hirschler Marić New Data, Old Town: The Case of Aquae Balissae (preliminary communication), we see an excellent introduction to the problems regarding the research of urban archaeology. The road from Mursa to the north - segments of the road on remote sensing data (preliminary communication), by Mislav Fileš and Miroslav Vuković uses remote sensing methodology to enhance the knowledge about known parts of the road. The results of the research performed in the Lower Pannonian part of the Danube Limes are presented in the article On the Road from Ad Herculem to Rittium (original scientific paper) by a group of authors Biljana Lučić, Alessandro De Rosa, Sara Zanni, Gorana Lamajić, Sonja Štefanski-Zorić, Radoslav Muždeka and Uroš Nikolić. Upper Moesia is presented by Nemanja Mrđić and Angelina Raičković Savić with the article Lifeline of the Frontier: The Road and Port Network and the Concept of Supply and Distribution in Moesia Superior (original scientific paper). v

This project also tried to explore a chance for a new perspective: how does life develop around Roman roads during the primary usage of the roads and after this phase is over? Part of this research question is answered in the articles Signposts of Settlement: Roman Communications and Avar-Age Sites in Continental Croatia: Overlap Pattern Analysis (original scientific paper) by Lorena Jurakić and Pia Šmalcelj Novaković and Small Items, Big Results (professional paper) by Anita Rapan Papeša. This publication has gone “public” and each article has been “blind” reviewed. So the editors would like to express their gratitude to all the reviewers (Martin Auer, Vladimir Petrović and Iva Kaić) and the scientific board (Maja Petrinec, Bartul Šiljeg and Pia Šmalcelj Novaković). Without their help, we would not have been able to complete the publication. Then we would like to thank all the authors, as they accepted the challenges we set as project goals and helped us meet them. Special thanks go to Pia Šmalcelj Novaković and Mislav Fileš, who took the time to help us proofread and edit the publication. Editors Ivana Ožanić Roguljić Jere Drpić Angelina Raičković Savić

vi

Šibenik’s Donje Polje Toni Brajković, Andrija Nakić and Željko Krnčević Abstract: Šibenik’s Donje polje is an extremely rich archaeological area, and we can say, the area with which life, in part, spilt over into Medieval Šibenik. Unfortunately, the archaeological research conducted so far within this area did not follow its potential and significance. Despite this, a whole series of Prehistoric, Roman and Medieval sites have been identified, along with a network of communications that connect them. In addition to the already mentioned in the literature, the routes from Grebaštica (Ad Pretorum), Šibenik’s Dubrava, the branch of the Salona - Scardona road, and from Gornje / Danilsko polje (Municipium Riditarum), in this paper, we also present some newly established routes of communications and the question of their importance and purpose, especially the connection of Donje polje with the sea, a probable port in Mandalina and a possible pier in Morinje. Among the more significant sites, the Roman sites of Grušine and the area of Jadrtovac stand out, which by all accounts could have been the centres of some kind of rural agglomerations. Keywords: ŠIBENIK’S DONJE POLJE, ŠIBENIK’S GORNJE (DANILSKO) POLJE, GRUŠINE, ST. LAWRENCE, MANDALINA, MORINJE, DANILO (MUNICIPIUM RIDITARUM), ROADS, PORTS

Šibenik’s Donje polje1 already in its name evokes the belonging and significance of the arable piece of land that starts northeast of the city and extends in a southeast direction to Vrpolje-its top or end. With the development of Šibenik and administrative divisions, the northwestern area of Donje polje narrowed and changed its character from agrarian to urban. In this paper, the term “Donje polje” includes its originalnatural scope, which extends all the way to the Channel of St. Anthony - the submerged estuary of the Krka River. Due to the area of about 1050 hectares of extremely fertile and arable land with an abundance of occasional and permanent water sources, Donje polje was, for centuries, along with salt production and trade, the backbone of Šibenik’s economy. In the Venetian books, it is mentioned under the Italian name “Campo di sotto”opposite the Danilo or Gornje polje,2 which is “Campo di sopra”.3 Both fields, owned by the Šibenik’s nobles, the Church and landowners, were extensively cultivated, and their crops filled the budget of their masters. Climatological and agrarian potentials have created conditions suitable for life inside the field, especially on its edges bordered by mountain ranges no higher than 250 m. During the late Neolithic, the first settlement appeared in the middle of the Danilo’s-upper field, the material remains of which gave the name to the eponymous Danilo culture. A little later, in the metal age, hilltops that flank these fertile plains sprouted hillfort settlements. There are six on the northern side Donje polje literally means “lower field”. 2  Gornje polje literally means “upper field”. Gornje (upper) polje today is called Danilsko polje. 3  The earliest undoubted mention of Donji polje can be found in Juraj Šižgorić’s work: On the settlement of Illyria and the city of Šibenik (Šižgorić 1981: 38-39). This book is a translation and facsimile of the manuscript: De situ Illiyriae et civitate Sibenici. 1 

of Donje polje (Vrpolje, Renje, Popelj, Đelalije, Ljubljan and Umina) on a stretch of eight km. Of course, not all “hillforts” existed at the same time. Their purpose of either refugium or habitat, as well as the exact period of use, has not been fully elucidated,4 but the undoubted number during the Prehistory points to the extensive exploitation of the fields, as well as to the inexorability of the traffic route that connected them. Zlatko Gunjača presented the most comprehensive synthesis of previously available data on Donje polje, supplemented by new archaeological knowledge with a valid interpretation, in the commemorative collection on the 900th anniversary of the first mention of Šibenik. In it, compiling written and other sources in one place, he collects random finds in addition to those from archaeological research and field surveys, architecture and burial sites (necropolises, cemeteries, tumuli) following the chronological sequence from the Prehistory to the Middle Ages. An archaeological framework of Donje polje has been established, within which all relevant sites are indicated in relation to the landscape, the surrounding area, resources and communications. Unfortunately, longer-term research within Donje polje was mostly not carried out. An exception is an area around the church of St. Lawrence. Back in the 15th century, the ruins of a larger Roman building were so imposing that Juraj Šižgorić attributes them to remains of the “Noric city.” This rustic villa in 4  Smaller archaeological researches were carried out only at the hillforts of Renje and Ljubljan. In 1955, Josip Korošec opened a test trench at Renje hillfort. Two small tranches were opened on the same hillfort in the mid-1970s. The eastern entrance of the ramparts was excavated in the Ljubljan hillfort. The prehistoric material observed on them, as well as the type of settlements on the heights (hillfort settlements), can belong to all periods of the metal age: copper, bronze and iron.

Life Along Communication Routes (Archaeopress 2023): 1–9

Toni Brajković, Andrija Nakić and Željko Krnčević

Figure 1: Map of Donje polje and its surroundings with prehistoric and ancient sites (author: T. Brajković)

Figure 2: Part of Donje polje, view from the northwest (source: Šibenik City Museum archive)

2

Šibenik’s Donje Polje

Figure 3: Part of Donje polje, view from the southeast (source: Šibenik City Museum archive)

the location of Grušine was partially explored during the campaigns of 1935-1938 and 1977.5 The discovered remains of the hypocaust, mosaic floors, stucco, painted plaster and different building styles speak of its use from the 1st-2nd until 6th century AD. That Grušine is not a modest and isolated building, but a more monumental complex, or several residential and commercial buildings, is indicated by the remains of Roman architecture north of the mentioned location, and around a hundred meters remote St. Lawrence where the walls are located in direct correlation with Grušine (Krnčević 2005: 211-212; Laszlo Klemar, Zeman 2010: 152). The surface remains of tegulae and Roman pottery on stone piles - clearings near Grušine, suggest

the probable existence of more contemporaneous buildings. An identical situation occurs in several locations within Donje polje. Thus, at the positions of Gomilice, St. George, Podgoričje, Zaton jadrtovački and Vruci-Mostir, along with fragments of tiles and pottery of Roman provenance, we find stone carvings, which, according to some local stories, were shaped during the cultivation of the land, and graves found near Ribnik, in Jadrtovac, Rujave, Podgoričje and Gomilice support the thesis about the continuity of settlement during antiquity (Stošić 1941: 46; Dujmović 1953: 241-242; Gunjača 1976: 43-46; Miletić 1991: 73; Laszlo Klemar, Zeman 2010: 150-151). As an additional confirmation of the population density and the possible existence of pagus, we point out the construction of probably two early Christian churches at a relatively short distance. The first one, based on the established material is found on the location of Vruci-Mostir, with an approximate dating to the 5th6th century (Dujmović 1953: 242; Gunjača 1976: 4344; Miletić 1991: 73). The second one may have been on the site of today’s St. Lawrence. This location was investigated on several occasions,6 but the most comprehensive and longest were those conducted from 1995 to 2005 (Krnčević 1995a: 52-55; Krnčević

Although in the years before the Second World War, the leader of the research was nominally Franciscan priest Lujo Marun, the works were managed by Frane Dujmović and Ivan Ostojić. From that time, the most extensive research report has been preserved-as Z. Gunjača writes, “The professorial work of Ivan Ostojić kept in the library of the Šibenik City Museum” (Gunjača 1978: 75-76, note. 70). Grušine is always mentioned in the context of the site of St. Lawrence, which is in the immediate vicinity, even sometimes the micro-location of Grušine includes the position where the medieval church of St. Lawrence is situated (Dujmović 1953: 242; Gunjača 1976: 47, 49; Gunjača 1977: 22-26). For details about Grušine, see: (Gunjača 1976: 41-43, 45-47; Krnčević 1994: 98; Krnčević 1995: 20-21; Laszlo Klemar, Zeman 2010: 150-152, 157; Stošić 1941: 49). Grušine is also mentioned by M. Zaninović in a short passage (Zaninović 1967: 367-368) but with the false claim that they were investigated by K. Stošić and the information that the site is located in a bay, although it is more than a km away from it.

5 

6 

3

With interruptions from 1935 to 1938, then 1953 and 1977.

Toni Brajković, Andrija Nakić and Željko Krnčević

Figure 4: St. Lawrence and nearby Grušine (source: Šibenik City Museum archive)

1997: 70-72; Krnčević 1999: 83-86; Krnčević 2000: 122125; Krnčević 2005: 211-212; Krnčević 2006: 339). Over 40 decorated stone fragments of church furniture have been found so far. Titulars, starting with St. Florian, through St. Michael to Ss. Cosmas, Damian and Lawrence, and St. Bartholomew, as well as the stylistic features of the lapide, speak of several sacral buildings or one that underwent numerous changes from the 9th to the 11th century (Jakšić, Krnčević 1997: 93-109; Krnčević 2008: 113-118, 229, 241-244; Laszlo Klemar, Zeman 2010: 156-157; Uglešić 2018: 53) and which is probably the successor of an even older late antique church built on this site (Gunjača 1976: 44; Gunjača 1978: 76; Laszlo Klemar, Zeman 2010: 152-153, Krnčević 2008: 99; Uglešić 2018: 53).

right in the bay of St. Peter, Luka Jelić described and sketched the remains of the Badnjina site. This object, defined as a water tank (13 x 5.60 m), was located along the coast, somewhere in the middle of the complex. At the beginning of the 20th century, the dimensions of this complex of 58.40 m were determined. We will never be able to know the full scope and purpose of these buildings, which, according to Jelić, form one unit with the architecture found inside the local cemetery of St. Peter’s, (Stošić 1941: 33; Gunjača 1976: 40; Krnčević 1995: 19). The question is whether Jelić was able to see everything and what the degree of devastation was before his arrival. Unfortunately, the entire area was archaeologically destroyed almost half a century ago by the construction of the Mornaričko tehnički remontni zavod Velimir Škorpik. However, the assumptions put forward by Frano Dujmović and Zlatko Gunjača that on the site of this complex, we can see a pier to which the population of Donje polje gravitated, seem to be justified for several reasons (Dujmović 1962: 1441; Gunjača 1976: 40-41). Firstly, it is the previously established area of this architectural complex, and its position next to the coastline itself suggests an orientation towards maritime activities. Location in a protected cove at the very end of a fertile field follows the road line that stretches from Vrpolje towards Mandalina (Stošić 1941: 57; Miletić 1991: 73 note 35). Furthermore, the

The northwestern edge of the field extends to the sea and ends at Mandalina.7 This narrow peninsula of modest dimensions divides the arm of the submerged Krka estuary into two bays; St. Peter and Vrnaža.8 In 1912,9 See more about Mandalina in Gunjača 1976: 37, 39-40, 44-45, 53; Krnčević, Krnčević 1994: 101; Krnčević 1995: 19 and Laszlo Klemar, Zeman 2010: 152-153. 8  Until recently, Vrnaža was called Furnaža in lat. Furnum-furnace (Stošić 1941: 34). Therefore, both names refer to the same bay. 9  Krsto Stošić states the year 1914 (Stošić 1941: 33), although Luka Jelić’s visit to the Badnjina site took place in 1912, which is evident from the documentation of the Archaeological Museum in Split (folder no. 35). 7 

4

Šibenik’s Donje Polje

Figure 5: Map of the road through Šibenik’s Donje polje (author: A. Nakić)

direction of Šibenik’s Dubrava and branched off with one branch towards Gornje and the other towards Donje polje between the hillforts of Ljubljan and Popelj, or it followed the relatively flat terrain along the left bank of Krka River towards the Šibenik hillfort and further towards Mandalina and Donje polje. From Donje polje, in addition to the direction from Šibenik’s Dubrava, there is also another communication towards Gornje polje (Danilo)-between the hillforts of Popelj and St. John, across Crno Brdo, not far from Stražbenica cave (Miletić 1991: 73-79). This route goes not far from the course of the Dabar stream, which until the second half of the 20th century, used to flood the area, even with torrential force. Of course, water is directly correlated with fertile soil, and in the area of Donje polje, we find dozens of ponds and wells and one inexhaustible spring called Ribnik. The assumption that Grušine could have been the nucleus of the pagus in Donje polje seems completely justified for several reasons. This is indicated by the volume of the partially established architectural complex. At least one part of it, with a mosaic floor and plaster–painting in different colours, speaks of the property status of the owners at that time. It was, we assume, based on agricultural products and their trade, perhaps even export. We have seen that one possible port was in the area of Mandalina, but we must not ignore the possible pier in Morinje either. It would certainly have a local character because the port in Mandalina is located on the main road - Krka River,

dating of Badnjina to the 1st century AD fits into the time frame of the most intensive trade through Krka River as a result of the establishment and necessity of the military camp in Burnum. In addition to Badnjina, the geostrategic position of the Mandalina peninsula includes several other positions with remains of Roman architecture and graves as evidence of uninterrupted life at least from the time of the Principate to the present day (Gunjača 1976: 44-45; Krnčević 1995: 19). Communication from Vrpolje, through Donje polje, towards the Mandalina peninsula was just one in a series of routes. These probably prehistoric routes used incessantly even in the Roman times are best elaborated by Ž. Miletić, citing several routes that connect Šibenik’s Donje polje with Danilo (Municipium Riditarum) in the northeast, Grebaštica (Ad Pretorum) in the south, and the Liburnian area with large centres - Scardona and Iader to the northwest across the Krka River. It was Krka, as the natural border of the Liburnian and Delmataean territories, that was the main obstacle in the land connection of the eastern Adriatic coast. In addition to the crossing near Skradin, Antonine’s Itinerary and material remaining (passages in the cliffs, ruts) suggest two more crossings, a little further southat the mouth of Guduča on Lake Prukljan and near the Tradanj cave close to Šibenik’s Zaton. From there, the communication flowed either through the valleys from Stubalj on the southeast side of Lake Prukljan in the 5

Toni Brajković, Andrija Nakić and Željko Krnčević road from Dubrava went down over the Bikarac hill, not far from today’s road, and exited at Grušine. Today this direction can be followed for a length of about 400 m (Figures 6 and 7). A small trench was made on one part of the established section. On that occasion, a limestone substrate was discovered with visible traces of the longterm use of this route as a convenient way to easily overcome the height difference between the fields and the surrounding hills. Traces on the bedrock indicate chiselling, while the rest of the surface is smooth. The width of the road is about 180 cm.

which carries all the major traffic towards Skradin and the interior. The Morinje channel, on the other hand, provides good protection and is the shortest route for products from Šibenik’s Dubrava and Danilo, i.e., Gornje polje. Furthermore, on both sides of Morinje Bay, there are remains from the Roman period. Of particular note is the probable early Christian sacral object at the location of Vruci, perhaps built on the site of an even earlier cult site, as suggested by a discovered Roman altar. The main prerequisite for the development of a larger rural agglomeration was, in addition to the fertile field, the only non-drying source of water, Ribnik. The population of Morinje certainly used the benefits of both the sea and the fields. Two piers, in Morinje and Mandalina certainly existed. The question is whether they existed at the same time or whether the pier in Morinje was of earlier origin and the main connection of the surrounding area with the sea before the construction of the port in Mandalina in the 1st century AD. That the communication between Dubrava - the coast and Danilo (Municipium Riditarum) - the coast, was in use, can be seen from the diverse material from Prehistoric to Modern Times, as well as from the remains of shells discovered during archaeological research in various cultural layers. Another, so far undetermined

There was probably a necropolis along the road, from which nine tombstones originated. All but one were found during archaeological research in 1935/36. The most famous and the only one published so far is the one dedicated to Recius Eumolpus, dated to the 2nd half of the 1st and the first half of the 2nd century AD (Stošić 1941: 50-51; Gunjača 1976: 42, T. XXVI, 2; Année épigraphique 1976: 00524; Jakšić, Krnčević 1997: 96; Epigraphische Datenbank Heidelberg: 012300, Epigraphische Datenbank Clauss–Slaby: 09300467). On the back side of the monument is carved the inscription Bartolo, which is related to the earlier titular of the church, the alleged Benedictine abbey, the royal endowment and the seat

Figure 6: The route of the road located near Grušine that led to the Gornji (Danilo) field (photo: A. Nakić)

6

Šibenik’s Donje Polje

Figure 7: The route of the road located near Grušine that led to Gornji (Danilsko) polje (photo: A. Nakić)

of the alleged Morinje diocese, which the people of Šibenik appealed to in their efforts to obtain their own episcopal jurisdiction (Stošić 1941: 48; Dujmović 1953: 242; Ostojić 1964: 257-259; Krnčević 1995: 15; Krnčević 2001: 33-34; Krnčević 2008: 115; Laszlo Klemar, Zeman 2010: 157-158; Uglešić 2018: 46). This tombstone was discovered by the peasants during the arrangement of the new cemetery, which prompted Franciscan priest Lujo Marun to start research on St. Lawrence and Grušine sites. Other tombstones, which we will not deal with here because they are the subject of another paper, indicate either the same time, even earlier, or laterthe early Christian time of its creation. This range of centuries leads to thinking about the possible existence of a larger and longer-lasting necropolis where the population of this region was buried. This will be clarified by some future research, but for now, there are enough credible indicators and indications based on which we can rightly assume that Grušine and Morinje were kind of centres in the Roman period and that the entire Donje polje was inhabited and agriculturally exploited during that period.

Conclusion Šibenik’s Donje polje is the largest arable area in the surroundings of Šibenik. It has been inhabited since Prehistoric Times, as evidenced by a series of hillfort settlements and tumuli on the peaks that separate it from Danilsko or Gornje polje. Fertile land and plenty of ponds, occasional and one permanent source of water, Ribnik, were the basic prerequisites of life. Direct contact with the sea and the river Krka, an intensive traffic route from the earliest times, had repercussions on trade and the connection of this part of Dalmatia and with the surrounding area. Most of the established Prehistoric communications probably continued to be used during the Roman Period, but new ones are certainly being built. The pacification of Illyricum by the Roman conquest was done by military force, which was based on the establishment of a good road network. One of the main routes from Salona to Scardona and on to Iadera went not far from Donje polje and connected it with several road branches. One branch stretched from Vrpolje, along Grušine to Mandalina, where, in the bay of St. Peter, was most likely a port. The second pier was probably in Morinje. Along this road, larger or smaller 7

Toni Brajković, Andrija Nakić and Željko Krnčević rural agglomerations of Podgoričje, Morinje, Grušine, Gomilice, and St. George sites were developed. Other road routes went over the mountain range that separates Donje from Gornje (Danilsko) polje and Šibenik’s Dubrava. One of them passed by the site of Grušine. This was established during a recent archaeological field survey. Grušine, considering everything that has been said, was one of the most significant rural agglomerations of Donje polje, probably the centre pagus to which the surrounding vicus gravitated. This is indicated by Šižgorić’s testimonies, traffic routes and the results of research so far-a large Roman complex with an associated necropolis. Indirectly, the importance of this micro-location during the Principate is also indicated by the remains of one or more sacred buildings that were located in the immediate vicinity at the beginning of the Medieval period. They were undoubtedly erected in a place of certain importance. In later times, this tradition took shape in the legend about the royal courtsBartulovština, the core of which we know to be true by finding the title of the church of St. Bartholomew. Only future archaeological research will remove all doubts and show the true size of the complex on Grušine, and its development from a probable rural-economic estate to the centre of the secular and spiritual life of Donje polje, which from earlier antiquity spilt over into later periods.

Krnčević, K. and Ž. Krnčević 1994. Bibliografija arheoloških radova šibenskog područja. Šibenik. Krnčević, Ž. 1995. Historijat arheoloških istraživanja na šibenskom području, in Stoljeće arheologije na šibenskom području (katalog izložbe): 9-56. Šibenik: Muzej grada Šibenika. Krnčević, Ž. 1995a. Sveti Lovre-šibensko Donje Polje, sustavna arheološka istraživanja. Obavijesti Hrvatskog arheološkog društva 27, br. 3: 52-55. Krnčević, Ž. 1997. Sveti Lovre-šibensko Donje Polje, sustavna arheološka istraživanja u godini 1996. Obavijesti Hrvatskog arheološkog društva 29, br. 2: 7072. Krnčević, Ž. 1999. Rezultati istraživanja srednjovjekovnih arheoloških lokaliteta na šibenskom području u godini 1997. i 1998. Obavijesti Hrvatskog arheološkog društva 31, br. 2: 83-86. Krnčević, Ž. 2000. Sustavna arheološka istraživanja na lokalitetu Sv. Lovre u šibenskom Donjem polju – rezultati kampanja 1999. i 2000. Obavijesti Hrvatskog arheološkog društva 32, br. 3: 122-125. Krnčević, Ž. 2001. Novija arheološka istraživanja srednjovjekovnih arheoloških lokaliteta šibenskog kraja, in Vilijam Lakić (ed.) Sedam stoljeća Šibenske biskupije. Zbornik radova sa znanstvenog skupa Šibenska biskupija od 1298. do 1998.: 25-42. Šibenik : Gradska knjižnica “Juraj Šižgorić”. Krnčević, Ž. 2005. Sveti Lovre-šibensko Donje polje. Hrvatski arheološki godišnjak 1/2004: 211-212. Krnčević, Ž. 2006. Sveti Lovre-šibensko Donje polje. Hrvatski arheološki godišnjak 2/2005, 339. Krnčević, Ž. 2008. Ranokršćansko i starohrvatsko razdoblje, in Ž. Krnčević, E. Podrug (eds) Arheološki tragovi kultova i religija na šibenskom području (katalog izložbe): 91-127, 138-142, 229-244. Šibenik: Muzej Grada Šibenika. Laszlo Klemar, K. and Zeman M. 2010. Naselja i organizacija prostora na srednjem istočnom Jadranu od antike do ranoga srednjeg vijekapitanje kontinuiteta (Bribir, Donje šibensko polje), in T. Šeparović (ed.) Zbornik radova sa Znanstvenog skupa „Stjepan Gunjača i hrvatska srednjovjekovna povijesno-arheološka baština“ u povodu 100. obljetnice rođenja akademika Stjepana Gunjače, Split 3.-6. studenog 2009.: 147-165. Split : Muzej hrvatskih arheoloških spomenika. Miletić, Ž. 1991. Rimska cestovna mreža između Arauzone i Tragurija, Radovi Filozofskog Fakulteta u Zadru. Razdio povijesnih znanosti 31, 1991./1992: 6388. Ostojić, I. 1964. Benediktinci u Hrvatskoj i ostalim našim krajevima: benediktinci u Dalmaciji. Split: Benediktinski priorat. Stošić, K. 1941. Sela šibenskog kotare. Šibenik: Tiskara ˝Kačić. Šižgorić, J. 1981. O smještaju Ilirije i o gradu Šibeniku. Šibenik: Muzej grada Šibenika.

Website Epigraphische Datenbank Heidelberg, Heidelberger Akademie der Wissenschaften, viewed 13 May 2022,

Epigraphische Datenbank Clauss-Slaby, Manfred Clauss, viewed 13 May 2022, Bibliography Dujmović, F. 1956. Arheološka rekognosciranja južno od Šibenika. Vjesnik za arheologiju i historiju dalmatinsku 55 (1953): 241-244. Dujmović, F. 1962. O urbanističkom razvoju šibenske luke, Pomorski zbornik II: 1439-1452. Gunjača, Z. 1976. O kontinuitetu naseljavanja na području Šibenika i najuže okolice, in S. Grubišić (ed.) Šibenik-Spomen zbornik o 900. obljetnici, Šibenik: 27-58. Muzej grada Šibenika: NIP Štampa. Gunjača, Z. 1977. Early medieval archaeological finds from Šibenik and its nearest surroundings. Balcanoslavica 5 (1976): 19-30. Gunjača, Z. 1978. Rezultati neobjavljenih i najnovijih arheoloških istraživanja antičkih i srednjovjekovnih lokaliteta na šibenskom području. Izdanja Hrvatskog arheološkog društva 3 (1978): 69-82. Jakšić, N. and Ž. Krnčević 1997. Predromanički reljefi iz crkvice Sv. Lovre u Morinju kraj Šibenika. Starohrvatska prosvjeta III, 24: 91-109. 8

Šibenik’s Donje Polje Uglešić, A. 2018. Najstariji kršćanski tragovi na prostoru Šibenika, in I. Kurelac (ed.) Šibenik od prvog spomena, Zbornik radova s međunarodnog znanstvenog skupa. 950 godina od prvog spomena Šibenika. Šibenik, 26. do 28. rujna 2016: 43-59. Šibenik-Zagreb: Muzej grada Šibenika, Hrvatska akademija znanosti i umjetnosti.

Zaninović, M. 1967. Neki primjeri smještaja antičkih gospodarskih zgrada u obalno - otočkom području Dalmacije. Arheološki radovi i rasprave 4-5: 357-371.

9

Discovering Roman Roads of the Trilj Area Domagoj Bužanić, Jelena Bužanić, Blaž Glavinić, Karla Ivak, Marko Jukić, Lucija Prusac, Domagoj Tončinić and Ivan Vidović Abstract: It has often been suggested that the prosperity of the Roman civilization was greatly impacted by its complex network of roads. This seems likely if we consider the sheer amount of work needed to construct miles and miles of roads in areas with difficult terrain. One such terrain was the Dalmatian karst. After claiming Dalmatia, the Romans were in a position to establish a new land communication route from Aquileia to Dyrrachium. Of course, many other local roads branched from it. North of the provincial centre, Salona, the road to Dyrrachium had to cross the river Cetina, a natural barrier. In Roman times, the crossing itself was controlled by the military camp Tilurium, which was located on a nearby hill. However, at the moment, differentiating between Roman and later roads proved problematic. To help tackle this, smaller field surveys were conducted in the Trilj area from 2017 to 2021. The research aimed to identify surviving road sections. A larger study of Roman roads and pathways in the wider area around Salona would greatly improve our understanding of life in ancient times and also the placement of local archaeological sites. This paper aims to be a step in that direction. KEYWORDS: ROMAN ROADS, TILURIUM, DALMATIA, FIELD SURVEY

Introduction This paper aims to familiarize the reader with the current state of research in surveying and mapping historic roads in the area that is today mostly a part of Trilj municipality (Croatia). The modern-day town is situated in an area that historically had strategic importance. This importance is mostly due to its proximity to a suitable crossing of the river Cetina, often used for traffic going from the major coastal cities of northern and central Dalmatia to the interior of the region (which is a part of Bosnia and Herzegovina today) and southern Dalmatia. Considering the fact that a river crossing, by its nature, presents a point of connection of roads that exist on both sides of the river, the area proved suitable for conducting this type of research. Interest in local historic roads was initially stimulated by the desire to determine the probable routes of Roman roads that connected the legionary camp with the surrounding area, and, finally, with larger cities in the province of Dalmatia, such as nearby Salona, the provincial capital. The importance of this area in Antiquity is indicated by the location of the Roman legionary camp Tilurium, which controlled the crossing of the river and the surrounding area from a hill. The strategic importance of the position of the Tilurium legionary camp at the crossing of the river Cetina has been known for many years (Wilkes 1969: 97; Zaninović 1996: 280-292) and the site has been a topic of archaeological research and excavation since 1996. The results of the research have been published continuously since 1998 and unquestionably prove the existence of a legionary camp at that place (Sanader 1998; Sanader 2001; Sanader 2003; Šimić-Kanaet 2010;

Sanader et al. 2014; Sanader et al. 2017; Sanader et al. 2021). The only other similar legionary fortress this close to Dalmatian towns was Burnum on the river Krka. Since the crossing remained of strategic importance in later times, there are, however, many remains of roads and paths from different historical periods in the general area of ​​Trilj. The sheer number of remains was what prompted the launch of a series of projects which have been primarily focused on mapping as many historic roads as possible using field surveys (since some of the roads were known only to the local population and some were completely forgotten), hoping it will help with locating and dating individual routes during future research of road infrastructure in this area. Research The number of remains and general difficulties in dating preindustrial roads strictly via field surveys excludes the possibility of focusing on a particular historical period and requires that finds of all periods be taken into account. In other words, for research to be successful in the future, firstly it was necessary to make a base map of all known routes. Without insight into all of the remains, even modern roads, it is not possible to distinguish and single out certain historic routes. In addition, the difficult karst terrain does not allow or rather significantly complicates the construction of alternate routes, so it is not uncommon for roads of different periods to follow the same patterns, some even to this day. However, it should immediately be emphasized that modern construction can nullify the influence of terrain, due to the availability of heavy

Life Along Communication Routes (Archaeopress 2023): 10–20

Discovering Roman Roads of the Trilj Area

Figure 1: First research area (source: Google Earth, author: D. Bužanić)

machinery. This further clouds our perception of the relationship between roads and the terrain they run through.

sides of the river Cetina. The area on the eastern side of the river included the northern parts of Trilj: villages Vedrine, Jabuka, and Velić. On the opposite side of the river (western), surveys were carried out on the part of the riverbank, i.e., in the area between the riverbank and the modern Trilj - Bisko road, at Gardun (site of the legionary fortress), both are parts of Trilj, and Vojnić Sinjski, within which lies the majority of the nearby karst field. GPS Waypoint Navigator mobile application was used to georeference points and routes during field surveys. That way, the points, and routes could easily be mapped on satellite images and connected with corresponding field photographs. The base map of roads is supplemented with newly discovered routes every year, and to 2020 a total of 19 different historic routes, which are still visible in the field, have been mapped. Many roads are no longer visible today, but their approximate position, as they formed a part of a complex network of interconnected historic roads with the already mapped ones, can be assumed.

To better understand the changes in the environment through time, preparatory research consisted of an analysis of modern satellite imagery and historical maps. The earliest sources confirming the existence of Tilurium on the route of historic roads are Roman itineraries.1 However, despite their value in interpretation, they present almost no relevant information for finding physical remains in the field. Therefore, more recent and more accurate maps were used to better reference data collected from satellite images and field surveys. The earliest relevant maps of this sort known to us were nineteenth-century cadastral and military maps, of which the third military survey of the Austro-Hungarian Empire, carried out from 1869 to 1887, at a scale of 1:25 000, proved to be especially valuable.2 Its value lies in the detailed representation of various routes and terrain features that are not visible on modern satellite images due to dense vegetation and/or disuse. The parallel use of these tools has greatly helped in preparing and carrying out field research.

First research area (Vedrine, Jabuka, Velić) Field surveys on the eastern side of the river concentrated on the remains of roads in two different areas.

The field research was broadly divided into two separate parts (or survey areas), which are located on different

The first research area on the eastern side of the river included primarily the area of modern-day villages Velić, Jabuka, and Vedrine (Figure 1). This area consists of a valley bordered by hills on the east, north, and south and on the west by the river Ruda, which flows

Tabula Peutingeriana mentions Tilurio (Tab. Peut., 6, 3), Antonine Itinerary Ponte Tiluri (It. Ant., 337, 5) and The Ravenna Cosmography Ponteluri (Rav. Cosm., IV, 16) 2  Digitized historical maps are available on the website: https:// maps.arcanum.com/en/. 1 

11

Domagoj Bužanić et al.

Figure 2: Known routes A1, A2 and A5 (source: Google Earth, author: D. Bužanić)

into Cetina just 500 m further south. The valley used to be periodically flooded, so most of the important routes passed through or along the edge of bordering hills, which makes this area even more suitable for finding and comparing roads of different eras. Historically, the most important route in the area was the one connecting modern-day Trilj and, consequently, Split, with Livno and Tomislavgrad, i. e. the area of today’s Bosnia and Herzegovina. Until recently, this road ran from a hill located on the southwestern part of the research area to the east. The protruding part of this southwestern hill is dominated by the prehistoric hillfort Križina, around and below which modern houses are located today (Milošević 1998: 261).

to rock formations similar to road ruts, which are actually made by precipitation or other geological and hydrological changes in the environment. Because of this, any survey done in karst is called to caution in interpreting all rut-like formations as road ruts. An additional, more general problem is the dense mid to low-maquis vegetation, widespread in the area, that reduces surface rock visibility (Šekrst et al. 2020). In March 2020, however, rescue excavations were carried out where the remains of the Roman road were assumed near Križina, in Vedrine.3 Excavations have shown that now excavated and confirmed ruts run continuously in the length of almost 70 m, the full length of the trench. Therefore, excavation of this segment proved to be extremely important as a test of our method of finding historic roads. It justified trying to identify roads based on an approach combining historical maps, satellite images, and maps done by our field surveys.

Route A5 It was to be expected that this point had similar importance in Antiquity. Near the hillfort, at a position about 180 m S-E and not far from today’s cemetery, signs of a Roman road were mentioned in the scientific literature (Abramić 1926-27: 14; Bojanovski 1974: 152153; Milošević 1998: 288-no. 512. Šipići; Radan 2016). Initial surveys showed that remains of ruts could be identified in a length ranging from only several meters to, perhaps, 30 m. The wide range and cautious interpretation depended on conditions in the field. Specifically, there are cases in which two possibly continuous ruts are only 20 cm apart, but it was often impossible to find their possible continuation for several or several dozen meters along the assumed route. Also, caution is needed in karst areas due

With a confirmed segment of what is most probably a Roman road (now called route A5, (Figure 2), judging from almost characteristic ruts, the field survey continued north, as such an extension of the route is suggested by historical maps. A 100 m north of the mentioned excavation trench, visible curbs of a road can be seen. Unfortunately, the modern cemetery and a construction waste site are located between these two points (the trench and the curbs), which made it 3  A rescue excavation was carried out by the Department of Archeology, Faculty of Humanities and Social Sciences, University of Zagreb (Tončinić 2020).

12

Discovering Roman Roads of the Trilj Area

Figure 3: Substructure of the route A2 (author: D. Bužanić)

Even though the rest of the A5, the assumed Roman road, could not be followed immediately further, other road segments which show a general orientation towards the south of the research area and the A5 were found. More confirmed road segments will lead us to a more informed assumption of its possible route.

impossible to find their direct physical connection. Nevertheless, judging by the available maps and the orientation of the segments, it seems highly likely they were parts of the same road. It was possible to follow the curbs further NE in a longer length, almost 130 m, with the road following the edge of the hill. Modern houses and dense vegetation cloud the continuation of the road further NE. At the moment, its exact route remains unclear. The same can be said for the now lost alternate route which, according to Austro-Hungarian maps, ran just above the A5, coming from around and below the hillfort and connected to roughly the same path along the edge of the hill to the northeast as the A5. At the moment, this is assumed to be a road constructed sometime after the Roman times, most probably during the rule of the Habsburgs. This, however, can’t be confirmed at the moment, mostly due to modern settlement and construction destroying the road, as well as the dense vegetation, making other parts of the road inaccessible. We hope that additional research will answer these questions.

Route A1 Today’s road Trilj - Livno passes by the early Christian site Velić (Bužanić et al. 2018; Tončinić et al. 2021), which is located on the northern edge of the survey area (Figures 1 and 2).4 There, following the edge of the hill, on the northern side of the modern road Trilj - Livno, there are well-visible and preserved curbs and substructures of the A1 route (Figure 2), which in some places are up to 15 m high. In one part, the road follows the modern road and, it seems, intersects with it at the entrance to the area of Jabuka, after which it passes to the south. According to the local population, the road 4  As the valley is not being flooded anymore, the main route to Bosnia and Herzegovina shifted to the valley, passing along hills on the northern part of the research area, where it passes until today.

13

Domagoj Bužanić et al.

Figure 4: Known routes B8 and B9, along with a present part of the Roman road at site Samoleč (source: Google Earth, author: D. Bužanić)

was completed in the 1920s. The style of construction also does not seem to be much older than that. To this day, no signs resembling the A5 segment have been found on this route.

focused on it. The long-term goal of exploring this area remains to map the routes and connect Tilurium with the determined Roman road at Samoleč (Figure 4). Surveying roads from the site Samoleč

Route A2

The presented part of the Roman road at Samoleč (Figure 5) passes along the slope of the hill and has visible substructures on its eastern side. However, to the north, it disappears in the terraces of arable land that rise above a modern road. Historical maps show the route used to extend to the north, but its remains, after the known section of the road ends, are no longer visible.

Route A2 is located near route A1 but mostly runs south of it. It partially follows today’s road Trilj Livno on its southern side and at one point disappears below it (Figure 2). The local population claims that its construction dates back to the time of the AustroHungarian Empire, i.e., that it is older than the A1 route. This should, of course, be taken with caution. This does point to a cautious conclusion that this is the oldest known route in the area for travelling east, to Livno. The memory of the local population may only provide us with a relative dating, dating its use but not its construction. The road segment is covered with dense vegetation. Its substructures are in poor condition and composed of roughly worked stone, (Figure 3).

Route B8 The maps show the road runs from Samoleč to the north, where it intersects with another road near a water source. The position of the spring and the public well linked to it can still be seen about 400 m east of the church of St. George. The remains of the route coming to that water source from the south are not well visible in the field. Although structures resembling road curbs can be noticed in some places under heaps of stone (Figure 6), many dry-stone walls and demarcations make it impossible to distinguish them from newer structures and thus confirm the remains. Following this route further south, there is a small slope after which all signs of the remains of the road disappear. Today’s agricultural plots stand in their way. After the slope,

The second research area (Vojnić, Gardun) The Roman military camp Tilurium and a previously known and excavated section of a Roman road, located on the site Samoleč, are both on the western side of Cetina. These two sites are separated by a field that stretches below the present-day villages of Vojnić and Gardun. It forms a kind of hinterland of the legionary camp and therefore the project’s field survey is specially 14

Discovering Roman Roads of the Trilj Area

Figure 5: Presented Roman road at site Samoleč (author: D. Bužanić)

route B8 seems to partly disappear. The cadastral map from 1832, however, marks the road at the site of a plot of land that is used today for agriculture. According to the map, the road once ran further south to modern land terraces, under which the remains of the road from the Samoleč position disappear.5

Surveying roads near the site of legionary camp Tilurium Route B12 At the very southern edge of Tilurium, the AustroHungarian military map shows a route (B12) leading from the camp, passing by the presumed legionary amphitheatre, to an intersection with routes B2, B6, B13 and B14 (Figure 8).6

Route B9 Along the slope where the remains of the B8 route disappear, the maps show a side road that runs east. This route is bordered by two substructures. On the north side, it passes under a substructure that supports the agricultural plot above it and there is a substructure that supports the road itself on the south side (Figure 7). These structures make this road about 2.5 m wide.

The curbs of route B12 are in some places quite obvious (Figure 9). Possible ruts, which are yet to be confirmed, can be spotted in some places along this route. On the road, it is noticeable that the surface is dominated by gravel and smaller rocks. This is in obvious contrast to other local roads in the area, where the remains of modern brick and other construction materials dominate. Judging by the maps and satellite images that show a similar situation today to the one from 150 years ago, route B12 runs to the intersection about 500

5  According to the legend of the Austro-Hungarian cadastral map, which is also available on the Arcanum web portal, the roads are marked in brown (https://www.arcanum.hu/media/uploads/ mapire/legend/cadastral.pdf, accessed 30 March 2021).

M. Buovac assumed the location of the camp’s amphitheatre was on the south side of the camp, near the S-W corner (Buovac 2012: 142143)

6 

15

Domagoj Bužanić et al.

Figure 6: Heaps of stone over assumed road curbs of route B8 (author: J. Bužanić)

Figure 7: Part of route B9 (author: J. Bužanić)

Figure 8: Known routes B2, B6, B10, B12 and B14 (source: Google Earth, author: D. Bužanić)

16

Discovering Roman Roads of the Trilj Area

Figure 9: Road curbs on route B12 (author: J. Bužanić)

m SW of the camp, where it meets the routes coming from the south, from the field of Vojnić, and the north, from the area of modern Trilj.

of the terrain. For these reasons, we cannot claim with certainty that route B14 has been identified successfully. Route B2

Route B13

The Austro-Hungarian military map shows that routes B2 and B6 formed the southern approach to the intersection with B12, B13 and B14. This was, at least in part, confirmed in the field. The remains of route B2 have been identified southwest of Tilurium. The route is visible at full width in some parts, with curbs on both sides (Figure 10). The remains of the road stretching south disappear under modern houses.

According to the Austro-Hungarian maps, there were two roads that formed the northern approach to the intersection with B12. The remains of the eastern of the two roads, which passed immediately west of the former rampart of the camp, remain unknown to us. Difficulty in identification is due to the modern use of the road and the decline of the terrain, as well as the general configuration of the area. Possible curbs have been found in two separate places, but due to the terraced nature of the terrain, used today for growing fruit and agriculture, and the absence of any other finds, they cannot be identified as traces of a historic road.

Route B6 Route B6 is in one part almost parallel to route B2 and runs immediately west of it. It seems that just before the intersection with routes B12, B13 and B14, they combined into a unified road. The curbs of this, most probably recent road, are well preserved and are still visible along the modern road.

Route B14 Identifying route B14, which is located west of the unknown route B13, proved to be quite difficult due to the proximity of today’s sand mine and the large amount of smaller stones that are scattered in the area. Possible substructures were found in some places, but it was not possible to follow them further because of the dense vegetation and the terraced configuration

Route B10 By examining the maps, one can notice the existence of another road south of routes B2 and B6. Judging by the depiction on the military map, the route somehow intersected with route B6. At the moment, their 17

Domagoj Bužanić et al.

Figure 10: Part of route B2 (author: D. Bužanić)

connection is not entirely clear in the field. On the other end, it stretched further to the south.

(Figure 12). To the south, the road can be followed to the modern road Trilj-Bisko, in the east of the karst field. Despite some difficulties in following the route, it has been documented in a length of one kilometre.

Further south, there are curbs and substructures that could belong to it, although they cannot be seen in any significant length because of more modern dry walls and stone heaps on top of them. Satellite images and Austro-Hungarian maps show that the route branches into two parts, which is also the case today, and that a modern road uses its eastern branch. On the western branch, which is not under a modern road, visible curbs can be seen. It seems that the curbs could continue further south, now under modern dry-stone walls, but these walls and dense vegetation made it impossible to identify them at the time.

Route B3 Along the banks of the river Cetina, preserved substructures of a road can be seen. The road can be traced in the length of just over 200 m, in the direction of today’s road Trilj - Bisko. According to the AustroHungarian military map, the road should continue towards an intersection with route B1, but a more detailed field survey was impossible due to extremely dense vegetation.

Surveying roads of the eastern part of the karst field

Route B4

Route B1

The Austro-Hungarian military map also shows a road leading from Route B3 to what is now St. Peter’s church, located east of the legionary camp, at a prominent strategic point overlooking Cetina and its crossings. While trying to find this route, a road of undefined date was found. That route cannot be interpreted with greater certainty as the one descending towards route B3. The curbs and substructures of this road are very

In the area of Gardun, about 800 m south of the eastern end of the mapped route B12, route B1 was identified (Figure 11). Its general course north is oriented approximately in the direction of a modern road leading to the walls of the legionary camp. The B1 road is preserved at full width, with curbs and substructures 18

Discovering Roman Roads of the Trilj Area

Figure 11: Known routes B1, B3, B4 and B7 (source: Google Earth, author: D. Bužanić)

Figure 12: Road curbs of route B1 (left) and descent to the river Cetina on the same route (right) (author: D. Bužanić)

19

Domagoj Bužanić et al. visible near the church but moving down the slope they get lost in the dense forest and increasingly steep terrain. Their connection has not yet been confirmed.

kontekstu povijesnih cestovnih komunikacija. Fragmenti 6: 33-46. Bužanić, D., K. Ivak, J. Šekrst and I. Vidović 2019. Mapiranje povijesnih cestovnih komunikacija triljskog područja. Fragmenti 7: 9-23. Milošević, A. 1998. Arheološka topografija Cetine. Split: Muzej hrvatskih arheoloških spomenika. Radan, M. 2016. Izvješće o provedenom arheološkom rekognosciranju u gradu Trilju u naselju Vedrine, zaseok Šipići. Professional report/paper. Sanader, M. 1998. Tilurij-rimski vojni logor-prethodno izvješće s arheoloških istraživanja u 1997. i 1998. Opuscula archaeologica 22: 243-255. Sanader, M. 2001. Tilurij-rimski vojni logor. Prethodno izvješće o arheološkim istraživanjima u sezoni 2000. Opuscula archaeologica 25: 183-194. Sanader, M. 2003. Tilurium I. Istraživanja: 1997.-2001. Tilurium I. Forschungen: 1997.-2001. Zagreb: Arheološki zavod Filozofskog fakulteta u Zagrebu, Golden marketing, Zagreb. Sanader, M., D. Tončinić, Z. Buljević, S. Ivčević and T. Šeparović 2014. Tilurium III. Istraživanja 2002.-2006. godine. Zagreb: Filozofski fakultet Sveučilišta u Zagrebu, Zavod za arheologiju. Sanader, M., D. Tončinić, Z. Šimić-Kanaet, S. Ivčević, Z. Buljević, T. Šeparović and I. Miloglav 2017. Tilurium IV. Arheološka istraživanja 2007.-2010. godine. Zagreb: Filozofski fakultet Sveučilišta u Zagrebu, Zavod za arheologiju. Sanader, M., D. Tončinić, Z. Šimić-Kanaet, Z. Buljević and S. Ivčević 2021. Tilurium V. Arheološka istraživanja 2010.-2018. Godine. Zagreb: Filozofski fakultet Sveučilišta u Zagrebu, Arheološki zavod. Šekrst, J., K. Ivak, M. Jukić and L. Prusac 2020. Izvješće o rezultatima studentskog projekta Mapiranje povijesnih cestovnih komunikacija triljskog područja II. Zagreb Šimić-Kanaet, Z., M. Sanader., D. Tončinić and I. Kaić 2010. Tilurium II. Keramika 1997.-2006. Zagreb: Arheološki zavod Filozofskog fakulteta u Zagrebu, Golden marketing-Tehnička knjiga. Tončinić, D. 2020. Izvješće o rezultatima zaštitnih arheoloških istraživanja na području Gubavica (Vedrine) u 2020. god. Professional report/paper. Zagreb Tončinić, D., V. Matijević and M. Vuković 2021. Imagebased modelling approach in documenting Early Christian memorial chapel in Velić, Croatia, in I. Miloglav (ed.) Proceedings from the 7th Scientific Conference Methodology and Archaeometry, Zagreb, 2nd-3rd December 2019: 43-53. Zagreb: Filozofski fakultet Sveučilišta u Zagrebu. Wilkes, J. J. 1969. Dalmatia. Cambridge, Massachusetts: Harvard University Press. Zaninović, M. 1996. Od Helena do Hrvata. Zagreb: Školska knjiga.

Route B7 According to the Austro-Hungarian military map, the B1 and B3 roads intersect with a road running in the direction of the medieval fortress Nutjak and the modern Bisko municipality (B7). By following the route further south, well-preserved and visible curbs and substructures can be found, with a length of about 600 m. Conclusion In conclusion, we can say that so far this research showed the extreme complexity of identifying and dating historic roads solely by field surveying. However, mapping these roads in the area of Trilj shows the potential in creating a base layer of data for any future research into the topic. Also, Trilj had great strategic importance to wider traffic of goods and people, which is linked to a nearby suitable crossing of the river Cetina, maintaining it throughout most of history, as evidenced by the remains of roads found during the surveys. The intertwined network of historic roads that gradually comes to light with each survey is the result of successive construction that has been going on since Roman times. Despite the complexity, this situation offers an opportunity to compare road construction and the general approach to the construction and reuse of roads in different periods, which, in addition to these non-invasive research methods, should be supplemented by excavations of important segments of the network. Therefore, we believe that mapping is just the first step toward the correct interpretation and dating of roads, as well as toward a more comprehensive study of them. Website Arcanum Maps, viewed 11 February 2022, Google Earth, viewed 11 February 2022, Bibliography Abramić, M. 1926-27. O novim miljokazima i rimskim cestama u Dalmaciji. Vjesnik za arheologiju i historiju dalmatinsku 49: 139-155. Bojanovski, I. 1974. Dolabelin sistem cesta u rimskoj provinciji Dalmaciji. Sarajevo: Akademija nauka i umjetnosti Bosne i Hercegovine. Buovac, M. 2021. Rimski amfiteatri na tlu istočnog Jadrana i zaobalja. Histria Antiqua 21: 129-155. Bužanić, D., B. Glavinić, K. Ivak, J. Šekrst and I. Vidović 2018. Mapiranje arheološkog lokaliteta Velić u 20

A Fort on the Road Andrej Janeš and Nikolina Vrančić Abstract: Situated on the coast in present-day Novi Vinodolski, fort Lopar represents a rare example of an early 4th-century AD fortification in the north Adriatic. With its pentagonal layout and massive walls, it stands as a credit to Roman builders in a time of crisis. It was constructed on the coastline itself probably to control the sea channel between the mainland and the island of Krk. Early 20th-century stray finds indicate the existence of a settlement or villa in the Novi environs. This suggests that the local road connecting the towns of Tarsatica and Senia, which crossed the Vinodol valley, passed near the Lopar fort. So, it seems that the role of the fort was dual, controlling the sea channel and the coastal road. Excavations conducted since 2011 brought to light numismatic finds of the early 4th century AD, with ceramic finds and radiocarbon data that show the usage of the fort till the 6th century AD. It seems that the fort was the farthest reach of Justinian’s rule in the north Adriatic mainland. KEYWORDS: LATE ANTIQUITY, FORTIFICATIONS, ROADS, VINODOL, CERAMIC FINDS

Introduction The fort of Lopar is situated on the eastern edge of Novi Vinodolski, on the sea coast between the bays of Lišanj and Muroska (Janeš 2014: 7).1 It has been known since the 19th century and was written about by G. Szabo, V. Klaić, and J. Brunšmid. (Laszowski 1902; Faber, Matejčić 1963: 44). Based on the construction technique of the fort, G. Szabo believed that it must have been built during Antiquity, as well as that this building technique was surely used in the Middle Ages as well, especially in the coastal areas (Szabo 1920: 190; Janeš 2014: 10). Historical sources reveal almost no data about the fort. It was mentioned only in AD 1598 when an attack under the command of Captain Giovanni Bembo was carried out by the Venetian navy (Laszowski 1902: 231). The fact that the fort was included in the AD 1573 map of Schlavoniae, Croatiae, Carniae, Istriae, Bosniae finitmarumque regionum nova descriptio by Abraham Ortelius shows that its position was significant (Maleković 1992: 45). On the said map, the fort was named Lopur and was included alongside all significant contemporaneous forts in the Vinodol area. Archaeological excavations Although certain researchers were interested in the fort at the beginning of the 20th century, after devastation, no serious research was conducted on it until 2011 (Janeš, Pleština 2015: 70). Since that year, systematic rescue excavations have been conducted, alongside the conservation interventions on the preserved architecture. This research revealed the remains of the northern portion of the fort and the degree of The fort used to be situated on the coast itself, which changed in the first decade of the 21st century when the coastal area was filled with land. 1 

preservation after it had been blown up. Despite the immense damage caused by the event, a site with layers dating from the 4th to the 13th century AD was discovered. Unfortunately, the material remains from the Late Middle Ages and Early Modern period were mostly lost during the removal of construction material after the site was blown up. Research within the northern courtyard revealed intact layers containing movable finds which could be dated to Late Antiquity. This was confirmed by several charcoal samples sent for radiocarbon dating (Janeš, Pleština 2015: 72; Janeš 2014: 20-21). The assumption that the fort was built at the beginning of the 4th century AD is supported by the numismatic finds as well; Emperor Galerius’ reduced follis from the year AD 313,2 Emperor Licinius’ nummus from the same year, and two barely intelligible coins from the late 3rd and early 4th century AD.3 The fort was comprised of two parts; a larger central courtyard and a smaller northern courtyard. The northern courtyard was a part of the fort that had a triangle-shaped layout (Figure 1). The courtyard itself was trapezoidal, while in the northern part, there was a triangular tower (Vrančić, Janeš 2020: 123). According to the earlier sketches of the remains, the fort had two massive rectangular towers on the coast side (Figure 2). Movable finds from Lopar Fort: Archaeological material from Late Antiquity An immense quantity of archaeological material of more than ten thousand artefacts was discovered during the seven years of excavation of Lopar Fort (Vrančić 2019: 18). They mostly include fragments of ceramic ware 2  When it was first published, this find was categorized as Emperor Maximinus Daia’s reduced follis (Janeš, Pleština 2014: 70; Janeš 2014: 17). 3  Numismatic analysis was conducted by Luka Štefan, mag. archeol.

Life Along Communication Routes (Archaeopress 2023): 21–28

Andrej Janeš and Nikolina Vrančić

Figure 1: The layout of fort Lopar (drawing CCI Archive)

Figure 2: View of the sea and Krk island from Lopar (photo by: T. Zojčeski)

22

A Fort on the Road that can be dated to the long time period from Antiquity to the Early Modern period. Besides that, several glass, stone, bone, and metal finds were discovered, alongside more recent finds appearing at the site as the result of devastation during the 20th century (Vrančić 2019: 18). Pottery mostly includes fragments of ceramic ware, but fragments of amphorae from Late Antiquity, as well as spindle whorls and ceramic pebbles also stand out (Vrančić, Janeš 2020: 125).

to Late Antiquity (and the Early Middle Ages), like the ones that were discovered at Lopar (SU29), have analogies in collections of ceramic ware from multiple sites in Slovenia, Istria, the Kvarner Gulf, and Serbia dating to Late Antiquity (Vrančić 2019: 57; Ciglenečki 1984: 313-328, Table 5:56, 59, 62; Ciglenečki 2000: 74; Boltin-Tome 1989/1990: 137-138, Table 2:2-3, 5 and 6, Table 1:1, 2 and 6, Table 2:1; Marušić 1986: 65, 71: Figure 8:5; Juroš-Monfardin 1986: 211, 230: Table 4:2, 4, 231: Table 5:2, 3 and 6, 232: Table 6:3; Ruffieux 2010: 247, 263: Plate 5: IV.A:4; Starac 2004: 32, 22: Table 1:2; Popović and Bikić 2009: 69: Figure 45: 24, 31). Although they were discovered in the layer that was clearly dated to Late Antiquity, several rim fragments from this site can be found at many sites, not only from Late Antiquity but also from significantly later periods (Vrančić 2019: 41, 45). Ceramic vessel fragments with a single wavy line ornament are common in this layer, and prevail at the site in general (Vrančić 2019: 50). However, due to the high fragmentation of ceramic material from Lopar, it is difficult to say whether the fragments originally contained more incised parallel wavy lines. Although this testifies to this ornament’s popularity in (Late) Antiquity, it was used throughout the Early and High Middle Ages and, therefore, cannot be dated precisely (Vrančić 2019: 50-53; Janeš et al. 2017: 350; Bunčić 2010: 69-110). Similar, long-lasting ornaments can be found on other fragments from SU29; several single-lined wavy ornaments, sometimes intertwined (Vrančić 2019: 53), the surface of a vessel horizontally decorated with rows of parallel wavy lines (Vrančić 2019: 54), rows of double wavy lines (Vrančić 2019: 57-58), and parallel wavy lines combined with oblique incised wavy lines sometimes crossed with rows of wavy lines (Vrančić 2019: 59). Fragments of ceramic ware decorated with a modelled plastic band also appeared at Lopar. Their analogies were discovered at the Late Antique site of Rogatica near Barban but were also used in the Early Middle Ages (Vrančić, Janeš 2020: 150). As far as rims are concerned, the ones curved outwards and obliquely cut outwards prevail, which were very common in Antiquity. They are usually dated to the 4th century AD or Late Antiquity, i.e., Early Byzantine period, but were used in the Middle Ages as well (Vrančić 2019: 34). Rims with a somewhat higher degree of curvature of the neck of the vessel are dated similarly and imprecisely (Vrančić 2019: 34, 41, 45). Fragments of rims curved inward, with edges obliquely cut outwards and thickening on the inner side of the edge wall are also dated to a wide period, although they were discovered in the stratigraphic unit dated from the 4th to the 6th century (Vrančić 2019: 45). The same goes for precise dating of rims curved outwards, with a groove on the other edge of the rim (Vrančić 2019: 45). Among the pottery finds, aside from the ware, fragments of amphora rims should also be emphasized. They were interpreted as early types of Lamboglia or Dressel

Several stratigraphic units were dated by the absolute dating method, and, therefore, the layer dating to Late Antiquity can be identified at the site. However, due to the abovementioned devastation of the site in later periods, the material dating to Late Antiquity was present in stratigraphic units dated by relative chronology methods to the Medieval period as well. The stratigraphic units dated with certainty and belonging to Late Antiquity are SU19 (a layer of grey-brown earth), SU31 (a layer of brown earth with crushed rocks), and SU47 radiocarbon dated to the 6th century AD (Vrančić 2019: 84, 165). Those finds date back to the construction of Lopar Fort as one of many forts in the Adriatic defence system used for the surveillance of routes towards Italy (Vrančić 2019: 87). Therefore, these stratigraphic units are dated to the period from the 4th to the 6th century AD. The colours of ceramic vessel fragments from the site vary from ochre, light orange, brick red, to grey and black. Their facture is mostly coarse, tempered with crushed calcite (Vrančić 2019: 23-25). Layer SU29 revealed fragments of ceramic ware and other pottery, alongside a few metal, glass, and stone finds (Figure 3:1-7) Fragment of a ceramic spindle whorl originates from stratigraphic unit SU29. It is dated to Late Antiquity, as well as the fragments of ceramic ware, with multiple analogies from Late Antiquity at the sites of Betiga near Barbariga, Forum of Flavium Fulfinum on the island of Krk, and the Basilica of St Cecilia in Guran (Vrančić 2019: 26, 42; Juroš-Monfardin 1986: 209-233; Konestra 2015: 172, 207; Ruffieux 2010: 247, 265). The same stratigraphic unit yielded a ceramic vessel rim fragment, analogous to the rim discovered at the site of Pul vele crikve in Rijeka, which could have been used up until the Early Middle Ages, i.e., the 6th-7th century AD (Vrančić 2019: 44; Bekić in press: cat. 10). The next fragment belonging to Late Antiquity is the ceramic vessel rim with a common typology, analogous to the ones in Tonovcov grad near Kobarid, the Guran basilica, and some Austrian sites from the period (Vrančić, Janeš 2020: 125; Ruffieux 2008: 254, 256, 262; Ruffieux 2010: 247, 263; Ruffieux 2006: 272-273, 278-279, Table 3:26, Rodriguez 1997: 153-177, Table 4:32; Ciglenečki 1984: 313-328, Table 1:4-5; Modrijan and Milavec 2011: 172-175, Table 4:1). Fragments of ceramic rims dating 23

Andrej Janeš and Nikolina Vrančić amphorae, light yellow or orange in colour, and coarse in facture. Their analogies can be found at the site of St. Nicholas’ Church in Zadar, dated between the 4th and 7th century AD (Bekić 2017: 39, Vrančić and Janeš 2020: 156; Vrančić 2019: 67). This stratigraphic unit also yielded fragments of building ceramics that are generally dated to the long period of Late Antiquity and the Middle Ages (Vrančić 2019: 26). This layer from Late Antiquity also yielded fragments of glass with motherof-pearl iridescence, used during Late Antiquity and the Middle Ages, making them impossible to be dated more precisely or their function to be determined. Their analogy can be found at Augusteum in Narona (Vrančić 2019: 70; Buljević 2005: 109-149). A chance find of a cannonball stands out among the metal finds from this stratigraphic unit. It probably ended up in this context during the devastation of the site in AD 1598 (Janeš 2013: 41). Besides, a metal band, i.e., a plate was discovered. It has semi-circular endings and was used as a tool for weapon maintenance in the restless times of this fort’s history (Vrančić, Janeš 2020: 167). Two finds of whetstones from this layer also testify to everyday life in the fort (Vrančić, Janeš 2020: 170).

Several fragments from this layer have facture made of refined clay with a reddish finish. They are interpreted as fragments of sigillata ware decorated with horizontal ribs on the surface of the vessel. According to analogies, they can be dated to the second half of the 5th century AD (Vrančić, Janeš 2020: 156). This layer also revealed a fragment of a ceramic spindle whorl with a very shallow groove on the top and bottom side, dated to Late Antiquity (Vrančić, Janeš 2020: 157). Since they have been used for so long, it is difficult to date the several fragments of construction material from this layer precisely. Nevertheless, they were assigned to the Late Antique layer based on the absolute dating of other finds. The Antique road in the context of the Vinodol Valley Vinodol includes a 45 km long valley which stretches from Trsat and Rječina in the N-W, along the seacoast, and all the way to Ledenice and Novi in the S-E (Kosanović 2012: 8). The valley is bordered by two slopes belonging to the Velika Kapela complex, parallel with the coastline (Kostrenčić 1923: 110). Small rivers and streams link the Vinodol Valley to the sea - Rječina near Trsat, Dubračina or Vinodolska rijeka near Crikvenica, and Suha or Bribirska Ričina near Novi (Klaić 1901: 51).

The other stratigraphic unit dated by the absolute dating method to the period from the 4th to the 6th century AD is SU31 (Figure 3:8). This layer revealed a fragment of an amphora handle, but its typology is difficult to determine (Vrančić 2019: 26). Rim, neck, and shoulder fragments have their analogies at the contemporaneous sites of Korinjski hrib in Slovenia, Teurnia-Bischofkirche in Austria, and the sites of Muntajana and Rogatica in Istria. However, these types of pottery were used in the Early Middle Ages as well, all the way to the Late Medieval period (Vrančić 2019: 41-42; Table 7:1). There is a similar problem with the dating of the ceramic ware fragments decorated with horizontal parallel incised lines (Vrančić, Janeš 2020: 151). The amphora handle fragment from this layer also indicates that the fort was inhabited in Late Antiquity (Vrančić 2019: 26, 62). Other amphora rim and body fragments actually belong to an earlier type, Lamboglia or Dressel. They have analogies at the site of St Nicholas’s Church in Zadar (Bekić 2017: 39: Figure 1). A small fragment of a thicker bronze wire stands out among the metal finds. Before any further analysis, it cannot be more closely determined (Vrančić 2019: 82).

In Antiquity, the Vinodol area was situated between two larger Roman urban settlements: Tarsatica in the north, and Senia in the south. The area was characterized by smaller settlements such as Volcera, alongside potential vici, business complexes, and road stations such as the one in Ad Turres. There are two potential communication routes, one along the coast and the other through the valley itself (Goss 1989: 92). This thesis is confirmed by the well-known map of Tabula Peutingeriana, in which two roads were charted between Tarsatica and Senia (Figure 4). No systematic research on the archaeological topography of the Vindol area had been conducted by the beginning of this century, but in the first 15 or so years, two works were published. In the first one, R. Starac compiled all previous finds about the site from the Roman administration (Starac 2006), and in the second, he and G. Lipovac Vrkljan updated the list with all known and newly-discovered sites in the area (Lipovac Vrkljan, Starac 2014). These publications revealed reflections on certain traces of potential Roman roads based on observations made during onsite visits. Due to the high urbanization of the coastal area, most data was collected in the Vinodol hinterland.

Finds dated to the period from the 4th to the 6th century AD were discovered in layer SU47 (Figure 3:9). Among the finds, multiple ceramic vessel fragments stand out. Their walls were light orange in colour and were interpreted as fragments of type Hayes 23 vessels originating from North Africa. The trade conducted between the 3rd and the 6th century AD enabled the vessels to appear here. Their analogies can be found in Zadar (Bekić 2017: 41; Vrančić and Janeš 2020: 156).

It is assumed that a section of the road of Aquileia - Senia is situated in the village of Bakarac, while the remaining portion of the branch of the coastal road is between 24

A Fort on the Road

Figure 3: Characteristic Late Antique finds, 1-7 from SU029, 8 from SU031 and 9 from SU047 (drawings by: A. Bendeković, T. Marketin)

the peaks of Gradac, Hreljin, Gomilica, and Medomišalj (Lipovac Vrkljan, Starac 2014: 95-96).

Sections of a 30 m long track were discovered in southern Vinodol, at the site of Humac near Pavlomir (Starac 2006: 82-83). All the abovementioned finds confirm the presence of the Roman road which stretched across the Vinodol Valley. Graves containing finds from Antiquity discovered at the site of the construction of Lišanj Hotel in Novi Vinodolski indicate the presence of a coastal road (Brunšmid 1895: 152-153). The question of the place where the two Vinodol roads merge remains.

The remains of the assumed Antique communication route were recorded at the site of Tribalj - Kloštar (Gašparovići) in present-day Vinodol and are considered to be a part of the road that connected the present-day Bakarac area to the Crikvenica area (Lipovac Vrkljan, Starac 2014: 97). 25

Andrej Janeš and Nikolina Vrančić

Figure 4: Sections of the Tabula Peutingeriana with the two roads in Vinodol (source: https://digitalmapsoftheancientworld.com/ancient-maps/tabula-peutingeriana/)

Figure 5: Late Antique forts in Vinodol and on Krk island (map by: A. Janeš, source: EU-DEM produced using Copernicus data and information)

26

A Fort on the Road

Figure 6: Viewshed analysis of fort Lopar (map by: A. Janeš, source: EU-DEM produced using Copernicus data and information)

Was it on the mouth of Suha Ričina in Novi (in which case, the traffic would continue down the coastal road towards Senia) or in the Povile area, east of the presentday Novi Vinodolski?

maritime routes, as well as using the resources of the nearby island of Krk.

The Lopar Fort’s position enabled it to hold two roles; it controlled the coastal road, but also a portion of the water area towards the eastern coast of the island of Krk. The establishment of the Late Antique settlement at the site of Orgul as well as the Veli Grad Fort on the nearby cape of Glavina surely gave Lopar Fort more significance in controlling the sea. The analysis of visibility range from the fort’s position indicates that the fort had better visual control of the sea than it had of the mainland. The direct surveillance over the coastal road would indicate that it was the only significant road in the area. Therefore, it should be assumed that the two Vinodol roads merged at the mouth of Suha Ričina. Since the fort is topped by the nearby hill of Osap with the remains of a Late Bronze Age hillfort (Matejčić 1981: 309), it can be concluded that the fort did not originally have a significant role in the defence of the wider area, but that it was used to control the narrow pass along the sea, i.e., the road towards the southeast. The considerable visual control of the sea and the position on the seacoast itself indicate that the fort relied on

Bekić, L. (in press), Gruba antička i ranosrednjovjekovna keramika, in N. Radić-Štivić, N. and J. Višnjić (eds) Arheološka istraživanja trga Pul vele crikve u Rijeci 1. Bekić, L. 2017. Antička, kasnoantička i srednjovjekovna gruba keramika i amfore kod crkve sv. Nikole u Zadru, in Bekić L., Vršić Š., Pešić M. And Surić R. (eds) Sveti Nikola u Zadru. Arheološko iskopavanje u samostanskom sklopu sv. Nikola u Zadru 2014.- 2016. Prvi dio: 37-62. Zadar: Međunarodni centar za podvodnu arheologiju Boltin-Tome, E. 1989/1990. Valovnica na keramiki iz Predloke, Histria Archaeologica 20/21: 128-138. Brunšmid, J. 1895. Arheološke bilješke iz Dalmacije i Panonije, Vjesnik Hrvatskog arheološkog društva sv. 1: 148-183. Buljević, Z. 2005. Stakleni inventar iz Augusteuma Narone, Vjesnik za arheologiju i historiju dalmatinsku 97(1): 109-149. Bunčić, M. 2010. Naseobinski pokazatelji kasnog srednjeg vijeka zagrebačkog nalazišta Stenjevec, Vjesnik Arheološkog muzeja u Zagrebu 43 (3.s.): 69-

Bibliography

27

Andrej Janeš and Nikolina Vrančić 110. Ciglenečki, S. 1984. Symposion zu Fragen der spätantiken und völkerwanderungszeitlichen Keramik, Archaeologia Austriatica 68: 31-328. Ciglenečki, S. 2000. Tinje nad Loko pri Žusmu. Poznoantična in zgodnjesrednjeveška naselbina. Ljubljana: Inštitut za arheologijo Znanstveno raziskovalni center Slovenske akademije znanosti in umetnosti. Faber, A. and R. Matejčić 1963. Izvještaj o stanju gradine Lopar u Novom Vinodolskom. Vijesti muzealaca i konzervatora Hrvatske 2: 44-47. Goss, V. P. 1989. Vinodolski teritorij i njegova prostorna organizacija. Radovi Instituta za povijest umjetnosti 12/13: 90-95. Janeš, A. 2013. Novi Vinodolski - utvrda Lopar, in Z. Wiewegh (ed.) Hrvatski arheološki godišnjak 9/2012: 513-515. Janeš, A. 2014. Nove spoznaje o utvrdi Lopar u Novom Vinodolskom. Preliminarni rezultati arheoloških istraživanja 2011.-2013. Vinodolski zbornik 15: 7-30. Janeš A. and I. Pleština 2015. Utvrda Lopar u Novom Vinodolskom - antička i srednjovjekovna utvrda: istraživanja 2011. i 2012. godine, in Z. Ettinger Starčić and D. Tončinić (eds) Istraživanja na otocima: znanstveni skup, Lošinj, 1.-4. listopada 2012. Izdanja Hrvatskog arheološkog društva 30: 67-75. Zagreb: Hrvatsko arheološko društvo, Lošinjski muzej. Janeš A., I. Hirschler Marić and A. Azinović Bebek 2017. Stari Perkovci-Sela, ruralno naselje 14. stoljeća, in T. Sekelj Ivančan, T. Tkalčec, S. Krznar and J. Belaj (eds) Srednjovjekovna naselja u svjetlu arheoloških izvora/ Mediaeval settlements in the light of archaeological sources (Serta Instituti Archaeologici 6): 337-388. Zagreb: Institut za arheologiju. Juroš-Monfardin, F. 1986. Pokušaj sistematizacije kasnoantičke i ranobizantske keramike grube fakture iz profanog objekta u Betigi kod Barbarige, in V. Jurkić (ed.) Rezultati i spoznaje arheoloških istraživanja u Istri i Hrvatskom primorju, Izdanja Hrvatskog arheološkog društva 11/2: 209-233. Zagreb: Hrvatsko arheološko društvo. Konestra, A. 2015. Keramika s Foruma Municipia Flavia Fulfinuma (otok Krk, Hrvatska) - istraživanja od 2007. do 2013. godine. Prilozi Instituta za arheologiju 32: 147-214. Kosanović, O. 2012. Državina krčkih knezova - Vinodol, Senj i Krk od početka 14. stoljeća do 1420. godine, Unpublished PhD dissertation, University in Zagreb. Kostrenčić, M. 1923. Vinodolski zakon. Rad Jugoslavanske akademije znanosti i umjetnosti 227: 110-230. Klaić, V. 1901. Krčki knezovi Frankapani. Zagreb: Matica hrvatska. Laszowski, E. 1902. Hrvatske povijesne građevine: mjestopisni i povjesni opisi gradova, kula, samostana, crkava i drugih povjesnih gradjevina domovine Hrvata. Zagreb: Emilij Laszowski.

Lipovac Vrkljan, G. and R. Starac 2014. Antička mikrotopografija Hrvatskog primorja, in G. Lipovac Vrkljan, B. Šiljeg, I. Ožanić Roguljić and A. Konestra (eds) Rimske keramičarske i staklarske radionice. Proizvodnja i trgovina na jadranskom prostoru. Zbornik II. Međunarodnog arheološkog kolokvija, Crikvenica, 28.-29. listopada 2011: 93-105. Crikvenica: Institut za arheologiju. Marušić, B. 1986. Prilog poznavanju kasnoantičkog Nezakcija. Starohrvatska prosvjeta 16: 51-76. Maleković, V. (ed.) 1992. Granice Hrvatske na zemljovidima od 12. do 20. stoljeća, Zagreb: Muzej za umjetnost i obrt. Matejčić, R. 1981. Pregled kulturno-povijesnih spomenika Vinodola, Vinodolski zbornik II: 309-336. Modrijan, Z. and T. Milavec 2011. Poznoantična utrjena naselbina Tonovcov grad pri Kobaridu. Najdbe/ Late Antique fortified settlement Tonovcov grad near Kobarid. Finds (Opera instituti archaeologici sloveniae 24). Ljubljana: Znanstveno raziskovalni center Slovenske akademije znanosti in umetnosti. Popović, M. and V. Bikić 2009. Vrsenice. Kasnoantičko i srpsko ranosrednjovekovno utvrđenje/ Vrsenice - Late Roman and Serbian Early Medieval Fortress. Beograd: Arheološki Institut. Rodriguez, H. 1997. Die Zeit vor und nach der Schlacht am Fluvius Frigidus (394 n.Chr.) im Spiegel der südostalpinen Gebrauchkeramik. Arheološki vestnik 48: 153-177. Ruffieux, P. 2006. La céramique de la basilique à trois nefs de Guran en Istrie (Croatie). Hortus Artium Mediaevalium 12: 271-279. Ruffieux, P. 2008. La cèramique de l’aglomération de Guran en Istrie (Croatie): essai de classification. Hortus Artium Mediaevalium 14: 249-264. Ruffieux, P. 2010. Céramique commune de l’antiquité tardive découverte sur le site de l’église Sainte Cécile a Guran, en Istrie, Hortus Artium Mediaevalium 16, 247-266. Starac, R. 2004. Prilog poznavanju materijalne kulture stanovnika tarsatičke Liburnije između petog i devetog stoljeća, Zbornik sv. Vid IX: 21-35. Starac, R. 2006. Antička topografija Hrvatskog primorja. Vinodolski zbornik 11: 73-84. Szabo, G. 1920. Sredovječni gradovi u Hrvatskoj i Slavoniji. Izvanredno izdanje Matice hrvatske. Zagreb: Matica hrvatska. Vrančić, N. 2019. Pokretni srednjovjekovni arheološki nalazi iz utvrde Lopar u Novom Vinodolskom: istraživačke kampanje od 2011. do 2017. godine. Unpublished MA dissertation. University in Zagreb. Vrančić, N. and A. Janeš 2020. Razvoj utvrde Lopar u Novom Vinodolskom kroz pokretne nalaze / Development of Fort Lopar in Novi Vinodolski through movable finds. Starohrvatska prosvjeta 47, Vol. III: 119-230.

28

An Island of Connectedness: Intra-Insular and Inter-Regional Communications of Rab Island (North-East Adriatic, Kvarner Gulf), Croatia Ana Konestra, Paula Androić Gračanin and Fabian Welc Abstract: Drawing from archaeological data on Prehistoric and Roman settlement patterns, and respective sepulchral features, just as on later historic and cartographic sources, in this contribution it will be shown how movement along with mobility influenced on and were an integral part of settlement dynamics on the island of Rab (NE Adriatic, Croatia). The employed approach sees the practice of movement as dependent on specific social and historical context, thus predicting a diachronic development of routes of interaction. Therefore, one of the aims was the identification of corridors of movement that persist within the island’s landscape, connecting its features and signalling persistent uses of locations. Such an approach allowed the reconstruction of a landscape of movement, where place and landscape are intertwined through spatiality and temporality. On the other hand, the connection of land and maritime networks of movement was approached by analysing how terrestrial connectivity interacted with maritime one and evidencing those sites that benefitted mostly from inter-island connectivity lasting throughout all tackled periods. KEYWORDS: ISLAND OF RAB, MOBILITY, SETTLEMENT DYNAMICS, LANDSCAPE OF MOVEMENT, MARITIME NETWORKS

Islands are often thought of as isolated entities that cannot be regarded without the sea (in general see Boomert, Bright 2007), thus mobility, communication networks, and connectivity of insular spaces are equally often viewed as depending exclusively on seafaring and sea routes. In this preliminary overview of Rab island’s connectedness in its long-term history both intra and inter-island communications, as essential structures in human interactions (Snead et al. 2009) will be discussed, emphasizing the diversity of an island’s long-term communication patterns reflected in its landscape. Using data collected through different research projects,1 a vibrant Roman settlement pattern was identified within the hinterland of Rab island. This network of settlements and related features hinted at the existence of a communication network spanning the island’s inland (Lipovac Vrkljan, Konestra 2015: 131; Jurković 2019) and functioning beyond Antiquity, and in part probably predating it. It is exactly this terrestrial network of communications, its reconstruction and persistence, that will be explored, showing that not all island communications depended upon the sea, but that the two networks worked in synchrony. Thus, it will be shown how movement and mobility influenced and were an integral part of settlement dynamics on the island (e.g., Aldred 2021: 36-77).

The theoretical framework within which we will be dwelling sees the practice of movement as dependent on specific social and historical context, thus predicting a diachronic development of routes of interaction (Gibson 2007: 62, 74). The latter is seen as corridors of movement that persist within the landscape, connecting its features, which can, on the other hand, also signal persistent uses of locations (van Lanen et al. 2018), that is a landscape of movement (Snead et al. 2009: 1). In this view place and landscape are seen intertwined through spatiality and temporality, that is movement and activity (see Mlekuž 2014: 8, 18 with earlier bibliography; Gibson 2007: 63-64). Such routes do not necessarily precisely follow the layout of actual roads, especially in regard to the earliest periods here examined, and are thus more dynamic and adaptive to a changing landscape, while being generally orientated as the actual road network, where one is present (van Lanen et al. 2015: 145). On the other hand, within traditional “island archaeology”, most often a network of connections is limited on intra-island network scrutiny, as islands are observed as “distinct units” (e.g., Anderson 2004; Renfrew 2004; Fitzpatrick 2007) in contrast to their surroundings. Therefore, both intra and inter-island communications over the long-term history of Rab island will be discussed here, emphasizing the diversity of this island’s various features, rather than observing it as a homogeneous unit of space.

Mainly the Archaeological topography of the Island of Rab (conducted through a collaboration of the Institute of Archaeology, Zagreb, and the Institute of Archaeology, Cardinal Stefan Wyszyński University in Warsaw, directed by A. Konestra and F. Welc) and RED - Roman economy in Dalmatia (2013 - 2018 grant by the Croatian science foundation, PI G. Lipovac Vrkljan).

In order to propose, and then analyse potential corridors of movement, firstly a look at the environment and geomorphology of the island is needed. Rab is a mediumsized island of the Kvarner bay, NE Adriatic, located just a few kilometres from the coast, that is the slopes of the

Introduction and data sources

1 

Life Along Communication Routes (Archaeopress 2023): 29–42

Ana Konestra et al.

Figure 1: Island of Rab vegetation cover and flood risk - modern data (ArchaeoTopoRab GIS; author: A. Konestra)

north Velebit mountain range. Throughout Antiquity and to this day, only one urban centre developed on the island. It is the town of Rab, persisting on the location of one of the numerous hillforts marking habitation during Prehistory (Androić Gračanin et al. 2020: 449-450). The island features large flysch fields where alluvial deposits have created suitable soil for cultivation, while the hilly parts of the island are characterized by carbonates and thus karst (Lončar 2012) and are suitable for grazing and silviculture. Climatically, the island falls within the Cfa zone, which is a hot-summer Mediterranean climate (Filipčić 1998). Although most of its surface is hilly, the overall elevations are not excessive, thus virtually all landscapes are reachable without serious

impediments. In particular, the west-east elongated plains in the central part of the island and on the Lopar peninsula stand out. Equally, although extrapolated from modern data, vegetation cover and water courses do not seriously influence mobility or render certain areas inaccessible. Nevertheless, swampy areas with higher flood risk might be regarded as unfavourable for both stable settlement and the creation of stable road infrastructure, and these are usually located within shallow bays at the fringes of said fields (Bay of Kampor, Sveta Fumia bay, Supetarska bay, Lopar bay etc.), (Figure 1). Rural settlement on the island is mostly sparse, with few small nuclei usually formed by a larger cluster of houses, lacking any urban features (Budak 1987: 30

An Island of Connectedness

Figure 2: The network of roads and paths on the island of Rab extrapolated from the Franciscan cadastre, with indicated the main settlement nuclei (ArchaeoTopoRab GIS; author: A. Konestra)

195), and isolated households, both marked by recent anthroponyms.2 Toponyms in use, at least since the Middle Ages, to define larger areas of the island (Barbat, Banjol, Palit, Kampor, Gonar, Mundanije, Supetarska draga, and Lopar) have today come to indicate the administrative borders of the island’s settlements.

and systematise data gathered in the field, just as a wide range of cartographic layers, comprising various modern environmental datasets (Lipovac Vrkljan et al. 2017). All archaeological evidence was gathered from available sources, including historical and bibliographic ones, both grey literature and published works, and museum materials. All such data allowed us to plan the conduction of remote and on-site surveys, including aerial imaging, geophysics, reconnaissance, underwater survey, and finally trial trenching. Nevertheless, due to natural and anthropic modifications to the landscape, certain difficulties arose in both data acquisition and the interpretation of collected data. Mostly, overgrown fields and pastures, and erosion, often hinder field data acquisition.

Secondly, a short overview of available data and methods should be laid out.3 The basic analytical tool used is an ad hoc GIS database developed to collect 2  Modern urbanisation of the island’s historic landscapes, sprung from the mid-20th century by tourism, has partly obliterated the traditional settlement pattern. 3  For methodologies see e.g., Patitucci, Uggeri 2007.

31

Ana Konestra et al.

Figure 3: An area at the border of Mundanije and Banjol-example of secondary paths branching from a main E-W communication corridor; data from FC (ArchaeoTopoRab GIS; author: A. Konestra)

Up until now no definitive evidence for Roman road infrastructure was discovered on the island, nevertheless, as mentioned, sites’ distribution and the combination of archaeological, historic, and cartographic data allow us to propose selected case studies based on which landscapes of movement and their persistence through time can be analysed, while in some cases historic data availability allows further discussion. The first step to the here proposed reconstruction was the digitalization of the road and paths network from the so-called Franciscan cadastre (FC) of the 1820s (Horović Vuković 2012), as the oldest record available. This was done by first georeferencing the sheets of the cadastre on the bases of features from the digital orthophoto of the island, and then each road segment, named or anonymous, was recorded with its main details as unmovable find, and finally, its layout was rendered as a linear feature (Figure 2). By combining this road network with the established settlement patterns for several historic periods, possible corridors of movement and their persistency through time are proposed.

the road system depicted in the cadastre it was first noticed that several continuous E-W roads exist, modelled on the bases of the island’s geomorphology, but also linking the main rural settlement nuclei to Rab town. An analysis of several branches stemming from such main corridors allowed us to suppose that most could be of probably recent date, as they form links to settlement units possibly dating to the early Modern period, and usually located on the borders of the field or the foothills, as is the case of e.g. Supetarska Draga and Mundanije fields, in the central part of the island (Figure 3), but also Kampor field and part of Barbat. Nevertheless, late Medieval historic sources provide insights into communications’ organisation within the island’s hinterland, allowing us to propose that the aforementioned pathways might originate from an earlier period. In fact, in the notary documents of Nicolo Curtarolo (active between 1372 and 1383) (Mlacović 1995: 85-86; Mlacović 2005: 525-529), three types of roads are mentioned: via comunis, via publica4 and via convicinalis (Mlacović 2005: 525). The first two indicate, according to Mlacović, public roads, while the third is used to indicate secondary paths and culde-sacs used by neighbouring households to reach their homes (Mlacović 2005: 525). This could also indicate that some of the sparse settlements and houses that are today visible in the countryside might have earlier precedents, or at least follow the same model, thus

Earliest historical data–the early Modern communication network as an echo of Medieval pathways Unfortunately, detailed cartographic data on the island of Rab is scarce and mostly of late date, and no comprehensive research regarding it has been carried out yet. As mentioned, the only detailed enough cartographic data source is the 1820s FC. By analysing

A term often mentioned in Italy in connection to previous, Roman roads (Patitucci, Uggeri 2007: 331).

4 

32

An Island of Connectedness being linked to habitation developed in junction with the field system (Budak 1987: 195; Mlacović 2012: 168). However, no archaeological data that would back such a claim exists.

actual infrastructure might have switched to higher grounds, that is away from flooded or swampy areas (Patitucci, Uggeri 2007: 324-325; see also Dall’Aglio 1997). A re-appropriation of lower laying areas and consequent descent of the actual road infrastructure is accompanied by investment in water management infrastructure, such as drainage channels, which is, at least in northern Italy, a feature dated from the early Middle Ages on, and often connected with landscape management conducted by the Benedictine monks (Patitucci, Uggeri 2007: 325). It is exactly the Benedictine monks of Supetarska draga who could have, once obtaining St. Peter’s and St. Cyprian’s churches, along with a wider landed possession in AD 1059 (Mlacović 2005: 513), reclaimed the lands and re-traced the roads, or rather re-modelled them to match their new needs, possibly retracing previous corridors of movement.9 In fact, the whole southern section of Supetarska Draga field is crisscrossed by channels leading water towards the bay, still well visible in the FC, but today mostly abandoned.10 These channelling networks might have been set up also in connection to the salt pans located in the bay to the west of the monastery, and mentioned in a document dated AD 1199 (Mlacović 2005). A hint to these Benedictine interventions might come from the road stretching from the monastery towards the northern shores of Supetarska bay and the monastery’s mill,11 named by the cadastre as “Strada al Molino” (Road to the Mill), and possibly also mentioned in the AD 1199 document. It is not, however, impossible, that these reclamations allowed, or included, repairs of an older road, the mentioned antiqua via comunis, but not knowing its actual layout, further elaborations are difficult. The church of St. Peter existed before its donation to the monks, thus its late antique date is possible (Jarak 2017: 51-52; Jurković 2019: 120-121). Also, earlier structures have been supposed in its environs (Lipovac Vrkljan, Konestra 2015) and have possibly been ascertained below the monastery as well (Starac 2017), thus communication leading to this settlement must have existed in Roman times.

Another, earlier document, that allows us to draw information about the road network on the island, and methods of its upkeep, is the Statute of Rab (AD 1328),5 which mentions the existence of vias and cararius, and of supervisors - iuduex carariarum - in charge of control of the public roads, also prescribing fines in case the roads are not kept in good shape (Mlacović 2005: 529530, n. 56). Furthermore, in a document dated to 1380, related to a real estate contract,6 a public road along Supeteraska Draga field is mentioned as antiqua via comunis (Mlacović 2005: 525). According to Mlacović, this would indicate, along with other data, and the fact that there is no mention of new roads construction anywhere in the documents, that in the 14th century, the whole road network was mostly formed, and that one of the eastwest corridors (and roads) traced along the Mundijsko polje and Supetarska Draga fields are of a much earlier date. According to this author, as early as Late Antique. These roads and paths are also often mentioned, along with water channels, as markers of property (Mlacović 2005; see also Degrassi 2002: 219; Patitucci, Uggeri 2007: 325; Francesconi 2011), which might signal not only that this road system in the 14th century is “frozen” as it was in a previous period, but also that the whole organisation of the landscape of Supetarska draga, and probably Mundanijsko polje, are by that time old and well established (Mlacović 2005: 528-529; Mlacović 2012: 169).7 Another document dealing with property boundaries, in this case of plots belonging to St. Mary’s church (ex-Cathedral of Rab),8 dated to AD 1086, mentions roads (cararia) that define the agricultural landscape of the island (CD I: 184-185 (n. 144)). How far back the actual roads mentioned in these sources could be dated is hard to tell due to the current lack of archaeological data, but if we are to draw comparisons from neighbouring areas, it seems that dating them in the early Middle Ages could be valid. In fact, data from northern Italy would seem to suggest a general lack of care for roads and other infrastructure after the 6th century AD, possibly accompanied by a generally harsher weather pattern, and thus while maintaining the same corridors of movement, the

The main communication corridor from the island’s largest fields towards the town of Rab might be reconstructed over the low hills dividing them from Kampor field (known as Monchi and e.g. monte S. Helye (see Vuletić 2020: 241-243)), as testified by the account of V. Brusić in 1926 (Brusić 1926: 179), and by the network of paths and roads shown on the FC descending 9  Worth mentioning is D. Mlacović’s opinion that the land donated to the Abbey by the Commune were parcels that the latter did not need, that is, did not feel their loss in its income (Mlacović 2012: 169). It is, thus, possible, that these were at least partly marginal lands in need of reclamation. 10  A further hint to problems of water management might come from some Rab toponyms, such as Palit (Paludo), Bagnol (Bagnolo) and perhaps also the very Valle, as Supetarska draga was named (see e.g. Dall’Aglio, Franceschelli 2012: 90). 11  For the mill (see Mlacović 2005: 529-531; Mlacović 2012: 168).

Many statutes of Medieval Communes in Italy also mention the road system and waterways, and their upkeep, see Patitucci, Uggeri 2007: 326; Dall’Aglio, Franceschelli 2012: 87-88. 6  The purchase of land by the Abbot of St. Peter in Supetarska draga. 7  Similar markers of property boundaries are known elsewhere on the island, (see e.g., Vuletić 2020: 241) for the 15th century description of properties at another location on the island. 8  On the basis of a few toponyms that can be more securely located, the areas in question span from Barbat to Supetarska draga/Gonar. 5 

33

Ana Konestra et al.

Figure 4: Pathways and archaeological evidence in the area of Snuga/Caplaca - Za Markovićem and Nad Matušani sites (ArchaeoTopoRab GIS; author: A. Konestra)

towards south from St. Elias church.12 Here, sources also mention the bridge of St. Gervasius (Mlacović 2012: 172).

have been determined on the bases of reconnaissance (Nad Matušani and LOK-177), (Lipovac Vrkljan, Konestra 2015), and another small church has been ascertained (St. Mark), (Figure 4).

Communications persistence from Roman times

Several segments of this path were excavated, 13 and finds confirmed that the section of cobbled pathway discovered under the contemporary path close to Za Markovićem is probably of early modern date, if not later (Figure 5), just as are the segments stretching onto terracing walls close to Nad Matušani site. A few unearthed fragments of tile and amphorae sherds are to be related to scattering from the Za Markovićem site. Thus, while the road bordering from south of the Za Markovićem site does not present traces of Roman infrastructure, its earlier counterpart might have been laid on either side of the site defining the development of mobility in this segment of Mundanije field. The presence of a necropolis, just as nearby contemporaneous sites more to the east, but also the west towards Supetarska Draga, certainly indicates the direction of movement, further confirmed by the late Medieval chapels, possibly signalling a persistence of place, movement, and organization of the rural territories.

Along with the settlement pattern, other built structures in the landscape allow to reconstruct persistent corridors of movement (Patitucci, Uggeri 2007). Such are the small churches and chapels that dot the island’s coasts and hinterland, mostly of (late) medieval date, though on some occasions traces of earlier structures have been ascertained (e.g., Brogiolo et al. 2018). One such case of persistent use of location and probably communications is the site of St. Anastasius/ Za Markovićem, in the area known as Snuga or Caplaca. Here, a Roman rural settlement has been ascertained through remote sensing and excavations, and its necropolis is ascertained by finds of Roman sepulchral monuments (Konestra et al. 2021). The site is located a few hundred meters to the east of St. Anastasius church and the two are connected by a narrow path, while further east along the latter, two other Roman sites 12  The FC is showing in this segment of the hill a more complex network of roads, two of which are named “Strada s. Elia”, but from the said church only one descends towards the valley, named “Strada per Valle”, which then forms the southern longitudinal E-W corridor bordering the field. A path, still in use today (“Strada per Duda” on the FC) might also be a remnant of earlier communications.

13  A rescue excavation was conducted in 2019 by the Institute of Archaeology, Zagreb (A. Konestra, G. Lipovac Vrkljan) during setup of a bike trail by Rab Municipality.

34

An Island of Connectedness

Figure 5: Examples of Early Modern (or later) path on Rab island: 1-Segment of Early Modern (or later) path excavated at Za Markovićem; 2-Segment of the path above Za Markovićem; 3- Segment of the possibly enlarged path above Mundanije; 4-Path and bridge at Barkunićeva voda (Banjol); 5-Segment of the path above Mundanije (original width?) (author: A. Konestra)

Similarly, settlement sites and evidence of necropolises within the Kampor field show the creation of a similar communication corridor in this part of the island (Figure 6). Here we can propose the connection of a private estate (that of the Beabii) (lastly in Jurković, Turković 2019) to the urban centre, passing by some of its satellite sites, or even another private estate, that of the Remii, possibly located in the proximity of the monastery in Kampor (Konestra, Cambi 2021).

(possibly all the way to the church of St. Barbatus on the easternmost tip of the island). Finally, to the furthest north of the island, on the Lopar peninsula, a similar situation of Roman sites located on the outskirts of a field, signal the location of possible communication corridors (Figure 7), further marked by the east-west road known from the FC. Although the persistence of corridors of movements, and on some occasions also of place, allow reconstructing communications and connectivity on the island, there are certain areas where data signals lacks, as in the case of two other sites on Lopar–the large Roman rural settlements at Podšilo bay and Zidine located in the north-east and north-west part of the peninsula (Figure 7). Several reasons might be proposed, such as lack of Medieval evidence, and major natural changes to the landscape, but also a later formation of the postRoman settlement (Budak 1987: 195), which might indicate Medieval depopulation of this area, and thus a remodelling of the corridors of movement.

An area equally rich in roman sites is that of Barbat, though here natural constraints are more felt, and settlement is restricted to the coastal zone. Nevertheless, the persistence of place and movement, are indicated by a document dated AD 1230 referring to various properties, including those of the Bishop of Rab, which are bordered by a road descending from the late antique fortress of St. Cosmas and Damian and onto a via pubblica14 (Mlacović 2005: 529; 2012: 172), which probably extended further east to reach the monastery of St. Stephen and the furthest fringes of Barbat 14  (...)Philippus iudex iuratus dixit se aufivisse a suo patre et a suis praedecessoribus, quod a via qua venitur a monte sancti Damiani usque ad viam publicam et inde eundo versus austrum usque ad vineam sancti Petri et vineam Cotopagne totum fuit de quarto;(...) (CD, III, 330).

35

Ana Konestra et al.

Figure 6: Pathways and archaeological evidence in Kampor field (ArchaeoTopoRab GIS; author: A. Konestra)

Figure 7: Pathways and archaeological evidence in Lopar: 1-area of Lopar field; 2-paths from the FC in relation to Podšilo and Zidine sites (ArchaeoTopoRab GIS; author: A. Konestra)

island by Kamenjak ridge. The modern state road that leads to this northwestern part of the island is the only paved road that leads to Lopar. This roadway follows a steep terrain contour above the northeast coastline of Supetarska draga so as to turn towards Lopar field through a small glen of a seasonal stream. This pathway

The earliest corridors of movement In light of connections, the Lopar peninsula is interesting from several points of view. Foremost its connection to the island’s mainland. Lopar peninsula appears to be physically fenced out from the rest of the 36

An Island of Connectedness

Figure 8: Fruga plateau: road segment of the FC’s “Strada per Loparo”, remains of Early Modern road and archaeological landscape of the area (ArchaeoTopoRab GIS; author: A. Konestra)

was certainly not the optimal route to reach the Lopar peninsula in the past, since for this modern roadway serious interventions in the landscape were to be done so it would serve its purpose adequately. As hinted by the early Modern road drawn on the FC (“Strada per Loparo”),15 and the actual evidence of a segment of it (Figure 8), the natural communication between Lopar and the rest of the island was well established elsewhere. A pathway across the flat area of the Fruga plateau and the hilly landscape of the western part of Kamenjak ridge was the main route to the Lopar peninsula at the beginning of the 19th century. However, there are strong indications of much earlier use of this corridor, since a rich prehistoric landscape at the northwest range of Kamenjak allows us to propose that the area functioned as a (terrestrial) junction from the earliest of times. Here a set of different types of prehistoric archaeological finds support this assumption. The earliest are the ones of scattered small finds, flint flakes and lames (blades) together with pottery fragments dating to a wide range from the Copper Age successively to the Iron Age, found in the vicinity of Karasić pound, in the southwest of Fruga plateau. But more tangible

are the ones of imposing, manmade prehistoric features, a set of Bronze and Iron Age hillforts and tumuli ascertained on the northeast slopes of the ridge (Konestra et al. 2020). While up until today’s date no direct evidence of prehistoric ridgeway was identified, the course of movement and main conjunction points on this part of Kamenjak can be easily presumed, mainly due to abundant archaeological remains (see Batović 1985; Glogović 1989; Brusić 1990; Blečić Kavur 2014; Blečić Kavur 2021), some of which rather clearly impose a communication function (Čače 1981; Bell, Lock 2000), and their proximity to supposed pathway, but also due to a terrain configuration, which does not allow a lot of astray. Rab and the world – inter-regional maritime communication The indented coasts of Rab offer plenty of suitable bays for harbours and landings, ensuring easy communication within the region and further afield. Maritime communication was of particular importance as the Commune of Rab had numerous properties off the island, that is on the mainland (Jablanac), and in the NW part of the island of Pag (Lun, Novalja, Caska), thus

This path is also mentioned by O. Badurina in the Chronicle of Kampor (I, 54) written in the 1930s.

15 

37

Ana Konestra et al.

Figure 9: Val Martina-Artić, Barbat: remains of Roman/late antique docking facilities and architecture (author: A. Konestra)

being not only a means to reach foreign lands but also, in a way, an extension of internal routes. Numerous sources testify to the interconnectedness of Rab with both nearby ports, mainly Senj on the mainland, but also distant trans-Adriatic networks (e.g., Hrabak 1991) throughout the Middle Ages. While the town’s port is often mentioned in the sources, and several bays along the island are mentioned as possible harbours too (e.g., Kampor bay and Lopar bay), no data is available on the actual port infrastructure.16

connectivity interacted with maritime one, possibly allowing interchangeable transport possibilities depending on seasons, weather and so forth. The very location of some of the protohistoric settlements identified on Rab speaks of their maritime vocation. There are several sea-oriented positions on the island, for which it can be assumed with relative certainty to have been occupied in the last centuries BC. The apparent maritime habituation can be read from their positions, these sites are all located on smaller peninsulas or capes above the sea. In East Adriatic, it is exactly the occupation of peninsulas that attributes to prosperous settlement, often a long-lasting one (Parica 2021: 184). When it comes to the island of Rab, one of those sites is foremost the old town of Rab. The continuity of settling the peninsula, which became the only urban centre on the island, has been ongoing at least since the last centuries of the old era (Suić 2003; Glavičić 2009). The city of Rab gained a leading role in the Iron Age, and consequently arguably later, due to its advantageous position-located in the midpoint of the island, this urban centre has developed into a kind of communication hub, both land and sea one (Tomičić 1988; Suić 2003; Konestra et al. 2020). Although only preliminary, underwater archaeological research in front of the town of Rab, corroborated active,

Similarly, only one landing dated to Roman times has been ascertained on the island (Lipovac Vrkljan et al. 2016) (Figure 9), connected to one of the coastal settlements (tentatively maritime villas) located along the shores of Barbat (Artić-Valmartina site). It is, of course, very probable that all coastal settlements had some sort of landing, used primarily to transport resources (e.g., as testified by allochthonous construction materials) though evidence has so far eluded us. In fact, it might be supposed that terrestrial 16  (See for e.g., Margetić 2001: IV, 28: Mlacović 2012: 56; Coronelli 1703: 295-300). Equally, in the Statute of Rab, several chapters deal with fines related to navigation, boats etc.

38

An Island of Connectedness

Figure 10: Proposed reconstruction of the main corridors of movement on the island (author: A. Konestra)

Iron Age settlement with a much shorter lifespan was established. Ongoing targeted research on Kaštelina at cape Stolac show that this hillfort was occupied in the 5th and 4th centuries BC. The position of this promontory inevitably ensures visual control over potential land resources, but even more pronouncedly here, a sea route communication. Earlier surface finds, like fragments of pottery of conceivable South Italian provenance, now corroborated by contextualised, excavated imported Hellenistic tableware and household ware testify to both cultural contacts and consequent local transformations of the obviously sea-oriented population that occupied this cape (Androić Gračanin et al. in press). Furthermore, from

prehistoric, maritime communication. At the shoal that spreads from the small islet of Sveti Juraj to the cape of north Padova II bay, a strip that virtually closes the bay and today’s port of Rab, a connecting Bronze age embankment structure can be assumed, overlayed by a modern breakwater (Parica 2021: 95). Although the character of these embankments cannot be fully determined unilaterally, it is possible to presume its military-trade, or even refugee role, which in this case testifies to active communication with non-islanders (Parica 2021: 157-169). As opposed to the continuously populated urban centre of the town of Rab, on the Lopar peninsula, the Late 39

Ana Konestra et al. Kaštelina it is easy to monitor two important local sea routes, one going through the Rab channel, and the other through the Velebit channel (Androić Gračanin et al. 2020: 451). Attesting to the importance of these routes are two shipwrecks, northwest and southeast of the island, one from the 3rd century BC and another from the 2nd-1st century BC (Dautova-Ruševljan 1975; Miholjek 2007; Glavaš et al. 2020). The setting of the hillfort at cape Stolac may have served many functions, including acting as a visual reference for the sea, as a form of sea beacon, as well as a node in the seaborne communication networks, or simply as a landmark that confirms possession of nearby land (see Čučković 2017; Čače 1981; Androić Gračanin et al. in press).

the whole Adriatic Sea routes network, in a constant negation of its supposed isolation. Earlier corridors of movement can so far be recognised in the area of the Fruga plateau, connecting the Lopar pre-/protohistoric settlements to the favourable area of Fruga, but with all probability with Lopar peninsula and, further SE, the rest of the island. Here, the role of the landscape in the patterns of movement is even more evident, as numerous ponds (e.g., Glavaš, Glavičić 2019: 126), flatlands, and areas suitable for pasture determined the directionality of movement from and to the hillfort settlements, with further data being provided by sepulchral features. Seaborn routes, just as in later periods, played a strong role in connectivity, probably at both the island and regional levels.

Concluding remarks Reconstructing long-term corridors of movement or persistence of the actual infrastructure has to consider not only the geomorphology of territory, but also the subsequent modes of organisation of the landscape (Patitucci, Uggeri 2007: 333), and ultimately changing routine practices (Mlekuž 2014: 7-8). By analysing available data on the actual road network provided by historic cartography and written sources, combined with archaeological topography, the persistence of movement through the landscape is inferred and a route network is proposed (Figure 10).

Acknowledgements The work presented in this article was supported by grant ID: 478202 NO. 2020/37/B/HS3/02458 from the Polish National Science Center (Narodowe Centrum Nauki). Sources CD I = Codex diplomaticus Regni Croatiae, Dalmatiae et Slavoniae, Vol. I, diplomata annorum 743.-1100, M. Kostrenčić, J. Stipišić, M. Šamšalović (eds), Academia Scientiarum et Artium Slavorum Meridionalium, Zagreb 1967.

In the context of Rab island, continuity from the Roman period to the Middle Ages can be observed as the rural landscape was used in virtually the same manner - as a locus of agricultural exploitation (in the broadest sense) and sparse habitation. An exception is, however, the possible shift of settlement nuclei to higher ground, which determined the formation of north-south corridors in an as-of-yet hard-to-pinpoint, but certainly post-Roman, time frame. Nevertheless, the persistence of place is determined by the numerous large and small ecclesiastic structures dotting the landscape and often marking the previous habitation, which allows us to identify those corridors with most probably the longest use. All these corridors converge on the town of Rab, as a place of dwelling, power and exchange, and where the land and maritime networks of movement join through the supposedly largest harbour of the island. Nevertheless, just as the land movement was disjoined from the main corridors to serve daily practices (e.g., Mlekuž 2014: 17-18), so was the seaborne one, connecting the most remote areas and serving private needs of communication. These seaborne pathways allowed connection with the main terrestrial communications, further linking the island to the coastal hinterland. Thus, not only the island, but each of its settlements intertwined with the wider area of the island’s satellite possessions, the region, and

Bibliography Aldred, O. 2021. The Archaeology of Movement. Boca Raton-London-New York: CRC Press, Taylor & Francis Group. Anderson, A. 2004. Islands of ambivalence, in S. M. Fitzpatrick (ed.), Voyages of Discovery: The Archaeology of Islands: 251-273. Praeger: West-port. Androić Gračanin, P., F. Welc, A. Konestra, and B. Nowacki 2020. An integrated geoarchaeological approach to late Iron Age settlement at Kaštelina hillfort (Lopar, Island of Rab, Croatia) using Amplitude Data Comparison (ADC) method and trial excavation, Polish Archaeology in the Mediterranean 29: 447-468. Androić Gračanin, P., M. Ugarković, F. Welc and A. Konestra in press. Preliminary insights on contextualized evidence of foreign fine ware appropriation from Kaštelina hillfort on Rab island (Kvarner gulf, NE Adriatic, Croatia), in Proceedings of the 5th Conference of the International Association for Research on Pottery of the Hellenistic Period “From East to West and Back Again: Societies, Economics and Ceramics in the Hellenistic World”, June 2021, 22nd–25th Seville, Spain (in press).

40

An Island of Connectedness Batović, Š. 1985. Rekognosciranje otoka Raba u godini 1984. Obavijesti hrvatskoga arheološkog društva XVII/1: 13-15. Bell, T. and G. Lock 2000. Topographic and cultural influences on walking the Ridgeway, in G. Lock (ed.), Beyond the map: archaeology and spatial technologies: 85-100. Ios Press. Blečić Kavur, M. 2014. Na razmeđu svjetova za prijelaza milenija: Kasno brončano doba na Kvarneru (Katalozi i monografije Arheološkog muzeja u Zagrebu, Svezak XI). Zagreb: Arheološki muzej u Zagrebu. Blečić Kavur, M. 2021. Kvarner in the Early Iron Age of the northern Adriatic. Arheoloski vestnik 72: 533-550. Boomert, A. and A. J. Bright 2007. Island Archaeology: In Search of a New Horizon. Island Studies Journal 2/1: 3-26. Brogiolo, G.P., A. Chavarria Arnau, F. Giacomello, M. Jurković and G. Bilogrivić 2018. The Late Antique Church of Saint Lawrence, Banjol (Island of Rab, Croatia) ‒ Results of the First Two Archaeological Campaigns (2015-2016). Hortus Artium Medievalium 23/2: 666-673. Brusić, V. 1926. Otok Rab. Kampor: Franjevački kamporski samostan sv. Eufemije. Brusić, Z. 1990. Otok Rab. Rekognosciranje gradina. Arheološki pregled 1988/29: 229-233. Budak, N. 1987. Neki elementi demografskoekonomskog razvoja i prostorne organizacije otoka Raba od VI. do kraja XIII. st., in A. Mohorovičić (ed.) Rapski zbornik. Zbornik radova sa znanstvenog skupa o otoku Rabu (od 25. do 27. listopada 1984. godine): 193198. Zagreb: JAZU/Skupština općine Rab. Coronelli, V. 1703. Biblioteca universale sacroprofana,antico-moderna, Tomo IV. Venezia: Antonio Tivani. Čače, S. 1981. Naselje i nekropola u prostoru zajednice. Dometi 5: 35-40. Čučković, Z. 2017. Claiming the sea: Bronze Age fortified sites of the north-eastern Adriatic Sea (Cres and Lošinj islands, Croatia). World Archaeology 49(4): 526546. Dall’Aglio, P. L. 1997. Il “diluvium di Paolo Diacono” e le modificazioni ambientali tardoantiche: un problema di metodo. Ocnus 5: 97-104. Dall’Aglio, P. L. and C. Franceschelli 2012. Bonifche e regimazioni idrauliche tra pianifcazione e gestione del territorio, in M. Martins, I. V. de Freitas and M. I. del Val Valdivieso (eds) Caminhos da Água. Paisagens e usos na longa duração: 77-104. Braga: CITCEM - Centro de Investigação Transdisciplinar «Cultura, Espaço e Memória». Dautova-Ruševljan, V. 1975. Zaštitno istraživanje podvodnog nalaza amfora na otoku Rabu (Rescue excavations of underwater amphora site on the island of Rab). Diadora 8: 89-102. Degrassi, D. 2002. Frontiere, confini e interazioni transconfinarie nel Medioevo: alcuni esempi

nell’area nordorientale d’Italia. Archivio Storico Italiano 160-2: 195-220. Filipčić, A. 1998. Klimatska regionalizacija Hrvatske po W. Köppenu za standardno razdoblje 1961.-1990. u odnosu na razdoblje 1931.-1960. Acta Geographica Croatica 33: 7-14. Fitzpatrick, S. M. 2007. Archaeology’s Contribution to Island Studies. Island Studies Journal 2/1: 77-100. Francesconi, G. 2011. Scrivere il contado. I linguaggi della costruzione territoriale cittadina nell’Italia centrale. Mélanges de l’École française de Rome - Moyen Âge 123-2: 499-529. Gibson, E. 2007. The Archaeology of Movement in a Mediterranean Landscape. Journal of Mediterranean Archaeology 20/1: 61-87. Glavaš, V. and M. Glavičić 2019. Transformation of prehistoric to historic landscape: the example of civitas Lopsica. Senjski zbornik 46: 119-136. Glavaš, V., A. Konestra, and A. Tonc 2020. Wine consumption in the Kvarner and sub-Velebit area (NE Adriatic) in the last centuries BCE: evidence from amphorae finds, in M. Ugarković, I. Kamenjarin (eds) 3rd IARPotHP Kaštela 2017. Exploring the neighbourhood: The role of ceramics in understanding place in the Hellenistic world: 271-286. Vienna: Phoibos Verlag. Glavičić, M. 2009. Magistrati i uglednici Arbe, Senjski zbornik 36: 57-66. Horović Vuković, B. 2012. Otok Rab u Franciskanskom katastru, in J. Andrić and R. Lončarić (eds) Rapski zbornik II: 219-234. Rab: Matica hrvatska. Glogović, D. 1989. Prilozi poznavanju željeznog doba na sjevernom Jadranu, Hrvatsko primorje i Kvarnerski otoci. Zagreb: Jugoslavenska akademija znanosti i umjetnosti, Zavod za arheologiju. Hrabak, B. 1991. Regionalna i međunarodna trgovina mlečana i dubrovčana drvetom iz Senja (XIV-XVIII stoljeće). Radovi Zavoda za hrvatsku povijest Filozofskog fakulteta u Zagrebu 24: 57-107. Jarak, M. 2017. Studije o kasnoantičkoj i ranosrednjovjekovnoj skulpturi otoka Raba. Zagreb: FF Press. Jurković, M. 2019. The transformation of the Adriatic Islands from Antiquity to the Early Middle Ages, in Cau Ontiveros M. A. Cau Ontiveros and C. Mas Florit (eds) Change & Resilience. The Occupation of Mediterranean Islands in Late Antiquit (Joukowsky Institute Publication 9): 111-138. Oxford: Oxbow Books. Jurković, M. and T. Turković 2019. Kaštelina - vila obitelji Bebija. Ars Adriatica 9: 7-20. Konestra, A. and N. Cambi 2021. Osteotheca from Rab: new insights on the sepulchral monument of Remmia Maxima held in the Franciscan monastery at Kampor on the island of Rab. Vjesnik za arheologiju i historiju dalmatinsku 113/2: 351-375. Konestra, A., F. Welc, P. Androić Gračanin, K. Rabiega, B. Nowacki, and A. Kukela 2020. Tipologija i organizacija 41

Ana Konestra et al. otočnih naselja Raba kroz dijakronijski pristup. Prvi podaci multidisciplinarnih istraživanja. Annales Instituti Archaeologici 16: 229-244. Konestra, A., P. Androić Gračanin and F. Welc 2021. Burialscapes of Rab Island (North-East Adriatic): The Role of Sepulchral Evidence in the Reconstruction of Roman and Late Antique Rural Settlement Pattern. Annales, Series Historia et Sociologia 31/3: 391-412. Lipovac Vrkljan, G. and A. Konestra 2015. Projekt Arheološka topograija otoka Raba- rezultati terenskog pregleda na području grada Raba u 2014. godini i izložba Arheološka topograija: putovanje kroz prošlost Lopara. Annales Instituti Archaeologici XI: 128-132. Lipovac Vrkljan, G., A. Konestra and I. Radić Rossi 2016. Rezultati aktivnosti projekta „Arheološka topografja otoka Raba“ u 2015. g.: terenski pregledi, obrada arheološke građe, popularizacija znanosti. Annales Instituti Archaeologici XII: 201-205. Lipovac Vrkljan, G., A. Konestra and N. Šegvić 2017. Felix Arba-reconstructing urban and rural economic capacities through GIS, in A. Plosnić Škarić (ed.) Mapping urban changes: 314-335. Zagreb: Institute of Art History. Lončar, N. 2012. Geomorfološka regionalizacija i kategorizacije reljefa otoka Raba, in J. Andrić and R. Lončarić (eds) Rapski zbornik II: 187-196. Rab: Matica hrvatska. Margetić, L. 2001. Statuto d’Arbe. Atti, Centro di Ricerche Storiche Rovigno XXX: 9-221. Miholjek, I. 2007. Podmorsko arheološko nalazište kod rta Sorinj na otoku Rabu. Histria antiqua 15: 377-384. Mlacović, D. 1995. The world of Dominchiellus Meçigna (The Testators, the Commune and Family on the Island of Rab in the 2nd Half of the 14th Century). Medium Aevum Quotidianum 35: 85-106. Mlacović, D. 2005. Supetarska draga na Rabu v srednjem veku, in N. Budak (ed.) Raukarov zbornik. Zbornik u čast Tomislava Raukara: 513‒535. Zagreb: Filozofski fakultet Sveučilišta u Zagrebu, Odsjek za povijest, FF-press. Mlacović, D. 2012. Plemstvo i otok. Pad i uspon rapskog plemstva. Zagreb: Leykam international. Mlekuž, D. 2014. Exploring the topography of movement, in S. Polla and P. Verhagen (eds) Computational

Approaches to the Study of Movement in Archaeology. Theory, Practice and Interpretation of Factors and Effects of Long Term Landscape Formation and Transformation (Topoi. Berlin Studies of the Ancient World, Vol. 23): 5-21. Berlin/Boston: De Gruyter. Parica, M. 2021. Prapovijesne maritimne konstrukcije Dalmacije i Kvarnera. Zadar: Sveučilište u Zadru. Patitucci, S. and G. Uggeri 2007. La viabilità nell’Italia medievale: questioni di metodo, in E. Catani and G. Paci (eds) La Salaria in età tardoantica e altomedievale: 323-358. Roma: L’Erma di Bretschneider. Renfrew, C. 2004. Islands out of time? Towards an analytical framework, in S. M. Fitzpatrick (ed.) Voyages of Discovery: The Archaeology of Islands: 275229. Praeger: West-port. Snead, J. E., C. L. Erickson and A.J. Darling 2009. Making Human Space: The Archaeology of Trails, Paths, and Roads, in J. E. Snead, C. L., Erickson and A. J. Darling (eds) Landscapes of Movement. Trails, Paths, and Roads in Anthropological Perspective: 1‒19. Philadelphia: University of Pennsylvania Museum of Archaeology and Anthropology. Starac, R. 2017. Lokalitet: Supetarska Draga-crkva sv. Petra. Hrvatski arheološki godišnjak 13/2016: 531-534. Suić, M. 2003. Antički grad na istočnom Jadranu, Zagreb: Golden marketing-Institut za arheologiju. Tomičić, Ž. 1988. Arheološka svjedočanstva o ranobizantskom vojnom graditeljstvu na sjeverojadranskim otocima. Prilozi Instituta za arheologiju u Zagrebu 5: 29-53. van Lanen, R. J., M. C. Kosian, B. J. Groenewoudt, T. Spek, and E. Jansma 2015. Best travel options: Modelling Roman and early-medieval routes in the Netherlands using a multi-proxy approach. Journal of Archaeological Science: Reports 3: 144-159. van Lanen, R. J., Groenewoudt, B. J., Spek, T., Jansma, E. 2018. Route persistence. Modelling and quantifying historical route-network stability from the Roman period to early-modern times (AD 100-1600): a case study from the Netherlands. Archaeological and Anthropological Sciences 10: 1037-1052. Vuletić, N. 2020. Bilješke iz rapske povijesne toponimije: toponimi bez suvremenih potvrda. Folia Onomastica Croatica 29: 237-259.

42

The Roadside Station at Žuta Lokva Ivana Ožanić Roguljić Abstract: The site Žuta Lokva is situated in the hinterland of the Roman province Dalmatia, in its northern part. Today it is situated in the region called Lika. A facility built there had two phases. The building of the 1st phase (dimensions 23 x 16 m) was built around the beginning of the 1st century and probably lasted until the third quarter of the 1st century AD. The building of a second phase was bigger (dimensions 30 x 20 m), and according to the finds, it was in function until the middle of the 2nd century AD. This paper will try to find analogies to the building. KEYWORDS: ROMAN DALMATIA, ŽUTA LOKVA, ROMAN ROADSIDE STATION

Introduction The site Žuta Lokva is located a few meters from the crossroad Senj - Otočac - Brinje. The existence was known from the end of the 19th and the beginning of the 20th century (Brunšmid 1898: 188, 189). It was refound during the building of the road in 1996 (Vekić 1996: 38) when several campaigns of rescue excavations took place until 2003 (Vekić 1996; Kolak 2012; Ožanić Roguljić, Kolak 2019). In Roman times this site was situated alongside a road mentioned in the Itinerarium Antonini Augusti Ad Aquileia per Liburniam Sisciam (Senia – Avendon – Arrupio – Bibium - Romula – Quadrata - Siscia). The section of the road from Senia (Senj) to Žuta Lokva goes over the Vratnik pass, the shortest natural connection from the sea to the continent’s interior (Glavaš 2010). Senia was the most important port in this part of the Adriatic coast in Roman times, which is confirmed by the fact that in the second half of the 2nd century AD, it was a collecting point for the publicum portorium Illyrici. The imperial customs office was passing through and the goods which entered the Senia were distributed to the hinterland (Glavičić 1994: 54; Glavaš 2010). The pottery evidence from Žuta Lokva showed a place equipped with numerous elaborate Roman drinking vessels fit to serve travellers on the road rather than the local population. The following groups of pottery are present on the site: Italian terra sigillata and thinwalled pottery, glazed ware, common ware jugs and bowls, products from the workshop of Sextus Metillius Maximus in Crikvenica, coarse ware, Pannonian blackslipped ware, African red slipped ware A, Eastern sigillata B. Amphora: Dressel 20, Dressel 2-4, Adriatic flat bottom amphorae, Dressel 6B, several types of small amphorae (Ožanić Roguljić, Kolak 2019; Ožanić Roguljić 2019; Ožanić Roguljić, Mantovani 2020), all in a very bad state of preservation. Because of the pottery, we can state that it was built soon after the Roman

Figure 1: Senia, Žuta Lokva, Kompolje, Siscia (Basemaps: WMS_Esri Shaded Relief http://server. arcgisonline.com/arcgis/rest/services/World_Shaded_ Relief/MapServer/tile/{z}/{y}/{x}, EU-DEM produced using Copernicus data and information, author: J. Drpić)

pacification. Therefore, it brought new typical Roman habits in serving and drinking to a facility probably used as a roadside station. Numismatic finds from excavation confirm the site’s longer duration than is visible from the pottery finds. They show the range

Life Along Communication Routes (Archaeopress 2023): 43–46

Ivana Ožanić Roguljić from quadrans Augustus 15-10 BC (RIC 227) to minimus of the 5th century AD. Altogether, 66 pieces were found. The majority of numismatic finds (51 coins) are dated from Augustus (15-10 BC) to Faustina I. (As, 138-141), and the rest (15 coins) are dated from Severus Alexander (D, 222-235) to minimus of 5th century AD. (Kolak, Šeparović 2020: 34-36).

century AD and probably lasted until the third quarter of the 1st century AD. The building of the second phase was bigger (dimensions 30 x 20 m), and according to the pottery finds, it was in function until the middle of the 2nd century AD (Figure 3). The burnt layer is found during the excavations, from which we can assume that, at the moment, invisible, a phase dated to the 4th and 5th century AD probably existed. Both phases reveal a similar building. The building of the first phase is not preserved as a whole but it seems that it might have four rooms at the front and the courtyard. The second phase had two main rooms at the front and a large courtyard at the back. Rooms of both phases were divided by a small, roofed hall. The rear part of the buildings was a courtyard that might have worked as a half-opened courtyard or stable (dimensions 13 x 14 m -first phase; 17 x 18 m - second phase). In the hall, two layers of the gravelled surface were discovered most probably connected to the second phase. 

The building The site revealed a building with two phases (Figure 2). The walls of the building are preserved in their foundation only, and the layers of soil above it were very thin. Moreover, the site is situated on a slope, and the stratigraphy is disturbed after a landslide. Most of the archaeological material was excavated in mixed stratigraphic units in the western part of the area. But even with this archaeologically unfortunate site situation, interesting finds were revealed. The two phases are difficult to date since the layers are filled with pottery finds from the beginning of the 1st century to the middle of the 2nd century AD. We can only assume that the building of the 1st phase (dimensions 23 x 16 m) was built around the beginning of the 1st

The complex’s ground plan is similar to the sites found at several points on the German frontiers and the Pannonian limes. In her work, Szilvia Bíró suggests that most of the similar structures from limes can be

Figure 2: Plan of Žuta Lokva (authors: J. Osterman and K.Turkalj)

44

The Roadside Station at Žuta Lokva

Figure 3: Timetable of Žuta Lokva (author: I. Ožanić Roguljić)

interpreted as road stations or horse stations, and all of them show signs of military presence (e.g., stamped bricks or belt fittings) (Bíró 2017: 180-185). These stations were part of the official postal system, the cursus publicus, used by the official administration (Bender 1975: 19-21). Some of them could also play the role of beneficiary stations. The Latin term for such a building is also problematic, as it is rarely possible to identify a station with a name from an itinerary or an inscription (Bender 1975: 22; Seitz 2005: 420-425), and also the translations of these categories may vary over time and space. The general terms, statio, mutatio, or praetorium, could perhaps be applied (e.g., Gabler 1991: 76-78). Buildings of smaller dimensions are also documented as a roadside station in Roman Gaul (Mauné 2016: 233, fig. 10). The dimensions of the Gaulish buildings vary from 10 x 24 m (Chameleux) to 15 x 30 m (Soulmaltre). Their common characteristic is a courtyard. Their plan differs by having four rooms at the front or even rooms in the courtyard. The functions of the room can be debated. The Gönyű example has a kitchen in one of the rooms. Some of the Gaulish examples have the adjacent thermae (Corne du Bois, Soulmaltre), (Mauné 2016: 234, T.1). Most of the Gaulish road stations are dated in the 1st century AD while limes road stations are dated in the 2nd to 3rd century AD.

Conclusion We suggest that the structure excavated at Žuta Lokva is also one of these road stations or horse-changing stations. According to Suetonius, Augustus established a transportation and information system in the Roman Empire. (SUET., Aug. 49.3) and he founded what is now commonly known as the cursus publicus. Augustus began to create a network of roadside stations (mutationes), built in regular intervals (modicis intervallis) along most major roads (viae militares  (Lemcke 2013: 12). But  terminus technicus-cursus publicus  appears in the sources after the year AD 300 and should be avoided for the early period in which Žuta Lokva operated.  We cannot rule out the possibility that it was a beneficiary station. At Japod territory, two beneficiary stations were confirmed: in  Avendo  (Crkvina near Kompolje), (Patsch 1900: 91) and Retinium (Golubić south of Bihać), and the third was probably in  Metulum  (Viničica near Josipdol), (Glavaš 2016: 13). But much more evidence is needed that this station was a beneficiary station.  Pottery and numismatic evidence show the uninterrupted timeline from the last decades of the 1st century BC to the middle of the 2nd century AD. This means that construction started very soon after Augustus occupied this area. The station most likely became a part of the road system that Augustus himself instituted. 45

Ivana Ožanić Roguljić to the Beginning of the 16th Century. Gospić: Muzej Like. Lemcke, L. 2013. Imperial Transportation and Communication from the Third to the Late Fourth Century, The Golden Age of the cursus publicus. Unpublished PhD dissertation, The University of Waterloo. Mauné, S. 2016. La voie entre Cessero (Saint-Thibéry, Hérault) et Segodunum (Rodez, Aveyron): Lieux d’étape de la moyenne vallée de l’Hérault. Gallia 73(1): 219-251. Ožanić Roguljić, I. 2019. Thinwalled pottery from Žuta lokva, Roman Dalmatia (Croatia), in H. Dolenz and K. Strobel (ed.) Chronologie und vergleichende Chronologien zum Ausgang der Römischen Republik und zur Frühen Kaiserzeit. Klagenfurt am Wörthersee: 427435. Klagenfurt: Landes museum Kärtnen. Ožanić Roguljić, I. and T. Kolak 2019. Excavations of the Roman Sites in Lika (Croatia): Žuta Lokva, Lički Ribnik, in M. Janežić (ed.), New Discoveries between the Alps and the Black Sea Results from the Roman Sites in the Period between 2005 and 2015 Proceedings of the 1st International Archaeological Conference: 115-125. Ptuj: Zavod za varstvo kulturne dediščine Slovenije. Ožanić Roguljić, I. and V. Mantovani 2020. Sarius cups from Žuta Lokva (roman Dalmatia- Croatia)preliminary report. Rei cretariae Romanae fautorum acta 46: 311-316. Bonn: RCRF. Patsch, C. 1900. Lika u rimsko doba. Gospić. Seitz, G. 2005. Strassenstation. Infrastruktur für die Weltherrschaft, in Imperium Romanum. Roms Provinzen an Neckar: 420-425. Rhein und Donau. Esslingen am Neckar: Archäologisches Landesmuseum BadenWürttemberg. Vekić, A. 1996. Zaštitno arheološko istraživanje u Žutoj Lokvi. Senjski zbornik 23: 35-40.

Sources Gaius Suetonius Tranquillus Hosu, S. (translator) 1979. Gaj Svetonije Trankvil, Dvanaest rimskih careva. Zagreb: Naprijed. Bibliography Bender, H. 1975. Römische Straßen und Straßenstationen. Kleine Schriften zur Kenntnis der römischen Besetzungsgeschichte Südwestdeutschlands 13. Stuttgart: Gesellschaft für Vor- und Frühgeschichte in Württemberg und Hohenzollern. Bíró, S. 2017. A Roman road station on the Pannonian Limes, in Nick Hodgson, in P. Bidwell and J. Schachtmann, Judith (eds) Proceedings of the XXI International Congress of Roman Frontier Studies (Limes Congress) held at Newcastle upon Tyne in August 2009: 180-185. Oxford: Archaeopress. Brunšmid, J. 1898. Arheološke bilješke iz Dalmacije i Panonije, Vjesnik Hrvatskoga arheološkoga društva, n.s. III: 150-205. Zagreb: Arheološki muzej, Gabler, D. 1991. Römische Strassenstation in der Gemarkung von Sárvár. Communicationes Archaeologicae Hungariae: 39-84. Glavaš, I. 2016. Konzularni beneficijariji u rimskoj provinciji Dalmaciji. Zagreb: Ministarstvo kulture. Glavaš, V. 2010. Prometno i strateško značenje prijevoja Vratnik u antici. Senjski zbornik 37: 5-8. Glavičić, M. 1994. Značenje Senije tijekom antike. Senjski zbornik 21: 41-58. Senj: Gradski muzej. Kolak, T. 2012. Antička postaja u Žutoj lokvi. Modruški zbornik 6: 17–28. Josipdol: Katedra Čakavskog sabora Modruš. Kolak, T. and T. Šeparović 2020. Numizmatička topografija Like Numismatic Topography of Lika from Earliest Times

46

Petrijanec-Aqua Viva Lovorka Štimac-Dedić Abstract: The paper brings an overview of the last 50 years of archaeological research in the village of Petrijanec as well as stray finds retrieved from the area in the course of the last two centuries. Since the paper is based on the poster Petrijanec-Aqua Viva, it provides only summary data, while a detailed analysis of artefacts in light of recent research compared to older research will be presented in another paper. The assumption that Aqua Viva, a Roman road station, was located in the area of presentday Petrijanec is based on numerous Roman finds and structures recorded in the entire area of the village and additionally supported by the data from ancient itineraries. Although recent research has supplemented the data on the appearance of the settlement, typological analogies for architectural features of the station cannot be provided without systematic archaeological investigations, which can be carried out in several locations in Petrijanec. KEYWORDS: PETRIJANEC, AQUA VIVA, STRAY FINDS, ARCHAEOLOGICAL RESEARCH

The village of Petrijanec is located along the ancient Poetovio-Mursa main road. The assumption that a Roman road station (mansio Aqua Viva) was located here is based on the finds of archaeological structures and numerous artefacts, including rich hoards. Ancient itineraries1 state different distances between the settlements of Poetovio and Aqua Viva, which has resulted in different opinions on the location of the latter. Most authors, however, consider Aqua Viva to be located in Petrijanec (Abramić 1925: 8; Pahič 1965: 315; Vikić, Gorenc 1969: 5-6; Fulir 1970: 8; Šimek 2002: 3031; Gračanin 2010: 20) and the Roman layer recorded in archaeological investigations throughout the entire area of the village (Šarić, Tomičić 1969; Šarić 1978: 177195; Štimac-Dedić 2016: 2-13; 2020: 2-12; 2021: 2-17) additionally supports this opinion, although the name of the Roman settlement has not yet been confirmed by epigraphic evidence. In the Roman period, the settlement of Petrijanec was located along the Poetovio-Mursa main road, which followed the course of the Drava River. The location of the settlement by the Drava River is a vital hydrographic factor that certainly influenced the appearance of the settlement, just like it has influenced all natural and social phenomena along the river to this day. How exactly the inhabitants of Petrijanec in antiquity adapted to this erratic river, described by Pliny the Elder (Domić Kunić 2012: 31) as quite a forceful current, and how they contributed to the changes in the river dynamics is hard to imagine without systematic archaeological research, as well as geological and hydrographic analyses that will hopefully be included The settlement of Aqua Viva is mentioned in the Antonine Itinerary (Itinerarium provinciarum Antonini Augusti et Hierosolymitanum), the Jerusalem (Bordeaux) Itinerary (Itinerarium Burdigalense), the Peutinger Map (Tabula Peutingeriana) and the Ravenna Cosmography (Ravennatis Anonymi Cosmographia et Guidonis Geographica).

1 

in the planned interdisciplinary research aimed at paleoenvironmental reconstruction.2 Until relatively recently, the Drava River flooded this area all the way to the highest point of the village, south of the church, while today its flow is regulated and significantly altered by the construction of hydroelectric power plants, which consequently slowed down the river, causing the disappearance of historic waterways. The Drava riverbed has always been exposed to changes, as torrential water washed away the old beds and at the same time created new ones, which caused the shifting of the banks, meandering and formation of wetlands. The evidence of meandering of the river in the past was recorded along the northern edge of Petrijanec, marked on the geological map of Croatia as an oxbow lake (Mioč, Marković 1998a: Sheet Čakovec L33-57; 1998b: 40). Due to prolonged periods of water retention, the sediment in the old riverbeds is silty and clayey; on a map from the 18th century, the oxbow lake is recorded as a pond next to the village where water lingered during rain and snow melting seasons (Valentić, Horbec, Jukić 2005: 129-130, Map: Section 3). The first archaeological finds from Petrijanec were recorded at the end of the 18th century and continued to emerge sporadically over the next two centuries. Most of them are dated to the Roman period, which supports the fact that this was an inhabited area in the period of the Roman Empire, while more recent finds, dating from the 16th to the 18th century, have been recorded only in recent research (Štimac-Dedić 2016: 2-13; 2020: 2-12; 2021: 2-17). In 1805 a rich hoard of gold jewellery and coins was found (Ljubić 1880: 111-113; Mirnik 2008: 433-452), after For an example of paleoenvironmental reconstruction, see: Van Dinter 2013.

2 

Life Along Communication Routes (Archaeopress 2023): 47–54

Lovorka Štimac-Dedić which stray finds were recorded in the entire area of the village, especially after the Second World War due to increased construction, but unfortunately many of them were lost, deliberately destroyed or built into the foundations of houses etc. (Klemenc 1936: 26; Kugly 1965; Tomičić, Šarić 1969: 3-5; Fulir 1969: 404; Vikić, Gorenc 1969: 5-6). The finds include sarcophagi, stone building elements, inscriptions, remains of Roman architecture and wells, a stone female head with traces of paint in the eyes etc. Some of the stray finds were collected by Stjepan Vuković before 1949, but regrettably without the data on the exact locations of the findings (Figures 1 and 2), while some (Figures 3 and 4) were sufficiently recorded (Vuković 1960). The recorded stray finds provided a basis for planning the first archaeological investigations in Petrijanec, conducted in 1969 and 1970 (Tomičić, Šarić 1969; Šarić 1977: 60) when a total of 19 trenches were opened (Figures 5 and 6). In 1969, a total surface of 230 m² was investigated in the area along and north of Vladimir Nazor Street, south and southeast of the village centre and on Saint Peter’s Square. The goal was to determine the boundaries of the archaeological zone to include

Figure 2: Cup (GMV AO 1342), Petrijanec, 1st- 2nd century AD (source: Varaždin City Museum photo-documentation, Archeological Department)

Figure 3: Sarcophagus (GMV 66343) at the time of its discovery in 1960 (source: Varaždin City Museum photodocumentation, Archeological Department)

them in future construction projects in Petrijanec. These investigations confirmed the existence of a Roman settlement and a segment of the Roman main road south of the village and in the village, southwest of the former alcohol factory (Tomičić, Šarić 1969; Šarić 1978: 178, 191). Due to the limitations of trial

Figure 1: Double handle jug (GMV AO 1341), Petrijanec, 1st - 4th century AD (source: Varaždin City Museum photodocumentation, Archeological Department)

48

Petrijanec-Aqua Viva

Figure 5: Bath area discovered in Petrijanec in 1969, foto orientated I-W direction (source: Varaždin City Museum photo-documentation, Archeological Department)

Figure 4: Altar dedicated to Jupiter Culminalis (GMV 66347), the second half of 2nd or first half of 3rd century AD (source: Varaždin City Museum photo-documentation, Archeological Department)

trenching in an inhabited area, no conclusive data on the appearance of the settlement was obtained, but the retrieved artefacts enabled the Roman settlement to be approximately dated from the 1st to the beginning of the 4th century AD (Tomičić, Šarić 1969; Šarić 1978: 177-195). When we sum up the last 50 years since the first archaeological investigations were conducted, when there was a plan to include archaeology in the urban development of Petrijanec, it is hard to conceive that archaeological finds and structures were never taken into account in planning construction projects. Consequently, most of the construction works in the period after the aforementioned research was carried out without archaeological supervision, even the construction and restoration of larger objects. From 1969 to 1995, when the first large-scale archaeological supervision was carried out during the installation of a gas pipeline in the village, archaeologists were only sporadically present at construction sites, and smallscale supervision was conducted in 1977 (Šimek 1977).

Figure 6: Bath area discovered in Petrijanec in 1969, foto orientated W-E direction (source: Varaždin City Museum photo-documentation, Archeological Department)

After a hoard of dishware (Figure 7) and coins were found in 2005 (Šiša-Vivek, Leleković, Kalafatić 2005: 231-345), attention was once more focused on this site, but after that, in practice and the field, there have been no concrete improvements in its protection, which is especially visible in more recent investigations, which often reveal layers destroyed by relatively recent infrastructure works. 49

Lovorka Štimac-Dedić the centre and south of the village that had not been previously investigated by archaeologists (Skelac 2017: 3-31). The survey was conducted on St. Peter’s Square, in the rectory courtyard and on the piece of land south of the village where large amounts of Roman artefacts have been recorded on the surface along the route of the Roman road. It revealed a potential archaeological zone in the rectory courtyard, west of the Church of St. Peter and St. Paul. An outline of a building with approximate dimensions of 22 x 10.5 m was recorded, to which several other buildings are attached. To confirm the results of the geophysical survey, the Varaždin City Museum conducted trial excavations in the rectory courtyard in the autumn of 2019. Two trenches were excavated (S I and S II) with the goal of uncovering building elements visible on geophysical images. In S I, Roman brick and pottery fragments mixed with postmedieval material such as stove tiles, pottery and iron nails were found, starting from the humus layer at five cm to a depth of 45 cm, indicating that this segment of the yard was filled in. In the Roman layer, starting at a depth of 45 cm, a layer with a collapsed Roman building was uncovered, containing a large amount of plaster, hollow brick and roof tile, under which a Roman floor was found at a depth of 90 cm (Figure 8). Waterproof plaster and a wall were recorded along the western edge of the trench extension, while the northern part of the trench was not lowered to the level of the Roman floor in that campaign, so the situation here remains unclear. Due to unfavourable financial conditions, only one more trench was excavated, northwest of S I. In S II, located on lower ground than S I, a layer with collapsed Roman architecture was uncovered under a thin layer with a very small amount of post-medieval material. Another floor, slightly cruder than the one in S I, was found under a large amount of roof tile and hollow brick mixed with fragments of pottery and glass. During the excavation of S I and S II, trenches for the rectory sewage were also dug. One contained a 30 cm thick opus spicatum wall built without a binder, while another massive, sturdy wall was found west of it, only partially visible in the trench. These trial excavations revealed that a larger heated building was most likely located in the area, possibly connected to the buildings discovered some hundred meters to the southwest in 1969, which points to a larger building complex or a group of buildings recorded in other stations as well.3 This research barely scratched the surface of archaeological layers in the so far unexplored archaeological zone, which points to the need for systematic archaeological research to preserve the remaining archaeology in Petrijanec from continuous destruction that has been going on for years.

Figure 7: Silver plate with a figural representation of goddess Epona (GMV 55665), Petrijanec, 3rd century AD (source: Varaždin City Museum photo-documentation, Archeological Department)

Recent research Archaeological investigations conducted in the last few years have supplemented the picture of the settlement, but the available data are still insufficient for concrete interpretations of its appearance. However, these investigations have defined new, archaeologically potentially interesting zones that will be included in future systematic excavations. The rescue excavation conducted in the yard of a family house in 2016, west of the bath complex discovered in 1969, attested to the continuity of settlement typical of archaeological sites in populated areas (Štimac-Dedić 2016: 2-13). After the demolition of the house, built in the 1930s, layers from the 16th to 18th century were discovered that overlaid Roman construction debris dated with a coin of the emperor Aurelian (AD 274-275). Along the southern edge of the trench, under the 18thcentury wall perpendicular to it, remains of a compact Roman wall were found, preserved 3.5 m in length, most likely connected to the aforementioned architecture discovered in 1969. A review of documentation from 1995 (Bilić 1995: 3) revealed that the zone with architectural remains probably continued further to the west of this location, but, unfortunately, only piles of stones from destroyed structures retrieved from the gas pipe trench, along with Roman artefacts, testified to this. In 2017 a geophysical survey was carried out on a surface of 12,000 m², aimed at examining the areas in

Mischka 2011: 8-13; Fingerlin 1976: 27–31; Liesen, Reuter 2009: 279312; Kovács 1999: T. LXV.

3 

50

Petrijanec-Aqua Viva

Figure 8: Roman floor found during the excavation in the rectory courtyard in 2019 (source: Varaždin City Museum photo-documentation, Archeological Department)

buried by the Roman wall in a pot covered with another pot, dated with a denar of Matthias II (AD 1608-1619).

The last archaeological supervision in Petrijanec was carried out in January 2021 during the reconstruction of the road and the construction of a storm drainage system on St. Peter’s Square. Due to the structures uncovered in the southern part of the square, the reconstruction works were temporarily suspended in that location. Two control trenches were excavated; the first, covering the area of 41 m², was to determine the layout of the discovered walls, while the second, a smaller trench was excavated along the southern edge of the street on the square that connects with Vladimir Nazor Street. In SI, walls were uncovered that formed a corner of a building that probably had a single-pitched roof covered with tegulae and imbrices, which were found only on its southern side. Remains of a gravel foundation of a wall or a fence were found by the eastern part of the building wall, which formed a quadrangular unit with the preserved walls (Figure 9). On the south side of the building, under Roman rubble with a large amount of roof tile, a screed was uncovered, part of which was also recorded along the western part of the object under a layer of burning. The layers along the western part of the building were destroyed during sewer works in 2005 when there was no archaeological supervision. In addition to the Roman material, there were several post-medieval finds, such as a hedgehog

The trench S II was smaller in size (four meters square) and yielded a structure that was probably a wall or a collapsed building, with a darker layer of burning in which only Roman material was found, including an antoninianus of Emperor Tacitus (AD 275-276) and two centenionales of Constans I (AD 337-350). With difficult winter conditions and, again, short deadlines that dictated the pace and conclusion of the research, it was not feasible to complete the excavation. For this reason, the interpretation of the site to the extent it deserves is not possible. The quadrangular object found in S I was only partially uncovered, which is also attested by the remains of a floor west and south of it, whereas in S II the remains of a structure, probably destroyed in a fire, are just barely visible. Both trenches were located near the trenches investigated in 1969 and they probably constitute a whole, only for now it is not possible to determine what kind. Conclusion The current state of research in Petrijanec does not allow determining the type of settlement and drawing 51

Lovorka Štimac-Dedić

Figure 9: Remains of a building found on St Peter’s Square in 2021 (source: Varaždin City Museum photo-documentation, Archeological Department)

analogies with other stations in the Roman Empire. For this reason, the connection of parts of the newly discovered buildings with the bath area discovered in 1969 also remains unclear. It is possible that this is a group of different buildings or a single complex that consisted of rooms of different purposes required for the functioning of the travel station.

today’s Vladimir Nazor Street along the southwest edge of the Fanon factory. The assumption that the silver, twisted-wire bracelet that was found in 1995 (Figure 10) and handed over by the workers together with a number of Roman coins to the archaeologist in charge several days later (Bilić 1995: 3-4), came from a destroyed hoard found three houses down from the silver hoard found in 2005, indicates that, together with the hoard from 1805, all three hoards were located very close to each other.

For now, it is only possible to define three separate segments of the settlement, which consist of excavated building elements that point only to the possible size of the settlement. Based on the research conducted so far, in which architectural remains were recorded in the centre of the village along Vladimir Nazor Street, south of the village along the Roman road, on St. Peter’s Square and in the rectory courtyard, combined with earlier records of stray finds and architectural remains and bricks, it can be concluded that the zone with these remains is located in the centre of the village, covering the area of nearly four hectares.

Every new research bears witness to the devastation of this archaeological site, the layers of which are slowly disappearing with the construction of modern infrastructure and the development of the village. The more recent research presented here supplements the data on the settlement, but still not enough to draw concrete conclusions. The data collected so far point to the need for systematic archaeological research which would enable detailed interpretation and supplement the existing data in combination with revision of older research and previously collected material. It is, therefore, necessary in addition to systematic

Zones with the finds of sarcophagi are documented in three locations, along the route of the Roman road south and southeast of the centre of the village and north of 52

Petrijanec-Aqua Viva Ljubić, Š. 1880. Arkeologičke crtice iz mog putovanja po njekojih predjelih Podravine i Zagorja, god. 1879. Vjesnik Arheološkog muzeja u Zagrebu Vol. 2, No. 1: 110-120. Mioč, P. and S. Marković 1998a. Osnovna geološka karta Republike Hrvatske i Republike Slovenije 1:100.000, List Čakovec L33-57. Zagreb: Institut za geološka istraživanja, Ljubljana: Inštitut za geologijo geotehniko in geofiziko. Mioč, P. and S. Marković 1998b. Osnovna geološka karta Republike Hrvatske i Republike Slovenije 1:100.000, Tumač za list Čakovec: 33-57. Zagreb: Institut za geološka istraživanja, Ljubljana: Inštitut za geologijo geotehniko in geofiziko. Mirnik, I. 2008. Maksimijanov aureus iz skupnog nalaza u Petrijancu 1808. god. Archaeologia Adriatica Vol. 2, No. 2: 432-452. Mischka, C. 2011. Die neu entdeckte mansio in der Aussensiedlung des Kastells Pfunz. Der Limes 5/1: 8-13. Pahič, S. 1965. K poteku rimskih cest med Ptujem in Središćem. Arheološki vestnik 15/16: 283-320. Skelac, G. 2017. Izvješće o geofizičkim istraživanjima na lokaciji Petrijanec, trg sv. Petra i područje župe Petrijanec. Documentation of the Archaeological Department of the Varaždin City Museum (unpublished field report): 2-31. Šarić, I. 1978. Antičko naselje u Petrijancu, in Ž. Rapanić (ed.) Novija arheološka istraživanja u sjeverozapadnoj Hrvatskoj: 177-195. Zagreb: Izdanja HAD 2. Šimek, M. 1977. Petrijanec: Izvještaj s terena. Documentation of the Archaeological Department of the Varaždin City Museum (unpublished field report). Šimek, M. 2004. Naseljavanje u prapovijesti i ranoj povijesti, in Život uz Dravu nekad i danas (exhibition catalog): 23-34. Varaždin: Gradski muzej Varaždin. Šiša-Vivek, M., T. Leleković and H. Kalafatić 2005. Ostava rimskog novca i srebrnog posuđa iz Petrijanca (Hoard of Roman Coins and Silver Dishware from Petrijanec). Opuscula Archaeologica 29: 231-345. Štimac-Dedić, L. 2016. Petrijanec-izvještaj o zaštitnom arheološkom istraživanju. Documentation of the Archaeological Department of the Varaždin City Museum (unpublished field report). Štimac-Dedić, L. 2020. Petrijanec-župni dvor, izvještaj o probnom arheološkom istraživanju. Documentation of the Archaeological Department of the Varaždin City Museum (unpublished field report). Štimac-Dedić, L. 2021. Petrijanec-Trg svetog Petra, izvještaj o arheološkom nadzoru, Documentation of the Archaeological Department of the Varaždin City Museum (unpublished field report). Tomičić, Ž. and I. Šarić 1969. Istraživanje arheološke podloge Petrijanca. Documentation of the Archaeological Department of the Varaždin City Museum (unpublished field report).

Figure 10: Silver twisted-wire bracelet (GMV 69758), Petrijanec, 4th century AD (source: Varaždin City Museum photo-documentation, Archeological Department)

archaeological research, to conduct geological and hydrographic analyses and, if possible, determine the type of the settlement and find an analogy among other researched stations in the territory of the Roman Empire in order to better understand life by the Roman station in the Varaždin area. Bibliography Abramić, M. 1925. Poetovio Führer durch die Denkmäler der römischen Stadt. Wien: Österreichiches Archäologisches Institut. Bilić, M. 1995. Izvještaj o arheološkom nadzoru u Petrijancu, travanj-lipanj 1995. god. Documentation of the Archaeological Department of the Varaždin City Museum (unpublished field report). Domić Kunić, A. 2012. Literary Sources before the Marcomannic Wars, in B. Migotti (ed.) The Archaeology of Roman Southern Pannonia: The state of research and selected problems in Croatian part of the Roman province of Pannonia (British Archaeological Reports International Series 2393): 29-69. Oxford: Archaeopress. Fingerlin, G. 1976. Zwei römische Straßenstationen im südlichen Oberrheintal. Denkmalpflege in BadenWürttemberg 5/1: 27–31. Fulir, M. 1969. Topografska istraživanja rimskih cesta na varaždinskom i međimurskom području (19601967). Rasprave SAZU 6: 363-431. Fulir, M. 1970. Osvrt na položaj današnjeg Varaždina u nizu nekadanjih rimskih postaja. Godišnjak Gradskog muzeja Varaždin 4: 5-20. Gračanin, H. 2010. Rimske prometnice i komunikacije u kasnoantičkoj južnoj Panoniji. Scrinia Slavonica 10: 9-69. Kovács, P. 1999. Vicus és a castellum kapcsolata az alsópannoniai limes mentén, Piliscsaba. Kugly, I. 1965. Urbanistička analiza naselja Petrijanec, prijedlozi za urbanističko rješenje, zaštitna arheološka zona, popis zaštićenih objekata. Documentation of the Archaeological Department of the Varaždin City Museum (unpublished report). Liesen, B. and M. Reuter 2009. Der Keller der mansio in Insula 38 und das Ende der Colonia Ulpia Traiana. Xantener Berichte 15: 279-312. 53

Lovorka Štimac-Dedić Valentić, M., I. Horbec and I. Jukić 2005. Hrvatska na tajnim zemljovidima 18. i 19. st. (Varaždinska županija). Zagreb: Hrvatski institut za povijest. Van Dinter, M. 2013. The Roman Limes in the Netherlands: how a delta landscape determined the location of the military structures, Netherlands Journal of Geosciences– Geologie en Mijnbouw 92/1: 11-32.

Vikić, B. And M. Gorenc 1969. Prilog istraživanju antiknih naselja i putova u SZ Hrvatskoj. Zagreb: Regionalni zavod za zaštitu spomenika kulture u Zagrebu. Vuković, S. 1960. Izvještaj o nalazu rimskog sarkofaga u Petrijancu kod Varaždina. Documentation of the Archaeological Department of the Varaždin City Museum (unpublished field report).

54

The Unknown about the Known: the Možđenec-Sudovec Roman Road Jere Drpić Abstract: Based on archaeological research of the two road segments with stone paving along the route of the asphalt road over the Kalnik mountain pass, as well as data collected from historical, and cartographic sources about this road, this paper aims to update existing knowledge, analyze current assumptions, and discuss the Roman origin of the remains of this stone-paved road in the “heart” of the Kalnik mountains. KEYWORDS: KALNIK HILL, ROMAN ROAD, STONE-PAVED ROAD, HILL PASS, MOŽĐENEC, SUDOVEC

Introduction If you head south from the direction of Varaždinske Toplice (Aquae Iasae) of the Bednja valley and want to reach the Kalnik foothills, in the Križevci region, by the shortest route, you must first negotiate the Kalnik mountains on your way. Today the pass, through which this can be done, is also the only preserved road pass over this mountain. It is situated in the western part of the mountain, on a narrow (five km) route, which connects what is today the village Možđenec in the Bednja valley, on the north side of Kalnik, and the village Sudovec, which is located on the southern slopes of this mountain. For a long time, this road was one of the few unpaved Croatian state roads, which changed during the reconstruction in the 2000s. It was noticed that this road has an older substructure built of regularly paved stone, under a layer of gravel, which has been excellently preserved in parts. This is especially true for one bend, which was then separated from the planned new route, to be saved for future valorization. By tracing a new bend, the older one was cut into two segments: a southern and a northern one (Figure 1). The southern segment was then partially presented, while the northern one was covered with an additional layer of gravel. The position where this bend is located in older cartographic sources is called Vražja Rit, while the toponym that denotes the wider area is known as Zverinjak, but also the toponym Mostišće is used. Although archaeological excavations on this road have never been carried out, due to its construction with stone paving, which invokes the Roman way of building, the road has so far been roughly dated to that period. Due to the lack of archaeological findings that would concretize such dating and enable a more accurate interpretation of the remains of this road, cleaning and minor archaeological research was undertaken at the end of 2020 as part of the LRR1 1 

Project named; Life on the Roman road: communications, trade and

project of the Institute of Archeology in Zagreb. As part of the mentioned activities, this site was observed in a broader historical-geographical context. Based on the data that we collected while conducting archaeological research on the remains of this road and comparing the obtained results with historical-cartographic data, it was possible to partially shed light on some earlier assumptions related to this road (Drpić 2021). Kalnik and its roads in older sources Hilly and mountainous areas, even today, with all available technology, are still inaccessible and consequently less explored, especially in the archaeological sense. The Kalnik mountains are no exception. The fact is that we have almost no archaeological data on the interior of the Kalnik mountains, which is especially true of the Roman period. Even the ancient sources themselves when mentioning the hilly areas in the whole area of the ​​ Empire are not, so to speak, as descriptive. For example, the only ancient mention of a mountainous area in the whole of today’s central Croatia refers to the Claudius mountains (Gaius Velleius Paterculus and Pliny the Elder), without specifying the exact area where it would be located. One assumption is that the source could be referring to the Kalnik mountains (Marković 2004: 202-203), which has not yet been concretely proven. Consequently, in ancient sources, we do not have data on possible Roman communications that would pass through this mountainous area.2 However, although it is not mentioned in ancient sources, one medieval source brings us information about a road that could be of ancient origin. The road is called via magna, which identities in Croatia from 1st-8th CE. LRR (UIP-05-2017-9768), funded by the Croatian Science Foundation. Results of this excavation were published in a form of a short report (Drpić 2021). The project leader is Ivana Ožanić Roguljić, PhD. 2  The basic hypothesis about the tracing of Roman roads is that they avoid larger relief obstacles and pass-through hilly areas only in cases where the obstacle cannot be adequately bypassed (Gallo 2006).

Life Along Communication Routes (Archaeopress 2023): 55–61

Jere Drpić

Figure 1: Position of the Možđenic-Sudovec road, arrows indicating the northern and southern segments of the rest of the stone-paved road in the area (author: J. Drpić, base maps: WMS www.geoportal.dgu.hr_DOF 2021, TK 25)

connected Varaždin and Križevci (Smičiklas 1905: 246; Bedenko 1975: 2, 4; Demo 1982: 76, 77) via Kozji Hrbat (a pass in the interior of Kalnik). It is also the only road known from the source that is mentioned as a pass over Kalnik.

(Valentić et al. 2004, 106, 107, 110)3 from Možđenac to Sudovac. Although it is shown in a stylized, more or less regular line, and we do not have information about the exact route, this presentation in the 18th century testifies to the existence of a hill pass of some kind in that area. With more detailed cadastral surveys in the 19th century (Slukan 1997), this area was for the first time both geodetically accurate and cadastral plotted. By comparing today’s route of this road, and the one before its reconstruction in the 2000s, there is an almost complete overlap with the cadastral and military maps from the 19th century. What still does not match, i.e., “jumps out,” of the route is just a bend at the site of Zverinjak. Namely, on the cartographic representations of the 19th century, this bend was drawn around the hill itself (elevation 293) nearby the group of houses, and it

Cartographic data Graphic representations and descriptions of roads in older cartographic sources can be useful source of data for older roads and their possible routes. So far, the earliest such depictions of the Kalnik mountains date from the 18th-19th century, i.e., the time of the first precise cadastral surveys (Slukan 1997) and known military valorization (Valentić et al. 2004). In the mentioned military-cartographic sources, when talking about the roads on Kalnik (which was back then called Gora), it is stated that the roads across Gora are mostly just bad forest roads, which can be driven by oxcarts, and it is noted that they are always usable for riding. It is also stated that these roads usually have muddy soil, a very narrow rut, which is washed out and corrupted by holes and can only be passed through bad, narrow bullock carts, in dry weather. No route is specifically mentioned by name as a pass over the mountain

“Road leading from Križevci to Grana (Novi Marof); it has completely solid ground, it is wide and usable for heavy vehicles, but it has to cross some steep hills, and in these places, it becomes difficult to pass. The slopes above the village of Možđenec on both sides are intersected by deep ravines. They cannot be crossed on foot or on horseback, and they are also very densely forested. For the village of Leskovec; “The roads that lead through the high and unnamed mountains are just forest roads, usable for exporting wood in dry weather with bad wooden cars. They end up on hiking trails.” (Valentić et al. 2004: 106, 107)

3 

56

The Unknown about the Known

Figure 2: Comparison of the route of the bend at the site Zverinjak (Vražja rid); On the left is a view of the route on a map from the 19th century, on the right is a view of a modern route of bends and the remains of a stone-paved road (edited by J. Drpić, layout: https://www.arcanum.com/en/)

Figure 3: Orthophoto image of the road on the site Zverinjak, overlayered with a digital cadastral plan (author: J. Drpić, base map: www.geoportal.dgu.hr, DOF)

was circumnavigated in an SE direction (Figure 2). It is still in function today as a side, unpaved road, located along the NE side of the mentioned small hill. It is up to five meters wide, with no visible structural elements of the road. On the other hand, the bend with the “S” line and the remains of stone paving is visible only on the 20th-century maps, not before (Figure 2 and 3).

The indicative data on the Možđenec-Sudovec road, which we collected during a conversation with local people from the surrounding settlements, are also interesting.4 Namely, we find out that the people of this area do not have a “living” idea or tradition about the older road on the route in question, i.e., about its Roman origin. Most of the older respondents pointed out that according to their memory, at least as far as the preserved pavement is concerned, it is a construction of a bend at the beginning of the 20th century, i.e., renovations. Although we have studied historical sources from that

time, so far we have not found specific historical data for these allegations.5 As for previous research on this road, as we mentioned earlier, it has never been archaeologically researched, and the first to write more specifically about it, i.e., the southern segment of the subject bend, is Ivan Milotić (2010), who dates it to the Roman period due to constructional characteristics. Although this road was a gravel road until the 2000s, none of the earlier researchers of Roman roads in the northwestern part of Croatia proposed or identified this road as a possible Roman road.6 In the context of constructional characteristics of the subject road and possible analogies, it is also important to point out that a similarly constructed road segment was found in the archaeological context during the rescue excavation in Slovenia nearby the village of Ivančna Gorica, dated from the 18th to the 20th century (Jerin, Nabergoj 2003: 196-198). Further, in most cases, Roman roads, even the major ones, were not paved or just paved in the section where they were passing through the settlements (Oleson 2008: 567).

Residents of Sudovec and Možđenec stopped on the main road during archaeological excavations and on that occasion shared the local tradition about this road orally, which of course should be taken with a dose of caution.

5  Of course, these data should be taken with a dose of caution, but the fact is that the tradition of the Roman road does not exist. 6  For example, Miroslav Fulir, wrote about Roman roads in the vicinity of Varaždinske Toplice. (Fulir 1969)

Local tradition and history of research

4 

57

Jere Drpić

Figure 4: The southern segment of the road, presented in situ (Milotić 2010: 56)

impossible to obtain a cross-section. During the cleaning, we did not come across any small archaeological finds, including those that would confirm the dating of this road to the Roman period. Also, by more detailed cleaning, no linear depressions were noticed, which would manifest themselves in the form of grooves on the stone surface of the track and would consequently indicate more intensive and long-lasting wear of the stone surface with the use of carts. This is precisely the case with most Roman roads with well-preserved stone paving in the Roman Empire (Gallo 2006).

Archaeological research To verify the previously presented findings and further support them with field research, as part of the project “LRR” mentioned before, minor archaeological excavations have been undertaken on isolated segments (north and southern) of the bend of the stone-paved road (Drpić 2021).7 The southern segment This segment contains well-preserved stone paving of the road in the length of 50 m and an average width of six meters (Figure 4). Although presented, this segment has been partially covered with a layer of humus, moss, and fine gravel over time due to the action of anthropogenic and anthropomorphic factors. For these reasons, it was not possible to spot the curbs of the road on the entire segment, and surface cleaning was approached to gain a better insight. Once cleared, the segment was also documented. During the excavation of the water drainage canal of the modern road, this segment was cut across the entire width in the north. This position seemed to be an ideal opportunity to see the intersection of the older road with minimal intervention. However, after cleaning the moss, it was noticed that the slope of the modern canal in that part was walled up with stones and mortar, which made it

The northern segment Unlike the southern one, which is relatively well presented and visible, the stone paving of the northern segment could not be seen on the surface of the segment, i.e., only the 55 m long and ten m wide gravel embankment was visible along the northern edge of today’s modern road. The research was approached from the westernmost part to the east because we estimated that according to the direction of the southern segment, there could be the best-preserved part of the northern, which was confirmed by research, in a shallow 80 m² trench. Removing the surface layer of a thin layer of humus, we came across a dark grey layer of gravel that in some places contained larger stones. The entire surface of the road was found to be covered with such a layer, and in parts, this layer was so thin, especially in the western part, that individual paving stones of the

Until these excavations, no archaeological excavations have been carried out on the road in focus.

7 

58

The Unknown about the Known

Figure 5: Photo mosaic of the northern segment of the road (author: M. Fileš, edited by: J. Drpić)

road could be seen on the surface. Gradual removal of the gravel layer, from west to east, began to fully reveal the stone paving of the road, made on one side of larger flattened pebbles with joints filled with clay soil and smaller pebbles, without binder plaster. Continuation of the excavation to the east revealed a larger segment of well-preserved paving of the road, while along the southern edge of the probe, the edge of the pavement began to appear, properly strung in a light arc with an east-west direction. To determine the existence of a stone curb on the opposite side of the road, i.e., on the northern edge, the trench was extended to the north. In that part there was a thicker layer of the previously mentioned gravel, backfilling the road. Removing this thicker layer of gravel, no clearly defined pavement edge was found in the southern part, but at the level of the stone paving a layer with brown earth and smaller pebbles appeared, with irregular extension in the eastwest direction along the northern paving edge. It has also been observed that sometimes this layer fills parts of the road where stone paving is missing. Expanding the excavation further to the north revealed that this layer extends below the layer with smaller stones, which extends linearly along the northern edge of the probe. This layer of stone is similar to the stone paving of the road but is visibly worse built. To determine the width of the extension and thus enable the dating of the northern drainage ditch, which was partially visible

before the research, three trenches were made along the northern edge of the probe (Figure 6). A drainage canal was identified, buried in a geological sandy layer, and recently buried with gravel (rubber and plastic found in the layer, at the bottom of the canal). We did not discover any small archaeological finds during the excavation. The northern edges of the road were not found during this excavation, but the edge consisted of a layer of smaller irregularly placed stones, which goes to the edge of the drainage channel. This same stratigraphic relationship was confirmed in the other two trenches excavated along the northern edge of the probe trench. To explore the southern drainage canal, which was seen in the northeast part of the road in the form of a slight line depression, a small probe trench was dug. A cross-section of the drainage canal revealed a clay fill, and at the bottom of the canal, there was a layer of black earth without stones and any small finds. Along the edge of the southern canal, it was also found that the curb line was lost, and an irregular road edge of larger stones appeared. During the excavation, there were also larger quantities of packed embankment in this part, which could indicate that due to the destruction of the pavement edge, it was repaired by an embankment. The stone paving of the road in this northern segment was discovered in a total length of 20 m in a stretch of EW direction with an average width of three metres, and if we include the stone extension 59

Jere Drpić

Figure 6: The northern segment of the road during archaeological excavations (photo: J. Drpić)

of the road to the north then the width goes up to five meters (Drpić 2021), (Figure 5).

sources, as well as local traditions. According to archaeological research and other evidence for now we cannot say with certainty that the segments of the subject paved road were once part of a Roman road. Namely, although we found well-preserved remains of the stone-paved surface of the road, so far, no small archaeological finds have been discovered that would undoubtedly date these segments to that period. What we have learned by studying cartographic sources is the fact that on the first precise cadastral survey from the 19th century, the “S’’ bend on which the stone base is well preserved was not shown while the rest of the route is precisely drawn. Also, from conversations with locals, we noticed that there is no local tradition of a Roman road on this route. On the contrary, they told a story that the “S” bend with a stone road was (re)built in the early 20th century, which would correspond with cartographic sources. From this arises the real possibility that the bend with the remains of stone paving was built or rebuilt in the late 19th and early 20th century. That possibility is additionally supported by the fact that a segment of the road with similar construction dating to the Modern period was found

Conclusion The Možđenec-Sudovec road is located in the westernmost part of Kalnik which is also one of the lowest parts of that hill, so it is not surprising that there could be continuity of using the road pass, especially since today there is the only asphalt road crossing Kalnik. When we talk about this possibility, we primarily mean the remains of a layer of the stone road on that route, which was found at the bend at elevation 293, in the area called Zverinjak. Based on the characteristics of construction it was dated back to Roman times, which is certainly an indicator of a potential Roman road, but not a final confirmation of it. As mentioned before most of the Roman roads were not paved or just paved in the areas passing settlements, i.e, there is a possibility that the subject road was constructed or reconstructed in another period of time. Therefore, to collect as much data as possible, we undertook archaeological research and included data from historical and cartographic 60

The Unknown about the Known in Slovenia. Today the location of the road between Možđenec and Sudovec is known for landslides; so it is not hard to imagine that at one point, the older bend (or even the whole section of this mountain road) could have been rebuilt with a hard, stone surface that would allow vehicles to overcome the climb more easily than the previous one, which could be muddy and difficult to pass for most of the year, according to the mentioned historical-military sources, that in the 18th and 19th century describe in more detail the roads in this mountain, which were important at the time for military-strategic reasons. There is no mention in these sources of any stone-built road on the route from Možđenac to Sudovac. The fact is that it simply had to be noticed by the military experts as a wider road with a hard surface, which allows a quick and efficient crossing of large and important relief obstacles such as Kalnik. Also, what is interesting about this road being a possible Roman road is that it’s not mentioned by any researcher who dealt with the topic of Roman roads, for example, by Miroslav Fulir (1969) who dealt closely with this area. Due to the previously stated facts, for now, we cannot confirm that this road is undoubtedly a Roman road because the data we obtained from the research suggests that it is a possible recent construction, at least as far as the bend in the area of Zverinjak.

antičko doba s osvrtom na numizmatičke nalaze, Križevački zbornik II: 75-92. Jerin, B. and T. Nabergoj 2003. Mrzlo Polje near Ivančna Gorica, in B. Djurić and D. Prešeren (eds), The Earth Beneath Your Feet: Archaeology on the Motorway in Slovenia: Guide to Sites: 196-198. Ljubljana: Institute for the Protection of the Cultural Heritage of Slovenia. Drpić, J. 2021. Arheološko istraživanje segmenta ceste Možđenec-Sudovec na položaju Zverinjak. Annales Instituti Archaeologici, XVII(1): 255-257. Fulir, M. 1969. Topografska istraživanja rimskih cesta na varaždinskom i međimurskom području. Razprave Slovenske akademije znanosti in umetnosti VI: 391-401. Gallo, I. M. 2006. Vias romanas: ingenieria y tecnica constructive. Madrid. Marković, M. 2004. Antička naselja i grčko-rimska zemljopisna imena na tlu današnje Hrvatske, Zagreb: Naklada Jasenski i Turk. Milotić, I. 2010. Rimska cestovna baština na tlu Hrvatske. Zagreb: Hrvatsko društvo za ceste-VIA VITA. Oleson, J. P. (ed.) 2008. The Oxford Handbook of Engineering and Technology in the Classical World. Oxford: Oxford University Press. Slukan, M. 1997. Katastarska dokumentacija Arhiva mapa u Hrvatskom državnom arhivu. Arhivski vjesnik 40: 139-155. Smičiklas, T. (ed.) 1905. Diplomatički Zbornik Kraljevine Hrvatske, Dalmacije i Slavonije. sv. 3. Zagreb: Hrvatska akademija znanosti i umjetnosti. Valentić, M., I. Horbec and I. Jukić (eds) 2004. Hrvatska na tajnim zemljovidima 18. i 19. st., Križevačka županija, svezak 9. Zagreb: Hrvatski institut za povijest.

Bibliography Bedenko, V. 1975. Križevci  –  razvoj grada, Glasilo Arhitektonskog fakulteta Sveučilišta u Zagrebu 2/3. Zagreb. Demo, Ž. 1982. Prilog topografiji križevačke regije u

61

New Data, Old Town: The Case of Aquae Balissae Andrej Janeš, Marina Matković Vrban and Ivana Hirschler Marić Abstract: The knowledge about the existence of the Roman town Aquae Balissae or municipium Iasorum begins in the 18th century. From that time a variety of inscriptions accidentally found in the town’s urban core attested to the existence of a thriving community that flourished from the 1st century AD till the Late Antique period. These finds corroborated the idea that the Roman town of Aquae Balissae was situated in present-day Daruvar.

Modern-day research brought to light new structures on the site of Stari Slavik in Daruvar. The site is situated east of the presentday Special hospital for medical rehabilitation, on a hill known as Stari Slavik. In the present-day forest, excavations unearthed parts of the presumed Late Antique town fortifications or fort. The excavation yielded the remains of a wall and foundations 26.39 m long, with a northwest-southeast direction. On both ends, the wall was destroyed by later actions. On the eastern part of the wall, a semicircular construction has been discovered with a 1.5 m thick wall. Geomagnetic and LiDAR surveys corroborated older finds, indicating that the Roman town stretched on the southern slopes of the Stari Slavik hill. Surface features indicate the existence of underground structures of rectangular layouts on the highest ridge of the hill. The area of the Roman town that was not encompassed by the urban development of Daruvar comprises a total of 3000 m² with possible remains of fortifications and housing. KEYWORDS: AQUAE BALISSAE, DARUVAR, ROMAN TOWN, LANDSCAPE, URBANIZATION, LIDAR

Introduction The modern city of Daruvar is best known for its thermal water springs and spa complex, currently called Aquae Balissae. This name can be found in several written and epigraphic sources depicting Aquae Balissae as a Roman town established on the territory of the Iasi people (ʹres publica Iasorumʹ, ʹmunicipium Iasorumʹ). According to historical sources (written sources include Pliny the Elder’s Naturalis Historia and Ptolemy’s Geographia, the Iasi were a Pannonian Celtic tribe that lived in the area between Drava and Sava Rivers, or, more precisely, between Varaždinske Toplice (Aqua Iasae) in the west and Daruvar (Aquae Balissae) in the east. The southern border of their territory might have spanned as far as Pakrac and Lipik, encompassing the whole interfluve as well as the areas of Kalnik, Bilogora, and Moslavina, all rich in hot springs (Domić Kunić 2006: 76). Based on the available sources, it is assumed that the Iasi were under the political administration of the Taurisci (Benac 1987: 861). Their first contact with the Roman army might have been in the 2nd century BC when the army first began penetrating the territory of present-day Croatia. During the first half of the 1st century AD, the Roman army set up a military camp on the territory of present-day Daruvar, on the spot where the Iasi oppidum had been situated up until then. During the process of Romanization, the Iasi people gained the title of res publica Iasorum. During Hadrian’s rule, veterans, settlers, and other Roman citizens, together with the autochthonous population, were organized

in which municipium Iasorum included the territory between Sava and Drava Rivers and had its centre in Aquae Balissae (Schejbal 2004: 99). Plenty of chance finds discovered in Daruvar and its surroundings, and especially the epigraphic finds, testify to the development of the settlement (and the life in it). The settlement could have been divided into two sections based on the terrain configuration; a luxurious public area including the forum, Temple of Jupiter, Temple of Silvanus, thermae in the lowland area, and a residential complex on the hill (the position of the Celtic fort and settlement, Roman military camp, and subsequently the fort and refuge in Late Antiquity), (Schejbal 2004: 99). Research in the Daruvar area to date Professionals in the field were aware of Daruvar’s ancient history due to chance finds, primarily stone monuments discovered at different sites in and around the city of Daruvar, mostly in the 19th century. Friedrich Wilhelm von Taube was the first to describe Daruvar in his depiction of Slavonia and Syrmia in 1777/1778. Among other things, he mentioned multiple strong thermal water springs, ʹRoman baths with multiple artefactsʹ discovered, but also two stone monuments with inscriptions mentioning the Iasi (Taube 1778: 183184). Gjuro Szabo conducted the first field research in 1905-1907, (Figure 1), (Schejbal 1994a: 14) and published a comprehensive paper on Daruvar’s finds in 1934. The paper included a short history of Daruvar, research and finds so far together with Szabo’s observations

Life Along Communication Routes (Archaeopress 2023): 62–68

New Data, Old Town

Figure 1: G. Szabo’s sketch of the ʹRoman forestʹ in Daruvar, 1905 (source: Archive of the Archaeological Museum in Zagreb)

and thoughts (Szabo 1934: 79-98). Szabo listed the first known archaeological finds discovered by chance during various construction works, (Figure 2). The earliest find is a diatretum discovered in 1785 in Daruvar. It was subsequently taken to Vienna where it is still kept today. The next find was discovered in 1877; it included fragments of a bronze horse statue discovered in the courtyard of pharmacist Kušević (present-day King Tomislav Square) (Szabo 1934: 82; Schejbal 1994a: 12). The finds also included fragments of military votive altars dedicated to Jupiter Dolichenus, due to which Szabo concluded that this spot used to be a shrine (Szabo 1934: 82-83). T. Mommsen (CIL III 3998 and CIL III 3999) and J. Brunšmid (Schejbal 1994a: 11) were the first to publish the inscriptions on the altars. Szabo wrote that ʹa robust wall, partially doubleʹ was discovered in R. Kuneš’s land in 1907 (Szabo 1934: 81, Schejbal 1994a: 14). On that occasion, stone fragments with inscriptions mentioning Emperor Gordian and his wife Sabinia Tranquillina, erected by ʹres publica Iasorumʹ were discovered as well. Szabo also mentioned a stone monument discovered in 1920 in the area known as Kantari. It was a 2nd-century AD tombstone depicting a soldier in a Phrygian cap with an inscription mentioning municipium Iasorum. He also recorded fragments of a tombstone with an inscription stating that the monument was erected by a mother for her fallen sons, a sarcophagus discovered in 1842 in the nearby village of Bastaji, and stone slabs with inscriptions, one of which was built into Count Janković’s manor (Szabo 1934: 81-85).

Figure 2: Sketch of the presumed Roman ruins in Daruvar (drawing by G. Szabo, 1905; source: Archive of the Archaeological Museum in Zagreb)

Other artefacts were discovered by chance in 1952. On the site of the Daruvar brickyard, a section of a wall, a fragment of Icarus’ statue, and three Roman coins were found. The site was visited by Ivica Degmedžić and the wall was interpreted as a bulwark from a later 63

Andrej Janeš et al. period (Schejbal 1994a: 15). During construction works in the 1960s, several stone fragments were discovered near the bathing resort building; a fragment of a tomb stele, statue of Jupiter, two capitals with acanthus leaves, two pillars, five stone blocks, a portal, and four altars dedicated to Silvanus (three of them containing inscriptions, one a relief), (Schejbal 1994a: 17). Danica Pinterović (1975: 123-157) and Duje Rendić-Miočević (1980: 105-123) wrote about these altars (Schejbal 1994a: 18; Schejbal 2006: 76). In 1973, brickyard workers discovered foundations of a 30 x 9 m building and a shallow pit containing pottery and bone fragments, as well as two coins. One of them was Constantine’s, the other Constantius’. These finds were published by Mirko Bulat in Glasnik slavonskih muzeja (1973: 24-26; Schejbal 1994a: 15). In the same year, 1973, remains of foundations and walls were discovered during excavations in F. Sabljić’s house at the site of Stari Slavik. Similar finds were recorded at the site of Ciglana (Minichreiter 1973: 1).

At the end of 2012, in the Spa park area of Daruvarske Toplice, in the so-called Julius park, a test excavation was conducted at the positions determined by geophysical research. Three positions did not reveal traces of history before the 18th century, but in one, the presence of an older structure was presumed (KuntićMakvić, Olujić, Osterman 2013: 211-215). With the help of geophysical research, it was assumed that these were the remains of an around ten meters long structure (Mušič 2012: 12). Excavation revealed the remains of a semi-circular wall, a layer of mortar following Roman standards, and small finds most interesting of which are coins from the first half of the 4th century AD. The discovered remains are assumed to be part of sacral structures in the Middle Ages, with possible earlier Roman substructions (Kuntić-Makvić, Olujić, Osterman 2013: 213). Underground anomalies were noted in the northern section of the spa, west of the restaurant building. They make up parallel lines and were not drawn into any of the existing infrastructural plans; therefore, they are considered archaeological structures (Mušič 2012: 11).

Rescue excavations at the site of Stari Slavik in 1983 revealed the remains of Roman architecture, sections of two building complexes, and many movable finds (iron nails, bolts, knives, Roman glass, ceramic lamps, and 54 coins dating from Marcus Aurelius to Constantine; Minichreiter 1984b: 19, Minichreiter, Sakara 1984: 3).1 Moreover, pits with archaeological material and various brick and stone paving were discovered and interpreted as evidence of multiple Roman villas in the area dating to the 3rd/4th century AD (Minichreiter 1984b; Schejbal 1994a: 16).

The site of Stari Slavik, Jewish cemetery Research conducted north of the Jewish cemetery revealed remains of architecture and it was assumed they belonged to fortification remains recorded by G. Szabo. A wall in the E-W direction was discovered, 18 m long, and 1.5 m wide, with an added form in the shape of an apse (actually a semi-circular tower) to the northern side of its eastern section. In the western section of the wall, there is a smaller wall in the N-S direction. North of the big wall, a plaster-tied pebble layer was discovered, interpreted as a potential walking surface, part of a courtyard, for instance (Jakovljević 2018: 177-182). The probable remains of a rectangular tower, only the cut of which is preserved, were discovered in the northwestern section (Figure 3).

B. Schejbal presented the results of new rescue excavations conducted in 1994 within Julius park (the sites of Terasa and Ivanov dom) for the Special hospital for medical rehabilitation and in cooperation with the Regional Institute for Protection of Cultural Monuments in Osijek. The site of Terasa revealed stone blocks which, as it was ascertained, were part of a monumental Roman building, and in the 19th century were used for constructing Arcadia (Schejbal 1994b: 9).2 The site of Ivanov dom revealed a stone block with inscription (monument base) informing us that the monument was erected by res publica Iasorum for Emperor Marcus Aurelius Antoninus. The first four rows of the inscription were chiselled. It is assumed that the missing section included the name of Emperor Commodus, Marcus Aurelius Antoninus’s son, which would date the monument to the second half of the 2nd century AD (Schejbal 1994b: 10).

It should be stressed that the corpus of the discovered architecture is made up of foundations sporadically constructed of flagstone in a herringbone pattern, while only the western part of the wall itself was preserved. Next to the added semi-circular structure, the foundation was made up of flagstone larger stones, and the whole length of the wall is preserved. It was noted that the structure was adapted to the terrain which slopes in the NW-SE direction and was constructed the same way. Relations of the walls reveal that the semi-circular structure was added to the wall foundation. Based on current finds we can assume that the excavated structure was a part of the fortification complex which included a semi-circular tower and a rectangular tower on the western edge. On the western section of the terrain, a sharp slope towards the west can be seen, which supports the hypothesis that this is the edge of the researched structure.

Excavation was conducted due to the construction of a building by housing cooperative Dalit. 2  Present-day Villa Arcadia in Daruvar. 1 

64

New Data, Old Town

Figure 3: Orthophoto mosaic of the defensive walls and towers in Stari Slavik, Jewish cemetery site (author: M. Mađerić)

Remains interpreted as stone blocks and flagstones were discovered south of the Jewish cemetery, stretching westwards. G. Jakovljević concluded that they can probably be linked to the perimeter wall which Gjuro Szabo noted in the layout from 1905 and which was also recorded in 1983 during the construction of the Dalit buildings (Jakovljević 2018: 176).

Zone 2 includes ridges in straight lines, which make up a grid-like layout west of the ending of Slavik street. Finds of Roman architecture sketched by G. Szabo during his visit to Daruvar in 1907 and the area excavated in 1983 indicate a great probability that these are the remains of Roman residential buildings. Zone 3 has terrain characteristics similar to the ones along the edge of Stari Slavik elevation, west of Ivan Zajec street, above the road, and the railway which destroyed the western part of the hill. The presence of Roman architecture can also be assumed here, as confirmed by G. Szabo’s sketches made at the beginning of the 20th century. Skeletal burials with artefacts from Emperor Constantine’s period were discovered in the vicinity as well (Schejbal 2003: 114).3

Roman forest The Roman forest is an area stretching N-E from the centre of the present-day settlement, east of the Special hospital for medical rehabilitation, north of the urban zone of Stari Slavik. ALS/LiDAR scanning was conducted during 2020 across the 25 ha area. Analysis of the obtained DTM revealed potential preserved underground remains that are visible in the terrain morphology in the southern area of the researched zone. The edge of the elevation sharply sloping towards the north makes up the central axis of the scanned area, while the not-as-steep southern side is the position of the present-day’s residential area of Stari Slavik. Three spots in the terrain morphology revealed indicators of possible structures (Figure 4).

Conclusion The number of finds discovered in the long period, from the 18th century till today, shows that Daruvar was a highly urbanized town and community during the Roman administration. The multitude of finds indicates the presence of a settlement organized according to Roman tradition, having a forum, public buildings such as temples, and thermae. Maps of current finds, results of archaeological excavations at the sites of Stari Slavik and the Jewish cemetery, and the analysis of spatial data enabled us to conclude that the Roman settlement stretched from the area of the present-day Special

Zone 1 is situated around the Jewish cemetery. These characteristics were recorded east of the trench and south of the cemetery, along the cemetery itself, and on the eastern edge of the forest. According to the results of field surveys conducted thus far, we can link most of the recorded characteristics to the assumed fortification complex of the former Roman settlement.

Several skeletal burials in built tombs were discovered in 1947 and a female burial was discovered in 1968 (Schejbal 2003: 114).

3 

65

Andrej Janeš et al.

Figure 4: Topography of Roman finds in the Roman forest (map by: A. Janeš, layout: www.geoportal.dgu.hr, DOF, HOK)

of the acropolis in Caričin grad (Ivanišević 2016). Although it is still too early to present steady conclusions, it can be assumed, with great caution, that some of the discovered architecture belongs to fortification structures. Due to the lack of architectural finds in its eastern part, it is possible that this was the entrance. Čezava (Castrum Novae) indicates that the whole fort was reinforced with circular and semi-circular towers in its sixth phase, dated to the beginning of the 4th century AD (Vasić 1984: 101-102, Figure 8). The southern entrance on the Diana Fort (Karataš) near Kladovo was reinforced at the end of the 3rd and the beginning of the 4th century AD when towers with elongated semi-circular layouts, i.e., layouts in the shape of a horseshoe were rebuilt (Rankov 1984: 8-9). A fort with semi-circular towers was excavated in Pilismarót; Castra ad Herculem, constructed in the 5th century AD (Visy 2000). The results of the radiocarbon dating of a mortar sample from the wall have been dated from the 2nd quarter of the 5th to the end of the 6th century corroborating the presumed analogies.4

hospital for medical rehabilitation towards the east and encompassed the Roman forest and Stari Slavik. Given the sharp terrain incline towards the north and Popov dol, i.e., present-day hospital, scarcity of the finds on that side, and the abundance of finds on the southern slopes of Stari Slavik, we can state that the larger portion of the Roman settlement of Aquae Balissae was situated on the southern, slightly sloping side of Stari Slavik Hill. Architectural finds discovered during the second half of the 20th century, as well as the results of spatial research, support this assumption. Remains discovered north of the Jewish cemetery were the north-eastern edge of the former settlement, (Figure 5). It can be ascertained that the structures discovered along the Jewish cemetery were fortification sections that can be dated to Late Antiquity, but given the extremely scarce finds, these conclusions should be considered with a degree of caution. However, some examples in the region could strengthen this thesis. There are known fortifications with reinforcement in the form of rectangular and semi-circular towers on the territory of present-day Serbia, i.e., former Roman limes on the Danube, and were dated to the period from the 4th to the 6th century AD. Entrances with semi-circular towers can be found in Čezava Fort on the Đerdap gore (Vasić 1994/1995), in Diana Fort in Kladovo, and the fortification

Translated by: K. Deskar The analysis was conducted at the Ruđer Bošković Institute in Zagreb, sample number Z-8287 (95,4 % probability cal AD 431 to 585, 68,3 % probability cal AD 532 to 571 (34,4 %), cal AD 437 to 463 (17 %), cal AD 476 to 499 (16,9 %)).

4 

66

New Data, Old Town

Figure 5: The presumed extent of the Roman town in present-day Daruvar according to archaeological finds (map by: A. Janeš; layout: EU-DEM produced using Copernicus data and information, www.geoportal.dgu.hr, HOK)

Minichreiter, K. 1984a. Otkrivanje rimskog municipija. Glas Slavonije god. XLI, br. 11919, 1. ožujka 1984. Minichreiter, K. 1984b. Daruvar-Stari Slavik, antički arheološki kompleks. Obavijesti Hrvatskog arheološkog društva XVI: 19-20. Minichreiter, K. and V. Sakar 1984. Daruvar-„Stari Slavikʹ, preliminarni izvještaj o zaštitnim arheološkim iskopavanjima. Glas slavonskih muzeja XVIII/XIX, 48/49: 1-4. Mušič, B. 2012. Izvješće o geofizikalnim istraživanjima na lokalitetu Prostor Julijevog parka (Kupališnog parka) Specijalne bolnice za medicinsku rehabilitaciju Daruvarske toplice. Maribor. Pinterović, D. 1975. Nepoznata Slavonija. Osječki zbornik 14-15: 123-155. Benac, A. (ed.) 1987. Praistorija Jugoslavenskih zemalja V. Sarajevo: „Svjetlost, „ OOUR Izdavačka djelatnost: Akademija nauka i umjetnosti Bosne i Hercegovine. Rankov, J. 1984. Karataš-statio Cataractarum Dianae. Đerdapske sveske II: 7-15. Rendić-Miočević, D. 1979-80. Neki ikonografski i onomastički aspekti Silvanove „panonsko-iliričke“ kultne zajednice, Résumé: Quelques aspects iconographiques et onomastiques de la communauté cultuelle »pannonico-illyrique« de Silvain. Vjesnik

Bibliography Bulat, M. 1973. Novi antički nalazi u Daruvaru. Glasnik slavonskih muzeja 22: 24-26. Domić Kunić, A. 2006. Bellum Pannonicum (12.-11. pr.Kr.) posljednja faza osvajanja južne Panonije. Vjesnik Arheološkog muzeja u Zagrebu 3.s. XXXIX: 59164. Ivanišević, V. 2016. Caričin grad (Justinijana Prima): A New-Discovered City for a „Newʹ Society, in S. Marjanović-Dušanić (ed.) Proceedings of the 23rd International Congress of Byzantine Studies, Belgrade 22-27 August 2016: 107-126. Belgrade: The Serbian National Committee of AIEB. Jakovljević, G. 2018. Rimski Daruvar u svjetlu najnovijih arheoloških istraživanja. Radovi Zavoda za znanstvenoistraživački i umjetnički rad u Bjelovaru 12: 169-195. Kuntić-Makvić, B., B. Olujić and J. Osterman 2013. Daruvar-Kupališni perivoj Daruvarske Toplice (Julijev park), in Z. Wiewegh (ed.) Hrvatski arheološki godišnjak 9/2012: 211-215. Minichreiter, K. 1973. Izvještaj sa službenog putovanja u Daruvar, 3.9.1973. g. Arhiv Konzervatorskog odjela MK u Bjelovaru. 67

Andrej Janeš et al. Arheološkog muzeja u Zagrebu 3 vol. XII-XIII: 105-123 + 2 Plates. Schejbal, B. 1994a. Aquae Balissae i njeno područje u antičko doba. Unpublished MA dissertation. University of Zagreb. Schejbal, B. 1994b. Zašitna arheološka istraživanja u Daruvaru 1994. godine. Vrela: glasilo Ogranka Matice hrvatske Daruvar za znanost, kulturu i umjetnost 1: 9-10. Schejbal, B. 2003. Nova razmatranja o Aquae Balissae i narodu Jaza: pejzaž-vode-etimologija-kultovimitologija, pitanje atribucije i kontinuiteta. Opvscvla archaeologica 27: 393-416. Schejbal, B. 2004. Municipium Iasorum (Aquae Balissae). International Conference on Norico-Pannonian

Autonomous Towns: Conference Proceedings, Situla 42: 99-129. Szabo, G. 1934. Iz prošlosti Daruvara i okolice. Narodna Starina 13: 79-98. Taube, F.W. 1778./2012. Povijesni i zemljopisni opis Kraljevine Slavonije i vojvodstva Srijema 1777./1778. (ed. S. Sršen). Osijek: Državni arhiv u Osijeku. Vasić, M. 1984. Čezava-castrum Novae. Starinar XXXIIIXXXIV (1982/83): 91-122. Vasić, M. 1994/1995. Le limes protobyzantin dans le province de Mesie Premiere. Starinar 45-46: 41-53. Visy, Z. 2000. Der pannonische Limes in Ungarn. Stuttgart: Theiss, Konrad.

68

The Road from Mursa to the North: Segments of the Road in Remote Sensing Data Mislav Fileš and Miroslav Vuković Abstract: This research paper aims to present the results of the remote sensing data analysis and the subsequent archaeological interpretation of the spatial data gathered in Bilje, close to the city of Osijek in Slavonija, Croatia. Croatia’s interpretation of roman road networks has predominantly focused on historical and epigraphical sources. This short survey combines several remote sensing methods with a field survey to establish a simple framework for ancient road research. KEYWORDS: ROMAN ROAD, LIMES, LIDAR, REMOTE SENSING, MURSA

Introduction The goal of this research paper is to present the results of the remote sensing data analysis and the subsequent archaeological interpretation of the spatial data gathered in the area of Bilje, close to the city of Osijek in Slavonija, Croatia. The focus is on a particular archaeological feature still present in the landscape, a large segment of the Roman road leading from what was once the Roman city of Mursa northward. This particular section of the road was confirmed by numerous smaller rescue excavations, especially in the vicinity of the village of Bilje. On its southward end where river Drava forms a natural border toward Osijek, the exact location and direction of the road is still a matter of debate. Our research was focused on the southern section of the road, and it included the analysis of aerial photographs, LiDAR data, a drone survey, and a field survey, with the goal of identifying the remains of the road and determining its general direction. This research will contribute to the ongoing scientific debate regarding the exact location of large roman infrastructural projects in Baranja, and more importantly help narrow down the exact route of the Roman road which began with the bridge which spanned the Drava River and served as a northern entrance to Roman Mursa. History of research One of the earliest cartographic surveys in the broader area of ancient Mursa, present-day Osijek, was conducted by Count Luigi Ferdinando Marsigli in the 18th century. It was done as a part of a survey of the frontier of the Habsburg monarchy towards the Ottoman empire. The only mention concerning this research paper from his maps is the road from Osijek to Darda, which he annotates as a Roman road (Vukmanić 2017: 32). In 1777 Franciscan monks from Osijek mentioned ruins of the ancient Roman bridge being visible in the riverbed (Gardaš 2002: 14),

which is a bridge that crosses the Drava River from the city proper to the northern bank. The location of the bridge has been well known and documented through several campaigns (Zubčić 2010: 33). Near ancient Mursa several milestones have been discovered, some of them as early as the 18th century, and they sparked the interest of many early archaeologists who worked in this region. The first comprehensive list of ancient Roman sites in and around Mursa was published by Matija Petar Katančić in the scope of his disertation, in the second half of the 18th century (Pinterović 1979: 102). His work focused on many aspects of Roman Pannonia but he most famously dealt with milestone inscriptions found near the Roman colony of Mursa and tried to establish connections with other Roman sites in the region. In 1785 the first drawing of the road that this article is dealing with has been put on a map by Hungarian cartographer Dost (Vukmanić 2017: 37). In 1798 and 1799 a map by Samuel K. Pávai was made for the northern bank of the Drava River showing the same road (Pavai 1798). Several other later maps done by Joseph Koller and other cartographers and historians in the 19th and early 20th century,mention and show the same road leading from Mursa to the village of Bilje (Vukmanić 2017: 185 – 186). In 2013 another milestone was found on the same route mentioning the distance to Aquincum (ancient Roman Budapest) (Vukmanić 2023). The same road was mentioned several more times in the 20th century by modern archaeologists, but it was never a primary focus of research until developmentled excavation in 2007. Historical maps The remains of the Roman road were noted by historical cartographers in various map iterations of the area through history from Count Marsigli to modern-day scholars. One of the most precise and detailed maps comes from Samuel K. Pávai who drew two maps of the area in 1798 and 1799 drawing the extent of the estate of Bilje and including the ancient roman road

Life Along Communication Routes (Archaeopress 2023): 69–76

Mislav Fileš and Miroslav Vuković on the map as well. The map from 1798 (Pavai 1798) contains a drawing of the Roman road, labelled as: “Ruderae Viae antiquae Romanorum per C. I. V. Maximinum Romanorum Imperatorem Anno Chr. 236 extructae”. The drawing is detailed and shows visible sections of the road as well as projected lines in the supposed direction of further sections. The map also shows the location of the stone bridge over the Drava River at the end of the southern section of the Roman road labelled as: “Ruderis pontis lapidei”, which overlaps with recent underwater archaeological surveys aiming to locate the exact location of the bridge. Another interesting road seemingly overlooked in previous research, is located to the east of the drawn Roman road, and it is labelled as: “Vestigium via publicae antique ab Esiek versus Belljam tendentis usuata usque ad annum 1778”. This road was obviously in use until the end of the 18th century, and although it can’t directly be attributed to the Roman period it is interesting that it seems to be branching out from the Roman road after it crosses the bridge to Baranja, and it also offers an alternate route towards Bilje and Kopačevo. Since the map from 1798 by Samuel K. Pávai provided us with many new details we decided to use his map and georeferenced it in QGIS to attempt to overlay it with our remote sensing data. For these purposes, we used the Georeferencer to identify points which were unchanged in the landscape in the past

200 years. Fortunately, we were able to use the existing modern road from Osijek to Bilje, which was built upon the Austrian road erected in the 18th century, as a reference. Remote sensing Aerial photographs There is a wealth of aerial photographs taken in various seasons of the year and varying states of the crops planted on the fields in our research area. The large number of available aerial photographs is due to the fact that our research area is located near a major city. Ploughed fields and marshlands reappropriated for agricultural use are significant factors that impact the state of preservation of archaeological sites. Nonetheless, these fields can be an excellent source of information due to the potential soil marks and crop marks left by archaeological features present below the surface. Our first task was to map out the locations where there was some archaeological research conducted. Since the northern section near the village of Bilje was more extensively researched we started with the northern section. The remains of the Roman road are visible as a straight line stretching from the banks of Stara Drava near Bilje toward the modern

Figure 1: A segment of the map by Samuel K. Pávai from 1798, showing the route of the three roads marked by coloured arrows (blue - Austrian road from the 18th century; red - Roman road from the 3rd/4th century AD; green - possible second road in use until the 18th century), (author: S.K. Pávai)

70

The Road from Mursa to the North

Figure 2: The northern section of the road visible on aerial images source: Google Earth, February 2021, (author: M. Vuković)

from light to dark and the road itself continues for 800 m before running into a modern floodwall. The change in the colouration of the soil mark could be a consequence of the changes in the soil (i.e., higher moisture in some areas), or it could also reflect a change in the material of the road construction. This particular satellite image also reveals traces of the second road present on the historical map from 1798 which confirms that it was present in the field and furthers the credibility of the map from Samuel K. Pávai.

road connecting Osijek and Bilje. The feature on the photographs is approximately 19 m wide and 1.28 km long, when seen as a soil mark it has a dry, white middle segment with darker ditch-like features on both sides. The colour of the midsection could be due to the fact that some of the original stone and gravel pavement was found in situ during rescue excavations conducted in 2007. During those excavations, the researchers have documented remains of a roman road dated to the 3rd/4th century AD, with the following description: “The extant remains of the road are built of compacted limestone and sandstone, and among the building elements there is a large quantity of Roman brick” (Skelac 2007: 29). Another interesting aspect of the aerial data interpretation is the fact that a part of the road is clearly overlaying a geological feature, a dried-out river meander. Although overlayed by modern infrastructure the feature continues in a straight line south from the modern Osijek-Bilje Road where the soil marks gradually change

LiDAR/ALS The data available for our research area encompasses an ALS survey conducted in 2008. The survey was completed for the Croatian public water management company Hrvatske vode d.o.o. and had the goal of identifying potential flood areas in the basin of Dunav, Drava and Vuka rivers. The metadata is incomplete, and we are still trying to ascertain the flight time and the 71

Mislav Fileš and Miroslav Vuković

Figure 3: The southern section of the road visible on aerial images; red arrows show the soil marks revealing the Roman road; green arrows show the soil mark related to the second road visible on the old maps (source: Google Earth, February 2021, (author: M. Vuković)

scanner that was used in the survey. Non the less we were offered access to the raw data by Hrvatske vode d.o.o. and the *.xyz files were organized in a standard ASCII grid with a final ground resolution of five points/ m2 for the entire dataset. Ideally, for archaeological interpretation, the data should have a ground resolution of at least 20 points/m2, as the lower resolution can often directly affect our ability to perceive potential archaeological features. The raw data were imported to a specialized software, the RVT - Relief visualization toolkit which enabled us to create terrain visualizations designed to accentuate potential archaeological features. We visualized the dataset using the LRM Local relief model, SvF - Sky view factor, Hillshade and subsequently imported the data into QGIS where it was overlayed with historical data and aerial images.

overlayed by modern infrastructure and fields erected in what was once a marshland (Figure 5). The northern segment of the road is completely invisible in the data confirming the conclusion of the excavation team that what remains are just traces of a road which was mostly destroyed by modern farming practices. The situation is a bit different in the southern section where the road seems to be present not only on aerial data as a soil mark but can also be seen as an expression in the relief on the ALS data. It was surprising as it showed that although the land was intensively worked on with modern machinery some archaeological features can still be identified on the LiDAR data. The feature is on average 19 m wide as was the case with the feature visible on the aerial data and it is elevated almost a full meter when compared to the surrounding terrain.

The data shows a dramatically influenced landscape with visible old streams, marshes, and river meanders,

The identification of this potential feature presented a perfect opportunity to confirm our findings in the field 72

The Road from Mursa to the North

Figure 4: A segment of the ALS data showing the old meanders, channels, and modern interventions in the landscape near the city of Osijek; QGIS, visualization: Local relief model, r = 20, (author: M. Vuković)

Figure 5: A segment of the LiDAR data showing the cross-section of the roman road and its elevation in comparison with the surrounding terrain, (author: M. Vuković)

73

Mislav Fileš and Miroslav Vuković and conduct a survey with the aim of confirming the southern section of the road.

data sources confirming the presence of the feature identified on the LiDAR data.

Drone surveys and field walking

Conclusion

A short survey was conducted to confirm the presence of the elevated feature in the field. The conditions for the survey were far from ideal as there was some vegetation still present on the field and it seems that the plot has not been worked on for quite some time. Despite the poor visibility and the fact that a proper grid survey could not be established, we managed to locate the elevated feature and confirm it as the remains of a Roman road. The feature is quite prominent in an otherwise flat field and fragmented roman bricks and tegulae can be found strewn across the full length of the road.

All the identified segments of the road present on all of our data sources have been overlaid in QGIS with the goal of providing us with a map of the exact route of the Roman road stretching from the bridge in Roman Mursa north towards Baranja. The features are completely aligned with rescue excavations which confirmed the presence of the road in the northern section (Skelac 2007: 30). Although the southern section is cut off by modern interventions its projected route would end at the left bank of the Drava River where segments of the stone Roman bridge were located by underwater surveys on the exact place where it was mentioned in the 18th century (Gardaš 2001: 9-12; Gardaš 2002: 9; Gardaš 2003: 18-19; Zubčić 2009: 44; Zubčić 2010: 36). Another interesting feature identified in the data is the presence of another road which seems to branch out from the southern section offering an alternate route towards Bilje and Kopačevo.

We conducted a drone survey using a DJI Mavic 2 Pro with an integrated Hasselbad sensor, from a height of 100m. The flight plan was designed to cover the full length of the feature and provide us with overlapping images for the creation of a photogrammetric 3D model of the surface beneath the drone. The 3D model was created using Agisoft Metashape software and the final result was a digital surface model (DSM) with a ground resolution of nine cm/pix. The image of the finished 3D model was imported to GIS and overlayed with other

This alternate route is drawn on a map of the area from the 18th century, and since it seems to branch out from what is now confirmed as a Roman road, it should be explored as a possible second route from Mursa towards

Figure 6: Top left: an oblique aerial photo showing the elevated segment of the road within dried-out crops; Bottom left: a shaded DSM made from vertical overlapping drone photos showing the same segment; Bottom right: a digital elevation model showing the elevated segment of the road, (author: M. Vuković)

74

The Road from Mursa to the North

Figure 7: Left: satellite photo showing soli marks of two roads; right: map from 1798 showing the branching point of the two roads, (author: M. Vuković)

Figure 8: Final interpretation of our research in QGIS; red dashed line marks the route of the Roman road on historical maps, an area in yellow represents the section that is visible in relief as well as on the aerial/ satellite images, an area in red represents the sections that are visible only on aerial images; green dashed line marks the route of the second road which was in use until the end of the 18th century and the area in red represents the section that is visible in aerial/satellite images, (author: M. Vuković)

75

Mislav Fileš and Miroslav Vuković Bilje and Kopačevo. Since there are traces of a potential Roman military outpost in the vicinity of these villages this potential new route could influence our future interpretations of the Roman Limes border on the Danube.

Bibliography Gardaš, M. 2001. Neki novi podaci o rimskom mostu pokraj Murse. Anali zavoda za znanstveni i umjetnički rad u Osijeku 17: 7-26. Gardaš, M. 2002. Ostaci rimskog mosta u koritu rijeke Drave pokraj Osijeka prije i nakon raščišćavanja 1985. godine. Anali zavoda za znanstveni i umjetnički rad u Osijeku 18: 7-21. Gardaš, M. 2003. Neke pretpostavke o izgledu rimskog mosta pokraj Murse. Anali zavoda za znanstveni i umjetnički rad u Osijeku 19: 7-15. Pavai, S. 1798. Map acquired from Croatian state archives in Osijek (HR-DAOS 26-8) Pinterović, D. 1979. Katančić – Inicijator istraživanja antičke Murse. Osječki zbornik 17: 95-106. Skelac, G. and K. Vodička 2007. Rimska prometnica. Hrvatski arheološki godišnjak 3/2006: 29-30. Vukmanić, I. 2017. Dunavski limes u Hrvatskoj. Unpublished PhD dissertation, University of Zagreb. From: http://darhiv.ffzg.unizg.hr/ id/eprint/9240/1/Vukmani%C4%87%20I.,%20 Dunavski%20limes%20u%20Hrvatskoj,%20dio%201. pdf (Accessed: 04.08.2023) Vukmanić, I. (2023). The recent discovery of three Roman inscriptions in Baranja and their interpretation, Osječki zbornik, Vol. 38(xx), p 9-15 From: https:// hrcak.srce.hr/300456 (Accessed: 04.08.2023) Zubčić, K. 2009. Osijek-rijeka Drava (rimski most). Hrvatski arheološki godišnjak 5/2008: 44-45. Zubčić, K. 2010. Osijek-rimski most. Hrvatski arheološki godišnjak 6/2009: 35-37.

Although the landscape has gone through significant changes our research has shown that archaeological prospection and remote sensing methods can give us good results and help identify new archaeological features. When comparing the methods, it has to be noted that in this particular landscape, aerial archaeology still provides us with the best results. This is a result of the nature of the soil marks and crop marks which are affected by remnants of archaeological features still present underground. As was expected methods which are focused on changes in microtopography on the surface, such as LiDAR and photogrammetry, gave us limited results. Obviously, this is a consequence of modern interventions, particularly farming which in most cases leaves us with flattened-out surfaces devoid of any microtopographic indicators. Fortunately, it seems that larger infrastructural interventions from the past can leave a big enough trace in the landscape that even modern farming practices can’t completely level them out, which was the case in the southern section of our Roman road. Although the results were limited when compared to aerial photography, where the entire length of the road was visible, this research has demonstrated that LiDAR data acquired for other purposes and with a smaller resolution can be successfully used to identify archaeological features even in these heavily influenced landscapes.

76

On the Road from Ad Herculem to Rittium (Segment Čortanovci-Surduk) Biljana Lučić, Alessandro De Rosa, Sara Zanni, Gorana Lemajić, Sonja Štefanski-Zorić, Radoslav Muždeka and Uroš Nikolić Abstract: In this paper, we will present the results of the research performed in the Lower Pannonian part of the Danube Limes with emphasis on the segment of the Roman road from the archaeological sites located in the area comprehended between Mihaljevačka šuma in Čortanovci (Ad Herculem) through Gradina in Stari Slankamen (Acumincum) to Gradina in Surduk (Rittium). We will focus our attention on the results obtained through non-destructive methods of remote field detection (from satellite images, historical and other more recent maps to LiDAR images) and their verification in the field. The paper will include an overview of the site’s spatial distribution from a diachronic perspective in order to monitor the evolution of settlement dynamics in the whole research area. Special attention will be dedicated to the analysis of materials from the archaeological site of Marina in Surduk. KEYWORDS: LOWER PANNONIAN LIMES, ROMAN ROAD, ACUMINCUM, RITTIUM, CENTURIATION, LIMITATION, MILITARY EQUIPMENT

Spatial framework and methodology The borders of the former Roman Empire, or Limes, stretch over 7500 km through 20 countries, making them one of the longest cultural heritage monuments in the world (Korać et al. 2014: 5). The length of the Limes in Serbia is 450 km, and the segment spanning from Čortanovci to Novi Banovci, which will be the focus of this paper, geographically belongs to the Lower Pannonian Limes and extends for about 50 km long. This area is one of the least explored parts of the Limes in Serbia. The latest research was conducted on a modest scale in the 1950s, 1960s and 1970s and was led by the Homeland Museum of Zemun, the Military Museum Belgrade, the Museum of Vojvodina and the Provincial Institute for the Protection of Cultural Monuments.1 At that time, systematic surveys were carried out in order to determine the priorities regarding the archaeological excavations undertaken in Stari Slankamen at the sites of Gradina, Humka and Dugorep, then at Gradina in Novi Banovci, site Mihaljevačka šuma in Čortanovci, as well as at Banoštor, Dumbovo near Beočin, Neštin, Čerević, Susek, and (on the left bank of the Danube) in Begeč (Onagrinum) and Bač.2 Recent research at the Petrovaradin Fortress has revealed important data about the Roman fortification of Cusum.3 The area that is The results of the survey were published (Grbić 1961). The synthesis of the executed works on the Limes in Lower Pannonia was published by Petrović (1995: 9-32). Revision of the results of the research was carried out at the Gradina site in Stari Slankamen by Marko Popović (1996: 155-168), together with the coordinator of the research Danica Dimitrijević (1996: 7-38). An extensive study of the Roman army in this area has been done within the catalogue of the exhibition carrying the same name (Dautova-Ruševljan, Vujović 2006). 3  The results of the research were published by N. Savić (2020: 33-43). 1  2 

the subject of our research is part of the project “Border of the Roman Empire - Danube Limes from Acumincum to Rittium”,4 which was carried out in 2019 and 2020. The project assessed the existing archaeological data and information regarding the archaeological heritage of the Roman defensive system in the administrative territory of Srem, 15 km inland from the Danube bank, and analyzed the topography of the wider urban area of ​​Basianae that gravitating around the Limes (Figure 1). As a result of reviewing the actual state of preservation of the heritage and the possibility of improving cooperation with the local community, three sites were selected to be included in the UNESCO tentative list: Mihaljevačka šuma in Čortanovci, Gradina in Stari Slankamen and Gradina in Surduk. Each of these sites will be the subject of special studies, and we will take this opportunity to look at the results of the analysis in the wider research area (Figure 1). The project involved non-destructive methods of archaeological research; remote sensing analysis of satellite images, collection and processing of available cartographic and other data, as well as LiDAR data and UAV aerial images. The obtained data were the basis for planning the surface terrain survey, to precisely georeference the sites and for determining the necessary steps to ensure legal protection of the involved archaeological sites and the further course of research. Also, this opportunity allowed us to continue the work on the study of Roman roads in Srem, which started within the RecRoad project (2017-2018) providing excellent results in terms of the application The project was financed by the Ministry of Culture and Information of the Republic of Serbia.

4 

Life Along Communication Routes (Archaeopress 2023): 77–107

Biljana Lučić et al.

Figure 1: Territory of Pannonia Secunda with a marked zone of the research area (Lučić 2021: Map 1)

of satellite remote sensing methods to the mapping of buried archaeological structures (Figure 2) (Zanni and De Rosa 2019: 1-18). Through the analysis of satellite images spatially defined position and extension of the archaeological sites located in the territory of Bassianae in the cadastral municipalities of Čortanovci, Krčedin, Beška, Novi Slankamen, Stari Slankamen, Novi Karlovci and Surduk. In order to increase our knowledge about the archaeological shreds of evidence in this part of the Danube Limes, we investigated the relationships of these fortifications with the hinterland, the city of Bassianae as the administrative centre of this part of the territory, and the fortification at the site of Solnok (Caput Bassianensae).5

protection institutions, so they were not mentioned in the various works published over the last two decades.

The first phase of the research included the gathering of archaeological material collected during several campaigns of terrain systematic survey performed in the territory of Srem by several generations of archaeologists (Popović 1996a: 53-92; Popović 1996b: 9-24; Popović 1997: 7-24; Đorđević 2007; Crnobrnja 2015; Lučić 2017).6 The archaeological materials collected during the 1970s and 1980s were mostly unpublished, but only inventoried and archived in competent

In accordance with examples of good practice in the use of new technologies, to get more details, we decided that after collecting archaeological, historical and cartographic materials, in the continuation of the project, we would apply methods of remote inspection using photographs taken by the drone and LiDAR field imaging. LiDAR was performed only in the area of the Gradina Fortress in Stari Slankamen (Figure 3).9

In the second phase of the project, our goal was to use non-destructive remote sensing methods to improve our knowledge about existing archaeological sites and to obtain new data about the area of interest. It turned out that the Srem plain, where more than 60% of the land is intended for agriculture, is especially suitable for remote sensing analysis, using different types of satellite photos from EarthExplorer7 and Google Earth Pro™ to multispectral Sentinel-2 images.8 Different types of historical maps were of great help to us in studying changes in the topography of the terrain.

https://earthexplorer.usgs.gov/ https://spacedata.copernicus.eu/ 9  Parts of the defence system can be clearly traced on a LiDAR recording. The first fortification includes the oldest stone ramparts, part of whose defence system consists of traces of platforms, ramps and earthen ramparts, which are concentrically arranged following the natural configuration of the terrain from north to west. The results of the archaeological excavations of Gradina in Stari Slankamen, the first phase of the fortification on the site are related to the Scordisci and the 1st century AD that is, for the period immediately preceding the Roman conquest (Popović 1996: 155-156). 7  8 

During the future phases of the project, it will focus on the settlements Belegiš, Stari Banovci, Novi Banovci (on the Danube bank), as well as Stara Pazova, Golubinci, Donji Petrovci, Popinci and Dobrinci (the area that connects the Danube bank and the administrative center of the area - Bassianae). Material for the second phase of the project has been prepared. 6  Information about every archaeological locality in the research area are gathered in the Table 2 of the article. 5 

78

On the Road from Ad Herculem to Rittium

Figure 2: Georeferenced route of the Roman road Sirmium- Singidunum by S. Zanni and A. De Rosa (Zanni 2021: Map 2)

Analysis of collected cartographic material and interpretation Geographical features of the area The research started by observing geomorphological, pedological and hydrogeographic maps of the area. The research area is located on the loess plain. There are large areas having very high agricultural production values that are suitable for growing crops (Figures 4 and 5). The Danube bank consists of this area of high and steep loess banks intersected by ravines. Streams, like Patka and Budovar, flowed, in the Roman period as today, through the flat part of the terrain and were once navigable. Along the canals and streams, as well as along the banks of the Danube, there were a number of settlements. The construction of the canal network was of exceptional importance for the development of the Bassianae territory. Regulatory work in the Sava River Basin began in Roman times when Jaračka Jarčina and Progarska Jarčina were excavated.

Figure 3: LiDAR image of the Gradina fortification in Stari Slankamen (GeoGIS Consultants 2019: Map 3)

Limes Road (Rippa Pannonica)

The final phase of the project contemplated the verification of the obtained results by surveying and precise georeferencing archaeological sites from all periods. A terrain surface survey was performed on 60 sites. Archaeological sites were mapped within the QGIS program, then georeferenced data were connected to the state cadastral network (GeoSerbia).10 Depending on the size of the area where the movable archaeological finds were discovered, the sites are presented as points or polygons. 10 

For the purposes of studying the organization of space in Roman times, data on the positions of roads are very important. The use of satellite photographs is essential in finding the routes of roads in our area. They proved to be extremely efficient this time as well, because today the Roman roads are extremely poorly preserved due to the intensive cultivation of the land, and on the ground, without prior preparation, they are practically unrecognizable. Road construction was extremely important for economic development, trade, and military transport. The army was used for their construction, as well as for the construction of

https://geosrbija.rs/

79

Biljana Lučić et al.

Figures 4 and 5: The area of southeastern Pannonia transferred to the modern hydrographic and pedological map with the area of interest highlighted. Base maps were produced by the Geographic Institute in Belgrade (Lučić 2021: Maps 4 and 5)

80

On the Road from Ad Herculem to Rittium aqueducts, canals and fortifications, as well as for the needs of other public works (Dautova-Ruševljan, Vujović 2006: 22). Roads in our area began to be built during the time of the first emperors of the Julian dynastyOctavian Augustus and Tiberius (Dautova-Ruševljan, Vujović 2006: 13). The end of the road construction activity, in the coastal area of ​​the Danube, is considered to be, by the majority of authors, the middle of the 1st century AD (Petrović 1986: 95; Dautova-Ruševljan, Vujović 2006: 14). However, it is also believed that the province of Pannonia Inferior, as well as the Srem section of the Danube Limes, was finally organized under Trajan (AD 103-107) (Petrović 1995: 10). The Limes road (Ripa Pannonica) stretched for a length of about 2300 km, connecting the fortifications on the bank of the Danube. It started from the fortification of Castro Reginae (today’s Regensburg in Germany), ending at the confluence of the Danube in the Black Sea (Crnobrnja 2015: 179). In the area of ​​southeastern Pannonia, it connected the fortifications on the right bank of the Danube from Taurunum, through Burgenae, Rittium, Acumincum, fortification in Čortanovci (Ad Herculis?), Cusum, Bonnonia, Cuccium, Cornacum. Judging by the itineraries, it was 67 Roman miles long (about 99 km). The oldest milestones found on its route date back to the time of Emperor Nerva (AD 96-98) (Mirković 2006: 137, cat. No. 10) and Marcus Aurelius (AD 161-180) and it was found at Petrovaradin (Cusum). The latest restorations are testified by a milestone from the time of Septimius Severus (AD 197-199) (Bruckner, Dautova-Ruševljan 2015: 25), Gordianus (AD 238-244) (Mirković 2006: 137, cat. No. 11) and Claudius II (AD 268-270) (Mirković 2006: 138, cat. No. 12), all of them found at Petrovaradin. Parallel with the terrestrial road, the river Danube was among the main waterways of the Empire. Ports at statio Confluentes (between Belgrade and Zemun) and Bonnonia (Banoštor) were built in the 1st century AD, Burgenae (Novi Banovci) and Onagrinum (Begeč) in the Diocletian period. They were very important for the development of the province of Pannonia, solving both trade and military functions. Thanks to anomalies observed on satellite images marked by rectangular red and purple polygons we could spot the remains of the roads (Figures 6, 7 and 8). The shreds of evidence of the road were registered in the area of ​​Beška, east of the Belgrade-Novi Sad highway (Figure 6), and then the route stretches along the ridge north of Krčedin and Novi Slankamen (Figure 7). The branch of the road leading to the present-day settlement of Stari Slankamen is clearly visible in the satellite image.

lines) intersect it leading to the fortification of Bonnonia in Novi Banovci (north-south or cardo line) and Rittium in Surduk (west-east or decumanus line). It is obvious that the line of decumanus extends to the west towards Bassianae in Donji Petrovci. The branch of the road from Bassianae to Rittium was clearly recognized during the archaeological survey of the Sirmium-Singidunum road (Figure 3). Limitation: centuriation and strigatio The consolidation of Roman rule in the newly conquered territory was accompanied, in addition to the construction of roads, by the establishment of settlements and the construction of fortifications along the Danube and in the interior of the province. The wider territory of the Limes had to be defined in detail in the cadastral, parallel with the construction of roads and the establishment of the fortifications. The Romans, in order to divide and allocate conquered land, in certain phases of the expansion of the Empire and certain regions of the Empire, with their conceptualization of space defined the areas they conquered by a special procedure called limitation (limitatio) (Libertini 2019: 5).11 It is assumed that the centuriation of the provinces, particularly in the 1st century AD, was primarily related to the allocation of land to veterans, which was probably the case in our country, keeping in mind that the study of epigraphic sources suggests that the settlement of veterans after the service, especially in Pannonia, was carried out in the surroundings of military fortifications (Ferjančić 2002: 52-54). The cadastral division of land was done following strict rules based on Roman laws, imperial decisions, and defined territories of certain areas and competencies (Marcu, Cupcea 2011: 547). The main orthogonal scheme of the systematic organization of the rural space that was detected in our region was The basic measuring unit was ten feet, pertica 2.957 m or decempeda. There were two basic types of limitation or delimitation (delimitationes), namely: stigation (strigatio, pl. strigationes) and centuriation (centuriatio, pl. centurationes). The first type was used to define the direction border roads–limits (limites) that were straight line, parallel, which distance was always equal to the value of the actus module (actus). Single actus value was 120 x 120 feet, or 35.48 x 35.48 m in the modern metric system. The strigation method defined the borders of the area that would be divided, but also the angle at which the limits were placed relative to the north-south axis. The angle could not be higher than 45⁰ (Libertini 2019: 5-6). Centuriation was done by tracing two groups of the limits lines that cut at the right angle on a distance equal to the actus module. If the distance between the two groups of limits was the same it was a square centuriation, and if it was different, than rectangular. Also, the centuriation depended on the angle at which the limits were set. In order to achieve the ideal centuriation scheme, in umbilicus, the central point of the field that was divided, the main axes were cardo maximus (oriented from south to north) and decumanus maximus (from east to west) cut. In this way, decumanus divided the land into the right (dextra) and the left (sinistra) half, and the cardo on area in front of (ultra) and behind (citra). The orientation of the centuriations depended on a number of factors, primarily on the shape and orientation of arable land. When measuring length, a special method called cultellatio (vertical differences in the height of the soil) was defined.

11 

The route of the road along the Limes was spotted in the area between Stari Slankamen and Surduk (Figure 8), between the Danube bank and the modern road that connects these two settlements. Parallel to the Roman road, internal communications (limitationes) are still visible on the ground. The two main communication roads representing the centuriation axes (yellow dashed 81

Biljana Lučić et al.

Figure 6: Satellite image of the area of the settlement of Beška (Google Earth Pro) with anomalies representing the remains of the Roman road (De Rosa 2020: Map 6)

Figure 7: Satellite image of the area of Krčedin, Stari and Novi Slankamen with anomalies representing the remains of the Roman road (De Rosa 2019: Map 7)

82

On the Road from Ad Herculem to Rittium

Figure 8: Satellite photography with a marked route (yellow solid line), internal communications (yellow dashed lines) and site positions (red circles and dots) (De Rosa 2020 and Lučić 2021: Map 8)

on an area of over 2500 km2 (Figure 8). The orthogonal scheme is still visible with the directions of the cardo and decumanus axes, while the networks of the more detailed division of space are only partially preserved (Figures 8 and 9). It is interesting to note that the cardo and decumanus, established in Roman times, are still in use, although the fortifications that formed the backbone of the spatial organization have been abandoned a long time ago. The importance of studying the remains of the patterns of the Roman agricultural land organization lies in the fact that, in proportion to the degree of preservation of the limitations, it shows that agricultural land has been cultivated from antiquity till today and also demonstrates that the areas and the land could not have been completely abandoned (Libertini 2019: 13). This is very important for interpreting the historical circumstances through which the area of interest went through over times. It turned out when studying some examples from the territory of Italy, such as the cities of Cales and Calatia, which are known to be completely abandoned in ancient times, that the territories of these cities actually show good preservation of limitations, which proves that land cultivation was not interrupted (Libertini 2019: 14).

Archaeological site Rittium in Surduk Site Gradina Unlike the archaeological sites Mihaljevačka šuma in Čortanovci and Gradina in Stari Slankamen, which have been explored for many years, Gradina in Surduk has remained on the sidelines in terms of research work, so its valorization and protection have not been possible so far. The historical significance of Rittium and the military units stationed in this fortification has been widely discussed in several important publications (Klemenc 1961: 21-22; Petrović 1995: 22; DautovaRuševljan, Vujović 2006: 62). We have limited data regarding the fortification in Surduk, primarily on its topography, and the existence of a civilian settlement and necropolis. The Gradina site, where the Roman fortification is located, is on arable land, behind houses along the east side of Brđanska Street. The topography of the area leads us to the conclusion that fortifications existed in this area; an elevated plateau with a high loess profile facing the Danube on the north side, a steep ravine on the west, and the outlines of defensive trenches on the south and east sides. 83

Biljana Lučić et al.

Figures 9 and 10: An area on the historical map and satellite image of centuriation (De Rosa 2019: Maps 9 and 10)

84

On the Road from Ad Herculem to Rittium

Figure 11: Panorama of the archaeological site Gradina in Surduk, view from the southeast (red black arrows indicate the outlines of the defensive trench) (Lučić 2019: Photo 11)

Figure 12: Aerial photo of the site Rittium (Marijanović 2021: Photo 12)

The approximate dimensions of the fortification are 175 × 220 m, considering also the defensive moat and part of the settlement outside of the ramparts, the length is much more impressive. Big quantities of finds of construction material have been found at the site, as well as large quantities of pottery fragments from the Roman period that can be dated from the 1st to the 3rd as well as from the 3rd to the 4th (5th/6th) century AD, and there is also evidence prehistoric pottery.

the time of Octavian (Dautova-Ruševljan, Vujović 2006: 58). From the second half of the 1st and the beginning of the 2nd century AD, fortifications were built of stone and became permanent (Dautova-Ruševljan, Vujović 2006: 58). Large military camps for the accommodation of legions (castra) such as Carnuntum (Petronell near Vienna), Vindobona (Vienna), Brigetio (Szőnyju) or Aquincum (Budapest) were not built in the Srem part of the Limes. Smaller-size fortifications (castella) were built there to accommodate parts of the legions and auxiliary troops. The fortification in Surduk once belonged to these castella. Based on better-studied fortifications in Hungary and Austria, their spatial plan can be assumed.

Probably the fortification in Surduk, like early wooden fortifications with a defensive moat and an earthen embankment, the so-called presidium, was built during 85

Biljana Lučić et al. In the fortifications from the second half of the first, to second and first half of the 3rd century AD, the plan was regular, the gates were placed on all four sides of the fortification, and the inner towers along the ramparts and at the corners. There was a defensive moat around the ramparts called fossa. From the second half of the 3rd century AD until the 6th century AD, their plan became irregular, circular or rectangular towers were placed on the outside of the ramparts and only one entrance gate was left (Dautova-Ruševljan and Vujović 2006: 58). From the period of Constantine’s wars against the Sarmatians, some fortifications got trapezoidal-shaped towers (Jovanović 2005: 84-85). Since its founding, military units recruited in the eastern provinces of the Empire have resided in this fortification. We know that the Ala I Augusta Ityreorum sagittatiorum was stationed in the fortress, which smaller or larger parts resided here from the 1st to the 2nd and 3rd century AD. In the 4th century AD, the crew consisted of horsemen (equites Dalmatae). Also, the crews of VII Breucorum civium Romanorum equitata whose presence is testified by a diploma, and of the cohors II Asturum et Callaecorum, cohors II Ityraeorum (testified by the bricks) are mentioned (Dautova-Ruševljan, Vujović 2006: 61).

well as from the Marina site about one kilometre from Gradina in northwestern direction. The Marina site is located on both sides of a ravine, which provided access to the interior from the Danube bank. Precisely in such places, in addition to ravines, at the mouths of streams or rivers into the Danube, smaller quadrangular watchtowers were erected, the remains of which were found during the surveys. Thanks to Lazar Čolaković, a resident of Surduk and antiquities enthusiast, we managed to get the material removed by the weather and dragged down the landslide loess profiles, which he has carefully collected over the last two decades. The material covers a broad period of time. The included chart (Figure 15) represents the approximate percentage of ceramic material pertaining to the various periods.13 Prehistoric pottery (cultural groups) constitute the most relevant category, according to the quantity of findings. The Roman period from the 1st to the end of the 4th century AD is significant (see Plates T. 1-5), while the Medieval period is represented only by materials dating to the Avaric period of domination (belt fittings and pendants), and it is present in ceramic material only from the 10th century AD onwards. The collected materials from the Roman period most probably originate from the necropolis, which presence was visible in some points of the loess sections. The necropolis can be attributed to the Rittium fortress and most probably to the settlement that developed next to it lying under the streets, houses and gardens of contemporary Surduk. Materials coming from the site include the characteristic types of fibulae, which are present in the fortifications of Pannonia and Moesia during 1st-4th century AD (T. 1: 1-16), Late

Findings from the Marina site in Surduk For the time being, we have only indirect evidence about the significance of the Gradina in Surduk and about the richness of its remains, thanks to accidental findings from settlements kept in the Archaeological Museum in Zagreb and the Belgrade City Museum,12 as 12  We know ofthe greatest number of findings from Surduk primarily thanks to numerous reports published in the Vjesnici Arheološkog muzeja u Zagrebu. The first mention of the Surduk findings is about two altars with the same inscription that M. Aurelius Apolinaris, the decurion of Municipia Mursele dedicated to Jupiter Dolihenus (Brunšmid 1895: 180-181; Ljubić 1883: 69; Brunšmid 1907: 107-109, kat. 22 and 226). It was noted that a ceramic lamp with the negroid image of a young man came from the archaeological site of Gradina, as well as a bronze application with a lion’s head, a glass veassel of a square cross section (Šaranović-Svetek 1986: 21, T. VIII/2; Gregl 2013: 65, kat. 90) dated in 2nd-3rd century AD, then silver fibula and ring, bronze ring and fibulae dated in the 4th century AD (Brunšmid 1901: 155-156, kat. 1-7). From the same area comes a stone with a Roman measure of two congii (6.57 l) (Brunšmid 1911: 109, kat. 710), a smaller fragment of a Venus’s marble torso (Brunšmid 1911: 113, kat. 729; Popović 2012: 60, etc. 3), which is dated at the beginning of the 2nd century AD. From the Orthodox cemetery in Surduk comes an altar dedicated to the gods by L(ucius) Ravon(ius) Portin(us) (Brunšmid 1911: 118-119, kat. 737). Part of the Covenant Monument was also found in the village-VICT)ORIN(US)? EX V(otum) M(ERITO L(IBENS)(SOLVIT), badly damaged, dated to the first half of the 3rd century AD; part of a scale armor of two different types, one of which was dated into the 1st century AD (Hoffiller 1912: 39-40, etc. 13), coins of Apollonia and Dirahium (Brunšmid 1912: 265), an amulet (Ljubić 1879a: 6-1). There is also evidence of several bronze objects coming from Surduk: an application from a bronze jar shaped like a ram’s head (Brunšmid 1914: 261, kat. 246), a deer-shaped application, (Brunšmid 1914: 261, kat. 247), a smaller pendant shaped like an elephant’s head (Brunšmid 1914: 261, kat. 248), a bronze mouse figurine standing on its hind legs and holding a round object (Brunšmid 1914: 261, kat. 251), numerous phallus-shaped pendants (Brunšmid 1914: 247, kat. 115-144), a

decorative rivet in the form of an Eros mask (Brunšmid 1914: 247248, kat. 155-156), a decorative application shaped in the form of the head of a young man wearing a Frigian hat (Brunšmid 1914: 248, kat. 159), an application with two theatre masks performed carved in relief (Brunšmid 1914: 251, kat. 177), one lion mask (Brunšmid 1914: 254-255, kat. 202, 205), a horse-jumping application and a horsehead application (Brunšmid 1914: 259, kat. 229, 230, 231), a figurine application shaped as a Hermes bust dated to the 1st century AD (Brunšmid 1914: 223, kat. 34), one bronze application shaped as the mask of Silenus (Brunšmid 1914: 228, kat. 49), one bust of a soldier (Brunšmid 1914: 240, kat. 80), two torsos of men (Brunšmid 1914: 243, kat. 89, 90), as well as a fragment of circular-shaped lead plate of a Danubian horseman (Hoffiller 1935: 62, kat. 2). At the Belgrade City Museum, a late antique belt phalera from Surduk is kept (Vujović 2003: 217, T. I, 6). In Serbia, there are evidences of this type of phalerae only in the area of upper-Moesian and lower Pannonian Limes from Surduk to Kostol. Their presence is related to equestrian and infantry crews and sites where the dignification of oriental deities is confirmed, i. e. presence of military units from the Orient (Vujović 2003: 222). From the same collection comes the fitting shaped in the form of the upper body of a bull (Krunić, Crnobrnja 1997: 75, kat. 70) dated into the 2nd-3rd century AD, as well as the figurine of Jupiter Dolichenus, which is dated to the very end of the 2nd century AD (Krunić, Crnobrnja 1997: 39, kat. 28). 13  The ceramic material from the Marina site in Surduk was processed by Ljiljana Muždeka, and its detailed analysis will be presented in a separate article.

86

On the Road from Ad Herculem to Rittium

Figure 13: The foundations of the watchtower at the top of the loess profile at the Marina site in Surduk, view from the east (De Rosa 2019: Photo 13) Figure 14: Ravine, on both sides of which is the Marina site, view from the east (Lemajić 2019: Photo 14)

its decorative and apotrophic role was very important, too. Roman horse equipment had five saddle joints with a belt that then forked to several sides. Based on the functional characteristics of the finds, the equipment can be divided into strap fasteners, junction loops and strap mounts (Plate T. 3: 3, 5-8) which are dated to the 1st to the 2nd century AD; phalerae (Plate T. 3: 4) (decorative and functional) which appear from the time of Claudius; decorative belt fittings (Plate T. 3: 1-2, 1012, 17), pendants (Plate T. 3: 9, 13-14) and applications/ fittings (Plate T. 3: 15-16). The archaeological evidence of Roman military equipment includes also; jaws and objects used for pulling and loosening reins, but these elements have not been found on our site. As pointed out, the strap fasteners, junction loops and strap mounts found at the Marina site are related to the period of the 1st and the beginning of the 2nd century

La Tène coins (Plate T. 1: 17-18), one 2nd century AD coin in secondary use as a rein attachment (Plate T. 1: 20), as well as the coin of Constantius II (Plate T. 1: 19). Weapons like the spear butt (Plate T. 2: 6) and iron trilobate tangled arrowhead (Plate T. 2: 5) can be dated in a wide time range from the 1st-4th century AD, while catapult bolt head (Plate T. 2: 4) can be dated to a more limited time span and, precisely, to the 1st century AD. Decorative parts of military belt sets are relatively rare in our collection, but they also cover the time span from the 1st to the 4th century AD (Plate T. 2: 1-3). The most common artefacts in our collection are horse harnesses (Plate T. 3 and 4), which can be dated from the 1st to the 3rd century AD, according to analogies. The main purpose of the horse equipment was to firmly tie the saddle to the horse, however, judging by the number and variety of decorative forms, we can conclude that 87

Biljana Lučić et al.

Figure 15: Chart of the percentage of ceramic material of certain time periods from the Marina site in Surduk (Lučić, Muždeka 2021: Chart 15)

AD. The decorative belt fittings, rectangular in shape and multi-profiled body, are dated to the same period (Plate T. 3: 1-2, 10). Leaf-shaped and oval-shaped belt fittings, as well as pendants of similar forms, are mostly dated to the second half of the 2nd and 3rd century AD (Plate T. 3: 11-13, 17). They are close in dating with a bronze pendant decorated with glass paste that belongs to the developed form of a lunar pendant (Plate T. 3: 14), as well as a decorative rivet with a circular head with a four-pointed ornament (Plate T. 3: 15). There is a noticeable lack of horse equipment that can be explicitly dated to the period of Late Antiquity. This should be expected considering that all other categories of findings suggest so (Plate T. 4).

considering their proximity. The archaeological evidence suggests the presence of the army in both fortifications from the 1st to the end of the 4th century AD, while some findings from Banovci speak in favour of a longer encampment of the army in this location. Also, some findings related to the presence of military units, according to the context, seem to demonstrate close contacts with troops located at the Lower Pannonian Limes both in Surduk and Novi Banovci (applications in the form of lion heads and masks, busts of soldiers, lead plates with a representation of the cult of the Danubian horseman, inscriptions dedicated to Jupiter Dolihenus, masks with the image of Eros, phallus-shaped pendants, mouse-shaped applications), Stari Slankamen (goatshaped applications, the cult of Jupiter Dolichenus) as well as on Duvarine site in Belegiš. A high percentage of archaeological finds that can be associated with horse equipment was found in Surduk and Novi Banovci. They can testify to the existence of auxiliary unitsmostly cavalry, which in addition to infantry troops, are attested in historical sources in the wider area of ​​the Lower Pannonian Limes (at Burgenae: cohors I Thracum civium Romanorum pia felix, ala I civium Romanorum legio V Iovia, equites Dalmatae, cuneus equitum Constantianorum; at Acumincum: Cohors I Campanorum voluntariorum, Cohors V Gallorum; at Titel: Cohors I Britanica milliaria civium Romanorum equitata; at Rittium: ala Augusta Ityraerorum sagittariorum; ala Pannoniorum- at Belegiš, Slankamen and Teutoburgium: ala Praetoria civium Romanorum; at Caput Bassianense: legio II Adiutrix pia fielis; at Onagrinum: legio VI Herculia and other). The capability of the military units to move quickly along the Limes was of special importance for the control of the border

Numerous other items related to everyday use were found at the site (Plate T. 5). Among the jewellery, there are bracelets, as well as hairpins that are dated to the 3rd and 4th century AD (Plate T. 5: 11-13). Several imported metal vessels can be attributed to the Romanization period: the first one is the handle of a simpulum of the Aislingen type (Plate T. 5: 1) (Feugère, Freycon 2014: 1719; Type 43 after Radnóti), and the second is a casserole (Plate T. 5: 2). Both are dated to the 1st and the first half of the 2nd century AD. Conclusion Comparing the findings from Surduk, with the items of military and horse equipment found at the nearby fortification in Novi Banovci (Radman-Livaja 2008: 295308; Radman-Livaja 2009: 1499-1508) we can notice a great similarity in the material, which is not surprising 88

On the Road from Ad Herculem to Rittium and the Danube waterway, and the terrestrial road made it possible. The presence of soldiers of oriental origin was confirmed among the army on the Limes. In Surduk, this is documented, both by inscriptions and historical sources. It is believed that the most efficient archers (sagittarii) were recruited from the eastern part of the Empire (Greece, Syria, Numidia and Crete), as well as slingers (funditores) (from Komagena, Rhodes and the Balearic Islands) (Vujović, Cvijetić 2018: 55). Also, the study of epigraphic sources suggests the conclusion that the settlement of veterans after the end of the military service, especially in Pannonia, was carried out in the immediate vicinity of military camps or large urban centres (Ferjančić 2002: 52-54). Perhaps this fact can explain the great migrations of the inhabitants of Oriental origin to Pannonia that have taken place since the time of the Severian dynasty (Mocsy 1974: 227). It is assumed that the migrations were encouraged after the wars led by Marcus Aurelius against the Marcomanni, during which this province especially suffered. It is also known that the Pannonian legions enrolled soldiers of oriental origin during the wars of Septimius Severus in the eastern provinces. However, epigraphic inscriptions suggest that in the proximity of Sirmium, Mursa, and Savaria several civilian settlements were entirely populated by people coming from the eastern provinces of the Empire (Mocsy 1974: 228). There is evidence of their presence in Brigetio, Intercisa and Aquincum. Inscriptions testify also the presence of people from Thracia, Africa and Italia but they were probably less relevant in numbers than the Oriental people. Their settlement, among other things, influenced the economic empowerment of the province, the development of trade and crafts, as well as the presence of the cults of Mithras and Jupiter Dolihenus.

5th century AD (Popović 1996: 155-168). The discovered remains of a Roman stone rampart also belong to this period. The fortification was destroyed or abandoned, before the end of the 4th century or in the first decades of the 5th century AD, and at the latest in the period that ended with the invasion of the Huns in AD 441. The castellum from Mihaljevačka šuma in Čortanovci (Ad Herculem?) is also dated to the Late Antique period, as well as the speculum discovered at the Humka site in Stari Slankamen. This period is known for wars fought against the Sarmatians, then the Goth invasion and the collapse of the Empire’s defensive system on the Danube. The general situation of political uncertainty obviously influenced the construction of new fortifications, and probably the restoration of the old ones. The results of archaeological excavations, as well as surveys, lead to the conclusion that the landscape along the Danube bank after the 6th century AD remained abandoned for a long time. A slightly different situation was observed in sites that are far from the main communication network. In the areas of Novi Slankamen (Čarevac, Grabovac) and Novi Karlovci (Vinjerac, Rupine, Vodice), the abundant medieval shreds of evidence found in the areas where the Roman sites are documented suggest the continuity of life. Catalogue of finds Plate I: FIBULAE, COINS 1. Late La Tène bronze fibula with the semicircular thickened bow. The bow has a circular cross-section. The foot is thinner with a small semicircular pin catch. Length 4.7 cm. Field inventory number C-33. Approximate dated: second half of the 1st century BC. 2. Highly profiled bronze fibula. The spring plate, spring and pin are missing. There is one knob on the bow, as well as at the end of the foot. The bow is thickened at the transition to the head. Length 4.5 cm, preserved width 1.2 cm. Field inventory number C-3. Literature: Bojović 1983: type 9, 3 (Ritopek, fortificationes in Srem); Petković 2010: type 13B (Castra Tricornia). Approximate dated: 1st century AD. 3. Highly profiled bronze fibula with partially preserved spring plate and spring. The pin is missing. The same type as the previous fibula, cat. 2 (Plate T. 1, 2). Length 4.2 cm. Field inventory number C-4. 4. Silver fibula fragment. A part of the foot, pin catch and bow is preserved. It cannot be typologically determined. At the end of the foot is an oval-shaped knob as well as at the transition to the bow. The needle holder is rectangular. Preserved length 3.6 cm. Field inventory number C-7. Approximate dated: Roman period, 1st to 3rd century AD. 5. Fragment of the highly profiled bronze fibula (?).

The majority of the archaeological sites discovered during the survey were concentrated along the bank of the Danube. On certain sections of the Danube bank, zones with archaeological shreds of evidence appear in a greater or lesser continuity in the length of about one kilometre. The analysis of the collected ceramic material from the sites along the Danube revealed a significant presence of Roman ceramics, which makes up 85-95% of the collected samples. The exceptions are the sites in Stari Slankamen, where the horizon of the late medieval period dominates. Based on preliminary data obtained from surface surveys, a certain continuity of settlement on the Danube bank can be traced only in relation to the prehistoric period, which is obvious at the Marina site in Surduk. The results of the analysis of ceramics from the archaeological excavations of the Gradina site at Stari Slankamen have shown that 90% of all excavated fragments from the Roman settlement phase belong to the Late Antiquity period, more precisely to the 4th and the first decades of the 89

Biljana Lučić et al. The foot with a pin catch and a part of the bow is preserved. There is a button-shaped knob at the transition of the bow to the foot, as well as bulbshaped knob at the end of the foot. Preserved length 3.5 cm, width 0.7 cm. Field inventory number C-8. Dated: 1st to 3rd century AD. 6. Highly profiled fibula with no spring plate. The spring and pin are missing. The bow is shallowly curved. There is one button thickening in the form of a ridge which is profiled several times. At the end of the foot, there are remains of the knob. The pin catcher is tall and rectangular. Analogy: Bojović 1983: type 10, variant 3. Field inventory number C-5. Dated: 2nd and first half of the 3rd century AD. 7. Bronze crossbow fibula fragment. A part of the foot and bow. At the transition from the bow to the foot, the body is thinned and decorated with incisions. The foot is also decorated with transverse incisions and longitudinal strips between which triangles are incised. Preserved length 3 cm, width 0.8 cm. Field inventory number C- 6. Analogy: Petković 2010: type 33C (Viminacium). Approximate dated: 4th century AD. 8. Highly profiled bronze fibula with a partly preserved foot, rectangular pin catch and part of the bow. There is a triple decorative knob at the bow and a stud-shaped knob at the end of the foot. Preserved length 3.6 cm, width 0.8 cm. Field inventory number C-9. Dated: 1st to 3rd century AD. 9. Fragment of the stripe-shaped foot with the root of the pin catch. At the end of the foot, a vertical bulbshaped knob is attached. Preserved length 3.0 cm, width 0.8 cm. Field inventory number C-10. Dated: 1st to 3rd century AD (?). 10. Small bronze fibula. It belongs to the type of knee fibulae with spring and semi-circular plate. Semi circular plate is decorated with zig-zag incisions around the rim. The bow has a semi-circular crosssection. The pin and pin catch are missing. Length 3.5 cm, width 1.9 cm. Field inventory number C-30. Analogies: Bojović 1983: 55, T. XIX, 173, 175, type 20, var. 1. (along the limes in Upper Moesia ); Petković 2010: tip 18A (Viminacium). Dated: 2nd to 3rd century AD. 11. Foot, bow and pin catch fragment of the highly profiled bronze fibula (?). There is a triple decorative knob at the bow and a stud-shaped knob at the end of the foot. The pin catch is triangular with three circular perforations. Dimensions: length 5.4 cm. Field inventory number C-32. Analogy: Petković 2010, type 33, cat. 585, Pl. XX, 1, earlier variants of highly profiled trumpet-shaped fibulae, var. 1 (?) (Šabac and Pocerina). Approximate dated: 2nd century AD. 12. Foot and bow fragment of the bronze crossbow fibula. The bow has a trapezoidal cross-section. There is a single decorative knob at the bow. The foot is

decorated with triangular stamped ornaments along the rim. The central part of the rectangular flat foot has two parallel incisions. Preserved length 4 cm, width 1 cm. Field inventory number C-2. Dated: 4th century AD. 13. Crossbar and bow fragment of the bronze crossbow fibula. There are three onion-shaped knobs preserved on it. The crossbar has a rectangular cross-section. It is decorated with notches at the transition to the bow and perforated on both sides of the central bulb. The bow has a trapezoidal cross-section. It has incised oblique lines at the top. Preserved length 3.2 cm, width 5.1 cm. Field inventory number C-31. Approximate dated: 4th century AD. 14. Crossbar and bow fragment of the bronze crossbow fibula. The foot and pin are missing. The crossbar has a six-sided cross-section with three ovoidshaped knobs on it. It is slightly thickened at the transition to the bow. At both ends of the crossbar, there are double ridge thickenings in front of the knobs. The bow has a trapezoidal cross-section. There are triangular incisions at the top of it. At the transition to the foot and to the crossbar, the bow is decorated with thickenings in the form of a ridge, and incisions. Preserved width 5.7 cm, length 4.8 cm. Field inventory number C-126. Analogies: Petković 2010, similar to type 34 A. Approximate dated: the end of the 3rd and the first quarter of the 4th century AD (from the first tetrarchy to the end of the joint rule of Constantine I and Licinius, 293-324). 15. Crossbar and bow fragment of the bronze crossbow fibula. There are two onion-shaped knobs preserved on it. The crossbar has a rectangular cross-section. It is decorated with notches and perforated on both sides of the central bulb. The bow has a trapezoidal cross-section. It has incised lines at the top of it. Preserved length 4 cm, width 3 cm. Field inventory number C-1. Analogy: Bojović 1983: type 34, variant 6 (350-380). Approximate dated: 4th century AD. 16. Ring-shaped bronze fibula with widened and thickened ends: type 29 (Petković 2010). The bronze fibula has an open-ringed body with a circular crosssection. Button-shaped ends are decorated with incisions. The bronze pin is completely preserved. It is believed that they were worn by men, mostly soldiers, on the right shoulder. They mostly derive from fortifications. Preserved length of the pin is 2.9 cm, R 2.7 cm. Field inventory number C-28. Approximate dated: Roman period, 4th century AD (from the second half of the 3rd to the beginning of the 5th century). 17. Celtic bronze coin. A stylized representation of the ruler turned to the right can be seen on the obverse. A stylized representation of a horse is barely visible on the reverse. R 1.5 cm. Field inventory number C-49. Approximate Dated: Celtic period of domination in Srem. 90

On the Road from Ad Herculem to Rittium 18. Bronze circular coin with two extensions in the form of oppositely placed leaf-shaped ornaments. Imitation of uroboros. R 2.2 cm. Field inventory number C-26. Dated: Celtic period of domination in Srem. 19. Roman bronze coin of Constantius II. On the obverse, there is a bust of the emperor facing to the right. On the reverse: in wreath XX MVLTXXX. R 1.3 cm. Field inventory number C-50. 20. Roman coin dated in the 2nd century that was secondarily used as a rein attachment. R 2.5 cm. Field inventory number C-94.

tip. Length 10 cm, width 2.7 cm, R 2,7 cm. Field inventory number: C-64. Analogies: Radman-Livaja 2010: 64, kat. 16 (Gardun - Tilurium). Dated: Roman period. Plate III: HORSE HARNESS 1. Bronze decorative rectangular fitting, plate-shaped. The central part of the fitting consists of two leafshaped ornaments. The ornaments end with two rectangular extensions. There are two pins on the back side and a bronze rectangular sheet tile with rounded corners which is applied. Length 3.5 cm, width 1.3 cm. Field inventory number C-76. Analogies: Radman-Livaja 2004: cat. 456 (Bishop type 6, subtype d). Dated: 1st century AD. 2. Bronze fitting fastened to the belt. The shape is rectangular with a convex central part. There are circular decorations on both ends. There are two pins on the back. Width 1.0 cm, length 4.2 cm. Field inventory number C-99. Analogies: Bishop1988: 170, Fig. 56, top 8b. Dated: 1st century AD. 3. Strap fasteners (female). Bronze alloy with traces of gilding on the surface. The female „keyhole“ fastener with a simple rectangular opening in the central part of the fitting. The body is rectangular, too. At one end, there is a circular extension, perforated in the middle. Width 1.4 cm, preserved length 3.2 cm. Field inventory number C-106. Analogies: Bishop: 1988, 167, Figs. 54, type 2c. Dated: 1st century AD. 4. Circular flat phalera, perforated in the middle. Bronze alloy with traces of silver on the surface. The surface is decorated with two incised concentric circles, while the central part of the item is perforated. There are remains of two rivets on the back side with which the object was attached to the base. Two extensions (rivets) are visible on the rim of the object, too. R 3 cm, thickness 3 mm. Field inventory number C-105. Analogies: Bishop: 1988, 137, Figs. 40, type 5c. Dated: Roman period, 1st century AD. 5. Junction loop. Bronze alloy. Rectangular body. The hook has a plastic rectangular thickening at the junction. Width 1.0 cm, preserved length 4.0 cm. Field inventory number C-127. Analogies: Bishop 1988: 157, Figs. 50, closest type 5 i. Dated: 1st century AD. 6. Strap mount. The body is elongated and profiled several times, with a circular central part perforated in the middle. Half of the object is missing. R 1.9 cm, preserved length 5.1 cm. Field inventory number C-108. Analogies: Bishop 1988: 170, Fig. 56, top 6d. Dated: 1st and early 2nd century AD. 7. Strap fasteners (female). The body is profiled several times and ends with a circular extension perforated in the middle. R 1.5 cm, preserved length 3.7 cm. Field inventory number C-109. Analogies: Bishop

Plate II: MILITARY PERSONAL EQUIPMENT AND WEAPONS 1. Bronze amphora-shaped strap ends. The upper part is missing, while the lower part ends with a knob. The surface is undecorated. Preserved length 3.3 cm, width 2.0 cm. Field inventory number C-72. Analogies: Krunić, Crnobrnja 1997: 290-291, cat. 494-497 (Taurunum); Radman-Livaja 2004: 99, cat. 362. Dated: Roman period, 4th century AD. 2. Part of military equipment (belt sets) - fittings made in the casting technique. The upper part is plate, rectangular. The central part is rounded and the end is circular. The face is embossed, while the back is flat. Two iron rivets are preserved on the fitting. Length 4.00 cm, width 2.2 cm. Field inventory number C-29. Analogies: Radman Livaja 2004: 94, cat. 282-283, the second half of the 2nd and part of the 3rd century AD (iron specimens); Bishop 1988: 164, type 2, subtype b, parts of horse equipment made in the same manner are dated to 1st to 2nd century AD. Dated: 1st to the 3rd century AD. 3. Bronze belt propeller-shaped fitting. Width: 1.2 cm, length 2.4 cm. Field inventory number C-107. Analogies: Radman-Livaja 2004: 97, cat. 370-382 (Siscia). Dated: Late Antiquity. 4. Catapult bolt head of square cross-section and the typical enlogated bodkin-shape. It is socketed. Length 6.8 cm, R 1.3 cm. Field inventory number C-68. Analogies: Radman-Livaja 2010: 30, cat. no. 2, 3 (Tilurium); Radman-Livaja 1998: 223, T. I. 5; Bishop & Coulston 2006: 88-89,135, 168-170. Dated: 1st century AD. 5. Iron trilobate tangled arrowhead. Length 6 cm, width 1.2 cm, R pin: 0.4 cm. Field inventory number C-67. Analogies: Bruckner 1976: 40, T. XVI, 3; Dautova-Ruševljan 1995: 127, T. XVI, 5 (Dumbovo near Beočin - 4th century); Radman-Livaja 2001: 124, T. I, 1. Dated: from the second half of the 2nd century BC to the 4th century AD. e. During the Migration period, they become more massive and heavier. 6. Spear butt. Iron reinforcement of the spear base (butt) in the shape of the socket with the pointed 91

Biljana Lučić et al. 1988: 167, Figs. 54, closest type 1d. Dated: Roman period, 1st and early 2nd century. 8. Junction loop. Bronze alloy. R 1.4 cm, preserved length 2.2 cm. The loop is profiled several times. The circular perforated extension is attached to it. Field inventory number C-111. Analogies: Bishop 1988: 157, Fig. 50, closest type 1p; Radman-Livaja: 2009, kat. 2 (Burgenae). Dated: 1st and early 2nd century AD. 9. Leaf-shaped pendant with four perforations in the central part. The loop and the very top of the object is damaged. Width 2.6 cm, length 4.4 cm. Field inventory number C-100. Analogies: Bishop 1988: 145, Figs. 45, type 4. Dated: Roman period, from Claudius throughout the 2nd century AD. 10. Terminal bronze fittings of the decorative straps. It is the final fitting of the decorative belt. Made in the casting technique. The rectangular body is profiled in the middle part of the object. The lower part of the fitting ends with a knomb. Two pins have been preserved on the back. Length 4.1 cm, width 0.9 cm. Field inventory number C-21. Analogies: RadmanLivaja 2004: nearest cat. 453, 457, 458; Bishop 1988: 164, fig. 52. type 6, subtype e. Dated: 1st to 2nd century AD. 11. Oval-shaped bronze fitting. The central part of the fitting is convex. There is a circular loop on the longer rim. Two pins have been preserved on the back. Width 2.9 cm, height 2.1 cm. Field inventory number C-22. Remarks: stylistically corresponds to the object in the catalogue T. 3, 12 (C-14). Dated: 3rd century AD. 12. Bronze alloy oval-shaped fitting. The central part of the fitting is convex. Two pins are preserved on the back. One part of the object has not been preserved. In the upper part, the fitting is thickened in the form of a ridge. Length 4 cm, width 1.3 cm. Field inventory number C-14. Analogies: Radman-Livaja 2004: closest to the fitting from catalogue no. 536. Dated: 3rd century AD. 13. Bronze pendant made by casting. The pendant is leaf-shaped with a three-pointed end. The central part of the object is convex and oval-shaped. Length 4.7 cm, width 1.8 cm. Field inventory number C-12. Dated: 3rd to 4th century AD (?). 14. Upgraded form of a lunar bronze pendant. Earlier forms are flat, while later, like ours, have profiled bodies. The ends of the pendant are joined in the form of two smaller knobs. The lower part continues in the form of a ring. A connecting loop, with which the pendant was attached to the horse equipment, has been preserved. The surface of the object is decorated in the nielo technique ( blue and yellow glass paste). R 2.2 cm, preserved length 4.2 cm. Field inventory number C-77. Analogies: Radman-Livaja 2009: 1501, kat. 14 (Bishop, type 9 m); Bishop 1988: 154, Fig. 48. Dated: 2nd century AD and later.

15. Decorative bronze fitting with a cross-shaped head. There is a pin on the back. Widespread during the 3rd century AD (from Dura Europos to Spain). Height with pin 1.1 cm, length 2.0 cm. Field inventory number C-74. Analogies: Radman-Livaja 2004: T.74, cat. 538 (Sisak). Dated: 3rd century AD. 16. The small rectangular fitting of a hollow crosssection. Most probably it could be placed on the reins. It was used for pulling and loosening them. Height 1.4 cm, length 1.6 cm, width 1.5 cm. Field inventory number C-60. Analogies: Radman-Livaja 2004: cat. 540. Dated: 3rd century AD. 17. Oval-shaped bronze fitting with a convex circular protrusion in the middle. Two pins are preserved on the back. Width 1.5 cm, preserved length 3.0 cm. Field inventory number C-103. Analogy: RadmanLivaja 2004: cat. 537 and 536 (Siscia). Dated: 3rd century AD. 18. Bronze wedge (caneus). It has a wide plate head and a pin at the back. R: 2.7 cm, height with pin 1.7 cm. Field inventory number C-78. Analogies: Krunić. Crnobrnja 1997: 210, cat. 311 (Singidunum, castrum). Dated: 3rd to 4th century AD. 19. Bronze wedge (caneus). Similar to the previous one. R 3.3 cm, height with pin 1.1 cm. Field inventory number C-89. Plate IV: ITEMS ATTRIBUTED TO HORSE EQUIPMENT 1. Decorative bronze propeller-shaped fitting. The central part of the object is circular with a circular perforation. There are three more circular extensions radially arranged around it which are perforated in the middle, too. All four segments of the object are decorated with stamped circles. R of each segment is 1.7 cm, length 6.3 cm. Field inventory number C-79. 2. Fragment of the bronze flat fitting. One part of the fitting ends with a stylized pelta. A ridge is placed vertically along the preserved part of the item. The missing part probably ends with a semi-circle. Perforation is visible in the middle of it. There are three fastening elements at the back. Preserved length 4.3 cm, maximum width 2.4 cm. Field inventory number C-18. 3. Three oval bronze pendants with a flat circular head. Horn ornaments are applied on the top of the head. R: 0.6-0.7 cm, length 0.7-1.0 cm. Field inventory number C-24abc. 4. Part of a bronze pendant. The body of the pendant has a hexagonal cross-section. Its top is multiprofiled, with a circular cross-section and horn decoration. R 1.2 cm, length preserved 3.0 cm. Field inventory number C-56. 5. Stripe-shaped application made of a bronze alloy sheet with a circular loop at one end. There is a lunular excrescent on the rim of the object, in its 92

On the Road from Ad Herculem to Rittium central part. Height 1.8 cm, preserved length 5.4 cm. Field inventory number C-84. 6. Bronze elongated pendant with a hollow upper part and a circular cross-section. The lower part has four decorative knobs at the top. R: 1.0 cm, length 4.0 cm. Field inventory number C-95. 7. Object of unknown purpose. The top of the item is conical with an oval knob at the top. The body (implant) is cylindrical (elliptical in cross-section). R:2.0 cm, length 4.7 cm. Field inventory number C-53. 8. Round iron sheet fitting. A central part of the calotte ends with a pointed tip. Traces of a silver coating have been preserved inside of the object, which probably covered the entire surface. R 1.9 cm, length 4.7 cm. Field inventory number C-73.

7. Bronze medical or cosmetic instrument. Length 3.2 cm, width 0.3 cm. Field inventory number: C-35. 8. Bronze medical instrument with a circular knob on it. The cross-section of the body is square. Preserved length 5.2 cm, R1 0.8 cm, R2 0.1 cm. Field inventory number C-19. Analogies: Milošević 2001: 153, T. 1 (the last item). Dated: Roman period. 9. A double bronze instrument (chisel). The central part of the body has a circular cross-section, while both ends are flattened and widened, with sharp edges. Length 4.9 cm, width 0.4 cm. Field inventory number C-34. 10. Small fragment of twisted bronze bracelet/ring made of wire. Dimensions: R 4 cm. Field inventory number C-15. Analogies: Krunić, Crnobrnja 1997: cat. 150. Dated: 3rd century AD. 11. Bronze bracelet with open ends made by casting, embossing and engraving, rectangular crosssection. One end is striped and decorated with a series of horizontal lines that form fields. The other end is narrowed. Field inventory number C-36. Dated: 4th century AD. 12. Bronze bracelet made by casting, embossing and engraving. Field inventory number C-97. Dated: Roman period, 4th century AD (?). 13. Fragment of a bronze bracelet of semi-circular cross-section (probably open type). Unadorned. Dimensions R 6.2 cm. Field inventory number C-11. Analogies: Bronze open bracelets are mostly given in the 4th century, Krunić, Crnobrnja 1997: cat. 131153; Dautova-Ruševljan, Vujović 2006: 103, cat. 102, 104. Dated: Late Antiquity. 14. Fragment of a hairpin with a biconical head. Preserved length 2.9 cm, R 1 cm. Field inventory number C-13. Analogies: Krunić, Crnobrnja 1997: cat. 427 (Ušće near Obrenovac, Plantaža site). Dated: 3rd to 4th century AD. 15. Fragment of a bronze hairpin. The head is biconical. Preserved length 2.3 cm, R: 0.9 cm. Field inventory number C-16. Analogies: Krunić, Crnobrnja 1997: cat. 427 (Ušće near Obrenovac, Plantaža site). Dated: 3rd to 4th century AD. 16. Part of a multi-profiled head and a pin of the hairpin or medical instrument (?). Circular cross-section. Preserved length 1.7 cm, R 0.1 cm. Field inventory number: C-17. Analogies: Dautova Ruševljan 1995: 122, T. IV, 17 (Vizić). Dated: 3rd to 4th century AD. 17. Fragment of the head and body of a hairpin. The head, made of a golden sheet, is deformed under pressure. The body has of circular cross-section. Found with golden leaves. Preserved length: 1.9 cm, preserved width: 0.7 cm. Field inventory numbers C-20, C-23. Dated: Roman period.

Plate V: VESSELS, JEWELLERY, MEDICAL INSTRUMENTS, UTILITY ITEMS 1. Fragment of the bronze simulum handle (Aislingentype). Only the upper part of the handle has been preserved. The handle has been profiled several times. It ends in the form of a spoon with a hanging hole. Vessels of this type were used for wine serving. Preserved length 4.3 cm, width 1.8 cm. Field inventory number C-110. Analogies: Feugère, Freycon 2014: 17-19. Dated: 1st to 2nd century AD. 2. Casserole handle with a circular end which is perforated in the middle. Bronze alloy (lead is added). The circular rim of the casserole and the perforation are decorated with ribbon. Such vessels were produced in Gallic workshops. R 6.3 cm, preserved length 8.7 cm. Field inventory number C-82. Analogies: Radman-Livaja 2010: 106, cat. 111 (Gardun - Tilurium); Jovanović 2010: 196, Sl. 2, T. I, 3 (Split), type 139-144 according to Eggers. Dated: from the 1st to the middle of the 2nd century AD. 3. A fragment of the rim and wall of a smaller silver vessel (probably patera). There are two decorative grooves below the rim. The wall is convex. Length 3.4 cm, preserved height 3.0 cm. Field inventory number C-37. 4. Bronze rim and wall of the bell. The rim is slightly open and the wall is flat. R 7.0 cm, preserved height 3.1 cm. Field inventory number C-80. 5. the Smaller handle of the bronze goblet with preserved part of the wall. Length 3.2 cm, width 2.0 cm. Field inventory number C-101. 6. Bronze key with a ring head, a rectangular body on which horizontal lines are engraved. Length 4.0 cm, width 1.8 cm. Field inventory number C-92. Dated: Roman period, 2-4. century AD.

93

Biljana Lučić et al.

Plate I

94

On the Road from Ad Herculem to Rittium

Plate II

95

Biljana Lučić et al.

Plate III

96

On the Road from Ad Herculem to Rittium

Plate IV

97

Biljana Lučić et al.

Plate V

98

On the Road from Ad Herculem to Rittium Table 1 Pottery card number

Locality and its ID number

Prehistory

Roman period

Middle Ages

16-18. century

No. of pottery fragments

B.1 i B.5

Beška, Ribarica-Slankamenac (45)

98,3%

-

1,67%

-

60

B.2

Krčedin, Pod visom (39)

18,82%

81.77% (1-4. century)

1,76%

0,59%

170

B.3

Krčedin, Livade (38)

76,26 % (1-3. century)

6,47%

17,26%

139

B.4

Stari Slankamen, Počenta (2)

46,15%

5,77 %

44,23%

1,9 %

52

B.6

Beška, Lokalitet Brdo (52)

32,5 %

25,8%

29,03%

12,9%

31

B.7

Novi Slankamen, Grabovac (10)

-

3,18%

93,64%

0,35%

283

B.8

Surduk, Gradina (15)

3.46

89,3%

-

7,23%

318

B.9

Surduk, Surdučak (18)

-

100%

-

-

20

B.10

Surduk, Čolićev čot (16)

-

81,8%

9,1%

9,1 %

11

B.11

Surduk, Marina (36)

Chart

Chart

Chart

Chart

Chart

B.12

Stari Slankamen, Prva gudura (60)

21,12 %

70,3% (1-3. vek)

9,6%

0,26%

374

B.13

Stari Slankamen, Čot, Umka (3)

43,9%

8,8%

47,25%

-

91

B.14

Stari Slakamen, Potes Despotova anta (8)

33%

50%

-

16,6%

6

B.15

Novi Slankamen, Čarevac (11)

21,91%

26%

45,2%

-

73

B.16

Novi Slankamen, Crkvena (9)

-

-

-

100%

38

B.17

Krčedin, Batinić salaš (40)

-

-

100%

-

-

B.18

Krčedin, Budžak (41)

-

-

40%

60%

5

B.19

Krčedin, Kupusare (42)

8,3 %

-

41,7%

50%

12

B.20

Čortanovci, Stara Tidija 1 (56)

100%

-

-

-

19

B.21

Čortanovci, Krasovićev salaš (59)

-

-

100%

B.22

Beška, Brdo (48)

25,9%

48,15% (1-2. century)

-

18,5%

27

B.23

Beška, Ribarica, Ribnjak (47)

80% (Laten, Early Iron Age)

20%

-

-

75

B.24

Novi Karlovci, Vinjerac (20)

6,9%

9,3%

62,8%

16,3%

43

B.25

Novi Karlovci, Rupine (21)

10%

2,5 (1. century)

40%

20%

40

99

5

Biljana Lučić et al. Table 2. ID Place name/site no.

Research type Typology

Chronology

References Dimitrijević 1958: 303-305; Dimitrijević 1959: 179-181; Dimitrijević 1961: 93-107; Dimitrijević 1970: 269-277; Dimitrijević 1996: 7-38; Piletić 1957: 59-73; Piletić, Rašić 1958: 339-341; Piletić 1962: 9-24; Popović 1996: 155-168; DautovaRuševljan, Vujović 2006: 63; Đorđević 2007: 72-73; Crnobrnja 2015: 270-271, kat. 138.

1

Stari Slankamen, GRADINA

Excavations

Prehistoric settlement, Late bronze age, Early Iron Age, oppidum, Roman and Late Iron Age, Roman period, Medieval fortification Late Medieval period

2

Stari Slankamen, POČENTA

Field survey

Necropolis

3

Stari Slankamen, ČOT, HUMKA, area Umka („Oduševac?“)

Excavations

Piletić 1958: 246-247; Piletić 1961: Prehistoric settlement, Roman period, Medieval period, 91; Piletić 1964: 9; Piletić 1989: 81-82; Roman speculum and Prehistory (Eneolitic periodCrnobrnja 2015: 271, kat. 140; Lučić, De necropolis Baden cultural groupe) Rosa, Muždeka, Lemajić, Nikolić 2019: unpublished.

4

Stari Slankamen, PRAVOSLAVNO GROBLJE

Field survey

Settlement

Kovačević, D. Popović 1965, 1968, 1970, 1985: unpublished.

D. Popović 1967: 178; Lučić,. De Rosa, Medieval period, Roman period, Muždeka, Lemajić, Nikolić 2019: Prehistory, 16-18. century unpublished.

Medieval period, Eneolitic period (Baden cultural groupe), La Tene

D. Popović 1980: unpublished.

5

Stari Slankamen, MANASTIRINE

Excavations

Partlly preserveed and presented remains of the church and the Medieval period, 13-14. century town gates, graves with no grave goods

6

Stari Slankamen, KIS

Excavations

Settlement

Medieval period

Piletić 1955: unpublished.

7

Stari Slankamen, PET BUNARA

Excavations

Pit-hearts of the medieval huts

Medieval period

Piletić 1955: unpublished.

8

Stari Slankamen, HUMKA, DESPTOVA ANTA

Field survey

Settlement with necropolis

Roman period and Prehistory

Piletić: 1955; Lučić, De Rosa, Muždeka, Lemajić, Nikolić 2019: unpublished.

9

Novi Slankamen, CRKVENA

Field survey

Settlement, local Popović 1969: 252; Lučić, De Rosa, tradition cosiders that 16-18th century, Medieval period Muždeka, Lemajić, Nikolić 2019: mosque used to be unpublished. situated thare

10

Novi Slankamen, GRABOVAC

Field survey

Settlement

La Tene, Roman period, Medieval period.

11

Novi Slankamen, ČAREVAC

Field survey

Settlement

Early Iron Age (Bosut Group), La Popović 1969: 251-252. Tene, Roman, Medieval period

12

Novi Slankamen, UNKA, area Komarevac

Field survey

Settlement

Roman period: 1st-2nd century

13

Novi Slankamen, PROPUST, area Komarevac

Field survey

architectural complex Roman period

14

Novi Slankamen, BEGLUK, area Komarevac

Field survey

Settlement

15

Surduk, GRADINA, Field survey, Castellum Rittium excavations

Castrum Rittium and probably a part of the civil settlement and protohistoric settlement

Popović 1969: 252; Lučić, De Rosa, Muždeka, Lemajić, Nikolić 2019: unpublished.

D. Popović 1976: unpublished.

D. Popović 1976: unpublished.

Roman period, Prеhistory, La Tene

D. Popović 1976: unpublished.

Roman period, Early Iron Age

Piletić, Rašić 1961: 88; Dimitrijević 1961: 95; Petrović 1995: 22, sl. 10; Dimitrijević 1996: 17-18; Dautova-Ruševljan, Vujović 2006: 62; Đorđević 2007: 73; Crnobrnja 2015: 296-297, kat.197. Lučić, De Rosa, Muždeka, Lemajić, Nikolić 2019: unpublished; Lučić, Mrđić 2022: unpublished.

100

On the Road from Ad Herculem to Rittium ID Place name/site no.

Research type Typology

Chronology

References

Field survey

Prеhistоric settlement, Roman necropolis, Bronze Age, Roman period, 16settlement 16-18. 18th century century

D. Popović 1971: unpublished; Lučić, De Rosa, Muždeka, Lemajić, Nikolić 2019: unpublished.

Surduk, PRAVOSLAVNO GROBLJE

Field survey

Tumulus

Prehistory

D. Popović 1971: unpublished; Lučić, De Rosa, Muždeka, Lemajić, Nikolić 2019: unpublished.

18

Surduk, SURDUČAK

Field survey

Not determined

Bronze Age, Iron Age (La Tene), Roman period

D. Popović 1982: unpublished; Lučić, De Rosa, Muždeka, Lemajić, Nikolić 2019: unpublished.

19

Novi Karlovci, VODICE, area Mala Field survey Anta - Čupić polje

Settlement with necropolis

Late Roman period, Medieval period

Popović 1968: 222; Crnobrnja 2015: 274275, kat. 149.

20

Novi Karlovci, VINJERAC

Field survey

Settlement

Medieval period, Roman period

D. Popović 1977: unpublished; Lučić, De Rosa, Muždeka, Lemajić, Nikolić 2019: unpublished.

21

Novi Karlovci, RUPINE

Field survey

Necropolis

Medieval period, 16-18th century, Roman period - rare

D. Popović 1977: unpublished; Lučić, De Rosa, Muždeka, Lemajić, Nikolić 2019: unpublished.

22

Novi Karlovci, SUTOROV/ SUDOROV SALAŠ

Field survey

Not determined

Roman period

D. Popović 1977: unpublished; Lučić, De Rosa, Muždeka, Lemajić, Nikolić 2019: unpublished.

23

Stara Pazova, STARO SELO

Field survey and accidental finds

Settlement (villa rustica) with necropolis

Roman period

Popović 1966: 189.

24

Stara Pazova, POLJOPRIVREDNO Accidental DOBRO finds, survey “NAPREDAK”, potes JARKOVCE

Settlement - pagus

Roman period - 2nd century

D. Popović 1975: unpublished.

25

Stara Pazova, KONOPISKA, potes Field survey STARO SELO

Smaller settlement

Roman period

D. Popović 1975: unpublished.

26

Golubinci, LIVADE Field survey

Settlement, 3-4 houses Medieval period

Popović 1967: 174.

27

Golubinci, SREDNJE NJIVE ELDE

Necropolis

Roman period

Popović 1967: 174.

28

Golubinci, CIGLA Field survey TAPAVICE

Settlement (Villa rustica)

Roman period, Prеhistory, La Tene

Popović 1967: 174; Zanni, Lučić, De Rosa 2018 (in print).

29

Golubinci, FRTALJICE

Villa rustica

Roman period

D. Popović 1973: unpublished.

30

Golubinci, SELIŠTE Field survey

Settlement - pagus

Roman period

D. Popović 1973: unpublished.

31

Golubinci, PAVINA Field survey ANTA

Artificial elevation (humka)

Not determined

D. Popović 1973: unpublished.

32

Golubinci, RIMSKI Field survey PUT

Infrastructure

Roman period

Popović 1980: 101-108.

33

Donji Petrovci, ŠEVAR

Late Bronze Age - Early Iron Age

Popović 1969: 249. D. Popović 1975: unpublished.

16

Surduk, ČOLIĆEV ČOT

17

34

Donji Petrovci, BASSIANAE, area Gradina

Field survey

Field survey

Accidental find, Hoard field survey

Excavations

Civil settlement (colonia) and Roman period necropolis around the fortified town

101

Ljubić 1883: 65-70; Grbić 1935: 483486; Grbić 1936: 19-31; Grbić 1937: 1-7; Dušanić 1967: 67-81; Dautova-Ruševljan 1983:19-22, kat. 81, 82, 85, 86, 88, 92, 93, 94, 96, 102, 109, 112; Milin 2004; DautovaRuševljan, Vujović 2006: 66; Đorđević 2007: 45-48; Crnobrnja 2015: 284-285, kat.172; Filzwieser, Ivanišević et all 2021: 1-29.

Biljana Lučić et al.

ID Place name/site no.

Research type Typology

Chronology

References

Roman period

Piletić 1961: 91; Crnobrnja 2015: 271-272, kat. 141.

35

Stari Slankamen, DUGOREP

Excavations

Speculum or tower

36

Surduk, MARINA, EASTERN AND WESTERN SIDE

Field survey

Stojić 2004: 129-143; Eneolithic period, Bronze Age, Prehistoric settlement, Stojić 2006: 21-28; Lučić, De Rosa, Iron Age, Roman period, 16-18th Roman necropolis Muždeka, Lemajić, Nikolić 2019: century unpublished.

37

Krčedin, RUPA KOD GROBLJA

Field survey

pits

La Tene period

D. Popović 1971: unpublished.

38

Krčedin, LIVADE

Field survey

Settlement

Roman period, 16-18th century

D. Popović 1971: unpublished; Lučić, De Rosa, Muždeka, Lemajić, Nikolić 2019: unpublished.

39

Krčedin, POD VISOM

Field survey

Architectural complex

Roman period, Medieval period and Prehistory

D. Popović 1975: unpublished; Lučić, De Rosa, Muždeka, Lemajić, Nikolić 2019: unpublished.

40

Krčedin, BATINIĆ SALAŠ

Field survey

Settlement?

Medieval period

D. Popović 1975: unpublished; Lučić, De Rosa, Muždeka, Lemajić, Nikolić 2019: unpublished.

41

Krčedin, BUDŽAK Field survey

Settlement?

Prehistory and Medieval period

Milošević 1972: 164. D. Popović 1975: unpublished; Lučić, De Rosa, Muždeka, Lemajić, Nikolić 2019: unpublished.

42

Krčedin, KUPUSARA

Field survey

Settlement?

Prehistory and Medieval period

D. Popović 1975: unpublished; Lučić, De Rosa, Muždeka, Lemajić, Nikolić 2019: unpublished.

43

Krčedin, BARAKA

Field survey

Settlement?

Prehistory and Medieval period

D. Popović 1975: unpublished.

44

Krčedin, VIS

Field survey

Settlement

Early Iron Age/ Bosut cultural group

Popović 1981: 15-16, T. IV, 8, T. V, 4.

45

Beška, RIBARICA – Field survey SLANKAMENAC

Settlement

Eneolitic period (BadenKostolac cultural group)

D. Popović 1972: unpublished; Lučić, De Rosa, Muždeka, Lemajić, Nikolić 2019: unpublished.

46

Beška, BREST

Necropolis

Roman period

Manojlović 1966: 138–139; Manojlović 1987; Đorđević 2007: 70; Crnobrnja 2015: 268, kat. 136.

47

Beška, RIBARICA Field survey RIBNJAK

Multilayered settlement

Eneolitic period (Baden and Kostolac cultural group), Bronze D. Popović 1972: unpublished; Lučić 2018: Age, Early Iron Age, La Tene, unpublished. Roman period

48

Beška, BRDO

Settlement - villa rustica?

Early Roman period

D. Popović 1972: unpublished; Lučić 2018: unpublished.

49

Beška, KALAKAČA Excavations

Settlement

Early Iron Age

Medović 1988; Jevtić 2011.

50

Beška, ČOKLJA

Excavations, field survey

Settlement

Prehistory, Roman, Medieval period, 16-18. century

D. Popović 1972: unpublished; Lučić 2019: unpublished.

51

Beška, BELA LAĐA Field survey

Settlement

Praehistory/ neolitic

D. Popović 1982: unpublished.

52

Beška, BRDO PREKO PUTA LOKALITETA BREST

Field survey

Settlement

Praehistory, Roman period, Medieval period, 16-18. century

Lučić, De Rosa, Muždeka, Lemajić, Nikolić 2019: unpublished.

53

Čortanovci, PROVALIJE

Field survey

Settlement

Eneolitic period (Vučedol D. Popović 1971: unpublished. cultural group); Medieval period

54

Čortanovci, MIHALJEVAČKA ŠUMA, AREA PROSJANICE

Excavations

Field survey

Excavations

Settlement with necropolis/ fortification

Roman period- 4th century

102

Manojlović 1961: 100-101; Manojlović 1962: 187-189; Manojlović 1962a: 123125; DautovaRuševljan, Petrović 1995: 13-33; Petrović 1995: 23-24; Bunardžić 2006: 37-53; Dautova-Ruševljan, Vujović 2006: 63; Đorđević 2007: 69-70; Crnobrnja 2015: 265-266, kat. 131.

On the Road from Ad Herculem to Rittium

ID Place name/site no.

Research type Typology

Chronology

References

55

Čortanovci, LAGUM

Field survey

Settlement

Late Medieval period

D. Popović 1971: unpublished.

56

Čortanovci, STARA Field survey TIDIJA I

Settlement

Prehistory - neolitic (Starčevo and Vinča cultural group)

D. Popović 1973: unpublished; Lučić, De Rosa, Muždeka, Lemajić, Nikolić 2019: unpublished.

57

Čortanovci, STARA Field survey TIDIJA II

Settlement

Early Iron Age (Bosut cultural group), Roman period

D. Popović 1973: unpublished.

58

Čortanovci, KRASOVIĆEV SALAŠ

Field survey

Settlement and necropolis (next to the Medieval period (Hungarian) site HUMKA)

59

Čortanovci, HUMKA

Field survey

Tumulus? and necropolis

D. Popović 1973: unpublished; Lučić, De Prehistoric and Medieval period Rosa, Muždeka, Lemajić, Nikolić 2019: unpublished.

60

Surduk, PRVA GUDURA

Field survey

Settlement

Roman period, Medieval period, Prehistory, 16-18th century

Lučić, De Rosa, Muždeka, Lemajić, Nikolić 2019: unpublished.

61

Belegiš, BUDOVAR, Field survey AREA BRESTOVI

Not Defined

Roman period

D. Popović 1972: unpublished.

62

Belegiš, GRADAC

Excavations, field survey

Settlement and necropolis

Early Iron age, Early bronze age, Tasić 1962: 47-48. Eneolithic period

63

Belegiš, ŠANČINE

Field surveys, excavations

Multilayered settlement

Eneolithic periodBodrogkereštur, Vučedol and Vatin cultural group

64

Belegiš, STOJIĆA GUMNO

Excavations

Prehistoric necropolis, early medieval kilns - Prehistoric and Medieval period Tasić 1963: 50-52. Slavic pottery

65

Belegiš, DUVARINE

Field survey

Multilayered settlement

Prehistory, Roman period and 16-18th century

D. Popović 1972: unpublished.

66

Belegiš, Field survey JUGOVIĆEV BREG

Multilayered settlement

Prehistory: eneolitic period - Vučedol cultural groupe, Medieval period, 16-18th century

D. Popović 1972: unpublished.

67

Belegiš, ŠMICIN BREG

Field survey

Multilayered settlement

Prеhistory and Medieval period

D. Popović 1972: unpublished.

68

Belegiš, RADNOVIĆ BREG

Field survey

Smaller settlement

Prehistory, Roman period, Medieval period

D. Popović 1972: unpublished.

69

Belegiš, BRODIĆ

Field survey

settlement

Prehistory, Medieval period

D. Popović 1972: unpublished.

70

Belegiš, BROD

Field survey

Necropolis

Prehistory/ Iron age and Bronze D. Popović 1973: unpublished. age

71

Stari Banovci, BELI BREG

Field survey

Settlement

Early Iron age - Bosut cultural group, Roman period

72

Stari Banovci, BREG PODUNAVLJE

Field survey

Prehistory/ Eneolitoc period: Settlement, necropolis Baden-kostolac, Vučedol; Roman D. Popović 1973: unpublished. period

73

Stari Banovci, PETRINJCI

Field survey

Multilayered settlement

74

Novi Banovci, GRADINA PURGER

Excavations, field survey

75

Novi Banovci, PRNJAVOR

76

Novi Banovci, DJERVIN

D. Popović 1973: unpublished; Lučić, De Rosa, Muždeka, Lemajić, Nikolić 2019: unpublished.

Tasić 1961: 34-38.

D. Popović 1973: unpublished.

Prehistory, Roman period, 1618th century

D. Popović 1973: unpublished.

Roman fortification Burgenae

Roman period

Brunšmid 1895: 172-180; Piletić, Rašić 1958: 338-339; , Piletić, Rašić 1961: 93-94; Popović 1969: 251; Piletić 1972; Petrović 1995: 21; Dautova-Ruševljan, Vujović 2006: 61-62; Đorđević 2007: 75; Crnobrnja 2015: 303-304, kat. 213.

Field survey

Necropolis

Roman period

Popović 1969: 251.

Field survey

Settlement

Prehistory - Late bronze age, Medieval period

D. Popović 1969: unpublished.

103

Biljana Lučić et al. Bunardžić, R. 2006. Ka ubikaciji Acumincuma, Glasnik Muzeja grada Novog Sada 1 (2005): 37-53. Crnobrnja, A. 2015. Antička naselja, fortigikacije i komunikacije u jugoistočnom delu Donje Panonije. Unpublished PhD dissertation, University of Belgrade. Ćurčić, S. 2012. Formiranje mreže naselja na teritoriji savremene Vojvodine, in Lj. Krkljuš (ed.) Atlas naselja Vojvodine - Knjiga I, Srem: 43-71. Novi Sad: Matica Srpska. Dautova-Ruševljan, V. 1981. Ostave varvarskog, rimskog i vizantijskog novca iz Vojvodine. Numizmatičar 4: 64-68. Dautova-Ruševljan, V. 1983. Rimska kamena plastika u jugoslovenskom delu provincije Donje Panonije. Novi Sad: Vojvođanski muzej savez arheoloških društava jugoslavije. Dautova-Ruševljan, V. 1987. Numizmatički nalazi i trgovčki promet, in M. Stojanov (ed.) Počeci romanizacije u jugoistočnom delu provincije Panonije: 45-64. Novi Sad: Matica srpska. Dautova-Ruševljan, V. and P. Petrović 1995. Rimsko utvrđenje kod Čortanovaca. Radovi vojvođanskih muzeja 36: 13-33. Dautova-Ruševljan, V. 1995. Zanatski proizvodi i numizmatički nalazi, in N. Tasić (ed.) Fruška gora u antičko doba: 111-169. Novi Sad: Matica srpska. Dautova-Ruševljan, V. and M. Vujović 2006. Rimska vojska u Sremu. Novi Sad: Muzej Vojvodine. Dimitrijević, D. 1958. Rad Zemunskog muzeja na rimskom limesu/Reserches du Musée National de Zemun sur le limes romain. Starinar VII-VIII: 303305. Dimitrijević, D. 1959. Gradina u Starom Slankamenu/ Gradina a Stari Slankamen. Starinar IX-X: 179-181. Dimitrijević, D. 1961. Nekoliko podataka o rimskom limesu u istočnom Sremu, in M. Grbić (ed.) Limes u Jugoslaviji I: 93-107. Beograd: Arheološko društvo Jugoslavije. Dimitrijević, D., J. Kovačević and Z. Vinski 1962. Seoba naroda; Arheološki nalazi jugoslovenskog Podunavlja. Zemun: Narodni muzej. Dimitrijević, D. 1969. Istraživanja rimskog limesa u istočnom Sremu s posebnim osvrtom na pitanje komunikacija. Osiječki zbornik 12: 87-121. Dimitrijević, D. 1970. Protoistorijski i antički Acumuncum. Zbornik za likovne umetnosti 6: 269-277. Dimitrijević, D. 1996. Pristaništa Classis Flaviae Pannonicae na Basijanskom sektoru Dunava. Zbornik Matice Srpske za istoriju 54: 7-38. Dušanić, S. 1967. Bassianae and Its Territory. Archaeologia Iugoslavica VIII: 67-81. Dušanić, M. 1988. Rimske opeke sa žigom sa panonskog limesa (odsek: Cuccium-Taurunum), Starinar 39: 8597. Dušanić, M. 1995. Stanovništvo Fruškogorske oblasti prema epigrafskim izvorima, in N. Tasić (ed.) Fruška

Bibliography Bishop, M.C. 1988. Calvary equipment of the Roman army in the first century A. D., in J.C. Coulston (ed.) Military Equipment and the Identity of Roman Sоldiers. Proceedings of the Fourth Roman Military Equipment Conference (British Archaeological Reports International Series 394): 67-195. Oxford: Arhaeopress. Bishop, M.C. and J.C.N. Coulston 2006. Roman Military Equipment: From the Punic Wars to the Fall of Rome. Oxford: OxbowBooks. Bojović, D. 1983. Rimske fibule Singidunuma. Beograd: Muzej grada Beograda. Brukner, O. 1982. Žirovac, Ruma - rimski put, Arheološki pregled 23: 95-97. Brukner, O. 1987. Importovana i panonska keramička produkcija sa aspekta društveno- ekonomskih promena, in M. Stojanov (ed.) Počeci romanizacije u jugoistočnom delu provincije Panonije: 23-44. Novi Sad: Matica Srpska. Brukner, O. 1988. Rimski grad Basijana - Podaci iz antičkih izvora i literature. Građa za proučavanje spomenika kulture Vojvodine 15: 89-94. Brukner, O. 1995. Importovana i panonska keramika, Prilog topografskim istraživanjima Fruške gore, in N. Tasić (ed.) Fruška gora u antičko doba: prilozi za staru istoriju i arheologiju: 57-119. Novi Sad: Matica srpska. Brukner, O. 2015. Gradske teritorije kolonija Sirmijum i Basijana, in O. Brukner and V. Dautova-Ruševljan (eds) Rimska poljoprivredna imanja - vile rustike u jugoistočnom delu provincije Panonije: 21-23. Novi Sad: Matica srpska.  Brunšmid, J. 1895. Arheološke bilješke iz Dalmacije i Panonije. Vjesnik Arheološkog muzeja u Zagrebu: 1 (1): 148-183. Brunšmid, J. 1901. Arheološke bilješke iz Dalmacije i Panonije IV. Vjesnik Arheološkog muzeja u Zagrebu 5 (1): 87-168. Brunšmid, Ј. 1902. Praistorijski predmeti iz Sremske županije. Vjesnik Arheološkog muzeja u Zagrebu 6: 6886. Brunšmid, J. 1904. Kameni spomenici hrvatskoga narodnoga muzeja u Zagrebu. Vjesnik Arheološkog muzeja u Zagrebu 7 (1): 209-240. Brunšmid, J. 1907. Kameni spomenici hrvatskoga narodnoga muzeja u Zagrebu. Vjesnik Arheološkog muzeja u Zagrebu 9 (1): 82-184. Brunšmid, J. 1911, Kameni spomenici hrvatskoga narodnoga muzeja u Zagrebu. Vjesnik Arheološkog muzeja u Zagrebu 11 (1): 64-144. Brunšmid, J. 1912. Nekoliko našašća novaca na skupu u Hrvatskoj i Slavoniji XXXIV do XXXVI. Vjesnik Arheološkog muzeja u Zagrebu 12 (1): 260-286. Brunšmid, J. 1914. Antikni figuralni bronsani predmeti u hrvatskom narodnom muzeju u Zagrebu. Vjesnik Arheološkog muzeja u Zagrebu 13 (1): 207-268. 104

On the Road from Ad Herculem to Rittium gora u antičko doba: 33-56. Novi Sad: Matica srpska. Đorđević, M. 2007. Arheološka nalazišta rimskog perioda u Vojvodini. Beograd: Institute for the Protection of Cultural Monuments of Serbia. Ferjančić, S. 2002. Naseljavanje legijskih veterana u Balkanskim provincijama I-III vek n. e. Beograd: Srpska akademija nauka i umetnosti, Balkanološki institut. Feugère, M. and Fl. Freycon 2014. Les manches de substitution des simpulums de type Aislingen. Instrumentum 40: 17-19. Filzwieser, R. et al. 2021. Integrating Geophysical and Photographic Data to Visualize the Quarried Structures of the Roman Town of Bassianae. Remote Sensing 13 (12): 1-29. https://doi.org/10.3390/ RS13122384 Gorenc, M. and D. Dimitrijević 1956. Gradina u St. Slakamenu. Rad Vojvođanskih muzeja 5: 150-154. Gračanin, H. 2010. Rimske prometnice i komunikacije u kasnoantičkoj Južnoj Panoniji. Scrinia Slavonica 10: 9-69. Grbić, M. 1935. Izveštaj o radu na iskopavanju rimske Basijane kod Donjih Petrovaca, Glasnik Istorijskog društva Novi Sad VIII-3: 483-486. Grbić, M. 1936. Arhitektura u Basijani (Sremski Petrovaci) I. Rekonstrukcija mesta prema situacijonom planu i avionskom snimku. Glasnik Istorijskog društva Novi Sad IX-1: 19-31. Grbić, M. 1937. Arhitektura u Basijani kod Donjih Petrovaca u Sremu II. Otkopane građevine u 1935. godini. Glasnik Istorijskog društva Novi Sad X-1: 1-17. Grbić, M. (ed.) 1961. Zbornik radova sa simpozijuma o Limesu 1960. godine, Limes u Jugoslaviji I. Beograd: Arheološko društvo Jugoslavije. Hoffiller, V. 1912. Oprema rimskoga vojnika u prvo doba carstva II. Vjesnik Arheološkog muzeja u Zagrebu 12 (1): 16-123. Hoffiller, V. 1935. Novi tračko-mitrički votivni relijefi. Vjesnik Arheološkog muzeja u Zagrebu 16 (1): 61-66. Jevtić, M. 2011. Čuvari žita u praistoriji. Vršac-Beograd: Gradski muzej Vršac and Filozofski fakultet Beograd. Јоvanović, A. 2005. Archaeological notes of the fortifications from Pannonia Secunda, in M. Mirković (ed.) Römische Städte und Festungen an der Donau: 83-87. Beograd: Filozofski fakultet. Jovanović, J. 2010. Rimsko brončano posuđe iz Arheološkog muzeja u Zagrebu. Vjesnik za arheolologiju i povijest dalmatinsku 103 (1): 191-232. Klemenc, J. 1961. Limes u Donjoj Panoniji, in M. Grbić (ed.) Limes u Jugoslaviji I: 5-34. Beograd: Arheološko društvo Jugoslavije. Korać, M., S. Golubović, N. Mrđić, G. Jeremić and S. Popo-Lazić 2014. Granice rimske imperije - rimski limes u Srbiji. Beograd: Arheološki institut Beograd. Kovacs, P. 2013. Territoria, pagi and vici in Pannonia, in W. Eck and P. Kovács (eds) Studia epigraphica in memoriam Géza Alföldy: 131-154. Bonn: Habelt Verlag. Kovačević, J. and D. Dimitrijević 1965. Stari Slankamen

(Acumincum), Inđija, Praistorijsko, antičko i srednjovekovno nalazište. Arheološki pregled 7: 115120. Krunić, S. (ed.) 1997. Antička bronza Singidunuma (katalog izložbe). Beograd: Muzej grada Beograda. Libertini, G. (ed.) 2019. Liber coloniarum, The book of the Colonies, Naples: Istituto di tudi Atellani, Frattamaggiore. Lučić, B. 2017. Roman viability through Srem region, In: S. Zanni (ed.) Proceedings of the Colloque international, Tiens bien la route! Routes, agglomerations et territories antiques et medievaux, 29-30 novembre 2017, Universite Bordeaux Montaigne (forthcoming). Ljubić, Š. 1879a. Dragulj iz predhistoričke dobe. Vjesnik Arheološkog muzeja u Zagrebu 1 (1): 6-12. Ljubić, Š. 1879b. Rittium (Surduk). Vjesnik Arheološkog muzeja u Zagrebu 1 (1): 97-101. Ljubić, Š. 1883. Arkeologička izkapanja na Petrovačkoj gradini u Sriemu, gdje tobož starorimska Bassianis (II). Vjesnik Arheološkog muzeja u Zagrebu 5 (1): 65-70. Manojlović, M. 1961. Prosjanice - Čortanovci, utvrđеnje. Arheološki pregled 3: 100-101. Manojlović, M. 1962a. Prosjanice-Čortanovci, rimsko utvrđrnje. Arheološki pregled 4: 187-189. Manojlović, M. 1962b. Rimsko utvrđenje kod Čortanovaca, Izveštaj o iskopavanju u 1961. godini. Rad Vojvođanskih muzeja 11: 123-125. Manojlović, M. 1966. Beška, Inđija, lok. “Brest” - antička nekropola. Arheološki pregled 8: 138-139. Manojlović, M. 1987. Rimska nekropola kod Beške u Sremu. Novi Sad: Vojvođanski muzej..  Marcu, F. and G. Cupcea 2011. The topography of Colonia Ulpia Traiana Aususta Dacica Sarmizegetusa and the first centuration in Dacia. Archäologisches Korrespondenzblatt Jahrgang 41(4): 543-560. Mainz: Römisch-Germanisches Zentralmuseum, Forschungsinstitut für Vor- und Frühgeschichte. Medović, P. 1988. Kalakača. Naselje starijeg gvozdenog doba. Novi Sad. Milošević, P. 1969, Arheološki pregled 11: 199-201. Milošević, P. 1972. Rekognosciranje rimskog puta Cuccium-Acumincum. Arheološki pregled 14: 161-164. Milošević, P. 2001. Arheologija i istorija Sirmijuma. Novi Sad: Matica srpska. Mirković, M. 1971. Sirmium-Its History from the I Century A.D. to 582 A.D. Sirmium I: 5-94. Mirković, M. 2006. Sirmium, istorija rimskog grada od I do kraja VI veka. Sremska Mitrovica: Blago Sirmijuma. Mirković, M. 2017. Sirmium, Its History from the First Century AD to 582 AD. Novi Sad: Filozofski fakultet. Mocsy, A. 1974. Pannonia and Upper Moesia, A History of the Middle Danube Provinces of the Roman Empire. London and Boston: Routledge & Kegan. Petković, S. 2010. Rimske fibule u Srbiji od I do V veka n. e. Beograd: Arheološki institut. Petrović, P. 1986. Odbrambeni sistemi u antici (jugoistočni sektor), in P. Medović (ed.) Odbrambeni 105

Biljana Lučić et al. sistemi u praistoriji i antici na tlu Jugoslavije: 91-100. Novi Sad: Savez arheoloških društava Jugoslavije. Petrović, P. 1995. Rimski limes na Dunavu u Donjoj Panoniji, in N. Tasić (ed.) Fruška gora u antičko doba: 9-32. Novi Sad: Matica Srpska. Piletić, D. 1957. Arheološka istraživanja na rimskom limesu u Starom Slankamenu, Vesnik Vojnog muzeja 4: 59-73. Piletić, D. 1958. Rekognosciranje rimskog limesa i njegove neposredne pozadine na terenu u okolini Starog Slankamena. Vesnik Vojnog muzeja 5: 245-252. Piletić, D. and B. Rašić 1958. Orijentaciono rekognosciranje dunavskog limesa na itinereru Novi Banovci-Sremski Karlovci, Vesnik Vojnog muzeja 5: 337-342. Piletić, D. and B. Rašić 1961. Pregled radova Vojnog muzeja JNA na limesu od Novih Banovaca do Sremskih Karlovaca, in M. Grbić (ed.) Limes u Jugoslaviji I: 87-92. Beograd: Arheološko društvo Jugoslavije. Piletić, D. 1962. Neki novi momenti u istraživanju Slankamena 1958. godine. Vesnik Vojnog muzeja 6-7: 9-24. Piletić, D. 1964. Speculum na lokalitetu Humka-primer rimskog stražarskog objekta na donjopanonskom Limesu. Vesnik Vojnog muzeja 10: 9-22. Piletić, D. 1972. Preliminarni rezultati istraživanja rimskog kastruma Burgenae u Novim Banovcima 1971. godine. Vesnik Vojnog muzeja 18: 7-16. Piletić, D. 1989. Rimske fortifikacije na teritoriji SR Srbije. Vesnik Vojnog muzeja 33: 79-104. Popović, D. 1966. Rekognosciranje u Sremu. Arheološki pregled 8: 186-190. Popović, D. 1967. Rekognosciranja u Sremu. Arheološki pregled 9: 172-180. Popović, D. 1968. Rekognosciranje u Sremu. Arheološki pregled 10: 215-224. Popović, D. 1969. Rekognosciranje u Sremu. Arheološki pregled 11: 249-259. Popović, D. and M. Vasiljević 1969. Rekognosciranje i sondiranje rimskog puta Sirmium-Bassianae. Arheološki pregled 11: 261-262. Popović, D. 1975. Ostava iz Šimanovaca, in M. Garašanin and N Tasić (eds) Praistorijske ostave u Srbiji I Vojvodini I: 43-52. Beograd: Srpska akademija nauka i umetnosti. Popović, D. 1980. Glavna antička komunikacija u Sremu u svetlu arheoloških istraživanja, in I. Curk (ed.) Putevi i komunikacije u antici (referati održani na Simpozijumu Praistorijske i Antičke sekcije Saveza arheoloških društava Jugoslavije, Peć 1978), Materijali XVII: 101-108. Priština: Savez arheoloških društava Jugoslavije, Musej Kosova. Popović, D. 1981. Keramika starijeg gvozdenog doba u Sremu. Beograd: Savez arheoloških društava. Popović, D. 1993. Stanovništvo mlađeg gvozdenog doba u Sremu. Sunčani sat 3: 117-125.

Popović, D. 1996a. Pregled zaštitnih iskopavanja u Sremu. Spomenica 1961-1996: 53-92. Popović, D. 1996b. Stanje istraženosti arheoloških nalazišta u Sremu. Zbornik Muzeja Srema 2: 9-24. Popović, D. 1997. Arheološko nasleđe na tlu Rume i šire okoline. Zbornik Zavičajnog muzeja Ruma 1: 7-24. Popović, I. 2012. Sirmium: Mermerne sculpture/Marble Sculptures. Beograd-Sremska Mitrovica: Arheološki institut - Blago Sirmijuma. Popović, M. 1996. La fortresse de Slankamen, Starinar 46: 155-168. Radman-Livaja, I. 1998. Rimska streljačka oprema nađena na Gardunu kod Trilja, Opuscula Archaeologica 22: 219-231. Radman-Livaja, I. 2001. Rimski projektili iz Arheološkog muzeja u Zagrebu. Vjesnik Arheološkog muzeja u Zagrebu 34 (1): 123-152. Radman-Livaja, I. 2005. Militaria Sisciensia nalazi rimske vojne opreme iz Siska u fundusu Arheološkog muzeja u Zagrebu. Zagreb: Arheološki muzej Zagreb Radman-Livaja, I. 2008. Roman belt-fittings from Burgenae. Journal of Roman military equipment studies 16: 295-308. Radman-Livaja, I. 2009. Roman horse harness fittings from Burgenae. Anejos de Gladius 13: 1499-1508. Radman-Livaja, I. (ed.) 2010. Nalazi rimske vojne opreme u Hrvatskoj/Finds Of The Roman Military Equipment (Exhibition Catalogue). Zagreb: Arheološki muzej Zagreb Soproni, S. 1980a. Limes, in A. Lengyel and G. T. B. Radan (eds.) The Archaeology of Roman Pannonia: 219-237. University Press of Kentucky/Akadémiai Kiadó Budapest. Soproni, S. 1980b. Roads, in A. Lengyel and G. T. B. Radan (eds.) The Archaeology of Roman Pannonia: 207-217. University Press of Kentucky/Akadémiai Kiadó Budapest. Savić, N. 2020. Rimsko utvrđenje Cusum na Petrovaradinskoj tvrđavi. Godišnjak Muzeja Grada Novog Sada 15: 33-43. Stojić, M. 2004. Surduk - Lokalitet Marina. Višeslojno naselje iz eneolota, bronzanog doba i gvozdenog doba. Rad Vojvođanskih muzeja 46: 129-143. Stojić, M. 2006. Nekoliko novih arheoloških nalaza iz Srema i Banata. Rad Vojvođanskih muzeja 47/48: 2128. Tadin, Lj. 1979. Sitna rimska bronzana plastika u jugoistočnom delu provincije Panonije. Beograd: Savez arheoloških društava Jugoslavije. Tasić, N 1961. Šančine, Belegiš, Stara Pazova – Vučedolsko naselje. Arheološki pregled 3: 34-38. Tasić, N. 1962. Belegiš, Stara Pazova – Praistorijsko naselje i nekropole. Arheološki pregled 4: 47-48. Tasić, N. 1963. Stojića gumno, Belegiš, Stara Pazova – nekropola ravnih polja sa urnama. Arheološki pregled 5: 50-52. 106

On the Road from Ad Herculem to Rittium Gregl, Z. (ed.) 2013. Transparentna ljepota: staklo iz Hrvatskih muzeja - od prapovijesti do srednjeg vijeka (Exhibition Catalogue). Zagreb: Arheloški muzej Zagreb. Vinski, Z. 1957. Arheološki spomenici velike seobe naroda u Srijemu. Ljubljana: Narodni muzej Slovenije; Opseg 2 iz Situla (glasnik Narodnega Muzeja v Ljubljani). Vujović, M. 2003. Kasnorimske pojasne falere na tlu Srbije, in: Rad Dragoslava Srejovića na istraživanju antičke arhologije, Zbornik radova 2: 211-223. Kragujevac: Centar za naučna istraživanja Srpske akademije nauka i umetnosti univerziteta u Kragujevcu. Vujović, M. and J. Cvijetić 2018. The slings and arrows of outrageous fortune: long-range fighting units

on the Roman limes in Serbia, in M. Korać, S. Golubović and N. Mrđić (eds) VIVERE MILITARE EST: From Populus to Emperors - Living on the Frontier: 35-63. Belgrade: Institute of Archaeology. Zanni, S., B. Lučić and A. De Rosa 2018. Seek and ye shall find. A spatial approach to mapping Roman roads and buried archaeological sites in Srem region. The case study of Tapavice site, in Proceedings of the 24. Limes Congress, ViminaciumBelgrade, September 2018. Belgrade: Archaeological Institute (forthcoming). Zanni, S. and A. De Rosa 2019. Remote sensing analyses on Sentinel-2 images: looking for Roman roads in Srem region (Serbia). Geosciences 9 (1) 25: 1-18.

107

Lifeline of the Frontier: The Road and Port Network and the Concept of Supply and Distribution in Moesia Superior Nemanja Mrđić and Angelina Raičković Savić Abstract: Defining the relationship between the Danube as a major river route and adjacent limes road with all its crossroads and shortcuts proved to be a challenging task. Both military and civilians were focused on both routes exploiting them to the maximum. From the earliest decades after forming Moesia as a province, legions were tasked to build roads through some of the harshest terrains. The importance of these actions can be traced through time with tablets, both imperial and legionary celebrating these enormous undertakings necessary for the frontier to function. Placing and coordination of military distribution centres were crucial for the supply of garrisons in smaller forts that had no facilities for long-term storage of food. In theory, major routes are simple and follow the course of a river. In the field, the situation is far more complicated. Viminacium as a case study had its network of roads directed towards all directions with multiple roads intersecting and merging. Following the distribution of villas, cemeteries and production centres for almost 20 years to this date-a vivid network came to life. The city and legionary fortress were in the centre of this road network with Mlava and Danube to encircle and open it worldwide towards both east and west.

Limes was the best economically developed region in Moesia Superior with most of its production potential focused on the Danube for distribution. Local production of pottery and coinage from Viminacium mint found its way to remote provinces along this line. KEYWORDS: ROMAN LIMES, ROMAN FRONTIERS, ROMAN ARMY, ROMAN PORTS, ROMAN ROADS, SUPPLY, LOGISTICS, DANUBE, IRON GATE.

Introduction

Port network

During the last decades, within Roman provincial archaeology, great attention has been paid to supplying the army. The concept of supply in Moesia Superior was not studied thoroughly enough. In this sense, this paper could represent an introduction to further study of the supply and transport of goods and other products. During the past few decades in Serbia, the question of supplying Roman troops in Moesia was most studied by Petar Petrović (Petrović 1980; Petrović 1983; Petrović 1991) and Viminacium research team (Ilić et al. 2011).

Generally, in Moesia, any commander would expect more ports judging by the importance of river transport and supply. The main concept of the classis/fleet to transport and supply is designed on the two bases. The first one is the force and tactics of opponents-there is no force of battleships to fight. The lack of opponents’ battleships is more than clear. Invasion troops have the core force of dugouts/monoxylons. The second one is the pure need to maintain their defensive system.

The backbone of the military supply system was the classis-navy. Ships could carry large quantities of provisions to all military posts along the river more effectively than any other means available on the frontier. But river navigation was subject to numerous potential problems that could slow, obstruct or fully stop supply. Can be interfered with by both human and natural factors. Bad weather and harsh winters with frozen rivers proved to make this means of transport highly unreliable. This fact brings back the focus on the limes road that in these situations had to take over the full circulation of men and supplies.

The Discovery of the flat bottom ships-barges that are in use even today proves the practical side of Danube navigation with the deployment of the ship type that doesn’t require complicated port infrastructure. These barges with low-depth gauges and flat bottoms were capable of landing almost anywhere along the Danube banks. Protected landing zones by the military posts, with low sandy banks, were all that was needed to load or unload, embark, or disembark cargo, provisions, and man. The flexibility of this concept was wonderful, simply functional and practical. If the river was not frozen, and there were no heavy winds that could make the river the backbone of the frontier maintenance. Within the project “Living Danube Limes” one of the activities was a connecting cruise of the Lusoria-type

Life Along Communication Routes (Archaeopress 2023): 108–116

Lifeline of the Frontier ship. Authors participated in this journey and had firsthand experience in the navigation with the rowing boat on the Danube.

Smaller ports were located in Tekija (Transdierna), Čezava (Novae), Hajdučka vodenica, Karataš (Statio Cataractarum Dianae), Brza Palanka (Egeta) and Kusjak near Prahovo (Aquae). Viminacium, Aquae and Novae had ports separated from the main fortification, while on the other sites, ports were situated next to the fortifications and were protected with perimetral walls relying on the main defensive wall of each of the fortifications. Such positioning of ports is known from Singidunum, Hajdučka vodenica, Diana, Egeta. In German literature, such fortified ports are described as Landeburg and they are often encountered along the Rhine, although less along the middle and the lower Danube valley (Petrović 1991: 207–216).

The good and bad sides of the different ship types were clear and understandable. Although with better combat characteristics Lusoria ships showed numerous problems that appeared during this journey. Limited docking capabilities made it hard to stop on many occasions. With the strong currents of the Danube, it is logical to assume that all heavy transport was dominantly directed downstream while the ships came back empty upstream. The main supply centre could be identified in Horreum Margi, located in the relatively safe hinterland and with the Morava River as the crucial transport route. Today, Morava is not considered to be a navigable river, only some 8 km upstream from its mouth to the Danube. But with the gauge of the Roman barges, we see no obstacle for the shipping to normally and regularly circulate between Horreum Margi (Ćuprija) and Margum (Dubravica). Other supply centres are identified at Porečka reka and Mala Vrbica Konopište near Pontes (Kostol) and modern-day Kladovo. Both legionary fortresses at Singidunum and Viminacium could also be considered to be regional supply and distribution centres.

Road network The Limes road is crucial communication for the frontier. It was set up to follow the river in the best possible way and connect military posts. Serbia has poorly explored Roman roads as until recently this was not an attractive topic for research, except for Iron Gate roads, always attractive with their unique concept. As it was noted road was traced to be as close to the Danube on high banks that were not endangered by flooding. At the approaches to Viminacium it had large and deep ditches on both sides for water drainage, and

Figure 1: Zone of Viminacium as the intersection of the major land routes (Basemap: Google Earth, author: N. Mrđić)

109

Nemanja Mrđić and Angelina Raičković Savić

Figure 2: Limes section between Taliata and Egeta with major land routes along the river and over the Miroč mountain (Basemap: Google Earth, author: N. Mrđić)

even at sections where pavement or substructions were missing the path could be easily followed by following these double ditches. But according to the itineraries on two locations, an exception was made and the main road went the shortest route leaving the banks of the Danube. Leaving the Viminacium if one should follow the Danube route, he would travel Viminacium Lederata - Pincum. But Lederata is located on the tip of the peninsula. Travelling to Dacia one should go in this direction and cross the Danube into the neighbouring province. The Danube route would be approximately 35 km compared to the direct line Viminacium - Pincum which should be 26 km long (Viminatio XIII Pinco 19.24 km), (Figure 1). The same principle is seen at the shortcut over the Miroč mountain. Danube Road goes from Taliata (Donji Milanovac) to Egeta (Brza Palanka) and instead of going along the Danube for 100 km, the shortcut road over the mountain is only 30 km (Tabula Peutengeriana: Faliatis VIII Gerulatis VI Unam VI Egeta 29.6 km), (Figure 2).

network concept seems to be very functional not just for longer-range transports, but also for sending couriers command post transfers etc. Discussion and Conclusion Often it was discussed how long military posts can survive on their own without resupply. Ancient sources claim that castra could hold provisions for a year. Our research and capacities of horrea discovered in legionary fortresses put this claim rather in the realm of wishful thinking. Quantities of provisions in Auxiliary forts seemed to be measured by several months while burgi and watchtowers could hold several weeks. This means that the supply system had to be continuous and always functional. Several horrea were discovered along the Upper Moesian Limes, within fortifications of different sizes and shapes: in Sapaja (Dimitrijević 1984: 29-71), Čezava (Novae) (Vasić 1984: 91-122), Boljetin (Smorna), (Zotović 1984: 211-225), Veliki Gradac (Taliata), (Popović 1984: 265-282). Forts of smaller dimensions did not have enough room for such buildings.

For the fast deployment of troops or rapid interventions, these shortcuts were of crucial importance and therefore they were marked in the itineraries. This 110

Lifeline of the Frontier Stopping the navigation along the Danube meant that the land route remained the only available option. And high cliffs and impassable terrain had to be overcome in some way. Building the road through the Iron Gate gorge lasted for 70 years. This marvellous undertaking was almost mission impossible. With no explosives, power tools, or heavy machines, but only with the chisels and pickaxes of the legionnaires, this was a long-lasting challenge. Building strategic land communication where there was no space even for the hiker’s path was a marvellous undertaking beyond engineering and normal human possibilities, (Figures 3-5). Although the road that was cut into the almost vertical cliffs, was safe and stable it still wasn’t wide enough for regular functions. That is why another effort was made.

Figure 5: Roman road in the Iron Gate cut into the cliff/ section 3 (Photo documentation of the Institute of Archaeology, Belgrade)

Below the path cut into the cliff square holes were made. In these holes, wooden beams were installed as consoles to carry the construction of a wooden path that extended the road over the river itself. Only then road through the gorge gain its full functionality and made the possible normal flow of people and goods.

Figure 3: Roman road in the Iron Gate cut into the cliff/ section 1 (Photo documentation of the Institute of Archaeology, Belgrade)

In total nine imperial tabulae and two legionary inscriptions testify to the enormous importance of these supply routes. Two tabulae were left by the Tiberius troops, one of the Claudius, two of Domitian, three of Trajan and one of Diocletian, (Figure 6). Two of Trajan’s were dedicated to river navigation and were dedicated to the building of the Sip canal, (Figure 7). and the port of Kusjak. Six remaining tabulae were dedicated to the building of the road through the Iron Gate. Both military inscriptions dedicated to the soldiers of Legio VII Claudia and IIII Flavia were related to the building of the roads above the Danube. One inscription is located below the Turkish Ram fortress and the other is close to Tabula Traiana in Kazan, (Figures 8 and 9).

Figure 4: Roman road in the Iron Gate cut into the cliff/ section at the Tabula Traiana (Photo documentation of the Institute of Archaeology, Belgrade)

111

Nemanja Mrđić and Angelina Raičković Savić

Figure 6: Imperial inscriptions with the road section at the Gospođin Vir site, Drawing from the 17th century (Source: Danubius Pannonico-mysicus : observationibus geographicis, astronomicis, hydrographicis, historicis, physicis, perlustratus et in sex tomos digestus. T. 2, [Antiquitates romanae militares ad ultramque ripam Danubii]Marsigli, Luigi Ferdinando 1726, Hagæ Comitum : Apud P. Gosse, R. Chr. Alberts, P. de Hondt ; Amstelodami : Apud Harm. Uytwerf & Franç. Changuion, http://resolve.ubsm.bg.ac.rs/HRS/1854, Page 183 http://ubsm.bg.ac.rs/latinica/dokument/1854/danubius-pannonico-mysicus-observationibus-geographicis-astronomicishydrographicis-historicis-physicis-perlustratus-et-in-sex-tomos-digestus-t-2-antiquitates-romanae-militares-adultramque-ripam-danubii)

The special functionality of roman limes roads was related to dragging ships upstream. Ships were towed against the river current by human or animal power. Traces of these actions remain as rope cuts in the rock cliffs, (Figures 10-12). The best example was located in the Iron Gate, but unfortunately, this site is now submerged under the water of the Danube after the building of the dam Djerdap I. Furthermore, part of the road below tower 6 was solely dedicated to this purpose and was made by the vexillation of Legio VII Claudia under the command of centurion Caius Rufinus during the reign of Emperor Trajan. This towing process lasted long after the fall of the Empire and was documented until the middle of the 20th century when boat engines become strong enough to overcome the force of the river current. Personal notes of Felix Kanitz from 1889 mention this towing process through the Sip canal where the boats were towed upstream by 20-40 oxen.

Figure 7: Tabula Traiana from Statium Cataractarum Diane dedicated to the building of the Trajan’s canal at the Sip, fort at present-day Karatas, near Kladovo (Photo: N. Mrđić)

112

Lifeline of the Frontier

Figure 11: Traces of rope cuts into the rock made during towing ships upstream, located below Ram fortress. Roman inscription located above the flattened rock surface (Photo: N. Mrđić)

Figure 8: Tabula Traiana near Kazan, Iron Gate, present-day situation (Photo: N. Mrđić)

Figure 9: Tabula Traiana near Kazan, Iron Gate, original photo in situ before relocation (Photo documentation of the Institute of Archaeology, Belgrade)

Figure 12: Traces of rope cuts into the rock made during towing ships upstream, located in the Iron gates. The man in the background is sitting on the Roman road cut into the cliff (Photo documentation of the Institute of Archaeology, Belgrade)

Danube cataracts at Sip were the major and unsolvable problem until emperor Trajan’s preparations for the Dacian wars. Digging canal between Karataš and Sip solved this problem and imperial Tabula of the Traianus testified on this undertaking exhibited at the entrance of the Djerdap I hydroelectric plant. Originally canal was 3225 m long/57 m wide and this marvellous achievement was celebrated by the imperial tabula of emperor Trajan, (Figure 7), which stood at the entrance of the canal by the Diana auxiliary fortress at Karataš (Kanic 1989: 499-500).

Figure 10: Traces of rope cuts into the rock made during towing ships upstream, located below Ram fortress. Roman inscription located above the flattened rock surface (Photo: N. Mrđić)

Groups of imperial tablets suggest the concentration of the hardest engineering works during the 1st century AD. The site of Gospođin Vir with three tabulae and another two on the cliff nearby clearly indicates where 113

Nemanja Mrđić and Angelina Raičković Savić most of the efforts from Tiberius until Domitian were made. Unfortunately today these tabulae are either underwater in the Danube accumulation lake, (Figures 13-14), or destroyed (except tabula Traiana in Kazan and the 2 tabulae at the entrance to Hydroelectric power plant Djerdap I that is presented, but none of them in situ). Within the UNESCO Limes nomination process efforts are made to present these underwater sites that still exist after more than 50 years of being submerged. According to inscriptions, four legions were the core of the engineering force for the road building in Moesia during the 1st and at the beginning of the 2nd century AD. As mentioned in the inscription of Tiberius (33/34 AD) Legio V Macedonica and Legio IIII Scythica did the leg work in the first decades, (Figure 15). Trajan’s works were spearheaded by Legio IIII Flavia and Legio VII Claudia.

Figure 15: Inscription of Emperor Tiberius 33/34 AD at Gospođin Vir site. The Roman road is fully submerged under the tabula. Present-day situation. (Photo: N. Mrđić)

The Tabula of emperor Tiberius was also the oldest known imperial inscription, as well maybe the first inscription documented in central Serbia and its Danube region. Chronologically the last imperial Tabula of emperor Diocletian from Donje Butorke fort, (Figure 16), is also presented at the Hydroelectric power plant Djerdap I – not in situ. It is the only one dedicated to the

Figure 16: Inscription of Diocletian from Donje Butorke, presented at Hydroelectric power plant Djerdap I, presentday situation (Photo: N. Mrđić).

fort itself and not to major infrastructural undertakings like all the other imperial tabulae.

Figure 13: Roman imperial inscription partially submerged in the Iron Gate at site Gospođin Vir. (Photo: A. Teodorović)

This road remained in use until the late 1960s by the local population and border patrols of the Yugoslav People’s Army. Today only a small section at Golubac fort is visible and a section that is partially submerged in Gospođin Vir is being prepared for presentation. The rest lies between five and 20 m below the waters of the Danube’s Djerdap accumulation lake. In his discussion of the problem of supplying and transporting goods for Roman troops, D. Breeze tried to answer by stating a hypothesis about four ways of a food supply (Breeze 2000: 59-64). Breeze states that possibilities that a municipium or a colony could collect and transport supplies with their means or engage transporters for military purposes. On the other hand, the military could collect, and transports supplies on

Figure 14: Roman imperial inscription submerged in the Iron Gate at site Gospođin Vir. (Photo: A. Teodorović)

114

Lifeline of the Frontier

Figure 17: Basic supply scheme along the Danube Limes in Province Upper Moesia (Moesia Superior) (Basemap: Google Earth, author: N. Mrđić)

its own or engages private transporters. Although this seems logical so far we did not find examples to confirm this collection and distribution system.

were visible. Shards of larger pottery vessels were also found, used for transporting and storing food: pithoi and amphorae.

Positioning supply centres on the major crossroads was also one of the strategic requirements. Thus, placing one at the mouth of the Porečka river was of enormous importance. It could send supplies both downstream of the Danube, or over the Miroč mountain towards Egeta over the land route (Petrović 1981: 53-62; Petrović 1983: 285-291). The site is situated in a very convenient position, in the vicinity of one of the biggest camps of the Danubian Limes-Taliata (Veliki Gradac), which had soldiers stationed permanently throughout the Roman and Early Byzantine periods (Petrović 1984: 265-282).

The second site identified as the supply centre was located at the site Konopište, 3 km downstream from Trajan’s bridge and the fort Pontes, foundations of irregular rectangular buildings were discovered (Popović 1996: fig. 1). Part of the site was recently excavated, but the port that was identified decades ago remains under the water of Danube. The site Kurvingrad, today under water, was located one kilometre downstream from Konopište. At the excavated area, buildings used for storing and keeping grain and other goods were discovered. Dimensions of the horrea could not be determined because of erosion. Few finds included amphorae shards, also from the end of the 1st and the beginning of the 2nd century AD (Popović 1996: 103).

The mouth of the Porečka river was additionally closed with a defensive wall,1 while on the right river bank, there were two rectangular towers. The most interesting features are two broad buildings of approximately the same size. Building A was made of bricks and stones, with no inner walls, with a broad entrance facing the south, while building B was made of stone, with two pillars in its interior. The building was filled with huge amounts of debris, with bricks on which fire traces

It is clear that the supply strategy goes in the full direction of sending provisions downstream from one supply centre downstream until the next one. Ships would be able to go back upstream empty as heavy loads were very hard to push upstream with small crews on board. Several points where ships had to be towed upstream from the land supporting servicemen confirm this long-lasting problem in Danube navigation, (Figure 17).

Closing mouths of Danube tributaries by walls was general practice in the Iron gate. With steep cliffs these water corridors were actually the only possible way inland.

1 

115

Nemanja Mrđić and Angelina Raičković Savić The Danube and the Limes road were also major trade routes. We observe limes mostly in the light of military strategy, operations and troop movements. But during peacetime and after the conquest of Dacia, the road remained in full function. Roads in Moesia become the backbone of the trade network.

hydrographicis, historicis, physicis, perlustratus et in sex tomos digestus. T. 1, in tres partes digestus : geographicam, astronomicam, hydrographicam. Hague. http://ubsm.bg.ac.rs/engleski/dokument/1853/ danubius-pannonico-mysicus-observationibusgeographicis-astronomicis-hydrographicishistoricis-physicis-perlustratus-et-in-sex-tomosdigestus-t-1-in-tres-partes-digestus-geographicamastronomicam-hydrographicam Mirković, M. 1981. Srednjobalkanski prostori u doba kasnog carstva, in S. Ćirković (ed.) Istorija srpskog naroda I: 89-104. Beograd: Srpska književna zadruga. Mirković, M. 1986. Inscriptions de la Mésie Supérieure, vol. II (Viminacium et Margum). Beograd: Centar za antičku epigrafiku i numizmatiku. Petrović, P. 1972. Nova Trajanova ploča u Đerdapu. Starinar XXXI: 31-39. Petrović, P. 1979. Inscriptions de la Mésie Supérieure, vol. IV, (Naissus, Remesiana et Horreum Margi). Beograd: Centar za antičku epigrafiku i numizmatiku. Petrović, P. 1981. O opskrbi rimskih trupa na đerdapskom limesu. Starinar XXXI: 53-62. Petrović, P. 1984. Rijeka Poreč, Sabrino središte opskrbe rimskih trupa na Đerdapu. Starinar XXXIII-XXXIV: 285-291. Petrović, P. 1991. Classis Flavia Moesica na Dunavu u Gornjoj Meziji. Starinar XL-XLI: 207-216. Petrović, V. 2004. Izleti o rimskim starinama u okolini Đerdapa u oblasti Bela de Gonda. Balkanika XXIV (2003): 71-95. Popović, M. 2006. Beogradska tvrđava, Beograd: Javno preduzeće „Beogradska tvrđava”. Popović, P. 1996. Konopište – Roman Architecural Complex, in P. Petrović (ed.) Roman Limes on the Middle and Lower Danube: 101-104. Belgrade: Archaeological Institute. Popović, V. 1984. Donji Milanovac - Veliki Gradac (Taliata), rimska i ranobizantska utvrda. Starinar XXXIII-XXXIV: 265-282. Rickman, G. 1971. Roman Granaries and Store Buildings. Cambridge: University Press. Roth, J.P. 1999. The Logistics of The Roman Army at War (264 B.C. - A.D. 235). Leiden-Boston-Köln: Brill. Šašel, J. 1961. Rimski natpisi u Đerdapu, in M. Grbić (ed.) Limes u Jugoslaviji I: 156-164. Beograd: Arheološko društvo Jugoslavije. Thomas, R. and S. Stallibrass 2008. For starters: producing and supplying food, in R. Thomas, S. Stallibrass (eds) Feeding the Roman Army the Archaeology of Production and Supply in NW Europe: 1-17. Oxford: Oxbow Books. Vasić, M. 1984. Cezava-Castrum Novae. Starinar XXIIIXXXIV: 91-122. Vasić, M. 1990. Horreum Margi (Ćuprija). Beograd: Arheološki institut. Zotović, Lj. 1984. Boljetin (Smorna), rimski i ranobizantski logor. Starinar XXIII-XXXIV: 211-225.

Neither of these communication systems could function individually and this need as well as the possibility for mutual backing up is what made it reliable and functional. Efforts that made it possible to speak for themselves on the importance of making it the true lifeline of the Frontiers of the Roman Empire. Bibliography Alföldi, M. R. 2001. PROVIDENTIA AUGUSTI. To the question of Limes fortifications in the 4th century, in Gloria Romanorum: Schriften zur Spatäntike: 154-167. Stuttgart: Franz Steiner Verlag. Bjelajac, Lj. 1996. Amfore gornjomezijskog Podunavlja. Beograd: Arheološki institut. Bohec, Y. 2000. The Imperial Roman Army. London: Routledge. Breeze, D. 2000. Supplying the Army, in G. Alföldy, B. Dobson, W. Eck (eds) Kaiser, Heer und Gesellschaft in der Römischen Kaiserzert: Gedenkschrift für Eric Birley: 59–64. Stuttgart: Steiner. Campbell, B. 1994. The Roman Army 31 BC-337 AD. London, New York: Routledge. Davies, R. W. 1989. Service in the Roman Army. Edinburgh. Herz, P. 2007. Finances and Cost of the Roman Army, in P. Erdkamp (ed.) A Companion to the Roman Army: 306– 322. Malden-Oxford-Carlton: Blackwell Publishing. Ilić, O., Golubović, S., Mrđić, N. 2011. Supplying and Transport Along Danube Limes in the Upper Moesia. Arheologija i Prirodne Nauke 6 (2010): 61-76. Jones, A. H. M. 1973. The Later Roman Empire II. Oxford: Basil Blackwell. Kanitz, F. 1892. Römische Studien in Serbien. Der DonauGrenzwall, das Strassennetz, die Städte, Castelle, Denkmale, Thermen und Bergwerke zur Römerzeit im Königreiche Serben. Wien: Tempsky. (Denkschriften der Kaiserlichen Akademie der Wissenschaften, Philosophisch-Historische Classe, 41, 2. Wien, 1893). Kanic, F. 1989. Srbija zemlja i stanovništvo or ramskog doba do kraja XIX veka, knjiga II. Beograd: Srpska književna zadruga. Kehne, P. 2007. War-and Peacetime Logistics: Supplying Imperial Armies in East and West, in P. Erdkamp (ed.) A Companion to the Roman Army: 323-338. MaldenOxford-Carlton: Blackwell Publishing. Korać, M., S. Golubović and Mrđić N. 2009. Itinerarivm Romanvm Serbiae. Belgrade: Center for New Technologies. Marsigli, L. F. 1726. Danubius Pannonico-mysicus: observationibus geographicis, astronomicis, 116

Signposts of Settlement: Roman Communications and Avar-Age Sites in Continental Croatia: Overlap Pattern Analysis Lorena Jurakić and Pia Šmalcelj Novaković Abstract: In Avar archaeology, the phenomenon of frequent overlapping of Avar-age and Roman sites in former Roman Pannonia has already been noticed. In this paper, the authors analyzed the pattern of overlap of (assumed) Roman communications and Avar-age sites in continental Croatia (Roman provinces of Pannonia Savia and Sirmiensis), with regard to the geographical characteristics of the area, trying to establish a certain regularity of the sample. This paper was created under the project Life on the Roman road: communications, trade and identities on Roman roads in Croatia from 1st - 8th CE (UIP-05-2017-9768) at the Institute of Archaeology, funded by the Croatian Science Foundation. KEYWORDS: (LATE) AVAR-AGE, ROMAN, PROJECT LIFE ON THE ROMAN ROAD, EARLY MEDIEVAL, SETTLEMENT PATTERNS, ROMAN COMMUNICATIONS, HYDROGRAPHY

Introduction Continental Croatia is part of the Republic of Croatia between the Drava River on the border with Hungary, the Una and Sava rivers on the border with Bosnia and Herzegovina and Sutla, Kupa and Žumberak Mountains on the border with Slovenia and the Danube River on the border with Serbia to Gorski Kotar and Lika (Brozović et. al. 2022). The region belongs to the SW part of the Pannonian Plain; throughout history, this area has been an area of turmoil ​​ and change for various populations, and its strategic defensive and border position is also crucial. The strategic importance today is most evident in the numerous traffic routes running through the region: it has always been a hub of important communication routes to Western Europe and the Baltics, as well as to the Adriatic, Aegean and Black Seas (Andrić 2002). It should be noted that several names are used for this area in the archaeological literature: Krešimir Filipec calls this area Lower Pannonia and defines it as an area south of the Drava to the Sava River, but also south and north of these rivers (Filipec 2015: 15), Željko Tomičić and Tajana Sekelj Ivančan define it as an area bordered by rivers, the Mura, Drava and Danube in the north and the Sava in the south and call it the SavaDrava interfluve (Tomičić 2000; Sekelj Ivančan 2010), while historian Hrvoje Gračanin, who often deals with archaeological topics, uses the name Southern Pannonia (Gračanin 2009). During Roman rule, the area of ​​continental Croatia belonged to the southern part of the Roman province of Pannonia. Pannonia (lat. Pannonia) is an ancient name

for the area of ​​the province of the Roman Empire whose border in the north and east was the right bank of the Danube, from Vienna to the confluence of the Sava and the Danube. The border in the south was slightly south of the Sava and Kupa, and in the west Pannonia bordered the province of Noricum. (Šišić 1920: 27) At the beginning of the 2nd century AD, during the reign of Emperor Trajan, Pannonia was administratively divided into Upper Pannonia (lat. Pannonia Superior) in the west and Lower Pannonia (lat. Pannonia Inferior) in the east. In AD 297, the Roman emperor Diocletian divided the whole of Pannonia into four areas following the new administrative structure of the state. He divided Upper Pannonia into the Pannonia Prima or Pannonia Superior, based in Savaria (now Szombathely) and the Sava Pannonia (lat. Pannonia Savia) based in Siscia (now Sisak); Lower Pannonia into Pannonia Valeria (lat. Pannonia Valeria) with its seat in Sopianae (today Pécs) and Pannonia Secunda or Pannonia Inferior with its centre in Sirmium (today Sremska Mitrovica) (Šišić 1920: 28). Continental Croatia would then consist of most of the province of Pannonia Savia, part of the Second Pannonia west of the Danube and the southernmost part of Pannonia Valeria along the Drava River, or southern Pannonia. In Roman times, southern Pannonia was an important transit area between the western and eastern parts of the Roman Empire, and its roads played a prominent role in the political circumstances of the time (Andrić 2002: 121-25). Roman communications According to the analysis conducted based on data on the sites in the database of antique archaeological sites of the Republic of Croatia (hereinafter: BASE), the

Life Along Communication Routes (Archaeopress 2023): 117–129

Lorena Jurakić and Pia Šmalcelj Novaković

Figure 1: Map of Roman communications in continental Croatia (Jurakić 2021: sl.10, 17)

population in Roman times was mostly concentrated around Roman roads and rivers Sava, Drava and Danube, but also smaller rivers such as Bosut, Kupa and Česma (Jurakić 2021: 10-11). The analysis and mapping of Roman communications in this paper is based on data from the BASE, additionally also Hrvoje Gračanin’s work (Gračanin 2010) and based on cartographic information from the Tabula Peuntingeriana.1

centuries AD, is available. The area of Croatia is shown from the fourth to the seventh segment of the map. The direction of the roads is drawn in thicker straight lines, and the names of the provinces are written in red. The names of the road stations are written above the road line, and the stations themselves and the cities and military camps are marked with smaller or larger city icons, depending on their importance. The most prominent are Rome, Constantinople and Antioch (Milotić 2010: 165-167).

Today, the most famous and extensive cartographic representation of places and roads during the Roman Empire is the aforementioned Tabula Peutingeriana, showing the road network of the whole Empire, Persia and part of India with road stations, a total of more than 4000 toponyms. It is based on earlier maps of the (known) world such as that of Mark Vipsanius Agrippa from the 1st century BC as a template: the original has not been preserved, but a medieval copy of a map with 11 preserved leaves, made between the 12th and 13th

Of course, as in modern times, roads were built with the logic of best profitability, so the Romans built roads in valleys, mountain valleys and along rivers. Continental Croatia is a great example of this logic: the main roads are located along the valleys of the great rivers Drava, Sava and Danube, which are interconnected transverse roads laid along smaller rivers and streams or in mountain valleys. It is safe to say that in modern times the hierarchy of roads in today’s continental Croatia is not much different from Roman times (Gračanin 2010: 10).

The database of antique archaeological sites of the Republic of Croatia (BASE) is one of the most comprehensive sources for aforementioned sites in continental Croatia today: it offers basic information about the site (including finds from other periods), state of research, bibliography and cartographic data. For more details, see: http://www.iarh.hr/en/projects/the-database-of-antique-archaeolo gical-sites-of-the-republic-of-croatia/

1 

Thus, in continental Croatia we recognize the three most important routes of Roman communications: 118

Signposts of Settlement Table 1: (Presumed) Sites of the Podravska route (author: P. Šmalcelj Novaković) Site

Source

Babinec

http://baza.iarh.hr/public/locality/detail/781

Križovljan

http://baza.iarh.hr/public/locality/detail/942

Petrijanec (Aquae Vivae)

http://baza.iarh.hr/public/locality/detail/683

Sračinec

http://baza.iarh.hr/public/locality/detail/3121

Varaždin

Gračanin, 2010: Rimske prometnice i komunikacije u kasnoantičkoj južnoj Panoniji (20), Scrinia Slavonica 10, Zagreb.

Jalžabet

http://baza.iarh.hr/public/locality/detail/701

Ludbreg (Iovia)

Gračanin, 2010: Rimske prometnice i komunikacije u kasnoantičkoj južnoj Panoniji (20), Scrinia Slavonica 10, Zagreb.

Kunovec Breg (Sunista)

http://baza.iarh.hr/public/locality/detail/2000

Draganovec (Piretis)

http://baza.iarh.hr/public/locality/detail/1998

Novigrad Podravski

http://baza.iarh.hr/public/locality/detail/2069

Mičetinac

http://baza.iarh.hr/public/locality/detail/2067

Čepelovac

http://baza.iarh.hr/public/locality/detail/2026

Budrovac

http://baza.iarh.hr/public/locality/detail/2024

Sirova Katalena

http://baza.iarh.hr/public/locality/detail/2074

Kozarevac

http://baza.iarh.hr/public/locality/detail/2061

Mala Črešnjevica

http://baza.iarh.hr/public/locality/detail/456

Velika Črešnjevica

http://baza.iarh.hr/public/locality/detail/483

Vukosavljevica

http://baza.iarh.hr/public/locality/detail/490

Golo Brdo

http://baza.iarh.hr/public/locality/detail/450

Virovitica

http://baza.iarh.hr/public/locality/detail/487

Gradina (Serota?)

http://baza.iarh.hr/public/locality/detail/451

Orešac (Bolentio?)

http://baza.iarh.hr/public/locality/detail/458

Podgajci Podravski (Berebae?)

http://baza.iarh.hr/public/locality/detail/2215

Petrijevci (Mursella)

http://baza.iarh.hr/public/locality/detail/2423

Osijek (Mursa)

http://baza.iarh.hr/public/locality/detail/1088

Nemetin

http://baza.iarh.hr/public/locality/detail/3446

Vinkovci (Cibaliae)

http://baza.iarh.hr/public/locality/detail/1103

Orolik (Cansilena)

Gračanin, 2010: Rimske prometnice i komunikacije u kasnoantičkoj južnoj Panoniji (24), Scrinia Slavonica 10, Zagreb. http://baza.iarh. hr/public/locality/detail/2362

The Podravska route, basically a branch of the famous Amber Road passed the most favourable route between the Drava river valley and the hills of Bilogora and Kalnik (Gračanin 2010: 20). From Sirmium, the road continues to Taurunum and Singidunum, today’s Zemun and Belgrade, which were also important hubs of Pannonia during the Roman Empire (Gračanin 2010: 18).

The Posavska route, along with the Podravina highway, is the most important road in southern Pannonia and the shortest traffic connection between Italy and the eastern provinces of the Roman Empire. Only the Tabula Peutingeriana fully depicts this route (Jurakić 2021: 20). The route of the Posavina Highway starts in Aemona, today’s Ljubljana and ends in Sirmium (Sremska Mitrovica), passing important traffic hubs such as Siscia 119

Lorena Jurakić and Pia Šmalcelj Novaković (Sisak), Servitium (Gradiška) and Marsonia (Slavonski Brod), (Gračanin 2010: 11, 18).

In Continental Croatia, during the Roman Empire, in addition to the main roads of the Podravina, Posavina and Dunav limes areas, there were local and regional roads. They connected colonies and municipalities with rural settlements and rural centres and settlements with economic facilities such as port docks, ceramic centres, quarries, mines, country villas, military fortifications, road stations, rest areas, as well as sacral and burial facilities such as temples and cemeteries

Danube Limes in Croatia or Pannonian Limes is a road that is only a small part of the large traffic and defence corridor of the Danube Limes that followed the flow of the Danube through the Roman Empire Noricum, Pannonia, and Scythia to the Black Sea (Gračanin 2010: 27-28).

Table 2: (Presumed) sites of the Posavska route (author: P. Šmalcelj Novaković) Site

Source

Lužnica

http://baza.iarh.hr/public/locality/detail/1907

Ježdovec

http://baza.iarh.hr/public/locality/detail/2564

Botinec

http://baza.iarh.hr/public/locality/detail/1838

Gornji Čehi

http://baza.iarh.hr/public/locality/detail/1857

Petrovina Turopoljska

http://baza.iarh.hr/public/locality/detail/1876

Okuje

http://baza.iarh.hr/public/locality/detail/1832

Mraclin

http://baza.iarh.hr/public/locality/detail/1833

Buševec

http://baza.iarh.hr/public/locality/detail/1843

Sisak (Siscia)

http://baza.iarh.hr/public/locality/detail/1087

Hrvatska Kostajnica

http://baza.iarh.hr/public/locality/detail/2141

Slabinja

http://baza.iarh.hr/public/locality/detail/2142

Baćin

http://baza.iarh.hr/public/locality/detail/2143

Hrvatska Dubica

http://baza.iarh.hr/public/locality/detail/2144

Gradiška (Servitium)

Gračanin, 2010: Rimske prometnice i komunikacije u kasnoantičkoj južnoj Panoniji (17).

Slavonski Kobaš

http://baza.iarh.hr/public/locality/detail/2345

Gornji Andrijevci

http://baza.iarh.hr/public/locality/detail/296

Završje

http://baza.iarh.hr/public/locality/detail/326

Slavonski Brod (Marsonia)

http://baza.iarh.hr/public/locality/detail/311

Rušćica

http://baza.iarh.hr/public/locality/detail/2368

Oprisavci

http://baza.iarh.hr/public/locality/detail/290

Sikirevci

http://baza.iarh.hr/public/locality/detail/301

Posavski Podgajci (Saldae)

Gračanin, 2010: Rimske prometnice i komunikacije u kasnoantičkoj južnoj Panoniji (17-18).

120

Signposts of Settlement Table 3: (Presumed) Sites of the Danube limes route (author: P. Šmalcelj Novaković) Site

Source

Ilok (Cuccium)

http://baza.iarh.hr/public/locality/detail/1086

Sotin (Cornacum)

http://baza.iarh.hr/public/locality/detail/2268

Dalj (Teutoburgium)

Gračanin, 2010: Rimske prometnice i komunikacije u kasnoantičkoj južnoj Panoniji (28-29), Scrinia Slavonica 10, Zagreb.

Aljmaš

Gračanin, 2010: Rimske prometnice i komunikacije u kasnoantičkoj južnoj Panoniji (29), Scrinia Slavonica 10, Zagreb.

Nemetin

http://baza.iarh.hr/public/locality/detail/3446

Kopačevo (Ad Labores)

http://baza.iarh.hr/public/locality/detail/3081

Lug (Albanum)

Gračanin, 2010: Rimske prometnice i komunikacije u kasnoantičkoj južnoj Panoniji (29), Scrinia Slavonica 10, Zagreb.

Jasenovac

http://baza.iarh.hr/public/locality/detail/2644

Kneževi Vinogradi (Donatianae)

Gračanin, 2010: Rimske prometnice i komunikacije u kasnoantičkoj južnoj Panoniji (30-31), Scrinia Slavonica 10, Zagreb.

Kamenac (Aureus Mons)

Gračanin, 2010: Rimske prometnice i komunikacije u kasnoantičkoj južnoj Panoniji (31-32), Scrinia Slavonica 10, Zagreb.

Popovac (Quadriburgium)

Gračanin, 2010: Rimske prometnice i komunikacije u kasnoantičkoj južnoj Panoniji (31-32), Scrinia Slavonica 10, Zagreb.

Zmajevac (Ad Novas)

Gračanin, 2010: Rimske prometnice i komunikacije u kasnoantičkoj južnoj Panoniji (31-32), Scrinia Slavonica 10, Zagreb.

Batina (Ad Militare)

http://baza.iarh.hr/public/locality/detail/2279

farther from city centres. The Romans used the name via mediterranea for internal roads. The most important side road of southern Pannonia was the regional road that went from Sisak (Siscia) through the central part of Continental Croatia to the city centres of Osijek (Mursa) and Vinkovci (Cibalae), (Gračanin 2010: 33-41).

control of the whole of Pannonia (Pohl 2018: 62-9). In the ensuing period (especially after the conquest of Sirmium, which provided control of communication routes to the Balkan Byzantine provinces and the Black Sea region) until the failed siege of Constantinople in AD 626, Avars undertook a series of attacks on Byzantine territory and continued to live nomadic and semi-nomadic lifestyles; archaeological material indicates the survival of part of the Germanic and Romanized population found by the newcomers; in Avar costumes and women’s jewellery, there is a strong emulation of Byzantine cultural influence, primarily by imitation of decorative elements in the elements of precious metal costumes in this period (Pohl 2018: 198-243, 294-305). After the decisive, unsuccessful siege of Constantinople and the consequent cessation of tribute, it seems that the change in the way of life and the social order within the Avar state can be traced to the middle of the 7th century AD when the transition to a sedentary lifestyle of the majority of the population seems to begin. They change their costumes and the cultural expression becomes homogeneous, without visible traces of Germanic and very rare traces of the presence of Romanized inhabitants; the main decorative elements of men’s costumes are now

The Avars The Avars2 entered the Carpathian Basin in the year AD 567, by allying with the Lombards against the Gepids, thus as an external factor in the multigenerational conflict of two Germanic peoples controlling Pannonia and the Carpathian Basin during the 6th Century (Pohl 2018: 58-69). Very quickly, the warriors under the leadership of Kagan Bayan confirmed their reputation, the Gepids were swiftly defeated, and their living space now belonged to Avars by right of conquest; in the spring of AD 568, the Lombards left the western part of Pannonia and moved to Italy, and the Avars took The term “Avar” in this paper is not used as an ethnic determinant, but as a designation of members of the community that under the Kagan Bayan in year AD 567 entered the Carpathian Basin and lived with the existing population for the next two and a half centuries within the Avar Khaganate (see Pohl 2018 in more detail: 1-5).

2 

121

Lorena Jurakić and Pia Šmalcelj Novaković bronze cast fittings with a range of decorative motifs from the Mediterranean-Byzantine cultural sphere, but seemingly selectively chosen and transmitted by Avar masters, thus creating their own aesthetic and symbolic expression (Pohl 2018: 335-352). Such a way of life and cultural expression persisted until the collapse of the Avar state organization caused by a series of Frankish attacks in the period AD 788-796; 822 is the last year the Avar embassy at the Frankish court is mentioned (Pohl 2018: 376-389).

Avar cultural circle (logically, including those sites that have only Slavic and / or Carolingian material, such as Medvedička (Šmalcelj Novaković 2022a: 10). Settlement sites were not included, although we know of several from this period, because most of these sites have no published finds of explicit Avar cultural affiliation and it is impossible to conclude with complete certainty that these are communities that were certainly under the control of the Avar state; there are two exceptions-a small researched part of the settlement in Stari Jankovci which certainly belonged to the community at the Avar-era cemetery in the immediate vicinity (Šmalcelj Novaković 2022b, Filipec 2017) and the site of VirjeSušine, where a late Avar belt-end was discovered inside a settlement object (Sekelj Ivančan 2014: 103). It should be emphasized that the list is composed exclusively of published sites.

At the time when the Avars conquered Sirmium (Sremska Mitrovica) in AD 582, Siscia (Sisak) had probably already fallen, until then it was presumably under the control of the Lombards (Bilogrivić, Gračanin 2014: 18, 19). It can be concluded that the Siscia episcopal centre was abandoned and the church organization disappeared, but the Christian cult of the population did not: it was maintained by the core Romanized population that continued to live in Siscia during the 7th century (Bilogrivić, Gračanin 2014: 20-21). With the Avar or Slavic conquest of Siscia, at the beginning it probably served as a Slavic stronghold from which they invaded the south (Bilogrivić, Gračanin 2014: 20). Archaeological finds from that period indicate the continuity of its population; It seems that Siscia gradually became more and more similar to a typical early Slavic settlement; finds of Byzantine coins and matrix for pressing horse harness ornaments belong to that era, but due to the very small number of such finds it is impossible to conclude to what extent life continued in the once famous Pannonian city; it is certain that Siscia degrades urbanistic, and population-wise (Bilogrivić, Gračanin 2014: 22-23).

A comparative map of Roman communications and Avar-age sites (Figure 2), which are shown with only one mark on the map in the case of several sites/ positions in the area of ​​one settlement, due to clarity, best illustrates the connection between their spatial distribution in the area. Their comparison leads to several conclusions: Avar-age sites mostly appear on the route of Roman communications: outside of them are only Velika Horvatska, Križevci, Novska-factory Kozara and Glina riverbed in Marinbrod, where accidental Avar-age finds have been registered (Table 4: 49, 24, 29, 27), while Draškovec, Kuzelin, Otok-Gradina, Prelog-Šljunčara, Privlaka and two positions in Zagreb (Kaptol-Bakačeva kula and Kruge) seem to be very close to Roman communications (Table 4: 15, 25, 32, 36, 59, 60).5 Such sites as Bapska, Gradac, Dalj, Busija and Daljska planina, Draž, Koprivnica, Šoderica, Kotlina and Slavonski Šamac, Sava riverbed (Table 4: 1, 11, 15, 16, 22, 23, 63) are so close to the communication infrastructure or rather, its remains, that it can be considered with great certainty they were also situated on Roman routes (this is well illustrated by the example of Dalj, Busija and even Dalj, Daljska planina located in the settlement area of ​​the former Roman limes fortress Teutoburgium).

The former Roman colony of Cibalae (Vinkovci) also continued to function after the collapse of the Western Roman Empire, but to a much lesser extent: after the destruction of the Kingdom of Gepids in AD 567 it seems that the Cibalae area was devastated, by the end of the 6th century the urban fabric degraded to a rural settlement-the revival of the settlement occurs in the 7th century, and most of the newcomers settled in the vicinity of the former city until the second half of the 8th century AD, completely rejecting the Roman urban tradition (Gračanin, Rapan Papeša 2011: 22-3).3

concentration of sites between the easternmost and the rest of continental Croatia: the majority of Avar-age sites in continental Croatia are situated on the Limes itself, in its background and along the Drava.

The map of Avar-age sites was created according to the Table 44 which lists all such sites in continental Croatia that are defined as cemeteries (excavated cemeteries, presumed cemeteries, individual graves) and those where a random find was registered; the sites that have been left out were those where no find has been recorded that can be unequivocally attributed to the

The discrepancy in the number of sites between the easternmost and the rest of continental Croatia is not easy to explain: it cannot be fully explained by the Some researchers believe that Prelog itself was on the route of two ancient communications, one that followed the bank of the Drava in the east-west direction and the other of north-south direction that crossed the Drava south of Prelog (Vučetić 2009: 179; Petrić 1995: 54, 56).

5 

3  However, new field research points to the Avar presence in the city itself, see: Rapan Papeša 2020 (What to expect when you’re expecting). 4  Table 4 was created on the basis of Šmalcelj Novaković 2022a: Tab. 1

122

Signposts of Settlement Table 4: List of (published) Avar-age sites in continental Croatia (author: P. Šmalcelj Novaković) Site

Bibliography

Bapska- Gradac

Dimitrijević, Kovačević, Vinski 1962: 73; Burić et al. 2011

Batina

Vinski 1958: 46, fn 114; Dimitrijević, Kovačević, Vinski 1962: 67

Belišće-Zagajci II

Filipec, Roksandić 2009: 19–20; Filipec 2015: 81–2, 91

Bijelo brdo II (Bijelo brdo- Bajer)

Ivaniček 1949: 111–4, pls1–43; Dimitrijević, Kovačević, Vinski 1962: 107–10, figs 1–11

Borovo-Gradac

Dimitrijević, Kovačević, Vinski 1962: 75–6, figs 1–4; ADAM 2002: 65

Brestovac, Požeški

Hampel 1905 2: 423–6; 1905 3: Taf 320

Brodski Drenovac- Plana

Vinski-Gasparini, Ercegović 1958

Čađavica- Branjiska

Fettich 1941: 55–61, Tafn 3–5; Vinski 1958: 27, pls 14–5

Dalj- Bogaljevci, ul. Đ. Đakovića (Đ. Đaković Street)

Bojčić 1984: 215–6, Fig. 2–5

Dalj- Bogojevci, vrt S. Blagojević (S. Blagojević’s yard)

Bulat 1978: 175

Dalj-Busija, vinograd M. Panišića (M. Panišić’s vineyard)

Dimitrijević, Kovačević, Vinski 1962: 111, fig 5

Dalj-Ciglana (Brickyard)

Dimitrijević, Kovačević, Vinski 1962: 110, figs 1–3, pl 30:1

Dalj-Daljska planina

Dimitrijević, Kovačević, Vinski 1962: 111, fig 6

Dalj-Dunavski bajer

Dimitrijević, Kovačević, Vinski 1962: 110–1

Draškovec

Lipp 1884: 9–10; ADAM 2002: 118

Draž

Dimitrijević, Kovačević, Vinski 1962: 68, figs 1–2

Đakovo-Župna crkva (Parish church)

Filipec 2003: 133

Hrvatska (Croatia)

Dimitrijević, Kovačević, Vinski 1962: 102–3, figs 1–3

Ilok

Hampel 1905 3: Taf 43:9

Jagodnjak-Krčevine

Dugonjić, Tresić Pavičić, Novak 2020

Kneževi vinogradi

Csallány 1956: 132, no 415; Dimitrijević, Kovačević, Vinski 1962: 68

Koprivnica-Šoderica

HiK 2 2000: 90, cat nos 2.10–11; Tomičić 2000: 154; Aralica, Ilkić 2012: 170–1, cat nos 1–3

Kotlina

Dimitrijević, Kovačević, Vinski 1962: 69, figs 1–3

Križevci

Vinski-Gasparini, Ercegović 1958: 152

Kuzelin

Sokol 1994: 48, 152

Lobor-Majka Božja Gorska

Filipec 2009

Marinbrod-korito rijeke Gline (riverbed of the Glina)

Filipec 2003: 133; 2010

123

Lorena Jurakić and Pia Šmalcelj Novaković

Site

Bibliography

Novo Čiče

Simoni 1981: 160, 163; Milošević 2000: 93–4

Novska-tvornica Kozara (Kozara factory, today Metaflex)

Aralica, Ilkić 2012: 172, cat no 5

Nuštar-dvorac Khuen-Belassy (Castle Khuen-Belassy)

Rapan Papeša 2011; 2012

Osijek-Zeleno polje

Horváth 1935; Vinski 1958: 26, pl 12; Dimitrijević, Kovačević, Vinski 1962: 112–4

Otok-Mandekov vinograd (Mandek’s vineyard)

Dimitrijević 1957: 21–38; Šmalcelj, Težak-Gregl 1992: 40–1; Rapan Papeša, Šmalcelj Novaković 2016

Podgrađe

Vinski-Gasparini, Ercegović 1958: 152

Popovac-ul. V. Nazora (V. Nazor Street)

Dimitrijević, Kovačević, Vinski 1962: 69, figs 1–2

Prelog-Šljunčara (Gravel pit)

Tomičić 1978: 212; HiK 2 2000, 100–4

Privlaka-Gole njive

Šmalcelj 1973: 117–9, pl 69; 1976: 127–8; 1981b: 143–4; 1992a: 44–5

Samatovci

Dimitrijević, Kovačević, Vinski 1962: 114, fig 1

Sarvaš-Vlastelinski brijeg (Kučka)

Vinski 1954: 74; Dimitrijević, Kovačević, Vinski 1962: 115

Sisak (Siscia)-unknown locations and the River Kupa

Vinski 1954: 74; 1958: 27; Šercer, Ždralović 1979: cat no 682; Filipec 2003; Bošković 2009: 125, cat no 72; Aralica, Ilkić 2012: 172, cat no 4, pl 17:1–4

Sotin (maybe Zmajevac)

Dimitrijević, Kovačević, Vinski 1962: 91, figs 2–4

Sotin-Višnjica

Vinski 1954: 75, fig H; Dimitrijević, Kovačević, Vinski 1962: 90

Sotin-Vrućak

Uglešić 1994: 145–52; Ilkić 2010: 384, cat nos 1–3

Sotin-Zmajevac, vinograd Luke Crnjanovića (Luka Crnjanović’s vineyard)

Dimitrijević, Kovačević, Vinski 1962: 90, fig 1

Stari Jankovci-Gatina

Šmalcelj 1981a; 1992b: 48–9

Struga Nartska- Šljunčara

Simoni 1981; Aralica, Ilkić 2012: 175, cat no 11

Suhopolje-Borova

HiK 2 2000: 104

Šarengrad-Klopare

Dizdar, Rapan Papeša, Rimpf 2017; Rapan Papeša, Rimpf, Dizdar 2018; Rimpf, Rapan Papeša, Dizdar 2020

Velika Gorica-Visoki brijeg

Hoffiller 1909: 120–34; Vinski 1960: 50–2; HiK 2 2000: 106

Velika Horvatska

Simoni 1986: 219

Vinkovci-Meraja

Dimitrijević, Kovačević, Vinski 1962: 97, fig 3; Dimitrijević 1966: 42

Vinkovci-Duga ulica 99

Sekelj Ivančan, Tkalčec 2006

Vinkovci-Town cemetery

Rapan Papeša 2020

Vučedol

Karmanski 1976; Mrkobrad 1980: 85, fig l

Vukovar-benzinska stanica Jugopetrola) (Jugopetrol gas station)

Dimitrijević, Kovačević, Vinski 1962: 102, fig 3

124

Signposts of Settlement

Site

Bibliography

Vukovar-crkva sv. Filipa i Jakova (Church of Sts Philip and James)

Demo 1996: 93

Vukovar-Lijeva bara

Vinski 1959: 103–4, pl 38:20–23; Dimitrijević, Kovačević, Vinski 1962: 102, figs 1–2; Demo 2009: 539, fig 53, 552

Vukovar-unknown location

Dimitrijević, Kovačević, Vinski 1962: 103

Vukovar-Švapsko brdo

Karmanski 1976: 5–6; Demo 1996: 15, fig 3

Zagreb-Kaptol, Bakačeva kula

Vinski 1952: 30, 51–2

Zagreb-Kruge (Trnje), prudište Štrpca

Vinski 1960: 52–3

Zmajevac (Vörösmart)

Garam 1982: 209–10, figs 14–5; ADAM 2002: 433–4

Zmajevac (Vörösmart), perhaps Várhegy?

Hampel 1905 2: 844–5; 1905 3; Tafn 498–9: SN 1962: 70–1, figs 2–5; ADAM 2002: 433

Site unknown (presumed to be found in Sava riverbed, near Bosanski and Slavonski Šamac)

Aralica, Ilkić 2012: 174–5, cat no 9

Site unknown

Aralica, Ilkić 2012: 175, cat no 10

Site unknown

Kovač 2003: cat no 29; Aralica, Ilkić 2012: 176, cat no 12

Site unknown

Aralica, Ilkić 2012: 177–8, cat no 14

Figure 2: Overlap of Roman communications and Avar-age sites in continental Croatia (Jurakić 2021: sl. 19, 52)

125

Lorena Jurakić and Pia Šmalcelj Novaković state of research, primarily due to major infrastructure works related to the construction of motorways part A3 and A5 of the European transport corridor and consequent protective excavations of significant scope at the end of the 20th and the beginning of the 21st century, which did not result in a significant number of discoveries of new sites of the period (the exception is Jagodnjak-Krčevine site (Dugonjić, Tresić-Pavičić, Novak 2020), although the possibility that they were somehow bypassed cannot be completely written off.

recognizable in the Avar period (Balen-Letunić, RadmanLivaja 2008: 422), but it is somewhat more difficult to conclude whether the late antique communication routes were visible in the area during the 7th and 8th century AD, primarily due to the fact that a great part of the region is prone to floods/floodplain space, which can greatly accelerate the deterioration of transport infrastructure (Milotić 2010: 75). The fact that the Avars, during their campaigns against Byzantium, used ancient communication routes in the Balkan Byzantine provinces until the end of the first third of the 7th century AD does not say much, because these areas were continuously under Byzantine control, unlike Pannonia (Pohl 2018: Map 3).

Furthermore, because they are not directly on the route of Roman communications and presumed communications, it does not necessarily mean that certain positions are certainly not related to Roman infrastructural remains: villa rustica in the Prelog area is also known, as well as stray Roman finds from a number of positions, one of them Draškovec, for example.6

The question also arises: given that the Romans built roads in the most strategically and relief-wise suitable positions and that the routes (often also Roman settlements) largely overlap with prehistoric ones (Milotić 2010: 72),7 is it really about occupying the positions along Roman communications or simply occupying the best strategic positions in a particular area? It seems that the finds from two graves from the Privlaka-Gole njive site, which, like the famous Čađavica find, point to the Black Sea production context and a very rich equestrian grave 88 (silver belt set with Byzantine-Mediterranean motifs and silver phalera, weapons) the earliest from the Stari Jankovci-Gatina site, which was located in the area of ​​the former limes road station, all part of the horizon of the end of the 7th and the beginning of the 8th century AD, allude to the use of supraregional communication network (Black Sea elements) and their possible control/supervision (Šmalcelj Novaković 2022b).

Continental Croatia is a region defined by Sava and Drava rivers; for a number of sites, it is already clear from their position that they must be connected with a river crossing. For example, Sava is suitable for crossing near Zagreb, Sisak and Bosanski Brod (Milotić 2010: 350); in the area of ​​Trnje there was a river crossing during the Middle Ages (Milotić 2010: 350, site Zagreb (Trnje-Kruge), and the accidental finding of a sabre presumed to be from Slavonski Brod, from the riverbed, could easily be related to the existence of a crossing at Bosanski Brod; near Marinbrod there is one of the crossings over Glina river, leading to Pokuplje (Filipec 2010: 135), and south of Prelog itself there is also a crossing over the Drava (Vučetić 2009: 179). ŠarengradKlopare is the best-researched site of this type so far; an Avar-age cemetery whose functioning is presumably related to the favourable crossing of the Danube at that position; it is interesting that the latest research results confirm the birituality of the cemetery (Rimpf, Rapan Papeša, Dizdar 2020).

Flooding of rivers is a big factor in the settlement of the Sava - Drava interfluve: comparing the map of Roman communications and Avar-age sites in the interfluve and the hydrographic map of the Carpathian Basin before major land reclamation works, it can be seen that both Podravina and Posavina communication routes were exposed to flooding. Avar-age sites occur on parts of the route out of the range of flooding, and in the case of Srijem and Baranja mainly on drained “enclaves”, most often elevated terraces near watercourses. The fact that Sava in continental Croatia is very prone to flooding is an additional argument for the theory that this river represented the southern border of the Avar state, an area presumed to be sparsely populated for easier control of space and a kind of “no man’s land” (Filipec 2003; Rapan Papeša 2015).

The phenomenon of the appearance of Avar-age sites on the route of Roman communications is not an isolated phenomenon in our interfluve; in Mezőföld it was noticed that 60% of sites from the early Avar period are within a radius of 1-2 km from Roman infrastructure, and in 50% of cases these buildings were made of stone (so it is possible that they were still visible in space when Avars came to the area) and it is considered that the settlement started by acquiring control of strategic points and communications (Szücsi 2015: 41). According to Josip Brunšmid, who in the 19th century recognized a probable watchtower on the ŠarengradGradac position, according to the appearance of the relief and clearly visible moats around the hill, it can be assumed that the remains of Roman architecture were 6 

The same is with Avar-age sites, of which we give only a few examples: Otok-Gradina, late Avar cemetery on the prehistoric tell (Dimitrijević 1957; Rapan Papeša, Šmalcelj Novaković 2016), equestrian grave on the Bapska-Gradac position also on the prehistoric tell (Burić, Hulina, Lapić 2012), the finds of late Avar mounts in the area of ​​Roman Siscia, prehistoric Segestica (Filipec 2003) and many others.

7 

See BAZA: Roman sites in the area of Prelog.

126

Signposts of Settlement WATER-WORKS, viewed 14 March 2022,

Conclusion There is no doubt that the remains of Roman architecture and roads were the basis for the settlement of many peoples in the early Middle Ages, as already defined strategically favourable roads and places. The settlements of the Avars on the territory of continental Croatia are also connected with this fact. Most Avarage sites are located along the remains of Roman roads and very close to the remains of Roman fortifications and settlements along the Danube in Eastern Slavonia. Like the population during Rome, the Avars settled at favourable river crossings and the crossroads of communication routes. The phenomenon of the overlap of Avar-age and Roman sites, often with a prehistoric background, is not surprising.

Bibliograpy Andrić, S. 2002. Južna Panonija u doba velike seobe naroda. Scrinia Slavonica, Vol. 2: 117-166. Aralica, T. and M. Ilkić 2012. Križnica sablje ili paloša Drugog avarskog kaganata iz Nina, Prilozi Instituta za arheologiju u Zagrebu, Vol. 29: 167-84. Balen-Letunić, D. and I. Radman-Livaja 2008. Izvorište arhivskih podataka: izvadci iz putnih bilježnica Josipa Brunšmida o rimskim nalazima s Dunavskog limesa. Vjesnik Arheološkog muzeja u Zagrebu, Vol. 41: 417-438. Bojčić, Z. 1984. Pregled istraživanja i rasprostranjenosti ranosrednjovjekovnih arheoloških nalaza u istočnoj Slavoniji i Baranji. Izdanja Hrvatskog arheološkog društva, Vol. 9: 211-22. Bošković, D. 2009. Slatkovodni nalazi u zbirci hladnog oružja Hrvatskoga povijesnog muzeja u Zagrebu, in L. Bekić (ed), Jurišićev zbornik. Zbornik radova u znak sjećanja na Marija Jurišića: 98-133. Zagreb: Hrvatski restauratorski zavod/Međunarodni centar za podvodnu arheologiju u Zadru. Brozović, D., A. Kovačec and S. Ravlić (eds.) 2022. Hrvatska enciklopedija (Republika Hrvatska), mrežno izdanje, https://www.enciklopedija.hr/ (14.3.2022.) Bulat, M. 1978. Terenska istraživanja arheološkog odjela Muzeja Slavonije u toku 1978. godine. Arheološki pregled, Vol. 20: 174-179. Burić, M., M. Hulina and J. Lapić 2011. Lokalitet: BapskaGradac. Hrvatski arheološki godišnjak, Vol. 8: 53-5. Csallány, D. 1956. Archäologische Denkmäler der Awarenzeit in Mitteleuropa: Schrifttum und Fundorte. Budapest: Verlag der Ungarischen Akademie der Wissenschaften. Demo, Ž. (ed.) 1996. Vukovar-Lijeva bara, (exhibition catalogue). Zagreb: Arheološki muzej u Zagrebu. Demo, Ž. 2009. Ranosrednjovjekovno groblje bjelobrdske culture: Vukovar-Lijeva Bara (X-XI. stoljeće) 1 i 2. Zagreb: Arheološki muzej u Zagrebu. Dimitrijević, D., J. Kovačević and Z. Vinski 1962. Seoba naroda. Arheološki nalazi jugoslovenskog Podunavlja. Zeman: Narodni muzej Zemun. Dimitrijević, S. 1957. Četiri groba iz slavenske nekropole u Otoku kod Vinkovaca s posebnim osvrtom na naušnice lunulasto-zvjezdolikog tipa, Opuscula Archaeologica Vol. 2: 21-38. Dimitrijević, S. 1966. Arheološka iskopavanja na području vinkovačkog muzeja, rezultati 1957-65., 20 godina muzeja Vinkovci, Acta Musei Cibalensis 1. Vinkovci: Gradski muzej Vinkovci. Dizdar, M., A. Rapan Papeša and A. Rimpf 2017. Rezultati zaštitnih istraživanja kasnoavarodobnoga groblja Šarengrad-Klopare. Annales Instituti Archaeologici, Vol. 13: 9-18.

Just like the Romans and various Germanic peoples after them (Ostrogoths, Gepids, Langobards), the Avars also settled in the strategically most favourable locations in continental Croatia; more specifically, at road junctions, favourable river crossings and along river valleys; most Avar-age sites outline the most important routes of (late) Roman communications. In future field research, special attention should be paid to the positions along the routes of the main Roman communications that are close to watercourses, i.e., outside or above the flooding zone, especially when nearby Drava and Sava rivers. With the current state of research, it is impossible to conclude with certainty whether the Avars settled in this pattern with the explicit intent to follow Roman communication routes or simply focused on settling strategically advantageous positions. The answer is probably in the middle; it is similar to the issue of a smaller number of sites along the Posavina route - the probable cause is that there was a border zone and that the area is prone to floods. In order to solve all the riddles that the past has left us as a legacy, we have no choice but to continue researching and gathering information for future knowledge. Further field research should try to further identify Roman roads and road stations, and new archaeological research should gain new insights into the Avar period in continental Croatia. Especially in central Croatia, where there is a lack of systematic archaeological research on the Avar period. Website The database of antique archaeological sites of the Republic of Croatia  of the Institute of Archeology (DATABASE), viewed 10 March 2022, The European Environment Agency, WATERLOGGED AREAS OF THE CARPATHIAN BASIN BEFORE THE GREAT 127

Lorena Jurakić and Pia Šmalcelj Novaković Dizdar, M., A. Rapan Papeša and A. Rimpf 2020. Rezultati zaštitnih istraživanja kasnoavarodobnog groblja Šarengrad-Klopare 2019. Godine. Annales Instituti Archaeologici, Vol. 16: 8-17. Dugonjić, A., D. Tresić Pavičić and M. Novak 2020. Early medieval cemetery in Jagodnjak (Croatian Baranja): First results of the archaeological research and anthropological analysis, in, Book of abstracts of the international scientific conference: Two Sides of a Belt Strap End ‘Avars on the North and South of the Khaganate: 9–9. Vinkovci: Municipal museum Vinkovci, Archaeological museum in Zagreb. Fettich, N. 1942. Der Fund von Čadjavica. Vjesnik Hrvatskog Arheološkog društva, Vol. 22-23 (1941-2): 55-61. Filipec, K. 2003. Kasnoavarski ukrasni okov (falera) u obliku veprove glave iz Siska. Godišnjak Gradskog muzeja Sisak, Vol. 3-4: 117-144. Filipec, K. 2009. A Slavic Cremation Grave from Lobor. Archaeologia Adriatica, Vol. 3: 347-57. Filipec, K., and D. Roksandić 2009. Lokalitet: BelišćeZagajci 2. Hrvatski arheološki godišnjak, Vol. 5: 19-20. Filipec, K. 2010. Ranosrednjovjekovna bojna sjekira iz Marinbroda (Banovina). Vjesnik Arheološkog muzeja u Zagrebu, Vol. 43: 135-142. Filipec, K. 2015. Donja Panonija od 9. do 11. Stoljeća. Sarajevo: Univerzitet u Sarajevu. Filipec, K. 2017. Ulomci keramičkih posuda iz ranosrednjovjekovnog naselja Stari Jankovci- Gatina, in T. Sekelj Ivančan, T. Tkalčec, T. Krznar and J. Belaj (eds.) Srednjovjekovna naselja u svjetlu arheoloških izvora: 77-109. Zagreb: Institut za arheologiju. Garam, É. 1982. Der Fund von Vörösmart im archäologischen Nachlass der Awarenzeit. Folia Archaeologica, Vol. 33: 187-213. Gračanin, H. 2009. Avari, južna Panonija i pad Sirmija. Scrinia Slavonica, Vol. 9: 7-56. Gračanin, H. 2010. Rimske prometnice i komunikacije u kasnoantičkoj južnoj Panoniji. Scrinia Slavonica, Vol. 10: 9-69. Gračanin, H. and A. Rapan Papeša 2011. Postrimski grad u južnoj Panoniji: primjer Cibala, Scrinia Slavonica, Vol. 11: 7-30. Gračanin, H. and G. Bilogrivić 2014. Postrimski grad u južnoj Panoniji: primjer Siscije. Zbornik Odsjeka povijesnih znanosti i Zavoda povijesnih društvenih znanosti Hrvatske akademije znanosti i umjetnosti, Vol. 32: 1-26. Hampel, J. 1905. Alterthümer des frühen Mittelalters in Ungarn (in drei Bänden). Braunschweig: Friedrich Vieweg und Sohn (reprint 1971, Farnborough: Gregg International Publishers Limited). Hoffiler, V. 1909. Staro groblje u Velikoj Gorici. Vjesnik Hrvatskog Arheološkog Društva, Vol. 10 (1908–9): 120134. Horváth, T. 1935. Az üllōi és a kiskōrösi avar temetõ / Die awarischen Gräberfelder von Üllõ und Kiskōrös. Archaeologia Hungarica 19: 1-128.

Ilkić, M. 2007. Nalazi seobe naroda i ranog srednjeg vijeka iz Sotina. Prilozi Instituta za arheologiju u Zagrebu, Vol. 14: 277-288. Ivaniček, F. 1949. Istraživanje nekropole ranog srednjeg vijeka u Bijelom Brdu. Ljetopis Jugoslavenske akademije znanosti i umjetnosti, Vol. 55: 111-144. Jurakić, L. 2021. Analiza uzoraka naseljenosti u avarskom periodu u Kontinentalnoj Hrvatskoj na temelju rimskih ostataka, Unpublished MA thesis, Geografski odsjek Prirodoslovno-matematičkog fakulteta Sveučilišta u Zagrebu, Geografski odsjek. Karmanski, S. 1976. Još nekoliko nalaza iz perioda seobe naroda koji se nalaze u Odžacima. Bilten Amaterske arheološke sekcije jugozapadne Bačke: 1-9. Kovač, M. 2003. Ubojite oštrice: hladno oružje na području Hrvatske od IX. do kraja XVIII. Stoljeća. Gornja Stubica: Muzeji Hrvatskog zagorja, Muzej seljačkih buna. Lipp, V. 1884. Goldschmiedgewerbe in Pannonien zur Zeit der Völkerwanderung. Szombathely. Milošević, A. (ed.) 2000. Hrvati i Karolinzi 1 i 2 (Rasprave i vrela; Katalog). Split: Muzej hrvatskih arheoloških spomenika. Milotić, I. 2010. Rimska cestovna baština na tlu Hrvatske. Zagreb: Hrvatsko društvo za ceste- VIA VITA. Mrkobrad, D. 1980. Arheološki nalazi seobe naroda u Jugoslaviji / Archäologische Funde der Völkerwanderungszeit in Jugoslawien. Fontes Archaeologiae Iugoslaviae 3, Monografije 6. Beograd: Muzej grada Beograda. Pohl, W. 2018. The Avars: A Steppe Empire in Central Europe, 567-822. Ithaca and London: Cornell University Press. Rapan Papeša, A. 2011. Lokalitet: Nuštar-avarsko groblje (park dvorca Khuen-Belassy). Hrvatski arheološki godišnjak, Vol. 8: 65-68. Rapan Papeša, A. 2012. Prvi nalazi s avarskog groblja na položaju Nuštar. Vinkovci: Gradski muzej Vinkovci. Rapan Papeša, A. 2015. Sava-granica kasnoavarske države: da ili ne?, in B. Ostajmer (ed.) Rijeka Sava u povijesti: 43-54. Slavonski Brod: Hrvatski institut za povijest-Podružnica za povijest Slavonije, Srijema i Baranje. Rapan Papeša, A. and P. Šmalcelj Novaković 2016. Kasnoavarodobno groblje na Gradini u Otoku, katalog izložbe. Vinkovci: Gradski muzej Vinkovci. Rapan Papeša, A., A. Rimpf and M. Dizdar 2018. Rezultati zaštitnih istraživanja kasnoavarodobnog groblja Šarengrad-Klopare 2017. Annales Instituti Archaeologici, Vol. 14: 13-19. Rapan Papeša, A., 2020. What to expect when you’re expecting?, in I. Ožanić Roguljić (ed.) Conference Roads and rivers 2: the transformation of life along the communications from Roman times to Middle ages, Institute of Archaeology, Zagreb, 30 November–1 December 2020, Book of abstracts: 17. Zagreb: Institut for archaeology in Zagreb. Rimpf, A., A. Rapan Papeša and M. Dizdar 2020. Rezultati zaštitnih istraživanja kasnoavarodobnoga groblja Šarengrad-Klopare 2019. godine. Annales Instituti Archaeologici, Vol. 16: 8-19. 128

Signposts of Settlement Sekelj Ivančan, T. and T. Tkalčec 2006. Slavensko paljevinsko groblje na položaju Duga ulica 99 u Vinkovcima. Prilozi Instituta za arheologiju, Vol. 23: 141-212. Sekelj Ivančan, T. 2010. Podravina u ranom srednjem vijeku, Rezultati arheoloških istraživanja ranosrednjovjekovnih nalazišta u Torčecu, Monographiae Instituti archaeologici 2, Zagreb: Institut for archaeology in Zagreb. Sekelj Ivančan, T. 2014. Lokalitet: Sušine, NaseljeVirje. Hrvatski arheološki godišnjak, Vol. 10: 181-183. Simoni, K. 1981. Skupni nalaz oruđa i oružja iz Nartskih Novaka. Vjesnik Arheološkog muzeja u Zagrebu, Vol. 15: 251-266. Simoni, K. 1986. Neobjavljeni okovi i jezičci Blatnica stila u Arheološkom Muzeju u Zagrebu. Vjesnik Arheološkog muzeja u Zagrebu, Vol. 19: 217-228. Sokol, V. 1994. Kasnoantičko i ranosrednjovjekovno razdoblje’(i kat. jedinice), in A. Rendić Miočević (ed.) Zagreb prije Zagreba; arheološka baština Zagreba od pretpovijesti do osnutka biskupije 1094. Godine: 4653, 152. Zagreb: Arheološki muzej u Zagrebu. Szücsi, F. 2015. Térhasználat az Avar kori Mezőföldön / Use of space in Avar age Mezőföld. Történeti Földrajzi Közlemények ser. 2, Vol 3: 41-60. Šercer, M. and M. Ždralović 1979. Sablje, Katalog muzejskih zbirki Vol. 18. Zagreb: Hrvatski povijesni muzej. Šišić, F. 1920. Pregled povijesti hrvatskog naroda od najstarijih dana do 1. decembra 1918., I. knjiga. Zagreb: St. Kugli. Šmalcelj, M. 1973. Privlaka, Gole njive (općina Vinkovci)-Nekropola 7-11. stoljeća-sistematska iskopavanja. Arheološki pregled, Vol. 15: 117-119. Šmalcelj, M. 1976, Privlaka-Gole njive, Vinkovci nekropola 8.-9. stoljeća. Arheološki pregled, Vol. 18: 127-128. Šmalcelj, M. 1981a. Stari Jankovci-Gatina (općina Vinkovci)-avaroslavenska nekropola. Arheološki pregled, Vol. 22: 143-145. Šmalcelj, M. 1981b. Privlaka-Gole njive (Opć. Vinkovci)avaroslavenska nekropola. Arheološki pregled, Vol. 22: 143-145. Šmalcelj, M. and T. Težak-Gregl 1992. Otok-Gradina, in Ministry of Education and Culture (ed.), The war in Croatia-archeological sites. Research undertaken by the Department of archeology at the Faculty of Philosophy: 40-41. Zagreb: University of Zagreb and Ministry of Education and Culture. Šmalcelj, M. 1992a. Privlaka-Gole njive, in Ministry of Education and Culture (ed.), The war in Croatia-

archeological sites. Research undertaken by the Department of archeology at the Faculty of Philosophy: 44-45. Zagreb: University of Zagreb, Ministry of Education and Culture. Šmalcelj, M. 1992b. Stari Jankovci-Gatina, in: Ministry of Education and Culture (ed.), The war in Croatiaarcheological sites. Research undertaken by the Department of archeology at the Faculty of Philosophy: 48-49. Zagreb: University of Zagreb and Ministry of Education and Culture. Šmalcelj Novaković, P. 2022a. Research Problems Concerning 8th-Century Material Heritage in Continental Croatia Revisited: New Directions, Challenges and Ideas (in publication). Šmalcelj Novaković, P. 2022b. Tri Avarodobne Mikrozajednice u Srijemu, Unpublished PhD dissertation, Odsjek za arheologiju Filozofskog fakulteta u Zagrebu. Tomičić, Ž. 1978. Rezultati ranosrednjovjekovnih arheoloških istraživanja u Međimurju i varaždinskoj regiji. Izdanja Hrvatskog arheološkog društva, Vol. 2: 209-222. Tomičić, Ž. 2000. Arheološka slika ranoga srednjeg vijeka na prostoru međuriječja Drave, Dunava i Save, in A. Milošević (ed.) Hrvati i Karolinzi 1: 142-163. Split: Muzej hrvatskih arheoloških spomenika. Uglešić, A. 1994. Nalazi seobe naroda iz zbirke Mate Ilkića. Radovi Filozofskog fakulteta u Zadru 33 (20), (1993/1994): 145-52. Vinski-Gasparini, K. and S. Ercegović 1958. Ranosrednjovjekovno groblje u Brodskom Drenovcu. Vjesnik Arheološkog muzeja u Zagrebu, Vol. 1: 129-161. Vinski, Z. 1952. Naušnice zvjezdolikog tipa u Arheološkom muzeju u Zagrebu s posebnim osvrtom na nosioce nakita Čađavica. Starohrvatska prosvjeta, Vol. 2: 29-56. Vinski, Z. 1954. Povodom našeg prvog priručnika slavenske arheologije s osvrtom na neke još novije publikacije iz tog područja. Peristil, Vol. 1: 199-207. Vinski, Z. 1958. O nalazima u 6. i 7. stoljeću u Jugoslaviji s posebnim obzirom na arheološku ostavštinu iz vremena prvog avarskog kaganata. Opvscvla archaeologica, Vol. 3: 13-55. Vinski, Z . 1959. Ausgrabungen in Vukovar. Acaheologica Iugoslavica, Vol. 3: 99-109. Vinski, Z. 1960. Ranosrednjovjekovni arheološki nalazi na užem i širem području Zagreba. Zbornik iz starog i novog Zagreba 2: 47-65. Vučetić, R. 2009. Trgovište Prelog-obilježja povijesnog razvoja. Radovi Instituta za povijest umjetnosti, Vol. 33: 179-219.

129

Small Items, Big Results Anita Rapan Papeša Abstract: The change in the methodology of field surveys conducted by Municipal Museum Vinkovci, as well as some private collections donated to the Museum, give us a new picture of the distribution of (Late) Avar period sites in the area of Vukovar-Srijem county. Apart from known (old), stray finds and (partly) investigated graveyards, which counted about ten sites known from the area, new finds raise these numbers to more than 50 potentially Avar period sites. Such a potential rise in sites shifts us closer to other, much better-known regions. This peripheral region at the south border of Avar influence hence becomes more important than previously considered. The paper shall present topographical distribution (with emphasis on the Roman background) within maps today and in the past, as well as a short overview of the typology of these finds. KEYWORDS: LATE AVAR PERIOD, STRAY FINDS, TOPOGRAPHY, TYPOLOGY

Introduction After a longer period of low interest for Avar period finds in continental Croatia in the last decade, partly due to some newly discovered sites and finds, partly due to a new generation of scientists, we are back in the picture. The Late Avar period at the south border of the Late Avar state (e.g., today’s Croatia) isn’t very well known (Filipec 2003: 129-131), although we have quite a few interesting finds. The very east of Croatia administratively belongs to Vukovar-Syrmium County, covering an area of about 2450 km2. It lies between the river Danube at the north, the river Sava at the south, the river Vuka at the west and partly streamlet Jošava and state border toward Serbia at the east. Geographically this area is a part of the Pannonian plain which is called the East Croatian plain (Bognar 1994). The latter consists of inflicted loess sediments and alluvial valleys (Njegač 2002: 255). Loess drifts are up to 20 m thick and form higher, arable surfaces of great importance (Roglić 1975: 18). Morphometric analysis shows that altitude in this area is mainly less than 200 m; the highest peak is 294 m, and the lowest is 78 m (Pokos, Turk 2012: 20). Because of its position on the junction between East and Central Europe, and due to geomorphologic reasons (loess/ clay ridge surrounded by oak forest and marshes), this area has been continuously settled since the Neolithic (Durman 2013). Since the end of the 19th century, Avar period sites have been recorded in the area. The earliest known finds ended up in Museums in Zagreb and Budapest, and after the Second World War and the founding of local museums, the finds were kept locally.

Up to the beginning of the 21st century, only 20 sites have been mentioned in the literature that deals with complete overviews of the Avar period sites in the Pannonian basin (Table 1). Some known finds have been published separately or not at all, so the number of known sites slightly rises to 23 (Table 2). However, there are many new small finds which indicate the Avar presence and now we can talk about 60 sites from the period in this area (Figure 1). Published sites and finds A. Bapska - Gradac The site is situated south of the village, on the slopes of Fruška gora. The site was found in the 19th century and since then several excavations have been conducted, but only recently with Avar finds; finds indicate a New Stone Age settlement. The Avar period pot was bought for the Archaeological museum in Zagreb in 1900 (Ivanković 2016: 11). Pot is made on a slow potter’s wheel from unrefined clay, the body is elongated, the edge drawn outwards and rounded, the bottom straight, and decorated with incised oblique and wavy lines, dark brown in colour (Dimitrijević, Kovačević, Vinski 1962: 69; Mrkobrad 1980: 102; Trbuhović 1982: 69; Ivanković 2016: 28). According to analogies, it can be dated to the 7th or beginning of the 8th century AD. During archaeological excavations in 2011 and 2015 a burial of a rider and his horse have been dug out (Burić, Hulina, Lapić 2012: 54-55; Burić, Podgorelec 2016: 84). Grave pit was not visible because it was dug into prehistoric layers (ruins of a Neolithic house). The horse is to the left of the rider. The horse laid on his

Life Along Communication Routes (Archaeopress 2023): 130–156

Small Items, Big Results Table 1: List of sites mentioned in overviews (chronological, when published), (Author: A. Rapan Papeša) SITE

1.

Borovo Gradac

BIBLIOGRAPHY

SITE

Csallany 1956: 90

Seoba 1962: 97

Seoba 1962: 75-76

Mrkobrad 1980: 146

Mrkobrad 1980: 146

Trbuhović 1982: 73

Trbuhović 1982: 71

6.

Sekelj Ivančan 1995: 243

Vinkovci (Meraja)

Szentpéteri 2002: 65

Gračanin 2011: 335

Seoba 1962: 90-91

Seoba 1962:102-103

Mrkobrad 1980: 146-147

Mrkobrad 1980: 146

Trbuhović 1982: 68-72 Sekelj Ivančan 1995: 243

7.

Tomičić 2000: 292 ADAM 2002: 325-326

Vukovar (unknown, Crkva, Lijeva bara, Šamac, Švapsko brdo, Jugopetrol)

8.

Mrkobrad 1980: 147

Vinkovačko Novo Selo

Mrkobrad 1980: 146

Szentpéteri 2002: 44 9.

Podgrađe

Gračanin 2011: 286

Filipec 2003: 134 Mrkobrad 1980: 146

Mrkobrad 1980: 146

Bojčić 1984: 212

Trbuhović 1982: 71

Šmalcelj 1992: 44-45

Tomičić 2000: 148

10.

Szentpéteri 2002: 169

Privlaka Gole njive

Sekelj Ivančan 1995: 240 Szentpéteri 2002: 295-296

Filipec 2003: 133

Filipec 2003: 134

Gračanin 2011: 300

Gračanin 2011: 319

Seoba 1962: 88-89

Mrkobrad 1980: 146

Mrkobrad 1980: 146

11.

Trbuhović 1982: 71

Vučedol

Filipec 2003:135 Gračanin 2011: 337

Bojčić 1984: 212 Otok Gradina

Szentpéteri 2002: 291 Gračanin 2011: 318

Seoba 1962: 77

5.

Trbuhović 1982: 74 Gračanin 2011: 335

Trbuhović 1982: 69 Filipec 2003: 132

Ilok

Szentpéteri 2002: 419-420

Seoba 1962: 97

Seoba 1962: 73

4.

Sekelj Ivančan 1995: 244

Gračanin 2011: 337

Gračanin 2011: 325-326

Bapska Gradac

Trbuhović 1982: 73

Filipec 2003: 135

Filipec 2003:134

3.

Sekelj Ivančan 1995: 242 Filipec 2003: 135

Csallany 1956: 230

Sotin (Višnjica, Vrućak, Zmajevac, unknown)

Bojčić 1984: 217 Szentpéteri 2002: 414-415

Gračanin 2011: 290

2.

BIBLIOGRAPHY

Bojčić 1984: 212

Težak Gregl, Šmalcelj 1992: 40-41

Šmalcelj 1992:48-49

Sekelj Ivančan 1995: 240 Tomičić 2000: 294

12.

Szentpéteri 2002: 275

Stari Jankovci

Sekelj Ivančan 1995: 241 Szentpéteri 2002: 329

Filipec 2003:133-134

Filipec 2003:134-135

Gračanin 2011: 316

Gračanin 2011: 329

131

Anita Rapan Papeša Table 2: List of finds published from sites (author: A. Rapan Papeša) SITE

DESCRIPTION

BIBLIOGRAPHY

Vinski 1954b: 207 parts of belt sets 1.

Borovo Gradac

Dimitrijević et al. 1962: 75-76 Mrkobrad 1980: 89, 91 Simoni 1986: 220 Dimitrijević et al. 1962: 76

pots

Mrkobrad 1980: 103 Trbuhović 1982: 71 Vinski 1954a: 74

2.

Vinkovačko Novo Selo

pot

Dimitrijević et al. 1962: 97 Mrkobrad 1980: 102 Dimitrijević et al. 1962: 90-91 Mrkobrad 1980: 86, 89

parts of belt sets

Simoni 1986: 220

Uglešić 1994: 146-152 Ilkić 2008: 283-284

3.

Ilkić 2010

Sotin (Višnjica, Zmajevac, unknown, Srednje polje)

Vinski 1954a: 75 Dimitrijević et al. 1962: 90 pot

Mrkobrad 1980: 102 Ilkić 2008: 280 Ivanković 2016: 29

arrow

Dimitrijević et al. 1962: 90

settlement

Ložnjak Dizdar, Kušić 2019 Dimitrijević et al. 1962: 69

4.

Bapska Gradac

pot

Mrkobrad 1980: 102 Trbuhović 1982: 69 Ivanković 2016: 28 Hampel 1905: taf. 43, 9

5.

6.

Ilok

Otok Gradina

finger ring

Garam 1993: 26, 109 Garam 2001: 86

pot

Dimitrijević et al. 1962: 77

4 graves

Dimitrijević 1957

finds

Dimitrijević et al. 1962: 88-89 Rapan Papeša 2007

site and finds

Rapan Papeša, Šmalcelj 2016

132

Small Items, Big Results

SITE

DESCRIPTION

BIBLIOGRAPHY

Dimitrijević et al. 1962: 97 Dimitrijević 1966: 42 7.

Vinkovci (Meraja, Glagoljaška 23)

strap end

Mrkobrad 1980: 87 Simoni 1986: 219 Milošević 2000: 109

pot

Dimitrijević 1979: 191

earring

Demo 1996: 93 Vinski 1959: 103-104

graves

Dimitrijević et al. 1962: 102 Karamanski 1976: 5 Demo 2009: 293

8.

Vukovar pots (Crkva, Lijeva bara, Šamac, Švapsko brdo, Jugopetrol)

Dimitrijević et al. 1962: 103 Demo 1996: 69 Ivanković 2016: 29 Dimitrijević et al. 1962: 102 Karamanski 1976: 4-6

parts of belt sets

Mrkobrad 1980: 86, 90 Demo 1996: 15, 69, 72 Demo 2009: 94, 293

9.

Podgrađe

part of belt set

Ercegović 1958: 291 Šmalcelj 1973

excavation

Šmalcelj 1976 Šmalcelj 1981

10.

Šmalcelj Novaković 2020

Privlaka

Heinrich Tamaska 2015 finds

SSiB Rapan Papeša 2007 Bognár, Rapan Papeša 2018

11.

Vučedol

side strap end, beads excavation, graves

12.

Stari Jankovci

Karamanski 1976: 7 Šmalcelj 1981a Rapan Papeša 2008 Rapan Papeša 2007

finds settlement

133

Bognár, Rapan Papeša 2018 Filipec 2017

Anita Rapan Papeša

13.

SITE

DESCRIPTION

BIBLIOGRAPHY

Jarmina

pots

Dimitrijević 1979: 195 Rapan Papeša 2012a Rapan Papeša 2012b Rapan Papeša 2014 Rapan Papeša, Grömer 2015 Jorgić 2015

Nuštar Dvorac

excavation, finds

Rapan Papeša 2016 Bognár, Rapan Papeša 2018 Rapan Papeša 2018 Ašperger et al. 2019 Rapan Papeša 2020a Rapan Papeša 2020b Dizdar et al. 2017

14.

Šarengrad Klopare

excavation

Rapan Papeša et al. 2018 Rimpf et al. 2020

finds

Bognár, Rapan Papeša 2018 Rapan Papeša 2021

15.

Vinkovci Groblje

Excavation, finds

Csuthy et al. 2021 Ostović 2021

right side, legs bent under the body. The rider laid in a supine position; arms next to the body. The orientation of the grave is W-E. Both rider (belt set folded next to his knees) and horse (functional elements-stirrups, bridle, harness phalerae, fittings) were equipped. According to analogies, it can be dated to the beginning of the 8th century AD.

Finds include two almost identical pots made on a slow potter’s wheel, unrefined clay, elongated body, edge drawn outwards and profiled, decorated with incised lines, brown in colour (Dimitrijević, Kovačević, Vinski 1962: 76); allegedly a pot made of so-called yellow pottery (Mrkobrad 1980: 103), currently unavailable; parts of different belt sets: buckle (Dimitrijević, Kovačević, Vinski 1962: 75-76; Mrkobrad 1980: 89), main belt mounts (Dimitrijević, Kovačević, Vinski 1962: 75; Mrkobrad 1980: 89, 91), side belt mounts (Dimitrijević, Kovačević, Vinski 1962: 75; Simoni 1986: 220), side belt strap ends (Dimitrijević, Kovačević, Vinski 1962: 75), almost complete belt set (Hutinec 2013). According to analogies, it can be dated to the 8th century AD.

B. Borovo Gradac The site is situated northwest of the village, on the right bank of the river Danube. The site was discovered in the 19th century; surface finds indicate Prehistory, the Roman period and the High Middle Ages. Parts of Avar period belt sets were bought for the Archaeological Museum in Zagreb in 1907, 1910, and 1911 (Vinski 1954b: 207; Dimitrijević, Kovačević, Vinski 1962: 75; Simoni 1986: 220). Finds were also collected for the Municipal museum Vukovar in 1956 (Bunčić 2007: 45, n. 37) and 1971 (Dorn 1973: 23). Since the Danube continues to undermine the site, new finds were donated to Municipal Museum Vukovar in 2013 (Hutinec 2013).

C. Ilok The exact position of the site is unknown; according to some notes, it could be situated in the southeast part of the town, in the vicinity of the brickyard. Avar period gold fingering was bought for Nemezeti Museum in Budapest in 1894 (Garam 1993: 109), and the Avar period pot was bought for the Archaeological 134

Small Items, Big Results

Figure 1: Position of known sites (letters) and new sites (numbers): A. Bapska Gradac; B. Borovo Gradac; C. Ilok; D. JarminaCiglana; E. Nuštar-Dvorac; F. Otok-Gradina, G. Privlaka-Gole njive; H. Sotin-Višnjica, Sotin-unknown, Sotin-Vrućak, SotinSrednje polje; I. Sotin-Zmajevac; J. Stari Jankovci-Gatina (cemetery); K. Stari Jankovci-Gatina (settlement); L. ŠarengradKlopare; M. Vinkovci Vinkovačko Novo Selo-by the tracks; N. Meraja, Vinkovci, Glagoljaška street 23; O. Vinkovci-Rimokatoličko groblje; P. Vučedol; R. Vukovar-unknown Vukovar- Church Vukovar-Šamac; S. Vukovar-Lijeva bara; T. Vukovar-Sajmište (Švapsko brdo); U. Vukovar-Priljevo (Jugopetrol); 1. Andrijaševci-Gornje njive; 2. Andrijaševci Lončarski poloj; 3. Banovci-Iza škole; 4. Bobota-Stara Bobota; 5. Gaboš-Sokolovac; 6. Ilača; 7. Ilok-Duboki potok; 8. Ivankovo-Dren; 9. Ivankovo-Slatina; 10. Karadžićevo-Grabina; 11. Ludvinci-Barvalo; 12. Markušica-Čeretinci; 13. Mirkovci-Malat; 14. Nuštar-Rimski most; 15. NuštarŽankovac; 16. Orolik-Krčevine; 17. Ostrovo-Gorica; 18. Pačetin-Pazarište; 19. Podgrađe-Donje livade; 20. Podgrađe-Grac; 21. Privlaka-Rudina Selište; 22. Privlaka-Zakota; 23. Retkovci-Opolenac; 24. Retkovci-Didove njive; 25. Sotin-vinograd Ivana Radića; 26. Sotin-Grabovo; 27. Tordinci-Močila; 28. Tovarnik-Orašje; 29. Vinkovačko Novo Selo-Lisičak; 30. Vinkovačko Novo Selo-Crkvište; 31. Vinkovačko Novo Selo-Mokro polje; 32. Vinkovci-unknown; 33. Vinkovci-Jošine; 34. Vinkovci-Ivana Gorana Kovačića 38 Street; 35. Vinkovci-68. Bojne vojne policije street bb; 36. Vođinci-Matakovica (map source: http://webgis.hgicgs.hr/gk300/default.aspx, adopted by A. Rapan Papeša)

Museum in Zagreb in 1901 (Dimitrijević, Kovačević, Vinski 1962: 77). Accidental find of a golden fingering made in Byzantine workshops dated to the last quarter of 7th century AD (Garam 2001: 86) and accidental find of a pot made by hand, unrefined clay, elongated body dated to the end of the 8th century AD.

Avar period pots (Dimitrijević 1979: 194-195) were donated to the museum; one pot was made by hand, unrefined clay, rounded body, decorated with incised horizontal and wavy lines, brown in colour and another pot, made on slow potter’s wheel, unrefined clay, rounded body, edge drawn outwards and profiled, decorated with incised horizontal and wavy lines, brown in colour. According to analogies, those can be dated to the 8th century AD.

D. Jarmina-Ciglana The site is situated west of the village, on the slopes of the Đakovo-Vinkovci loess plateau. The site was found during clay exploitation in 1958, but excavations have not been conducted; finds indicate prehistorical pits and a Middle age graveyard (Korda 1960: 56, 58).

E. Nuštar-Dvorac The site is situated on the southeast outskirts of the village Nuštar. The site was found during construction 135

Anita Rapan Papeša works for a new football field; some graves were dug out without archaeologists being present at the site (Rapan Papeša 2012a: 4). In autumn 2011 altogether 196 graves were excavated at the site, leaving the south and west border of the graveyard undefined (Rapan Papeša 2012b).

Age. Altogether 230 graves were excavated in four campaigns between 1973 and 1980 under the leadership of M. Šmalcelj (Šmalcelj 1973; Šmalcelj 1976; Šmalcelj 1981a). According to grave finds the cemetery can be dated throughout the 8th century AD; only some finds have been mentioned or published from this, up to the date, the biggest cemetery from the period excavated in Croatia.

According to analogies, this graveyard can be dated throughout the 8th century AD. The complete publication is being prepared, but some finds have already been processed (spindle whorls (Rapan Papeša 2014), an overview of pottery finds (Rapan Papeša 2016), yellow pottery (Rapan Papeša, Bognár 2018), bone and stone objects (Rapan Papeša 2018), “antiquities” (Rapan Papeša 2020a), an interesting and rare type of belt set (Rapan Papeša 2021c) and some have been presented but not published), as well as numerous interdisciplinary types of research. So far anthropological analysis has been conducted, with preliminary results published (Premužić et al. 2018), as well as a special case of trepanation with so far only C14 date from the site (Premužić et al. 2016). Also, zooarchaeological analysis of animal bones has been done (Premužić et al. 2016a), which resulted in a detailed analysis of horse skeletons (Trbojević Vukičević et al. 2017) and horse genomes (Orlando et al. 2019). Also, an analysis of eggshells has confirmed domestic chicken (Rapan Papeša 2021b: 603). Archaeobotany has been conducted on selected samples on two levels: macro (earth from grave fills) and micro (earth from pots) (Rapan Papeša et al. 2015). Also, remains of leather and textile on metal objects have been processed (Rapan Papeša, Grömer 2015), as well as analysis of glass beads (Ašperger et al. 2018). Selected samples have been chosen for stable isotope analysis (Vidal-Ronchas et al. 2019), and the remaining samples are being processed at the moment. Also, partial human genome analysis with interesting relative pairs is about to be published.

H. Sotin-Višnjica The site known by the find was not exactly and precisely located within today’s Sotin, but most likely it is called Majcanov vinograd today; it is situated in the middle of today’s village, just above the high right Danube bank. A ceramic pot from the first half of the 8th century AD was found at the site; made by hand from unrefined clay, elongated body, edge drawn outwards and profiled, decorated with incised lines, brown in colour (Vinski 1954: 75; Dimitrijević, Kovačević, Vinski 1962: 90; Mrkobrad 1980: 102; Ivanković 2016: 20; Szentpéteri 2002: 325). H. Sotin-Vrućak The site is located on the low right bank of the river Danube and finds from this site most probably come from the collapsing high Danube bank. Some finds have been lost in the Homeland war, but different parts of belt sets and earrings (Uglešić 1994; Ilkić 2008; Ilkić 2010) dated to the 8th century AD come from the site. I. Sotin-Zmajevac

F. Otok-Gradina

The site is located east of the village of Sotin. It is a multilayered site with finds from the Stone age, Bronze age, Late iron age and Middle age periods.

The site Gradina situated southeast of the town Otok is an artificial tell site from prehistory (New Stone Age, Late Iron Age) and Avar period graves have been dug into these older layers. It was excavated in 1957 by S. Dimitrijević and in 1970 by S. Dimitrijević and M. Šmalcelj. Altogether 22 graves have been unearthed in four probes about 150 m2 big (Rapan Papeša, Šmalcelj Novaković 2016).

Different parts of belt sets, mount and buckle, dated to the middle 8th century AD are known from the site (Dimitrijević, Kovačević, Vinski 1962: 90-91; Mrkobrad 1980: 86, 89; Simoni 1986: T.2: 15). H. Sotin-unknown

The site has been dated to the 8th century; some finds can be dated to the early 8th and some to the late 8th century AD (Šmalcelj 2013).

From an unknown position in Sotin for the Archaeological Museum in Zagreb from Đ. Haupt gilded mount was purchased.

G. Privlaka-Gole njive

Mount used as a horse harness part in the shape of a knot can be dated to the 8th century AD (Simoni 1986: 220; Milošević 2000: 104; AiS 2019: 78).

The site is located next to prehistoric tell settlements from the New Stone Age, Bronze Age and Late Iron 136

Small Items, Big Results The find is a belt strap end, dated to the early 9th century AD, made of gilded bronze and decorated in Blatnica style (Dimitrijević, Kovačević, Vinski 1962: 97; Dimitrijević 1979: 194; Mrkobrad 1980: 87; Simoni 1986: 219; Milošević 2000: 109).

H. Sotin-Srednje polje The site is located southeast of the village centre and during systematic excavations, parts of Early Medieval settlement were explored (Ložnjak Dizdar, Dizdar, Kušić 2018: 31).

M. Vinkovačko Novo Selo-by the tracks

The processing of the finds is still ongoing within the project of the Institute of archaeology in Zagreb.

Next to the tracks of the Vinkovci - Slavonski Brod railway line prehistorical and medieval finds were recovered in the 19th century (Ljubić 1888: 2).

J. Stari Jankovci-Gatina (cemetery) The site is located on the high bank of streamlet Gatina on the north edge of the village of Stari Jankovci. During clay exploration in 1978, an unknown number of graves was destroyed; afterwards 15 graves and finds were identified. Systematic excavations were conducted later on in 1986 and 1988; on the surface of approximately 600 m2, 88 graves were excavated (Šmalcelj 1981b; Rapan Papeša 2008).

A vessel, a pot made by hand, was found there (Vinski 1954: 74; Dimitrijević, Kovačević, Vinski 1962: 97; Mrkobrad 1980: 102) dated to the 8th century AD. N. Vinkovci-Glagoljaška street 23 The site is situated west of Vinkovci city centre, on a higher right Bosut riverbank. The find was collected by Professor S. Dimitrijević and that piece is nowadays unavailable.

According to finds graveyard can be dated from the end of the 7th to the second half of the 8th century AD.

The find is a fragment of a ceramic vessel, a pot, dated to the 8th century AD (Dimitrijević 1979: 191).

K. Stari Jankovci-Gatina (settlement) The settlement is located north of the cemetery, divided by a deep depression and a creek. Trial trenches were excavated in 1988, and the material is partly published (Filipec 2017).

O. Vinkovci-Rimokatoličko groblje The site is situated west of Vinkovci centre, in the area of today’s Roman Catholic and the town cemetery. The site was found during the construction of a new grave field; some skeletal graves have been damaged during the works, but they stopped as machines stumbled upon the brick tomb. Eventually, nine graves have been unearthed, some of them being plundered shortly after burial. Finds include parts of belt sets, jewellery, pottery and horse equipment and harness. According to finds, this site can be dated to the 8th century AD.

Due to deficient excavation documentation and mixed finds, it is hard to determine the exact objects and material within them. L. Šarengrad-Klopare The site was noticed during earth exploration in 2006 when an unknown number of graves was destroyed. Only two vessels were preserved for the Ilok town museum (Dizdar, Rapan Papeša, Rimpf 2017: 9). Protective excavations were carried out in 2016, 2017 and 2019. A surface of about 1300 m2 was excavated and 77 graves were unearthed, of which seven were cremation burials (Dizdar, Rapan Papeša, Rimpf 2017; Rapan Papeša, Rimpf, Dizdar 2018; Rimpf, Rapan Papeša, Dizdar 2020).

P. Vučedol The site is situated on the high right bank of the Danube River, known as the eponymous site of the Copper age Vučedol culture. Numerous excavations conducted have not revealed Early Medieval layers but finds from Prehistory and High Middle Ages have been recorded (Balen 2006: 300; Bunčić 2018: 138-141). Finds from this site were given to S. Karamanski in the 1970s. Finds include a side belt strap end and three beads (Karamanski 1976: 7). According to analogies the site can be dated to the end of the 8th century AD.

The site can be dated from the end of the 7th century up to the end of the 8th century AD. N. Vinkovci-Meraja The site is situated west of Vinkovci city centre, on a higher right Bosut riverbank. It is the site with an older and a younger medieval church and graveyard around it, but with finds of Roman period graves from the Western necropolis of Cibalae as well.

R. Vukovar-unknown From an unknown position in Vukovar, one find is known. 137

Anita Rapan Papeša Find of a side strap belt end (Dimitrijević, Kovačević, Vinski 1962: 102) that can be dated to the middle 8th century AD, according to analogies.

Ceramic pots made on a slow potter’s wheel from the site are dated to the end of the 8th and early 9th century AD (Dimitrijević, Kovačević, Vinski 1962: 103; Demo 1996: 69; Ivanković 2016: 25).

R. Vukovar-Church

Podgrađe - mystery or not?

The site is situated on the high right bank of the Danube River, around the church of St. James and St. Philip. This is presumed to be a Roman outpost (Pinterović 1968: 69-70), a Medieval and Ottoman period fortress. Find from this site was bought in the 19th century.

The site Podgrađe in Vukovar-Srijem County is first listed in an overview made by D. Mrkobrad in 1980, and later on, it was taken for granted (Szentpéteri 2002: 291; Filipec 2003: 134; Gračanin 2011: 318). He took data from an article published in 1958, but in that paper, the authors list Podgrađe within sites in Dalmatia and the Adriatic area (Vinski-Gasparini, Ercegović 1958: 152153). Hence there are several places in Croatia with that name; this is a classic example of the confusion that was later on taken for granted. The coastal site Podgrađe, with finds from Christian graves, was published as such in 2009; that horizon is dated from the 8th to 11th century (Petrinec 2009: 93) and particular finds to the late 9th and 10th century AD (Petrinec 2009: 209, 220), but there is no mention of belt set parts (as described in original publication).

The find is an earring, dated to the beginning of the 9th century AD (Demo 1996: 59). S. Vukovar-Lijeva bara The site is situated on a small hilltop in the east part of the town Vukovar. Bijelo brdo culture cemetery and prehistoric settlement and graves were excavated in 1951-1953 (Demo 1996: 28-31). In two medieval graves, secondary used Avar period items were recovered. Finds include belt strap ends (Demo 2009: 293-294) which originally date to the third quarter of the 8th century AD.

Anyway, although data listed about Podgrađe in overviews is an unfortunate misinterpretation according to some new finds (listed below) we can now state that Podgrađe is an Avar period site.

U. Vukovar-Priljevo (Jugopetrol)

Unpublished sites and finds

The site is located in the western part of the town Vukovar, at the place where rivers Vuka and Danube approach the most and used to interflow during high waters, before the regulation in the 19th century. The finds were recovered from two destroyed graves during the construction of a gas station in 1955.

1. Andrijaševci - Gornje njive The site is situated northwest of the village, on the right bank of river Bosut. The site was found during a field survey; surface finds indicate Late Iron Age, Roman period and High Middle Age. According to aerial archaeology, it is a New Stone Age site as well (Kalafatić, Šiljeg 2018: 92).

A pot and two side belt strap end were preserved from the site, dating to the third quarter of the 8th century AD.

Avar period side strap end was found among the material collected in 2016. According to analogies, it can be dated to the second half of the 8th century AD.

T. Vukovar - Sajmište (Švapsko brdo) The site is located on the high right bank of the river Vuka. Finds from a destroyed grave were given to S. Karamanski in the 1970s, first kept in Odžaci and afterwards in Municipal Museum Vukovar, but nowadays lost due to Homeland war in Croatia.

2. Andrijaševci - Lončarski poloj The site is situated southeast of the village, on the left bank of river Bosut. The site was found during a field survey; surface finds indicate the Roman period.

Finds include an incomplete belt set dated to the end of the 7th and early 8th century AD (Karamanski 1976: 5).

Two Avar period side strap ends were found among the material collected in 2018. According to analogies, it can be dated to the first half of the 8th century AD.

R. Vukovar - Šamac

3. Banovci-Iza škole

The site is located on the high bank of the Danube River, between the church and Lijeva Bara. Two pots were bought in the early 20th century from this site.

The site is situated northeast of the village. A local associate Ivan Kelava pointed the Museum to this site 138

Small Items, Big Results which was included in field surveys; surface finds indicate Prehistory and the Roman period. In October 2016 magnetic prospecting was conducted, with visible archaeological structures, but excavation is needed to prove their origin. The site is in the vicinity of Roman road Cibalae-Sirmium, on a slightly erected field; in the foothill, there used to be a swamp which is drained today.

8. Ivankovo-Dren The site is situated east of Vinkovci, on the left bank of river Bosut. A local associate pointed the Museum to this site in 2014 when donated a large amount of finds collected for more than 40 years on the site. Since then, several field surveys have been conducted (Vulić 2015: 99), as well as magnetic prospection. Surface finds indicate Bronze Age, Late Iron Age, Roman period production centre, and High Middle Age.

Avar period side strap ends, and mounts were found among the material collected in 2013, 2014 and 2017. According to analogies, it can be dated to the 8th century AD.

Avar period mount pendant, horse harness fitting, belt buckle and earring were found at the site. According to analogies, it can be dated to the first half of the 8th century AD.

4. Bobota-Stara Bobota The site is situated west of the village, on the slightly elevated ground surrounded in the past with swamps, drained today. The site was found during field works in 1985 when a Roman brick tomb was damaged; finds include glass and two coins of Constantius II and a recent field survey proves remains of Roman architecture (Hutinec 2009: 111).

9. Ivankovo-Slatina The site is situated east of Vinkovci, divided by a modern canal as well as paleo flow visible on satellite images from above mentioned Dren site. The site was found during a field survey (Vulić 2015: 99); surface finds indicate the Early and Late Iron Age, the Roman period and the High Middle Age.

The Avar period mount pendant was donated by Josip Štipak in 2014. According to analogies, it can be dated to the first half of the 8th century AD.

Avar period side strap ends, parts of horse harness ornaments and an arrow were found among the material collected in 2014-2017. According to analogies, it can be dated to the 8th century AD.

5. Gaboš-Sokolovac The site is situated west of the village, on the right bank of stream Vučica. The site was found during a field survey; surface finds indicate the Early Roman period.

10. Karadžićevo-Grabina The site is situated west of the village, in a plain that was surrounded by swamps in the past, but now drained. Finds from the site were donated by Steve Gaunt and Domagoj Jovanić in 2012. Avar period side strap ends and a mount were found. According to analogies, it can be dated to the 8th century AD.

Avar period side strap end and damaged main strap end were found among the material collected in 2016. According to analogies, it can be dated to the first quarter of the 8th century AD.

11. Ludvinci-Barvalo

6. Ilača

The site is situated southeast of the village, divided by a modern canal from the site Močilo in Tordinci. Finds were donated by Mirko Adžaga in 2012. An Avar period side strap end was found at the site. According to analogies, it can be dated to the end of the 8th century AD.

From an unknown site in the vicinity of the village, an iron arrow point was donated to the Museum in 1954. According to analogies, it can be dated to the 8th century AD. 7. Ilok-Duboki potok

12. Markušica-Čeretinci

The site is situated southeast of town, high (150 m) above stream Lisičak. The site was found during a field survey; surface finds indicate the Roman period and the High Middle Ages.

The site is situated southwest of the village, on the slopes of the Đakovo-Vinkovci loess plateau; at the bottom of the site stream Vučica springs. Finds were donated by Mirko Adžaga in 2012. Avar period side strap ends, and mounts were found at the site. According to analogies, it can be dated to the second half of the 8th century AD.

Avar period side strap end was found among the material collected in 2016. According to analogies, it can be dated to the first half of the 8th century AD. 139

Anita Rapan Papeša 13. Mirkovci-Malat

New stone age and Middle Ages have been recorded there (Hutinec 2009: 113). The find was donated to the museum by M. Adžaga. A piece of main strap end dated to the 8th century comes from the site.

The site is situated east of the village, on the slope above stream Vidor. The site was found during a field survey (Vulić 2015: 99); surface finds indicate New Stone Age, Late Iron Age, Roman period and High Middle Ages. Aerial archaeology points to several New Stone Age sites (Kalafatić, Šiljeg 2018: 96). Avar period side strap end, belt buckle, and 2 earrings were found among the material collected in 2014-2016. According to analogies, it can be dated to the second half of the 8th century AD.

19. Podgrađe-Donje livade The site is located southwest of the village Podgrađe and south of river Bosut. Surface finds indicate a Roman period site, but among collected material also Avar period items were noticed. The main strap end mount pendant comes from this site and can be dated to the 8th century.

14. Nuštar-Rimski most The site is situated northwest of the village, along the river Vuka. The site was found and reported to the Museum by Mirko Adžaga. During low water levels wooden pylons of an old bridge are visible (Krznarić Škrivanko, Rapan Papeša, Vulić 2013: 57), and on the banks finds from the Roman period are prevailing. An incomplete belt set originates from the site. According to analogies, it can be dated to the second half of the 8th century AD.

20. Podgrađe-Grac The site is located on the left bank of the river Bosut on a settlement from the new Stone Age and Medieval fortress St. Lawrence (Szentlőrinc). The site was found by a museum curator in 1954 (Korda 1960: 51). He collected an arrowhead there. The arrow can be dated to the Early Medieval period. 21. Privlaka-Rudina Selište

15. Nuštar-Žankovac

The site is located south of the village Privlaka; surface finds indicate the Roman period and Middle age site. It was found during a field survey in 2017.

The site spreads south from road Vinkovci-Vukovar; it has been known since 1977 as a Roman period burial ground belonging to a villa rustica or smaller settlement. Finds from the Avar period have been donated by Mirko Adžaga. A broken side strap end and belt mount come from the site. According to analogies, it can be dated to the first half of the 8th century AD.

Several pieces of belt sets come from the site. According to analogies those date into the middle 8th century AD. 22. Privlaka-Zakota

16. Orolik-Krčevine The site spreads south of village Orolik, on the west banks of streamlet Savak. Avar period finds were donated to the Museum by I. Kelava and M. Adžaga. Two broken items come from the site: main and side strap ends according to analogies can be dated to the second half of the 8th century AD.

The site is next to the site Selište, just a little more southern. It was found during a field survey in 2019, and surface finds indicate Late Iron Age, Roman period and Middle age period. Finds from this site include a broken side strap end, a broken buckle and 2 earrings. According to analogies, it can be dated to the third quarter of the 8th century AD.

17. Ostrovo-Gorica

23. Retkovci-Opolenac

The site is situated on the slopes of the Đakovo-Vinkovci loess plateau, southwest of the village of Ostrovo. The site belongs to the High Medieval period (medieval estate Hegyfalu). Finds from the Avar period have been donated by S. Gaunt. A piece of the belt loop and an earring come from the site. According to analogies, the site can be dated to the second half of the 8th century AD.

The site is situated northeast of the village Retkovci. It was unknown prior to the field survey conducted in 2016. Surface finds indicate the Roman period and Middle age period. The find from the site is a piece of buckle dated to the third quarter of the 8th century AD. 24. Retkovci-Didove njive

18. Pačetin-Pazarište

The site is situated east of site Opolenac. It was discovered during a field survey in 2019 and surface finds indicate the Roman and Middle age period.

The site is located southeast of the village Pačetin, on a higher level surrounded by low waters. Finds from the 140

Small Items, Big Results Three different main belt mounts from the site can be dated around the middle 8th century AD.

The finds include a vessel and an iron arrow blade tip. According to analogies, it can be dated to the late 8th / early 9th century AD.

25. Sotin-vinograd Ivana Radića

31. Vinkovačko Novo Selo-Mokro polje

The site is located on the high right bank of the Danube in the eastern part of the village Sotin. The Avar period cemetery is presumed to be in this location (Dizdar 2010: 18).

The site is situated between villages Vinkovačko Novo Selo and Ivankovo. It was found during field surveys, with surface finds from Prehistory and the High Middle Ages (Rapan Papeša, Petković 2017: 502-503). Finds include broken earring and side strap belt end. According to analogies, it can be dated to the second half of the 8th century AD.

Finds from this site (belt set parts, tinder, knives) were lost during the Homeland war in Croatia, but according to some handmade preserved sketches they can be dated to the first half of the 8th century AD.

32. Vinkovci-unknown

26. Sotin-Grabovo

From the initial museum collection collected and sold by M. Medvedović, we have some finds from this period; unfortunately, the exact site was not noted. Finds include three iron arrow blade tips and a spear dated to the 8th century AD.

The site is located southwest of Sotin, in the vicinity of a Roman road. Finds from this site were donated to the Museum by S. Andrijašević and J. Štipak. Finds from this site include a broken belt strap end and a boar head appliqué. According to analogies, the site can be dated to the middle 8th century AD.

33. Vinkovci-Jošine

27. Tordinci-Močila The site is situated northeast of the village Tordinci; in the past, this area was seasonally floated by swamp Palača during high waters. Finds were donated to the Museum by D. Jovanić. The broken main strap belt end comes from this site. According to analogies, the site can be dated to the second half of the 8th century AD.

The site is situated southeast of Vinkovci on the Bosut river bank. This multilayered site was partly excavated in 2011 (some Early Medieval pits were found (Koprivnjak 2012) and Avar period finds were noticed during a field survey in 2015 (Vulić 2015: 99). Find from this site includes a side belt strap end. According to analogies, the site can be dated to the end of the 8th century AD.

28. Tovarnik-Orašje

34. Vinkovci-Ivana Gorana Kovačića 38 Street

The site is situated northwest of the village Tovarnik, on slopes and it is presumed to be the Roman site Ulmo on the Cibalae-Sirmium road. Finds from this site were donated to Museum by J. Štipak. The broken main strap belt end comes from this site. According to analogies, the site can be dated to the second half of the 8th century AD.

The site is situated west of Vinkovci city centre, on a higher right Bosut river bank. The gardens on the right side of that street are situated on the Roman moat, and the Western necropolis lies under houses on that side of the street. The find is an iron arrow blade tip, dated to the 8th century AD. 35. Vinkovci-68. Bojne vojne policije street bb

29.Vinkovačko Novo Selo-Lisičak

The site is situated southwest of Vinkovci city centre, on the higher right bank of river Bosut which was settled in continuity since the New Stone age up to today. The solitary grave of an older girl was excavated in the vicinity of a Roman tegulae oven. Finds include a pair of bronze earrings dated to the 8th century AD.

The site is situated underneath the site BorinciCrkvište, on the southern slopes. It was found during a field survey in 2015 (Vulić 2015: 99). The broken side strap belt end comes from this site. According to analogies, the site can be dated to the second half of the 8th century AD.

36. Vođinci-Matakovica

30. Vinkovačko Novo Selo-Crkvište

The site is situated between the railway Vinkovci-Zagreb and road Ivankovo-Vođinci. It is a multilayered site with three hilltops with finds from prehistory and the Medieval age (Petković 2006: 44-47, 286; Šiljeg, Kalafatić 2015: 138). Some finds were donated by S. Gaunt and

The site is situated on a hilltop of the Đakovo-Vinkovci loess plateau, and finds were collected during orchard planting and brought to the Museum. It is a multilayered site with finds from prehistory and the Middle Ages. 141

Anita Rapan Papeša some were collected during a field survey. Finds include side belt strap end and broken belt mounts. According to analogies, the site can be dated to the first half of the 8th century AD.

2. Andrijaševci Lončarski poloj 2.1. Upper part of a side belt strap end, cast in bronze, straight sides, slit for the fastening decorated on both sides with tendrils, back side flat, front side central decoration plate with a tendril, l. 21 mm, w. 13 mm (Plate I:2). Type: Zábojník 87 Closest analogy: Nuštar grave 34

Concluding remarks Why are these individuals’ finds important? Many archaeologists would claim that such surface finds without archaeological context have no worth at all. Finds that do not originate from field excavation, according to their opinion, should not be included in the modern museum concept and archaeologists should take such items with caution, due to lawbreaking (Bunčić, Solter 2020: 224-225). I do agree that missing archaeological context and lack of research from surface finds represent the main weakness of such finds. We do not know whether we deal with graves, settlements or hoards, or are those just pieces lost or reused? On the other hand, this is also the main strength of surface finds: they provide new data which allows a better knowledge of the period. They also enable a good and solid basis for future projects and excavations; hence they reduce the surface for research. If not picked up and brought to museums those finds would be lost in agricultural works or stolen by malicious individuals and sold on the black market. Many items presented below show traces of damage done by agriculture and they were “saved at the last minute”.

2.2. Upper part of one side of two-pieced side belt strap end, cast in bronze, straight sides, animal heads on the top, central decoration plate with a flat-leaf tendril, l. 22 mm, w. 13 mm (Plate I:3). Type: Zábojník 113 Closest analogy: Nuštar grave 34 3. Banovci; Iza škole 3.1. Main belt mount, open cast in bronze, longish, straight sides and rounded top, hinge and pendant missing, central decoration plate with fleuer-de-lis motif framed with imitation of hollow beaded wire, l. 23 mm, w. 11 mm (Plate I:4). Type: Zábojník 258 Closest analogy: Nuštar grave 83 3.2. The Lower part of a side strap end, cast in bronze, longish, straight sides, rounded bottom, back side flat, front side central decoration plate with vine, l. 32 mm, w. 11 mm (Plate I:5). Type: Zábojník 113 Closest analogy: Nuštar grave 34

Ostensibly small items from the Avar period collected in the area of Vukovar-Srijem County in Croatia provide many important data. New finds and sites raised in number due to more research, somewhat changed methodology (usage of metal detectors) and deeper and deeper ploughing of arable areas. The distribution of the Avar period finds provides a more complex picture of sites density and their connection to the Roman road network. Obtained results are very well compared with other parts of the Avar state. For some sites this is most likely the last chance to obtain archaeological information and those should be incorporated into some future projects and research.

3.3. Side strap end, cast in bronze, longish, narrowed in the middle, bottom tongue-shaped, undecorated slit for fastening, on both sides’ central decoration plate with tendrils, l. 29 mm, w. 10 mm (Plate I:6). Type: Zábojník 105 Closest analogy: Otok 4. Bobota Stara Bobota 4.1. Main belt mount pendant, cast in bronze, decorated with palmette and grape motifs, l. 27 mm, w. 23 mm (Plate I:8). Type: Zábojník 248

Catalogue of new finds (l.-length, w.-width) Belt set parts

5. Gaboš Sokolovac

1. Andrijaševci Gornje njive

5.1. Upper part of the main belt strap end, cast in bronze, straight sides, slit for fastening decorated on one side with griffon, on the other with a tendril, central decoration plate with animal fight motif (griffons in a row) on the front side and tendril on the back side, l. 49 mm, w. 27 mm (Plate I:10). Type: Zábojník 14 Closest analogy: Otok grave 4

1.1. Main belt mount, cast in bronze, longish, straight sides and rounded top, central decoration plate with a tendril, l. 25 mm, w. 14 mm (Plate I:1). Type: Zábojník 548 Closest analogy: Nuštar grave 5

142

Small Items, Big Results 10. Karadžićevo Grabina

5.2. Half of the main belt mount, propeller-shaped, imitation of hollow beaded wire on edges, l. 35 mm, w. 13 mm (Plate I:9). Type: Zábojník 157 Closest analogy: Nuštar grave 21

10.1. Side belt mount, cast in bronze, shield-like, decorated with tendrils, l. 14 mm, w. 12 mm (Plate I:17). Type: Zábojník 337 Closest analogy: Nuštar grave 196

5.3. Side belt strap end, cast in bronze, longish, straight sides, rounded bottom, undecorated slit for fastening, back side flat, central decoration plate with scale motif, l. 37 mm, w. 11 mm (Plate I:7). Type: Zábojník 92 Closest analogy: Ivankovo Slatine

10.2. One side of two-pieced side strap belt end, open cast in bronze, longish, straight sides, rounded bottom, animal heads on top, central decoration plate with a tendril, l. 31 mm, w. 13 mm (Plate I:18). Type: Zábojník 113 Closest analogy: Nuštar grave 2

7. Ilok Duboki potok

10.3. Side strap belt end, open cast in bronze, longish, sides straight, bottom rounded, slit for fastening undecorated, central decoration plate with fleur-de-lis motifs, l. 29 mm, w. 13 mm (Plate I:19). Type: Zábojník 113 Closest analogy: Nuštar grave 179

7.1. Side strap belt end, cast in bronze, longish, narrowed in the middle, bottom tongue-shaped, undecorated slit for fastening, on both sides’ central decoration plate with tendrils, l. 22 mm, w. 10 mm (Plate I:11). Type: Zábojník 91 Closest analogy: Otok

11. Ludvinci Barvalo

8. Ivankovo Dren

11.1. Side strap belt end, open cast in bronze, longish, sides straight, bottom rounded, slit for fastening undecorated, central decoration plate with fleur-de-lis motifs, l. 22 mm, w. 11 mm (Plate II:1). Type: Zábojník 113 Closest analogy: Nuštar grave 179

8.1. Belt buckle mount, open cast in bronze, frame and hinge missing, tongue-shaped, straight sides, central decoration plate with double tendril motif, l. 31 mm, w. 19 mm (Plate I:12). Type: Zábojník 144 Closest analogy: Nuštar grave 12

12. Markušica Čeretinci

8.2. Main belt mount pendant, open cast in bronze, triangular, stylised animal face, l. 15 mm, w. 12 mm (Plate I:13). Type: Zábojník 250

12.1. Main belt mount, open cast in bronze, longish, straight sides, rounded top, hinge and pendant missing, central decoration plate with fleur-de-lis motif, l. 22 mm, w. 12 mm (Plate I:21). Type: Zábojník 257

9. Ivankovo Slatine

12.2. Upper part of a side strap belt end, cast in bronze, straight sides narrow towards bottom, undecorated slit for fastening, backside flat, central decoration plate with a tendril, l. 12 mm, w. 14 mm (Plate I:20). Type: Zábojník 91

9.1. Main belt mount, cast in bronze, shield-shaped, undecorated, pendant decorated with stylised faces, l. 31 mm, w. 22 m (Plate I:14). 9.2. Side strap belt end, cast in bronze, longish, sides straight, bottom rounded, slit for fastening decorated with incised lines, central decoration front plate with scales, back side plate decorated with tendrils, l. 26 mm, w. 10 mm (Plate I:15). Type: Zábojník 92 Closest analogy: Gaboš Sokolovac

13. Mirkovci Malat 13.1. Belt buckle mount, cast in bronze, frame and hinge missing, rounded end, straight sides, central decoration plate with tendril motif, l. 32 mm, w. 20 mm (Plate I:22). Type: Zábojník 132 Closest analogy: Nuštar grave 135

9.3. Side strap belt end, open cast in bronze, longish, sides straight, bottom rounded, slit for fastening undecorated, central decoration plate with fleur-de-lis motifs, l. 34 mm, w. 13 mm (Plate I:16). Type: Zábojník 107 Closest analogy: Otok grave 13

13.2. One half of two-pieced side strap belt end, open cast in bronze, longish, straight sides, rounded bottom, animal heads on top, central decoration plate with a vine motif, l. 34 mm, w. 14 mm (Plate II:2). Type: Zábojník 113 Closest analogy: Nuštar grave 8 143

Anita Rapan Papeša 14. Nuštar Rimski most

16. Orolik Krčevine

14.1. Main belt strap end slit for fastening, cast in bronze, decorated on both sides with animal motif, l. 18 mm, w. 29 mm (Plate II:3). Type: Zábojník 14 Closest analogy: Vukovar Sajmište

16.1. Middle part of a broken main strap belt end, cast in bronze, longish, straight sides, central decoration plate with flat tendril motif, l. 46 mm, w. 29 mm (Plate II:16). Type: Zábojník 57 Closest analogy: Nuštar grave 12

14.2.-5. Four pieces of main belt mounts, cast in bronze, longish, straight sides, rounded top, hinge and pendants, central decoration plate with tendrils in a frame of a zig-zag motif, l. 30 mm, w. 11 mm (Plate II:47). Type: Zábojník 250 Closest analogy: Nuštar grave 79

16.2. the Lower part of a side belt strap end, open cast in bronze, longish, sides straight, bottom rounded, central decoration plate with fleur-de-lis motifs, l. 21 mm, w. 15 mm (Plate II:17). Type: Zábojník 107 Closest analogy: Otok grave 13 17. Ostrovo Gorica

14.6. Main belt mount, like previously described, but without a pendant, l. 20 mm, w. 11 mm (Plate II:8). Type: Zábojník 544 Closest analogy: Nuštar grave 79

17.1. Bronze casted belt loop decoration plate, oblong, central decoration plate with tendril motif, l. 26 mm, w. 13 mm (Plate II:19). Type: Zábojník 271

14.7.-9. Three pieces of side strap mounts, cast in bronze, triangular, decorated with scale motifs, l. 12 mm, w. 12 mm (Plate II:9-11). Type: Zábojník 357 Closest analogy: Nuštar grave 79

18. Pačetin Pazarište 18.1. Belt end mount, cast in bronze, straight sides, round bottom, undecorated, l. 26 mm, w. 25 mm (Plate II:20).

14.10. Side belt strap end, cast in bronze, double, longish, straight sides, rounded bottom, central decoration plate on both sides with flat tendrils in a frame of a zigzag motif, l. 31 mm, w. 12 mm (Plate II:12). Type: Zábojník 99 Closest analogy: Nuštar grave 79

19. Podgrađe Donje Livade 19.1. Main belt mount pendant, open cast in bronze, rectangular with tongue-shaped bottom, l. 14 mm, w. 12 mm (Plate II:18). Type: Zábojník 250

15. Nuštar Žankovac

21. Privlaka Rudina selište

15.1. Main belt mount, open cast in bronze, shield-like, hinge and pendant missing, decorated (motif unclear), l. 24 mm, w. 22 mm (Plate II:15). Type: Zábojník 230

21.1. the Lower part of a main strap belt end, cast in bronze, longish, straight sides, tongue-shaped bottom, central decoration plate with animal fight motifs, l. 38 mm, w. 26 mm (Plate II:21). Type: Zábojník 16

15.2. Side belt strap end, cast in bronze, longish, straight sides, tongue-shaped bottom, undecorated slit for fastening, backside flat, central decoration plate with a tendril, l. 35 mm, w. 14 mm (Plate II:13). Type: Zábojník 91 Closest analogy: Privlaka

22.2. Side belt strap end, cast in bronze, longish, straight sides, tongue-shaped bottom, undecorated slit for fastening, back side flat, central decoration plate on the front side, motif unclear l. 41 mm, w. 12 mm (Plate II:22).

15.3. The Lower part of a side belt strap end, cast in bronze, longish, straight sides, tongue-shaped bottom, backside flat, central decoration plate with a tendril, l. 16 mm, w. 11 mm (Plate II:14). Type: Zábojník 91

23.3. Side belt strap end, cast in bronze, longish, narrowed in the middle, bottom tongue-shaped, slit for fastening decorated with dots, on both sides central decoration plate with tendrils, l. 29 mm, w. 14 mm (Plate III:1). Type: Zábojník 91 Closest analogy: Otok

144

Small Items, Big Results 24.4. The Lower part of a side belt strap end, cast in bronze, straight sides, tongue-shaped bottom, decorated, motif unclear, l. 17 mm, w. 11 mm (Plate III:3).

25. Sotin Ivan Radić vineyard 25.1. The main belt mount, cast in bronze, shield-like, hinge and pendant, central decoration plate with double tendril motif, l. 40 mm, w. 25 mm. Type: Zábojník 246 Closest analogy: Nuštar grave 10

25.5. One half of two pieced side belt strap end, cast in bronze, longish, straight sides, rounded bottom, central decoration plate with tendril motif, l. 27 mm, w. 11 mm (Plate III:2). Type: Zábojník 113 Closest analogy: Nuštar grave 12

27. Tordinci Močila 27.1. Upper part of one half of double main belt strap end, cast in bronze, straight sides, two animal heads opposite one another on the top, central decoration plate with flat tendril motif, l. 36 mm, w. 22 mm (Plate III:11). Type: Zábojník 57 Closest analogy: Nuštar grave 12

22. Privlaka Zakota 22.1. The Lower part of a side strap belt end, cast in bronze, longish, straight sides, rounded bottom, central decoration plate with tendril motif on both sides, l. 14 mm, w. 12 mm (Plate III:4). Type: Zábojník 113

28. Tovarnik Orašje 28.1. Lower half of main strap end, cast in bronze, longish, straight sides, rounded bottom, central decoration plate on both sides with flat tendril motif framed with zig-zag motif, l. 42 mm, w. 20 mm (Plate III:12). Type: Zábojník 29 Closest analogy: Nuštar grave 79

22.2. Main belt mount, cast on bronze, rectangular, rounded top, hinge and pendant missing, central decoration plate with double tendril motif, l. 33 mm, w. 27 mm (Plate III:5). Type: Zábojník 211 23. Retkovci Didove njive 23.1. Main belt mount, open cast in a bronze, gilded, round, motif of a wild beast, Ø 19 mm (Plate III:6). Type: Zábojník 531

29. Vinkovačko Novo Selo Lisičak 29.1. The main belt mount, open cast in bronze, hinge and pendant missing, longish, straight sides, rounded top, central decoration plate with fleur-de-lis motif, l. 26 mm, w. 12 mm (Plate III:13). Type: Zábojník 248 Closest analogy: Nuštar grave 72

23.2. Damaged main belt mount, cast in bronze, only the middle part in the shape of a griffon preserved, l. 30 mm, w. 19 mm (Plate III:7). Type: Zábojník 205/239 23.3. Main belt mount, open cast in bronze, gilded, shield-like, decorated with griffon motif, pendant, l. 35 mm, w. 28 mm (Plate III:8). Type: Zábojník 575

31. Vinkovačko Novo Selo Mokro polje 31.1. A damaged side belt mount, cast in bronze, rhombic, imitation of hollow beaded wire, l. 16 mm, w. 14 mm (Plate III:14). Type: Zábojník 173 Closest analogy: Nuštar grave 37

24. Retkovci Opolenac 24.1. Belt buckle mount, open cast in bronze, frame and hinge missing, tongue-shaped end, straight sides, central decoration plate with tendril motif, l. 37 mm, w. 19 mm (Plate III:9). Type: Zábojník 144

33. Vinkovci Jošine 33.1. One half of double side belt strap end, cast in bronze, longish, straight sides, rounded bottom, central decoration plate with palmette and lily motifs, l. 27 mm, w. 11 mm (Plate III:15). Closest analogy: Nuštar grave 59

26. Sotin Grabovo 26.1. Upper part of a side belt strap end, cast in bronze, straight sides, slit for fastening decorated on both sides with vine motifs, central decoration plate with a vine motif, l. 26 mm, w. 15 mm (Plate III:10). Type: Zábojník 87 Closest analogy: Nuštar grave 34 145

Anita Rapan Papeša 36. Vođinci Matakovica

17. Ostrovo Gorica

36.1. A broken propeller-shaped belt mount, cast in bronze, decoration in form of imitation of hollow beaded wire on the edges, l. 15 mm, w. 11 mm (Plate III:17). Type: Zábojník 157 Closest analogy: Nuštar

17.2. Crescent of an earring, cast in bronze, imitation of filigree wire along edges, star-shaped pendant and 2 additional globular pendants, l. 19 mm, w. 21 mm (Plate IV:5). Type: Čilinska, IIC Closest analogy: Otok grave 17

36.2. A mount pendant, casted, triangular, decorated with grapes, l. 14 mm, w. 13 mm (Plate III:18). Type: Zábojník 258

22. Privlaka Zakota 22.3. Crescent of an earring, cast in silver, decoration in form of imitation of filigree double wire along edges, three globular pendants, l. 18 mm, w. 24 mm (Plate IV:6). Type: Čilinska, IIC Closest analogy: Otok grave 2D

36.3. A-side belt strap end, cast in bronze, longish, straight sides, rounded bottom, undecorated slit for fastening, back side undecorated and flat, central decoration plate with tendril motif, l. 47 mm, w. 15 mm (Plate III:16). Type: Zábojník 536

22.4. Damaged cresent of an earring, cast in silver, decoration in form of imitation of filigree double wire along edges, l. 14 mm, w. 20 mm (Plate IV:7). Type: Čilinska, IIC Closest analogy: Otok grave 2D

Jewellery 8. Ivankovo Dren

31. Vinkovačko Novo Selo Mokro polje

8.3. Half of a crescent from an earring, made of sheet silver, decorated with filigree double wire on the edges, two globular pendants preserved, l. 20 mm, w. 9 mm (Plate IV:1). Type: Čilinska, IIC Closest analogy: Otok grave 2D

31.2. Half of a crescent from an earring, cast in bronze, imitation of beaded filigree wire around edges, five globular pendants preserved, l. 15 mm, w. 15 mm (Plate IV:8). Type: Čilinska, IIC Closest analogy: Privlaka grave 202

12. Markušica Čeretinci

35. Vinkovci 68. Bojne vojne policije street bb

12.3. Crescent of an earring, cast in bronze, imitation of beaded filigree wire around edges, three globular pendants, l. 13 mm, w. 16 mm (Plate IV:2). Type: Čilinska, IIC Closest analogy: Privlaka grave 202

35.1.-2. A pair of earrings, cast in bronze and decorated with filigree wires and pressing, oval hoop continues into a crescent decorated with tendril motif; a starshaped pendant with a globe in the middle, l. 36 mm, w. 21 mm (Plate IV:9-10). Type: Čilinska, IIC Closest analogy: Otok grave 2D

13. Mirkovci Malat 13.3. Crescent of an earring, open cast in silver, decoration in form of imitation of filigree double wire along edges, five globular pendants, l. 16 mm, w. 16 mm (Plate IV:3). Type: Čilinska, IIC Closest analogy: Otok grave 2D

Weapons 6. Ilača 6.1. An iron arrowhead, flat, hooked and triangular shaped, with a long socket, l. 70 mm, w. 25 mm (Plate IV:11). Type: Holešćák AVIb1 Closest analogy: Šarengrad grave 42

13.4. Half of a crescent from an earring, cast in bronze, imitation of beaded filigree wire around edges, four globular pendants preserved, l. 16 mm, w. 15 mm (Plate IV:4). Type: Čilinska, IIC Closest analogy: Privlaka grave 202

146

Small Items, Big Results 20. Podgrađe Grac

30. Vinkovci - Crkvište

20.1. An iron arrowhead, flat, hooked and triangular, with a long socket, l. 90 mm, w. 27 mm (Plate IV:12). Type: Holešćák AVIb1 Closest analogy: Šarengrad grave 42

30.1. An iron arrowhead, flat, rhombic in shape, with socket, l. 78 mm, w. 23 mm (Plate V:4). Type: Holešćák AV1 Closest analogy: Stari Jankovci, Z-Nović

32. Vinkovci

Horse harness

32.1. An iron arrowhead, trilateral, tear-shaped, with thorn, l. 81 mm, w. 20 mm (Plate IV:13). Type: Holešćák BII1 Closest analogy: Šarengrad grave 52

8. Ivankovo Dren

32.2. An iron arrowhead, flat, rhombic in shape, with socket, l. 89 mm, w. 23 mm (Plate IV:14). Type: Holešćák AV1 Closest analogy: Stari Jankovci, Z-Nović

9. Ivankovo Slatine

8.4. Horse harness fitting, glided bronze, rectangular with rounded corners, l. 15 mm, w. 15 mm (Plate V:5).

9.4. Horse harness divider, cast in bronze, rectangular with rounded corners and an oval loop at the bottom, l. 20 mm, w. 14 mm (Plate V:6).

32.3. An iron arrowhead, flat, rhombic in shape, with socket, l. 70 mm, w. 22 mm (Plate V:1). Type: Holešćák AV1 Closest analogy: Stari Jankovci, Z-Nović

26. Sotin Grabovo 26.2. Horse harness fitting, gilded bronze, in the shape of a boar head, l. 24 mm, w. 13 mm (Plate V:7).

32.4. Iron spearhead, narrow leaf-shaped blade with a ridge in the middle, l. 24 mm, w. 27 mm (Plate V:8). Closest analogy: Stari Jankovci grave 88

Pottery 30. Vinkovci-Crkvište

34. Vinkovci Ivana Gorana Kovačića street 38

30.2. A pot made by hand, rounded body, slightly outwards drawn edge, slightly emphasized bottom, decorated with shallow incised lines, dark in colour, h. 100 mm, ø 120 mm.

34.1. An iron arrowhead, flat, hooked and triangular shaped, with a long socket and twisted neck, l. 69 mm, w. 22 mm (Plate V:2). Type: Holešćák AVIb1 Closest analogy: Šarengrad grave 42 9. Ivankovo Slatine 9.4. An iron arrowhead slightly damaged, flat, hooked and triangular shaped, with a long socket, l. 79 mm, w. 20 mm (Plate V:3). Type: Holešćák AVIb1 Closest analogy: Šarengrad grave 42

147

Anita Rapan Papeša

Plate I: 1. Andrijaševci Gornje njive; 2.-3. Andrijaševci Lončarski poloj; 4.-6. Banovci; 7., 9., 10. Gaboš Sokolovac; 8. Bobota, Stara Bobota; 11. Ilok Duboki potok; 12.-13. Ivankovo Dren; 14.-16. Ivankovo Slatine; 17.-19. Karadžićevo Grabina; 20.-21. Markušica Čeretinci; 22. Mirkovci Malat (photos: I. Sokol, Gradski muzej Vinkovci, adapted by A. Rapan Papeša)

148

Small Items, Big Results

Plate II: 1. Ludvinci Barvalo; 2.-3. Mirkovci Malat; 4.-12. Nuštar Rimski most; 13.-15. Nuštar Žankovac; 16.-17. Orolik Krčevine; 18. Podgrađe Donje livade; 19. Ostrovo Gorica; 20. Pačetin Pazarište; 21.-22. Privlaka Rudina selište (photos: I. Sokol, Gradski muzej Vinkovci, adapted by A. Rapan Papeša)

149

Anita Rapan Papeša

Plate III: 1.-3. Privlaka Rudina selište; 4.-5. Privlaka Zakota; 6.-8. Retkovci Didove njive; 9. Retkovci Opolenac; 10. Sotin Grabovo; 11. Tordinci Močila; 12. Tovarnik Orašje; 13. Vinkovačko Novo Selo Lisičak; 14. Vinkovačko Novo Selo Mokro polje; 15. Vinkovci Jošine; 16.-18. Vođinci Matakovica (photos: I. Sokol, Gradski muzej Vinkovci, adapted by A. Rapan Papeša)

150

Small Items, Big Results

Plate IV: 1. Ivankovo Dren; 2. Markušica Čeretinci; 3.-4. Mirkovci Malat; 5. Ostrovo Gorica 6.-7. Privlaka Zakota; 8. Vinkovačko Novo Selo Mokro polje; 9.-10. Vinkovci 68. Bojne vojne policije; 11. Ilača; 12. Podgrađe Grac; 13.-14. Vinkovci (photos: I. Sokol, Gradski muzej Vinkovci, adapted by A. Rapan Papeša)

151

Anita Rapan Papeša

Plate V: 1., 8. Vinkovci; 2. Vinkovci Ivana Gorana Kovačića; 3. Ivankovo Slatine; 4. Vinkovci Crkvište; 5. Ivankovo Dren; 6. Ivankovo Slatine; 7. Sotin Grabovo (photos: I. Sokol, Gradski muzej Vinkovci, adapted by A. Rapan Papeša)

152

Small Items, Big Results Demo, Ž. 2009. Ranosrednjovjekovno groblje bjelobrdske culture: Vukovar-Lijeva Bara (X-XI. Stoljeće)/An Early Medieval Cemetery of the Bijelo Brdo Culture: VukovarLijeva Bara (10th-11th Centuries), Musei Archaeologici Zagrabiensis Catalogi et Monographiae Volume VI/I and II. Zagreb: Arheološki muzej u Zagrebu. Dimitrijević, S. 1957. Četiri groba iz novootkrivene slavenske nekropole u Otoku kod Vinkovaca s posebnim osvrtom na naušnice lunulastozvjezdolikog tipa. Opuscula Archaeologica 2: 21-38. Dimitrijević, D. 1966. Der Fund von Zemun polje im Lichte der spätawarenzeitlichen Problematik, Archaeologia Iugoslavica VII: 53-75. Dimitrijević, S. 1979. Arheološka topografija i izbor nalaza s vinkovačkog tla/Archäologische Topographie und Auswahl archäologischer Funde vom Vinkovcer Boden, in Ž. Rapanić (ed.) Corolla Memoriae Iosepho Brunšmid Dicata, Izdanja Hrvatskog arheološkog društva 4: 133-282. Vinkovci: Hrvatsko arheološko društvo. Dizdar, M. 2010. Arheološka istraživanja u Sotinu zadaci i ciljevi, in M. Hutinec, D. Ložnjak Dizdar (eds) Arheološke spoznaje o Sotinu. Katalog izložbe: 15-19,. Vukovar: Gradski muzej Vukovar. Dizdar, M., A. Rapan Papeša, A. Rimpf 2017. Rezultati zaštitnih istraživanja kasnoavarodobnog groblja Šarengrad-Klopare. Annales Instituti Archaeologici III: 9-18. Dorn, A. 1973. Rekognosciranje limesa na području vukovarske općine. Glasnik slavonskih muzeja 23 (prosinac 1973): 22-30. Dugonjić, A. and A. Rapan Papeša (eds) 2019. Avari i Slaveni /Avars and Slavs. Exhibiton catalogue. Zagreb: Arheološki muzej u Zagrebu. Durman, A. (ed.) 2013. The oldest town in Europe. Vinkovci: Gradski muzej Vinkovci, Privlačica d.o.o. Filipec. K. 2003. Kasnoavarski ukrasni okov (falera) u obliku veprove glave iz Siska. Godišnjak Gradskog muzeja Sisak III-IV (2002-2003): 117-146. Filipec, K. 2017. Ulomci keramičkih posuda iz ranosrednjovjekovnog naselja Stari Jankovci Gatina, in T. Sekelj Ivančan, T. Tkalčec, S. Krznar and J. Belaj (eds) Srednjovjekovna naselja u svjetlu arheoloških izvora/Mediaeval settlements in the light of archaeological sources, Serta Instituti Archaeologici 6: 77-109. Zagreb: Institut za arheologiju. Garam, É. 1993. Katalog der awarenzeitlichen Goldgegenstände und der Fundstücke aus dem Fürstengräbern im Ungarischen Nationalmuseum. (Catalogi Musei Nationalis Hungarici. Seria Archeologica 1). Budapest: Verlag Ungarisches Nationalmuseum. Garam, É. 2001. Funde byzantischer Herkunft in der Awarenzeit vom Ende des 6. Bis zum Ende des 7. Jahrhunderts (Monumenta Avarorum Archaeologica Vol. 5.) Budapest: Magyar Nemzeti Múzeum. Gračanin, H. 2011. Južna Panonija u kasnoj antici i ranom

Bibliography Ašperger, D., A. Rapan Papeša, S. Jorgić, S. Fazinić and M. Trkmić 2019. Destructive and Non-Destructive Methods in the Analysis of Glass Jewellery from the Archaeological Site Nuštar, in D. Stasšikova Štukovská (ed.) Historické Sklo. The Historical Glass. A Multidisciplinary Approach to Historical Glass III: 119125, 389-390. Bratislava: Slovenská archeologická spoločnosť pri SAV. Balen, J. 2006. Vučedol, in A. Durman (ed.) Stotinu hrvatskih arheoloških lokaliteta: 300-301. Zagreb: Leksikografski zavod Miroslav Krleža. Bognar, A. 1994. Na vukovarskoj lesnoj zaravni, in I. Karaman (ed.) Vukovar vjekovni hrvatski grad na Dunavu: 25-48. Varaždin: Nakladnička kuća dr. Feletar Koprivnica Bognár, K. and A. Rapan Papeša 2018. Gelbe Keramik im Stadtmuseum Vinkovci. Hadak útján XXVII: 32-57. Bojčić, Z. 1984. Pregled istraživanja i rasprostranjenosti ranosrednjovjekovnih arheoloških nalaza u istočnoj Slavoniji i Baranji, in N. Majnarić Pandžić (ed.) Arheološka istraživanja u istočnoj Slavoniji i Baranji, Izdanja Hrvatskog arheološkog društva 9: 211-222. Zagreb: Hrvatsko arheološko društvo. Bunčić, M. 2007. Topografija prapovijesnih nalazišta u Vukovaru, Vjesnik Arheološkog muzeja u Zagrebu 3.s. XL: 35-72. Bunčić, M. 2018. Srednjovjekovni nalazi s vučedolskog Gradca, in D. Tončinić (ed.) Arheologija na Dunavu: znanstveni skup, Vukovar, 7.-11. listopada 2013, Izdanja Hrvatskog arheološkog društva 31: 135-157. Zagreb: Hrvatsko arheološko društvo. Bunčić, M. and A. Solter 2020. Ta ova je godina dobra počela samo da nije B. i G.’a. Trgovina arheološkim nalazima početkom 20. stoljeća na primjeru Srijemske Mitrovice / This year would have started well if only it were not for B. and G. The antiquities trade in Sremska Mitrovica at the turn of the 20th century, Vjesnik Arheološkog muzeja u Zagrebu 53: 203227. Burić, M., M. Hulina and J. Lapić 2012. Redni broj 20: Bapska-Gradac, in Z. Wiewegh (ed.) Hrvatski arheološki godišnjak 8/2011: 53-55. Zagreb: Ministarstvo kulture Republike Hrvatske. Burić, M., and M. Podgorelec 2016. Redni broj 37: BapskaGradac, in Z. Wieweigh (ed.) Hrvatski arheološki godišnjak 12/2015: 83-84. Zagreb: Ministarstvo kulture Republike Hrvatske. Csallany, D. 1956. Archäologische Denkmäler der Awarenzeit in Mitteleuropa: Schrifttum und Fundorte. Budapest: Akadémiai Kiadó. Čilinská, Z. 1975. Frauenschmuck aus dem 7.8. Jahrhundert im Karpatenbecken. Slovenská Аrcheológia ХXIII: 63-96. Demo, Ž. 1996. Vukovar Lijeva bara. Katalog izložbe. Zagreb: Arheološki muzej u Zagrebu. 153

Anita Rapan Papeša srednjovjekovlju (od konca 4. do konca 11. stoljeća). Zagreb: Plejada. Csuthy, A., V. Zsombor Győrffy, V., and A. Papp, 2021., Avar-age horse harnes fitting in Shape of Boar head from Balatonalmádi-Felsőhegy, New Parallels and Notes in Connection with Preliminary Researchof the Avar-age Burial Ground, in Z. Robak and M. Ruttkay (eds.), Slovenská archeológia – Supplementum 2, Celts – Germans – Slavs. A Tribute Anthology to Karol Pieta. Nitra: 401– 408. Hampel, J. 1905. Alterthümer des frühen Mittelalters in Ungarn I-III. Braunschweig: F. Vieweg und Sohn. Heinrich Tamaska, O. 2015. Studien zu den awarenzeitlichen Tauschierarbeiten (Monographien zur Frühgeschichte und Mittelalterarchäologie 11). Innsbruck: Romisch-Germanisches Zentralmuseum, Mainz und Institut Fur Ur- und Frühgeschichte Der Universitat Wien. Milošević, A. (ed.) 2000. Hrvati i Karolinzi. Katalog. Split: Muzej hrvatskih arheoloških spomenika. Holeščàk, M. 2015. Early medieval arrowheads from the area of today’s Slovakia, in C. Cosma (ed.)Warriors, weapons, and harness from the 5th-10th centuries in the Carphatian Basin: 299-306. Cluj-Napoca: Mega Publishing House. Hutinec, M. 2009. Redni broj 39: Vukovarsko područje - zapadni dio, in J. Mesić (ed.) Hrvatski arheološki godišnjak 5/2008: 110-115. Zagreb: Ministarstvo kulture Republike Hrvatske. Hutinec, M. 2013. Novi nalaz avarske pojasne garniture s borovskog Gradca, Poster-presentation.Znanstveni skup „Arheologija na Dunavu“, Vukovar 08.-11.10.2013. Ilkić, M. 2008. Nalazi seobe naroda i ranog srednjeg vijeka iz Sotina/Funde der Völkerwanderung und des Frühmittelalters aus Sotin. Prilozi Instituta za arheologiju u Zagrebu 24: 277-288. Ilkić, M. 2010. Arheološka topografija srednjovjekovnog Sotina, in T. Šeparović (ed.) Stjepan Gunjača i hrvatska srednjovjekovna povijesno - arheološka baština: 383-401. Split: Muzej hrvatskih arheoloških spomenika. Ivanković, A. 2016. Pojedinačni i slučajni nalazi avarodobne keramike iz Istočne Slavonije i Srijema u Arheološkom muzeju u Zagrebu. Unpublished Professional paper/ report. Jorgić, S. 2015. Destruktivne i nedestruktivne metode u analizi nakita od stakla iz arheoloških nalazišta. Unpublished MA dissertation, University of Zagreb. Kalafatić, H. and B. Šiljeg 2018. Krugovi parnjaci: novi uvidi u neolitičke obrasce naseljavanja / Twin circles: new insights in the Neolithic settlement pattern, Prilozi Instituta za arheologiju u Zagrebu 35: 71-111. Karamanski, S. 1976. Još nekoliko nalaza iz perioda seobe naroda koji se nalaze u muzejskoj zbirci u Odžacima. Bilten 6-7: 1-9. Korda, J. 1960. Tragom prošlosti Vinkovaca. Vinkovci: Vinkovci Municipal Museum.

Krznarić Škrivanko, M., A. Rapan Papeša and H. Vulić 2013. Redni broj 26: Vinkovačko područje,in Z. Wiewegh (ed.) Hrvatski arheološki godišnjak 9/2012: 55-60. Zagreb: Ministarstvo kulture Republike Hrvatske. Ložnjak Dizdar, D., M. Dizdar and G. Kušić 2018. SotinSrednje polje-Istraživanje višeslojnog nalazišta u Podunavlju 2017. /Sotin-Srednje polje-Research of multilayer site in the Danube Region in 2017. Annales Instituti Archaeologici XIV: 28-33. Ložnjak Dizdar, D. and G. Kušić 2019. Sotin - istraživanja višeslojnog nalazišta 2018. na položaju Srednje polje, Knjiga sažetaka 4. Okruglog stola „Recentna arheološka istraživanja u Vukovarsko-srijemskoj županiji“, Vinkovci, 20. ožujka 2019 (book of abstracts). Ljubić, Š. 1888. Prvo otkriće predhistoričkog selišta iz kamene dobe na našoj zemlji. Vjesnik hrvatskog arheološkog društva 10: 1-2. Mrkobrad, D. 1980. Arheološki nalazi seobe naroda u Jugoslaviji. Beograd: Savez areholoških društava Jugoslavije: Muzej grada Beograda. Njegač, D. 2002. Istočna Hrvatska, in I. Borovac (ed.) Veliki atlas Hrvatske: 254-270. Zagreb: Mozaik knjiga. Orlando, L., Fages, A., K. Hanghøj, N. Khan, C. Gaunitz, A. Seguin-Orlando, M. Leonardi, C. Mccrory Constantz, C. Gamba, K. A.S. Al-Rasheid, S. Albizuri, A. H. Alfarhan, M. Allentoft, S. Alquraishi, D. Anthony, N. Baimukhanov, J. H. Barrett, J. Bayarsaikhan, N. Benecke, E. Bernáldez-Sánchez, L. Berrocal-Rangel, F. Biglari, S. Boessenkool, B. Boldgiv, G. Brem, D. Brown, J. Burger, E. Crubézy, L. Daugnora, H. Davoudi, P. de Barros Damgaard, M. de los Ángeles de Chorro y de Villa-Ceballos, S. Deschler-Erb, C. Detry, N. Dill, M. do Mar Oom, A. Dohr, S. Ellingvåg, D. Erdenebaatar, H. Fathi, S. Felkel, C. FernándezRodríguez, E. García-Viñas, M. Germonpré, J. D. Granado, J. H. Hallsson, H. Hemmer, M. Hofreiter, A. Kasparov, M. Khasanov, R. Khazaeli, P. Kosintsev, K. Kristiansen, T. Kubatbek, L. Kuderna, P. Kuznetsov, H. Laleh, J. A. Leonard, J. Lhuillier, C. L. von Lettow-Vorbeck, A. Logvin, L. Lõugas, A. Ludwig, C. Luis, A. Margarida Arruda, T. Marques-Bonet, R. Matoso Silva, V. Merz, E. Mijiddorj, B. K. Miller, O. Monchalov, F. A. Mohaseb, A. Morales, A. NietoEspinet, H. Nistelberger, V. Onar, A. H. Pálsdóttir, V. Pitulko, K. Pitskhelauri, M. Pruvost, P. Rajic Šikanjic, A. Rapan Papeša, N. Roslyakova, A. Sardari, E. Sauer, R. Schafberg, A. Scheu, J. Schibler, A. Schlumbaum, N. Serrand, A. Serres-Armero, B. Shapiro, S. Sheikhi Seno, I. Shevnina, S. Shidrang, J. Southon, B. Star, N. Sykes, K. Taheri, W. Taylor, W. R. Teegen, T. Trbojević Vukičević, S. Trixl, D. Tumen, S. Undrakhbold, E. Usmanova, A. Vahdati, S. Valenzuela-Lamas, C. Viegas, B. Wallner, J. Weinstock, V. Zaibert, B. Clavel, S. Lepetz, M. Mashkour, A. Helgason, K. Stefánsson, E. Barrey, E. Willerslev, A. K. Outram and P. Librado 154

Small Items, Big Results 2019, Tracking five millennia of horse management with extensive ancient genome time series. Cell 177: 1419-1435. Ostović, N. 2021. Rezultati zaštitnih arheoloških istraživanja i drugih aktivnosti Arheološkog odjela Gradskog muzeja Vinkovci u 2020. godini. Godišnjak ogranka Matice hrvatske Vinkovci 38: 165-182. Petković, D. 2006. Srednjovjekovna naselja sjeverozapadnog dijela vinkovačkog kraja (Acta Musei Cibalensis 3). Vinkovci: Gradski muzej Vinkovci. Petrinec, M. 2009. Groblja od 8. do 11. stoljeća na području ranosrednjovjekovne hrvatske države. Split: Muzej hrvatskih arheoloških spomenika. Pinterović, D. 1968. Limesstudien in der Baranja und in Slawonien, Archaeologia Iugoslavica IX: 55-82. Pokos, N. and I. Turk 2012. Geografska obilježja Vukovarsko-srijemske županije, in D. Živić (ed.) Vukovarsko-srijemska županija - Prostor, ljudi, identitet: 11-26. Zagreb-Vukovar: Institut društvenih znanosti Ivo Pilar. Premužić, Z., P. Rajić Šikanjić and A. Rapan Papeša 2016. A case of Avar period trepanation from Croatia. Anthropological Review 79/4: 471-482. Premužić, Z., T. Trbojević Vukičević, P. Rajić Šikanjić and A. Rapan Papeša 2016. Životinjske kosti kao prilozi na avarodobnom groblju iz Nuštra, in S. Krznar, T. Sekelj Ivančan, T. Tkalčec and J. Belaj (eds.), Groblja i pogrebni običaji u srednjem i novom vijeku na prostoru sjeverne Hrvatske. 37-48. Zagreb: Institut za arheologiju. Premužić, Z., P. Rajić Šikanjić and A. Rapan Papeša 2018. Bioarheološka analiza avarodobnog groblja u Nuštru, in D. Tončinić (ed.) Arheologija na Dunavu, Izdanja Hrvatskog arheološkog društva 31: 117-122. Zagreb: Hrvatsko arheološko društvo. Rapan Papeša, A. 2007. Nova građa u srednjovjekovnoj zbirci Arheološkog odjela Gradskog muzeja Vinkovci, Godišnjak ogranka Matice Hrvatske Vinkovci 25: 263-272. Rapan Papeša, A. 2008. Stari Jankovci - Gatina - naselje i groblje avaro-slavenskog razdoblja, in I. Karaula (ed.) Starojankovčani kroz stoljeća 1301.-2007.: 831-840. Vinkovci-Slakovci: Naklada Karaula. Rapan Papeša, A. 2012a. Prvi nalazi s avarskog groblja na položaju Nuštar / The first findings from the avar cemetry in Nuštar/ Die ersten Funde des Awarenfriedhofs an der Fundstelle Nuštar. Katalog izložbe.Vinkovci: Gradski muzej Vinkovci. Rapan Papeša, A. 2012b. Redni broj: 28. Lokalitet: Nuštar-avarsko groblje, in Z. Wiewegh (ed.) Hrvatski arheološki godišnjak 8/11: 65-68. Zagreb: Ministarstvo kulture Republike Hrvatske. Rapan Papeša, A. 2014. Keramički i olovni pršljenci za vretena iz avarodobnih grobova na lokalitetu Nuštar-Dvorac/Clay and lead spindle whorls from Avar period graves at Nuštar- Dvorac site, Prilozi Instituta za arheologiju u Zagrebu 31: 159-180.

Rapan Papeša, A. 2016. Prethodna razmišljanja o prilozima keramičkih lonaca i njihovom sadržaju u avarodobnim grobovima na lokalitetu NuštarDvorac, in S. Krznar, T. Sekelj Ivančan, T. Tkalčec, J. Belaj (eds) Groblja i Pogrebni Običaji u Srednjem i Ranom Novom Vijeku na Prostoru Sjeverne Hrvatske. Zbornik radova prvog međunarodnog znanstvenog skupa srednjovjekovne arheologije Insituta za arheologiju Zagreb, 4. Lipnja 2014. Serta Instituti Archaeologici 4: 2335. Zagreb: Institut za arheologiju. Rapan Papeša, A. 2018. Predmeti od kosti, roga i kamena iz avarodobnih grobova na lokalitetu Nuštar/ Dvorac,in D. Tončinić (ed.) Arheologija na Dunavu: znanstveni skup, Vukovar, 7.-11. listopada 2013, Izdanja Hrvatskog arheološkog društva 31: 107-115. Zagreb: Hrvatsko arheološko društvo. Rapan Papeša, A. 2020a. Magic or practicality? „Antiquities“ in; Avar period graves in Eastern Croatia, in S. Krznar (ed.) Life and death in mediaeval and early modern times. Proceedings of the 5th International Scientific Conference of Mediaeval Archaeology of the Institute of Archaeology. Serta Instituti Archaeologici 14: 17-31. Zagreb: Institut za arheologiju. Rapan Papeša, A. 2020b. Seoba naroda i Rani srednji vijek, in: H. Vulić (ed.) Arheološki odjel, Vodič kroz postav: 102-117. Vinkovci: Gradski muzej Vinkovci. Rapan Papeša, A. 2021a. Groblje na groblju, Moji Vinkovci 31: 24-26. Rapan Papeša, A. 2021b. Avarodobni arheološki krajobraz međuriječja Bosuta, Vuke i Dunava. unpublished PhD dissertation, University of Zagreb. Rapan Papeša, A. 2021c. The Lion does not care about the Opinion of the Sheep, in L. Keve, D. BăcuețCrișan, I.Stanciu and F. Mărginean (eds) Exploring Dwellings and Manufacturing Spaces in Medieval Context (7th – 14th centuries): 123-135. Cluj - Napoca: Editura mega. Rapan Papeša, A., Á. Kenéz and Á. Pető 2015. Arheobotanička analiza uzoraka iz kasnoavarodobnih grobova iz Nuštra (istočna Hrvatska). Prilozi Instituta za arheologiju u Zagrebu 32: 261-288. Rapan Papeša, A. and K. Grömer 2015. Jednostavne tkanine i žigosana koža: organski nalazi s avarodobnog groblja u Nuštru (istočna Hrvatska) / Simple cloth and stamped leather: organic finds from the Avar graveyard in Nuštar (Eastern Croatia). Vjesnik Arheološkog muzeja u Zagrebu 48: 51-83. Rapan Papeša, A. and P. Šmalcelj Novaković 2016. Kasnoavarodobno groblje na Gradini u Otoku. Katalog izložbe.Vinkovci: Gradski muzej Vinkovci. Rapan Papeša, A. and D. Petković 2017. How Archaeology Helps History: the Case of Estate Zablaće in the Vicinity of Vinkovci, in T. Sekelj-Ivančan, T. Tkalčec, S. Krznar and J. Belaj (eds) Srednjovjekovna naselja u svjetlu arheoloških izvora/Mediaeval settlements in the light of archaeological sources. Serta Instituti Archaeologici 6: 499-508. Zagreb: Institut za arheologiju. 155

Anita Rapan Papeša Rapan Papeša, A. and K. B. Bognár 2018. Gelbe Keramik im Stadtmuseum Vinkovci, in P. Forisek, A. Szabó, J. Szakács (eds.) „Hadak útján” a népvándorláskor fiatal kutatóinak xxvii. Konferenciája: 32-57. Debrecen: Debreceni Egyetem. Rapan Papeša, A., A. Rimpf and M. Dizdar 2018. Rezultati zaštitnih istraživanja kasnoavarodobnog groblja Šarengrad-Klopare 2017. godine. Annales Instituti Archaeologici XIV: 13-19. Rimpf, A., A. Rapan Papeša and M. Dizdar 2020. Rezultati zaštitnih istraživanja kasnoavarodobnog groblja Šarengrad-Klopare 2019. godine. Annales Instituti Archaeologici XVI: 8-17. Roglić, V. 1975. Istočna Hrvatska, in A. Cvitanović (ed.) Geografija SR Hrvatske, knjiga 3: p. 175. Zagreb: Školska knjiga. Sekelj Ivančan, T. 1995. Catalogue of Medieval Sites in Continental Croatia. Oxford: Tempus Reparatum and Hadrian Books. Dimitrijević, D., J. Kovačević and Z. Vinski (eds) 1962. Seoba Naroda-arheološki nalazi jugoslovenskog podunavlja. Katalog izložbe. Zemun: Narodni muzej Zemun. Simoni, K. 1986. Neobjavljeni okovi i jezičci nakitnog stila Blatnica iz Arheološkog muzeja u Zagrebu. Vjesnik Arheološkog muzeja u Zagrebu 19: 217-228. Szentpéteri, J. (ed.) 2002. Archäologische Denkmäler der Awarenzeit in Mitteleuropa (Varia archaeologica Hungarica 13). Budapest: Archäologisches Institut der Ungarischen Akademie der Wissenschafte. Šulc, B. (ed.) 2009. Slavonija, Baranja i Srijem, vrela europske civilizacije. Katalog izložbe. Zagreb: Ministarstvo kulture Republike Hrvatske. Šmalcelj, M. 1973. Privlaka - „Gole njive“ (općina Vinkovci) - Nekropola VII-IX stoljeća - sistematska iskopavanja. Arheološki Pregled 15: 117-119. Šmalcelj, M. 1976. Privlaka - „Gole njive“, Vinkovci Nekropola VIII-IX st. Arheološki Pregled 18: 127-128. Šmalcelj, M. 1981a. Privlaka-Gole njive (opć. Vinkovci)avaroslavenska nekropola. Arheološki Pregled 22: 143-144. Šmalcelj, M. 1981b. Stari Jankovci-Gatina (općina Vinkovci)-avaroslavenska nekropola. Arheološki Pregled 22: 142-143. Šmalcelj, M. 1992. Gole njive, Stari Jankovci-Gatina, in V. Girardi-Jurkić (ed.) Arheologija i rat/ Archeology and War: 44-49. Zagreb: Ministarstvo kulture i obrazovanja Republike Hrvatske.

Šmalcelj, P. 2013. Ranosrednjovjekovno groblje Otok kod Vinkovaca. Unpublished MA thesis, University of Zagreb. Šmalcelj Novaković, P. 2020. What are they doing here?, in S. Krznar (ed.) Life And Death in Mediaeval and Early Modern Times. Proceedings of the 5th International Scientific Conference of Mediaeval Archaeology of the Institute of Archaeology Zagreb, 6th and 7th June 2018, Serta Instituti Archaeologici 14: 32-46. Zagreb: Institut za arheologiju. Tomičić, Ž. 2000. Arheološka slika ranoga srednjeg vijeka na prostoru međurječja Drave, Dunava i Save, in A. Milošević (ed.) Hrvati i Karolinzi. Rasprave i vrela: 142161. Split: Muzej hrvatskih arheoloških spomenika. Trbuhović, L. 1982. Avar finds from Sirmium and the surrounding region. Sirmium IV: 61-74. Uglešić, A. 1994. Nalazi seobe naroda iz zbirke Mate Ilkića / The finds from the Great Migration period in Mato Ilkić’s collection. Radovi 20: 145-152. Vinski, Z. 1954a. Gibt es frühslawische Keramik aus der Zeit der südslawischen Landnahme? Archaeologia Iugoslavica I: 71-82. Vinski, Z. 1954b. Povodom našeg prvog priručnika slavenske arheologije s osvrtom na još neke novije publikacije s tog područja. Peristil I: 199-207. Vinski, Z. 1959. Ausgrabungen in Vukovar. Archaeologia Iugoslavica III: 99-109. Vinski-Gasparini, K. and S. Ercegović 1958. Ranosrednjovjekovno groblje u Brodskom Drenovcu/ Das frühmittelalterliche Gräberfeld von Brodski Drenovac. Vjesnik Arheološkog muzeja u Zagrebu 3.s. 1: 129-161. Trbojević Vukičević, T., A. Rapan Papeša, I. Alić, A. Ekert Kabalin, M. Ostović and S. Kužir 2017. Contribution to understanding Avar burials with equids in Croatia: detailed archaeozoological analysis. Revue de médecine vétérinaire 168 4/6: 73-80. Vulić, H. 2015. Redni broj: 45, Lokalitet: Vinkovačko područje, in Z. Wiewegh (ed.) Hrvatski arheološki godišnjak 12/2015: 98-99. Zagreb: Ministarstvo kulture Republike Hrvatske. Zábojník, J. 1991. Seriation von Gürtelbeschlaggarnituren aus dem Gebiet der Slowakei und Österreichs (Beitrag zur Chronologie der Zeit Des Awarischen Kaganats), in Z. Čilinská (ed.) K problematike osídlenia srednodunajskej oblasti vo včasnom stredoveku: 219‐321. Nitra: Archeologický ústav Slovenskej akadémie vied.

156