Roman Pottery and Glass Manufactures: Production and Trade in the Adriatic Region and Beyond Proceedings of the 4th International Archaeological Colloquium (Crikvenica, 8-9 November 2017) 1803273682, 9781803273686

Citation preview

Archaeopress Roman Archaeology 94

Roman Pottery and Glass Manufactures Production and trade in the Adriatic region and beyond

Proceedings of the 4th International Archaeological Colloquium (Crikvenica, 8-9 November 2017) edited by

Goranka Lipovac Vrkljan, Ana Konestra and Anamarija Eterović Borzić

ROMAN ECONOMY DALMATIA

IN

HRZZ-eng.jpg (JPEG Image, 800 × 377 pixels)

http://www3.hrzz.hr/UserDocsImages/logo/HRZZ-eng.jpg

Roman Pottery and Glass Manufactures Production and trade in the Adriatic region and beyond Proceedings of the 4th International Archaeological Colloquium (Crikvenica, 8-9 November 2017)

edited by

Goranka Lipovac Vrkljan, Ana Konestra, Anamarija Eterović Borzić

Archaeopress Roman Archaeology 94

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

ISBN 978-1-80327-368-6 ISBN 978-1-80327-369-3 (e-Pdf) © Archaeopress and the individula authors 2022

Cover illustration: Dora Vukičević Peer reviewers of the Proceedings of the 4th International Archaeological Colloquium (Crikvenica, 8-9 November 2017): Dott. ssa Stefania Mazzocchin University of Padova | UNIPD · Department of Cultural Heritage: Archaeology and History of Art, Cinema and Music DBC, Padova, Italy Associate professor Željko Miletić University of Zadar Department of Archaeology, Zadar, Croatia

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 Preface

1

1. Invited contribution

2

Inside fashion fusion: Fibulae from Crikvenica����������������������������������������������������������������������������������������������� 3 Martina Blečić Kavur 2. The role of landscape features, waterways and seaborne routes in pottery production and distribution 20 Distribution possibilities and transport infrastructure in northern Dalmatia’s pottery workshops: evidence and interpretations�������������������������������������������������������������������������������������������������������������������������21 Ana Konestra, Goranka Lipovac Vrkljan, Fabian Welc Distribution possibilities and transport infrastructure in northern Dalmatia’s pottery workshops���������23 The Amber route between Caput Adriae and Emona basin: the ceramic evidence on inner road - and water - communications ��������������������������������������������������������������������������������������������������������������������������������36 Iris Bekljanov Zidanšek, Petra Vojaković, Tina Žerjal Merci e sedi di mercato: lana e produzione ceramica a Mutina e ai Campi Macri (Modena, I)���������������������47 Carla Corti Pottery and brick production in northeastern Italy in the Roman Period: the exploitation of clay quarries�������������������������������������������������������������������������������������������������������������������56 Cristina Mondin Recent excavation of an amphorae kiln in the artisanal workshop complex at Loron (Tar-Vabriga, Croatia): results of the archaeological and anthracological study��������������������������������������������������������������65 Corinne Rousse, Christophe Vaschalde, Gaetano Benčić, Davor Munda Amphorae and tiles from the island of Pag: indirect evidence of the exploitation of local resources in the early Roman imperial period�������������������������������������������������������������������������������������������������������������������������77 Maja Grisonic, Nikolina Stepan An underwater archaeological survey of Havišće and Lokvišća Coves and Cape Ertak ������������������������������94 Igor Borzić, Mate Parica, Mato Ilkić, Tea Rosić The Roman shipwreck of Sv. Petar, Ilovik: old research, new possibilities������������������������������������������������106 Irena Radić Rossi, Bridget Buxton 3. Ceramic building materials production: production site location, production organisation, distribution channels, regional specificities 116 An overview of the CBM industries of ancient Viminacium������������������������������������������������������������������������117 Ljubomir Jevtović, Ilija Danković Brick by brick, tile by tile. The ceramic building material in Labraunda, Turkey��������������������������������������125 Alexandra Dolea Examples of ceramic building materials of Roman and late antique date from Cickini near Malinska on the island of Krk (NE Adriatic, Croatia)��������������������������������������������������������������������������������������������������������140 Ranko Starac Stamps on ceramic and glass finds from the Velika Mrdakovica site near Šibenik������������������������������������149 Toni Brajković

i

4. The study of material evidence of pottery, ceramics and glass production and distribution

172

HILARVS, from the Po valley to the Rhone valley?��������������������������������������������������������������������������������������� 173 Armand Desbat, Valentina Mantovani, Eleni Schindler Kaudelka, Roby Stuani Terre sigillate galliche. La collezione “storica” del Museo Archeologico Nazionale di Aquileia (Friuli Venezia Giulia – Italia)����������������������������������������������������������������������������������������������������������������������������������� 191 Patrizia Donat La ceramica metallescente di Treviri: nuovi dati da Verona (Italia)����������������������������������������������������������� 203 Valentina Mantovani Le anfore galliche nella Regio X: dati preliminari���������������������������������������������������������������������������������������� 212 Andrea Cipolato, Giuseppe Indino Terra sigillata norditalica decorata a matrice intenzionalmente nera: alcune considerazioni���������������� 222 Valentina Mantovani, Eleni Schindler Kaudelka, Roby Stuani, Émilie Mannocci, Carla Corti Canale Anfora, recupero 1988: contributo alla conoscenza della ceramica fine orientale ad Aquileia����� 239 Paola Maggi, Paola Ventura One man’s trash is another man’s treasure. Household waste deposit from Porta Pisana and trends of fine ware consumption within the settlement of Krk (in the last centuries BCE) ��������������������������������� 254 Marina Ugarković, Ranko Starac Pots for storage of food from Crikvenica������������������������������������������������������������������������������������������������������ 285 Ivana Ožanić Roguljić Roman pottery from Siculi – a short overview��������������������������������������������������������������������������������������������� 289 Ivanka Kamenjarin First results of the excavations of the Roman Villa of Bunje (Brač Island, Croatia)����������������������������������� 297 Emmanuel Botte, Kristina Jelinčić Vučković, Ana Konestra, Ivana Ožanić Roguljić Production and distribution of late antique glass and fine wares in Romagna (2nd-7th cent. CE)����������� 304 Tania Chinni, Enrico Cirelli The imports of high quality early Imperial glass in Romula (Pannonia)���������������������������������������������������� 321 Irena Lazar Glass beads from the archaeological site of Burnum (Amfiteatar, Campus)����������������������������������������������� 331 Ivana Jadrić-Kučan Glass askos from the ancient necropolis of Zadar���������������������������������������������������������������������������������������� 340 Timka Alihodžić Glass finds from Tomb 7 from the Late Iron Age necropolis of the Kopila hillfort on the island of Korčula�������������������������������������������������������������������������������������������������������������������������������������������� 344 Anamarija Eterović Borzić, Igor Borzić Roman glass from Lokvišća - Jadranovo site������������������������������������������������������������������������������������������������ 354 Bartul Šiljeg, Kristina Turkalj Experience in the conservation and restoration of archaeological glass at the Archaeological Museum of Istria in Pula��������������������������������������������������������������������������������������������������������������������������������������������������� 362 Monika Petrović Experimental archaeology in the Museum of Ancient Glass in Zadar��������������������������������������������������������� 367 Marko Štefanac, Berislav Štefanac 373

List of Contributors

ii

Preface Since 2004, when the first campaign of systematic excavations at the Roman pottery workshop located in modern-day Crikvenica (NE Adriatic, Croatia) took place, and after the recognition of almost 100 types of commonwares, several shapes of ceramic building materials, but especially the 12 types of specific amphorae produced within it, the search for analogies and comparanda sprung a fertile dialogue with colleagues and pottery experts, in particular those working on the Western Adriatic and in Istria. That is how the idea to establish a regular series of colloquia to be held at Crikvenica to discuss issues of pottery workshops and production, kilns, amphorae, glass and the many related issues within an Adriatic context was born. These were, also, accompanied by the relative Proceedings, allowing a more lasting impact of the topics and issues discussed.1

presented at the 2nd colloquy at Crikvenica in 2011, and here updated), so to enriching and round up contextual and chronological data about the site. In the slightly more than a decade marked by the excavations at Crikvenica and the accompanying colloquia, knowledge on pottery and ceramics production in late Hellenistic and Roman Dalmatia made a huge leap forward thanks to the discovery of several kiln sites, workshops and a more widespread application of archaeometry, while scholarly interest not only in production, but also in the study of pottery, ceramic and glass finds has increased to until recently unimaginable levels. Several recent conferences dealt with different aspects of pottery studies, journal articles devoted to related topics are regularly published, just as are MA and PhD theses. Hopefully, regional research on these fundamental materials which, apart from providing basic chronological information can be interpreted within traditional or novel theoretical frameworks, allowing us to bring to life the everyday and mundane of the past, will continue with the same rhythm. Our own work on workshops, pottery and glass was aided by numerous colleagues, and without them the colloquia of Crikvenica would have certainly been less successful, thus on this occasion we would like to thank them all, especially Eleni Schindler Kaudelka, Susanne Zabehlicky-Scheffenegger, Stefania Pesavento Mattioli, Tamás Bezeczky (†) and Boško Lugović (†), along with a large group of colleagues working in Friuli Venezia Giulia, Veneto and Emilia Romagna, both at Universities, Museum and as independent researchers. Without the aid of Crikvenica Town Museum and Crikvenica Municipality, our work there, just as these colloquia, would probably be impossible, thus we thank them for their constant support. Organization of the IV Colloquium in Crikvenica and editing of these Proceedings were carried out within the framework of project RED - Roman Economy in Dalmatia: production, distribution and demand in the light of pottery workshops (IP -11-2013- 3973) of the Croatian Science Foundation, which was implemented at the Institute of Archeology from 2013 to 2018.

The last of these colloquia, held in 2017 and organised by the Institute of Archaeology in Zagreb and the Crikvenica Town Museum, was dedicated to the usual topics, but with an aim to span a larger area, hence the “and beyond” in the title. This was aided by the joining of Gent University, the Museum of Ancient Glass in Zadar and the University of Zadar as co-organisers. In this Proceedings volume, talks and posters presented on the occasion have been collected, spanning diverse topics related to the production and commercialisation of Roman pottery and ceramics, and glass vessels. The contributions have been arranged in three sections dealing respectively with landscape and communication features related to production and distribution of pottery and ceramics, the specificities of production and distribution of ceramic building materials and, lastly, a section dedicated to various issues of Roman pottery, amphorae and glass. We are pleased that several papers deal with the immediate environs of Crikvenica, providing us, among others, with new data on its products’ distribution and a regional framework, but especially that several continental sites and more distant provinces are represented in the volume as well. Similarly, a fair amount of papers is dedicated to site assemblages and regional overviews of glass finds, and to aspects of its conservation and experimental approaches. As with the 2017 colloquium the series of scientific gatherings at Crikvenica was brought to an end, and therefore this being the last of the Proceedings, we have invited Martina Blečić Kavur to present a contribution which approaches the fibulae finds from the workshop in Crikvenica (originally

Finally, we would like to thank the authors of the here collected contributions for bearing with us to see their work published, unfortunately this time occurring a little later than the usual three years time-frame established for the previous volumes.

The Editors

Proceedings of the previous colloquia were published by the Institute of archaeology and the Crikvenica Town museum in 2011, 2014 and 2017. 1 

1

1.

Invited contribution

Inside fashion fusion: Fibulae from Crikvenica Martina Blečić Kavur Abstract During the archaeological excavations of the figlina at the Igralište site in Crikvenica, among other things, seven very wellpreserved bronze fibulae were discovered, and are today accompanied by an older find of a fibula from the Kaštel site. According to their formal features, they belong to two groups — fibulae of the Middle and Late La Tène schemes are significant forms of the Late La Tène cultural traditions, while the cast fibulae of Aucissa type, strongly profiled fibulae and those with the multiply segmented bow are characteristic elements of Roman provincial culture. Their morphological and stylistic features, as well as the adequate context of their discovery, allow us a precise typological classification, chronological determination and cultural interpretation that is the basis of this discussion. Considered within the material culture of Kvarner and the wider northern Adriatic region, the second half of the 1st century BCE and the course of the 1st century AD, they represent valuable first finds in the archaeological record of the area, and some of them are presented here for the first time. Their presence is associated with cosmopolitan culture and various identities related with the early establishment and greatest flourishing of this significant and first explored ceramic workshop in the province of Dalmatia belonging to Sextus M(e/u)tillius Maximus in Ad Turres.

Key words Crikvenica – Igralište, Ad Turres, ceramic workshop of Sextus M(e/u)tillius Maximus, fibulae, gender, identity.

Introduction The Kvarner basin, stretching from Podvelebit to the eastern coast of Istria, and from the Kastav region to the entrance to Kvarnerić, is not a geographically homogeneous region.1 On the contrary, a highly heterogeneous area that dissipates at its ends and merges with other, neighbouring ecological and cultural milieus (Figure 1). Nevertheless, this territory acted and reacted coherently and cohesively, that is, as a semantically meaningful entity in all its natural and anthropogenic manifestations, especially during the Late Bronze and Iron Age. However, in the last centuries BCE fundamental changes took place which changed irreversibly the picture of the seemingly established and ‘idyllic’ world of the Iron Age – the nearby Histrian wars, along the western extension of Kvarner, the conquest of Nesactium in 177 BCE and, at least formally, the end of long prehistory of the Istrian peninsula inaugurated the Roman Republic in that region (Starac 2000; Matijašić 2017). These key historical events were, due to their physical closeness, directly reflected in the socio-political circumstances on the entire Kvarner Basin. Consequently, we have to consider already the period of the 2nd century BCE as the century of great changes and the beginning of ‘Romanization’ on the wider stretch of North Adriatic. The term ‘Romanization’ is certainly not presented here in an administrative or binary sense, but in the broadest The article is a supplemented and expanded paper presented at II. International Congress of the Roman Ceramic and Glass Workshop in Crikvenica in 2011 (Blečić Kavur 2011). 1 

perspective indicates dynamic processes of intertwined economic and, of course, military-political intentions on that territory (Miškec 2003; Horvat 2009; Matijašić 2009; 2017; cf. Versluys 2014). Unlike the Histrian Nesactium, the ‘cities’ of Kvarner were not militantly conquered, destroyed and demolished. Since the beginning of the 2nd century BCE, material culture actually supports the thesis of their political orientation and strategic involvement with the acquired status of aleates or foederates — as demonstrated by the archaeological remains discovered not only in Osor, but also in the to the north located Krk, Baška and Rijeka (Tarsatica) (Lipovac 1991: 37-46; Šašel 1992: 615; Starac 2000: 16, 22, 81; Blečić 2001: 74; Višnjić 2009a; Blečić Kavur 2015: 231-233). During the 1st century BCE, especially from its middle, Kvarner, like most of Liburnia, was peacefully integrated into the Empire (Starac 2000: 10-18; Konestra et al. 2020: 88-89). And the organization of a military alliance with local forces and the very arrangement of Caesar’s military troops is also indicated by the celebrated battle of Krk in the year 49 BCE (Bilić Dujmušić 2014; cf. Vitelli Casella 2018). This was a period of economic development and prosperity, and it is thought that many private entrepreneurs preceded the establishment of military offices and constituted production, together with the organization of their estates — most likely as early as in the middle of the 1st century BCE. In this context, the figlina explored at the site Igralište in Crikvenica – Ad Turres, established saltus as a suburban territory

roman pottery and glass manufactures (Archaeopress 2022): 3–19

Martina Blečić Kavur

Figure 1. The position of the Kvarner Bay with the most significant sites mentioned in the text (made by M. Blečić Kavur).

between the municipi of Tarsatica and Senia (Lipovac Vrkljan and Konestra 2018; Konestra et al. 2020: 89).

well-known archaeological context. Namely, all fibulae originating from the Kvarner area are mostly of older date and almost as a rule without closer circumstances of their discovery. Crikvenica fibulae show such an interesting spectrum both in a typological-stylistic and in a chronological sense, although not in a quantitative one (Figure 2). In total, only seven bronze fibulae were discovered, of which three belong to fibulae of Middle and Late La Tène schemes, and four to Roman provincial fibulae. Nevertheless, even this seemingly modest sample will enable us to know and understand them

Many years of research at this site have provided exceptional results about the first systematically investigated Roman pottery workshop in the province, as evidenced by the already extensive literature as well as this volume.2 Exceptionally are the discoveries of bronze fibulae, which are available for the first time in a E.g. Lipovac Vrkljan 2011; Lipovac Vrkljan and Konestra 2018; Konestra et al. 2020 – with earlier literature. 2 

4

Inside fashion fusion: Fibulae from Crikvenica

Figure 2. Site layout with of the Crikvenica–Igralište ceramic workshop complex with more important structures and marked positions of discovered fibulae (Archive of the Institute of Archaeology, elaboration: A. Konestra, M. Blečić Kavur).

better in that coastal area, to set certain interpretations and evaluate them in the wider historical environment. In the discussion, all fibulae will be presented in a typological-chronological order, their basic characteristics and circumstances of discovery will be analysed for the purpose of more precise chronological dating and possible clarification of the cultural identity of their users.

They are quantitatively and qualitatively distributed in the wider area of Caput Adriae in a number of local variants. This applies in particular to, on one sided, the area of Friuli and Veneto, and on the other to the area of Istria and Kvarner (Figure 4), even though their topographic distribution is much wider (cf. Spânu 2020: fig. 13). In the Kvarner basin, a large collection of fibulae originating from the old excavations of the Kavanela in Osor stands out, while on other sites they are represented mostly in small numbers or as individual finds. From the multitude of fragments of this type from Osor (Figure 3: 8-9), it is also possible to distinguish those belonging to the Picugi and Aquileia variants, while two almost preserved fibulae from nearby Bakar can be undoubtedly attributed to the Aquileia variant (Figure 3: 6-7). However, they also differ from each other – some are extremely massive and have a profiled button on the foot, i.e. strongly profiled joints, while some are smaller and have a dissected plate bow as well as their joints on the bow, such as the fibula from Crikvenica (Figure 3: 2). Certainly, this astonishing amount of over 20 fibulae, as well as the distinctiveness between them, elevates Osor above other sites in the region where such items were discovered. In this way, it may suggest a possible production and/or at least finishing in this important and significant northern Adriatic ‘city’ (Blečić Kavur 2015: 217-219, fig. 77-78). The most closely related examples to Crikvenica fibulae, in addition to those from Osor and Bakar, should certainly be seen in the finds from Krk and Baška (Figure 3: 3-5), then Pula and Picugi (Orlić 2011: T. III: 3; Guštin 1987: Fig. 3: 2), all the way to Sermin and Socerb near Koper (Cunja and Mlinar 2010: 110: 103, 111: 108), as well as finds in the south-eastern Alpine hinterland (Figure 4).

On fibulae Fibulae of the Picugi type The oldest fibulae are of the so-called Middle La Tène schemes represented by the Picugi type – discovered were two variants of the later in two different contexts. Namely, the first example is an old find from the 19th century collected at the location of Kaštel (Figure 3: 1) (Konestra 2016: 135: 144), while the second fibula was found during excavations of the ceramic workshop on the Igralište site (Figure 3: 2) (Konestra 2016: 120: 99). It comes from a mixed layer of Roman-era rubble above the Roman architecture (SU 004) (Figure 2: 1). Both fibulae are smaller, damaged and partially preserved consisting of a bow, part of the foot and joint, and one half of the spiral on the head. Due to their typological definition, poor preservation and lack of the indicative foot, they can be determined in general as fibulae of Picugi type according to the typology of M. Guštin (Guštin 1987: 51-53, fig. 12; 1991: 38-39). They are considered being, given the available closed contexts, more significant for the male costume of the end of Late Iron Age and the early Roman period.

5

Martina Blečić Kavur

Figure 3. Fibulae of Picugi type from 1-2: Crikvenica, 3: Krk (according to Lo Schiavo 1970), 4-5: Baška (according to Glogović 1989), 6-7: Bakar, 8-9: Osor (according to Blečić Kavur 2015, M 1: 1).

Despite the morphological tradition of fibulae made according to the Middle La Tène scheme, in chronological aspect they marked the last periods of the end of Late Iron Age in the area of the northern Adriatic and its hinterland. Based on graves from Posočje, they are chronologically determined to IVa and IVb phases of the Idrija group, thus indicating their use during the entire Late La Tène period (Guštin 1987: 52-53; 1991: 38). Certainly, such dating will be acceptable for our fibulae described here, with the difference that the

phase VIb of the Histrian culture and the territory of Kvarner should be considered in terms of Romanization of these territories and not in the sense of Late La Tène cultural intertwining (Božič 2008: 84-85; Blečić Kavur 2015: 221, 224). Fibulae of the Almgren 65 type The most numerous group consists of four cast bronze fibulae (Figure 5, 7-9). Unlike the previous ones, all of 6

Inside fashion fusion: Fibulae from Crikvenica

Figure 4. Distribution map of Picugi type fibulae, Aquileia and Picugi variants: 1: Bologna, 2: Este, 3: SiebeneichTerlano, 4: Reichenhall bei Karlstein, 5: Piave di Cadore, 6: S. Floriano di Polcenigo, 7: Lagole, Vittorio Veneto, 8: Porcia, 9: S. Martino di Campagna, 10: Aquileia, 11: Terzo d’Aquileia, 12: Strassoldo, 13: Medea, 14: Pozzuolo del Friuli, 15: Cividale, 16: Most na Soči, 17: Idrija pri Bači, 18: Reka, 19: Kranj, 20: Vrhnika, 21: Ljubljanica-Bevke, 22: Magdalenska gora, 23: Šmihel, 24: Kovačevše, 25: Razdrto-Mandrga, 26: Stari grad nad Uncem, 27: Socerb, 28: Sermin, 29: Izola-Simonov zaliv, 30: Piran-Fornače, 31: Nova VasKaštelir, 32: Picugi, 33: Limska Gradina, 34: Pula, 35: Nesactium, 36: Rim, 37: Sv. Katarina, 38: Metlika, 39: Vinica, 40: Bakar, 41: Crikvenica, 42: Krk, 43: Baška, 44: Osor, 45: Prozor, 46: Obrovac-Cvijina gradina,47: Nadin, 48: Dragišić, 49: Velika Mrdakovica, 50: Solin (Salona) (supplemented after Guštin 1991; Blečić Kavur 2015).

7

Martina Blečić Kavur

Figure 5. Fibulae of Almgren 65a type from 1: Crikvenica and 2-4: Osor (according to Blečić Kavur 2015, M 1: 1).

them are extremely well preserved, lacking only the pin or a possible part of it. They can be divided in to two types – Almgren 65 and Almgren IV or strongly profiled fibulae according to the typology of O. Almgren (Almgren 1923). Since there are significant differences between them regarding their morphology and style, chronology and culture, they are divided in to two separate categories.

Božič 2008: 86, 145, 130; Meller 2012: 70-71; Drnić and Tonc 2014: 191; cf. Istenič and Šmit 2007). The precise context of the Crikvenica fibula will also support the proposed dating. Namely, radiometric dating of the layer confirmed the erection of structures during the middle of the 1st century BCE, confirming again the already argued fact that the architectural constructions of the first workshop rooms and fence walls were built exclusively using stone building materials (Lipovac Vrkljan 2011: 7; Konestra et al. 2020: 77).

The first fibula is determined as the Almgren 65a1 type (Figure 5: 1). It was discovered in the northern part of the so-called Kiln 3 (SU 025) used for the production of household and tableware (Figure 2: 2), of which only the lower part of the structure has been preserved (Lipovac Vrkljan 2007: 34; Lipovac Vrkljan and Šiljeg 2008: 91; Ožanić Roguljić 2012; Konestra et al. 2020: 78-79, Fig. 3). The bronze fibula, made in the style of Late La Tène fibulae, was preserved intact. It is 6 cm long and as such, it belongs to a somewhat rarer class of this type (Demetz 1999: 29-30). Its rounded massive head with a spiral construction is shaped with a profiled edge. The spiral is shorter and smaller with six coils and an external cord. The bow with oval cross-section ends with a trapezoidal shaped foot. A thickening of three transverse and profiled ribs with a nub further adorns the bow. They were equally represented in male and female costumes, not only as a functional but also as a decorative element of Late La Tène fashion of Lt D1b in terms of the Central European chronology (Demetz 1999: 31; Božič 2008: 62-63, 86; Meller 2012: 66). More precise contexts of their discoveries define their use in the first half and middle of the 1st century BCE. The most impressive prevalence is in the period of Caesarean rule and the associated dynamic process of Romanization, although, especially outside the Italian area, it will remain in use until the Proto-Augustan period (Demetz 1999: 37; 2008: 28-29; Ivčević 2001: 333-334; 2003: 129-130; 2017: 254;

Given the quantity of finds, fibulae of Almgren 65 type would denote the topographic-cultural area of northern Italy, where they developed from fibulae of Middle La Tène schemes, but also of Central Europe, where they were worn singly or in pairs with other fibulae (Figure 6). They were discovered on the territory of the eastern Adriatic coast and in its hinterland, with a higher concentration in the area of Aquileia and Notranjska, Karst and all the way to Posočje (Demetz 1999: 32, Maps 1; 2008; Ivčević 2001: 327-334; Dizdar and Tonc 2013: 54; Drnić and Tonc 2014: 190-191, Map 2; Mlinar 2020: 73, 94). Most closely related examples, although usually without detailed information about their discoveries come from Osor (Figure 5: 2-4), which, besides Crikvenica, is the only site in Kvarner with three examples of this specific variant. The fragment from Lopar will most likely belong to variant 65b (Konestra et al. 2021: fig. 9: 1). Identical fibulae were discovered in graves from Socerb above Koper and further north in Idrija near Bača (Figure 6). Further south, we know their highest concentration from Dalmatia, but unfortunately, lacking any information about the contexts of their discoveries – the only exception is the fibula from Gardun (Tilurium) (Ivčević 2001: T. II; 2017: 254, 300, T. 15: 249). Apart from trade with northern Italy, especially Aquileia, their presence 8

Inside fashion fusion: Fibulae from Crikvenica

Figure 6. Distribution map of Almgren 65 type fibulae: 1: Ripatransone, 2: PortorecantiAncona, 3: Este, 4: Torcello, 5: San Giorgio di Nogaro, 6: Aquileia, 7: Starssoldo, 8: Sevegliano, 9: Tissano, 10: Galleriano, 11: Tomba, 12: Sammardenchia, 13: Chiópris, 14: Pavia di Udine, 15: Udine, 16: Moruzzo, 17: Zugglio, 18: Misincinis-Paularo, 19: Gurina, 20: Magdalensberg. 21: Tonovcov grad-Kobarid, 22: Most na Soči, 23: Idrija pri Bači, 24: Ljubljana-Tribuna, 25: MariborMiklavž, 26: Ptuj-Vičava, 27: Celje-Savinja, 28: Kostanjevica, 29: Mihovo, 30: Novo mesto, 31: Šumenje pri Podturnu, 32: Žerovinšček, 33: Dolenja vas, 34: Stari Grad nad Uncem, 35: Ambroževo gradišče, 36: Čepna pri Knežaku, 37: Socerb, 38: Trieste, 39: Izola-Korte, 40: NovaVasKaštelir, 41: Pula, 42: Crikvenica, 43: Lopar, 44: Osor, 45: Gračac, 46: Dragišić, 47: Gardun (Tilurium), 48: Solin (Salona), 49: Vid (Narona), 50: Blato-Vinkovci (supplemented after Demetz 1999; Drnić and Tonc 2014).

9

Martina Blečić Kavur

Figure 7. Aucissa type fibulae from 1: Crikvenica, 2: Grobnik-Grobišće, 3–4 Fulfinum (Mirine, Sepen) (M 1: 1).

along the eastern Adriatic coast can be associated with the expansion of the Roman army and earlier business or economic activities during the Late Republic and the Early Empire (Ivčević 2001: 332-333; 2003: 130; 2017: 254-255; cf. Demetz 2008).

the imprinting of the stamp was significantly more prevalent than on the previous, older variants. We can follow them distributed throughout the geography of the Empire, although the accurately described example is closely associated with the large corpus of such finds from the province of Dalmatia. There they are mostly interpreted as parts of male costumes; this is of military units and activities of regional craft centres (Koščević 1980: 15-16; Feugere 1985: 319-323, fig. 46; Šeparović 1998: 185-186; Demetz 1999: 164–167; Ivčević 2002: 235; 2003: 131; 2007: 239-240; Marović 2006; Buora 2008: 30; Jadrić-Kučan and Zaninović 2015: 29-34). However, discoveries from graves as well as from other contexts indicate that such fibulae were also used in the socalled civil fashion in both male and female costumes – making their presence and distribution one of the keys to understand the variable processes of Romanization (Demetz 1999: 206; Buora 2008: 32; Ivčević 2007: 230; 2011: 169; Sedelmayer 2017: 131-134). Probably a reflection of this process is the long duration of their use throughout the century, from the Augustan period until the end of the 1st century AD (Riha 1994: 100-103; Šeparović 1998: 185-186; Marović 2006; Ivčević 2002: T. VII: 64-66; 2007: 239-240, 245-246; 2017: 257-258; Sedelmayer 2017: 135).

Fibula of the Aucissa type Only one example of the most widespread and most numerous early Roman provincial fibulae of Aucissa type was discovered at the Crikvenica workshop complex (Figure 2: 3, 7: 1). It is a massive but damaged fibula, 5.7 cm long, of standardized construction. The semi-circular bow is of semi-circular cross-section decorated with several short parallel-incised lines. It is characterized by a two-part construction with a hinged fastening mechanism. It ends with a square header that connects it to the head. The foot of the fibula is damaged, with a triangular pin holder ending with a decorative profiled button set straight. The short crossbar and header plate are made of iron and therefore extremely corroded – a state that damaged any possible decoration or inscription that may have been present on the fibula. These are standard features of younger variants of fibulae of the Aucissa type (type Feugere 22b) on which 10

Inside fashion fusion: Fibulae from Crikvenica

Figure 8. Southern Pannonian strongly profiled fibulae from 1-3: Crikvenica and 4: Stinica (according to Koščević 1999, M 1: 1).

Although we can observe the closest comparisons in the area of northern Dalmatia, especially in Podgrađe (Asseria) and Nin (Aenona), then in Solin (Salona), Gardun (Tilurium), Ivoševci (Burnum) or Vid (Narona) (Ivčević 2002, T. VI: 51-T. VIII: 75; 2003: 134, 160, Fig. 9-10; 2007: 250-262; 2014: 164-165, T. 6: 59-65; 2017: 301, T. 15: 253-255; 2021: 266-267, T. 2: 20; Jadrić-Kučan and Zaninović 2015: 36-37, cat.n. 60), Aucissa fibulae are also known from the Kvarner basin. Their largest number, where we can observe the closest parallels in as many as 17 examples, come from the Kavanela necropolis in Osor (Apsorus) (Klodič 1885: III, fig. 2; Težak-Gregl 1982: Pl. 3: 4-8). Similar examples were discovered in Fulfinum (Mirine, Sepen) on the island of Krk (PPMHP 117512: https://digitalni.ppmhp.hr/?pr=i&id=33720) (Figure 7: 3-4), and on the necropolis Grobišće near Grobnik (PPMHP 104314: https://digitalni.ppmhp. hr/?pr=i&id=19344) (Cetinić 1996: 196-197) (Figure 7: 2) but most of them, like the fragment from Tarsatica (Višnjić 2009b: 155-156, T. I: 1), do not belong to the form of fibulae adequate to the Crikvenica example.

South Pannonian strongly profiled fibulae The next group of three fibulae belongs to an extremely large and spatially impressively distributed group of highly profiled fibulae from the Almgren IV group, which are classified into a multitude of regional types with multiple developmental phases (Almgren 1923: 34-47; cf. Gugl 1995; 2008; Knauseder 2014). From the Crikvenica ceramic workshop comes a local variant (68) of South Pannonian highly profiled fibulae (Peškař 1972: 80; Koščević 1980: 22-23; 1997: 45). Their development is derived precisely from the older, already described, fibulae of Almgren 65 type. They feature a characteristic profiled and curved bow – the form of the most common Roman fibula during the 1st century AD, in the area of Noricum, the southern province of Pannonia and the surrounding area (Koščević 1980: 20-22; 1997: 45-46; Bojović 1983: 35; Gugl 1995: 11-12, appendix 1; 2008: 38; Gregl 1997: 68; Ivčević 2003: 130-131; 2009: 94-95; Vulić 2012: 102-103). 11

Martina Blečić Kavur Crikvenica fibulae were constructed from a single piece (Figure 8: 1-3). They have an elliptical and thickened cross-section of the head, elongated in the shape of a deltoid. On the head, there is a spiral spring, which is fastened to the head with a holder and additionally fastened with a long and narrow rectangular head plate and a front chord. Two fibulae have eight (Figure 8: 2-3) and one ten coils (Figure 8: 1). They are characterized by a massive curved, saddle-shaped and high bow with a horizontally placed decorative profiled button of polygonal or circular cross-section in the middle. The low and long foot ends with a round or slightly profiled knob – only one fibula has the foot perforated with three small and irregular circular perforations (Figure 8: 3). These elements classify them into a younger variant that represents the final phase of strongly profiled fibulae with a head plate (Koščević 1980: 22; Ivčević 2009: 95).

63; Lipovac Vrkljan and Konestra 2018: 23; Konestra et al. 2020: 78-79). The remaining two fibulae (Figure 8: 2-3) were discovered at the bottom of the layer with burned clay (SU 088) covering the area from the western circumferential wall of Kiln 1, to the western boundary of the investigated space (Figure 2: 5-6). The layer is contemporary with the filling in the smaller room (SU 272), thus the context of discovery supports the dating of the remaining two fibulae in to the same time span. Based on the stratigraphy of their discovery as well as their typological and stylistic features, it will not be wrong to consider all the described fibulae being contemporary, not only in their production, but also during their use, i.e. disposal. Strongly profiled fibulae are not a common find in Kvarner, but also not an isolated discovery which may be attributed to insufficient research or, more likely, partial publication. For now, they are known from the necropolis of nearby Bakar (Ljubić 1882: T. IV: 49), the principia of Tarsatica (Višnjić 2009b: 156, T. I: 2), from Senia (Ljubović 2000: 31) and from the necropolis of Osor (Težak-Gregl 1982: 106-107, fig. 1: 4-5), while the southern examples originate from Stinica near Jablanac (Figure 8: 4) and Karlobag (Brunšmid 1891: 169, fig. 73: 2; Koščević 1980: 21-22; 1999: 29, 32, T. I: 5-6). However, it is probable that none of the fibulae will represent closer typological-stylistic parallels to the Crikvenica ones, which thus remain the only known examples of the South Pannonian strongly profiled fibulae in the region.

We know the largest collection of this variant of fibulae from Sisak (Siscia) — discovered were as many as 164 examples together with numerous semi-finished products. Based on the concentration of finds, as well as specific rough workmanship, sharp contours and inadequate finishing and simplified technical details, they were considered being a separate group designated as typical fibulae of ancient Siscia, i.e. its artistic craft (Koščević 1980: 22; 1997: 45; Ivčević 2003: 131; 2009: 95; Vulić 2012: 101-102). We will therefore find very similar examples to those of Crikvenica in Sisak (Koščević 1980, Pl. XI-XV), in Velika Gorica at the Visoki brijeg necropolis (Gregl 1990, Pl. 1: 3; Pl. 3: 4; Pl. 4: 1-2 ), at the Osječenica hillfort near Dvor na Uni (Ožanić 1998: 36, fig. 3: 2) and at Cvijina gradina hillfort near Obrovac (Čondić 2014: 52), then in Asseria, Salona and Narona (Ivčević 2002, 256, T: XI : 97-98; 2003: 133, 159, fig. 3; 2009: 84, 103, 108: 11), and in the gravitating Iapodian necropolises of Pounje (Busuladžić 2010: 5254). In addition to the province of Dalmatia, they were also found in the area of Dacia and in Noricum (Koščević 1995: 45; Ivčević 2002: 237; Gugl 2008: 38; Vulić 2012: 102). Closed contexts, mostly burial inventories from rich Roman necropolises, where they are often found with other variants of strongly profiled Almgren-type fibulae, enabled their closer dating to the second half of the 1st and the beginning of the 2nd century AD. They were attributed to male, as a rule, military costume of the Pannonian area (Koščević 1997: 45-46; Božič 2008: 111-112; Ivčević 2009: 95). This context is joined by the Crikvenica fibula (Figure 8: 1), discovered in a layer of a smaller room (SU 272), along the eastern fence wall and located in the immediate vicinity of Kiln 4 and drainage system with amphorae (Figure 2: 4). The filling of this smallest Crikvenica kiln was composed exclusively of pyramidal loom-weights and was associated with the greatest flourishing of the workshop complex Ad Turres during the 1st century AD (Ožanić Roguljić 2016:

Fibula with a multiply segmented bow The last fibula from Crikvenica can be attributed to the typological group of fibulae with multiply segmented bow (Figure 9: 1). It was made of bronze, while the short wire in the tube, partially preserved on the head of the fibula, was made of iron, so it was corroded at that point. The slightly curved bow is divided into six differently shaped and irregularly distributed transverse ribs along the length of the bow. The decoration decreases towards the foot, and the two sharper ribs are finished with incised short oblique lines. Traces of circular ornaments are visible on the foot of the fibula, as well as a part of a semi-circular, originally perforation at the end of the triangular foot. It is a two-part type of fibula with an articulated mechanism that is significant for the area of the Roman province of Gaul, where they most likely developed. Although represented in the fashion of the whole Empire, they are more representative for the western European areas. Due to their prevalence, they are characterized by many local deviations, which divides them into many types and variants. Numerous authors have therefore paid great attention to their classification and interpretation, assuming, most likely, their origin 12

Inside fashion fusion: Fibulae from Crikvenica

Figure 9. Fibulae of a multiply segmented bow from 1: Crikvenica and 2: Asseria (according to Ivčević 2009, M 1: 1).

in the tradition or fusion of constructive and decorative elements of fibulae of the Aucissa, Nertomarus or Langton Down type. Among others, they are classified as a variant of Feugere 23b, Riha 5.6 or Gaspar 30a (Feugere 1985: 331-333; Riha 1994: 110-112; Gaspar 2007: 57: 30a; Ivčević 2009: 90; Petković 2010: 64) and dated, according to finds from graves, in to the 2nd quarter of the 1st century, with a possible duration throughout the 2nd century AD. Unlike Aucissa fibulae, their use is more significant for the civil fashion of that time, which flourished precisely in the time of the Flavians (Feugere 1985: 333-334; Ivčević 2002: 238; 2003: 132; 2009: 89-90; 2011: 169).

circle of reproduction restores both relations, between people and those between objects. The action of people on objects, that is, objects on people, as long as we create them, is in fact equivalent to the action between people (Sahlins 1999: 227-284; Gosden and Marshall 1999: 172-177; Tilley 1999: 76). All of these processes occur in a certain continuity of cultural circumstances that are less constant and more changeable, and are usually in a constant vortex of social change, which forces us to adapt infinitely. Material culture proves and teaches us about the intertwining of countless social relations, identities and economic processes within heterogeneous cultural communities that were geographically further or closer to each other and even in different hierarchical positions. The processes of Romanization, understood as globalization and cosmopolitanism, reflect this very plastically (Rothe 2012; 2013; Versluys 2014). However, it is precisely the inequality, peculiarity as well as certain ideologicalsymbolic values, that enable individual objects to make the area in which they are special and unique in the complex circle of universal production, circulation, use, storage – living.

The Crikvenica fibula can be attributed to the type C1, fibulae of small dimensions according to the typology of S. Rieckhoff-Pauli (Rieckhoff-Pauli 1977: 48-55, 63-64). They are quite modestly represented in the province of Dalmatia with one specimen in Asseria (Figure 9: 2), Salona and Narona (Ivčević 2002: 258, T. XIV: 128; 2003: 134, 160, fig. 11; 2009: 101, T. 1: 5), and present, although rarely, in the area of southern Pannonia, e.g. in Siscia (Koščević 1980: 29-30, T. XXVIII: 244) and in Sapaja near Vršac (Petković 2010: 67, T. IX: 4). They should be considered as imported objects that could adorn both male and female costumes. Since this fibula was also discovered in an alluvial layer, mixed with fragments of pottery (SU 056) (Figure 2: 7), its contextual determination as well as more precise chronological dating is not possible, although it can be narrowed to the second half of the 1st century AD, i.e. at the period of their greatest popularity.

Understood from this perspective, the Crikvenica fibulae indeed originate from an exceptional site. In addition to their typological-stylistic and culturalchronological value, as indeed, first such fibulae in Kvarner discovered in optimal conditions and known contexts of the archaeological record, they provide many more opportunities to understand their users, i.e. people within the Ad Turres ceramic complex and their economic activities.

Conclusion Archaeological science has been arguing for decades that the connection between material culture, i.e. the goods and functional necessities of the social system is causal. The relationship of people and objects in the

With the exception of numismatic finds (Ilkić 2018), it is unusually interesting to note that fibulae represent the most numerous group of bronze objects on this 13

Martina Blečić Kavur site. Moreover, their primary purpose is clear — used as a practical and functional part of the costume adhered to by the upper layers of clothing, mostly cloaks. However, fibulae also had a decorative function, and, as always visible, they were a kind of cultural and status symbol, i.e. a code of representation and recognition of many identities, cultural as well as social (Rothe 2012: 60-62; cf. Ivleva 2017). In chronological terms, older examples from already the first half and possibly the middle of 1st century BCE are represented by fibulae of the Picugi and Almgren 65a1 types. Their centres of production are located in the area of Aquileia where, most likely, we have to look for their origin (Demetz 2008). How and in what way they arrived cannot be explained from this context of the workshop complex. However, they could certainly be direct imports from either the indigenous population or part of the costumes of various newcomers, or a means of cultural exchange, so they may silently testify to this adaptive procedural course of the early Romanization of the area. Picugi-type fibulae are numerous in the northern Adriatic and its hinterland (Figure 4), with the highest concentration of finds in Osor in Kvarner, which demonstrates their relatively great popularity in the costume of the local population in the late 2nd and first half of the 1st century BCE. The picture is quite different with italic fibulae of Almgren 65 type which are much rarer (Figure 6), but where mostly older bronze specimens are more common in the hinterland – especially Notranjska, Karst and all the way to Posočje and Dolenjska (Laharnar 2012: 151; cf . Istenič and Šmid 2007: 145; Mlinar 2020: 73). It will most likely not mark the fashion of domestic costumes, but will suggest recognizable imported elements or costumes of the newly arrived (Italic) population. In any case, their presence certainly tells us about the earlier economic engagement of that rural area and most likely, the first plants of the workshop complex between Tarsatica and Senia, which were still connected with the area of Friuli and northern Italy. It is there, as in the western part of the south-eastern Alpine region, that graves with non-homogeneous material culture are often found – and Late La Tène costume elements, especially fibulae, are often documented with early Roman imports (Turk et al. 2009: 48-54; Istenič 2013: 24-25; Sagadin 2020: 206-207) and therefore, like the examples from Crikvenica, they should be understood in the concept of intertwining the diversity of material culture and people in a unique cultural framework (Versluys 2014: 14-19; Ivleva 2017).

workshops in Siscia, while the Aucissa-type fibula is already classified as a connection with a workshop in the Dalmatian area. And although they are mostly correlated with the costumes of soldiers, they are also found in various civilian contexts, which Crikvenica examples once again support. Therefore, they are interpreted in a more regional orientation of trade along the Adriatic coast, especially with the area of northern Dalmatia and strong coastal centres with which the Ad Turres figlina had established business relations (Borzić et al. 2018). It cannot be ruled out that the strongly profiled fibulae may have arrived precisely through these southern areas of the province, where we find them more numerous. Direct contacts with the northern parts of the province of Pannonia are not excluded, if we understand them as a consequence of the development of interregional and/or interprovincial trade, in the context of which one should certainly understand the fibula of multiply segmented bow imported from the western part of the Empire. The peak of its power Sextus M(e/u)tillius Maximus figlina fulfilled during the 1st century AD which, in addition to the typology and distribution of ceramics (Konestra et al. 2020), is confirmed by a set of numismatic finds with the highest concentration in the period from the Julian-Claudian to the Flavian dynasty (Ilkić 2018). The discoveries of the provincial fibulae presented here are in complete temporal agreement with this. Taken as a whole, and within the collection of finds from all over Kvarner, Crikvenica fibulae will certainly support the thesis of an early connection with Roman economic intentions, which was reflected monumentally, with the construction of ‘Roman style’ structures along the region in both urban and rural contexts (Konestra et al. 2020: 8990), and miniature – associated with material culture. Acknowledgments The processing and publication of ‘Crikvenica’ fibulae would not have been possible without the selfless transfer of material made possible by the head of the figlina research in Crikvenica, Goranka Lipovac Vrkljan. I thank her most sincerely and with deep respect for this, as well as for the endless instructions, discussions, suggestions and encouragements. Suzana Čule and Miljenka Galić signed the drawings, and the photos Damir Doračić, Miroslav Matejčić and Dragan Pelić, for which I am very grateful. I would also like to thank my colleagues Ana Konestra, Sanja Ivčević, Boris Kavur and Ranko Starac, who, each in their own way provided a valuable contribution to this research.

All early imperial fibulae from Crikvenica also represent imported, but widespread and very significant elements of the Roman costume. The situation with these younger fibulae at the turn of the millennium is changing, though. Namely, the strongly profiled fibulae of the southern Pannonian variant should be associated with possible

The study was made within the scope of the ‘Communities of the dead, societies of the living. Late Bronze Age of Eastern Slovenia (J6-9363)’ project financed by the Slovenian Research Agency of the Republic of Slovenia.

14

Inside fashion fusion: Fibulae from Crikvenica

Catalogue3 1.

2.

3.

4.

5.

6.

Fragment of a bronze fibula of the Picugi type, significantly damaged during burning. The bow of the fibula is of a thinned rectangular cross-section, ending with three-coil spring. A plate with one decorated rib and a spherical thickening was accepted on the bow with a joint. Foot and pin missing. L. 4.9 cm, H. 1.5 cm. MGC 2571 – Figure 3: 2. The bronze fibula of the Almgren 65a1 type. The massive head with a spiral construction is shaped rounded with a profiled edge. The spiral is shorter with six coils and an external cord. The bow with oval cross-section ends with a trapezoidal shaped foot. A thickening of three transverse and profiled ribs with a tongue further adorns the bow. L. 6 cm, H. 3 cm. MGC 2569 – Figure 5: 1. Bronze, massive and damaged Aucissa type fibula. The semi-circular bow is of semi-circular cross-section decorated with several short parallel-incised lines. It ends with a square header that connects it to the head. The foot of the fibula is damaged, with a triangular pin holder and decorative profiled knob set straight. The short crossbar and header plate are made of iron. Pin is partially preserved. L. 5.7 cm, H. 2.9. MGC 4997 – Figure 7: 1. Bronze strongly profiled fibula, with massive curved, saddle-shaped and high bow with a horizontally placed decorative profiled button of circular cross-section. The cross-section of the head is elliptical and thickened, with tencoil spiral fastened to the head with a holder and additionally fastened with a long and narrow rectangular head plate and external cord. The low foot ends with a round knob. L. 5.6 cm, H. 2.9 cm. MGC 2570 – Figure 8: 1. Bronze strongly profiled fibula, with massive curved, saddle-shaped and high bow with a horizontally placed decorative profiled button of biconical cross-section. The cross-section of the head is elliptical and thickened, with eightcoil spiral fastened to the head with a holder and fastened with a long and narrow rectangular head plate and external cord. The low foot ends with a profiled knob. The pin is partially preserved. L. 5.4 cm, H. 2.5 cm. MGC 4998 – Figure 8: 2. Bronze strongly profiled fibula, with massive curved, saddle-shaped and high bow with a horizontally placed decorative profiled button

7.

8.

of circular cross-section. The cross-section of the head is elliptical and thickened, with fourcoil spiral fastened to the head with a holder and fastened with a long and narrow rectangular head plate and external cord. The low foot ends with a profiled knob. It has three perforations on the foot and the pin missing. L. 6.3 cm, H. 2.4 cm. MGC 4999 – Figure 8: 3. Partially preserved bronze fibula of a multiply segmented bow. The short wire in the tube, made by an outward twist, was made of iron. The slightly curved bow is divided into six differently shaped and irregularly distributed transverse ribs along the length of the bow. The decoration decreases towards the foot, and the two sharper ribs are finished with incised short oblique lines. Traces of circular ornaments are visible on the foot of the fibula, as well as a part of a semi-circular, originally perforation at the end of the triangular foot. Pin is not preserved. L. 5 cm. MGC 4996 – Figure 9: 1. Fibula of the Picugi type from the Kaštel site/ Pavlinski samostan. The fragment of the bronze fibula was significantly damaged and corroded. The bow is knee-shaped, of flat cross-section, which is narrowed towards the foot and ends with five-coil spring. At the highest part of the bow is attached a joint of separated ends, with a central larger and two lateral smaller ribs. Foot and pin missing. L. 3.8 cm; D. of the bow 0.7 cm. AMZ A-9307 – Figure 3: 1.

References Almgren, O. 1923. Studien über Nordeuropäische Fibelformen der ersten nachchristlichen Jahrhunderte mit Berücksichtigung der provinzialrömischen und südrussischen Formen (Mannus Bibliothek 23). Lepizig: Kabitzsch. Bilić Dujmušić, S. 2014. Bitka kod Krka godine 49. pr. Kr., in: G. Lipovac Vrkljan, B. Šiljeg, I. Ožanić Roguljić and A. Konestra (eds) Roman pottery and glass manufactures. Production and trade in the Adriatic region. Proceedings of the 3rd International Archaeological Colloquy (Crikvenica, 4-5 November 2014): 107–124. Crikvenica – Zagreb: Institut za arheologiju, Muzej Grada Crikvenice. Blečić, M. 2001. Prilog poznavanju antičke Tarsatike. Vjesnik Arheološkog muzeja u Zagrebu 34/1: 65–122. Blečić Kavur, M. 2011. Facile est inventis addere: fibule kasnorepublikanskog doba iz Ad Turres ili prilog poznavanju kulturno-povijesne baštine Kvarnera / Fibule di epoca tardorepubblicana da Ad Turres ovvero un contributo alla conoscenza del patrimonio storico-culturale del Quarnero, in: G. Lipovac Vrkljan (ed.) Program kolokvija i knjiga sažetaka II. Međunarodni arheološki kolokvij: Rimske keramičarske

The fibulae numbers in the catalogue correspond to the fibulae numbers on the workshop site layout (Figure 2). The fibulae are kept in the Crikvenica Town Museum (MGC) and in the Archaeological Museum in Zagreb (AMZ). 3 

15

Martina Blečić Kavur i staklarske radionice / Programma del colloquio e libro degli abstracts II Colloquio archeologico internazionale: Officine per la produzione di ceramica e vetro in epoca romana, Crikvenica, 28. i 29. 10. 2011. Crikvenica: Muzej Grada Crikvenice, Institut za arheologiju. Blečić Kavur, M. 2015. A coherence of perspective: Osor in cultural contacts during the Late Iron Age. Koper – Mali Lošinj: Univerza na Primorskem, Lošinjski muzej. Bojović, D. 1983. Rimske fibule Singidunuma. Beograd: Muzej Grada Beograda. Borzić, I., M. Čelhar, G. Zaro and V. Glavaš 2018. New finds of Crikvenica 1-type amphorae in the area of southern Liburnia, in: G. Lipovac Vrkljan and A. Konestra (eds) Pottery Production, Landscape and Economy of Roman Dalmatia. Interdisciplinary approaches (Archaeopress Roman Archaeology 47): 50–56. Oxford: Archaeopress. Božič, D. 2008. Late La Tène-Roman cemetery in Novo mesto. Ljubljanska cesta and Okrajno glavarstvo. Studies on fibulae and on relative chronology of the Late La Tène period (Katalogi in monografije 39). Ljubljana: Narodni muzej Slovenije. Brunšmid, J. 1898. Arheološke bilješke iz Dalmacije i Panonije: II. Dalmatia. Vjesnik Hrvatskoga arheološkoga društva 3: 150–205. Buora, M. 2008. Diffusione delle fibule Aucissa nell’area altoadriatica, in: M. Buora and S. Seidel (eds) Fibule antiche del Friuli (Cataloghi e monografie archeologiche dei Civici musei di Udine 9): 30–32. Udine – Roma: L’Erma di Bretschneider. Busuladžić, A. 2010. The morphology of Antique Fibulae in the collection of the National Museum of Bosnia and Herzegovina. Sarajevo: Zemaljski muzej Bosne i Hercegovine. Cetinić, Ž. 1996. Rezultati novijih arheoloških istraživanja Grobišća. Grobnički Zbornik 4: 188–204. Crismani, A. and J. Righi 2002. Le sepolture protostoriche e il catalogo dei materiali, in: M. Viduli Torlo (ed.) La necropoli di San Servolo. Veneti, Istri, Celti e Romani nel territorio di Trieste: 63–88. Trieste: Civici musei di Storia ed Arte. Cunja, R. and M. Mlinar (eds) 2010. S fibulo v fabulo: fibule iz Istre, s Krasa, iz Notranjske in Posočja med prazgodovino in zgodnjim srednjim vekom / Con la fibula nella storia : fibule dall’Istria, dal Carso, dalla Carniola Interna e dall’Isontino tra preistoria e alto medioevo. Koper: Pokrajinski muzej. Čondić, N. 2014. Nova istraživanja – nekropola, in: M. Jurjević and N. Čondić (eds) Cvijina gradina – tragom zaboravljene prošlosti (Katalozi i monografije 9): 46 – 57. Zadar: Arheološki muzej Zadar. Demetz, S. 1999. Fibeln der spätlatène- und frühen römischen Kaiserzeit in den Alpenländern (Materialen und Forschungen 4). Rahden/Westfahlen: M. Leidorf. Demetz, S. 2008. Fibule di tipo Almgren 65, in: M. Buora and S. Seidel (eds) Fibule antiche del Friuli (Cataloghi e

monografie archeologiche dei Civici musei di Udine 9): 27–29. Roma: L’Erma di Bretschneider. Dizdar, M. and A. Tonc 2013. Finds of fibulae from 1st century Croatia: Trade and exchange between Eastern Alps, the Danube and the northern Adriatic area before and during Roman conquest, in: G. Grabherr, B., Kainrath and T. Schierl (eds) Verwandte in der Fremde. Fibeln und Bestandteile der Bekleidung als Mittel zur Rekonstruktion von interregionalem Austasch und zur Abgrenzung von Gruppen vom Ausgreifen Roms während des 1. Punischen Kriegs bis zum Ende des Weströmischen Reiches: 50–75. Innsbruck: Innsbruck University Press. Drnić, I. and A. Tonc 2014. Late La Tène and Early Imperial Fibulae from the Iapodian Territory. Prilozi Instituta za arheologiju u Zagrebu 31: 181–214. Feugere, M. 1985. Les fibules en Gaule méridionale de la conqête à la fin du Ve s. ap. J.-C. Paris: Éditions du centre national de la recherche scientifique. Gaspar, N. 2007. Die keltischen und gallo-römischen Fibeln vom Titelberg (Dossiers d’Archeologie du Musée National d’Historie et d’Art XI). Luxemburg: Musée national d’Histoire et d’Art. Glogović, D. 1989. Nalazi iz okolice Baške na otoku Krku, in: Ž. Rapanić (ed.) Arheološka istraživanja na otocima Krku, Rabu i Pagu i u Hrvatskom primorju (Izdanja Hrvatskog arheološkog društva 13): 97–102. Zagreb: Hrvatsko arheološko društvo. Gregl, Z. 1990. Rimskodobna nekropola Velika Gorica – Visoki brijeg. Opvscvla archaeologica 14 (1989): 67‒73. Gregl, Z. 1997. The Roman cemeteries of northern Croatia. Zagreb: Arheološki muzej u Zagrebu. Gugl, C. 1995. Die römischen Fibeln aus Virunum. Klagenfurt: Landesmuseum für Kärnten. Gugl, C. 2008. Le ‘kräftig profilierte Fibeln’ dal Friuli. Uno sguardo di insieme, in: M. Buora and S. Seidel (eds) Fibule antiche del Friuli (Cataloghi e monografie archeologiche dei Civici musei di Udine 9): 33–41. Roma: L’Erma di Bretschneider. Guštin, M. 1987. La Tène fibulae from Istria. Archaeologia Iugoslavica 24: 43–56. Guštin, M. 1991. Posočje in der jüngeren Eisenzeit (Katalogi in monografije 27). Ljubljana: Narodni muzej Slovenije. Gosden, C. and Y. Marshall 1999. The cultural biography of objects. World Archaeology 31/2: 169–178. Horvat, J. 2009. Selected aspects of Romanisation in Western and Central Slovenia, in Aspetti e problemi della romanizzazione. Venetia, Histria e arco alpino orientale (Antichita Altoadriatiche 68): 355–381. Ilkić, M. 2018. Numismatic finds from the area of the workshop of Sextus Metillius Maximus in Crikvenica, in: G. Lipovac Vrkljan and A. Konestra (eds) Pottery Production, Landscape and Economy of Roman Dalmatia. Interdisciplinary approaches (Archaeopress Roman Archaeology 47): 44–49. Oxford: Archaeopress. Istenič, J. 2013. Early Roman graves with weapons in Slovenia: an overview, in M. Sanader, A. Rendič16

Inside fashion fusion: Fibulae from Crikvenica

Miočević, D. Tončinić and I. Radman-Livaja (eds) Weapons and Military Equipment in a Funerary Context: 23–35. Zagreb: Filozofski fakultet Sveučilišta u Zagrebu, Arheološki muzej u Zagrebu. Istenič, J. and Ž. Šmit 2007. The beginning of the use of brass in Europe with particular reference to the southeastern Alpine region, in: S. La Niece, D. R. Hook and P. T. Craddock (eds) Metals and mines. Studies in archaeometallurgy: 140–147. Cambridge: Cambridge University Press. Ivčević, S. 2001. Fibule tip Almgren 65 i Nova Vas iz Arheološkog muzeja Split. Vjesnik za arheologiju i historiju dalmatinsku 94: 325–345. Ivčević, S. 2002. Fibule, in: E. Marin (ed.) Longae Salonae, I: 229–277; II: 124–146. Split: Arheološki muzej u Splitu. Ivčević, S. 2003. Antički metalni predmeti iz Narone, in: E. Marin (ed.) Arheološka istraživanja u Naroni i dolini Neretve (Izdanja Hrvatskog arheološkog društva 22): 129–167. Zagreb – Metković – Split: Hrvatsko arheološko društvo, Arheološki muzej u Splitu. Ivčević, S. 2009. Unpublished fibulae from Asseria in the Archaeological Museum in Split. Asseria 7: 81–114. Ivčević, S. 2011. Antičke fibule iz vojničkog logora Tilurij (Gardun), in: A. Librenjak and D. Tončinić (eds) Arheološka istraživanja u Cetinskoj krajini (Izdanja Hrvatskog arheološkog društva 27): 161–186. Zagreb: Hrvatsko arheološko društvo, Muzej Cetinske krajine, Filozofski fakultet Sveučilišta u Zagrebu. Ivčević, S. 2014. Metalni nalazi, in M. Sanader, D. Tončinić, Z. Buljević, S. Ivčević and T. Šeparović (eds) Tilurium III. Istraživanja 2001.-2006. godine: 147‒223. Zagreb: Filozofski fakultet Sveučilišta u Zagrebu. Ivčević, S. 2017. Metalni nalazi, in: M. Sanader, D. Tončinić, Z. Šimić Kanaet, S. Ivčević, Z. Buljević, T. Šeparović and I. Miloglav (eds) Tilurium IV. Arheološka istraživanja 2007.-2010. godine: 239–358. Zagreb: Filozofski fakultet Sveučilišta u Zagrebu. Ivčević, S. 2021. Metalni nalazi, in: M. Sanader, D. Tončinić, Z. Šimić Kanaet, Z. Buljević and S. Ivčević (eds) Tilurium V. Arheološka istraživanja 2010.2018. godine: 261–299. Zagreb: Filozofski fakultet Sveučilišta u Zagrebu. Ivleva, T. 2017. ‘Active brooches’: Theorising brooches of the Roman north-west (first to third centuries AD), in: T. F. Martin and R. Weech (eds) Dress and Society: Contributions from Archaeology: 69–93. Oxford: Oxbow Books. Jadrić-Kučan, I. and J. Zaninović 2015. Fibulae from the Burnum Arcaheological Collection (BURNUM – Katalozi i monografije VII). Drniš – Šibenik: Javna ustanova nacionalni park ‘Krka’, Gradski muzej Drniš. Klodič, A. R. 1885. Die Ausgrabungen auf Ossero. Mittheilungen der KK-Central Commission XI: I–VII. Knauseder, D. 2014. Überlegungen zu den kräftig profilierten Fibeln in: E. Trinkl (ed.) Akten des 14. Österreichischen Archäologentages am Institut für Archäologie der Universität Graz vom 19. bis 21. April 2012.

Veröffentlichungen des Instituts für Archäologie der Karl-Franzens-Universität Graz 11: 153–166. Wien: Phoibos. Konestra, A. 2016. Katalog, in: A. Konestra and T. Rosić (eds) 845˚C. Ad Turres (Exhibition catalogue): 120, 135. Crikvenica: Muzej Grada Crikvenice. Konestra, A., G. Lipovac Vrkljan and B. Šiljeg 2020. The assortment of ceramic building materials from the pottery workshop of Sextus Me(u)tillius Maximus at Crikvenica (Croatia). Prilozi Instituta za arheologiju u Zagrebu 37: 73–98. Konestra, A., F. Welc and P. Androić Gračanin 2021. Lokalitet na rtu Zidine u Loparu u kontekstu obalnih rezidencijalno-gospodarskih kompleksa otoka Raba. Annales Instituti Archaeologici 17: 152-170. Koščević, R. 1980. Antičke fibule s područja Siska. Zagreb: Odjel za arheologiju Centra za povijesne znanosti Zagreb. Koščević, R. 1997. Metalna produkcija antičke Siscije. Prilozi Instituta za arheologiju u Zagrebu 11-12 (19941995): 41–62. Koščević, R. 1999. Neobjavljene fibule iz rimske Dalmacije i Mezije. Prilozi Instituta za arheologiju u Zagrebu 15-16 (1998-1999): 29–39. Laharnar, B. 2012. Notranjska med prazgodovino in antiko, unpublished PhD Dissertation. Ljubljana: University of Ljubljana. Lipovac, G. 1991. Razmatranje o problemima antičkog bedema grada Krka – povodom novih nalaza. Prilozi Instituta za arheologiju u Zagrebu 8: 37–46. Lipovac Vrkljan, G. 2007. Ad Turres – Crikvenica: a ceramic manufacture of Sextus Metillius Maximus. Crikvenica: Grad Crikvenica. Lipovac Vrkljan, G. 2011. Lokalna keramičarska proizvodnja radionice Seksta Metilija Maksima u Crikvenici – crikvenička varijanta rimskih amfora ravnog dna, in: G. Lipovac Vrkljan, I. Radić Rossi and B. Šiljeg (eds) Rimske keramičarske i staklarske radionice. Proizvodnja i trgovina na jadranskom prostoru. Zbornik I. Međunarodnog arheološkog kolokvija Crikvenica, 23.-24. listopada 2008: 3–18. Zagreb – Crikvenica: Institut za arheologiju, Muzej Grada Crikvenice. Lipovac Vrkljan, G. and B. Šiljeg 2008. Exploration of the Igralište site in Crikvenica in 2007. Annales Instituti Archaeologici IV: 88–92. Lipovac Vrkljan, G. and A. Konestra 2018. Approaching the Roman economy of Province Dalmatia through pottery production – the Liburnia case study, in: G. Lipovac Vrkljan and A. Konestra (eds) Pottery Production, Landscape and Economy of Roman Dalmatia. Interdisciplinary approaches (Archaeopress Roman Archaeology 47): 14–36. Oxford: Archaeopress. Ljubić, Š. 1882. Arkeologičko izkapanje u Bakru (I). Viestnik hrvatskoga arheološkoga društva IV: 1–9. Ljubović, B. 2000. Senj u prapovijesti, antici i ranom srednjem vijeku. Senj: Gradski muzej Senj. 17

Martina Blečić Kavur Lo Schiavo, F. 1970. Il gruppo liburnico-japodico, per una definizione nell’ambito della protostoria balcanica. Atti della Accademia Nazionale dei Lincei VIII/14: 363‒524. Marović, I. 2006. Aucissa Fibulae with inscriptions kept in the Archaeological museums in Zagreb, Zadar and Split. Vjesnik za arheologiju i povijest dalmatinsku 99: 81–98. Matijašić, R. 2009. Città e territorio: aspetti della romanizzazione dell’Istria, in: G. Cuscito (ed.) Aspetti e problemi della romanizzazione. Venetia, Histria e arco alpino orientale: 383–400. Trieste: Editreg. Matijašić, R. 2017. Romanization of the Histri in the Early Roman Period, in: D. Demicheli (ed.) Illyrica Antiqua in Honorem Duje Rendić-Miočević. Proceedings of the International Conference, Šibenik, 12th-15th September 2013: 379–390. Zagreb: Faculty of Humanities and Social Sciences, University of Zagreb. Meller, H. 2002. Die Fibeln aus dem Reitia-Heiligtum von Este (Ausgrabungen 1880–1916) 1. Studien zu den Spätlateneformen (Studien zu vor- und frühgeschichtlichen Heiligtümern 2,1). Mainz: Nünnerich-Asmus. Meller, H. 2012. Die Fibeln aus dem Reitia-Heiligtum von Este (Ausgrabungen 1880–1916) 2. (Studien zu den Spätlatèneformen, Studien zu Vor-und Frühgeschichtlichen Heiligtümern 2,2). Mainz: Nünnerich-Asmus. Miškec, A. 2003. The early Romanization of the Southeastern Alpine region in the light of numismatic finds. Arheološki vestnik 54: 369–379. Mlinar, M. 2020. The 2000−2016 archaeological investigations on the left bank of the Idrijca (Opera Instituti Archaeologici Sloveniae 43). Ljubljana: ZRC SAZU Institute of Archaeology. Orlić, L. 2011. Željeznodobne fibule s nalazišta Četvrt Sv. Teodora u Puli. Histria Archaeologica 42: 185–215. Ožanić, I. 1998. Gradina Osječenica – antičko razdoblje. Opvscvla archaeologica 22: 27‒80. Ožanić Roguljić, I. 2012. Klasifikacija i tipologija rimske keramike s lokaliteta Crikvenica Igralište – proizvodi keramičarske radionice Seksta Metilija Maksima, unpublished PhD Dissertation. Zagreb: University of Zagreb. Ožanić Roguljić, I. 2016. Keramičko posuđe, in: A. Konestra and T. Rosić (eds) 845˚C. Ad Turres (Exhibition catalogue): 59‒69. Crikvenica: Muzej Grada Crikvenice. Peškař, I. 1972. Fibeln aus der römischen Kaiserzeit in Mähren. Praha: Academia Praha. Petković, S. 2010. Rimske fibule u Srbiji od I do V veka n.e. (Posebna izdanja, Knjiga 50). Beograd: Arheološki institut. Rieckhoff-Pauli, S. 1977. Die Fibeln aus dem römischen Vicus von Sulz am Neckar (Saalburg Jahrbuch XXXIV). Berlin: De Gruyter. Riha, E. 1994. Die römischen Fibeln aus Augst und Kaiseraugst (Forschungen in Augst 18). Augst: Römermuseum.

Rothe, U. 2012. Dress and cultural identity in the Roman Empire, in: M. Harlow (ed.) Dress and Identity (University of Birmingham IAA Interdisciplinary Series: Studies in Archaeology, History, Literature and Art 2): 59–68. Oxford: Archaeopress. Rothe, U. 2013. Die norisch-pannonische Tracht − gab es sie wirklich?, in: G. Grabherr, B. Kainrath and T. Schierl (eds) Verwandte in der Fremde. Fibeln und Bestandteile der Bekleidung als Mittel zur Rekonstruktion von interregionalem Austasch und zur Abgrenzung von Gruppen vom Ausgreifen Roms während des 1. Punischen Kriegs bis zum Ende des Weströmischen Reiches: 33–48. Innsbruck: Innsbruck University Press. Sagadin, M. 2020. Carnium – Kranj, in J. Horvat, I. Lazar and A. Gaspari (eds) Minor Roman settlements in Slovenia (Instituti Archaeologici Sloveniae 40): 201– 212. Ljubljana: ZRC SAZU Institute of Archaeology. Sahlins, M. 1999. Ekonomika kamene dobe. Rdeča zbirka. Ljubljana: Založba CF. Sedelmayer, H. 2017. Aucissafibeln – Typische Bestandteile der Frauentracht in einem Milieu geringer Romanisierung, in: E. Droberjar and B. Komoróczy (eds) Římské a germánské spony ve střední Europě (Archeologie barbarů 2012, Spisy archeologického ústavu Čr Brno 53): 129–142. Brno: Archeologický ústav AV ČR. Spânu, D. 2020. Practici depoziţionale non-funerare La Tène târziu în sud-vestul României. Descoperirile de la Glogova şi Desa. Materiale şi cercetări arheologice XVI: 87–117. Starac, A. 2000. Rimsko vladanje u Histriji i Liburniji. Liburnija, II. Monografije i katalozi 10/II. Pula: Arheološki muzej Istre. Šašel, J. 1992. Siscia, in Opera selecta (Situla 30): 600–620. Šeparović, T. 1998. Aucissa fibule s natpisom iz zbirke Muzeja hrvatskih arheoloških spomenika. Starohrvatska prosvjeta III/25: 177–187. Težak-Gregl, T. 1982. Rimske provincijalne fibule iz Arheološke zbirke Osor, in Ž. Rapanić (ed.) Arheološka istraživanja na otocima Cresu i Lošinju (Izdanja Hrvatskog arheološkog društva 7): 99–110. Zagreb: Hrvatsko arheološko društvo. Tilley, C. 1999. Metaphor and Material Culture. Oxford: Blackwell. Turk, P., D. Božič, J. Istenič, N. Osmuk and Ž. Šmit 2009. New pre-Roman inscriptions from western Slovenia: the archaeological evidence, in: G. Tiefengraber, B. Kavur and A. Gaspari (eds) Keltske študije II: papers in honour of Mitja Guštin (Protohistoire Europeenne 11): 47–64. Montagnac: Éditions Monique Mergoil. Versluys, M. J. 2014. Understanding objects in motion. An archaeological dialogue on Romanization. Archaeological Dialogues 21: 1–20. Višnjić, J. 2009a. Antički povijesni okvir, in: L. Bekić and N. Radić Štivić (eds) Tarsatički Principij. Kasnoantičko vojno zapovjedništvo: 25‒33. Rijeka: Grad Rijeka.

18

Inside fashion fusion: Fibulae from Crikvenica

Višnjić, J. 2009b. Antički metalni nalazi, in: L. Bekić and N. Radić Štivić (eds) Tarsatički Principij. Kasnoantičko vojno zapovjedništvo: 155‒172. Rijeka: Grad Rijeka. Vitelli Casella, M. 2018. La battaglia del Canale di Maltempo/Tihi kanal (49 a.C.) nel golfo del Quarnaro/Kvarner: note topografiche. Arheološki vestnik 69: 335–349.

Vulić, H. 2012. Fibule s lokaliteta Plandišta kod Cerića u kontekstu proizvodnje i uloge fibula u rimskom razdoblju. Acta musei Cibalensis 5: 95–133.

19

2.

The role of landscape features, waterways and seaborne routes in pottery production and distribution

Distribution possibilities and transport infrastructure in northern Dalmatia’s pottery workshops: evidence and interpretations Ana Konestra, Goranka Lipovac Vrkljan, Fabian Welc Abstract Transport infrastructure has been recognised as a fundamental feature in the choice of location for setup of craft activities. In fact, as part of the production cycle it allows resources to be brought in and products to be dispatched to the market. Thus, it was a key feature in the establishment of Roman pottery and ceramics workshops as well, securing, at least in rural settings, communication within the wider estate. This paper draws on results of recent research on pottery production sites in coastal and insular settings of the Roman province of Dalmatia with the aim to discuss availability and typology of transport infrastructure and better understand the mechanism of raw materials sourcing, production organisation and products marketing. On the bases of various evidence, including actual remains as well as other indirect data, and analogies with similar sites of the Roman world, the role of transport infrastructure and its features can shed light on not only on craft activities, but also on rural properties organisation and their role in the regional economy.

Key words Roman province of Dalmatia, pottery production, transport infrastructure, rural properties, regional economy.

Introduction1 The possibility of products, resources and raw materials transport is an aspect connected with the very setup of Roman pottery workshops (Mannoni and Giannichedda 1996: 220‒227; Martin 2006: 325; Goodchild 2013: 123; Sternini 2016: 147; 2019: 491; cf. also Arthur 2020: 10), their production organisation and the success of the distribution of theirs, but also of the wider estates’ products, the commercialisation of which often gave way, at least in rural contexts, to pottery production in the first place (Palazzo 1994; Santoro 2006: 173; Lewit 2013; Peña 2017: 215; Menchelli 2018: 25; cf. Corti in this volume). On the example of several production sites located on the northern coastal strip and islands of Roman province Dalmatia, the aspect of availability and typology of transport infrastructure will be tackled, in order to understand the mechanism of raw materials sourcing and products marketing. In fact, apart from the various natural and economical factors that influence location choice and production organisation (Poblome et al. 2002; Wilson 2012: 137‒138, 150), the possibility to use existing transport infrastructure or to create a dedicated one is crucial if the workshop is meant to market its products to a wider consumer area (Kiiskinen 2013; Arthur 2020: 10). Transport infrastructure, usually on a smaller scale, is also necessary to acquire especially some resources or raw materials needed for production, which are, such as wood, seldom available on site (eg. Wilson 2012: 149; Veal 2017), but it can also This paper stems from the activities carried out within project RED — Roman economy in Dalmatia: production, distribution and demand in the light of pottery workshops (2014-2018) financed by the Croatian Science Foundation (IP-11-2013-3973). 1 

in some cases reveal connections between individual production centres (Sternini 2019: 485). We will try to examine transport infrastructure in the case of pottery production centres established in province Dalmatia on the bases of different sets of evidence, that can broadly be defined as direct and indirect. In fact, the very distribution of products on consumer sites is in itself evidence of a movement of goods, and while telling us that an object produced in a certain location reached another, often leaves us guessing by which means this happened. On the other hand, finds of infrastructure remains, such as roads and docking facilities, but also shipwrecks, provide us with more tangible (direct) evidence of the methods of transportation and commercial routes. We can than combine this dedicated, purposely built infrastructure with routes of a more regional/public character, gaining further knowledge on both distribution and raw materials sourcing possibilities. Similarly, the very location of some of the identified workshops provides a glimpse in their distribution possibilities, even when no certain traces of transportation infrastructure have been established. Finally, accounting for marketing possibilities and the provisioning of infrastructure aimed at it can help explain the organisation and modes of production of each site as well as if and how it tried to approach a market flooded with foreign (in the timeframe in question mostly Italic) products. In this regard it should also be kept in mind that the identified infrastructure was, at least partly, at the disposal of the fundus within which the figlinae operated (Arnaud 2014: 164, 174‒175), thus providing us with further evidence

roman pottery and glass manufactures (Archaeopress 2022): 21–35

Ana Konestra, Goranka Lipovac Vrkljan, Fabian Welc on rural properties organisation and their role in regional economy.

the north Adriatic key emporium of Aquileia (Gaddi and Maggi 2017; Maggi 2018) (Figure 1).

The case studies

In the case of Crikvenica’s workshop, clear evidence of transport infrastructure is provided by the find of a road or even a road system (Lipovac Vrkljan et al. 2016; see also Welc et al. 2016) (Figure 2), developed within the workshop but extending also to the north of it, towards the Vinodol valley, where it could have joined the main road Aquileia-Salona crossing it in north-south direction (Figure 3). Also, the wider area of Crikvenica can be identified with the road station Ad Turres mentioned in several ancient itineraria (Suić 2003: 426435; Lipovac Vrkljan and Starac, R. 2014). In fact, the internal paths’ network identified in the central and western part of the figlina exited its perimeter through a passageway located in the northern enclosure wall (Figure 2). This communication network could have provided the means to transport products inland and along the via pubblica, but also raw materials, in particular firewood, into the production area (Goršić 2018). Within the workshop, it functioned as an internal paths’ system allowing communication between the various segments of the complex, in particular connecting the major kilns and the largest covered area (Figure 2), as detected in other workshops (cf. Vitali 2007: 23; Pallecchi 2017: 30‒31).

In recent years our knowledge on pottery production and figlinae in province Dalmatia has grown significantly, spanning a time-frame from the Hellenistic to at least the mid-Roman period (Lipovac Vrkljan et al. 2018). Several locations have been identified as certain production centres, while others are still determined on the bases of indirect evidence or archaeometry, nevertheless allowing a provisional grouping on the bases of production organisation and setting (Konestra and Lipovac Vrkljan 2018). Only one production centre, that of Sex. M(e/u)tillius Maximus at Crikvenica (NE Adriatic), has until recently been regarded as a large manufactory complex (Lipovac Vrkljan 2009; Konestra and Lipovac Vrkljan 2018: 133), but with new evidence from the site located in Plemići bay in the Ravni kotari area (northern Dalmatia) being gathered (Parica and Ilkić 2017; Lipovac Vrkljan and Konestra 2018; Welc 2018), we should rethink its scale and production organisation as it might be well comparable to that of Sextus’ figlina. For this reasons, but also due to the presence of direct transport evidence on both sites, we will concentrate on these two case studies, nevertheless tackling indirect evidence from minor production centres as well. Consequently, we will be dealing with production centres located in rural settings.

On the other hand, sea transport in Crikvenica’s case is far less directly evident. We are, on the one hand, aided by underwater finds of both shipwrecks with tegulae produced at the figlina (Radić Rossi 2011)2 as well as a series of underwater finds of amphorae (Figure 1), mostly of the Crikvenica Type 13 (see also Romanović 2017; Zubin Ferri et al. 2019). Equally indicative of the wide use of sea routes are the finds of the figlina’s products on island sites. These are clear, though indirect indicators of seaborne travel taken by the material produced at Sextus’ workshop, but also of the goods supplied from his saltus. On the other hand, finds such as ship plating nails, underwater finds in Crikvenica’s modern harbour (Konestra 2016: 118, cat. n. 106-108; Lipovac Vrkljan 2016: 91, n. 29) and accounts of finds recorded during infrastructure building in the area of the site, help us reconstruct a picture of a more direct connection to the sea. In particular the latter accounts, provided by A. Dračić, who based on them his reconstruction of the workshop and its connection to the sea (Dračić 1991) could point us into the right direction. Although dedicated studies to landscape reconstruction around the site of the figlina have not yet been performed, and are severely

Crikvenica’s figlina The figlina established at Crikvenica (NE Adriatic, Kvarner gulf) is located at the mouth of the Dubračina river, exploiting the silty sediments deposited by the Dubračina river course (Benac et al. 1993; Crmarić et al. 2007) and probably the resources-rich hinterland of the Vindol valley, where the extension of the owner’s saltus should be sought (Lipovac Vrkljan and Konestra 2018: 26‒30). Production, with its largest output probably to be placed within the 1st and possibly part of the 2nd century AD (Konestra and Lipovac Vrkljan 2018: 133) included amphorae, ceramic building materials, a large array of common ware shapes and implements, i.e. loom-weights (Lipovac Vrkljan 2009; Ožanić Roguljić 2012; Konestra et al. 2020). While the established distribution of Crikvenica’s products is still being enriched with new finds, we can propose an area spanning from Tarstatica in the north to Šibenik, or rather Scardona in the south, including both the coastal hinterland and the islands of the region Liburnia (Lipovac Vrkljan, Ožanić Roguljić 2013; Borzić et al. 2018; Konestra et al. 2020). A significant exception are the recently established finds of Crikvenica produced amphorae (certainly Type 1) in

These have been recognised thanks to the tile stamp De Salt. Sex. Mtll. Max. but also through archaeometry (Zubin Ferri et al. 2019). 3  Type 1 amphorae (Lipovac Vrkljan 2011) are typologically very peculiar vessels with a unique rim shape allows for easy identification, which aided in the recognition of such a wide distribution, as opposed to less typologically peculiar types produced by the figlina, such as Dr. 2-4 or others. 2 

22

Distribution possibilities and transport infrastructure in northern Dalmatia’s pottery workshops

Figure 1. Product distribution of the figlina at Crikvenica (Map: A. Konestra, data: Project RED GIS, basemap: EU DEM and ICES).

hindered by recent town’s development, geology and hydrological data on the high sedimentation of the Dubračina river (Benac et al. 1992; Crmarić et al. 2007), are informative on the changes that both the shoreline and Dubračina river estuary went through in the past. Two possible scenarios thus emerge: one which includes the navigability of the river — supposed again on written accounts mentioning it (Dračić 1991) — with a river harbour perhaps set up not far from the figlina’s location combined with an external dock located in the area of the modern harbour, the latter testified by underwater finds (Lipovac Vrkljan 2016); the second supposes a much deeper bay, on whose shores the figlina was erected. Available data collected so far seem to favour the second of the presented possibilities. When considering the topography of the region

during the Roman period, we must take into account relative changes in sea level and the accumulation rate of sediments deposited by the Dubračina river. Research of tidal notches created ca. 2000 years ago along the Croatian coast seem to prove at least 0,5 m lower sea level comparing to the modern times (Surić 2009; Faivre and Fouache 2003; Fouache et al. 2000, 2004). Location of submerged roman architecture on the Adriatic also indicate much lower sea level in the Roman period. Correlation of location of tidal notches with archaeological remains (mostly fishponds and harbors) preserved on the seashore in Istria and the Kvarner region, proved that in Roman times sea level was lower ca. 0.6-0.5 m (Surić 2009; Faivre and Fouache 2003; Fouache et al. 2000; 2004). On the other hand, submerged archaeological sites dated in the 1st - 2nd 23

Ana Konestra, Goranka Lipovac Vrkljan, Fabian Welc

Figure 2. Layout of the Crikvenica workshop based on excavation, magnetic survey and data from Dračić 1991 (Image: A. Konestra, F. Welc).

c. in the region of Zadar point to a sea level ca. 1.5 m lower than today (Faivre and Fouache 2003). According to Faivre and Fouache (2003) and Fouache et al. (2005), in this case we should also take into account regional tectonic activity which has caused coastline changes in this region. Geomorphological studies performed on the northern Adriatic coast (Italy, Slovenia and Croatia) (Antonioli et al. 2007) indicated that ca. 2000 years ago relative sea level was up to 2,5 m below the modern one. Taking into account all these estimates, it can be argued that the roman topography of the Dubračina river estuary in Crikvenica was much different than it is today. We can assume that the river valley was at that time more incised and most probably the pottery workshop itself was located on the relatively high and wide terrace (ca. 100 m) directly attached to the river course. This terrace could also protect the workshop

from periodical violent flooding. Because sediment transported by river is composed mostly by sandy and silty - clay material originated form Paleogene flysch and carbonate rock units (Šušnjar et al. 1970), sediment rate of the Dubračina river can be estimated from several centimeters to dozens of centimeters per year, especially during humid intervals, like in post-Antique times (Welc 2019). In result, the mouth of the Dubračina is presently filled with sediments and most probably shifted some distance towards the sea (Benac et al.,1992). The location of the Roman necropolis (Brunšmid 1901: 8-9) and recent land finds of a Roman pottery deposit4 provide us with information on the widest possible extent of the ancient river’s mouth or bay (Lipovac We thank Tea Rosić, curator of the Crikvenica Town museum for the information. The deposit includes La. 2 amphorae and black-gloss fine wares, thus indicating a (late?) 1st c. BC-early 1st c. AD dating. 4 

24

Distribution possibilities and transport infrastructure in northern Dalmatia’s pottery workshops

Figure 3. The wider area of the Vinodol valley and its transport infrastructure (Map: A. Konestra, data: Project RED GIS).

Vrkljan and Konestra 2018: fig. 10). The port of nearby ancient Tarsatica, though never archaeologically documented, is also supposed in the estuary of Riječina river which was used for docking throughout the Middle ages (Karković Takalić et al. 2014 with earlier bibliography) and is subject to similar sedimentation processes (Benac et al. 1992), thus providing us with a regional analogy. In the context of the saltus of which the workshop in Crikvenica was part, the bay of nearby Selce should also be taken in consideration for mooring purposes (Lipovac Vrkljan and Konestra 2018: 25-27, fig. 10) (Figure 2).

amphorae of a very recognizable type, but also loomweights and possibly common wares (Bekić and Pešić 2015; Parica and Ilkić 2017). Scanty data puts life on the site within the 1st c. AD, but it is indicative that so far there is no evidence of its frequentation in later centuries. Recently published finds from Plemići bay could also identify it as the production centre of roof tiles baring two different stamps: Muttieni and Ex of(ficina) L(ucii) Tetti Desedes (Juras and Jurković Pešić 2016: 60-62; Ilkić and Parica 2017). The find of both tegulae stamps at Plemići, but also on tegulae found within distribution and having very similar characteristic to the mass of stampless wasters, allows us to propose Plemići bay as the production site of both. The analysis of published tegulae stamps allows to reconstruct the distribution network of certain and supposed consumer sites for the two stamps (Juras and Jurković Pešić 2016) (Figure 4). It is interesting to note that on certain site the two stamps occur in relation, while at others with several examples (Colnago and Kiel 1905 (2008): 136 and n. 12‒13; Juras and Jurković Pešić 2016: 60‒62).

Plemići bay The location of the pottery production centre has been established along the southern coast of Plemići bay (nowadays Ražanac Municipality), located to the north of Roman Aenona, at the western outskirts of the fertile Ravni kotar plains in the hinterland of Zadar (Iader). The vast amount of ceramic wasters (discards, overfired objects) and kiln fragments, as well as the stratigraphy of the coastal profiles (Welc 2019), allow us to place without doubt a figlina at this location (Lipovac Vrkljan and Konestra 2018; Welc 2018). The aforementioned wasters generally belong to ceramic building materials,

Regarded as setup within a large rural property, the figlina (or some of its parts) might at a certain 25

Ana Konestra, Goranka Lipovac Vrkljan, Fabian Welc

Figure 4. Distribution of the MVTTIEN and Ex. Of. L. Titti. Desedes stamps (Map: A. Konestra, data: Project RED GIS, basemap: EU DEM and JU ZZPU PGŽ).

point have been subcontracted, which would explain the two different stamps (Lipovac Vrkljan and Konestra 2018; Konestra et al. 2021), although more data is necessary, especially to frame this evidence chronologically.

vessels to mourn and envisaging a dedicated export infrastructure at the site (Parica and Ilkić 2018: 74). To this, the identification, by remote sensing, of a large building on the shore of the bay, and which might be interpreted as a warehouse (horreum), can be added (Welc 2018; Lipovac Vrkljan et al. 2019: 196‒197).

While at Crikvenica we still face the problem of locating possible harbour infrastructure and define its type, at Plemići, thanks to underwater research undertaken within the bay, the evidence is much clearer (Parica and Ilkić 2017; 2018). In fact, a harbour composed of two piers has been discovered exactly in front of the workshop’s location, while finds associated with them clear any doubt on their contemporaneity with the evidence on the shore (Figure 5). Moreover, it has been proposed that the two piers might have had different purposes, allowing different types of

As for the availability of terrestrial transport infrastructure, the wider area lacks precise data on the Roman roads network, although some is available for Aenona, Cvijina gradina as well as for Argyruntum (Colnago and Kiel 1905 (2008): 137-138; Ilakovac 199596 with earlier bibliography; Dublonić 2007), a road leading to Aenona supposedly passed along the nearby rural site Glavičine - Podvršje (Uglešić 2017: 664), just as a N-S road traversing Iader’s hinterland (Chapman et al. (eds) 1996: 82, 278, fig. 52, 177). 26

Distribution possibilities and transport infrastructure in northern Dalmatia’s pottery workshops

Figure 5. Layout of the site at Plemići bay based on visible architecture, underwater survey and remote sensing (Image: A. Konestra, drawings: M. Grgurić, F. Welc).

transports, as expensive as usually thought (Sternini 2016: 152).

Transport modes and production organisation The aforementioned data, while being enlightening on the different sites and locations on which the products of both workshops can be found, opens a series of questions on the methods of transport and marketing of the products of the figlinae and those of the fundi they were inserted within. What emerges is that both sites (see also infra) had the availability of both maritime and road infrastructure, although it is still too early to tell on which and if any, they mostly relied. In fact, many studies have tried to establish the cost of transport methods in Roman times (e.g. Scheidel 2014 with earlier bibliography), and while waterway transportation is always deemed as the most convenient, especially for long distance transport (see Horden and Purcell 2000: 151; Kolb 2011-2012: 57; Leitch 2011; Kiiskinen 2013: 70-71), we will see how, although our case studies are all located on the shore or not far from it, we should take into consideration a combination of transport methods (Sternini 2019: 485, 491 passim), i.e. an integrated transport system (Adams 2012: 225, 231) or mixed-mode transport (Kiiskinen 2013: 130‒131), allowing us to better understand production and distribution organisation within the northern Dalmatian micro-region. In favour of the mix-mode method is the dense network of roads that characterises the region, especially its southern part, i.e. the Ravni kotari (Suić 1996: 441; Miletić 2004 with earlier bibliography). In fact, it is exclusively by road that products could reach the internal areas of both Lika and Ravni Kotari (Borzić et al. 2018: fig. 1), which should not have been, thanks to the mixed-mode of

When observing the reconstructed sea routes, both regional, internal to the Kvarner gulf (Matijašić 2001; Čaušević Bully, Bully 2012: fig. 1), and their connection to the main Adriatic N-S route (Vrsalović 2011: map 3), it is evident that the marketers of Crikvenica’s goods made very good use of the possibilities offered by the sea. We can, in fact, on the bases of its distribution, suppose further tramping and cabotage5 connections especially within the Vinodol channel, usually deemed as very dangerous for shipping (hence its Italian name Mal tempo). Similarly, some consumer sites enlighten us on the infrastructure available for products redistribution, such as the harbour infrastructure at Fulfinum on the island of Krk (Faber 1981; Zelić 1993: 8-9), where plenty of Crikvenica pottery has been found (Konestra 2014; 2015), as well as the harbour of Senia, which could have been used, alternatively to the road, as the main distribution hub for inland traffic (Glavaš 2010) and a regional market place. Another redistributive hub of a certain importance must have been that of Zaton - Aenona’s main harbour (Romanović 2017) (Figure 4), and on the bases of recently established evidence, the coastal town of Argyruntum, and Iader (Borzić et al. 2018: 50, 53, fig. 1). Even more indicative of the wide use of sea routes are the finds of the figlina’s products on islands other than Krk, i.e. Rab and Ugljan (Jurjević 2011: 63-69). For tramping and cabotage see Arnaud 2012. For evidence of pottery from Crikvenica see Borzić et al. in this volume. 5 

27

Ana Konestra, Goranka Lipovac Vrkljan, Fabian Welc Having in mind the distribution map which indicates numerous findspots well inland in today’s Vinodol and Lika (Figure 1), we cannot ignore the possibility of land transport provided by the via publica Aquileia - Salona, which did pass along Ad Turres, perhaps modern day Crikvenica or a wider area in its environs (Lipovac Vrkljan and Starac 2014), but which could have also been accessed through Senia, major port and seat of the publicum portorium Illyrici (Glavičić 1994; Glavaš 2010). Also, the road system of public and local roads allowed the arrival of products in the hinterland of northern Dalmatia, with all probability through a network of harbour-road connections (Miletić 2004; Brusić 2007).

the east and today filled by eroded land sediments, thus being very shallow, while preliminary geoarchaeological research allows to propose a somewhat different appearance of both the bay and its hinterland in antiquity (Welc 2018: 63). Diverse evidence, such as the presence of allochthonous building materials (limestone) at the site, but also other imports as well as the ‘logic’ of the location, allow to suppose some kind of docking facility within the bay. The size of kilns, the mono-production of only CBM and the absence of stamps, might indicate that the setup of these kilns was intended to fulfil the needs of the estate and only to a minor extent to be marketed, though further research on the island’s CMB assemblages is need. Similarly, we might propose a maritime product’s shipment for the nearby kiln site of Mahućine bay (Lipovac Vrkljan and Konestra 2018), located in an isolated bay on mainland Rab, with no viable land connection, but possibly producing for the villa at Zidine or a possible site at Lopar’s harbour (Konestra et al. 2021). A recently proposed production location, that at Gonar, on the SW part of the island, might support this model of estate production sites scattered along the island.

In the case of Plemići and its environs we can suppose the same mixed-mode method of distribution, for example at Nin, through its port in Zaton and then via road (Brusić1968: T. 1; Ilakovac 1995-96). Via a system of local roads, supposed on the bases of the dense network of sites (Chapman et al. (eds) 1996: 263-264) and the fairly plain terrain, Plemići’s products could have arrived to Cvijina gradina, while direct maritime connections can be supposed for finds at Iader and Argyruntum (Figure 4).

Perhaps a final note is possible on the Smrdelje site, in the hinterland of roman Scardona in northern Dalmatia, connected to tile production of the legionary camp Burnum (for a summary and most recent data on production indicators see Borzić 2020). The trading relationship of Scardona, important port on the mouth of the river Krka (Titius), and Burnum has been stressed several times just as that of the road network within which the camp operated (and its legions helped build) and through which it was connected to Scardona (Miletić 1991/92; 2011: 264, 267; Borzić 2007: 177; 2020). Smrdelje, located half-way between the two sites and within the territory under military jurisdiction (Čače 2012: 29-32 with earlier bibliography), must have then used land traffic to bring its products to the legionary camp, but also further, as its tiles are known from several Dalmatian hinterland sites (Miletić 2011: 269, 271).

While latter data remains very speculative, evidencing both land and marine movements, the aforementioned more direct evidence of transport infrastructure discovered during research at both case-study sites provides firmer grounds for discussion on the local level. These two examples provide us with evidence on the endeavours that were undertaken in production organisation and facility setup required by workshops of these size and complexity. They also point us in the direction that these complexes were not built to fulfil local needs, but that wider distribution was already planned at production setup. A fact that can be grasped from the variety of their products assortment, but that gains visibility and scale when these kind of evidence is analysed, indicating that they were, in both cases, established on locations where, among others, easy access to the local and main regional transport routes was possible.

Pottery production and the regional market

On the other hand, those sites where a mix-mode of communication was not available, seem to present a somewhat different production organisation. Among these we might regard the site of Podšilo bay on Rab island (Lopar Peninsula), where a system of kilns has been identified associated with a Roman rural site (Lipovac Vrkljan and Šiljeg 2012; Konestra et al. 2019; Welc et al. 2019) 6. The site is located within a bay open to

Such data, apart from providing us with a better understanding of the local situation(s) allows us to compare it with production centres throughout the Roman world. Both peri-urban and sub-urban production centres were often carefully located in vicinity of transport arteries, being that waterways or roads (Sternini 2019), leading sometimes to the creation of artisanal quarters, in certain regions connected to the elevation of the urban settlements to municipia, and thus to a definitive urban organisation of the settlements (Mondin 2017). A good example is that of the regio X, where at Padua kiln sites

A road connecting Lopar to Rab town must have existed, and its ‘memory’ is preserved in the local populations’ naming the old road connecting the two settlements through the Fruga plateau as ‘Roman road’, although no remains have been recorded thus far. A system of rural roads or paths can be reconstructed on the bases on sites’ distribution within the island’s fields of Mundanijsko and Kamporsko polje (Konestra et al. 2016: 107). 6 

28

Distribution possibilities and transport infrastructure in northern Dalmatia’s pottery workshops

were located both close to roads and along the river, which provided water and transportation possibilities, while other craft activities have been noted to follow the same logic, a situation present elsewhere in Cisalpina (Santoro 2006: 173; Cipriano and Mazzocchin 2017: 149‒150; Mondin 2017: 47). Similarly, at Altino the new peri-urban pottery production quarter followed the course of the Sioncello in close proximity to the Via Annia, and at Verona smaller workshops and figlinae dotted the area along the via Postumia (Mondin 2017: 48‒49).

other production centres of the ager Cosanus and the Albegna valley seem to favour proximity to transport infrastructure and waterways, rather than that to the villae producing wine (Vitali et al. 2005). Apart from the sub-urban production centre at Arezzo, central Italian sigillata production is connected to several rural centres, whose connectivity has also been tested (Kiiskinen 2013; Sternini 2016: 145; 2019). On the other hand, in the hinterland of Aquileia several rural kiln sites are located in proximity of the main and regional viability network (Degrassi and Maselli Scotti 2011; Buora and Magnani 2011), just as were several workshops in the ager of Tergeste localised in relation to the via Flavia (Žerjal 2011), in both cases producing mainly tiles. Although the distribution pattern of the Dr. 6B amphorae produced within the villa at Loron in the ager of Parentium (W Istria) (Rousse 2011; Marchiori and D’Incà 2011 with earlier bibliography) speaks of it shipping by sea (Marion and Starac 2001), the network of roads following the centuriation pattern of this area (Tassaux 2012) indicates the possibility of mixmode connections with its immediate environs. We could also add here some new evidence provided by the Fažana amphorae workshop (W Istria), property of the Laecanii and later Imperial, that could illustrate even more the importance of communication networks not only in distribution but also in raw materials sourcing, in this case proposed to have been brought to the figlina from its owner’s saltus, located more than 50km to the north and inland (Szakmány et al. 2017). Further examples from Gaul and Britain, and different production sites (Buckland et al. 2001; Martin 2006: 324325; Goodchild 2013), further evidence the relationship of communication networks and figlinae with larger commercial output. Interesting comparisons on products marketing strategies can be gained from other crafts as well, such as the organisation of transport within the stone cutting sites on northern and central Dalmatian islands, i.e. Zadar and Split area (Parica 2012; 2018). Harbour structures or simple piers have in fact been detected on several island sites, functioning in relation to quarries often located far afield in the island’s hinterland. Cut stone was transported from the quarry by road/ramps built within minimally adapted natural gullies, extending towards the shore and sometimes ending within the very harbour (Parica 2012: 351; 2018: 943, 948). These roads/ramps could have been simple paths or rock-cut passageways with paved or merely adapted walking surfaces (Parica 2018: 943, 945). Quarries’ harbours and lowering ramps thus indicate the provisioning of dedicated transport infrastructure and a mixed-mode of transport, as early as the 1st c. AD (Parica 2018: 948). The harbours were, in some cases, the only infrastructure available for other logistic operations (e.g. incoming goods) related to the quarry sites, often being the location of stonemason workshops (Parica 2012: 352).

A key example is that of Arezzo and its many production centres located at the outskirts and to a certain distance from the town, but along roads and/ or in proximity of rivers (Fülle 1997: 145; Goodchild 2013: 123‒124; Sternini 2019), while its distribution possibilities, and that of the Pisa workshop, have been thoroughly tested revealing the importance of roads for regional marketing, but also the relying on a system of mixed transportation with rivers playing a central role (Kiiskinen 2013; Sternini 2019). Moreover, Pisae used this system to acquire goods from its hinterland to then ship by sea, but also mediating in the opposite direction and producing, with town and rural areas becoming part of “the same production processes” (Menchelli 2018: 24). Further evidencing the relationship between production and transport is the case of ancient Sagalassos (Turkey), where the boom of fine ware production within its Potter’s quarters has been linked, among others, to the establishment of larger urban settlements in the region and the road system connecting them, but also to its relationship with the port of Perge (Poblome 1996; Waelkens et al. 1997: 246‒247). Here, nevertheless, the road system was also key for providing the best clays for fine ware production (Poblome et al. 2002: 880). The relationship between pottery production and the transport network is perhaps even more evident in the case of workshops in rural settings, where the organisation of production can be planned more freely, without impediments that a peri/sub-urban area must have had (cf. Mondin 2017), and thus better accommodate to the needs of the producer. In fact, from just a few examples, we can gather that the system of resource availability and road connections is followed in many, sometimes disparate cases, as well as that of a combination of transport system including road-river-sea. The resources acquisition/ marketing mixed-system of communications is evident in the organisation proposed for the fundus of Visellio at Giancola (Manacorda and Pallecchi (eds) 2012) in Apulia, and within the very workshop (Pallecchi 2017: 30‒31), while similarly the Albegna river and the via Aurelia Antica enclose the figlina at Albinia (Vitali 2007) at close proximity from north and west, and also 29

Ana Konestra, Goranka Lipovac Vrkljan, Fabian Welc The importance of a transport system allowing resources to be brought in and commodities to be sent out, thus becoming part of the production cycle (Mannoni and Giannichedda 1996: 226), has been emphasised by scholars dealing with craft production (Peña 2017: 216), and particularly ceramics (e.g. Menchelli 2018; Sternini 2019). Nevertheless, in the case of Roman Dalmatia, where pottery production and its setting characteristics have only recently become a topic of scientific interest and discussion, it opens a set of new questions on production organisation, communication networks, and ultimately on trade and markets existing within the province. In fact, while the interdependence of figlinae, therefore also of the properties they were inserted in, and transport infrastructure is evident, we still lack information on the actors involved in the distribution of the products, i.e. whether they were agents of the fundi or detached merchants (Poblome 2006: 357‒358; Sternini 2016: 151, 155) operating within the NE Adriatic region. Certainly, a complex property such as the saltus of Sextus M(e/u)tilius Maximus, producing both foodstuffs and ceramics, must have had a system of distribution that did not rely solely on the efforts of its owners or managers. If a similar organisational complexity is to be read from the tile stamps from Plemići bay, it might additionally support a market oriented production, for which its substantial harbour seems more fitting then for just fulfilling the needs of the estate.

References Adams, C. 2012. Transport, in W. Scheidel (ed.), The Cambridge Companion to the Roman Economy (Cambridge Companions to the Ancient World): 218‒240. Cambridge: Cambridge University Press.

Antonioli, F., M. Antidei, L. Lambeck, R. Auriemma, D. Gaddi, S. Furlani, P. Orrù, E. Solinas, A. Gaspari, S. Karinja, V. Kovačić and L. Surace 2007. Sea-level change during the Holocene in Sardinia and in the northeastern Adriatic (central Mediterranean Sea) from archaeological and geomorphological data. Quaternary Science Reviews 26: 2463‒2486. Arnaud, P. 2012. Ancient sailing routes and trade patterns: the impact of human factors, in D. Robinson and A. Wilson (eds) Maritime Archaeology and Ancient Trade in the Mediterranean (Oxford Centre for Maritime Archaeology: Monograph 6): 61‒80. Oxford: Oxford Centre for Maritime Archaeology, University of Oxford. Arnaud, P. 2014. Maritime Infrastructure Between Public and Private Initiative, in A. Kolb (ed.) Infrastruktur und Herrschaftsorganisation im Imperium Romanum, Herrschaftsstrukturen und Herrschaftspraxis III, Akten der Tagung in Zürich 19.–20. 10. 2012: 161‒179. Berlin: Akademie Verlag GmbH, De Gruyter. Arnaud, P. 2016. Cities and Maritime Trade under the Roman Empire, in C. Schäfer (ed.) Connecting the Ancient World. Mediterranean Shipping, Maritime Networks and their Impact (Parhos. Studien zur griechisch-römischen Antike, Band 35): 117‒173. Rahden/Westf.: Verlag Marie Leidorf GmbH. Arthur, P. 2020. The Changing Geography of Artefact Production in the Late Antique and Early Byzantine Mediterranean, in S. Cosentino (ed.) Ravenna and the Traditions of Late Antique and Early Byzantine Craftsmanship. Labour, Culture, and the Economy (Millennium-Studien / Millennium Studies 85): 5‒32. Berlin: De Gruyter. Bekić, L. and M. Pešić 2015. Evidence of maritime trade on the eastern Adriatic coast based on underwater finds from two Roman ports in Croatia. Histria archaeologica 45: 95‒112. Benac, Č., Ž. Arbanas and B. Jardas 1992. The morphogenesis and the evolution of the river mouths in the Kvarner area, in A. Bognar (ed.) Proceedings of the international symposium ‘Geomorphology and sea’ and the Meeting of the geomorphologic commission of the carpatho-balcan countries, Mali Lošinj, September 22-26, 1996: 37‒45. Zagreb: Department of geography, Faculty of science, University of Zagreb. Borzić, I. 2007. Geografski položaj Burnuma na razmeđi liburnskog i delmatskog područja, in D. Marguš (ed.) Simpozij ‘Rijeka Krka i Nacionalni park ‘Krka’. Prirodna i kulturna baština, zaštita i održivi razvitak’. Zbornik radova: 163‒179. Šibenik: Nacionalni park Krka.

Finally, what emerges from products’ distribution is a regional NE Adriatic - Liburnian7 - market (cf. Reger 2011) that relied on and made good use of the local road network, while criss-crossing the sea not only on the main S-N route but especially along regional and local routes, connecting the coast and the islands in a dynamic exchange system using a mixed-mode of transport. Within it, a northern and a southern local market might perhaps emerge, based on the preliminary evidence from Plemići, and that of a major concentration of finds in Crikvenica’s environs (Figures 1, 4). The Liburnian towns, both large and smaller ones, while being consumers8, emerge as hubs for the redistribution of local products (cf. Poblome 2006: 361; Van Oyen 2015: 288‒291; Arnaud 2016: 118‒119, 166; Sternini 2016: 147, 153; Menchelli 2018). The system certainly functioned in relation to different types of trade networks (Papaioannou 2011: 200) evident from the numerous imports, with local pottery production shedding light on the regional commercial one. Liburnia as a region emerges from historic sources which provide us with geographic boundaries (rivers Arsia and Titius), and based on ethnic factors that emerged at a certain point during late pre-Roman times, but also by later political boundaries determined by the conventus Scardonitanus which, though, sees them grouped with the Japodes (Starac, A. 2000: 7-8, 13-14, 57-61; Čače 2012; 2013 with earlier bibliography). 8  So far no evidence of urban pottery production has been established in the region. 7 

30

Distribution possibilities and transport infrastructure in northern Dalmatia’s pottery workshops

Borzić 2020. Keramički nalazi iz legijskog logora Burnum. Šibenik-Puljane: Nacionlani park Krka. Borzić, I., M. Čelhar, G. Zaro and V. Glavaš 2018. New finds of Crikvenica 1-type amphorae in the area of southern Liburnia, in G. Lipovac Vrkljan and A. Konestra (eds) Pottery Production, Landscape and Economy of Roman Dalmatia. Interdisciplinary approaches: 50‒56. Oxford: Archaeopress. Brusić, Z. 2007. The port of Pakoštane and other coastal Liburnian settlements in the Pašman channel in relation to hill-fort settlements in the hinterland and Asseria. ASSERIA 5: 11‒38. Brusić, Z. 1968. Istraživanje antičke luke kod Nina. Diadora 4: 203‒210. Buckland, P. C., K. F. Hartley, and V. Rigby 2001. The Roman Pottery Kilns and Rossington Bridge. Excavations 1956-1961 (Journal of Roman Pottery Studies, Volume 9). Oxford: Oxbow Books. Čače, S. 2012. Secus flumen Titium: on boundaries and changes along the river Krka before and at the beginning of Principate. Miscellanea Hadriatica et Mediterranea I: 15‒35. Čače, S. 2013. Notes on the relations between the Liburnian communities. ASSERIA 11: 11‒50. Chapman, J., R. Shiel and Š. Batović 1996 (eds). The Changing Face of Dalmatia. Archaeological and Ecological Studies in a Mediterranean Landscape. London: Leicester University Press, The Society of Antiquaries of London. Cipriano, S. and S. Mazzocchin 2017. Le aree artigianali e produttive di Padova romana: prima mappatura, in J. Turchetto and M. Asolati (eds) Paesaggi in movimento. Ricerche dedicate a Guido Rosada: 139‒155. Padova: Padova University Press. Crmarić, R., M. Juračić, Č. Benac and I. Ružić 2007. Sedimentacija na ušću Dubračine i novljanske Ričine, Vinodolski kanal, in Gereš, D. (ed.) 4. Hrvatska konferencija o vodama. Hrvatske vode i Europska unija — izazovi i mogućnosti. Zbornik radova: 297‒302. Opatija: Hrvatske vode. Dračić, A. 1991. Naselje na ušću Dubračine od II do VI vijeka, Vinodolski zbornik 6: 235‒247. Faber, A. 1981. Luka antičkog Fulfina na otoku Krku. Pomorski zbornik 19: 293‒314. Faivre, S. and E. Fouache 2003. Some tectonic influences on the Croatian shoreline evolution in the last 2000 years. Zeitschrift für Geomorphologie, N. F., 47/4: 521‒537. Fouache, E., S. Faivre, J.J Dufaure, V. Kovačić, F. Tassaux and P. Tronche, P. 2004. Morska razina u rimsko doba na području Istre. Vjesnik arheološkog muzeja u Zagrebu, 3.s., XXXVII: 173‒190. Fouache, E., S. Faivre, J. J. Dufaure and V. Kovačić 2000. New observations on the evolution of the Croatian shoreline between Poreč and Zadar over the past 2000 years, Zeitschrift für Geomorphologie, N. F., suppl.Bd. 122: 33‒46.

Fülle, G. 1997. The Internal Organization of the Arretine Terra Sigillata Industry: Problems of Evidence and Interpretation. The Journal of Roman Studies 87: 111‒155. Gaddi, D. and P. Maggi 2017. Anfore italiche, in P. Maggi. F. Maselli Scotti, S. Pesavento Mattioli and E.Zulini (eds) Materiali per Aquileia. Lo scavo di Canale Anfora (2004-2005): 263‒328. Trieste: Editreg. Glavaš, V. 2010. Prometno i strateško značenje prijevoja Vratnik u antici. Senjski zbornik 37: 5‒18. Glavičić, M. 1994. Značenje Senije tijekom antike. Senjski zbornik 21: 41‒57. Goodchild, P. J. 2013. The production centres: settlement hierarchies and spatial distribution, in M. Fulford and E. Durham (eds) Seeing Red. New economic and social perspectives on Gallo-Roman terra sigillata (Bulletin of the Institute of classical studies supplement 102): 121‒136. London: Institute of classical studies, School of advanced study, University of London. Goršić, E. 2018. Anthracology and wood analysis from the pottery workshop sites of Crikvenica – Igralište and Plemići Bay: first results of genus identification, in G. Lipovac Vrkljan and A. Konestra (eds) Pottery Production, Landscape and Economy of Roman Dalmatia Interdisciplinary approaches: 76‒80. Oxford: Archaeopress. Horden, P. and N. Purcell 2000. The Corrupting Sea. A study of Mediterranean history. Oxford - Maldon (MA) - Carlton: Blackwell. Ilakovac, B. 1995-96. Postanak i razvoj ninskih (Aenona) mostova. Vjesnik Arheološkog muzeja u Zagrebu s. 3/ XVIII-XXIX: 73‒95. Juras, I. and F. Jurković Pešić 2016. Stamped tegulae from the Roman antiquity collection of the Archaeological museum Zadar. Diadora 30: 31‒75. Jurjević, M. 2011. Krovne opeke (tegulae) s pečatom iz Preka na otoku Ugljanu. Diadora 24: 57‒80. Karković Takalić, P., P. Predoević Zadković and J. Višnjić 2014. The new archaeological map of Ancient and early Medieval Tarsatica. Histria archaeologica 45: 113‒136. Kolb, A. (2011-2012). The conception and practice of Roman rule: the example of transport infrastructure. Geographia Antiqua XX: 53‒69. Konestra, A. 2014. La ceramica dai corredi funerari della necropoli Kurilovo conservata nella collezione archeologica del JANAF, in G. Lipovac Vrkljan, B. Šiljeg, I. Ožanić Roguljić, A. Konestra (eds) Officine per la produzione di ceramica e vetro in epoca romana. Produzione e commercio nella regione adriatica. Tema del colloquio: Archeologia sperimentale. Atti del II Colloquio Archeologico Internazionale, Crikvenica (Croazia), 2829 ottobre 2011: 259‒277. Crikvenica: Muzej Grada Crikvenice, Institut za arheologiju. Konestra, A. 2015. Pottery from the Forum of Municipium Flavium Fulfinum (Krk Island, Croatia) – research between 2007 and 2013 / Keramika s Foruma Municipia Flavia Fulfinuma (otok Krk, 31

Ana Konestra, Goranka Lipovac Vrkljan, Fabian Welc Hrvatska) – istraživanja od 2007. do 2013. godine, Prilozi Instituta za arheologiju u Zagrebu 32: 147‒214. Konestra, A. 2016. Metalni nalazi, in A. Konestra and T. Rosić (eds) 845˚C. Ad Turres (exhibition catalogue): 118. Crikvenica: Muzej Grada Crikvenice. Konestra, A. and G. Lipovac Vrkljan 2018. Pottery workshops in the coastal area of Roman Dalmatia: landscape, spatial organization, ownership, in M. Janežic, B. Nadbath, T. Mulh and I. Žižek (eds) 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, Ptuj, 8th and 9th October 2015: 127‒141. Konestra, A., N. Šegvić, P. Androić Gračanin and R. Starac 2016. Arheološka topografija otoka Raba: geofizička, sondažna i topografska istraživanja u 2016. godini. Annales Instituti Archaeologici XIII:103–110. Konestra, A., F. Welc, A. Dugonjić, P. Androić Gračanin, K. Rabiega, R. Solecky and B. Nowacki 2019. Istraživanja projekta ‘Arheološka topografija otoka Raba’ u 2018. godini na području Lopara: nova saznanja o prapovijesnim i kasnoantičkim lokalitetima. Annales Instituti Archaeologici XV: 187‒194. Konestra, A., G. Lipovac Vrkljan and B. Šiljeg 2020. The assortment of ceramic building materials from the pottery workshop of Sextus Me(u)tillius Maximus at Crikvenica (Croatia), Prilozi Instituta za arheologiju u Zagrebu 37: 73‒99. Konestra, A., A. Kurilić and G. Lipovac Vrkljan 2021. Tiles and amphorae in the Roman province of Dalmatia: evidence of stamps, in D. Rigato, M. Mongardi and M.Vitelli Casella, Mattia (eds) Adriatlas 4. Produzioni artigianali in area adriatica: manufatti, ateliers e attori (III sec. a.C. – V sec. d.C.) (collection PrimaLun@ 8): 145‒166. Pessac: Ausonius éditions. Konestra, A., F. Welc and P. Androić Gračanin 2021. Lokalitet na rtu Zidine u Loparu u kontekstu obalnih rezidencijalno-gospodarskih kompleksa otoka Raba. Annales Instituti Archaeologici XVII: 152‒170. Leitch, V. 2011. Location, location, location: characterizing coastal and inland production and distribution of Roman African cooking wares, in D. Robinson and A. Wilson (eds) Maritime Archaeology and Ancient Trade in the Mediterranean (Oxford Centre for Maritime Archaeology: Monograph 6): 169‒195. Oxford: Oxford Centre for Maritime Archaeology, University of Oxford. Lewit, T. 2013. The mysterious case of La Graufesenque? Stimuli to large-scale fine pottery production and trade in the Roman Empire, in M. Fulford and E. Durham (eds) Seeing Red. New economic and social perspectives on Gallo-Roman terra sigillata (Bulletin of the Institute of classical studies supplement 102): 111‒120. London: Institute of classical studies, School of advanced study, University of London.

Lipovac Vrkljan, G. 2009. L’officina ceramica di Crikvenica, in S. Pesavento Mattioli and M.-B. Carre (eds) Olio e pesce in epoca romana: Produzione e commercio nelle regioni dell’alto Adriatico - Atti del Convegno (Padova, 16 Febraio 2007): 309‒314. Padova: Edizioni Quasar. Lipovac Vrkljan, G. 2016. Amfore, in A. Konestra and T. Rosić (eds) 845˚C. Ad Turres (exhibition catalogue): 91‒96. Crikvenica: Muzej Grada Crikvenice. Lipovac Vrkljan, G. and A. Konestra 2018. Approaching the Roman economy of Province Dalmatia through pottery production – the Liburnia case study, in G. Lipovac Vrkljan and A. Konestra (eds) Pottery Production, Landscape and Economy of Roman Dalmatia Interdisciplinary approaches: 14‒36. Oxford: Archaeopress. Lipovac Vrkljan, G. and B. Šiljeg 2012. Prilog antičkoj topografiji otoka Raba – rimska keramičarska peć na Loparu. Senjski zbornik 39: 5‒33. Lipovac Vrkljan, G. and I. Ožanić Roguljić 2013. Distribucija crikveničke keramike kao prilog poznavanju rimskog gospodarstva. Senjski zbornik 40: 255‒270. Lipovac Vrkljan, G. and R. Starac 2014. Antička mikrotopografija Hrvatskog primorja. Prilog poznavanju arheološkog krajolika Vinodola prema nalazima crikveničke keramike, 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 (ZIA 2): 93‒105. Crikvenica: Muzej Grada Crikvenice, Institut za arheologiju. Lipovac Vrkljan, G., A. Konestra and M. Ugarković 2018. Eastern Adriatic Hellenistic and Roman pottery and ceramics production sites, in G. Lipovac Vrkljan and A. Konestra (eds) Pottery Production, Landscape and Economy of Roman Dalmatia. Interdisciplinary approaches: 1‒6. Oxford: Archaeopress. Lipovac Vrkljan, G., A. Konestra, F. Welc, M. Ilkić and M. Parica 2019. Multidisciplinarni terenski radovi projekta RED u 2018. godini: istraživanja u uvali Plemići (Ražanac). Annales Instituti Archaeologici XV: 195–199. Maggi, P. 2018. Crikvenica amphorae from Canale Anfora: first evidence of Liburnian wine at Aquileia, in G. Lipovac Vrkljan and A. Konestra (eds) Pottery Production, Landscape and Economy of Roman Dalmatia Interdisciplinary approaches: 57‒61. Oxford: Archaeopress. Mannoni, T. and E. Giannichedda 1996. Archaeologia della produzione. Torino: Einaudi. Marchiori, A. and C. D’Incà 2011. Le fornaci di Loron (Istria, Croazia), in G. Lipovac Vrkljan, I. Radić Rossi and B. Šiljeg (eds) Rimske keramičarske i staklarske radionice. Proizvodnja i trgovina na Jadranskom prostoru. Zbornik I. Međunarodnog arheološkog kolokvija, Crikvenica, 23—24. listopada 2008: 83‒88. Crikvenica: Muzej Grada Crikvenice, Institut za arheologiju. 32

Distribution possibilities and transport infrastructure in northern Dalmatia’s pottery workshops

Marion, Y. and A. Starac 2001. Les amphores, in F. Tassaux, R. Matijašić and V. Kovačić (eds) Loron (Croatie). Un grand centre de production d’amphores à huile istriennes (Ier - IVe S. P. C.): 97‒126. Bordeaux: Institut Ausonius. Martin, T. 2006. Les ateliers de potiers gallo-romains de Montans, in: S, Menchelli and M. Paquinucci (eds) Territorio e produzioni ceramiche. Paesaggi, economia e società in età romana. Atti del Convegno internazionale, Pisa 20-22 ottobre 2005: 323‒346. Pisa: Edizioni Plus. Matijašić, R. 2001. I porti dell’Istria e della Liburnia, in C. Zaccaria (ed.) Strutture portuali e rotte marittime nell’Adriatico di Età Romana. Atti del Convegno Internazionale, Aquileia 20-23 maggio 1998. Antichità Altoadriatiche XLVI: 161‒174. Melis, R., S. Furlani, F. Antonioli, S. Biolchi., V. Degrassi and K. Mezgec 2012. Sea level and paleoenvironment during Roman times inferred from coastal archaeological sites in Trieste (northern Italy). Alpine and Mediterranean Quaternary 25 (1): 41‒55. Menchelli, S. 2018. Beyond the consumer city: il caso di Pisae, in F. Cantini and C. Rizzitelli (eds) Una città operosa. Archeologia della produzione a Pisa tra Età Romana e Medioevo. Atti del Convengo. Pisa, 26 maggio 2016: 21‒27. Firenze: All’Insegna del Giglio Miletić, Ž. 1991/92. Rimska cestovna mreža između Arauzone i Tragurija. Radovi Filozofskog fakulteta u Zadru 31 (18): 63‒88. Miletić, Ž. 2004. O rimskim cestama na aserijatskom području. ASSERIA 2: 7‒21. Miletić, Ž. 2011. Production of tegulae in Burnum in the context of building activities, in G. Lipovac Vrkljan, I. Radić Rossi and B. Šiljeg (eds) Rimske keramičarske i staklarske radionice. Proizvodnja i trgovina na Jadranskom prostoru. Zbornik I. Međunarodnog arheološkog kolokvija, Crikvenica, 23—24. listopada 2008: 263‒277. Crikvenica: Muzej Grada Crikvenice, Institut za arheologiju. Ožanić Roguljić, I. 2012. Pottery from the workshop of Sextus Metilius Maximus (Crikvenica-Igralište/Ad Turres, Northern Dalmatia). Rei cretariae Romanae fautorum Acta 42: 125‒132. Palazzo, P. 1994. Insediamenti artigianali e produzione agricola. I siti di Apani, Giancola, Marmorelle, La Rosa, in C. Marangio and A. Nitti (eds) Scritti di antichità in memoria di Benita Sciarra Bardaro: 53‒60. Fasano (Brindisi): Schena. Pallecchi, S. 2017. The Roman kilns at Giancola, Brindisi. Clues on the reconstruction of the work organisation system, 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 III. međunarodnog arheološkog kolokvija, Crikvenica, 4.-5. studenoga 2014. (ZIA 12): 29‒38. Crikvenica: Institut za arheologiju, Muzej Grada Crikvenice.

Papaioannou, T. 2011. A reconstruction of maritime trade patterns originating from western Asia Minor during Late Antiquity, on the bases of ceramic evidence, in D. Robinson and A. Wilson (eds) Maritime Archaeology and Ancient Trade in the Mediterranean (Oxford Centre for Maritime Archaeology: Monograph 6): 197‒210. Oxford: Oxford Centre for Maritime Archaeology, University of Oxford. Parica, M. 2012. Nekoliko primjera lučkih instalacija antičkih kamenoloma na dalmatinskim otocima. Histria Antiqua 21: 345‒353. Parica, M. 2018. The Remains of Infrastructural Facilities of the Ancient Quarries on Zadar Islands (Croatia), in D. Matetić Poljak and K. Marasović (eds) Interdisciplinary Studies on Ancient Stone Proceedings of the XI ASMOSIA Conference (Split 2015): 941‒949. Split: University of Split, Arts Academy in Split and Faculty of civil engineering, architecture and geodesy. Parica, M. and M. Ilkić 2017. Podmorski arheološki nalazi u okolici Ljupča, in J. Faričić and J. Lenkić (eds) Župa Ljubač – zrcalo povijesnih i geografskih mijena u sjeverozapadnom dijelu Ravnih Kotara: 98‒111. Zadar: Sveučilište u Zadru, Župa sv. Martina u Ljupču. Parica, M. and M. Ilkić 2018. Harbour installations in the context of the pottery workshop in Plemići Bay, in G. Lipovac Vrkljan and A. Konestra (eds) Pottery Production, Landscape and Economy of Roman Dalmatia. Interdisciplinary approaches: 71‒75. Oxford: Archaeopress. Peña, J. T. 2017. Issues in the Study of Rural Craft Production in Roman Italy, in T.C.A. de Haas, G.W. Tol (eds) The Economic Integration of Roman Italy (Mnemosyne Supplements, volume 404): 203‒230. Leiden, Boston: Brill. Poblome, J. 2006. Made in Sagalassos. Modelling regional potential and constraints, in: S, Menchelli and M. Paquinucci (eds) Territorio e produzioni ceramiche. Paesaggi, economia e società in età romana. Atti del Convegno internazionale, Pisa 20-22 ottobre 2005: 355‒363. Pisa: Edizioni Plus. Radić Rossi, I. 2011. Brodski tereti krovne opeke i proizvodi radionice Sexta Metilija Maksima u jadranskom podmorju, in G. Lipovac Vrkljan, I. Radić Rossi and B. Šiljeg (eds) Rimske keramičarske i staklarske radionice. Proizvodnja i trgovina na Jadranskom prostoru. Zbornik I. Međunarodnog arheološkog kolokvija, Crikvenica, 23—24. listopada 2008: 19‒30. Crikvenica: Muzej Grada Crikvenice, Institut za arheologiju. Reger, G. 2011. Inter-Regional Economies in the Aegean Basin, in Z. Archibald, J. K. Davies and V. Gabrielsen (eds) The Economies of Hellenistic Societies, Third to First Centuries BC: 368‒389. Oxford: Oxford University Press. Romanović, D. 2017. Amfore ravnog dna tipa Crikvenica iz slojeva antičke luke u Zatonu kod Zadra, in G. Lipovac Vrkljan, B. Šiljeg, I. Ožanić Roguljić and 33

Ana Konestra, Goranka Lipovac Vrkljan, Fabian Welc A. Konestra (eds) Rimske keramičarske i staklarske radionice. Proizvodnja i trgovina na jadranskom prostoru,  Zbornik III. međunarodnog arheološkog kolokvija, Crikvenica, 4.-5. studenoga 2014. (ZIA 12): 387‒405. Crikvenica: Institut za arheologiju, Muzej Grada Crikvenice. Rousse, C. 2011. Il sito di Loron (Istria, Croazia). L’organizzazione del complesso produttivo, in G. Lipovac Vrkljan, I. Radić Rossi and B. Šiljeg (eds) Rimske keramičarske i staklarske radionice. Proizvodnja i trgovina na Jadranskom prostoru. Zbornik I. Međunarodnog arheološkog kolokvija, Crikvenica, 23—24. listopada 2008: 75‒82. Crikvenica: Muzej Grada Crikvenice, Institut za arheologiju. Santoro, S. 2006. La localizzazione delle attività produttive della Cisalpina: commento ai dati di una ricerca in corso sull’artigianato romano, in Territorio e produzioni ceramiche. Paesaggi, economia e società in età romana. Atti del Convegno internazionale, Pisa 20-22 ottobre 2005: 165‒178. Pisa: Edizioni Plus. Scheidel, W. 2014. The shape of the Roman world: Modelling imperial connectivity. Journal of Roman Archaeology 27: 7‒32. Sternini, M. 2016. Italian Sigillata: Some Insights into a Complex Reality. Herom 5/2: 143‒160. Sternini, M. 2019. The production centres and river network of Italian terra sigillata between the Arno and Tiber valleys: a geographical point of view. Journal of Roman Archaeology 32: 485‒494.

Suić, M. 1996. Odnos grad-selo u antici na istočnoj jadranskoj obali, in N. Cambi (ed.) Odabrani radovi iz stare povijesti hrvatske — opera selecta: 427‒454. Zadar: Matica hrvatska, Arheološki muzej Zadar. Suić, M. 2003. Antički grad na istočnom Jadranu. Zagreb: Golden marketing — tehnička knjiga, Institut za arheologiju. Surić, M. 2009. Reconstructing Sea-Level Changes on the eastern Adriatic Sea (Croatia) – An Overview. Geoadria 14/2: 181‒199. Šušnjar, M., J. Bukovac, L. Nikler, I. Crnolatac, A. Milan, D. Šikić, I. Grimani, Ž. Vulić and I. Blašković 1970. Osnova geolška karta SFRJ, 1: 1:100 000, list Crikvenica. Beograd: Institut za geološka istraživanja, Zagreb, Savezni geološki zavod. Szakmány, G., S. Józsa and T. Bezeczky, New data on provenance and technology of the Fažana Amphora Workshop. A case study of Laecanii and Imperial amphorae, 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 III. međunarodnog arheološkog kolokvija, Crikvenica, 4.-5. studenoga 2014. (ZIA 12): 145‒158. Crikvenica: Institut za arheologiju, Muzej Grada Crikvenice. Tassaux, F. 2012. Obalne villae i njihov fundus, in M.B. Carre, V. Kovačić and F. Tassaux (eds) Sjeverno

priobalje poreštine u antici: 188‒190. Poreč: Zavičajni muzej poreštine. Uglešić, A. 2017. Podvršje - Glavičine: results of the most recent archaeological research. Hortus Artium Medievalium 23/2: 653‒665. Van Oyen, A. 2015. The Roman City as Articulated through Terra Sigillata. Oxford journal of archaeology 34: 279–299. Veal, R. 2017. Wood and Charcoal for Rome: Towards an Understanding of Ancient Regional Fuel Economics, in T.C.A. de Haas, G.W. Tol (eds) The Economic Integration of Roman Italy (Mnemosyne Supplements, volume 404): 388–406. Leiden, Boston: Brill. Vitali, D. (ed.) 2007. Le fornaci e le anfore di Albinia. Primi dati su produzioni e scambi dalla costa tirrenica al mondo gallico. Bologna: Dipartimento di archeologia. Vitali, D., F. Laubenheimer, L. Benquet, E. Cottafava and C. Calastri 2005. Le fornaci di Albinia (GR) e la produzione di anfore nella bassa valle dell’Albegna, in A. Camilli e M. Letizia Gualandi (eds) Materiali per Populonia, vol. 4: 259‒279. Firenze: All’Insegna del giglio. Vrsalović, D. 2011. Arheološka istraživanja u podmorju istočnoga Jadrana. Split: Književni krug Split, Arheološki muzej Split. Waelkens, M. et al. 1997. Interdisciplinarity in classical archaeology. A case study: the Sagalassos archaeological research project (southwest Turkey), in M. Waelkens and J. Poblome (eds) Sagalassos IV. Report on the survey and excavation campaigns of 1994 and 1995: 225‒252. Leuven: Leuven University Press. Welc, F. 2018. Geoarchaeology of pottery workshop sites in Roman Dalmatia, in G. Lipovac Vrkljan and A. Konestra (eds) Pottery Production, Landscape and Economy of Roman Dalmatia Interdisciplinary approaches: 62‒70. Oxford: Archaeopress. Welc, F. 2019. Geoarchaeological evidence of late and post-Antiquity (5th–9th c. AD) climate changes recordedat the Roman site in Plemići bay (Zadar region, Croatia). Studia Quaternaria 36/1: 3–17. Welc, F., A. Konestra, A. Dugonjić, P. Androić Gračanin, K. Rabiega and B. Nowacki 2019. Multidisciplinary insight into late Roman rural settlement on the north-eastern Adriatic coast. Island of Rab case study. Polish archaeology in the Mediterranean. 28/2: 481–508. Wilson, A. 2012. Raw materials and energy, in W. Scheidel (ed.) The Cambridge Companion to the Roman Economy  (Cambridge Companions to the Ancient World): 133‒155. Cambridge: Cambridge University Press. doi:10.1017/CCO9781139030199.010 Zelić, D. 1993. Nastanak urabanih naselja na otoku Krku. Radovi Instituta za povijest umjestnosti 17: 7‒17. 34

Distribution possibilities and transport infrastructure in northern Dalmatia’s pottery workshops

Zubin Ferri, T., S. Rončević, G. Lipovac Vrkljan and A. Konestra 2019. Post-depositional alterations of terrestrial and marine finds of Roman ceramics from Crikvenica production centre (NE Adriatic, Croatia) –

A contribution towards chemometric classification. Journal of cultural heritage: 12‒25. .

35

The Amber route between Caput Adriae and Emona basin: the ceramic evidence on inner road - and water - communications Iris Bekljanov Zidanšek, Petra Vojaković, Tina Žerjal Abstract The Prehistoric Amber route connected the emporia of Caput Adriae through Karstic passes and Ljubljana (Emona) Basin with the Baltic Sea. Some new archaeological discoveries with important ceramic assemblages and the dissemination of the imported wares from the end of the 2nd century BC till the beginning of the 1st century AD indicate trade routes and its organization just before and at the beginning of the Roman occupation of the area.

Key words Slovenia, Amber route, road - and water - communications, Piranon (Piran), Nauportus (Vrhnika), Emona (Ljubljana), Early Italic imports, 2nd and 1st century BC.

Figure 1. Trade routes between Caput Adriae and Emona basin in the 2nd century and the first half of the 1st century BC (made by Petra Vojaković; upgrade after Gaspari 2017, fig. 112).

Through the modern territory of Slovenia passes important transcontinental traffic routes between east and west or south and north (Figure 1). The south-eastern foothills of the Alps and the northern Adriatic always connected the central Danube Basin

and the northern Balkans with Italian Peninsula and the Mediterranean. Prehistoric Amber route, attested in the Argonaut Myths and mentioned by Strabo and other antique authors, connected the emporia of Caput Adriae through Karstic passes and Emona Basin with the

Roman Pottery and glass manufactures (Archaeopress 2022): 36–46

The Amber route between Caput Adriae and Emona basin

Baltic Sea (Buora 1996; Guštin and Gaspari 2005; Horvat and Bavdek 2009; Šašel 1974: 15−16; 1977; Šašel Kos 1990; 2017). North Istrian emporia The earliest imports from Italian Peninsula in the region are, among others North Adriatic sites, present also at the North Istrian emporia on the southwestern coast of Trieste gulf. An early emporium is supposed at the port of Piran, ancient Piranon (Stokin and Karinja 2004). At Bolniška ulica 1, excavated in 1993, an assemblage connected to the port facilities revealed material dated to the 2nd and mostly to the 1st centuries BC till the mid-1st century AD (Bekljanov Zidanšek 2011). Black-slip pottery sherds of forms Lamboglia 28, 5 and Lamboglia 7/16 (Plate 1: 1−4) were produced mainly in North Italian fabric, some were of Arretine origin (cf. Buora 1995; Dobreva and Griggio 2011; Donat 2011; Frontini 1985; Horvat and Bavdek 2009: 58−67; Schindler 1967; 1986). Thin-walled tall spindle-shaped beakers produced from depurated oxidized fabrics occur in forms Marabini 1 and 3 (Plate 1: 5−7; Figure 2), typical for the context of the 2nd till mid or third quarter of the 1st century BC (Donat 2007: 120−121; Horvat and Bavdek: 68−72; Marabini Moevs 1973: 49−53). Bowls and mortaria in Grey Venetic Ware (Plate 1: 8, 9) remained in common use until the end of their production in the mid-1st century AD (Bolzoni 2014; Cassani et al. 2007; Donat 2007: 117−118; Horvat and Bavdek 2009: 67). Central Italian Cooking Ware was produced on the Thyrrenian coast of Campania and Latium and in Late Republican period spread across the Italian peninsula, western Mediterranean and the newly conquered regions to the east and north. Baking dishes with a groove on the rim edge (orlobifido type) and lids (Plate 1: 10, 11) were appreciated for their high quality, resistant to temperature changes (Di Giovanni 1996; Peacock 1977: fabric Pompeian Red Ware; Schindler Kaudelka 1986). Wine amphorae sherds mainly belong to GrecoItalic and Lamboglia 2 amphorae (Plate 1: 12, 13; Figure 3). Both were produced on the Adriatic coasts, the earlier type in the 3rd and first half of the 2nd centuries BC (Horvat 1997: 60−62, 72−74; Kirigin 1994; Kirigin et al. 2006; Williams, Panella and Keay 2014), when the later form Lamb. 2 evolved from it. It was produced on many sites from the last quarter of the 2nd to the third decade of the 1st century BC (Bezeczky 2019: 41−42; Carre et al. 2014; Horvat 1997: 60−62, 72−74) and then developed into the later Dressel 6A form (Cipriano, Carre 1989, 85−88; Cipriano, Mazzocchin 2018), also present in minor quantities. Smaller amounts of other amphorae, such as Dressel 6B for olive oil transport were also identified (Bekljanov Zidanšek 2011).

Figure 2. Thin-walled pottery from Piran-Bolniška (photo: Iris Bekljanov Zidanšek).

Figure 3. Amphora Lamb. 2 from Piran-Bolniška (photo: Iris Bekljanov Zidanšek).

Esquilino type and amphorae Lamb. 2 were discovered at neighbour coastal location at Fornače (Stokin 1992: 83–85, Pl. 1–4; Horvat 1997: 73–74, 120–121). An important coastal emporium at Sermin (supposed ancient Aegida) near Koper revealed approximately same Late Republican imported wares, forms and productions, represented with black-slip pottery bowls and dishes such as forms Lamboglia 28 and 36, Italic thin-walled pottery beakers Marabini 1 and 3, Central Italian Cooking Ware, belonging mainly to the 2nd and 1st centuries BC. In the same period Italo-megaric cups were produced on the Aegean islands, Minor Asian coast, in Greece and on both Adriatic coasts. Adriatic wine trade is attested with Greco-Italic and Lamb. 2 amphorae, Rhodian amphorae reflect imports of the Aegean wine (Horvat 1997). The ceramic evidence shows even earlier long-trade connections of the

Similar material: black-slip pottery form Lamboglia 28, Grey Venetic Ware, Late Republican oil lamp of 37

Iris Bekljanov Zidanšek, Petra Vojaković, Tina Žerjal

1

2

3

4

5

6

7

8

9

10

11

12 13 Plate 1. PIRAN-BOLNIŠKA: 1–4 Black-slip pottery; 5–7 Thin-walled pottery; 8–9 Grey Venetic Ware; 10–11 Central Italic Cooking Ware; 12 Greco-Italic amphora; 13 Amphora Lamb. 2. M = 1 : 3 (drawings: Iris Bekljanov Zidanšek).

38

The Amber route between Caput Adriae and Emona basin

1

3

2

4

5

6

7

8

9

10

11

12

13

14

16 15 Plate 2. VRHNIKA-STARA POŠTA: 1–7 Black-slip pottery; 8–10 Thin-walled pottery; 11–12 Grey Venetic Ware; 13–14 Central Italic Cooking Ware; 15–16 Amphorae Lamb. 2. M = 1 : 3 (drawings: Suzana Puhar, Jerica Brečić).

39

Iris Bekljanov Zidanšek, Petra Vojaković, Tina Žerjal prehistoric settlement: fragments of Athic Red-paint pottery (end of 5th or 4th centuries BC), South Italic or Adriatic Gnathia pottery (4th to 2nd centuries BC) and similar Hellenistic productions (Horvat 1997). Ocra pass On the road to the continent the trade post on the relatively easily accessible mountain pass Ocra (600 m asl, present-day Razdrto over Nanos) appeared as a gateway between Mediterranean and Central Europe. Here different routes from the North Adriatic ports, i.e. Aquileia and Tergeste on the west and Tarsatica on the south converged (Guštin and Gaspari 2005; Horvat and Bavdek 2009; Šašel 1977). At the pass different locations with finds from different periods were investigated (Horvat and Bavdek 2009).

Figure 4. Black-slip pottery from Vrhnika – Stara pošta (photo: Jašar Skorupan).

beginning of the navigable river route Ljubljanica – Sava – Danube (general overview Horvat 1990; 2020). In 2016 similar imported finds from the end of the 2nd and beginning or first half of the 1st centuries BC were found at Nauportus at location Stara pošta on the left bank of Ljubljanica river (Vojaković et al. 2019). Traces of the earliest riverbank engineering, some disposal layers, a wooden container (supposedly a water tank) and a pavement of small stones were discovered.

The site of Mandrga, on the eastern side, is important for the earliest trade connections. Amounts of imported ware from the end of the 2nd and first half of the 1st centuries BC differentiate it from the surroundings villages with prehistoric characteristics. Therefore, a probable Italic trade post is assumed. Black-slip pottery sherds belonged to several productions, mainly C entral Italian, and forms: dishes Lamboglia 5 and 5/55, bowls Lamboglia 36, 6, 28, 31. Thin-walled pottery beakers Marabini 3 and 4 and hemispheric cups Marabini 36 originated from Central and Northern Italy. Some Grey Venetic Ware bowls and mortars, Venetic coarse cooking pots, Esquilino type oil lamps and wine amphorae Lamb. 2 were present. Italian Cooking Ware occurred in various Central and North Italian fabrics produced prior to popular Pompeian Red Ware. Beside the common orlobifido baking dishes or those with plain everted rim and their lids, some pots with almond rims and pots with emphasized rim edge were produced in similar Central Italian fabrics. Graphite pottery is more or less a unique indigenous element, if it could be supposed as such (Horvat and Bavdek 2009: 39−96).

Black-slip pottery (Figure 4) forms found here were common in Northern Italy at the end of the 2nd and first half of the 1st centuries BC. It includes a Lamboglia 6 dish, several dishes with rounded rim belonging to the group of Lamboglia 5 and fragments of conical bowls form Lamboglia 28 or Morel F 2640/2650 (Plate 2: 1−7; Lamboglia 1952: 147; Morel 1981: 113−115, pls. 62−64). The round oil lamp with a black slip is of the Late Republican Esquilino type. Similar lamps are known from Mandrga and from Grave 21 of the Celtic necropolis at Valeggio sul Mincio and dated to LTit D2 (Salzani 1995, 32-33, pl. 14: 5). Thin-walled pottery with spindle-shaped beakers Marabini 1 and 3 (Plate 2: 8−10), Grey Venetic Ware bowls and mortars (Plate 2: 11−12), Central Italian Cooking Ware (Plate 2: 13−14), fine Table Ware and amphorae Lamb. 2 (Plate 2: 15−16) represent trading goods from Italian peninsula. Scarce La Téne pottery, few Graphite sherd and four Tauriscan coins confirm trade with local indigenous communities. The predominance of imported pottery suggests a strong presence of Italic merchants (Vojaković et al. 2019).

The most natural route leading from Razdrto towards the east was a part of the main prehistoric Amber route and the later via publica Aquileia-Emona. It crosses the Pivka basin, goes through the Postojna Gate or across Studeno, past the Planina plain into Logatec and towards Vrhnika and Ljubljana marshes. Another old route across the Lugeon marsh (most likely Lake Cerknica) and cross the Bloke plateau towards the Krka or Kolpa river was oriented directly towards Pannonian lowlands and Segestica / Siscia (nowadays Sisak), which lies on the water routes of the river Sava, with its tributaries Ljubljanica, Krka and Kolpa (Horvat and Bavdek 2009: 130).

These finds are contemporary to the period from which Strabo’s mentions of the settlement rulled by celtic Taurisci originated (Strabo 7, 5, 2; Šašel Kos 1990: 17–20, 143–147; 2017). The discussion about the origin of the merchants, that were trading these goods, is still opened, as Strabo mainly mentioned indigenous population who traded with Aquileia and were changing imported goods such as wine for slaves, animals, horses and other perishable materials (Strabo 5,1,8; Horvat and Bavdek 2009, 145–146). Although at the moment the location of the indigenous village is not known. The location of Stara pošta is suitable for river crossing and port – used even in Middle ages (Turk et al. 2009, cat. numb. 50).

Nauportus Nauportus (today Vrhnika) was located in Ljubljana basin on the other side of the Alpine passes and at the 40

The Amber route between Caput Adriae and Emona basin

A logboat (14.4 m long) dated to the end of the 2nd century BC and remains of a sewn boat dated broadly to the 2nd century BC, discovered in the riverbed of the Ljubljanica between Stara pošta and Dolge njive, may thus be associated with the traffic that ran from Italy towards the east (Gaspari 2017).

They encompass fragments of black-slip pottery bowls Lamboglia 28, dishes Lamboglia 7/16 (Plate 3: 1–2) and thin-walled beakers Marabini 1 and Magdalensberg 35 (Plate 3: 3–4) (Novšak et al. 2017: 20–30). Even in other La Téne settlements in the region (Western and Central Slovenia) small amounts of imported ceramics were found (Gaspari 2017: sl. 112; Laharnar 2013; Horvat and Bavdek 2009: 138, 145: Bandera near Senožeče, Grad near Šmihel, Unec, Laze on the Planina Plain, Kovačevše, Vrtovin, Golo brdo, Most na Soči, Gradič near Kobarid), usually some fragments of black-slip pottery, thin-walled spindle-shaped beakers, Italian Cooking Ware or wine amphorae such as Lamb. 2. The sites are mostly located on the Kras plateau and in the Notranjska region (the area around the Amber route), rare hilltop settlements are along the Sava river or routes towards Segestica (Novo Mesto and Stična: Grahek 2016: 260–264, pl. 79, pl. 80: 14, pl. 95: 4). In the first half or the mid-1st century BC Roman influence is perceptible, but not as marked as at Nauportus.

At Nauportus, near the springs of the Ljubljanica river, was the beginning port, the entrance to an important inner waterway, that enabled the crossing of the dangerous Ljubljana marshes. It was important from the prehistoric times onwards as huge amounts of finds prove (Turk et al. 2009), and several mentions from the antique authors, especially associated with the legend of the Argonauts (Horvat 2020; Šašel Kos 1990; 2017). The river was even connected with an important river cult (Gaspari 2009; Gaspari 2017; Šašel Kos 2017). According to inscriptions Nauportus was a vicus within the Aquileian territory from the mid-1st century BC onwards, with the freemen of Aquileian merchant families in the leading roles (Šašel Kos 1990; 2017; Horvat 2020). A vast fort with storehouses and a river port was constructed along the river at the site Dolge njive on the right bank of Ljubljanica river either in the pre-Augustan or the Early Augustan period and operated as a reloading and supply post (Mušič and Horvat 2007; Horvat et al. 2016; Horvat 2020). The ships recovered from the Ljubljanica and also the small finds from the river reveal that traffic was particularly heavy in the Augustan period. The storage complex and a large share of the traffic were probably connected with supplying the legions deployed to the central Danube Basin (Horvat 1990; Istenič 2009a; 2009b, 2019; Gaspari 2017; Gaspari and Erič 2007; 2008).

With the consolidation of the Roman rule in the middle of the 1st centuries BC and especially in the Augustan period, with the deduction of the Colonia Iulia Emona on the left bank of the Ljubljanica river in AD 14/15 the picture changed. The presence of the Roman Army and newcoming Italic population gave an impulse to trade on larger scale. Research at Gornji trg revealed that Roman merchandise in the settlement below Grajski grič increased from the mid1st century BC onwards (Vičič 1994: 27–30; Županek, Ravnik 2017). The occupation and annexation of the area of Ljubljana to the province of Gallia Cisalpina is believed to have taken place in the time of Caesar’s proconsulship between 59 and 49 BC, with Cisalpina becoming part of Italy in 42 BC. The community of Roman citizens (a conventus) in the settlement below Grajski grič could thus have become a vicus and formed part of the vast territory of Aquileia (Slapšak 2014: 26–27; Gaspari 2014: 110–115; Gaspari et al. 2014; cf. Šašel Kos 2012; 2014). It would appear, however, that the prehistoric settlement in that period shrunk westwards. At Tribuna site the remains of two Roman military camps were excavated in 2008 and 2018/2019 by our team and unfortunately have not yet been fully published (Vojaković et al. 2011; Gaspari et al. 2014; Novšak et al. 2017). In the last two decades BC (20/10) and the first two decades AD (10/15), the Roman army established two successive camps on top of the ruins of the La Téne settlement and to the east of the Early Roman settlement. The multitude of Roman imports is certainly connected with the needs of the soldiers, a well organised supply of military and other goods, also provided by the inhabitants of the settlement below Grajski grič. The forts probably hosted a small army unit. It strategically protected the settlement, the river crossing (bridge or ferry dock), the harbour and

Emona On the other end of the Ljubljanica river and Ljubljana marshes, at the most important junction of communications in the south-eastern Alps stands Ljubljana – the ancient Emona. The latest discoveries enable us to observe the assemblages of an indigenous village at locations Tribuna (Vojaković 2014; Novšak et al. 2017) and Gornji trg (Vičič 1993; 1994; 2002; Županek, Ravnik 2017). The La Téne emporial settlement with a supposed river harbour spread on the terrace and southern slopes of Grajski grič (Castle hill) on the right bank of Ljubljanica river, in a vicinity of the first easy river crossing after the wetlands, away from dangerous floods and with sources of drinking water. A large part of the La Téne settlement was excavated at Tribuna site (Vojaković 2014; Novšak et al. 2017). Before the mid-1st century BC the imports from Italian Peninsula are rare, but present in very small quantities from the end of the 3rd and beginning of the 2nd centuries BC onwards (LT I, first La Téne phase). 41

Iris Bekljanov Zidanšek, Petra Vojaković, Tina Žerjal

1

2

3

4

5

6

7

9 8

10

11

12

13

14

15

Plate 3. LJUBLJANA-TRIBUNA: 1–4 La Tene settlement; 5–9 humus layer, 10–15 first Roman military camp: 1–2, 5–7, 10–12 Black-slip pottery; 3–4, 8–9, 13–15 Thin-walled pottery. M = 1 : 3 (drawings: Larisa Skalerič, Jerica Brečić).

42

The Amber route between Caput Adriae and Emona basin

Figure 5. Pottery finds from Ljubljana – Prule 9 (photo: Marko Pečovnik).

the continuation of land routes towards Siscia where military operations were underway at the time. The presence of the army has been related to the threat to Italian security during the Pannonian Wars in 12–9 BC and the Pannonian-Dalmatian rebellion in AD 6–9. The control of the communications’ node at the Ljubljana Gate was of crucial importance for the steady supply of the army engaged in battles in the east. Apart from that, the military presence in Emona has also been connected with the task of holding hostages here and, in the second phase, the construction of the Roman colony on the opposite, left bank of the Ljubljanica.

Conclusion Soon after the foundation of colonia Latina Aquileia in 181 BC Romans oriented to the east and to the southeastern Alpine region as was recognised by the coin hoards and individual coin finds such as victoriati (end of the 3rd to 141BC) and later Republican coins at Dutovlje, Gradišče near Knežak, Baba near Slavina, Stari grad near Unec, Žerovnišček, Cerkno and Ljubljanica river (Horvat and Bavdek 2009: 137–139; Miškec 2003; 2007). New archaeological discoveries with important ceramic assemblages confirm it. Early trade posts at Ocra - Mandrga (Razdrto) and Nauportus – Stara Pošta (Vrhnika) indicates trade routes and its organization before the Roman occupation of the area. By the late 2nd century BC the Romans controlled the whole route, from the North Adriatic ports (Caput Adraiae) and emporia to the Emona Basin. Probably this happened at the latest after the military campaign of C. Sempronius Tuditanus in 129 BC or the campaigns of M. Aemilius Scaurus in 115 BC (Horvat and Bavdek 2009: 133–136; Gaspari 2017).

The arrival of the army brought about a change in the importance of the river harbour, and the structures unearthed at Prule 9 may be connected with the military camp and the harbour in the vicinity. The same assemblages are known from these nearby excavations near the river bank (Žerjal 2017). Huge amounts of North Italian sigillata are present, oil lamps, balsamaria, Italian Cooking Ware, wine amphorae from Italy and Eastern Mediterranean, fish sauce amphorae from Iberian Peninsula, olive oil amphorae from north Adriatic and Istria (Figure 5). Although imports prevail, some products of local potters occur too: Coarse Cooking Ware and a pot (fig. 5 – no. 3) in fine grey depurated La Téne pottery that in Emona was produced at least till the end of the 1st century AD (Plesničar Gec 1977: 15). The pot is decorated with feather or rouletting decoration – a feature copied from the imported sigillata. The local potter fused local traditions and newcoming knowledge and taste.

The dissemination of the imported wares in indigenous villages in Western and Central Slovenia is scarce. The earlier sites are mostly located on the Kras plateau and in the Notranjska region (the area around the Amber route). Hilltop settlements along the Sava river or routes towards Segestica revealed scarce pre-Augustan material. In the middle of the 1st centuries BC and especially in the Augustan period the consolidation of the Roman 43

Iris Bekljanov Zidanšek, Petra Vojaković, Tina Žerjal rule, the presence of the Roman Army and Italic newcomers gave an impulse to trade on larger scale.

della Venetia. Rei Cretariae Romanae Fautorum Acta 45: 261−271. Di Giovanni, V. 1996. Produzione e consumo di ceramica da cucina nella Campania romana (II a.C.- II d.C.), in M. Bats (ed.) Les céramiques communes de Campanie et de Narbonnaise (Ier s. av. J.-C. – IIe s. ap. J.-C.). La vaiselle de cuisine et de table (Collection Centre Jean Bérard 14): 65−103. Naples: Centre Jean Bérard. Dobreva, D. and A. M. Griggio 2011. La ceramica a vernice nera dai fondi ex Cossar ad Aquiliea: problematiche e prospettive di ricerca. Quaderni Friulani di Archeologia 21: 77−100. Donat, P. 2007. La ceramica nella Cisalpina orientale dalla fondazione di Aquileia ad Augusto. Antichità Altoadriatiche 68:109−146. Donat, P. 2011. Ceramica a vernice nera nel Friuli Venezia Giulia. L’edito. Prima parte: la diffusione della classe. Quaderni Friulani di Archeologia 21:17−39. Erič, M., Gaspari, A. Čufar, K. Solina, F, Verbič, T. 2014. Zgodnjerimska ladja iz Ljubljanice pri Sinji Gorici / Early Roman barge from the Ljubljanica River at Sinja Gorica. Arheološki vestnik 65: 187–254. Frontini, P. 1985. La ceramica a vernice nera nei contesti tombali della Lombardia (Archeologia dell’Italia Settentrionale 3). Como: New Press. Gaspari, A. 2009. Keltski bojevniki in Ljubljanica, in Turk et al. 2009: 68‒73. Gaspari, A. 2014. Prazgodovinska in rimska Emona. Vodnik skozi arheološko preteklost predhodnice Ljubljane. Ljubljana: Muzej in galerije mesta Ljubljane. Gaspari, A. 2017. Deblak s konca 2. stoletja pr. n. št. iz Ljubljanice na Vrhniki/ The late 2nd Century BC Logboat from the Ljubljanica River at Vrhnika. Ljubljana: Znanstvena založba Filozofske fakultete, Muzej in galerije mesta Ljubljane = University Press, Faculty of Arts: Museum & Galleries of Ljubljana. Gaspari, A. and M. Erič 2007. Preliminarno poročilo o arheološkem podvodnem pregledu strug potokov Bistra in Ljubija na Ljubljanskem barju v letih 2005 in 2006. ZVKDS OE Ljubljana, Skupina za podvodno arheologijo. Gaspari, A. and M. Erič 2008. Arheološke raziskave struge Ljubljanice med Verdom in Vrhniko. Annales. Series historia et sociologia 18/2: 407-430. Gaspari, A., I. Bekljanov Zidanšek, J. Krajšek, R. Masaryk, A. Miškec and M. Novšak 2014. Novejša arheološka spoznanja o Emoni med zatonom prazgodovinske skupnosti in gradnjo rimskega mesta (druga polovica 1. stol. pr. n. št. in začetek 1. stol. n. št.) / New Archaeological Insights about Emona between the Decline of the Prehistoric Community and the Construction of the Roman Town (second half of the 1st century BC and early 1st century AD), in M. Ferle (ed.), Emona: mesto v imperiju / Emona: a City of the Empire: 135–165. Ljubljana: Muzej in galerije mesta Ljubljane.

The Augustan period brought about significant changes in the communication network. The construction of the road, via publica from Aquileia to Emona, through Hrušica and the northern fringes of the Ljubljana marsh, which was finished in the early Tiberian period, shifted a part of the main cargoes from the river to the land routes (Istenič 2009a; 2009b). Nauportus shrunk to its core along the road on the left bank (Horvat 1990; 2020). Consolidation of the Roman rule in the central Danube Basin presumably caused it to lose its role as an important post on the military supply line, and the regional centre of traffic and commerce shifted to the nearby colony of Emona. The transcontinental importance of the Ljubljanica river ceased (Istenič 2009a; 2009b; 2019; Gaspari 2017). The road across Razdrto became a minor road (Horvat and Bavdek 2009: 146). References Bekljanov Zidanšek, I. 2011. Bolniška ulica 1 – Piran: rezultati predhodnih arheoloških raziskav iz leta 1993. Unpublished MA Thesis. University of Ljubljana, Faculty of Arts, Department of Archaeology. Bezeczky, T. 2019. Italian and Istrian amphorae, in T. Bezeczky (ed.) Amphora Research in Castrum Villa on Brijuni Island (Denkschriften/Österreichische Akademie der Wissenschaften, PhilosophischHistorische Klasse 509; Archäologische Forschungen 29): 41−10. Wien: Austrian Academy of Sciences Press. Bolzoni G. 2014. La ceramica grigia nell’Italia Settentrionale come indicatore di fenomeni di acculturazione: il contatto con il mondo romano. Rei Cretariae Romanae Fautorum Acta 43: 241−250. Buora, M. 1995. Ceramica a vernice nera da Sevegliano. Alba Regia 25, Rei Cretariae Romanae Fautorum Acta 34: 155−163. Buora, M. (ed.) 1996. Lungo la via dell’Ambra. Udine: Arti Grafiche Friulane. Carre, M. - B., P. Monsieur and S. Pesavento Mattioli 2014. Transport amphorae Lamboglia 2 and Dressel 6A: Italy and/or Dalmatia? Some clarifications. Journal of Roman Archaeology 27: 417−428. Cassani G., S. Cipriano, P. Donat and R. Merlatti 2007. Il ruolo della ceramica grigia nella romanizzazione dell’Italia nord-orientale: produzione e circolazione. Antichità Altoadriatiche 65: 249−281. Cipriano, M. T. and M. B. Carre 1989. Production et typologie des amphores des Côte Adriatique et l’Italie, in Amphores romaines et histoire économique, dix ans de recherche (Collection de l’École française de Rome 114): 67−104. Rome: Ecole française de Rome; Paris: De Boccard. Cipriano, S. and S. Mazzocchin 2018. Sulla cronologia delle anfore Dressel 6A: novità dai contesti di bonifica 44

The Amber route between Caput Adriae and Emona basin

Grahek, L. 2016. Stična. Železnodobna naselbinska keramika / Stična. Iron Age Settlement Pottery (Opera Instituti Archaeologici Sloveniae 32). Ljubljana: Založba ZRC SAZU. Guštin, M. and A. Gaspari 2005. Ocra: il passo tra il mondo romano e le comunita protostoriche continentali, in G. Bandelli and E. Montagnari Kokelj (eds) Carlo Marchesetti e I castellieri 1903-2003 (Fonti e Studi per la Storia della Venezia Giulia 9): 353−360. Trieste: Editreg. Horvat, J. 1990. Nauportus / Vrhnika (Dela I. razr. SAZU / Opera Academia scientiarum et artium Slovenica, Classis I: Historia et sociologia 33). Ljubljana: Slovenska akademija znanosti in umetnosti. Horvat, J. 1997. Sermin. Prazgodovinska in zgodnjerimska naselbina v severozahodni Istri./Sermin. A Prehistoric and Early Roman Settlement in Northwestern Istria (Opera Instituti Archaeologici Sloveniae 3). Ljubljana: Založba ZRC SAZU. Horvat, J. 2020. Nauportus – Vrhnika, in J. Horvat, I. Lazar, A. Gaspari (eds), Manjša rimska naselja na slovenskem prostoru / Minor Roman settlements in Sloveina (Opera Instituti Archaeologici Sloveniae 40): 93-112 Ljubljana: Založba ZRC SAZU. Horvat, J. and A. Bavdek 2009. Okra. Vrata med Sredozemljem in Srednjo Evropo./Ocra. The gateway between the Mediterranean and Central Europe (Opera Instituti Archaeologici Sloveniae 17). Ljubljana: Založba ZRC SAZU. Horvat, J., P. Peterle Udovič, T. Tolar, B. Toškan 2016. Območje pristanišča v Navportu / The port area of Nauportus. Arheološki vestnik 67: 177-258. Istenič, J. 2009a. Ljubljanica – rimska trgovska in prometna pot, in Turk et al. 2009: 74‒80. Istenič, J. 2009b. The early Roman military route along the river Ljubljanica (Slovenia), in Á. Morillo, N. Hanel, E. Martín (eds) Limes XX. Estudios sobre la frontiera romana/ Roman frontier studies (Anejos de Gladius, 13): 855–865. Madrid, Merida. Istenič, J. 2019. Roman military equipment from the river Ljubljanica: typology, chronology and technology = Rimska vojaška oprema iz reke Ljubljanice: arheološke in naravoslovne raziskave (Katalogi in monografije 43). Ljubljana: Narodni muzej Slovenije. Kirigin, B. 1994. Grčko-italske amfore na Jadranu. Arheološki vestnik 45: 15-24. Kirigin, B, T. Katunarić and L. Šešelj 2006. Preliminary notes on some economic and social aspects of amphorae and fine pottery from Central Dalmatia, 4th-1st BC, in Rimini e l’Adriatico nell’età delle guerre puniche. Atti del Convegno Internazionale di Studi: 25‒27. Bologna: Ante Quem. Laharnar, B. 2013. Early Roman military equipment from the fortified settlements in the Notranjska region (SW Slovenia), in M. Sanader, A. Rendić-Miočević, D. Tončinić and I. Radman-Livaja (eds) Rimska vojna oprema u pogrebnom kontekstu: radovi XVII. ROMEC-a = Weapons and military equipment in a funerary context: proceedings of the XVIIth Roman Military Equipment

Conference (Roman Military Equipment Conference 17): 379–392. Zagreb: Arheološki muzej u Zagrebu, Odsjek za arheologiju Filozofskog fakulteta u Zagrebu. Lamboglia, N. 1952. Per una classificazione preliminare della ceramica campana, in Atti del Io congresso internazionale di studi Liguri: 139–206. Bordighera: Istituto Internazionale di Studi Liguri. Marabini Moevs, M. T. 1973. The Roman thin walled pottery from Cosa (1948-1954) (Memoirs of the American Academy in Rome 32). Rome: American Academy in Rome. Miškec, A. 2003. The Early Romanization of the Southeastern Alpine Region in the Light of Numismatic Finds. Arheološki vestnik 54: 369‒379. Miškec, A. 2007. Monetary circulation in the Posočje Region in Antiquity, in M. Chiabà, P. Maggi and C. Magrini (eds) Le Valli del Natisone e dell’Isonzo tra Centroeuropa e Adriatico (Studi e ricerche sulla Gallia Cisalpina 20): 275–281. Trieste - Roma: Edizioni Quasar. Miškec, A. 2009. The Augustan conquest of southeastern Alpine and western Pannonian areas: coins and hoards. Arheološki vestnik 60: 283–296. Morel, J. - P. 1981. Céramique campanienne: les forms (Bibliothèque des Écoles française d’Athènes et de Rome 244). Rome: École française de Rome. Mušič, B. and J. Horvat 2007. Nauportus – an Early Roman trading post at Dolge njive in Vrhnika: the results of geophysical prospecting using a variety of independent methods. Arheološki vestnik 58: 219–283. Novšak, M., I. Bekljanov Zidanšek and P. Vojaković 2017. Zaton predrimske naselbine na Tribuni. Razumevanje morebitne diskontinuitete poselitve med zadnjo fazo latenskega naselja in rimskim vojaškim taborom / The decline of the pre-Roman settlement at Tribuna. Deliberations on the possibility of settlement discontinuity between the final phase of the La Tène settlement and the Roman military camp, in: B. Vičič and B. Županek (eds) Emona MM, Urbanizacija prostora – nastanek mesta / Emona MM, Urbanisation of Space – beginning of a Town): 9–52. Ljubljana: Muzej in galerije mesta Ljubljane. Peacock, D. P. S. 1977. Pompeian Red Ware, in D. P. S. Peacock (ed.) Pottery and early Commerce. Characterization and Trade in Roman and Later ceramics: 147−162. London - New York: Academic Press. Plesničar Gec, L. 1977. Keramika emonskih nekropol / The pottery of Emona necropolises (Dissertationes et monographiae 20). Ljubljana: Mestni muzej; Beograd: Savez arheoloških društava Jugoslavije. Salzani, L. (ed.) 1995. La necropoli gallica di Valeggio sul Mincio (Documenti di archeologia 5). Mantova: Società Archeologica Padana. Schindler, M. 1967. Die ‘Schwarze Sigillata’ des Magdalensberges (Kärnter Museumsschriften 43, Archäologische Forschungen zu den Grabungen auf 45

Iris Bekljanov Zidanšek, Petra Vojaković, Tina Žerjal dem Magdalensberg 1). Klagenfurt: Geschichtsverein für Kärnten. Schindler, M. 1986. Die ‘Schwarze Sigillata’ des Magdalensberges 2. Neufunde seit 1965, in H. Vetters, G. Piccottini (eds) Die Ausgrabungen auf dem Magdalensberg 1975 bis 1979 (MagdalensbergGrabungsbericht 15): 345−390. Klagenfurt: Geschichtsverein für Kärnten. Schindler Kaudelka, E. 1986. Die Backplatten vom Magdalensberg, in H. Vetters and G. Piccottini (eds) Die Ausgrabungen auf dem Magdalensberg 1975 bis 1979 (Magdalensberg-Grabungsbericht 15): 279−337. Klagenfurt: Geschichtsverein für Kärnten. Slapšak, B. 2014. Na sledi urbanega: poti do prve izkušnje mesta v prostoru Ljubljane / Unravelling the Townscape: Tracing the First Urban Experience on the Location of the Present-day Ljubljana, in M. Ferle (ed.) Emona: mesto v imperiju / Emona: a City of the Empire: 17–42. Ljubljana: Muzej in galerije mesta Ljubljane. Stokin, M. 1992., Naselbinski ostanki iz 1. st. pr. n. št. v Fornačah pri Piranu. Arheološki vestnik 43: 79–92. Stokin, M. and S. Karinja 2004. Rana romanizacija i trgovina u sjeverozapadnoj Istri s naglaskom na materijalnu kulturu. Histria Antiqua 12: 45–54. Šašel, J. 1974. Okra. Kronika 22: 9−17. Šašel, J. 1977. Strabo, Ocra and archaeology, in Ancient Europe and Mediterranean. Studies presented in honour of Hugh Hencken: 157−160. Warminster: Aris & Phillips. (republished in Šašel, J. 1992. Opera Selecta (Situla 30): 630-633. Ljubljana: Narodni muzej Slovenije). Šašel Kos, M. 1990. Nauportus: antični literarni in epigrafski viri / Nauportus: Literary and Epigraphical Sources, in Horvat 1990: 17–33, 143–159. Šašel Kos, M. 2012. Colonia Iulia Emona – the genesis of the Roman city. Arheološki vestnik 63: 79–104. Šašel Kos, M. 2014. Kaj se je leta 14/15 dogajalo v Emoni – cesarski napis in upor panonskih legij / What was happening in Emona in AD 14/15? An imperial inscription and the mutiny of the Pannonian legions, in M. Ferle (ed.) Emona: mesto v imperiju / Emona: a City of the Empire: 79–96. Ljubljana: Muzej in galerije mesta Ljubljane. Šašel Kos, M. 2017. Antično ime za Ljubljanico / The ancient name(s) for the Ljubljanica River, in B. Vičič, B. Županek (eds) Emona MM, Urbanizacija prostora – nastanek mesta/ Emona MM, Urbanisation of Space – beginning of a Town), 225–234. Ljubljana: Muzej in galerije mesta Ljubljane. Turk, P., J. Istenič, T. Knific and T. Nabergoj 2009. Ljubljanica – kulturna dediščina reke. Ljubljana: Narodni muzej Slovenije. Reprint in English (Ljubljanica – a river and its past).

Vičič, B. 1993. Zgodnjerimsko naselje pod Grajskim gričem v Ljubljani. Gornji trg 15. Arheološki vestnik 44: 153–201. Vičič, B. 1994. Zgodnjerimsko naselje pod Grajskim gričem v Ljubljani. Gornji trg 30, Stari trg 17 in 32. Arheološki vestnik 45: 25–80. Vičič, B. 2002. Zgodnjerimsko naselje pod Grajskim gričem v Ljubljani. Gornji trg 3. Arheološki vestnik 53: 193–221. Vojaković, P. 2014. Predrimska Emona v luči novih arheoloških odkritij / Pre-Roman Emona in the Light of New Archaeological Discoveries, in M. Ferle (ed.) Emona: mesto v imperiju / Emona: a City of the Empire: 65–78. Ljubljana: Muzej in galerije mesta Ljubljane. Vojaković, P., M. Novšak, T. Žerjal, T. Verbič, J. Krajšek and J. Hrustel 2011. Poročilo o predhodnih arheoloških raziskavah na lokaciji Ljubljana – stanovanjska soseska Tribuna. Unpublished report. Ljubljana. Vojaković, P., I. Bekljanov Zidanšek and B. Toškan 2019. Poznorepublikanski Navport: območje Stare Pošte / Late Republican Nauportus: the Stara pošta site. Arheološki vestnik 70: 93–126. Williams, D., C. Panella and S. Keay 2014. Graeco-Italic Amphorae. Roman Amphorae: a digital resource. University of Southampton, 2005 (updated 2014), last viewed 30.12.2019. https://archaeologydataservice. ac.uk/archives/view/amphora_ahrb_2005/details. cfm?id=141&CFID=f5752a1d-b004-4a41-bf160df26fecd927&CFTOKEN=0 Žerjal, T. 2017. Obrežje Ljubljanice na Prulah (Ljubljana) v avgustejskem obdobju / The bank of the Ljubljanica at Prule (Ljubljana) in the Augustan period, in B. Vičič and B. Županek (eds) Emona MM, Urbanizacija prostora – nastanek mesta/ Emona MM, Urbanisation of Space – beginning of a Town), 53–69. Ljubljana: Muzej in galerije mesta Ljubljane. Županek, B. and M. Ravnik 2017. Nekaj novih ugotovitev o poselitvi pod Grajskim gričem v Ljubljani: raziskave na Starem in Gornjem trgu v letih 2009–2011 / New insights into the settlement below Grajski grič in Ljubljana: results of the 2009–2011 investigations at Stari trg and Gornji trg, in B. Vičič and B. Županek (eds) Emona MM, Urbanizacija prostora – nastanek mesta/ Emona MM, Urbanisation of Space – beginning of a Town), 71–89. Ljubljana: Muzej in galerije mesta Ljubljane.

46

Merci e sedi di mercato: lana e produzione ceramica a Mutina e ai Campi Macri (Modena, I) Carla Corti Abstract To understand the productive and commercial success of some goods, such as ceramics, it is necessary to observe more generally the economic context in which they are located. In order to balance the overall picture, particular attention should then be given to the goods that are most difficult to trace in the archaeological record, such as wool and textiles. Finally, the presence of supra-regional marketplaces can make the difference in outlining the economic dynamics. The case represented by Mutina (Modena), well-known for its wool and pottery production in Roman times, and by the Campi Macri (Magreta, Modena), seat of an important cattle fair in the Republican and Early Imperial age, offers the opportunity to investigate the relationships that have been established between production and marketing of these two goods as well as their evolution over time.

Key words Goods, roman fears and markets, roman trade, lamps production, wool roman economy.

Nella nascita di centri di produzione ceramica sono indispensabili le opportunità offerte dal territorio (risorse naturali e ubicazione in relazione a porti e alle vie di transito), ma per comprendere il successo commerciale di alcune produzioni è necessario valutare anche altri fattori, in primo luogo la presenza di sedi di mercato e di settori di grande impatto economico, come indubbiamente furono i prodotti tessili,1 che possano aver fatto inizialmente da traino. Categorie merceologiche come la ceramica fine da mensa e le lucerne fittili viaggiarono infatti soprattutto come merce di accompagno (Andreau 1991: 379; Tchernia 2011: 111‒112). Si tratta di prodotti che possiamo considerare “di massa”, più economici e facilmente vendibili, ma che garantivano al contempo minori ricavi. Una loro commercializzazione autonoma necessitava quindi di una distribuzione economicamente sostenibile, garantita dai quantitativi di merce venduta e da un consolidato successo commerciale. La presenza in ambito norditalico in età repubblicana e nella primissima età imperiale di un importante mercato, i Campi Macri (Magreta, Formigine, MO), in un ambito geografico che vede la contestuale presenza di produzioni di conclamato successo commerciale (lana e ceramica), fa di Mutina (Modena) il caso ideale per indagare questa relazione e le dinamiche produttivocommerciali e socio-economiche che si sono verificate

Ringraziamenti: desidero ringraziare il Museo Civico di Modena, nella persona di Cristiana Zanasi, per aver messo a disposizione per questo studio la documentazione fotografica dei materiali archeologici di proprietà civica provenienti da Magreta. 1 Vedi quanto osservato in Andreau 1991: 373‒374. *

La produzione ceramica a Mutina Plinio il Vecchio colloca la città di Mutina fra i più importanti centri di produzione ceramica del mondo romano, i cui vasi furono ampiamente esportati per terra e per mare (Nat. hist., XXXV, 160‒161). Plinio, nel I sec. d.C., fa tuttavia esplicito riferimento a una produzione realizzata al tornio (rotae officinis) (Buonopane e Corti 2017: 209), non ancora individuata.2 Nota da tempo è invece la produzione di lucerne, in particolare di imitazioni di Herzblattlampen pergamene, di Dressel 3 e soprattutto di Firmalampen, in questo caso documentate sia da matrici, che dalla specifica Mutinae fecit associata su alcuni marchi ai fabbricanti Cherinthus, C. Fadius, Fortis, Menander e Priscus (Buchi 1975: xxxiv; Parra 1983; Forte 1988: 110‒111; Michelini 1988: 536‒537; Giordani 2001: 263‒264; Labate 2016; Buonopane e Corti 2017: 209, 212‒213; Capellini e Corti 2020: 32-35, 43-54). I marchi presenti su matrici sono, per le lucerne di tradizione ellenistica e tardorepubblicane, VAL = Val(eri?), ALIX = Alix(ander?), PM = P(ubli) M(- - -) e, per quanto riguarda le Firmlampen, EVCARPI = Eucarpi (Parra 1983: 91; Labate 2016: 19, 23, 30).3 Noto da tempo è inoltre il frammento Ci si riferisce ovviamente alle produzioni che hanno avuto valenza commerciale nei traffici a medio-lungo raggio, le sole oggetto di questo contributo. 3 Altri produttori sono stati collocati a Mutina in base al rinvenimento di materiali interpretati come scarti di produzione, da viale Reiter e Novi Sad (Labate 2016); le alterazioni, tutte di lieve e lievissima entità, possono tuttavia aver avuto anche altre cause; prendiamo ad es. gli esemplari di viale Reiter, un contesto estremamente eterogeneo e edito solo in minima parte (Labate e Raimondi 2010): qui, su lucerne, ceramica comune, anfore e mattonelle pavimentali sono state riscontrate superfici annerite e su alcune lucerne fessurazioni e occasionali danni maggiori (vedi Labate 2016; Labate 2017; riscontro autoptico), che possono però essere compatibili anche con una ricottura per esposizione a fuoco secondario (ad es. 2 

roman pottery and glass manufactures (Archaeopress 2022): 47–55

Carla Corti

Figura 1. Carta dell’Italia settentrionale con l’ubicazione di Mutina (Modena).

di tegola su cui compare l’iscrizione Âd forn(acem) Cât(ianam vel -iorum?) / L(uci) Âemili / Fortis, databile ai primi decenni del I sec. d.C., forse posta come insegna dell’impianto produttivo.4 Accanto ai rinvenimenti di matrici modenesi e ai marchi con la specifica Mutinae fecit occorre menzionare i produttori individuati su base archeometrica, tra questi Strobilus, Communis, Phoetaspus e Atimetus (Schneider e Wird 1992; Ceci e Schneider 1994; Schneider e Daszkievic 2011).

ellenistico orientale con le produzioni pergamene ed efesine, grazie alla quale si può agevolmente aumentare la produzione e garantire uno standard qualitativo costante. Si tratta di una primogenitura che in Italia settentrionale sempre più concordemente viene riconosciuta all’ambito emiliano, in particolare modenese, e qui alle produzioni avviate nell’area dei Campi Macri (Grassi 2013: 217; Labate 2016: 19). Questo clima pionieristico, documentato a Mutina tra tarda età repubblicana e prima età imperiale, è contraddistinto dalla sperimentazione di nuove forme e dei rapporti di produzione (come farebbero sottintendere gli esemplari a doppia firma). È in tale clima che pare essersi verosimilmente realizzata anche la creazione di un nuovo tipo, la Firmalampe,5 il cui ampio successo commerciale è testimoniato dalla distribuzione dei rinvenimenti e dall’avvio di produzioni provinciali con marchi modenesi, attestate dal rinvenimento di matrici e/o dall’analisi archeometrica.6 Questo successo

L’avvio della produzione di imitazioni delle Herzblattlampen pergamene, la prima per la quale si può parlare a Mutina di successo commerciale, si colloca dalla metà del I sec. a.C. (Parra 1983: 101; Grassi e Mandelli 2012: 132; Grassi 2013: 215‒219; Labate 2016: 23). Essa documenta l’introduzione in Italia settentrionale di un’innovazione tecnologica, la produzione a matrice, già pienamente affermata nel III sec. a. C. in ambito l’incendio di un magazzino in cui erano stivati prodotti diversi, come farebbe supporre l’anfora cretese di forma Dressel 43/AC4 indicata come scarto di produzione in Labate 2017: 229, fig. 1); piú in generale sull´argomento cfr. quanto considerato in Schindler Kaudelka e Zabehlicky-Scheffenegger 2000. 4  Da ultimi Buonopane, Corti 2017: 209; una diversa proposta di lettura in Labate 2017: 229.

Si rimanda a quanto recentemente osservato in Capellini e Corti 2020: 43-51. 6 Già Ezio Buchi metteva ben in evidenza questo successo commerciale e l’articolazione delle produzioni (Buchi 1975: XL‒XLI); il lavoro di identificazione su base archeometrica è stato portato avanti in primis da Gerwulf Schneider. 5

48

Merci e sedi di mercato: lana e produzione ceramica a Mutina e ai Campi Macri (Modena, I)

Figura 2. Ceramica pergamena con rilievi applicati da Magreta (Modena, Museo Civico, Archivio fotografico).

coinvolse anche l’ambito italico, come attesta una cassa ancora imballata scoperta a Pompei con terra sigillata sud-gallica e Firmalampen di Strobilius, Communis e Fortis (Pucci 1977: 16).

I Campi Macri: una fiera di importanza economica panitalica, i contatti con il Mediterraneo orientale e l’indotto. Le prime produzioni di lucerne a Mutina Aveva sede ai Campi Macri una rinomata fiera extraurbana annuale del bestiame (mercatus o conventus), specializzata in ovini, di “importanza economica panitalica” (Gabba 1975: 156), che affondava molto probabilmente le sue radici della realtà economica locale del substrato preromano (Corti 2016). La particolarità della situazione appare evidente dal contesto socio-economico, con aspetti religiosi, e storico in cui si collocano i Campi Macri.7 La località, posta presso l’attuale Magreta (Formigine, MO), si trovava poco a sud-ovest di Mutina, nelle vicinanze del fiume Secchia, e fu direttamente collegata alla via Aemilia. La fiera ebbe luogo dall’età

Ma perché proprio Mutina divenne un centro così importante e attivo nella produzione ceramica? A cosa si deve il fervore produttivo-commerciale documentato dal dato archeologico? La disponibilità di risorse idonee all’impianto di una fiorente attività (argilla, acqua, legname) non distingue la città dai centri limitrofi e più in generale dal resto della pianura padana. Anche la posizione topografica, al centro della pianura emiliana (Figura 1), di per sé non aiuta a comprendere il successo commerciale delle sue ceramiche. A questo punto occorre guardare ai Campi Macri. È qui infatti che si collocano le prime produzioni di lucerne documentate nel Modenese.

Sui Campi Macri si vedano da ultimi Ortalli 2012, ivi bibliografia precedente, e Corti 2016; su fiere e mercati nel mondo romano si rimanda a De Ligt 1993, Frayn 1993 e Lo Cascio 2000. 7 

49

Carla Corti

Figura 3. Matrici di lucerne di tipo pergameno (Herzblattlampen) da Magreta (Modena, Museo Civico, Archivio fotografico).

repubblicana ad almeno la prima età augustea e risulta cessata da tempo nel 56 d.C., quando un Senatoconsulto autorizzò la demolizione degli edifici ormai fatiscenti ancora esistenti nell’area (S.C. Volusianum) (cfr. Calzolari 2008: 78‒79). Intorno all’evento annuale, in grado di richiamare l’attenzione anche dell’élite romana — qui infatti si recava per acquistare ovini pure Turranius Niger, cui Varrone dedicò il suo libro sull’ars pecuaria (Rust. 2, preaf. 6) — si svilupparono apprestamenti stabili con caratteristiche permanenti e semipermanenti (Buonopane e Corti 2017: 208, 210), tanto che Strabone (5, 1, 11) annovera i Campi Macri tra le mikrà polísmata. Ci troviamo in un ambito geografico, l’Aemilia centro-occidentale, tra Mutina e Parma, in cui il settore economico legato alla lana assunse un ruolo di primissimo piano, estremamente redditizio — erano in assoluto le lane più costose tra quelle elencate nell’Editto dioclezianeo — e longevo, documentabile tra la tarda età repubblicana e l’inizio del IV sec. d.C. Delle rinomate lane di questi luoghi e delle attività ad esse legate ci parlano gli autori antichi, supportati

dai dati epigrafici (Vicari 2001: 45‒47 e da ultimi, Corti 2017a e Buonopane e Corti 2017: 208‒209), anche se solo attraverso un uso integrato delle fonti (scritte, archeologiche, paleobotaniche, risorse ambientali e resti faunistici) emerge al meglio la portata di questo tipo di economia (Corti 2012). Tra i materiali rinvenuti ai Campi Macri un gruppo di ceramiche pergamene ben documenta la natura della frequentazione di questa sede di mercato, del tutto eccezionale nel contesto norditalico, testimoniando al contempo il suo pieno inserimento nella rete commerciale ad ampissimo raggio attraverso, con ogni probabilità, l’intermediazione della piazza commerciale di Delo (Parra 1983: 95‒96, 101). Si tratta di una decina di frammenti di ceramica verniciata con rilievi applicati prodotta a Pergamo dalla metà del II sec. a.C. agli anni 60 del secolo successivo (Figura 2). Tra queste ceramiche compare anche un frammento di matrice con scena erotica (Parra 1983: fig. 89). Questi prodotti non risultano al momento diffusi al di fuori dei Campi 50

Merci e sedi di mercato: lana e produzione ceramica a Mutina e ai Campi Macri (Modena, I)

Figura 4. Lucerne di tipo pergameno (Herzblattlampen) da Magreta (Modena, Museo Civico, Archivio fotografico).

Macri, ma la matrice mette bene in evidenzia il carattere imprenditoriale di tale presenza.

Portorecanati, ben testimonia il successo commerciale che ebbe, non solo in Italia settentrionale, questo tipo di lucerna (Grassi 2013: 215‒219). È stato poi messo in evidenza, in via indiziaria e senza poter istituire un nesso diretto, come membri delle gentes Valeria (VAL = Valeri?) e Acutia, altra famiglia attestata nell’area dei Campi Macri, compaiano tra i negotiatores italici presenti in ambito greco ellenistico e a Delo tra il II e il I sec. a.C. (Ortalli 2009: 84‒85).

Ai Campi Macri il dato archeologico attesta la nascita di un fiorente indotto legato alla produzione ceramica (Parra 1983 e da ultimi Labate et al. 2009). In particolare, se l’avvio di una manifattura a imitazione delle ceramiche pergamene pare proprio non aver avuto successo, altrettanto non si può dire per le lucerne. Qui sono stati infatti realizzati dalla metà I sec. a.C. esemplari chiaramente derivati dalle Herzblattlampen pergamene, come ben documentano frammenti di matrici e alcuni scarti di produzione (Figure 3‒4). Aveva in particolare sede ai Campi Macri l’attività ceramica di un personaggio che si firma VAL, come documentano l’unico frammento di matrice con marchio finora rinvenuto e alcuni positivi (Figura 5). L’ampia diffusione di questa produzione, che raggiunse Milano, Gazzo Veronese, Adria, Aquileia e

È nell’ambito dell’indotto dei Campi Macri che, in base ai rinvenimenti e ai dati cronologici finora disponibili, venne quindi avviata la produzione di lucerne nel Modenese. La funzione dei Campi Macri non fu però solo quella di centro precursore, ma anche di motore economico-imprenditoriale. In stretta relazione topografica e successione cronologica si collocano infatti le produzioni di Cittanova (Ortalli 2009: 83-84; 51

Carla Corti Labate 2016: 19, 23). Il sito, che ha restituito ben tre fornaci e numerose matrici, era posto nei pressi di un santuario ubicato direttamente sulla via Aemilia in strettissima relazione, anche temporale, con i Campi Macri (Ortalli 2012: 200, 205‒206). Da questo punto della strada consolare, importantissimo asse viario regionale, un cardine della centuriazione conduceva infatti direttamente alla sede del mercato, che distava solo 5 km. Solo successivamente compaiono le altre produzioni variamente distribuite sul territorio di Mutina, come ad es. testimoniano le matrici di Firmalampen di Savignano sul Panaro (Michelini 1988). La produzione di lucerne continuò verosimilmente fino alla cessazione del mercato, che dovette avvenire prima della metà del I d.C. (vedi supra). Sono stati infatti rinvenuti sia un frammento di matrice di lucerna a volute (Labate et al. 2009: fig. 472), che un frammento di matrice di Firmalampe (Figura 6). In particolare, quest’ultimo rinvenimento pare attestare infine come dovette avere precocemente sede qui anche una delle manifatture di maggior successo del Modenese, le Firmalampen, appunto (Parra 1983: 91‒92). Lana e produzione ceramica: possibili dinamiche economiche e commerciali Ai Campi Macri per tutta l’età repubblicana ebbe luogo una fiera del bestiame di importanza economica panitalica, in grado di richiamare un vasto pubblico, come indirettamente documenta la presenza di ceramica pergamena, disposto anche a lunghi viaggi per acquistare i rinomati ovini dalla pregiata lana, come nel caso di Turranius Niger (vedi supra). Membri della gens Turrania ricoprirono a più riprese la prefettura dell’Annona, vantando legami di amicizia con la famiglia imperiale (Buonopane e Corti 2017: 210). Si tratta di pochi ma significativi indizi che ci consentono di intravedere quello che doveva essere stato il «giro di affari» legato all’economia della lana, l’inserimento nei grandi circuiti commerciali, il livello della frequentazione del mercato e la sua capacità di attrarre capitali, anche umani. In tal senso deve essere verosimilmente letto il fatto che nel 56 d.C. l’area in cui si teneva il mercato risulta, almeno in parte, in mano a possessores italici (Ortalli 2012: 207; Corti 2012: 214). Non stupisce quindi che abbia potuto avere luogo proprio qui, nel vasto indotto generato dal mercato,8 l’avvio della produzione fittile di lucerne a Mutina, sfruttando le risorse disponibili localmente (argilla, acqua e legname) e la rete commerciale esistente per la distribuzione delle merci. Si tratta di un’operazione

Figura 5. La produzione di lucerne di tipo pergameno (Herzblattlampen) con marchio VAL: frammento di matrice (disegno) e positivi da Magreta (Modena, Museo Civico, Archivio fotografico; disegno tratto da Labate et al. 2009, fig. 473, 1).

Oltre alle produzioni ceramiche, tra cui anche coroplastica, nessuna delle quali ebbe però il successo commerciale delle lucerne, fece molto probabilmente parte dell’indotto dei Campi Macri anche l’attività della fornace di Torre Oche, sorta nelle vicinanze, in cui si produssero anfore per la commercializzazione del vino locale (da ultimo Corti 2017b). 8 

Figura 6. Frammento di matrice di Firmalampe da Magreta (Modena, Museo Civico, Archivio fotografico).

52

Merci e sedi di mercato: lana e produzione ceramica a Mutina e ai Campi Macri (Modena, I)

spiccatamente imprenditoriale, come testimonia anche il tipo di prodotto su cui si è scelto di puntare (con l’adozione della tecnica della produzione a matrice, vedi supra). Alla cessazione del mercato, non più funzionale nel riassetto post-augusteo (Ortalli 2012), la sua rete commerciale e di relazioni venne assorbita da Mutina, che rimase leader soprattutto nel settore tessile, fino alla tarda età imperiale, come testimonia l’Editto dioclezianeo, dove le lane di Mutina sono le più care, costando una sola libbra 300 denari, quanto 750 lucerne fittili (Giacchero 1974). Proprio questo sbilanciatissimo rapporto tra i costi di lana e lucerne, quindi anche tra i ricavi e le possibilità di guadagno, tracciato per l’inizio del IV secolo ma che possiamo ritenere sostanzialmente valido anche per l’epoca precedente, deve essere tenuto ben presente nel valutare l’economia legata alla produzione e commercializzazione di queste merci,9 dato l’incolmabile squilibrio che presenta la loro tracciabilità nella documentazione archeologica.

consideriamo anche l’elevato fattore di rischio che accompagnava i commerci nell’antichità. Questo processo di dislocazione ha riguardato in particolare le Firmalampen prodotte nel Modenese, come ben documentano il rinvenimento di matrici con marchi di produttori attivi a Mutina e le analisi archeometriche.10 Le produzioni si avvicinarono così ai mercati di consumo transalpini attraverso l’avvio di filiali, come prospettato per Eucarpus (Auer e Sitz 2014: 101‒105), o di produzioni indipendenti. Un ruolo di primo piano ebbero senz’altro le necessità legate all’approvvigionamento dell’esercito (Auer e Sitz 2014: 100), come testimonia più in generale anche il recente rinvenimento di una matrice di Firmalampe in una fornace di Vindonissa (Vienna) in cui si producevano mattoni con marchi legionari (Mosser 2018: 177, fig. 16). In conclusione, la rete commerciale e il vasto circuito di relazioni generato dai Campi Macri hanno creato nella tarda età repubblicana il clima imprenditoriale idoneo per l’avvio di nuove produzioni, gettando le basi del successo della produzione ceramica a Mutina. Inizialmente al traino del settore laniero, per essa dovettero così realizzarsi le condizioni ideali per un proficuo inserimento nella rete degli scambi e il consolidamento del favorevole riscontro sui mercati, fino ad arrivare in molti casi a un’indipendenza produttiva-commerciale con la dislocazione in ambito provinciale della fabbricazione (Firmalampen). La produzione di lucerne dovette continuare con l’adozione di nuove forme ancora nel pieno II sec. d.C., come documenta un raro esemplare di forma Loeschke XX rinvenuto a Heidelberg (D), attribuito su base archeometrica a Mutina (Schneider e Hensen 2009: 82, fig. 3). In questo caso tuttavia non si può certo parlare di successo commerciale, ma tutt’al più di merce occasionale di accompagno. Il clima commerciale sui mercati di consumo per questo tipo di prodotto realizzato in ambito mutinense appare ormai, nella piena età imperiale, sostanzialmente cambiato.

Pur nell’impossibilità di seguire in parallelo la diffusione dei prodotti, è possibile tuttavia fare qualche osservazione sulle dinamiche economico-commerciali che si sono generate. Molto interessante a questo proposito è il rinvenimento a Gazzo Veronese di un cospicuo gruppo di lucerne (21 esemplari derivati dalle Herzblattlampen pergamene) ricondotte alle prime produzioni modenesi legate ai Campi Macri, come testimonia anche il marchio VAL (Grassi e Mandelli 2012). Si tratta infatti di lucerne chiaramente destinate alla commercializzazione, possibile sia verso est (Este, Altino, Adria e Aquileia), che verso ovest (Cremona, Milano, Ivrea e Aosta) (Grassi 2013: 218). Ci troviamo lungo il frequentato asse viario Mutina-Verona, il cui prolungamento porterà direttamente sul Limes retico, la c.d. via Claudia Augusta Padana (da ultimi Basso et al. 2016). Lungo questa direttrice circolarono anche tessuti, come documenta l’etichetta in piombo rinvenuta ad Auerberg (D) (Ulbert 1975: 422, fig. 15). Non fu certo l’ambito modenese a produrre in via esclusiva questo tipo di lucerne, ma se ad esso va riconosciuta, come sembra, una primogenitura (Grassi 2103: 216‒218), è molto probabile che alla base del processo di dislocazione geografica delle produzioni ci sia la scarsa sostenibilità economica nella distribuzione sulla lunga distanza di questo tipo di prodotti. L’incidenza del costo del trasporto, soprattutto se terrestre, sulla redditività del commercio era infatti molto alta per merci di bassissimo valore economico, come le lucerne (cfr. quanto osservato in Tchernia 2011: 124‒131). I margini di guadagno si riducevano poi notevolmente e rischiavano di annullarsi se

Sul rinvenimento di matrici in ambito provinciale già Buchi 1975: XL‒XLI, 70 (Fortis), 149 (Strobilus); sull’individuazione delle produzioni provinciali con marchi di produttori modenesi, tra cui Fortis, Strobilus e Eucarpus, su base archeometrica si veda quanto osservato da Gerwulf Schneider in Auer e Sitz 2014: 92‒99, Schneider e Wirz 1992 e Schneider e Hensen 2009. 10 

Il mercato per i prodotti tessili si amplia notevolmente se si considera che accanto ai prodotti di prima qualità vi erano le seconde scelte (dal vello di una pecora si ricavano infatti lane di qualità diversa). 9 

53

Carla Corti Forte, M. 1988. Lucerne, in Modena dalle origini all’anno Mille. Studi di archeologia e storia, II: 105‒123. Modena: Edizioni Panini. Frayn, J.M. 1993. Markets and fairs in Roman Italy. Their social and economic importance from the second century B.C. to the third century A.D. Oxford: Clarendon Press. Gabba, E. 1975. Mercati e fiere nell’Italia romana. Studi classici e orientali 24: 141‒163. Giacchero, M. (ed.) 1974. Edictum Diocletiani et collegarum de pretiis rerum venalium. Genova: Istituto di storia antica e scienze ausiliarie. Giordani, N. 2001. L’economia del territorio di Mutina in età romana, in C. Corti e N. Giordani (eds) Pondera. Pesi e Misure nell’antichità: 253‒270. Modena: Libra 93. Grassi, E. 2013. Le lucerne, in C. Guarnieri (ed.) Vivere a Forum Livi. Lo scavo di via Curte a Forlì: 213‒225. Bologna: Ante Quem. Grassi, E. e C. Mandelli 2012. Una produzione “ellenistica” in Norditalia: imitazioni di Herzblattlampen pergamene, in Le luminaire antique. Lychnological Acts 3. Actes du 3e Congrès International d’études de l’ILA Université d’Heidelberg (Monographies Instrumentum 44): 129‒136. Montagnac: Éditions Mergoil. Labate, D. 2016. Mutina fecit. Dalle Herzblattlampen alle Firmalampen: nuovi dati sulla produzione di lucerne a matrice dal territorio di Modena, in G. Lipovac Vrkljan, I. Ožanić Roguljić e M. Ugarković (eds) Rimske i kasnoantičke svjetiljke. Proizvodnja i distribucija, kontakti na Mediteranu: 18‒37. Zagreb: Institut za arheologiju. Labate, D. 2017. Mutina: la produzione ceramica, in L. Malnati, S. Pellegrini, F. Piccinini e C. Stefani (eds) Mutina splendidissima. La città romana e la sua eredità: 228‒231. Roma: De Luca Editori d’Arte. Labate, D., R. Mussati e C. Stoppani 2009. FO 40, FO 47, FO 104, FO 106, FO 107, FO 158, FO 159. Podere Ceci, Campo Consiglio (1844), Podere Gazzuoli-Magiera (1870), in A. Cardarelli e L. Malnati (eds) Atlante dei Beni Archeologici della Provincia di Modena, III. Collina e Alta Pianura, 2: 283‒290. Firenze: All’Insegna del Giglio. Labate, D. e N. Raimondi 2010. Modena, Viale Reiter. Impianti produttivi di età romana. Atti e Memorie della Deputazione di Storia Patria per le Antiche Provincie Modenesi XXXII: 334‒336. Lo Cascio, E. (ed.) 2000. Mercati permanenti e mercati periodici nel mondo romano. Atti degli incontri capresi di storia dell’economia antica (Capri 1997). Bari: Edipuglia. Michelini, C. 1988. Il territorio di Savignano sul Panaro in età romana: le testimonianze ottocentesche e i materiali, in Modena dalle origini all’anno Mille. Studi di archeologia e storia, I: 534‒547. Modena: Edizioni Panini. Mosser, M. 2018. Neues zur römischen Legionsziegelei in Hernals, Die Grabung Wien 17, Steinergasse 17. Fundort Wien. Berichte zur Archäologie 21: 166‒181. Ortalli, J. 2009. Modena e il suo territorio: fisionomia e peculiarità di una colonia romana, in A. Cardarelli

References Andreau, J. 1991. Mercati e mercato, in G. Clemente, F. Coarelli e E. Gabba (eds) Storia di Roma, 2, L’impero mediterraneo, II. I principi e il mondo: 367‒385. Torino: Giulio Einaudi Editore. Auer, M. e S. Sitz 2014. Die Firmalampenproduktion des EVCARPVS. Eine Studie zur römischen Keramikproduktion. Germania 92: 85‒125. Basso, P., V. Grazioli, M. Pavoni e E. Zentilini 2016. La via Claudia Augusta: recenti indagini archeologiche dell’Università di Verona a Gazzo Veronese (Verona).

Buchi, E. 1975. Lucerne del Museo di Aquileia, I-II. Aquileia: Associazione Nazionale per Aquileia. Buonopane, A. e C. Corti 2017. Produzioni e commerci a Mutina, in L. Malnati, S. Pellegrini, F. Piccinini e C. Stefani (eds) Mutina splendidissima. La città romana e la sua eredità: 208‒213. Roma: De Luca Editori d’Arte. Calzolari, M. 2008. Città dell’Aemilia: Mutina. Le fonti letterarie di Modena romana. Modena: Editore G. S. B. M. Capellini, S. e C. Corti 2020. Fiat Lux. Le lucerne del Museo “Gaetano Chierici” di Paletnologia di Reggio Emilia (Fonti e studi 9). Reggio Emilia: Deputazione di Storia Patria per le Antiche Province Modenesi. Ceci, M. e G. Schneider 1994. Analisi chimiche su gruppi di lucerne bollate di fabbricazione urbana, in Epigrafia della produzione e della distribuzione, Actes de la VIIe Rencontre franco-italienne sur l’épigraphie du monde romain (Rome 1992): 433‒446. Rome: École Française de Rome. Corti, C. 2004. L’ager nord-occidentale della città di Mutina (Studia archaeologica 126). Roma: L’«Erma» di Bretschneider. Corti, C. 2012. L’economia della lana a Mutina, in M.S. Busana e P. Basso (eds) La lana nella Cisalpina romana. Economia e società. Studi in onore di Stefania Pesavento Mattioli, Atti del Convegno (Padova-Verona 2011): 213‒229. Padova. Corti, C. 2016. L’economia della lana nella Gallia Cisalpina, in D. Vitali e C. Goudineau (eds) Il mondo celtico prima e dopo la conquista romana, Studi in onore di Jean-Paul Guillaumet: 175‒188. Bologna-Dijon. Corti, C. 2017a. Lanam fecit. Economia della lana e società a Mutina, in L. Malnati, S. Pellegrini, F. Piccinini e C. Stefani (eds) Mutina splendidissima. La città romana e la sua eredità: 239‒243. Roma: De Luca Editori d’Arte. Corti, C. 2017b. Le anfore della fornace di Torre delle Oche e la commercializzazione del vino in età repubblicana, in L. Malnati, S. Pellegrini, F. Piccinini e C. Stefani (eds) Mutina splendidissima. La città romana e la sua eredità: 285‒287. Roma: De Luca Editori d’Arte. de Ligt, L. 1993. Fairs and markets in the Roman Empire. Economic and social aspects of periodic trade in a preindustrial society (Dutch Monographs on Ancient History and Archaeology, Volume 11). Leiden: Brill. 54

Merci e sedi di mercato: lana e produzione ceramica a Mutina e ai Campi Macri (Modena, I)

e L. Malnati (eds) Atlante dei Beni Archeologici della Provincia di Modena, III. Collina e Alta Pianura, 1: 76‒86. Firenze: All’Insegna del Giglio. Ortalli, J. 2012. I Campi Macri. Un mercato panitalico sulla via della lana, in M.S. Busana e P. Basso (eds) La lana nella Cisalpina romana. Economia e società. Studi in onore di Stefania Pesavento Mattioli, Atti del Convegno (Padova-Verona 2011) (Antenor quaderni 27): 195‒211. Padova : Padova University Press. Parra, M. C. 1983. La fornace di Magreta, in S. Settis e M. Pasquinucci (eds) Misurare la terra: centuriazione e coloni nel mondo romano. Il caso modenese. Modena: Edizioni Panini. Pucci, G. 1977. Le terre sigillate italiche, galliche e orientali, in L’instrumentum domesticum di Ercolano e Pompei nella prima età imperiale (Quaderni di cultura materiale 1): 9‒21. Roma: L’«Erma» di Bretschneider. Schindler Kaudelka, E. e S. Zabehlicky-Scheffenegger 2000. Es war der Schadensbrand und nicht der Fehlbrand. Rei Cretariae Romanae Fautorum Acta 36: 571‒574.

Schneider, G. e E. Wird 1992. Chemische Analyse von Firmalampen aus Vindonissa. Jahresbericht/ Gesellschaft Pro Vindonissa 1992: 35‒49. Schneider, G. e A. Hensen 2009. Chemische Analysen an Öllampen, in A. Hensen, Das römische Brandund Körpergräberfeld von Heidelberg, 1 (Forsch. u. Ber. Baden-Württemberg 108): 79‒85. Stuttgart: Landesamt für Denkmalpflege im Regierungspräsidium Stuttgart. Schneider, G. e M. Daszkievic 2011. Imported and local Firmalampen in Aquieia: chemical analyses by WDXRF. Aquileia Nostra 82: 261‒282. Tchernia, A. 2011. Les Romains et le commerce. Neaples: Centre Jean Bérard. Ulbert, G. 1975. Der Auerberg, in Ausgrabungen in Deutschland 1950-1975, 1: 409‒433. Mainz: Verlag des Römisch-Germanischen Zentralmuseums. Vicari, F. 2001. Produzione e commercio dei tessuti nell’Occidente romano (British Archaeological Reports International Series 916). Oxford: Archaeopress.

55

Pottery and brick production in northeastern Italy in the Roman Period: the exploitation of clay quarries Cristina Mondin Abstract In the territory of the Decima regio, numerous pottery and brick production centres were unearthed in the last two centuries. Some of them revealed further information on the production process adopted in the Roman period. In this paper, I will consider the quarrying of clay for the production of pottery and bricks. However, few are the archaeological remains of this activity, and, thus, information from iconographic and literary sources that talk about the craftsmanship practice are also presented. Classical literary sources are limited, and they described clay quarrying activities in two different occasions for material collection: Herodotus mentioned clay collected by the Nile river, while Varro underlined all the potentials of the fields have to be tapped, including exploiting clay quarries. Pottery and bricks production did not change significantly in the centuries, especially regarding the exploitation of the clay, so it is helpful to consider also the Medieval sources and Renaissance accounts. Among them, the most detailed description of the quarries was written by Cipriano Piccolpasso around the mid-1500 in Li tre libri dell’arte del vasaio. The author mentioned both river and the plain area quarries, as they are also described by classical sources. Archaeological data in the Decima regio area confirm the use of these two types of quarries. River quarries are the most difficult to recognise because of the changing landscape of the bottomland over time. However, in the Lunardelli production site at Chiavornicco, in the municipality of Cordenons (Udine, Italy), traces of Roman structures has been discovered, close to the site, in the bottomland of Meduna river and they could be associate to the activity of exploitation of the clay. More attested are the plain area quarries and some of them have been partially excavated. Usually, these quarries are quite close to the production sites, however they covered a big area based on dimension of the production site and the long life of the production. The biggest quarry areas discovered in Decima regio are the sites of Olmeneta (Cremona, Italy), Mirano (Venezia, Italy) and Carlino (Udine, Italy). In these contexts the quarries closer to the structures have been filled up with production waste. The farther pits have been left empty. My paper will discuss literary and archaeological evidence from the Decima regio in order to shed new light on this important section of pottery and brick production. Key words Pottery, brick, production, Roman period, clay quarry, classical sources, Renaissance accounts.

Introduction In the territory of the Decima regio Venetia et Histria (western part of Lombardy, Veneto, Friuli Venezia Giulia and Istrian peninsula), numerous pottery and brick production centres were excavated in the last two centuries.1 In some cases, the extensive archaeological investigations also revealed further information on the structures adjacent to such centres. In this work, I will consider the evidence available for clay quarries in the Roman period. Indeed, archaeological remains of quarries are limited, and some hypothesises can be proposed with the aid of ancient literary sources as well as Renaissance accounts dedicated to pottery and brick craftsmanship. The geographical context taken This paper is based on my Phd thesis discussed at the University of Padua in 2010 (Mondin 2010). The production sites studied are dated between the 2nd century BC and the 5th century AD. I would like to thank Paolo Maranzana for proofreading the text and English. 1 

into consideration in this paper is the Decima regio area, modern north-eastern Italy. It could be divided in three morphological macro areas: the mountains and hills on the North and East, the high plain characterized by gravel deposits, and the low plain characterized by clay and silt deposits. The border and geographical characteristics of Decima regio’s area are described by Pliny the Elder, in the third book of the Naturalis Historia (III, 130–132). Regarding the East border, the limit described by Pliny, between Italy and Pannonia, was the Carso Triestino and also included part of the Istrian peninsula (in today’s Slovenia and Croatia). However, the discovery of a Roman inscription close to Ljubljana allows us to extend the limit of the Decima regio to the East and to include ancient Emona, from the 1st century AD (Šašel 1992; Šašel Kos 2002: 253; Šašel Kos 2003: 13-14). Slovenian scholars argue that the apparent discrepancies in literary and epigraphic accounts (Pliny the Elder, Naturalis Historia, III, 147; Ptolemy, Geographia,

Roman Pottery and glass manufactures (Archaeopress 2022): 56–64

Pottery and brick production in northeastern of Italy in the Roman Period

Figure 1. Pinax F871B (from: Vidale 2002: 242, fig. 44). II, 14, 7 and VIII 7, 6) should be explained as follows: literary sources consider the geographic border of the Decima regio, while epigraphic documents defined the administrative border. In this paper, I will consider the geographic definition.

the 7th to the first quarter of the 6th century BC (Vidale 2002: 237). One of the earliest testimonies available is the pinax F871B (Figure 1). The interpretation is not clear, but the image shows four workers: two senior men are extracting raw material from a pit, and two apprentices are moving baskets (Cuomo di Caprio 1984; Stissi 2002: 78-79; Vidale 2002: 238-241). The image focuses on the workers, but it is interesting to observe the structure of the quarry; the access to the pit was guaranteed by stairs, which are visible on the left side of the image. The roof is a simple black line, and an amphora tied with ropes hangs from it. This element allows two different interpretations. The first is that this is the representation of a metal quarry (Vidale 2002: 238, note 7); the second is that the clay pit was covered with a canopy. Other pinakes show the same activity, however they are incomplete (Vidale 2002: 240-241).

Iconographic and literary sources Clay quarries are usually forgotten in studies dedicated to pottery and brick productions. This is probably due to two different reasons: the first is that literary sources rarely mention this type of use for the fundus; the second is that rare are the archaeological remains of these activities. Thus, I will start my contribution by reviewing the evidence available for quarrying activities in the Roman period in this region. Greek and Roman literary sources rarely mention clay quarries. However, several terracotta slabs named pinakes, found close to Corinth in the 1800s, were dedicated to the representation of pottery production (Vidale 2002: 237-256). Among them, some terracotta slabs were interpreted as the representation of clay quarries. These pinakes are dated between the half of

Another representation of possible clay quarries is dated to the Roman period. It is a funerary monument found at Aquileia, which scholars associate to the production practices of the amphorae type Dressel 6B (Buchi 1987: 158; Zaccaria and Pesavento Mattioli 2009: 286). The three façades of the stone show one 57

Cristina Mondin

Figure 2. Funerary monument found at Aquileia (from: Buchi 1987: 158).

amphora transported by a worker or a dealer, a storage of amphorae, and two hoes with short handles (Figure 2). In detail, the hoes were interpreted as elements connected with clay mining. The shape of the two instruments is similar to those represented in the described pinakes.

clay with water and kneading it, have been carried out in the same way (Peacock 1982; Cuomo di Caprio 2007). For this reason, it is essential to read also the Medieval/ Byzantine sources and the Renaissance essays in order to shed new light on quarrying of clay in the ancient past. During the 10th century, the Byzantine emperor Constantin VII instructed Cassianus Bassus to write a book regarding the use of fields. In the second book of Geoponica, he wrote that “in every land you can find clay”, thus underlining that pottery and brick productions could be organised everywhere. In the description, there is also some information on quarries; in fact, Cassianus Bassus wrote that clay for pots could be found on the surface, deep down or in hidden nooks (Cassianus Bassus, Geoponica, II, 49, 3. Comments in Cuomo di Caprio 2007: 700). In the fourth book, Cassianus Bassus wrote that not all soils are suitable for pottery. He also specified that some craftsmen prefer red or white clay, while others mixed both. He even wrote a few phrases on the exploitation of fields for pottery and brick production (Cassianus Bassus, Geoponica, VI, 3, 1). The description of soils for pottery production is brief and without details. This could mean that the topic was quite well known and did not need deepening.

Among the written sources, Herodotus reported what he read on a king inscription dedicated to the construction of a pyramid in Egypt: …for they struck a bole down into a marsh and collected what mud clave to the pole; therewith they made bricks, and thus was I built (Herodotus, II, 136, 4, translation Godley 1920: 441). The Greek writer described the use of a pole for mining mud from the Nile River or its canals. Even today pottery and brick productions in Egypt uses the Nile silt for bricks, and also pottery production. Few information about clay quarries come from De re rustica written by Varro (I, 2, 23) in the 1st century BC. In his book on the fundus, he illustrated the methods for exploiting the fields. In detail, he discerned what to consider as proper for agriculture, and he underlined that all the potentials of the fields have to be tapped, including exploiting clay quarries, named figlinae. Other latin authors like Cato (De agri cultura, I, 7) or Palladius (Opus agriculturae, VI, 12) do not mention clay quarries even though they consider the production of bricks.

There are two Renaissance essays about craftsman production in Italy which mentioned terracotta production too. The first is De la pirotechnia, written by Vannoccio Biringuccio and published after his death in 1540. The main important topic of this book is metal work, but it concentrates mainly on cooking and fusion processes. For this reason, pottery, named l’arte figulina, and brick production is also included in the treatise. Few are the notes about quarries. The author specifies that clay is named terra, terra creta, or arzilla, and it is cavata or extracted when the craftsmen need it (Biringuccio

The production of pottery and bricks has a long tradition and the two most important technical developments introduced were the use of the wheel for production of vessels (Cuomo di Caprio 2007: 175190) and the evolution of kilns (Mondin 2017: 50-59). On the other hand, the other craftsmanship activities, such as exploitation and crumbling of clay, softening 58

Pottery and brick production in northeastern of Italy in the Roman Period

Figure 3. Lunardelli kiln of Chiavornicco in the town of Cordenons (PN): reconstructive section of the site (from: Spanghero 1999: 451-452).

1540: 145). Even in this case, the lack of details allows us to hypothesise that the clay could be easily found.

conducted in several sites of Lombardia, Veneto and Friuli Venezia Giulia allowed us to recognise two different types of quarries: river and plain area quarries. The same type of quarries described by sources and essays.

Entirely dedicated to pottery production is the book Li tre libri dell’arte del vasaio, written by Cipriano Piccolpasso around the mid-1500s. In the first chapter of the first book, he described four different methods to mine clay (Piccolpasso 1976: 35-38). The first method is to collect it from river banks as clay accumulated along the riverside after the rain season, especially frequent during the summer. The second method is to collect earth from the surface clay deposits, which are more common in the plain areas. In particular, Cipriano Piccolpasso mentioned one area close to Padua named Battaglia. The author underlined that the best clay is the one that is collected already clean. In this case, the clay does not need a long process of purification. The third way to collect clay is one used in Corfù. After the rain season, clay accumulates at the base of the marl hills (badlands), which are characterised by the absence of vegetation and trees. Clay was then broken up, and it became useful for pottery production.2 The last method for clay collection is a mix of soil collection and purification. Piccolpasso described a system of cisterns placed at different elevations connected by water channels which allowed to carry clay dissolved in water. The author said that this system was in use in Spello, in Umbria. Based on this book in the Renaissance in Italy three different types of soils were in use for pottery and brick productions: river clay, plane areas, and marl hills where clay is collected in various methods (Piccolpasso 1976: 35-38).

River quarries Traces of river quarries are extremely rare in the archaeological record. Just as discussed in the previous section, clay quarries located along river fade away with new rain. Cipriano Piccolpasso underlined that clay collected from the river is better than the other types of quarries. This is due to the characteristics of the soil. First of all, these types of clay depsits are simple to recognise in the bottomland. Moreover, this kind of clay is fine, soft, breaks up naturally and there are fewer plants inside the clay than in the other types of quarries. Finally, the mining area will be regenerated after every new rain season. In the Decima regio area, only in the case of the Lunardelli production site it is possible to hypothesise the use of the Meduna river clay collected from a bottomland. The site is in Chiavornicco, in the municipality of Cordenons, where a Roman kiln was excavated at the beginning of the 1900s. The kiln was located on a Meduna river dump modelled during the post-glacial era. Few archaeological remains were discovered in the area of the kiln; only traces of a worn bridge made with Roman bricks were found in the river bottomland, between the river and the production site. In the same bottomland, two artificial water channels were identified. The biggest, close to the Meduna river, was parallel to the river in the first part, and the water channel goes back to the river at the end of the bottomland (Figure 3). Thanks to the Roman infrastructure found in this context, it is possible to hypothesise that clay for pottery production was picked up form the river. The passage of the water after the overflow of the Meduna was probably regulated by the channels (Serafini 1992: 66; Spanghero 1999).

The archaeological sites in North-East Italy Archaeologically recognised quarries are rare, which depends on a variety of reasons; first of all, quarries are just big pits in the countryside, and therefore it is hard to identify them in absence of structures built in the vicinity. Additionally, they could be quite far from the production sites, which makes it hard to predict their locations. However, recent archaeological research

Looking at the map of Padua in Roman time, it is very likely that the pottery production sites were built

In the Alps and pre-Alps area, badlands are rare because the composition of rocks is different from those in the Apennine mountains. 2 

59

Cristina Mondin

Figure 4. Pottery production sites at Padua: 1, Via San Pietro (Balista, Ruta Serafini 2001: 101–103); 2, Piazza Castello (Ruta Serafini, Sainati and Vigoni 2006); 3, Via Montona (Cozza and Ruta Serafini (eds) 2007); 4, Via Giustiniani (De Vanna and Ruta Serafini 1995: 21–24); 5, the clay pits in Via Gattamelata (from: Cipriano et al. 1998: 161–164).

along to the river banks (Figure 4),3 even though no archaeological evidence can confirm this. However, in the site of via San Pietro (Balista and Ruta Serafini 2001: 101-103), a series of rectangular basins were excavated along the riverside, which was interpreted as part of the clay production area. In this case, it is possible to hypothesise that the clay quarry was the river. These quarries were probably used together with the Padana plain quarries, which provide clay suitable for pottery production.

size”, as clay quarries due to the waste from the firing process and empty amphorae for drainage (De Vanna and Ruta Serafini 1995: 21; De Vanna 1998: 181-182).4 In Gattamelata street (Cipriano et al. 1998: 161–164), some pits were interpreted as quarries for pottery, but in this area, no production evidence was documented: pits had been filled with empty amphora for the drainage system (Cipriano et al.1998: 161-162). Clay deposits in the low plain in North Italy are actually common. Clay quarries in this areas were not used only for pottery and brick production; they were also used to seal the old levels of the town from the new life levels.5

Plain area quarries At Padua, in the archaeological excavation at Via Giustiniani (De Vanna and Ruta Serafini 1995: 21-24), a series of squared and rounded pits were found. In the same area, a burned platform was excavated, and it was interpreted as the remains of a pottery kiln. Archaeologists interpreted the pits, of “considerable

Few are also the quarries identified in the extra-urban areas. In the site of Olmeneta, in the hinterland of Cremona, archaeological research allowed to identify two kilns and numerous pits interpreted as quarries (Passi Pitcher 2003; Bonora Mazzoli 2008: 75). During the first decades of the 1900s, these archaeological remains were levelled. The Roman floors were destroyed, so we do not have the dimensions of the pits, and the organisation of the quarry area is still unclear. The bestpreserved quarries were located in the north sector

Not only at Padua pottery and brick productions are attested along the rivers. At Cremona the pottery production site of via Platina was built along the Po river (Bertolone 1960: 266). At Altino the pottery district was placed along the Sioncello river (Cipriano and Sandrini 1998). Many brick production sites studied in Friuli Venezia-Giulia were built along rivers which were probably used both for quarries and as commercial routes (some examples: Ronchis, Rivignano, Carlino, etc. – Ventura et al. 2011). I would like to thank the editor of the text Dr. Paolo Maranzana. 3 

Urban plain area quarries are also attested at Bologna: Ortalli 1998: 72-73. 5  Some examples are attested at Padua and Altino (Cipriano et al. 1999: 34; Cozza and Ruta Serafini (eds) 2007: 76, 93). 4 

60

Pottery and brick production in northeastern of Italy in the Roman Period

Figure 5. Site of Carlino with the quarry area: 1. Fields; 2. Woods; 3. Clay quarries; 4. Pits; 5. Roman pottery production site (from: Fontana 2006: 235, fig. 8.23).

of the production area, which were almost 30 meters north of kilns. Probably the quarry area continued beyond the limit of the archaeological excavation, but the size of the trenches did not allow to dig them all. Only two quarries were partially excavated, which revealed vertical sides filled with production waste.6

only filled when the area was reclaimed for agricultural purposes after the Roman period (Valle 2010: 75-76). Only one pit in the plain was identified in the brick production area of Adro, in the province of Brescia. In this case, a rectangular pit was found 65 meters away from the kiln; it was interpreted as a quarry and partially excavated. It was filled with production waste such as slag and part of the kiln’s walls. Chemical analysis were carry out both on bricks and clay from the pit; they confirm that local clays were used for brick production (Breda and Finzi 1991; Breda (ed.) 1993). In the municipality of Borgoricco, close to Padova, in Santa Maria delle Badesse, a circular pit filled with production waste was excavated. The diameter of the quarry was about 10 meters (Bonomi and Cipriano 2012). Quarries are also reported at Vallenoncello, in the province of Pordenone (Cividini and Ventura 2003; Ventura 2004: 218-223; Amato et al. 2007). At Ronchis, close to Latisana in the province of Udine, a regular series of rectangular pits were identified as quarries. In the bibliography, other pits were interpreted as

More information comes from the excavation of the production area of Vetrego, close to Mirano, in the province of Venice. In an area of 1500 square meters, five quarries and part of a canopy were dug. Quarries were located in the West area of the site and they had different dimensions and shapes. They always had some meters between them; space was probably functional to the passage of carts. All pits were excavated one meter and a half in depth, and quarries closest to the manufacturing area were filled with production waste. The other quarries had probably been left empty and

Regarding the information about quarries I would like to thank Dr. Lynn Passi Pitcher. 6 

61

Cristina Mondin quarries, but identification is not sure (Cividini and Ventura 2009; Ventura and Cividini 2011).

Friuli Venezia-Giulia hill areas. Anyway, only plain area quarries were investigated with archaeological methods and river quarries can be hypothesise thanks to the location of the production sites. However, based on literary sources, river quarries were largely used in the past, but do not leave archaeological remains.

Thanks to further analysis of aerial photos, other quarries in plains were identified in North-East Italy. In 2006, Alessandro Fontana published an aerial photo survey on low plains in territory of Veneto and Friuli Venezia-Giulia (Fontana 2006: 234-237). The main production site where clay quarries were identified is Chiamana, in the municipality of Carlino (Udine, Italy). The production area was excavated in the 1970s and an area of 2000 m2 was dug. This site produced common and glazed ware, and it was in use from the Mid-Roman period until Late Antiquity (Magrini and Sbarra 2005). Close to this site, at a distance ranging from 80 meters to 800 meters, more than a hundred pits were identified with aerial photography. Pits were rectangular (approximately 10 to 15 meter long sides), and ca. 1 meter deep (Figure 5). The topographic survey of this area enabled scholars to realise that the quarries changed the morphology of the hill, and these photos showed the organisation of the quarry area. The quarries area had been divided into two rectangular zones, and the pits had been organised in parallel lines. Fontana argues that the exploitation of the clay started in the part closer to the production area, and it was then extended to North-East in the following centuries.7

Ancient sources Cassianus Bassus, Geoponica. Cato, De agri cultura. Herodotus, Volume I (Books 1 and 2). With an English Translation by A. D. Godley, 1920. London: Loeb Classical Library edition, Harvard University press. Varro. De re Rustica. References Amato, F. Fabbri, B. and Ventura, P. 2007. Produzione di laterizi e terrecotte architettoniche nel pordenonese: il materiale dalla fornace di Vallenoncello, in B. Fabbri, S. Gualtieri and A. N. Rigoni (eds) Materiali argillosi non vascolari: un’occasione in più per l’archeologia (Atti della 9a Giornata di Archeometria della Ceramica – Pordenone 18–19 aprile 2005): 109–122. Pasian di Prato (UD): Lithostampa. Balista, C. and A. Ruta Serafini (eds) 2001. Lo scavo di una parte di un’insula perifluviale: l’area ex Ardor a Padova. Quaderni di Archeologia del Veneto XVII: 99–115. Bertolone, M. 1960. Notiziario, III. 17. Fornace romana a Cremona. Sibrium V: 266. Biringuccio V. 1540. De la pirotechnia. A. Caguro (ed.) 1977. Milano: Edizioni il Polifolio. Bonomi, S. and S. Cipriano 2012. Gli Scavi Archeologici, in C. Mengotti and S. Bortolami (eds) Antico e Sempre Nuovo. L’agro Centuriato a Nord-Est di Padova dalle Origini all’età Contemporanea: 55–59. Caselle di Sommacampagna (VR): Cierre Edizioni. Bonora Mazzoli, G. 2008. Fornaci di età romana per la produzione ceramica e laterizia in Lombardia.Agri Centuriati 4/2007: 63–78. Breda, A. and E. Finzi 1991. Adro (BS). Località Fornaci Quattro Vie. Soprintendenza pe i beni archeologici della Lombardia. Notiziario 1991: 27–28. Breda (ed.) 1993. La fornace romana di Adro, A. Breda (ed.) (Exhibition catalogue). Iseo: Vela Laeterizi. Buchi, E. 1987. Assetto agrario, risorse e attività economiche, in E. Buchi (ed.) Il Veneto nell’età romana I Storiografia, organizzazione del territorio, economia e religione: 103–184. Verona: Banca popolare di Verona. Cipriano, S., M. G. Sandrini, F. Ferrarini and E. Pujatti 1999. L’abitato di Altino in età tardorepubblicana: i dati archeologici, in G. Cresci Marrone and M. Tirelli (eds) Vigilia di romanizzazione. Altino e il Veneto orientale tra II e I sec. a.C.: 33–65. Roma: Quasar.

The layout of the quarrying areas can also be recognised in the site of Casali Pedrina, in the Rivarotta del Teor municipality (Maggi 1998: 82-90). Thanks to another set of aerial photos, numerous pits were identified at this site. Quarries were rectangular and located approximately 100 meters away from the production site (Fontana 2006: 236). Concluding remarks From this overview some interesting information came to light about pottery and brick production in the NorthEast of Roman Italy. Only two types of clay quarries could be identified in the Decima regio: river quarries, where clay is collected from the river bottomland, and plain area quarries. These are also the only types of quarries mentioned in the Greek and Roman literary sources, as seen in Herodotus and Varro. A third method to collect clay is from the hills, which is mentioned only by Cipriano Piccolpasso, a writer from the 16th century AD. It is important to underline that we do not have archaeological remains of this type of quarries in the Roman period, and our understanding of this process comes entire from limited literary accounts. However, today, this method is still in use in the Veneto and Thanks to this method of research it is possible to confirm that the clay quarries have the same characteristics, at least, from the Roman time until the 20th century. In some sites of Friuli Venezia-Giulia, Roman quarry areas were used until the last century (Fontana 2006: 235-236). 7 

62

Pottery and brick production in northeastern of Italy in the Roman Period

Cipriano, S., S. Mazzocchin and P. Pastore 1998. Padova. Tre casi in aree a diversa funzionalità, in S. Pesavento Mattioli (ed.) Bonifiche e drenaggi con anfore in epoca romana: aspetti tecnici e topografici. atti del Seminario di studi, Padova, 19-20 ottobre 1995 (Materiali d’Archeologia 3): 161–174. Modena: Panini. Cipriano, S. and G. M. Sandrini 1998. La villa suburbana e gli impianti produttivi lungo il Sioncello ad Altinum. Quaderni di Archeologia del Veneto XIV: 125–139. Cividini, T. and P. Ventura 2003. Pordenone, località Vallenoncello. Area produttiva di età romana. Aquileia Nostra LXXIV: 810–818. Cividini, T. and P. Ventura 2009. Ronchis (UD). Impianto produttivo. Notiziario della Soprintendenza per i beni archeologici del Friuli Venezia Giulia 2/2007: 57–62. Cozza, F. and A. Ruta Serafini (eds) 2007. I colori della terra. Storia stratificata nell’area urbana del Collegio Ravenna a Padova. Archeologia Veneta XXVII–XXVIII 2004–2005. Cuomo di Caprio, N. 1984. Pottery kilns on pinakes from Corinth, in Ancient Greek and Related Pottery, in H.A.G. Brijder (ed.) Proceedings of the International Vase Symposium (Amsterdam 12 – 15 April 1984): 72–82. Amsterdam: Allard Pierson Museum. Cuomo di Caprio, N. 2007. Ceramica in archeologia 2. Antiche tecniche di lavorazione e moderni metodi di indagine. Roma: L’Erma di Bretschneider. De Vanna, L. 1998. Padova. Le evidenze dell’area della clinica pediatrica di via Giustiniani, in S. Pesavento Mattioli (ed.) Bonifiche e drenaggi con anfore in epoca romana: aspetti tecnici e topografici. atti del Seminario di studi, Padova, 19-20 ottobre 1995 (Materiali d’Archeologia 3): 181–184. Modena: Panini. De Vanna, L. and A. Ruta Serafini 1995. Padova, via Giustiniani. Nuovo Padiglione Pediatrico. Nota preliminare sulle indagini 1993 e 1994. Quaderni di Archeologia del Veneto XI: 18–25. Fontana, A. 2006. Evoluzione geomorfologia della bassa pianura friulana e sue relazioni con le dinamiche insediative antiche. Udine: Edizioni del Museo Friulano di Storia Naturale. Maggi, P. 1998. Presenze romane nel territorio del Medio Friuli. 5. Teor. Tavagnacco: Arti Grafiche Friulane. Mondin, C. 2010, Impianti di produzione ceramica e laterizia in epoca romana: analisi morfologica delle strutture e relazioni territoriali nella decima regio. Unpublished PhD dissertation, University of Padua. Mondin, C. 2017. Impianti produttivi per ceramica nella decima regio: distribuzione topografica ed evoluzione strutturale, in G. Lipovac Vrkljan, B. Šiljeg, I. Ožanić Roguljić and A. Konestra (eds) Roman ceramic and glass manufactures Production and trade in the Adriatic region. Proceedings of the 3rd International Archaeological Colloquium Crikvenica (Croatia), 4-5 November 2014: 41–65. Crikvenica: Crikvenica Town Museum, Institute of Archaeology.

Ortalli, J. 1998. Assetto distributivo e funzionalità dei luoghi di produzione fittile nella Cispadana romana: ‘Bononia’ e il suo territorio, in V. Righini (ed.) Le fornaci romane produzione di anfore e laterizi con marchi di fabbrica nella Cispadana orientale e nell’Alto Adriatico (Atti delle Giornate internazionali di studio, Rimini 17-18 ottobre 1993): 69–87. Rimini: Musei comunali. Passi Pitcher, L. 2003. Archeologia della colonia di Cremona: la città e il territorio, in P. Tozzi (ed.) Storia di Cremona l’Età Antica: 130–229. Azzano San Paolo: Bolis. Peacock, D. P. S. 1982.  Pottery in the Roman World: an ethnoarchaeological approach. London: Longman. Piccolpasso, C. 1976. Li tre libri dell’arte del vasaio, G. Conti (ed.). Firenze: All’insegna del Giglio. Ruta Serafini, A., C. Sainati and A. Vigoni 2006. Lo scavo urbano pluristratificato di Piazza Castello n. 18 a Padova. Quaderni di Archeologia del Veneto XXII: 150– 167. Šašel, J. 1992. Zur verwaltungstechnischen zugehörigkeit Emonas. Opera Selecta (Situla 30): 707–714. Ljubljana: Narodni muzej. Šašel Kos, M. 2002. Il confine nord-orientale dell’Italia romana. Riesame del problema alla luce di un nuovo documento epigrafico. Aquileia Nostra LXXIII: 245– 260. Šašel Kos, M. 2003. Emona was in Italy, not in Pannonia, in M. Šašel Kos and P. Scherrer (eds) The autonomous towns of Noricum and Pannonia (Situla 41): 11–19. Ljubljana: Narodni muzej. Serafini, F. 1992. Rivisitazione di una antica area produttiva di epoca romana: la fornace Lunardelli. Caput Adriae. Bollettino della Federazione Archeologica Friulana II/2: 59–66. Spanghero, T. 1999. Cordenons, loc. Chiavornicco. Controlli archeologici di emergenza in prossimità della fornace romana “Lunardelli”, 1997 e 1999. Aquileia Nostra LXX: 451–455. Stissi, V. V. 2002. Pottery to the people. The production, distribution and consumption of decorated pottery in the Greek world in the Archaic period (650480 BC), PhD thesis, Universiteit van Amsterdam (viewed 31 December 2019). Valle, G. 2010. Lo scavo di impianti produttivi per laterizi (epoca romana e rinascimentale): Mirano (VE), località Vetrego, fondo Bettin, via Porara; Salzano (VE), località Robegano; Preganziol (TV), via Rio Serva km 24.500, in A. Lalli (ed.) Il passante autostradale di Mestre. Una infrastruttura chiave per l’Europa. Il progetto Il territorio L’ambiente: 75–77. Campodarsego: La Grafica Faggian. Ventura, P. 2004. Nuovi dati sulle fornaci del territorio di Iulia Concordia. Histria Antiqua 12/2004: 217–226. Ventura, P. and T. Cividini 2011. L’impianto produttivo di Ronchis di Latisana (UD, Italia), in G. Lipovac Vrkljan, I. Radić Rossi and B. Šiljeg (eds) Roman ceramic and glass manufactures Production and trade in 63

Cristina Mondin the Adriatic region. Proceedings of the 1st International Archaeological Colloquium Crikvenica (Croatia), 23–24 October 2008: 125–132. Crikvenica: Crikvenica Town Museum, Institute of Archaeology. Ventura, P., T. Cividini, P. Maggi and C. Magrini 2011. Il progetto “Antiche fornaci in Friuli” (Provincia di Udine), in G. Lipovac Vrkljan, I. Radić Rossi and B. Šiljeg (eds) Roman ceramic and glass manufactures Production and trade in the Adriatic region. Proceedings of the 1st International Archaeological Colloquium Crikvenica (Croatia), 23–24 October 2008: 115–124. Crikvenica: Crikvenica Town Museum, Institute of Archaeology.

Vidale, M. 2002. L’idea di un lavoro lieve. Il lavoro artigianale nelle immagini della ceramica greca tra VI e IV secolo a.C. (Saltuarie dal laboratorio del Piovego 5). Padova: Imprimitur. Zaccaria, C. and S. Pesavento Mattioli 2009. Uomini e merci, in F. Ghedini, M. Bueno and M. Novello (eds) Moenibus et portu celeberrima. Aquileia: storia di una città: 275–287. Roma: Libreria dello Stato, Istituto poligrafico e Zecca dello Stato.

64

Recent excavation of an amphorae kiln in the artisanal workshop complex at Loron (Tar-Vabriga, Croatia): results of the archaeological and anthracological study Corinne Rousse, Christophe Vaschalde, Gaetano Benčić, Davor Munda Abstract The ceramic production workshop at Loron (Tar-Vabriga, Croatia) is one of the largest known workshops in the northern area of the Adriatic. In 2017, a new excavation lead by the museum of the Poreč territory and the centre Camille Jullian (university of Aix Marseille / CNRS) on the large amphorae kilns of the workhop revealed the best-preserved kiln structure on the site, both in plan and in elevation. Measuring a total of 9 x 5.10m, the FR ξ-8000 kiln is characterized by a long praefurnium (2.5 m) leading to a rectangular heating chamber (6.5 x 5.10m), built of brick and with two parallel corridors. The floor — built according to a particular system of T-shaped bricks — rests on ten series of arches built out of bricks, the dimensions of which reveal a predetermined system of modules and forms, perfectly adapted to the construction of the kiln.  The research also clarifies the fuel supply strategies by applying a systematic sampling of charcoal to allow a dendro-anthracological study, accompanied by sampling for archaeomagnetic dating of the final phase of the kiln, as well as radiocarbon dating. These studies are the first one on the site. The complete excavation of the fourth large amphora kiln at the Loron workshop, combined with the anthracological study and several dating methods, renews the knowledge of the functioning of the workshop, especially for the last phase of production. It is now demonstrated that the large kiln FR ξ-8000 was in use until the end of the 3rd century to the beginning of the 4th, with a production that  corresponds to the late Dressel 6B amphora from Loron. This dating clearly confirms the chronology of the late Dressel 6B amphorae, relying so far on a relatively limited number of contexts, mainly identified in northern Italy. The anthracological study reveals also a well-thought out management system for the fuel, based on the exclusive use of deciduous oak for the supply of the amphora kiln. The upkeep of a sufficient quantity of wood after more than three hundred years of intensive production at the workshop indicates a forest management strategy at the property, maintained beyond the second century.  In an environment that seems to have been dominated by an oak grove, with a certain floristic diversity, the masters of the estate and the workshop must have favoured the exploitation of a single species, oak, for the supply of the large kilns, and reserved the exploitation of other species for the secondary kilns. 

Key words Loron, ceramic workshop, amphorae, kiln, fuel, anthracology, archaeomagnetic dating.

The context The Loron workshop The ceramic production workshop at Loron (TarVabriga, Croatia) is part of a  vast  senatorial and then imperial property, set up along the sea front around AD10 on the territory of the colony of Parentium (Poreč, Croatia). Excavations conducted by several international teams (1994–2011)1  revealed the extraordinary nature Known since the 19th century as the probable location of a large imperial workshop (Kandler 1849: 187), the site of Loron was identified in the early 1990s by F. Tassaux and V. Kovačić, with the establishment of a programme of excavations run by the museum of the Poreč territory/ Zavičajni muzej Poreštine and the Ausonius Institute — University Michel de Montaigne of Bordeaux. Following a first publication (Tassaux et al. (eds) 2001), a new research programme was launched from 2003 with the University of Padua (G. Rosada, A. Marchiori), then the Ecole française de Rome in 2007 (C. Rousse), with the support of the Croatian Ministry of Culture and the Italian and French Foreign Ministries. Running until 2011, it allowed for the discovery of the overall plan of the workshop, specifying the range of productions, beyond the Dressel 6B amphorae (wine amphorae, small Dressel 6B amphorae, sigillata, lamps, coarse wares, building materials). These 1 

of this large artisanal complex, mainly dedicated to the production of Dressel 6B oil amphorae exported to the Po Vally and the Danubian limes.2 Loron is thus one of the largest known workshops in the northern area of the Adriatic. The amphoric stamps (more than excavations are currently being published (2003-2011), but have been the subject of annual reports in the journals Histria Antiqua, Hrvastski Arheološki Godišnjak, Quaderni di Archeologia del Veneto et Mélanges de l’Ecole française de Rome. Since 2012, a new research programme has been launched with the museum of the Poreč territory/Zavičajni muzej Poreštine (G. Benčić, D. Munda), Ecole française de Rome and the Centre Camille Jullian at the University of Aix-Marseille (C. Rousse) which focuses on residential housing (i.e. a large villa 400 m north to the workshop, called Santa marina villa) and environmental studies. It is in this context that the excavations of the Loron workshop were resumed in 2017, with a campaign targeted at the sector with the large amphora kilns, aimed at comprehensively studying the structure of the kiln which had not yet been cleared. The campaign was coordinated by C. Vaschalde with, from the outset, the implementation of a protocol of anthracological study associated with the research. See Rousse et al. 2016, Rousse et al. 2018, Vaschalde et al. 2021 for archeological and anthracological studies on the kilns ; Rousse et al. 2021 for the context. 2  On the typology of Dressel 6B amphorae from Loron and the ceramic productions of the workshop (including a local production of stamped sigillata), see: Maggi 2001; Marion 2009; Marion and Starac 2001; Marion and Tassaux 2008; Maggi and Marion 2011.

roman pottery and glass manufactures (Archaeopress 2022): 65–76

Corinne Rousse, Christophe Vaschalde, Gaetano Benčić, Davor Munda

LORON CERAMIC WORKSHOP

Amphorae Kilns

Drying Yard

Drying Yard

LORON

48

49

45 Water basin

Kilns

Courtyard

Western Module : residential district

Eastern Module : Workshop Western amphorae kiln FR 8000 (excavation 2017)

0

50

N

Figure 1. Planimetry of the Loron workshop (CAO V. Dumas, C. Taffetani, AMU, CNRS, CCJ).

1800) reveal a list of prestigious owners: the first is  T. Statilius Taurus Sisenna, also called Sisenna Statilius Taurus, consul in AD16 and son of the great T. Statilius Taurus, a friend of Augustus. The workshop then passes into the hands of other  members  of the senatorial order, including  women  belonging to the imperial entourage such as Calvia Crispinilla.3 From the time of Domitian (AD 81), Loron becomes the property of the emperors, with a continuous series of stamps until Hadrian.  Beyond this, the amphorae are no longer stamped, but production continues until the end of the 3th century or the beginning of the 4th century, at which time the artisanal installations are gradually abandoned and exploited as a quarry, until the final abandonment of the site, a century later (late 5th century AD).

world (Figure 1). This plan is based on the construction of two building modules aligned on the coast line for 171m. The two modules are separated from each other by an access path down to the sea, 5.80m wide.  The road connects the workshop to the inland estate, while facilitating, from the western drying yard, the transport of amphorae to the sea, at a loading point that could not be located. The western module, studied mainly on the access passage to the beach, is characterized by the presence of a small bath complex: it is interpreted as a modest residential district, for the personnel, slaves or free men, who worked in the workshop.4 The eastern module corresponds to the production unit. It is 90m long and 80m wide, it occupies two superimposed terraces facing the sea: the lower terrace has a series of cellars – probably storage units – bounded to the north by an access passage; the upper terrace is occupied by the workshop, organized around a large rectangular courtyard (4500m²), framed by three long  buildings that  hold the structures of  production.  Four large battery kilns occupy the central space of the north building (room 45), framed on both sides by drying yards (rooms 48–49), that spill over to the east and west wings of the workshop  (Figure 1: Nos. 48–49).  These kilns constitute the main firing structures of the workshop, known for its large-scale production of

Beyond the volume of production, what characterizes the Loron workshop is the particularly functional arrangement of the buildings for forming, firing and storage of the amphorae.  The workshop follows an architectural plan, defined from the outset, and unparalleled for the moment in the Roman Amphoric and lapidary epigraphy was mainly studied by F. Tassaux (Tassaux et al. (eds): 32-44, 309-324). See also Manacorda 2010, whose proposal to integrate Messaline into the list of property owners at Loron is not retained in fine by F. Tassaux: Carre et al. (eds) 2011: 173. 3 

4 

66

Tassaux et al. (eds) 2001: 57-85.

Recent excavation of an amphorae kiln in the artisanal workshop complex at Loron

Dressel 6B oil amphorae, supplemented by a secondary production of wine amphorae, sigillata and coarse wares, as well as construction materials.

2001;  Degrassi and Maggi 2011;  Auriemma  et al.  2012; Maggi et al. 2017). Two small firing units, intended for the production of coarse wares, were also installed in the access pits of the eastern sector of Room 45, with a production dating to the end of the 3rd to early 4th century (Mondin 2017).

Past studies on the large amphora kilns (2007–2011) The kilns were recognized and partially excavated between 2007 and 2010 by the Italian-Croatian team5.  Only three structures were studied.  The existence of the fourth one (western kiln FR   ξ – 8000) was attested by a trench made at the praefurnium. 

A new study on the amphora kilns: the excavation of the western kiln FR ξ – 8000

The kilns are installed in a room 24.6m long and 18.6m wide (room 45), opening onto the courtyard with three entrances, whose traces are preserved in the south wall of the room. The kilns are arranged in pairs on either side of a central wall in small regular blocks, built in limestone rubble bonded with lime mortar.  They are separated by a thin corridor (1m wide), closed to the north and south by a masonry pillar base.  Perfectly aligned, the rear facades of the kilns are located 5m from the north wall of the room (MR 3288), leaving a communal loading space opening onto both drying yards from doors whose threshold is preserved. At the front of the kilns, a wide space of just over 5m wide serves as an access pit.

In 2017, the opening of a new excavation at the fourth largest kiln, located on the west side of room 45 and hitherto simply known through a trial trench, answered several requirements, both scientifically and in terms of valorisation.  It was decided to apply a targeted excavation protocol, integrating the methodological advances made in the study of the artisanal structures over the last fifteen years6. The main purpose of the excavation was to clarify the fuel supply strategies by applying a systematic sampling of charcoal to allow a dendro-anthracological study (the first one on this site), accompanied by sampling for archaeomagnetic dating of the final phase of the kiln, as well as radiocarbon dating (also for the first time on this site). The complete study of the kiln had also to be carried out in a single campaign and link to a process of valorisation.

According to these first studies, the kilns, oriented north-south, have the same characteristics: a long  praefurnium  (2m) facing south and a rectangular heating chamber of 7 x 5m, whose brick walls are reinforced on three sides by  tegulae  elevations  laid flat over one another and bonded with earth.  The remaining bases of the arches identified on the three kilns first studied suggest a single-channel heating chamber, close to type IIb of Cuomo di Caprio (Cuomo di Caprio 1972). The general scheme, confirmed by some stratigraphic elements, suggests that the setting up of the kilns goes back to the first phase of the complex (AD10–40).  As for recognized structures, they are more difficult to date: the two eastern kilns could have been abandoned in the course of the 2nd century, as indicated by the establishment of a small rectangular kiln in front of one of the two praefurnia; but the third large kiln, located west of the median wall, could have continued to function during the 3rd century (Kovačić et al.  2011;  Marchiori and D’Incà 2014).  Indeed, the typology of Dressel 6B amphora productions from Loron seems to continue until at least the end of the 3rd century  with a ‘late’ type, unstamped and posterior to the reign of Hadrian, well attested in Loron and on other sites north of the Adriatic in contexts of the late 3rd to early 4th century (Marion and Starac

The excavation revealed the best-preserved kiln structure on the site, both in plan and in elevation (Figure 2).  Measuring a total of 9 x 5.10m, the FR ξ-8000 kiln is characterized by a long praefurnium (2.5 m) leading to a rectangular heating chamber (6.5 x 5.10m), built of brick,with two parallel corridors: the final phase of the structure thus presents a type slightly different from that previously identified for the three other kilns (type IIc of the typology of Cuomo di Caprio 1972). Finally, the floor — built according to a particular system of T-shaped bricks — has been, exceptionally, preserved. It rests on ten series of arches built out of bricks, the dimensions of which reveal a predetermined system of modules and forms, perfectly adapted to the construction of the kiln. The firing chamber, located above the heating chamber, measures 5.20 x 3.90m externally. It is accessible through a door located north, in the axis of the kiln, measuring 80cm  wide.  This opening from the back of the kiln allows for loading from the communal space for all the four kilns, in the northern part of the room, which connects to the east and west with the drying yards.  The access pit (4.3 x 4m) located south of the kiln and its  praefurnium, is also accessible from the courtyard through one of three doors in the south wall of Room 45. Particularly in the south of France: Bigot and Vaschalde 2017-2018; Bourgeon 2018; Chabal 2001; Chabal et al. 2012; Denti and Villette (eds) 2019; Desbonnets 2018; Laubenheimer 2001 ; Mauné and Carrato 2012; Mauné et al. 2014; Mauné et al. 2017-2018; Thernot et al. 2004; Vaschalde 2018. 6 

V. Kovačić (Zavičajni muzej Poreštine — museum of the Poreč territory) —A. Marchiori, C. D’Incà, G. Rosada (university of Padova) as part of the international agreement described in n. 1. See Kovačić et al. 2011; Marchiori and D’Incà 2011; 2014. 5 

67

Corinne Rousse, Christophe Vaschalde, Gaetano Benčić, Davor Munda

N= 5015. 864

N

0

2m

N= 5015. 863

MR

328

8

N= 5015. 862

E sp ace ar

rièr e

N= 5015. 861

N= 5015. 860

N= 5015. 859

MR

814 0

ST ST

ST ST N= 5015. 854

ST ST

MR 809 4

815

1

MR

815

0

813

MR

2

813

3

813

4

814

1

MR

814 2

813

MR

N= 5015. 855

813 1

098

5

ST ST

MR

0

T8

812

N= 5015. 856

813

il S

809 6

MR 812 8

807 330 1

MR

ST ST

MR

N= 5015. 857

Seu MR

5

N= 5015. 858

5

813

6

S ST ole 800 4

813 7

813

8

813

9

N= 5015. 853

MR

814

3

Esp

Ala ST ndier 805 4

N= 5015. 852

Fou

ace ν

111

rF Rκ

MR 8

N= 5015. 851

FR

808 1

MR

814

4

814 5

N= 5015. 850

Fos

se d

MR

N= 5015. 849

’ac cès

N= 5015. 848

N= 5015. 847

MR

328

4

N= 5015. 846

Figure 2. Loron. Orthophotography of the kiln ξ -8000 (Orthophotography: V. Dumas, AMU,CNRS, CCJ; CAD: V. Vaschalde).

68

E= 5390. 528

E= 5390. 527

E= 5390. 526

E= 5390. 525

E= 5390. 524

E= 5390. 523

E= 5390. 522

E= 5390. 521

E= 5390. 520

E= 5390. 519

E= 5390. 518

E= 5390. 517

N= 5015. 844

E= 5390. 516

N= 5015. 845

Recent excavation of an amphorae kiln in the artisanal workshop complex at Loron

Figure 3. Loron. FR ξ -8000 kiln seen from the south. In the foreground, the praefurnium and its façade. On the left, the small kiln FR 8081 (photo: P. Ružić).

compact and homogeneous red clay filling was placed between this façade wall and the beginning of the heating chamber: it covers the praefurnium, perhaps for the purpose of exerting a weight on the vault and to ensure its stability.

The architecture of the kiln The praefurnium and its facade On the south side, the kiln FR ξ -8000 is equipped with a praefurnium, 2.50m long, which has only been partially cleared (Figure 3).  It forms a generally cylindrical tunnel, 1.35m wide and 1.58 m high at its southern end.  It is made of unfired clay bricks in the shape of voussoirs, bonded with clay, which are 40cm long and about 8cm thick and were fired when the kiln was in operation.  Inside, the praefurnium is partially filled with ash (sometimes hardened) and charcoal (US 8085 [US=stratigraphic unit]) from the last firing. This filling even includes intact charred logs (collected), as well as a cluster of limestone blocks with traces of heating. One of these blocks, placed on the ground, seems to be a threshold.

The heating chamber The built-up structure of the kiln FR ξ -8000 forms a quadrilateral of approximately 9 x 5.10m, the main body and praefurnium included. Preserved in places up to nearly 2m in height, this structure reveals a complex organization of masonry, allowing for periodic rebuilding of parts damaged by heat, or to reinforce others.  Among the permanent elements, pillar bases were found (MR 8141;  MR 8144), constructed out of  tegulae  laid flat and bonded with lime mortar, that define the north-east and south-east angles of the heating chamber (Figure 2).  To the west, the perpendicular wall of the room (MR 3301) marks out the path to the kiln. To the north, the MR wall   8151, built of limestone rubble and tegulae fragments bonded to lime mortar,  serves as a permanent foundation for the rear elevations of the kiln.

The facade wall leaned on the praefurnium. Cut through by several robber trenches, it could be observed only on the east side of the praefurnium. It consists of limestone rubble bonded with lime, and then tegulae fragments laid flat and bonded to lime mortar for the upper courses. A 69

Corinne Rousse, Christophe Vaschalde, Gaetano Benčić, Davor Munda

Figure 4. Loron. FR ξ -8000 kiln seen from the north. In the foreground, the back wall of the kiln, with the traces of a threshold accessing the firing chamber (photo: P. Ružić).

The east and west walls of the heating chamber are supported on these permanent elements (MR 8075 and MR 8125), built using tegulae bonded to clay. The tegulae are installed in a transverse (east-west) direction on the west wall, while they are aligned on the longitudinal axis of the kiln (north-south) for the east wall. Both walls show signs of heating and have been subjected to intense restoration activity, up until the complete disappearance of certain sections of the elevations.

(tiles bonded with clay) and MR 8096 (clay-bonded bricks) suggest that the kiln functioned, originally, with a single corridor. Above the rear wall, a threshold ST 8098 consisting of three fragments of bricks taken from a homogeneous beige clay matrix and framed by the traces of white carbonate lime (dimension 8.12m), marks the location of the back door of the kiln, which allowed access to the hearth and the firing chamber.  This opening, 75cm wide, was accessible from the rear space of the kiln (not cleared), north of the room, which also connected with the western drying yard (threshold preserved to the level of wall MR 3301).

At the back of the kiln, two walls were uncovered, supported by the foundation MR 8151 (Figure 4).  The northernmost wall is composed of two interconnected walls, with   traces of exposure to a very high heat: west side, a wall MR 8140 in tegulae bonded with clay, linked to the west wall of the kiln (MR 8075); east side, a wall MR 8096 made of bricks (25 x 25 x 7/8cm) bonded with clay. A second wall MR 8094 was added to the front of this wall, with the addition of a homogeneous beige clay backfill to fill the gap. This second wall, also exposed to a very strong heat, is built of clay-bonded bricks (25/27 x 25/26 x 12/13cm). Its installation reduces the length of the heating chamber by 40cm and could correspond to a change of plan, with the establishment of two corridors instead of only one in the initial phase, as seen for large eastern kilns.  Differences between the construction techniques used for the MR 8140 wall

The space between the east, west and north walls, and the southern base of the praefurnium forms the firing chamber.  It is occupied by two parallel north-south corridors bounded by brick walls bonded with clay (MR 8127 to the west, MR 8128 to the centre, between the two channels and MR 8129 to the east), on which are ten sets of arches supporting the floor (Figure 2 and 4).  These walls, seen only in the northern part of the kiln, are 60cm wide for 3 or 4 preserved courses (45cm high). They use two brick modules (43 x 29 x 9.5 cm; 26 x 6.5 cm), with traces of exposure to a strong heat. The east and west walls rest directly on the back wall of the kiln (MR 8096) 70

Recent excavation of an amphorae kiln in the artisanal workshop complex at Loron

Figure 5. Loron. Kiln FR ξ -8000. Detail of the kiln floor in place, with the T-bricks forming the heat-vents, the clay scraps of screed, tegulae  and stones blocking the heat-vents (photo and CAD: C.Vaschalde).

while the central base, which separates the two channels, stops 10cm from this rear wall. It also bears the traces of a repair and a change of arrangement of the arches supporting the floor: these elements make it possible to reconstruct a phase previous to the final stage of the kiln, which must have had nine series of arches, instead of the ten which currently support the walls.

while others have a single vertical half heat-vent. These bricks are placed between each arch, face to face, so as to form complete heat-vents. Once the bricks are laid, there remains a 10cm wide space between each row of bricks remains, filled with fragments of rectangular bricks and clay. A clay screed about 2 to 3.5cm thick is still present as scraps on the bricks of the kiln floor: it retains the traces of a pavement, either in the form of fragments of tegulae sealed in the clay, or in the form of negatives of these same fragments.  Finally, several heat-vents are still blocked by flat stones placed on the top as a sort of plug.

The kiln floor Ten series of two arches support the kiln floor, of which the front third is well preserved (Figure 5).  In order not to cause the collapse of the kiln floor, only the northernmost arches have been cleared and studied. They are made of voussoir bricks (43 x 30cm; the small side measuring 6cm, and the large one, 9cm), resting at the base on two courses of parallelepipedic bricks (41/42 x 30 x 10cm). The extrados of the arches does not rest directly on the outer walls MR 8075 and MR      8125, but on a filling composed of fragmented bricks bonded to clay. This same technique is used on the upper part of the extrados, where the level of the resting surface of the T-bricks of the kiln floor sits on a filling of recycled tegulae fragments.

The access pit of the kiln To the south of the kiln FR ξ -8000 extends the access pit to the praefurnium (Figure 2 and 3). It is installed in a quadrangular space bounded on the south and west by the walls MR 3284 and 3301, and on the north by the front of the kiln FR ξ -8000.  To the east, no structure separating this pit from the one associated with the large neighbouring kiln FR κ has been unearthed. The access pit measures approximately 5.40 x 4.40m, on an area of approximately 24m2.  The occupation level consists of a layer of red clay loam sediment rich in calcareous inclusions, with some fragments of tegulae, recognized as the layer of clay covering the

The kiln floor ST 8004 is built with T-bricks with forkshaped half heat-vents.  On the reverse, some are flat, 71

Corinne Rousse, Christophe Vaschalde, Gaetano Benčić, Davor Munda calcareous geological substratum.  The praefurnium and the walls of the kiln are based on this level.  The surface dips towards the entrance of the praefurnium (upper dimensions 6.40m; lower dimensions 5.80m). To the south, it is 70cm below the level of the restored threshold of the east gate of Room 45 (dimensions: 7.10m). This gate communicates with the courtyard: a small staircase, no longer visible today due to late antique destructions, was to facilitate access to the loading pit.

century AD (the accepted chronology for the so-called late amphora productions).  As for the FR-8000 kiln, carbon-14 dating now makes it possible to establish the end of its operation to between the middle of the 3rd and the end of the 4th century  (Poz 100663: 1735 +/- 30 BP).  The full analysis of the results of the archaeomagnetic study are still awaited, but so far suggests that the kilns stopped functioning around the end of the 3rd century.  This dating is perfectly consistent with the pottery associated with the working levels of the kiln.  Regarding the kiln FR8081, the two carbon-14 dates provide calibrated intervals between the second third of the 1st and the middle of the 3rd century (Poz-100664: 1835 +/- 30 BP; Poz-100788: 1875 +/- 30 BP). The latter kiln therefore began in an earlier phase but functioned at the same time as the large kiln FR-8000 later on.

The charcoal layers and ashen waste associated with the furnace of the kiln covered this occupation level, buried in turn under the demolition layers corresponding to the destruction of the kilns and the north building (4th–5th century AD).  The complete excavation of the  access pit did not reveal a dump, which could have informed us about the productions of the large kiln during this last firing phase.

Regarding the previous phases of the large kiln FR ξ -8000, the observations made about the arrangements of the firing chamber allow us to suggest at least two phases prior to its construction. In a first phase, the firing chamber could have operated with a single channel (according to the first hypothetical model of eastern kilns) or had already two channels. A first transformation (phase 2) consisted of reducing the length of the firing chamber by doubling the rear wall of the kiln and introducing two channels surmounted by 9 series of arches. A second redevelopment (phase 3) led to the modification of the placement of the arches and the introduction of an additional series, to arrive at the ten series of arches observed in the final phase.  These restructurings are not surprising if one considers that the implantation of the structures goes back to the initial phase of the workshop (AD 10–40), as indicated by the organization in the battery of kilns integrated into the overall plan of the workshop, and some stratigraphic elements observed east of Room 45. Just south of the kiln FR ξ -8000, outside room 45, a Sisenna amphora dump (AD 10–40), associated with damaged kiln elements, in the backfill that supported the occupation level of the courtyard, was interpreted as evidence of this first phase of production (Rousse and Tassaux 2008).

A small secondary kiln FR 8081 attached to the praefurnium of the large kiln In the western part of the access pit, several secondary firing structures seem to have been installed, as evidenced by negative traces of overheating observed on the inner wall of the west wall of the room, and a small kiln – this time well preserved – attached to the west wall of the praefurnium of the large kiln (Figure 3). This small kiln FR 8081 is a quadrangular structure with exterior measurements of 2.30m long and 70cm wide.  It has been preserved to a height of 65cm.  Its walls are built of adobe bricks fired during the use of the kiln. They preserve the traces of two phases: a first kiln 70cm wide, has been replaced by a second kiln with smaller dimensions (50cm wide). The stratigraphy of the filling also testifies to two successive phases, with two levels of ash separated from each other by a clay filling. The pottery associated with these ashy layers is composed of amphora stoppers (two sizes: 5 and 7cm in diameter) and fishing weights, which seem to have been the main production of this small kiln. A cluster of weights, which are wasters, has also been found in the upper filling of the access pit, on the east side of the pit.

The dendro-anthracological study: first results              

The dating of the kilns FR ξ -8000 and FR 8081

During the excavation, several levels of charcoal and ash were discovered in the two kilns FR ξ -8000 and FR 8081. In kiln FR ξ -8000, two types of samples were taken: a sampling of the carbonized logs left in situ, accompanied by a mass sampling of the carbonized layers.  In the kiln FR 8081, no logs were preserved, so only a bulk sample was taken from two successive carbonized levels.  In total, 125 litres of sediment and 8 logs were collected.  The bulk samples were then sieved under water on a sieve column (mesh: 4, 2 and 0.5 mm). The carbonized logs are to undergo a dendro-

The dating of the abandonment of kilns FR ξ -8000 and FR 8081 is difficult to establish, due to the intense restoration activity of the structures and for kiln FR ξ -8000 the absence of pottery directly associated with it. If we compare the general stratigraphy of the site and the typochronology of the productions with the results of the Italian team, the dating of the last phases of firing presents a wide range, between the 2nd century  (abandonment of the large eastern kiln in Room 45) and the end of the 3rd century/early 4th 72

Recent excavation of an amphorae kiln in the artisanal workshop complex at Loron

Structure

FR ξ-8000

FR 8081

Volume (L.)

45

20

Stratigraphic unity

8085

Nature

Taxinomic identification

Chronology

Fireplace of kiln

Fraxinus Ostrya carpinifolia Quercus deciduous Rosaceae maloïdeae Rosaceae prunoïdeae Tilia Ulmus Bark Knot Indeterminable TOTAL

Ash Hop-hornbeam Deciduous oak     Linden Elm

[285; 375] AD 98

1 1

8113

8148

Fireplace of kiln

Fireplace of kiln

[70; 250] AD 10 20 31 1 5 3 49 1

100

120

15

20 19

8 2 1 50

TOTAL Occ. 10 40 148 1 5 3 57 1 4 1 270

%

3,7% 14,8% 54,8% 0,4% 1,9% 1,1% 21,1% 0,4% 1,5% 0,4% 100,0%

Table 1. Charcoal analysis from kiln FR ξ -8000 and kiln FR8081 (C. Vaschalde).

chronological study, while the other samples have been subjected to an anthracological analysis (taxonomic determination and analysis of morphological and taphonomic parameters).  The study of the charcoal was carried out at ISEM UMR 5554 in Montpellier. The dendro-chronological analysis (Vaschalde et al. 2021) is being carried out by B. Brossier (IE, ISEM UMR 5554), and is based on the application of a particular method of stabilization of charcoal allowing for the dendrological study (Brossier and Poirier 2018).

regular cleaning of exploited afforestation, which could provide a diversified spectrum of species.  The established spectrum includes only species belonging to a managed forest.  The absence of thermophilic species, and the presence of taxa accustomed to semi-shaded environments (elm, lime tree), suggest that the exploited environment had a relatively open treeline. Note also the absence of fruit, which suggests that waste, particularly from olive trees, obviously abundant in the area of Loron, was not reused as fuel in the potters’ kilns.

The anthracological analysis was carried out on two US, one from kiln FR ξ -8000 (US 8085) and the other from phase 2 of kiln FR 8081 (US 8113). In both cases, these are the remains of the firing chambers from these two structures.  The study involved 220 charcoal samples (Table 1), identifying seven different taxa: Fraxinus (Ash), Ostrya carpinifolia (hop-hornbeam), Deciduous Quercus (Deciduous Oak), Rosaceae maloïdeae  (Rosaceae maloïdeae), Rosaceae prunoïdeae (Rosaceae prunoïdeae), Tilia (Linden) and Ulmus (Elm).

Conclusion The  complete excavation of the  fourth  large amphora kiln at the Loron workshop, combined with the anthracological study and several dating methods,  renews  our knowledge of the functioning of the workshop, especially for the last phase of production.  We have demonstrated that the large kiln was in use until the end of the 3rd  century to the beginning of  the  4th,  with a production that  corresponds  to the late Dressel 6B amphora from Loron, although a dump has not yet been located near the large kiln.  This dating confirms the chronology of the late Dressel 6B amphorae, relying so far on a relatively limited number of contexts, mainly identified in northern Italy. The interruption of the stamping of the amphorae, after the reign of Hadrian, prevents us from understanding the management of the workshop during this last phase of operation. Excavations carried out in the various units of the workshop show that certain structures were maintained and functioned (sheds, water reserve) while others were transformed (abandonment of at least one of the four large amphora kilns  in the central room 45; abandonment of small kilns located at the end of the east wing of the workshop).

The anthracological analysis of US 8085 and 8113 shows that fuel supply practices differ significantly between FR-8000 and FR 8081. The wood burnt in FR 8081 comes from a consociation of the type Ostryo-Quercetum pubescentis, without any apparent selection.  On the other hand, deciduous oak is the only species used in the FR ξ –8000 kiln. This difference makes it possible to hypothesize that a deliberate selection was carried out by the artisans and managers of the ancient domain of Loron.  The dendrological study shows that the artisans used logs aged 7 to 19 years. Some were cut 4 years before the last baking of the oven, suggesting a practice of fuel storage (Vaschalde et al. 2021). Unlike the supply of the large amphora kiln, the fuel of the small kiln FR 8081 could have come from a simple 73

Corinne Rousse, Christophe Vaschalde, Gaetano Benčić, Davor Munda The anthracological study reveals a well-thought out management system for the fuel, based on the exclusive use of deciduous oak for the supply of the amphora kiln. The upkeep of a sufficient quantity of wood after more than three hundred years of intensive production at the workshop  indicates  a  forest  management strategy at the property, maintained beyond the second century.  A sampling of the ash levels that cover the western drying yard suggests the exclusive use of deciduous oak as fuel for the large kilns, from the very foundation of the workshop.  In  an environment that seems to have been dominated by an oak grove, with a certain floristic diversity, the masters of the estate and the workshop must have favoured the exploitation of a single species, oak, for the supply of the large kilns, and reserved the exploitation of other species for the secondary kilns.  This observation opens up new questions, in a field still under-represented in the literature,  on the management of fuel resources, the space dedicated to them within the property, and the ability of the various owners to maintain  these resources in the long working life of the workshop.

References Auriemma, R., V. Degrassi and E. Quiri 2012. Produzione e circolazione di anfore in Adriatico tra III e IV secolo: dati da contesti emblematici, in C.S. Fiorello (ed.), Ceramica romana nella Puglia adriatica : 255‒289. Bari: Sedit. Bourgeon, O. 2018. La production d’amphores oléicoles dans la basse vallée du Genil (Ecija, Séville, Espagne). Contribution à l’histoire socio-économique de la Bétique romaine (Ier s.-IVe s. ap. J.-C.). Unpublished PhD dissertation, University of Montpellier Paul Valéry 3. Bigot, F. and Chr. Vaschalde 2017-2018. Le four de l’atelier de potiers d’Espeyran (Saint-Gilles-duGard)  : un témoin du développement économique de la basse vallée du Rhône durant la période augustéenne précoce, in Mauné et al. (eds) 20172018 :115‒141. Brossier, B. and Ph. Poirier 2018. A new method for faciliting tree-ring measurement on charcoal form archaeological and natural contexts. JAS Reports 19: 115-126. Carre, M. - B., V. Kovačić and F. Tassaux (eds) 2011. L’Istrie et la mer. La côte du Parentin dans l’Antiquité (Mémoires 25). Bordeaux: Ausonius. = Carre, M. - B., V. Kovačić and F. Tassaux (eds) 2012. Sjeverno priobalje poreštine u antici. Poreč : Zavičajni muzej Poreštine. Chabal, L. 2001. Les Potiers, le bois et la forêt à Sallèlesd’Aude (Ier-IIIe s. ap. J.-C.), in Laubenheimer  (ed.) 2001: 93‒110. Chabal, L., I. Figueiral, C. Pellecuer and I. Bermond 2012. Evidence of paleogeographic constraints on woodlands on the shores of a coastal lagoon during Antiquity: charcoal analysis of the PrésBas villa and Le Bourbou (Loupian, Hérault), in E. Badal, Y. Carrion, M. Macias and M. Ntinou (eds) Wood and charcoal. Evidence for human and natural history (Saguntum Extra 13): 115‒124. València; Universitat de València, Departament de Prehistòria i Arqueologia de la Facultad de Geografía i Història. Cuomo di Caprio, N. 1972. Proposta di classificazione delle fornaci per ceramica e laterizi nell’area italiana dalla preistoria a tutta l’epoca romana. Sibrium IX: 371‒461. Degrassi, V. and P. Maggi 2011. Il pozzo occidentale del foro di Aquileia. Commerci e consumi attraverso l’evidenza dei materiali di riempimento. Antichità Altoadriatiche 70 : 231‒240.  Denti, M. and M. Villette (eds) 2019. Archéologie des espaces artisanaux : fouiller et comprendre les gestes des potiers : actes du colloque international de Rennes (27-28 novembre 2014). Lattes : Editions de l’Association pour le Développement de l’Archéologie en LanguedocRoussillon. Desbonnets, Q. 2018. Les ateliers d’amphores à huile du conventus d’Hispalis (Séville, Espagne). Caractérisation et étude d’une zone de production de la province romaine de

Finally, the exceptional state of preservation of the large kiln  FR ξ -8000 with its floor in place, its praefurnium and  rectangular  heating chamber  with two channels, beautifully completes our knowledge of Loron’s battery kilns. It opens up new perspectives in terms of creating models to evaluate the production capacity of kilns, and more widely, combined with the study of amphorae, expands our knowledge of the production of this large workshop, one of the most important in the northern Adriatic area. Translated by Victoria Leitch7

7 

We very thank Victoria Leitch for reading our paper and translating it.

74

Recent excavation of an amphorae kiln in the artisanal workshop complex at Loron

Bétique (Ier s. av. J.-C.-Ve s. ap. J.-C.). Unpublished PhD dissertation, University of Montpellier Paul Valéry 3. Kandler, P. 1849. Qualche bollo su cotti. L’Istria A.IV, 47: 187. Kovačić V., A. Marchiori, Y. Marion, G. Rosada, C. Rousse and F. Tassaux 2011. Loron-Lorun, Parenzo-Poreč, Istria. Una Villa Maritima nell’agro parentino  : la campagna di ricerca 2010. Histria Antiqua 20: 515‒525. Laubenheimer, F. (ed.) 2001. 20 ans de recherches à Sallèlesd’Aude. Besançon — Paris: Presses universitaires Franc-Comtoises, Les Belles Lettres. Maggi, P. 2001. La ceramica fine da mensa, in Tassaux et al. (eds) 2001: 127‒176. Maggi, P. and Y. Marion 2011. Le produzioni di anfore e di terra sigillata a Loron e la loro diffusione, in G. Lipovac Vrkljan, I. Radić Rossi and B. Šiljeg (eds) Rimske keramičarske i staklarske radionice. Proizvodnja i trgovina na jadranskom prostoru, zbornik I. međunarodnog arheološkog kolokvija (Crikvenica 2008) / Roman pottery and glass manufactures. Production and Trade in the Adriatic Region, Proceedings of the 1st International Archaeological Symposium (Crikvenica 2008): 175‒187. Crikvenica : Muzej Grada Crikvenice, Institut za arheologiju. Maggi, P., F. Maselli Scotti, S. Pesavento Mattioli and E. Zulini (eds) 2017. Materiali per Aquileia. Lo scavo di Canale Anfora (2004-2005) (Scavi di Aquileia 4). Trieste: Editreg. Manacorda, D. 2010. Il misterioso MESCAE. Donne imprenditrici nell’Istria romana. Rei Cretariae Romanae Fautores Acta 41: 217‒227. Marchiori, A. and C. D’Incà 2011. Le fornaci di Loron (Istria, Croazia), in G. Lipovac Vrkljan, I. Radić Rossi and B. Šiljeg (eds) Rimske keramičarske i staklarske radionice. Proizvodnja i trgovina na jadranskom prostoru, zbornik I. međunarodnog arheološkog kolokvija (Crikvenica 2008) / Roman pottery and glass manufactures. Production and Trade in the Adriatic Region, Proceedings of the 1st International Archaeological Symposium (Crikvenica 2008): 83-90. Crikvenica: Muzej  Grada  Crikvenice , Institut za arheologiju. Marchiori, A. and C. D’Incà 2014. Le fornaci e il centro produttivo di Loron : nuovi dati, nuove rifessioni, 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 2011) / Roman pottery and glass manufactures. Production and Trade in the Adriatic Region, Proceedings of the 2nd International Archaeological Symposium (Crikvenica 2011) (Zbornik Instituta za arheologiju, 2): 127-139. Crikvenica: Institut za arheologiju, Muzej Grada Crikvenice. Marion, Y. 2009. Les Dressel 6B de petites dimensions de Loron, in S. Pesavento Mattioli and M. - B. Carre (eds) Olio e pesce in epoca romana. Produzioni e commercio nelle

regioni dell’alto Adriatico. Atti del Convegno (Padova, 2007) (Antenor Quaderni 15): 281‒287. Roma: Quasar. Marion, Y. and A. Starac 2001. Les amphores, in Tassaux et al. (eds) 2001: 97‒125. Marion, Y. and F. Tassaux 2008. Tuiles et amphores estampillées de Loron (Croatie), in M. Hainzmann and R. Wenedig (eds) Instrumenta inscripta latina II. Akten des 2 Internationalen Kolloqiums (Klagenfurt 2005): 209‒222. Klagenfurt: Verlag des Geschichtsverein für Kärnten. Mauné, S. and C. Carrato (eds) 2012. Le complexe domanial et artisanal de Saint-Bézard (Aspiran, Hérault) au début du Ier s. ap. J.-C. Fondation et genèse. Revue archéologique de Narbonnaise 45: 21‒38. Mauné, S., E. Garcia Vargas, O. Bourgeon, S. Corbeel, C. Carrato, S. Garcia Dils, F. Bigot and J. Vasquez Paz 2014. L’atelier d’amphores à huile Dr. 20 de Las Delicias à Ecija (Prov. de Séville, Espagne). SFECAG. Actes du Congrès de Chartres: 419‒444. Mauné, S., F. Bigot and S. Corbeel (eds) 2017- 2018. Recherches récentes sur les ateliers de production et les amphores vinaires de Gaule Narbonnaise et de Tarraconaise. Revue archéologique de Narbonnaise 5051: 17‒274. Mondin, C. 2017. La produzione di ceramica tarda a Loron / Lorun (Poreč / Parenzo – Croazia), in J. Turchetto and M. Asolati (eds) Paesaggi in movimento. Ricerche dedicate a Guido Rosada: 237‒248. Padova: University press. Rousse, C., V. Kovačić, K. Gergeta Sotončić, V. De Leonardis, V. Dumas, P. Maggi and C. Taffetani 2015. Loron / Santa Marina — Busuja (Tar-Vabriga, Poreč, Croatie), Chronique des activités archéologiques de l’École française de Rome [En ligne], Balkans. Rousse, C., D. Munda, G. Benčić, K. Gergeta Sotončić, V. De Leonardis, V. Dumas, P. Maggi, M. Tillier and C. Vaschalde 2016. Loron / Santa Marina (Tar-Vabriga, Poreč, Croatie). Chronique des activités archéologiques de l’École française de Rome [En ligne], Balkans. Rousse, C., G. Benčić, D., Munda, N. Garnier, K. Bartolić  Sirotić,, V. Dumas, N. Basuau, P.  Maggi 2021. La villa de Santa Marina (Tar Vabriga – Torre Abrega, Croatie), Bulletin archéologique des Écoles françaises à l’étranger [En ligne], Balkans. http:// journals.openedition.org/baefe/5098; DOI : https:// doi.org/10.4000/baefe.5098 Rousse, C., D. Munda, G. Benčić, V. Dumas and P. Maggi 2019. La villa de Santa Marina. Campagne 2018. Chronique des activités archéologiques de l’École française de Rome [En ligne], Balkans. Rousse C. and F. Tassaux 2008. Loron (Tar-Vabriga, Croatie). Mélanges de l’École française de Rome — Antiquité 120.1: 184‒190. Tassaux, F., R. Matijašić and V. Kovačić (eds) 2001. Loron (Croatie). Un grand centre de production d’amphores à 75

Corinne Rousse, Christophe Vaschalde, Gaetano Benčić, Davor Munda huile istriennes (Ier-Ive s. P.C.) (Ausonius- Mémoires 6). Bordeaux: Ausonius. Thernot, R., V. Bel and S. Mauné 2004. L’établissement rural antique de Soumaltre (Aspiran, Hérault). Ferme, auberge, nécropole et atelier de potier en bordure de la voie Cesseron-Condatomagus (Ier-IIe s.). Montagnac: Editions Mergoil. Vaschalde, Chr. 2018. Fours à chaux et chaufourniers en France méditerranéenne du Moyen Âge à l’époque moderne. Approche interdisciplinaire autour des techniques, des savoir-faire et des artisans (Europe médiévale 14). Montagnac: Editions Mergoil Vaschalde, Chr., M. Tillier, N. Rovira, Fr. Guibal, D. Kaniewski, Ch. De Bruxelles, M. - B. Carre and C. Rousse C. in press. Landscape, resources management and vegetal economy in the Parentium area (Croatia) between the Imperial period and the beginning of the Middle Ages. First results of a Archaeobotanical investigation, in I. Borzić, E. Cirelli, K. Jelinčić Vučković, A. Konestra and I. Ožanić Roguljić (eds) TRADE – Transformations of Adriatic Europe (2th-9th century). Proceedings of conference, Zadar, 11-13 February 2016.

Vaschalde, Chr., C. Rousse, B. Brossier and G. Benčić 2021. Nouvelle étude d’un four à amphore dans le complexe artisanal de Loron (Tar Vabriga Torre Abrega, Croatie). Premiers résultats de l’étude archéologique et anthracologique, in D. Van Limbergen and D. Taelman (eds), The Exploitation of Raw Materials in the Roman World: A Closer Look at Producer-Resource Dynamics: Panel 4.4. Proceedings of the 19th International Congress of Classical Archaeology: Archaeology and Economy in the Ancient World (Cologne/ Bonn 2018) (Propylaeum Band 27): 31‒45. Heidelberg: Propylaeum.

76

Amphorae and tiles from the island of Pag: indirect evidence of the exploitation of local resources in the early Roman imperial period Maja Grisonic, Nikolina Stepan Abstract In this article we discuss the possibility of local production of Dressel 6B, Dressel 2-4 and the newly discovered Caska 1 type amphorae in the areas of Caska and Novalja on the northern part of the island of Pag (Croatia). The hypothesis of local production of these containers is based on their frequency, morphological characteristics, stamps and macroscopic analysis of clay composition. The latter should be further researched by petrographic analyses in the future. These containers provide an indirect indicator of the exploitation of local resources in the early Roman imperial period, among which we can enumerate oil, wine and maybe (tuna) fish products, some of which might have been exported on local or regional markets. Consular stamps, for the moment unique in the Roman world, impressed on both amphorae and tiles, suggest the dating of these productions in the era of the emperors Augustus and Tiberius.

Key words Pag, Caska, Novalja, Croatia, amphorae, tiles, stamps, local resources.

Introduction The island of Pag was part of Liburnia, the northernmost part of the Roman province of Dalmatia. The island was named Cissa (Pliny, NH, III, 140) after its most important settlement, which is thought to have been located by the largest field on the northern side of the island (Blečić and Radić Rossi 2004: 21-22; Radić Rossi and Zmaić 2009). This field is known as Novaljsko polje and it is surrounded by three natural coves: Novalja to the west, Stara Novalja (Old Novalja) to the northwest and Caska to the southeast (Figure 1). Novalja was an important port, set on the main eastern Adriatic seafaring route, which connected the Dalmatian coastal settlements with the northern Adriatic colonies and municipia, reaching the colony of Aquileia. Stara Novalja was an internal harbor, open on the Velebit Channel, which separates the island of Pag from the mainland. Caska Bay is the northernmost part of the deep Gulf of Pag which ends in Caska Cove, far from the sea routes (Boetto and Radić Rossi 2017; Radić Rossi and Boetto 2020). The village of Caska is located at the border with the Novalja field, devoted to viticulture for many centuries. Its traditional source of sustenance was fishing, especially that of tuna. The Tunera or Turanj, the tuna fish watching tower, almost the symbol of Caska, testifies to this fervent activity of the past (Figures 2, 5). This tower was built in the 19th century by the rich Palčić family (Radić Rossi and Boetto 2010: 300, n. 16; Hilje 2011: 115). During Antiquity a multi-stratified settlement developed in the northern part of the Caska Cove

(Boetto and Radić Rossi 2017; Radić Rossi and Boetto 2020), the history of which is gradually unveiling with the progress of the archaeological research. In the early Roman imperial period a branch of the Calpurnii Pisones, one of the most influential Roman senatorial families, chose Caska as the spot for one of their estates (Boetto and Radić Rossi 2017: 279; Radić Rossi and Boetto 2020), as attested by votive and funerary inscriptions (Šašel 1963; Kurilić 2004; 2011a). The destiny of the property in later times is not yet clear at this point of research (Boetto and Radić Rossi 2017: 279; Radić Rossi and Boetto 2020). The underwater excavations in Caska Cove and the recording of the numerous archaeological remains both underwater and on the coast are part of the project CissAntiqua, directed by Irena Radić Rossi (University of Zadar) and Giulia Boetto (CNRS, Université Aix-Marseille, Centre Camille Jullian). During systematic and multi-year researches in Caska Cove, different structures have been discovered and documented, including four ships that were sunk on purpose and reused in the 1st and 2nd centuries AD for the construction of piers and wharfs (Radić Rossi and Boetto 2010; 2011; 2020; Boetto and Radić Rossi 2012; 2017). Anamarija Kurilić (University of Zadar) led the excavations of the necropolis and one part of the Late Roman settlement (Kurilić 2011a with previous bibliography; 2011b; Kurilić and Serventi 2018). In 2005 and 2006 the company Geoarheo d.o.o., directed by Goran Skelac, unearthed a cella vinaria with five dolia, and a drainage channel, probably passing underneath an open court (Figure 3). This storage and productive part of a Roman villa is located in a private property right beyond the beach, about five meters above

roman pottery and glass manufactures (Archaeopress 2022): 77–93

Maja Grisonic, Nikolina Stepan

Figure 1. N part of the island of Pag, with supposed differentiation of exploitation of natural resources and local manufactures during the early Roman imperial period (author: N. Stepan).

present-day sea level (Skelac 2006; 2007). These remains can be dated to the early Roman imperial period. The 3rd century transformations of the original plan of the complex, and Late Roman vessels and coins recovered in the upper layers, attest to the continuation of life on the site during later periods (Grisonic and Stepan 2017).

them as storage rooms (Grisonic 2017; Grisonic and Stepan 2017), but based on analogies from Catalonia it is now clear that these two pits were built to contain arca lapidum counterweights used to operate two beam presses for winemaking (torcularium) (Lipovac Vrkljan and Konestra 2018: 21; Martín i Oliveras 2012; Martín i Oliveras et al. 2017: 198-199, 201). Arca lapidum means (wooden) box filled with stones. Archaeological evidence from Catalonia points to the fact that this innovative technology used in winemaking, described by Pliny the Elder (NH, XVIII, 317), started to be applied from the Augustan age on (Brun 2004; Martín i Oliveras 2012: 93-94; Martín i Oliveras et al. 2017: 198199). According to Pliny, thanks to this innovation, the

The two pits inside the villa in Caska and their contents In the northeastern excavated area of the pars fructuaria of the villa, two identical square underground rooms have been discovered (Figure 3). Their sides measured 3.6 m and were 2.4 m deep. We originally interpreted 78

Amphorae and tiles from the island of Pag

Figure 2. Caska and the Tunera: view towards the gulf and town of Pag (photo: M. Grisonic, 2013).

beams of the presses and therefore also the dimensions of pressrooms were reduced.

belonged to north-italic and aretine mould-made relief pottery, thin walled pottery, coarse and cooking ware, ceramic balsamaria, oil lamps, stamped tegulae and amphorae. One of the deepest layers in the filling, besides the numerous pottery fragments, contained a large number of animal bones of cattle, horse, deer, pig, sheep, goat, fish, as well as the tibia of a wolf of impressive dimensions (Miculinić 2011). The northern pit has been just partially excavated. It was filled with the same type of archaeological material found in the southern one.

In Caska the two pits for the arca lapidum were later reused as rubbish dumps. Thanks to the archaeological material, their oldest filling layers could be dated to the late Augustan or the Tiberian period (Grisonic and Stepan 2017). The southern pit has been entirely excavated. Its southern wall was built of roof tiles and reached the bottom of the cellar, while the other three sides were built with a wall of blocks and bricks, only 70 cm deep. The three sides of the pit were further directly excavated in the soil, to reach the bottom at 2.4 m depth, where six stone blocks, measuring 45x45x40 cm, were placed in two parallel rows on two sides of the space. The blocks were meant to support the arca lapidum structure, stabilizing it and distributing the weight during pressing operations (Martín i Oliveras 2012: 59, 96). The vertical stratigraphy of the pit showed the succession of 17 layers, which contained numerous ceramic vessels, bricks, animal bones and shells, stones and charcoals (Skelac 2007). The most numerous were fragments of Italian terra sigillata cups and plates, imported during the Augustan age mainly from the northern Italian area (Po Valley), but also from central Italy (Grisonic and Stepan 2017). Other fragments

The amphorae fragments recovered in the two pits inside the villa belong substantially to three amphora types: Dressel 2-4, Dressel 6B and a previously unknown type. For the moment, the latter amphora type is attested only in Caska, where it is present with a certain frequency. Therefore, following earlier examples of amphorae appellations (e.g. the well-known Fažana and Crikvenica productions – see Bezeczky 1998, Lipovac Vrkljan and Konestra 2018), we decided to name this container Caska 1 type amphora. For each amphora type recovered in the pits of the villa in Caska the minimum number of individuals (MNI) was calculated, summing the number of the rim sherds: Dressel 6B (12 MNI), Caska 1 type (10 MNI) and Dressel 2-4 (4 MNI). Eight amphora stoppers were also 79

Maja Grisonic, Nikolina Stepan

Figure 3. Plan of the villa in Caska (author: V. Pellegrino, LabEx Archimède UPV Montpellier).

found, five of them were obtained out of reused parts of amphorae bodies.

findings in his hometown, to prevent the loss of traces of history (Kunkera 1982: 12, 18). Because of the housing boom in Novalja, which already started during his lifetime, his numerous (and sometimes creative) suppositions are difficult to verify. Nevertheless, at the end of the 1980s more than twenty Dressel 6B amphorae necks and an entire specimen (Plate 1: 9) were found underneath a modern house in the center of Novalja (Gluščević 1989: 80-81; 2017), located near the termination of Novalja’s Roman aqueduct, which was bringing water from the Novalja field to its harbor (Radić Rossi and Zmaić 2009). The owner of the modern house also found a ceramic water tube, spilling into the lower half of a dolium, and this had an exit channel made by tegulae, one of which bore the stamp AFAESONI (Gluščević 1989: 82). Unfortunately, these facts cannot be verified, but the information of the stamp is important. Eight tegulae with the stamp A.FAESONI.AF are walled up in the underground channel carved in the limestone of the nearby Roman aqueduct, constructed most likely during the Augustan age (Radić Rossi and Zmaić 2009). The correct reading of the stamp should be A(uli) Faesoni Af(ricani), a workshop from coastal northern-central Italy, operating during the first half of the 1st century AD (Matijašić 1989: 64; Juras and Jurković Pešić 2016: 43), which greatly contributed

Out of the Dressel 6B amphorae (Carre 1985; Carre and Pesavento Mattioli 2003; Cipriano 2009 with bibliography) found in the pits inside the villa, at least one third were most likely imported from northern Italy, while supposingly the rest of them could be local products. Morphologically, the rims of at least two Dressel 6B amphorae found in the southern pit are identical to the many Dressel 6B amphorae rims recovered in 1993 in the nearby presumed workshop in Novalja. Also, the clay of the examples from Novalja (Plate 1: 1) and Caska (Plate 1: 2 and 3) was examined under the optical microscope, which strengthened the idea of the similarity in fabrics, although we would like to specify that no petrographic analyses have yet been performed on these pieces. Dressel 6B oil amphorae from the presumed Novalja workshop (Plate 1) The first to suppose the existence of a ceramic workshop in Novalja was the local priest and passionate collector of antiquities Josip Kunkera (1914-1983), whose goal was to record all possible information about archaeological 80

Amphorae and tiles from the island of Pag

Plate 1. 1-6: macro photos of the clays from a Dressel 6B amphora from Novalja (1), two Dressel 6B amphorae recovered in Caska (2-3), two Caska 1 type amphorae (4-5) and a Dressel 2-4 found in Caska (6) (authors: N. Stepan, S. Govorčin); 7-8: Dressel 6B amphorae from Novalja stored in the Archaeological Museum Zadar (courtesy of D. Taras and D. Romanović); 9: Entire Dressel 6B amphora from Novalja (from Gluščević 1989: 77, fig. 11) (author: N. Stepan).

81

Maja Grisonic, Nikolina Stepan to the datation of the aqueduct itself and could thus, jointly with the typology of the amphorae (see below), provide additional data for the datation of the finding context of the Dressel 6B amphorae.

In Loron, the second most famous pottery production center of the Istrian peninsula (Tassaux et al. (eds) 2001; Rousse et al. 2019), Dressel 6B amphorae started to be produced from AD 10 and they already had a higher and rather straight rim (Maggi and Marion 2011: 175-176). Because of these comparisons the chronology of the Dressel 6B from Novalja can be therefore anticipated to include the late Republican period and should not exceed the Augustan age.

In 1993, archaeologists from the Archaeological Museum Zadar carried out a rescue excavation next to the same house cellar in the center of Novalja, discovering layers containing huge numbers of Dressel 6B amphorae sherds, together with fragments of tegulae, spicae and a few other ceramic vessels (Gluščević 2017: 65). In the layers filled with Dressel 6B amphorae, kiln waster of the rim of one amphora and clumps of semifired clay were found, leading to the interpretation of the presence of an amphorae workshop located in the immediate vicinity and dated presumably to the 1st century AD (Gluščević 2017).

Archaeometric analyses were performed on samples of the Dressel 6B amphorae and the kiln waster of the rim, and compared to clay samples collected in the Novalja field, the closeby Caska, and Dinjiška in the southern part of the island, where a possible Roman (brick?) kiln existed along the modern road Zadar-Pag (Gluščević 1989: 73-74; Kurilić 2013: 274). The petrological and chemical analyses, as well as the X-ray powder diffraction, showed a similar clay composition of the samples (Gluščević 2017: 66). Therefore, although a kiln has not yet been found, the existence of a local ceramic workshop in Novalja is very probable.

The examination of the material, stored in the Archaeological Museum Zadar, allowed us to determine that all fragments recovered in 1993 belonged to the same type of Dressel 6B (Plate 1: 7). The calculated MNI of specimens recovered is 60. Most Dressel 6B amphorae found in Novalja have the same rim morphology: rounded, with a line running around the external part. The rim diameter is 15-16 cm, the rim height is 3.5-4.5 cm. The neck is conical; the handles are short and vertical, ovoid in section. The whole specimen recovered in 1988 has a pyriform body, it is 77 cm high and 36 cm wide (Plate 1: 9). The thickness of the wall of the amphora measures about 2.5 cm. The morphology of the spikes is different, most of them are knobbed, with a bigger or less pronounced knob (Plate 1: 8). The colour of the clay is mainly reddish-yellow (Munsell 5YR 6/6), but there are also some yellow (Munsell 10YR 7/6) specimens (Gluščević 2017: 65). Unfortunately, stamps were not found on the recovered specimens.

It is likely that the need for a local manufacture of Dressel 6B amphorae was connected with the trade of olive oil, produced on the adjacent Lun peninsula, where the cultivation of olive trees probably started in Roman times (Gluščević 1989; 2017). Nowadays the Lun peninsula is known for possessing one of the largest wild olive groves in the world, with many hundreds of olive trees being more than 1000 years old (Radić Rossi and Zmaić 2013: 8). The Caska 1 type amphora (Plate 2) This type of amphora for the moment appears to have been found only in Caska: at least ten fragmented specimens (10 MNI) were found in the two pits inside the villa (Plate 2: 14), and three almost complete ones were recovered from the seabed in Caska Cove (Grisonic 2017) (Plate 2: 15-17). For this reason, we are proposing here to name this newly discovered, and probably locally manufactured, type of amphorae Caska 1 type. We are certain that with the continuation of research and the typo-chronological studies on pottery shapes recovered from this important archaeological site, several additional local forms will be distinguished in the future. Because of the current lack of kiln findings in Caska and its surroundings, the origin of the clay used to produce the Caska 1 type amphorae could be (at least partially) solved with archaeometric analyses, which have not yet been performed.

The typology of the Dressel 6B amphorae from Novalja corresponds to the earliest phase of production of this type, which started at the end of the Republican period: morphologically, with their rounded rim, ovoid and stocky body and knobbed spike they seem to be influenced by the earlier ovoid amphorae from the mid-western coast of the Adriatic (Carre and Pesavento Mattioli 2003: 461-462; Cipriano 2009: 179-180). Toniolo defined this type as ante Dressel 6B (Toniolo 1991), but her classification has not been accepted by all scholars. Morphologically, the Dressel 6B amphorae from Novalja can be compared to the ante Dressel 6B (or Dressel 6B from the first phase) amphorae from the kiln of Fažana in Istria, which were produced on the senatorial property of the Laecanii (Bezeczky 1998, with earlier bibliography): the beginning of this production can be dated starting from the late Republican period (c. 40 BC), while during the Augustan age the classical shape of the Dressel 6B develops (Bezeczky 2016: 73-74, 92).

The first Caska 1 type amphora (Plate 2: 17) was found during the underwater research campaign in 2005, in the ancient harbor zone on the NE side of Caska Cove, about 100 m SE of the villa. It laid broken at about 2.5 m 82

Amphorae and tiles from the island of Pag

Plate 2. 10: interior coating of pitch on a Caska 1 type amphora (P. Groscaux, CCJ/CNRS, CissAntiqua project 2013); 11: Graffito XXXXV on the C120 Caska 1 type amphora (P. Groscaux, CCJ/CNRS, CissAntiqua project 2013); 12: Graffito XXV on the C115 Caska 1 type amphora (T. Seguin, CissAntiqua project 2015); 13: M∙IVNIO∙SILANO stamp mold on the C120 Caska 1 type amphora (N. Lete, Geoarheo d.o.o); 14: Fragmented Caska 1 type amphora from the villa in Caska (drawing: Z. Bakić, photo: N. Stepan); 15-16: two Caska 1 type amphorae (C115 and C120) recovered in Caska Cove in 2013 (T. Seguin, CissAntiqua project 2015); Caska 1 type amphora recovered in Caska Cove in 2005 (N. Lete, courtesy of I. Radić Rossi) (author: N. Stepan).

83

Maja Grisonic, Nikolina Stepan

Figure 4. Caska 1 type amphora in situ (photo: P. Groscaux, CCJ/CNRS, CissAntiqua project 2013, courtesy of G. Boetto and I. Radić Rossi).

depth, in the mooring zone on the northern side of the breakwater, built to protect the coast from the southern winds (Radić Rossi 2006: 285-286; Boetto and Radić Rossi 2017). Inside the same archaeological trench, numerous fragments of amphorae, but also of other ceramic vessels and building material, as well as shells, snail shells, and the bones of fish and terrestrial animals, were found. After its recovery, the Caska 1 amphora was fully reconstructed by Neven Lete (Geoarheo d.o.o.) and it is now on display in the Novalja City Museum.

cm wide is made, under which elliptical ribbon handles are molded. The cylindrical neck is short and wide, continuing into a prolonged pear-shaped body. The mean height of the neck is 19 cm, with the diameter of the inner rim measuring 11-12 cm and the diameter of the outer rim 13-14 cm. The morphology of the base is unknown. The heights of the three almost complete Caska 1 amphorae recovered from the sea bottom range from about 65 to 80 cm, and their width is 2535 cm. The amphora recoverd in 2005 is a little smaller than the two amphorae found in 2013 in front of the Tunera. The colour of the amphorae can vary from yellow (Munsell 10YR 7/6) to red (Munsell 2.5YR 6/6). They have a powdery but relatively refined fabric, with some calcareous, quartz and mica inclusions. After the microscopic examination of various samples, it is clear that the fabrics of the Caska 1 (Plate 1: 4 and 5) and the Dressel 6B amphorae found in the two pits inside the villa (Plate 1: 2 and 3) look very similar to the fabric of the Dressel 6B amphorae from the supposed Novalja workshop (Plate 1: 1), with the fabric of the Caska 1 type being slightly coarser.

Two almost complete Caska 1 type amphorae (Plate 2: 15 and 16) were discovered during the 2013 excavation campaign, included in the framework of the CissAntiqua project (Boetto and Radić Rossi 2017; Radić Rossi and Boetto 2020). They were found on the seabed in front of the watching tower for tuna fishing, next to a submerged structure made of rectangular wooden caissons, at 1.4 m depth (Figures 4 and 5). The caissons, composed of horizontal logs with rectangular mortises, through which vertical poles were driven, were filled with rocks. This was most likely the base of a pier-like wooden construction (Boetto and Radić Rossi 2017; Radić Rossi and Boetto 2020).

The two Caska 1 type amphorae found on the seabed in front of the tuna-watching tower were stamped and internally coated with pitch (Plate 2: 10). The rectangular stamps were applied on the neck, just

The Caska 1 type amphora has a lightly out-curving collar rim, under which a characteristic groove 1.5-2 84

Amphorae and tiles from the island of Pag

the empty amphorae, as stated for the graffiti on the Dressel 6B amphorae from Fažana (Bezeczky 1998: 6) or for the tituli picti on the Tyrrhenian fish amphorae Dressel 21-22 (Botte 2009: 128-130; Carre et al. 2009: 224). The weight of the two empty Caska 1 amphorae would thus be 45 librae or 14.715 kg (1 libra = 0.327 kg). The contents of Caska 1 amphorae remains uncertain, but we believe that it was designed for fish products (see the discussion below). Dressel 2-4 wine amphora and tegulae stamped SEX APPVLEIO∙COS (Plate 3) The name in the dative or ablative case was also used for a stamp applied on a Dressel 2-4 amphora handle, recovered in the southern pit of the Roman villa: SEX APPVLEIO∙COS, which can be read Sex(to) Appuleio∙co(n) s(ule) (Grisonic 2017) (Plate 3: 18). Again, the fabric of this amphora (Plate 1: 6) is very similar to those of the Caska 1 and the Dressel 6B amphorae from Novalja, pointing toward a local manufacture of Dressel 2-4 amphorae as well, strengthened by the presence of the stamp. As a matter of fact, the same stamp SEX APPVLEIO C[---] was found on a now lost tegula fragment that was walled in the middle section of the aqueduct constructed to bring water to Caska from the source near Kolan, about 8 km south of Caska (Ilakovac 2008: 132, 138) (Plate 3: 19). This aqueduct is considered to be older than the one in Novalja and it was probably financed by the Calpurnii Pisones to bring water to their estate (Radić Rossi and Zmaić 2009). Another roof tile with the same stamp of uncertain provenance (maybe Zaton north of Zadar?) is preserved in the Archaeological Museum of Zadar (Juras and Jurković Pešić 2016: 65-66). Sextus Appuleius can equally be the consul in 29 BC (who was governor of the province of Illyricum in 8 BC, cf. Syme 1989: 316-317, 30, 44-45; Dzino 2010: 141), as well as his son, who bore the same name and held a consulate in AD 14. Similar lettering of the stamp, indicating their contemporaneity and perhaps the same production center, is found on the brick stamped [--- C]AESAR III COS, which was found during the excavation of the Late Roman settlement in Caska (published in Kurilić 2016). According to the pattern in which the emperors are referred to in naming the consular pair, the stamp can be restored as [Imp(erator) C]aesar III co(n)sul (cf. CIL III 74; CIL X 3790, 5162; AE 1993, 1751a) or [Ti(berius) C]aesar III co(n)sul (cf. CIL IV 1885; CIL X 1333; CIL XI 1356). Thus, the emperor mentioned in the stamp could equally be Augustus (or better to say Octavian), who held his third consulate in 31 BC (Kienast et al. 2017: 57), or Tiberius, who held his third consulate in AD 18 (Kienast et al. 2017: 72).1

Figure 5. The Tunera and the finding spot of two Caska 1 type amphorae during the CissAntiqua 2013 excavation campaign (author: N. Stepan).

underneath the rim. The stamp on the first Caska 1 type amphora (C115 — Plate 2: 15) is illegible. The stamp of the second amphora (C120 — Plate 2: 16) occupied two rows, but only the first line is readable: M(arco)∙IVNIO∙SILANO, followed by a depiction of a fish/tuna fish? (Plate 2: 13 and 16). It is curious to note that his name is given in the dative or ablative case. Because of the same epigraphical content and similar lettering with the other two stamps discussed below, it is equally possible that the stamp refers to M. Iunius Silanus (cos. 25 BC) or M. Iunius Silanus (cos. AD 19; cf. Syme 1939: 432, 495, 527). The two amphorae both have numerical graffiti incised after the firing. They were applied on two very visible spots, in the middle right above and below their shoulders: the first amphora (C115) bears the sign XXV (which was later scratched out) (Plate 2: 12 and 15), while the second amphora (C120), with the stamp M∙IVNIO∙SILANO, is denoted XXXXV (Plate 2: 11 and 16). We noticed that on the first graffito, the second grapheme X has a vertical stroke on both sides, connecting the apices of diagonal strokes of letter X (Plate 2: 12). We could not find analogies of this sign, but because of the same dimensions and weight of the two amphorae, perhaps it indicated number 30. The two incised numerals could represent the weight in librae of

We are thankful to Dr. Nikola Cesarik (Croatian Academy of Sciences and Arts) who drew attention to this issue. 1 

85

Maja Grisonic, Nikolina Stepan

Plate 3. 18: Dressel 2-4 handle from the villa in Caska stamped SEX APPVLEIO∙COS (drawing: M. Rončević, photo: R. Mosković, courtesy of G. Skelac); 19: tile stamped SEX APPVLEIO C[---] from the Roman aqueduct Kolan — Caska (from Ilakovac 2008: 138, fig. 10) (author: N. Stepan).

86

Amphorae and tiles from the island of Pag

After the identification of the two pits for the arca lapidum in the villa in Caska, local manufacture of Dressel 2-4 wine amphorae seems needed. Regional eastern Adriatic variants of standard Roman Dressel 2-4 amphorae were also produced in the well-known Crikvenica workshop (Lipovac Vrkljan and Konestra 2018: 22, with previous bibliography) as well as at Loron in Istria (Maggi and Marion 2011).

in Salona, the capital of the province of Dalmatia (CIL III, 2111), while Plocianus is attested on a funerary inscription from Pola in Histria (CIL V, 126). Indeed, because of its paleographic characteristics, we can shrink the chronological frame of the PLOC stamp and put it in the close time span after the appearance of the three above-mentioned consular stamps. As the in-depth analysis of the different stamps from Caska is being carried out by our colleagues (Konestra et al. 2021), we would just like to give an example: in both the SEX APPVLEIO∙COS and PLOC stamps the P is still slightly open. The open P, with the curved bowl not touching the vertical bar, is a trait of republican tradition, which can persist in the Julio-Claudian age and even later, until the unification of the two segments (Salomies 2015: 169; Panciera et al. 1991). The PLOC stamp of a libertus should logically follow in time the earliest consular stamps, but should not exceed the limit of the early imperial period.

Tegulae with PLOC stamp (Plate 4) Another stamp should be added to the abovementioned recently discovered stamps in Caska. Four tegulae stamped PLOC have been found during the excavations of the villa (Skelac 2006), in different layers more widely datable before the age of Gallienus (AD 253268) (Grisonic and Stepan 2017: 77). Three roof tiles bearing the stamp PLOC were recovered from the area of Gramače in Caska Cove (Šonje 1958: 312-314), right next to the excavated part of the Roman villa. According to the same author, tegulae with the PLOC stamp used to be quite a frequent find in Caska in the past. A tile with the same stamp was seen by Sticotti (1940: 180) in one of the houses of the Palčić family. Both Šonje and Sticotti read the stamp POC, not noticing the ligature between the letters P and L. Matijašić notes that the same stamp is mentioned in the catalogue of stamps of the Archaeological museum of Poreč (n. 137), but he did not see it personally. Part of this material was brought to the museum after the founding of the Società Istriana di Archeologia e Storia Patria in 1884, originating from private collections of Istrian collectors of antiquities, who probably gathered the artifacts from different territories (personal communication of R. Matijašić). The frequency with which this stamp appears in Caska is curious, pointing toward the fact that they might have been manufactured locally.

Discussion The Caska 1 and Dressel 2-4 types of amphorae, as well as the roof tiles bearing three different stamps attested only in Caska and its surroundings, could all be local productions from Caska and its wider area, which included the possible Dressel 6B amphorae workshop in Novalja. The dolia from the villa in Caska, which because of their dimensions and weight were likely to have been produced locally, also show a similar fabric. Both the similar clay fabric and the formulations used in the stamps, together with the similar paleography of the capital letters, corroborate the statement that they could have been manufactured in the area of Caska and Novalja for local needs, and during the same or very close time period. The stamps (on the Caska 1 and Dressel 2-4 amphorae) and the typology of the amphorae (Dressel 6 B from Novalja and Caska 1) all point toward the period ranging from the Late Republic (c. 40 BC) to the early years of the reign of Tiberius (c. AD 20). For the moment, due to the fragmented evidence available, this timeframe should be enlarged to include the whole reign of Tiberius until AD 37.

Tegulae bearing the PLOC marks are of good fabric, made out of purified light yellow clay (Munsell 2.5 Y 8/4). The stamp PLOC is included in a concave rectangular cartouche measuring 6.3 x 2.3 cm (Plate 4: 20-22). The letters are in relief, their height being 1.8 cm. The letter L ligates with the preceding letter P. We have here a single-name formulation that we can probably read Ploc(ami) or Ploc(iani). Plocamus is a name of Greek origin, in the Roman world usually attested for slaves and freedmen (Solin 2003: 1234-1235) and he could be identified with the officinator of the workshop that was producing the tiles. The presence of slaves or freedmen of Greek origin in Caska during the Augustan period has already been proposed based on the graffiti incised on the Italian terra sigillata vessels recovered in the villa (Grisonic and Stepan 2017: 70, 97). Plocamus, divi Aug(usti) l(ibertus), is attested in Aquileia in the Julio-Claudian period (AD 14-41) (CIL V, 1319). In the second half of the 2nd century AD a Plocamus is attested

From a morphological point of view, the Caska 1 amphora resembles very closely the Dressel 6B amphorae from Novalja: they both have a characteristic groove underneath the rim and an ovoid body, a height ranging from 65 to 80 cm and width of about 35 cm. The two types differ because of the morphology of the rim (rounded in the Dressel 6B and an out-curving collar in the Caska 1), the neck (which in the Caska 1 is wider) and the different form of the handles. The Caska 1 amphora should be a local production, inserted inside the morphological tradition of the Adriatic Dressel 6B amphorae. 87

Maja Grisonic, Nikolina Stepan

Plate 4. 20-23: tiles stamped PLOC from the villa in Caska (drawings: S. Čule, M. Rončević, photos: R. Mosković, courtesy of G. Skelac) (author: N. Stepan).

88

Amphorae and tiles from the island of Pag

At this point of research, the primary content of the Caska 1 amphora remains unknown. The depiction of the fish/tuna fish? on the M∙IVNIO∙SILANO stamp is hypothetical, while the second line of the stamp, which probably mentioned the content of the amphora, is illegible. The two Caska 1 amphorae found on the seabed in front of the Tunera tower present abundant pitch on the internal surfaces. Usually, in the archaeological bibliography, this excluded the content of oil (see for example Carre 1985: 218, 225, with earlier bibliography; Carre et al. 2009: 223, 227, 230). The latest research on the topic, validated by chemical analyses (Garnier et al. 2011), showed that because of the porosity of ceramics, amphorae need to be internally impermeable, and that pitch was widely used for this purpose for amphorae transporting oil. According to the authors, the presence or absence of pitch on ceramic containers depends primarily on the preservation capacities of the context of finding (Garnier et al. 2011: 411-413). Furthermore, these studies have confirmed the common practice of reuse of amphorae in the Roman world, leading to the consideration that the content of amphorae should be punctually verified by chemical analyses.

different types of fish to obtain a sauce of poorer quality compared to garum (Borzić 2011: 72, n. 24). For the moment, we are lacking archaeological evidence of fish processing in Caska Cove during Antiquity. The Tunera tower was built during the 19th century above the remains of a destroyed Roman wall, which belonged to the sea façade of the maritime villa in Caska (Radić Rossi and Boetto 2020). In two spots of the cove we have found the remains of structures in opus signinum from Roman times: one is located in front of the Roman breakwater (Radić Rossi and Boetto 2020) and another was built on the nearby Zrće peninsula to the SW of Caska Cove, on the opposite side from where the Tunera tower stands (Radić Rossi and Brkan Klarin 2018; Radić Rossi and Boetto 2020). The function of these structures remains unknown (Radić Rossi and Boetto 2020), but due to their vicinity to the sea and the lining in waterproof opus signinum, we can suppose that they were connected to processing of fish and/or other marine resources. Fishing for migratory fish such as tuna or mackerel produced a great amount of fish all at once and the surplus had to be sold or processed immediately (Bekker-Nielsen 2010: 203; Marzano 2013: 66-79), hence the need for a local production of ceramic containers in which fish could be packed. We believe that fish products as primary content of the Caska 1 amphorae can be suggested by exclusion: the archaeological evidence from Novalja and Caska hints to the local production of Dressel 6B (for oil from the Lun peninsula) and Dressel 2-4 (for wine produced in the villa in Caska). In this portfolio, we are lacking containers for the harvest from fishing, which must have been a fervent activity in Caska long before it is mentioned in archival documents. For the moment, the Caska 1 type amphora seems to be attested only at Caska: it may have been used for local storage and consumption, but because it appears quite frequently, we can suppose that it was also designed for trade, the scale of which is still to be determined.

Aware of all these problematics, we would like to avoid suprainterpretations of the possible contents of the Caska 1 amphorae, which deserves further analyses. Still, we would like to promulgate our hypothesis on their possible content of fish products, most likely related to tunas. Considering that at least from the 17th century the main economic activity in Caska Cove was tuna fishing and that Caska was known for this in the Adriatic, we can assume that fishing and fish processing was already a common activity in Roman times. The 1637 statute of the Pag Commune gives regulations about the trade of mackerel in Caska Cove, not mentioning tunas. Tuna trade in Caska is remarked a little later, in 1662: it was regulated by the Venetian Republic, which shipped tunas mainly to Venice (Basioli 1962: 47). In his Viaggio in Dalmazia, Alberto Fortis (1741-1803), the Venetian monk who travelled throughout this region, wrote that the Gulf of Pag was frequented by schools of tuna that were then unable to leave (Fortis 2004: 267, 279280). Tuna watching towers, among which the Tunera in Caska is the oldest, were located in at least five different spots on the island and were in function until the mid-20th century (Basioli 1962: 47; Magaš 2011: 41). Record tuna catches in Caska occurred during the 18th and 19th centuries (Basioli 1962: 48). The net used to close Caska Cove to catch the tunas, called preteg, was the longest on the Adriatic Sea, measuring 920 m, while smaller tuna fishing nets were called trate (Basioli 1962: 47; Radić Rossi and Zmaić 2013: 61).

For the early imperial period, it is generally thought that fish sauces were imported to the Adriatic from other regions of the Mediterranean, and that they were stored in reused local wine or oil amphorae (Carre et al. 2009: 232). In Caska, this could be true if the Caska 1 amphorae were originally made for oil storage: we could think that, compared to the Dressel 6B from Novalja, their morphological differences were due to their manufacture in different ceramic workshops or maybe Caska 1 amphorae followed chronologically the earlier Dressel 6B. Yet, Dalmatia and Liburnia had all the natural characteristics to develop an important fish processing industry, of which, with the exception maybe of the

It is interesting to note that the beloved Roman sauce muria, of which Pliny says that Dalmatia was known (NH, XXXI, 93-95), was made of tuna or by mixing 89

Maja Grisonic, Nikolina Stepan structures in Kumenat (see Parica 2017) we are lacking direct archaeological evidence (Borzić 2011; Lipovac Vrkljan and Konestra 2017: 54-55). If Dalmatia was known for muria in the second half of the 1st century AD, it is obvious that the tradition of this industry had started earlier. It is noteworthy that very few Hispanic fish amphorae from the 1st century AD or earlier have been recovered from Dalmatian sites, which could be partially because of the poor state of research of certain sites (Borzić 2011: 82; Borzić and Ožanić Roguljić 2018). Still, this points toward the existence of local fish processing industries, which needed their own ceramic containers to ship their products in the province to supply local markets. Caska 1 amphorae could therefore represent one of these containers, one of the oldest ones in Dalmatia, before the beginning of production of the small-sized Adriatic fish amphorae after the first half of the 1st century AD (Carre et al. 2009). The Adriatic fish amphorae, which probably held fish sauces like liquamen, indirectly indicate the implantation of fish processing industries in the Adriatic (Auriemma 2000; Carre et al. 2009). Fish amphorae of very small scale (3040 cm of height) have also been produced in Crikvenica, probably for the trade of liquid fish sauces (Lipovac Vrkljan and Konestra 2017: 53-54).

cultivated with vines, and maybe fruit and vegetables, while wheat was probably imported from the area of Ravni kotari on the mainland, north of the colony of Iader (Zadar). The rocky landscapes, which today cover 86% of the surface of the island (Magaš 2011: 39), served as pastures for small cattle, mainly sheep. Pliny (NH, VIII, 191) and Martial (Epigr. 14, 140) both mention the typical Liburnian sheep wool coats cuculli liburnici, and Pag is at present very famous for sheep cheese, which probably constituted an important economic activity even before the arrival of Romans. In Caska Cove, the remains of different Roman structures covered with waterproof mortar hint to various activities connected to the exploitation of marine resources, probably mainly fish processing, which is the traditional source of sustenance of Caska, documented by the archival documents from the 17th century on. One of the fundamental ingredients for fish preservation and processing was salt, for which the island of Pag is known. The Pag saltpans, the biggest in Croatia, are located on the southern edge of the Gulf of Pag (Figure 1). They are attested from the early Middle Ages, but most likely already developed in Antiquity (Zaninović 1991: 261; Peričić 2001: 46). Although we are lacking archaeological evidence of them, we can imagine that during Roman times salt from Pag might already have been shipped to Caska on the northern side of the gulf, or smaller saltpans might have been located in the coves closer to Caska (or in Caska itself). In the 12th century, saltpans existed in Stara Novalja Bay (Peričić 2001: 47), just 1 km north of Caska.

Conclusions The presence of three different types of amphorae (local Dressel 2-4, Dressel 6B from the presumed Novalja workshop and the Caska 1 type) in the same archaeological layer buried in the two pits inside the villa in Caska shows the contemporaneity of disposal (and use – assumed on the basis of stamps and typochronology of amphorae) of three different types of ceramic containers, which were probably produced locally, on a very small geographic area. The local manufacture of these amphorae is very likely connected to the self-sufficiency and food storage needs of various villae gravitating around the fertile Novalja field, but maybe also to the maritime trade of the surplus of resources produced on the island, among which we could enumerate oil packed inside the Dressel 6B amphorae in Novalja, very close to its port, and possibly tuna? and/ or other fish products? stored inside the Caska 1 type amphoras. Following the hypothesis of the continuity of exploitation of the landscape and natural resources on the island, and based on the available archaeological and historical documentation, we can suppose that the inhabitants of the island, maybe even before the arrival of Romans but certainly under the Roman principate, already chose distinctive areas for specific agricultural and marine exploitation, to optimize the human efforts and to obtain the maximum from the available natural resources (Figure 1). The Lun peninsula, like today, was most likely chosen for oil production; the Novalja field was probably, like nowadays, intensively

Although many questions remain open, the goal of this article was to contribute to the interdisciplinary discussion on the economy of the island of Pag during Roman times. The presumed local ceramic containers, which we have supposed based on stamps, amphorae typology and frequency, provide indirect evidence of different economic activities in the past. We discussed especially our hypothesis on the content of the Caska 1 type amphorae, which will hopefully be evaluated by new findings and future archaeometric analyses. The diffusion of these products from the island of Pag, the ancient Cissa, is yet to be determined.2

We would like to thank many international scholars with whom we discussed the topics exposed in this paper. Our special gratitude goes to Asst. Prof. Igor Borzić (University of Zadar), Dr. Laëtitia Cavassa (CNRS/CCJ) and Dr. David Ruff (Institute of Nautical Archaeology). 2 

90

Amphorae and tiles from the island of Pag

romana. Produzione e commercio nelle regioni dell’Alto Adriatico. Atti del convegno, Padova, 16.2.2007: 215–238. Roma: Edizioni Quasar. Cipriano, S. 2009. Le anfore olearie Dressel 6B, in S. Pesavento Mattioli and M. - B Carre (eds) Olio e pesce in epoca romana. Produzione e commercio nelle regioni dell’Alto Adriatico. Atti del convegno, Padova, 16.2.2007: 173–189. Roma: Edizioni Quasar. Dzino, D. 2010. Illyricum in Roman Politics. 229 BC — AD 68. Cambridge/New York: Cambridge University Press. Fortis, A. 2004. Put po Dalmaciji / Viaggio in Dalmazia dell’abate Alberto Fortis. Venezia (1774), Split: Marjan Tisak. Garnier, N., T. Silvino and D. Bernal Casasola 2011. L’identification du contenu des amphores: huile, conserves de poissons et poissage. SFECAG. Actes du Congrès d’Arles 2011: 397‒416. Gluščević, S. 1989. Antička keramika s otoka Paga, in Ž. Rapanić (ed.) Arheološka istraživanja na otocima Krku, Rabu i Pagu i u hrvatskom primorju. Znanstveni skup, Krk, 24.–27.9.1985 (Izdanja HAD-a 13): 73–87. Zagreb: Hrvatsko arheološko društvo. Gluščević, S. 2017. Amfore Dressel 6B iz Novalje na otoku Pagu (Dressel 6B amphorae from Novalja on the Island of Pag), in G. Lipovac Vrkljan, I. Radić Rossi and A. Konestra (eds) AdriAmphorae. Amphorae as a resource for the reconstruction of economic development in the Adriatic region in Antiquity: local production. Proceedings of the workshop, Zagreb, 21st April 2016: 63– 67. Zagreb: Institut za arheologiju. Grisonic, M. 2017. Amphorae from Caska in the AugustoTiberian period: imports and local productions?, in G. Lipovac Vrkljan, I. Radić Rossi and A. Konestra (eds) AdriAmphorae. Amphorae as a resource for the reconstruction of economic development in the Adriatic region in Antiquity: local production. Proceedings of the workshop, Zagreb, 21st April 2016: 68–79. Zagreb: Institut za arheologiju. Grisonic, M. and N. Stepan 2017. Italska terra sigillata iz rimske vile rustike u uvali Caska na otoku Pagu / Terra sigillata italica dalla villa rustica romana nella baia di Caska sull’isola di Pag. Archaeologia Adriatica 11: 69–143. Hilje, E. 2011. Spomenici povijesnoga graditeljstva na otoku Pagu, in V. Skračić (ed.) Toponimija otoka Paga: 93–180. Zadar: Sveučilište u Zadru, Centar za jadranska onomastička istraživanja. Ilakovac, B. 2008. Roman aqueducts on the Island of Pag. Vjesnik Arheološkog Muzeja u Zagrebu 41: 129–166. Juras, I. and F. Jurković Pešić 2016. Tegule s pečatom iz antičke zbirke Arheološkoga muzeja Zadar /Stamped tegulae from the Roman antiquity collection of the Archaeological Museum Zadar. Diadora 30: 31–75. Kienast, D., W. Eck and M. Heil 2017. Römische Kaisertabelle. Grundzüge einer römischen Kaiserchronologie (6. überarbeitete Auflage). Darmstadt: wbg Academic.

References Auriemma, R. 2000. Le anfore del relitto di Grado e il loro contenuto. Mélanges de l’Ecole française de Rome. Antiquité 112: 27–51. Basioli, J. 1962. Tunolov na Jadranu. Zagreb: Jadranski Institut Jugoslovenske Akademije Znanosti i Umjetnosti. Bekker-Nielsen, T. 2010. Fishing in the Roman World, in T. Bekker-Nielsen and D. Bernal Casasola (eds) Ancient Nets and Fishing Gear. Proceedings of the International Workshop on ‘Nets and fishing gear in Classical Antiquity: a first approach’, Cádiz, 15–17.11.2007: 187–203. Cadiz – Aarhus: Servicio de Publicaciones de la Universidad de Cádiz, Aarhus University Press. Bezeczky, T. 1998. The Laecanius Amphorae Stamps and the Villas of Brijuni. Wien: Österreichische Akademie der Wissenschaften. Bezeczky, T. 2016. Az isztriai olívaolaj gyártás és a Laecanius amphorák/ Olive oil production in Istria and the amphorae of Laecanii. Archeometriai Műhely 13/2: 69‒94. Blečić, M. and I. Radić Rossi 2004. Novaljsko područje u antici, in I. Radić Rossi (ed.) Skriveno blago Novalje: 21–32. Novalja – Zagreb: Grad Novalja, Hrvatski ronilački savez. Boetto, G. and I. Radić Rossi 2012. Šivani brod u uvali Caski na otoku Pagu: rezultati istraživačke kampanje 2011. Histria Antiqua 21: 609–622. Boetto, G. and I. Radić Rossi 2017. Ancient ships from the bay of Caska (Island of Pag, Croatia), in J. Litwin (ed.) 14th International Symposium on Boat and Ship Archaeology: Baltic and Beyond. Change and continuity in shipbuilding. Proceedings of the Fourteenth International Symposium on Boat and Ship Archaeology, Gdańsk 2015: 279‒288. Gdańsk: National Maritime Museum. Borzić, I. 2011. Hispanski garum na Burnumskom stolu/ Hispanic garum at a Burnum table. Archaeologia Adriatica 5: 65–88. Borzić, I. and I. Ožanić Roguljić 2018. Hispanic products in Dalmatia and the Croatian part of Roman Pannonia. Rei Cretariae Romanae Fautorum Acta 45: 511–523. Botte, E., 2009. Salaisons et sauces de poissons en Italie du sud et en Sicile durant l’Antiquité. Naples: Publications du Centre Jean Bérard. Brun, J.-P., 2004. Archéologie du vin et de l’huile dans l’Empire romain. Paris: Editions Errance. Carre, M. - B. 1985. Les amphores de la Cisalpine et de l’Adriatique au début de l’Empire. Mélanges de l’École française de Rome – Antiquité 97: 207–245. Carre, M. - B. and S. Pesavento Mattioli, S. 2003. Tentativo di classificazione delle anfore olearie adriatiche. Aquileia nostra 74: 453–476. Carre, M. - B., S. Pesavento Mattioli and C. Belotti 2009. Le anfore da pesce adriatiche, in S. Pesavento Mattioli and M.-B Carre (eds) Olio e pesce in epoca 91

Maja Grisonic, Nikolina Stepan Konestra, A., A. Kurilić and G. Lipovac Vrkljan 2021. Tiles and amphorae in the Roman province of Dalmatia: evidence of stamps, in D. Rigato, M. Mongardi, M. Vitelli Casella (eds) Adriatlas 4. Produzioni artigianali in area adriatica: manufatti, ateliers e attori (III sec. a.C. – V sec. d.C.) (collection PrimaLun@ 8): 145‒165. Pessac: Ausonius éditions Konestra, A. and G. Lipovac Vrkljan 2018. Pottery workshops in the coastal area of Roman Dalmatia: landscape, spatial organization, ownership, in M. Janežic, B. Nadbath, T. Mulh, I. Žižek (eds) Nova odkritja med Alpami in Črnim morjem. Rezultati raziskav rimskodobnih najdišč v obdobju med leti 2005 in 2015. Zbornik 1. mednarodnega arheološkega simpozija, Ptuj, 8. in 9. oktober 2015 (Monografije CPA 6): 127–141. Ljubljana: Zavod za varstvo kulturne dediščine Slovenije. Kunkera, J. 1982. Novalja kroz 30 stoljeća / Novalja through 30 centuries. Novalja. Kurilić, A. 2004. Epigrafski spomenici na prostoru Novalje / Epigraphic Monuments in the Area of Novalja. Novalja – Zadar: Grad Novalja, Arheološki muzej Zadar, Sveučilište u Zadru, Odjel za povijest. Kurilić, A. 2011a. Otok Pag od prapovijesti do kraja antičkog razdoblja, in V. Skračić (ed.) Toponimija otoka Paga: 51–91. Zadar: Sveučilište u Zadru, Centar za jadranska onomastička istraživanja. Kurilić, A. 2011b. Kasnoantički stambeni objekt iz Caske (Cissa) na otoku Pagu. Histria Antiqua 20: 405–413. Kurilić, A. 2013. Antičko i kasnoantičko razdoblje, in I. Oštarić and A. Kurilić (eds), Arheološka karta otoka Paga: 215–277. Novalja: Ogranak Matice hrvatske u Novalji, grad Novalja. Kurilić, A. 2016. Roman tile stamped [C]AESAR III COS. Antichità altoadriatiche 85: 377–386. Kurilić, A. and Z. Serventi 2018. The Caska necropolis– exceptions, rituals and ‘deathscapes’, in V. Nizzo (ed.) 3° Incontro Internazionale di Studi di Antropologia e Archeologia a confronto. Archeologia e antropologia della morte: 765–777. Lipovac Vrkljan, G. and A. Konestra 2017. Crikvenička riblja amfora − indikator proizvodnje/trgovine ribljim prerađevinama? / Crikvenica fish amphora − indicator of production/commerce of fish product?, in G. Lipovac Vrkljan, I. Radić Rossi and A. Konestra (eds) AdriAmphorae. Amphorae as a resource for the reconstruction of economic development in the Adriatic region in Antiquity: local production. Proceedings of the workshop, Zagreb, 21st April 2016: 48–62. Zagreb: Institut za arheologiju Lipovac Vrkljan, G. and A. Konestra 2018. Approaching the Roman economy of Province Dalmatia through pottery production. The Liburnia case study, in G. Lipovac Vrkljan and A. Konestra (eds) Pottery Production, Landscape and Economy of Roman

Dalmatia. Interdisciplinary approaches: 14–36. Oxford: Archaeopress. Magaš, D. 2011. Zemljopisna obilježja otoka Paga u funkciji upoznavanja njegove toponimije, in V. Skračić (ed.) Toponimija otoka Paga: 5–49. Zadar: Sveučilište u Zadru, Centar za jadranska onomastička istraživanja. Maggi, P. and Y. Marion 2011. Le produzioni di anfore e di terra sigillata a Loron e la loro diffusione/ Proizvodnja i distribucija amfora i terrae sigillatae u Lorunu, in G. Lipovac Vrkljan, I. Radić Rossi and B. Šiljeg (eds) Rimske keramičarske i staklarske radionice. Proizvodnja i trgovina na jadranskom prostoru. Zbornik I. međunarodnog arheološkog kolokvija, 23. — 24. listopada 2008., Crikvenica : 175‒187. Crikvenica: Muzej Grada Crikvenice, Institut za arheologiju. Martín i Oliveras, A. 2012. Anàlisi tecnofuncional d’estructures productives i vitivinícoles d’època romana. Identificació i localització a Catalunya de fosses de maniobra de premses de biga amb contrapès tipus arca lapidum. Pyrenae 43/2: 59–98. Martín i Oliveras, A., D. J. Martín-Arroyo Sánchez and V. Revilla Calvo 2017. The wine economy in Roman Hispania. Archaeological data and Modellization, in J. Remesal Rodriguez (ed.) Economia romana. Nuevas perspectivas/The Roman economy. New perspectives: 189–237. Barcelona: Universitat de Barcelona Edicions. Marzano, A. 2013. Harvesting the sea. The exploitation of Marine Resources in the Roman Mediterranean. Oxford: Oxford University Press. Matijašić, R. 1989. Rimske krovne opeke s radioničkim žigovima na području sjeverne Liburnije / Le tegole bollate romane nel territorio della Liburnia Settentrionale, in Ž. Rapanić (ed.) Arheološka istraživanja na otocima Krku, Rabu i Pagu i u hrvatskom primorju. Znanstveni skup, Krk, 24.–27.9.1985 (Izdanja HAD-a 13): 61–71. Zagreb: Hrvatsko arheološko društvo. Panciera, S., M. Bertinetti, I. Di Stefano Manzella, R. Friggeri, G.L. Gregori, C. Ricci, S. Gentili, G. Vergantini, M.L. Caldelli, S. Priuli and N. Petrucci 1991. Inscriptiones latinae liberae rei pubblicae, in Epigrafia. Actes du colloque international d’épigraphie latine en mémoire de Attilio Degrassi pour le centenaire de sa naissance. Actes de colloque de Rome (27-28 mai 1988) (Publications de l’École Française de Rome 143(1)): 241–491. Rome: École Française de Rome. Parica, M. 2017. Mogućnosti antičke riboprerađivačke industrije u Pašmanskom kanalu / The possibility of ancient fish-processing industry in the Pašman channel, in G. Lipovac Vrkljan, I. Radić Rossi and A. Konestra (eds) AdriAmphorae. Amphorae as a resource for the reconstruction of economic development in the Adriatic region in Antiquity: local production. Proceedings of the workshop, Zagreb, 21st April 2016: 86–98. Zagreb: Institut za arheologiju 92

Amphorae and tiles from the island of Pag

Peričić, Š. 2001. Proizvodnja i prodaja paške soli u prošlosti. Radovi Zavoda za povijesne znanosti HAZU u Zadru 43: 45–83. Radić Rossi, I. 2006. Lokalitet: Caska – podmorje. Hrvatski arheološki godišnjak 2/2005: 285–286. Radić Rossi, I. and G. Boetto 2010. Arheologija broda i plovidbe: šivani brod u uvali Caski na Pagu – istraživačka kampanja 2009. Histria Antiqua 19: 299– 307. Radić Rossi, I. and G. Boetto 2011. Šivani brod u uvali Caska na Pagu – istraživačka kampanja 2010. Histria Antiqua 20: 505–513. Radić Rossi, I. and G. Boetto 2020. The Roman Scuttled Ships and Harbour Structures of Caska, Pag Island, Croatia in their Cultural and Historical Context. International Journal of Nautical Archaeology 49/2: 263–290. Radić Rossi, I. and Z. Brkan Klarin 2014. Uvala Caska na otoku Pagu. Stručno izvješće o arheološkom istraživanju u podmorju i priobalju uvale, studeni 2014, unpublished excavation report. University of Zadar, Department of Archaeology. Radić Rossi, I. and Z. Brkan Klarin 2018. Izvješće o snimanju postojećeg stanja objekta na rtu Zrće, unpublished excavation report. University of Zadar, Department of Archaeology. Radić Rossi, I. and J. Zmaić 2009. Podzemni antički vodovod u Novalji na otoku Pagu / Underground Roman aqueduct in Novalja at the island of Pag. Histria Antiqua 18/2: 147–166. Radić Rossi, I. and J. Zmaić 2013. Sveobuhvatna prezentacija prirodne, arheološke i tradicijske baštine na području Grada Novalje, unpublished report. Rousse, C., D. Munda, G. Benčić, K. Gergeta Sotončić, C. Vaschalde, P. Maggi and V. Dumas 2019. Loron/Santa Marina (Tar-Vabriga, Poreč, Croatie). Chronique des activités archéologiques de l’École française de Rome [En ligne], Balkans (mis en ligne le 07 août 2019)

Salomies, O. 2015. The Roman Republic, in Ch. Bruun and J. Edmondson (eds) The Oxford Handbook of Roman Epigraphy: 153–177. Oxforf: Oxford University Press. Skelac, G. 2006. Lokalitet: Caska – gospodarski kompleks. Hrvatski arheološki godišnjak 2/2005: 281–283. Skelac, G. 2007. Lokalitet: Caska – gospodarski kompleks. Hrvatski arheološki godišnjak 3/2006: 315–316. Solin, H. 2003. Die Griechischen Personnennamen in Rom, 3. band. Berlin – New York: De Gruyter. Sticotti, P. 1940. Pag. Serta Hoffilleriana (= VHAD, ns, 18– 21/1937–1940): 179–181. Syme, R. 1939. The Roman Revolution. Oxford: Oxford University Press. Syme, R. 1989. The Augustan Aristocracy. Oxford: Oxford University Press. Šašel, J. 1963. Calpurnia L. Pisonis auguris filia. Živa antika 12/2: 387–390. Šonje, A. 1958. Nalaz rimskog natpisa na Caski kod Novalje na otoku Pagu. Živa antika 8: 311–322. Tassaux, F., R. Matijašić and V. Kovačić (eds) 2001. Loron (Croatie): un grand centre de production d’amphores à huile istriennes : Ier — IVe s. P.C. (Ausonius-Mémoires 6). Bordeaux: Ausonius. Toniolo, A. 1991. Le anfore di Altino. Altino: Società archeologica veneta. Zaninović, M. 1991. Sol u antici naše obale / Salz in der Antike der Adriaküste, in B. Čović (ed.) Zbornik radova posvećenih Alojzu Bencu (Posebna izdanja ANU BiH 95, Posebna izdanja Odjeljenja društvenih nauka 27): 255–264, Sarajevo: Akademija nauka i umjetnosti Bosne i Hercegovine.

93

An underwater archaeological survey of Havišće and Lokvišća Coves and Cape Ertak Igor Borzić, Mate Parica, Mato Ilkić, Tea Rosić Abstract In 2015 an underwater survey of the area comprising Havišće and Lokvišća Coves and Cape Ertak in Jadranovo (NE Adriatic, Croatia) was carried out through a collaboration of the Crikvenica Town Museum and the Department of Archaeology of the University of Zadar. An area rich in archaeological evidence dating back to the Bronze Age, Lokvišća Cove hosts the remains of early Roman structures and a necropolis, and a late Roman complex, with evidence of use of the area well into the 17th century. The survey allowed to recover several amphorae and pottery finds, and recognise the remains of a possible Roman or late Antique docking structure. These findings are herein discussed and placed within the immediate local and a wider regional context.

Key words Jadranovo, Lokvišća, archaeological survey of seafloor, dock from Antiquity, amphorae, pottery workshop in Crikvenica.

Introduction At the end of October 2015, underwater archaeological reconnaissance of the seafloor of Havišće Cove, Ertak (Rtac) Peninsula and the area from Lokvišća to Trstena Coves was conducted.1 This area is part of the Havišće Peninsula and is situated roughly one kilometre south of the town of Jadranovo, in the Kvarner gulf, the northen most large inlet of the eastern Adriatic. The peninsula juts into the strait between the mainland and the island of Krk, which gives it an outstanding strategic/surveillance position. Its highest point is a hillock called Čelo (62 m/asl) which descends westward to the coves of Lokvišća and Trstena. Lokvišća Cove is circular and resembles a pond with a narrow opening on its western side which provides sound protection against bora and sirocco winds (Figure 2). The lower portions of the peninsula are covered with deposits of sand and loamy soil, which facilities water retention, so there are many springs in the vicinity.

off Lokvišća Cove, where the remains of an early Roman necropolis from the 1st and 2nd centuries AD were ascertained in the late 19th and early 20th century (Brunšmid 1900: 118-120; Jovanović 1984: 54-56), and they were later verified by more recent excavations (Ujčić Grudenić 2014: 401, 402).2 Near the necropolis in the sheltered cove of Lokvišća is the presumed location of a settlement from Antiquity. Individual scholars have tied this area to the civil war between Pompey and Caesar and the blockade of the island of Krk in 49 BC, when the cove may have been temporarily occupied by

The entire peninsula is an archaeological zone that was inhabited from the early Bronze Age to the 17th century, and it consists of several sites. The first phase of human habitation can be found on Čelo, with traces of a fortified early Bronze Age settlement. At this hillock’s foot, on the meadows next to Lokvišća and Trstena Coves, traces of coastal harbour activities were found dating to the early and late Iron Age, particularly the 2nd and 1st centuries BC (Starac 2002: 189-194). From the coastal side, Ertak Peninsula closes Participants in this research were: Tea Rosić (Crikvenica Town Museum), Mate Parica and Mato Ilkić (Archaeology Department, University of Zadar), Sara Iglić, Paolo Iglić, Maja Kaleb, Stjepan Špalj and Mario Sajatovski (Figure 1). 1 

Figure. 1. The archaeology team before a dive (photo: DC Mihurić Selce). Lokvišća Cove has been explored, with interruptions, since 1993. The Maritime and History Museum of the Croatian Littoral in Rijeka has been conducting systematic excavations for years, led by Ranko Starac, and in 2009 the Crikvenica Town Museum also became involved in this research. 2 

Roman Pottery and glass manufactures (Archaeopress 2022): 94–105

An underwater archaeological survey of Havišće and Lokvišća Coves and Cape Ertak

Figure. 2. North-western slope of Cape Ertak (photo: M. Parica).

Roman legions (Starac 2002: 194; Bilić Dujmušić 2014: 107-124). The visible walls still present today in the cove’s northern part are from a somewhat later era, a part of a fortification complex from Late Antiquity, the 3rd or early 4th century, on which a harbour warehouse stood in the Early Modern era, from the 15th through 17th centuries (Starac 2002: 195-198). A variety of coins were found in the peninsula’s coastal belt adjacent to Lokvišća Cove (Starac 2002: 193, 197198). Among these numismatic materials, in which the standard Roman coins from the mid-1st century to Late Antiquity predominate, the copper-alloy coin from northern Africa dating to the pre-imperial period stands out (Figure 3). It originated in Numidia, and has been ascribed to Masinissa and his successors (208-148 BC and later).3 Such north African coins, together with Carthaginian and, partly, Ptolemaic coins made their way to Liburnia by sea, where an enormous quantity was in circulation as a means of payment, whence their use also expanded to Iapodia by further trade across the pass over Velebit.4 The Numidian coin from the shoreline at Lokvišća Cove is thus far the northernmost such numismatic find in Liburnia.

Figure. 3. Coin from Numidia (Masinissa and his successors, 208-148 BC and later) (photo: M. Ilkić).

Underwater survey results A survey of the wider underwater zone from Havišće Cove through Rtac Peninsula (Figure 4) and Lokvišća to Trstena Cove has ascertained the existence of individual finds, mostly amphora sherds. They are mostly situated in a belt at depths of 15 to 25 meters. Besides movable archaeological materials, a sunken docking structure was also found. There is a submerged stone embankment that has the form of a small pier in the north-western end of Havišće Cove, more specifically, it is in an inlet called Kut, where a sandy beach ends and the limestone rocks of Cape Ertak begin (Figures 4-6). It was made by packing amorphous stones with an average diameter of 50 cm, although there are

The obverse features a laureate head of a bearded ruler in left profile, while the middle of the reverse contains a horse galloping leftward. On such coins from Numidia, see: SNG Cop no. 504-517; SNG Ita Milano XIV, no. 1-17. 4  On the circulation of coins in the Hellenistic and Late Republic periods in Liburnia and its deeper hinterland, see: Šešelj and Ilkić 2014: 43-53.; Ilkić and Šešelj in press. 3 

95

Igor Borzić, Mate Parica, Mato Ilkić, Tea Rosić

Figure. 4. Aerial view of Lokvišća Cove (photo: I. Pervan).

several larger stone chunks. The pier is semi-circular with a width of roughly 10 meters, while its tip consists of a flat plateau which merges with the natural seashore at the same level. The pier’s end is at a depth of roughly 1.5 meters, while the deepest part of the embankment is 2.8 meters below today’s sea level.

Among the shoreline structures, the entrance to Lokvišća Cove should also be underscored, as it is one of the rare fully protected coves in the wider area. The seaside entrance to the cove is 35 m wide, while the greatest depth in the middle of the entry section is approximately 2.2 m below the modern sea level. The seafloor in this section consists of tiny stones with numerous sherds of amphorae, tegulae and pottery. There is no trace of bedrock, so that the passage may have been deeper in Antiquity, and the harbour entirely functional for even larger vessels. The central section of Lokvišća Cove contains thick deposits of dark mud, and it is difficult to reconstruct the seafloor’s level here.

No archaeological materials were noted in the area of the embankment and its immediate vicinity; several examples of ballast stones constitute the sole evidence of the pier’s use. Since there are no datable archaeological materials, the pier may be dated on the basis of its position in relation to the present-day sea level, and this would generally be Antiquity or Late Antiquity.5 The absence of materials, and the location at a place that is exposed to southerly winds tells us that the pier did not belong to the classical berthing facility of a villa, rather it probably played a specialized role with rarer use. It is notable that the seawater around the pier mixes with freshwater, so one possibility is that it was built to supply vessels from a spring, even though the position is considerably less secure than Lokvišća Cove.6

The archaeological materials found during an underwater inspection are rather poorly preserved, but several indicative diagnostic fragments very clearly indicate a period when this place was the scene of a certain level of activity which will only be fully elucidated by future systematic excavations. The earliest amphora sherd belongs to the so-called Greco-Italic type produced by the southern Italic workshops in the 4th and 3rd centuries BC (Will 1982: 338-356; Vandermersch 1994; Olcese 2004: 173-192).

In general, on sea level changes since Antiquity: Benac and Šegota 1990; Antonioli et al. 2007; Auriemma and Solinas 2009; Faivre et al. 2010; Lambeck et al. 2004. 6  A very similar parallel can be drawn with the dock on the island of Silba dated to Late Antiquity, as it was situated in a less safe cove 5 

partly for similar reasons (Parica 2016: 307-322). On the importance of freshwater sources in ports during Antiquity: Blackman et al. 2013.

96

An underwater archaeological survey of Havišće and Lokvišća Coves and Cape Ertak

Figure. 5. View of the north-western section of Havišće Cove with position of stone embankment marked (photo: I. Pervan).

Figure. 6. Underwater photograph of submerged stone embankment (photo: M. Parica).

Their great popularity, i.e., the intense wine trade throughout the Mediterranean, thus including eastern Adriatic coast (Kirigin 1994: 15-24; Jurišić 2000: 5-6), also led to the assumption of its production in workshops outside of the core region (for eastern Adriatic Issa, see: Kirigin et al. 2005: 13). A sherd from the seafloor at Jadranovo (Plate I,1), which, incidentally, was on the route by which Greco-Italic amphorae expanded

into the northern Adriatic (cf. the distribution map in Kirigin 1994: 17, fig. 3), based on its elegantly elongated bottom, probably belongs to their younger variant from the end of the 3rd and early 2nd century BC (cf. Form C based on Will 1982: 347-348). Four Lamboglia 2 (Plate I, 2-5) amphorae sherds date to a somewhat more recent period in the 2nd and 1st 97

Igor Borzić, Mate Parica, Mato Ilkić, Tea Rosić

Plate I. Lokvišće, underwater finds of amphora fragments (photo: T. Rosić, author: I. Borzić).

centuries BC. This was the best-known wine amphora produced in the entire Adriatic, particularly the central zones of its western coast (concisely and topically in: Carre et al. 2014: 417-428; Cipriano and Mazzocchin 2017: 33-47, esp. 40, fig. 2). Contextualization of epigraphic evidence with indicators from field

excavations presumes their production on the eastern coast as well, specifically in Issa and Tasovčići in the Neretva Valley (Cambi 1991: 55-65; Kirigin 1996: 152; Kirigin et al. 2005: 10-11; Čargo and Miše 2010: 16). The exceptional quantity of this amphora type at undersea (Radić Rossi 1993; Jurišić 2000: 6) and near-coastal 98

An underwater archaeological survey of Havišće and Lokvišća Coves and Cape Ertak

terrestrial sites (Kirigin et al. 2005: 14-15) testifies to the lively wine trade throughout the eastern Adriatic seaboard, trade oriented toward indigenous, but also Hellenistic and Roman settler communities (Šuta 2017: 26-32). Near Jadranovo, they can be found at sites such as the Crikvenica harbour (Lipovac Vrkljan 2016: 91, cat. no. 29) and Omišalj Bay and Voz Cove on the island of Krk (Radić Rossi 1993: 39-41, no. 17 and 19), but the topography of finds along the entire shore of the Velebit Channel (Radić Rossi 1993; Glavaš et al. 2020: 271-286) and the historical events that played out precisely in this area (Bilić-Dujmušić 2014: 107-124) make these finds entirely expected.

with typically moulded rims, necks, ribbed handles and ring-shaped flat bottoms belong to the Crikvenica variant of flat-bottomed wine amphorae (Crikvenica 1 type), dated to the end of the first half of the 2nd century (Lipovac Vrkljan 2011: 3-18; 2016: 57; 2017: 188-189; Lipovac Vrkljan and Konestra 2018: 21-22) (Plate II, 3-5). The direction of export of the Vinodol wine conveyed in them can for now be charted on the basis of undersea and terrestrial finds throughout Liburnia, so their discovery in Jadranovo came as no great surprise (Gluščević 2008: 411; Radić Rossi 2011: 23-24; Pešić 2012: 44-47; Lipovac Vrkljan and Ožanić Roguljić 2013: 255-270; Mustaček et al. 2013: 63 and 65; Romanović 2017: 387-406; Borzić et al. 2018: 50-56). But the fact that exports exceeded even these boundaries is demonstrated by the thus far few, but very indicative examples in the continental hinterland of the Liburnian coast in Žuta Lokva (Ožanić Roguljić 2017, 81) and especially in the then world-renowned mercantile metropolis of Aquileia (Maggi 2018: 57-61).

Two sherds of Dressel 6A (Plate I, 6, Plate II, 1) amphorae are chronologically tied to the preceding type. In both cases, they are part of the massive vertical, in one case also gently rounded, rim (Plate II, 1) resting on a high neck. Such morphological overlap with the preceding type is a result of mutual genealogical links arising from the fact that Dressel 6A amphorae appeared in the same area, and in some cases even in the same workshops in which the Lamboglia 2 amphorae had been previously produced. The centre of their production during the final third of the 1st century BC and first half of the 1st century AD were Italic Venetia and Emilia-Romagna (Carre et al. 2014: 422-423; Cipriano and Mazzocchin 2017: 33-47, esp. 41, fig. 3), and they met the demand for export of the widely renowned regional wines (Tchernia 1986). Interest in them was recorded in the eastern Adriatic as well, where amphorae of this type are, besides undersea contexts (Radić Rossi 1993: 121128; Jurišić 2000: 12), increasingly being discovered in residential sites of various types (Šimić-Kanaet 2010: 44; Topić 2004: 386, cat. no. 392-394; ZabehlickyScheffeneger 1979: 31-32, taf. 12/5).

One more product of Crikvenica workshop is the sherd of amphora that belongs to the Crikvenica 3 type (Plate II, 6). This is the upper half of a small amphora with belly-shaped/cylindrical body, a broad cylindrical neck and a high rim articulated into multiple steps below which two ribbon-like handles vertically rest on the sharply reclining shoulders. The product carried by these types of amphorae has not been determined with any certainty, but the prevailing view seems to be that they transported processed fish products or olive oil. Also noteworthy is that it is one of the three most common types among the amphora finds at the actual site of the Crikvenica workshop, which should also be evident in their general but still poorly known distribution.

The sherd of the upper section of an amphora with massive ring-shaped rim, short conical neck and two short, arched handles set on the shoulders (Plate II, 2) most likely dates to roughly the same time as the transition of eras or the very beginning of the 1st century AD. Based on its morphological features, it most likely belongs to the group of so-called ovoid oil amphorae such as those of Brindisi type (cf. Bezeczky 1998: 233, fig. 4/1) or even Oberaden 83 type (González Cesteros 2016). Some of the best known olive oil vessels of the Roman era would regionally emerge from them, e.g., Dressel 6B (Carre and Pesavento Mattioli 2003: 453-476), and Dressel 20 in Hispania (Berni and Vargas 2016), but we can see that even before they appeared olive oil imports in the eastern Adriatic were vigorous.

Little – much of it the same – may be said about the sherd of a Crikvenica 5 type amphora (Plate III, 1) quite similar to the preceding one. In this case it is the lower half of a small oblong egg-shaped amphora with a prominent foot having the typical mushroom-shaped end also visible on the Crikvenica type 2 amphora (a local variant of Dressel 2-4). Based on known examples, type 5 amphorae had a wide cylindrical neck and a simplyformed ring-shaped rim below which the ribbon-like handles emerged slightly askew, resting vertically on the amphora’s shoulders. The purpose of this type is also still uncertain, and besides the Crikvenica workshop an example has also been recorded in Roman Senia (Lipovac Vrkljan – Ožanić Roguljić 2013: 265, fig. 6; Lipovac Vrkljan 2016: 96, cat. no. 39).7

A particularly valuable set of amphorae from Jadranovo consists of 5 sherds of those whose production site has been located in the Crikvenica workshop of Sextus Mu/etilius Maximus (Lipovac Vrkljan 2011: 3-18; on amphorae in Lipovac Vrkljan 2016: 53-58). Three sherds

We would like to extend our sincerest gratitude to our colleague Goranka Lipovac Vrkljan for unselfishly providing information on this Crikvenica amphora type. 7 

99

Igor Borzić, Mate Parica, Mato Ilkić, Tea Rosić

Plate II. Lokvišće, underwater finds of amphora fragments (photo: T. Rosić, author: I. Borzić).

To later centuries belong sherds of Tripolitana III (Keay XI) (Plate III, 2) and Cretoise AC1d (Plate III, 3) type amphorae. The former originated in the north

African workshops situated in the territory of today’s Libya (Panella 1973: 566; Keay 1984: 133-136; Bonifay 2004: 105), while during the production phase from the 100

An underwater archaeological survey of Havišće and Lokvišća Coves and Cape Ertak

Plate III. Lokvišće, Lokvišće, underwater finds of amphora fragments and bowl (photo: T. Rosić, author: I. Borzić).

mid-2nd to 4th centuries, they were used to transport Tripolitan olive oil. Amphorae of this type are not numerous on the eastern Adriatic; in fact, besides their

appearance in several shipwrecks (Duboka, Oruda, Kaprije, Piruzi, Sveti Petar A at Kupari and Veliki Piruzi), they have also been recorded on the seafloor of certain 101

Igor Borzić, Mate Parica, Mato Ilkić, Tea Rosić port facilities (Savudrija, Vrulje-Zadar and Žunac) (with the complete scholarly literature in: Pešić 2019: 195), to which that of Jadranovo can now be added.

the Adriatic during the late Republic and early Empire, and significant insight into regional mercantile events are provided by the discovery of three amphora types made in the nearby ceramic workshop in Crikvenica. This is certainly a very promising archaeological zone, and more systematic research into it would offer a wealth of data for a more detailed reconstruction of the passage of time and events in the otherwise exceptionally dynamic territory of Kvarner.

The other sherd of an amphora with narrow ringshaped rim and short conical neck from which two short arched handles extend and rest on the shoulders of an egg-shaped ribbed body probably belongs to the repertoire of the Cretan pottery workshops (MarangouLerat 1995: 5-34; Robinson 1959: 43). Specifically, this should be the D variant of Cretoise AC1 type amphorae (Marangou-Lerat 1995: 73, fig. 44), thus a late issue of this typical series, whose production in variant A already began in the latter half of the 1st century, and ended precisely with variant D, most often associated with the 3rd and early 4th centuries based on archaeological contexts (Marangou-Lerat 1995: 74-75). What makes the example from Jadranovo particularly important is the fact that this amphora variant can rarely be found outside of Crete as its core production area (Robinson 1959: 68, K112; Majcherek 2007: 11-13 and 27, fig. 2/78). On the other hand, from the standpoint of trade ties between Crete and the Adriatic market, as an isolated find it confirms that period’s divergence from the early imperial era, when the quantity of Cretan products here was nonetheless somewhat greater (cf. Radić and Jurišić 1993: 114-138; Auriemma et al. 2015: 139-160; Taras 2015: 262).

References Antonioli, F., M. Anzidei, K. Lambeck, R. Auriemma, D. Gaddi, S. Furlani, P. Orrù, E. Solinas, A. Gaspari, S. Karinja, V. Kovačić and L. Surace 2007. Sea level change during Holocene from Sardinia and northeastern Adriatic (Central Mediterranean Sea) from archaeological and geomorphological data. Quaternary Science Reviews 26: 2463‒2486. Auriemma, R. and E. Solinas 2009. Archaeological remains as sea level change markers: A review. Quaternary International 206/1-2: 134‒146. Auriemma, R., V. Degrassi and E. Quiri 2015. Eastern Amphora imports in the Adriatic Sea: Evidence from terrestrial and underwater contexts of the Roman Imperial Age, in S. Demesticha (ed.) Per Terram, Per Mare: Seaborne Trade And The Distribution Of Roman Amphorae In The Mediterranean: 139‒160. Uppsala: Åströms Förlag. Benac, Č. and T. Šegota 1990. Potopljena ris-virmska abrazijska terasa u podmorju ispred Rijeke. Geološki vjesnik 43: 43‒52. Berni, P. and E. García Vargas 2016. Dressel 20 (Guadalquivir Valley), in Amphorae ex Hispania. Landscapes of production and consumption, viewed 21. February 2020, . Bezeczky, T. 1998. Amphora types of Magdalensberg. Arheološki vestnik 49: 225‒242. Bilić Dujmušić, S. 2014. Bitka kod Krka godine 49. pr. Kr., in G. Lipovac Vrkljan, B. Šiljeg, I. Ožanić Roguljić and A. Konestra (eds) Roman Ceramic and Glass manufactures. Production and trade in the Adriatic region, Proceedings of the 2nd International Archaeological Colloquium (2829 October 2011): 107‒124. Crikvenica: Institut za arheologiju and Muzej Grada Crikvenice. Blackman, D., B. Rankov, J. Pakkanen, H. Gerding, K. Baika and J. McKenzie 2013. Shipsheds of the Ancient Mediterranean. Cambridge: Cambridge University Press. Bonifay, M. 2004. Études sur la céramique romaine tardive d’Afrique (British Archaeological Reports International Series 1301). Oxford: Archaeopress. Borzić, I., M. Čelhar, G. Zaro and V. Glavaš 2018. New finds of Crikvenica 1-type amphorae in the area of Southern Liburnia, in G. Lipovac Vrkljan and A. Konestra (eds) Pottery Production, Landscape

Besides amphora from Antiquity, an Early Modern-era bowl (Plate III, 4a, 4b) was also found which probably came from the Argenta workshop in Emilia Romagna from the end of the 16th or 17th century (Gelichi 1992: 21, fig. 11). The closest analogy can be found at the Mala Jana site on the island of Krk, which contains Early Modern finds of ship cargo (Ujčić Grudenić 2017: 181200). Conclusion The presented archaeological repertoire gathered during underwater reconnaissance of the seafloor of Havišće and Lokvišća Coves and Cape Ertak, together with the known terrestrial archaeological topography, suggest that this area has been in continual use. The reasons for this should be sought primarily in its strategic advantages. Unfortunately, the ambiguous archaeological context of the amphorae and other pottery analysed herein does not allow for their use as archaeological evidence to the fullest extent. Nevertheless, just their presence, as well as the presence of ascertained underwater structures, suggest that the aforementioned advantages were based primarily on the cove’s sound maritime features on the highlyfrequented Kvarner navigation route down the eastern Adriatic coast. The chrono-typological spectrum of finds is generally standard in the maritime picture of 102

An underwater archaeological survey of Havišće and Lokvišća Coves and Cape Ertak

and Economy of Roman Dalmatia. Interdisciplinary approaches (Roman Archaeology 47): 50‒56. Oxford: Archeopress. Brunšmid, J. 1900. Arheološke bilješke iz Dalmacije i Panonije IV. Vjesnik Arheološkog muzeja u Zagrebu 5/1: 118‒120. Cambi, N. 1991. Amfore kasnorepublikanskog doba i njihova produkcija u Dalmaciji. Akademija nauka i umjetnosti BiH, Posebna izdanja knj. XCV, Odjeljenje društvenih nauka knj. 27: 55‒65. Carre M. - B. and S. Pesavento Mattioli 2003. Tentativo di classificazione delle anfore olearie adriatiche. Aqvileia Nostra LXXIV: 453‒476. Carre, M. - B., P. Monsieur and S. Pesavento Mattioli 2014. Transport amphora Lamboglia 2 and Dressel 6A: Italy and/or Dalmatia? Some clarification. Journal of Roman Archaeology 27: 417‒428. Cipriano, S. and S. Mazzocchin 2017. Western Adriatic amphorae productions: the research status, in G. Lipovac Vrkljan, I. Radić Rossi and A. Konestra (eds) Adriamphorae, Amphorae as a resource for the reconstruction of economic development in the Adriatic region in Antiquity: local production. Proceedings of the workshop Zagreb, 21st April 2016: 33‒47. Zagreb: Institut za arheologiju u Zagrebu. Čargo, B. and M. Miše 2010. Lončarska proizvodnja u Isi. Vjesnik za arheologiju i historiju dalmatinsku: 7‒40. Faivre, S., E. Fouache, V. Kovačić and S. Gluščević 2010. Geomorphological and archaeological indicators of Croatian shoreline evolution over the last two thousand years. GeoActa – Special Publication 3, Geology of the Adriatic Area: 125‒133. Gelichi, S. 1992. Argenta e la ceramica nel XVII secolo in Emilia Romagna, in S. Gelichi (ed.) La produzione ceramica in Argenta nel XVII secolo: 11‒24. Firenze: All’Insegna del Giglio. Glavaš, V., A. Konestra and A. Tonc 2020. Wine Consumption in the Kvarner and the Velebit Area (NE Adriatic) in the Last Centuries BCE: Evidence from Amphora Finds, in I. Kamenjarin and M. Ugarković (eds) Exploring the Neighborhood. The Role of Ceramics in Understanding Place in the Hellenistic World (IARPotHP V): 271‒286. Wien: Phoibos Verlag. Gluščević, S. 2008. Zadar-Kolovare. Hrvatski arheološki godišnjak 4/2017: 411‒412. González Cesteros, H., E. García Vargas and R. R. de Almeida 2016. Oberaden 83 (Guadalquivir Valley), in Amphorae ex Hispania. Landscapes of production and consumption, viewed 21 February 2020, . Ilkić, M. and L. Šešelj in press. Broken pre-Imperial coins from the region of Liburnia and Iapodia (Croatia), in Monete frazionate. Quadri regionali, questioni cronologiche, aspetti economici. International conference, Milano, 16-17 settembre 2019, Università Cattolica del Sacro Cuore, in press.

Jovanović, A. 1984. Rimske nekropole na teritoriju Jugoslavije. Beograd: Univerzitet u Beogradu. Jurišić, M. 2000. Ancient Shipwrecks of the Adriatic, Maritime transport during the first and second centuries AD (BAR International Series 828). Oxford: Archaeopress. Jurišić, M. and I. Radić Rossi 1993. Das Antiken Schiffwrack von Mljet, Kroatien. Germania 71: 114‒138. Keay, S. J. 1984. Late Roman amphorae in the Western Mediterranean. A typology and economic study: the Catalan evidence (British Archaeological Reports International Series 196). Oxford: Archaeopress. Kirigin, B. 1994. Grčko-italske amfore na Jadranu. Arheološki vestnik 45: 15‒24. Kirigin, B. 1996. Issa: Grčki grad na Jadranu. Zagreb: Matica Hrvatska. Kirigin, B., T. Katunarić and L. Šešelj 2005. Amfore i fina keramika (od 4. do 1. st. pr. Kr.) iz srednje Dalmacije: preliminarni ekonomski i socijalni pokazatelji. Vjesnik za arheologiju i povijest dalmatinsku 98(1): 7‒21. Lambeck, K., F. Antonioli, A. Purcella and S. Silenzi 2004. Sea-level change along the Italian coast for the past 10,000 yr. Quaternary Science Reviews 23: 1567–1598. Lipovac Vrkljan, G. 2011. Lokalna keramičarska radionica Seksta Metilija Maksima u Crikvenici – Crikveničke amfore ravnog dna, in G. Lipovac Vrkljan, I. Radić Rossi and B. Šiljeg (eds) Roman Ceramic and Glass manufactures. Production and trade in the Adriatic region, Proceedings of the 1st International Archaeological Colloquium (23-24 October 2008): 3-18. Crikvenica: Institut za arheologiju and Muzej Grada Crikvenice. Lipovac Vrkljan, G. 2016. Proizvodi keramičarske radionice: Amfore, in A. Konestra and T. Rosić (eds) 845°. Ad Turres (exhibition catalogue): 53‒58, 91‒96. Crikvenica: Muzej Grada Crikvenice. Lipovac Vrkljan, G. 2017. Tipologija lokalne proizvodnje amfora crikveničke radionice, in G. Lipovac Vrkljan, B. Šiljeg, I. Ožanić Roguljić and A. Konestra (eds) Roman Pottery and Glass manufactures. Production and trade in the Adriatic region, Proceedings of the 3rd International Archaeological Colloquium (4-5 November 2014, Crikvenica): 185‒196. Crikvenica: Institut za arheologiju, Muzej Grada Crikvenice. Lipovac Vrkljan, G. and I. Ožanić Roguljić 2013. Distribucija crikveničke keramike kao prilog poznavanju rimskog gospodarstva. Senjski zbornik: prilozi za geografiju, etnologiju, gospodarstvo, povijest i kulturu 40: 255‒270. Lipovac Vrkljan, G. and A. Konestra 2018. Approaching the Roman economy of Province Dalmatia through pottery production – the Liburnia case study, in G. Lipovac Vrkljan and A. Konestra (eds) Pottery Production, Landscape and Economy of Roman Dalmatia. Interdisciplinary approaches (Roman Archaeology 47): 14‒36. Oxford: Archeopress. Maggi, P. 2018. Crikvenica amphorae from Canale Anfora: first evidence of Liburnian wine at Aquileia, in G. Lipovac Vrkljan and A. Konestra (eds) Pottery 103

Igor Borzić, Mate Parica, Mato Ilkić, Tea Rosić Production, Landscape and Economy of Roman Dalmatia. Interdisciplinary approaches (Roman Archaeology 47): 57‒61. Oxford: Archeopress. Majcherek, G. 2007. Aegean and Asia Minor Amphorae from Marina El-Alamein, in S. Marchand and A. Maragound (eds) Amphores d’Egypt de la BAsse epoque a l’epoque arabe. Cahiers De La Ceramique Egyptienne 8: 9-32. Cairo: Institut Francaise d’archaologie Orientale. Marangou-Lerat, A. 1995. Le vin et les amphores de Créte: de l’époque classique à l’époque imperiale (Études Crétoises 30). Athens/Thessalonique/Paris: Ecole française d’Athènes, Fondation Fany Boutari, De Boccard. Mustaček, M., M. Ćurković and I. Martinović 2013. The conservation and Restoration of Three Amphora from Pakoštane, Janice and Rovinj – Veštar. Submerged Heritage 3: 63 – 68. Olcese, G. 2004. Anfore greco-italice antice: alcune osservazioni sull’origine e sulla circolazione alla luce di recenti ricerche archeologiche e archeometriche, in E. C. De Sena and H. Dessales (eds) Metodi e approcci archeologici, l’industria e il commercio nell’Italia antica (BAR International Series 1262): 173‒192. Oxford: Archaeopress. Ožanić Roguljić, I. 2017. Amphoras from Žuta Lokva, in G. Lipovac Vrkljan, I. Radić Rossi and A. Konestra (eds) Adriamphorae, Amphorae as a resource for the reconstruction of economic development in the Adriatic region in Antiquity: local production. Proceedings of the workshop Zagreb, 21st April 2016: 80‒85. Zagreb: Institut za arheologiju u Zagrebu. Panella, C. 1973. Appunti su un gruppo di anfore della prima, media e tarda età Imperiale, in Ostia III: Le terme del Nuotatore: scavo dell’ambiente V et di un saggio dell’area (Studi miscellenei 21): 460‒633. Roma: De Luca. Parica, M., 2016. Kasnoantičko pristanište u uvali Pocukmarak na otoku Silbi / Late Antique quay in Pocukmarak bay on the island of the Silba. Archaeologia Adriatica 9: 307‒322. Pešić, M. 2012. Survey of the seabed of Zadar county. Submerged heritage 2: 44‒47. Pešić, M. 2019. Afrički import iz hrvatskog podmorja od I. do V. stoljeća, unpublished PhD dissertation. Zadar: Sveučilište u Zadru. Radić Rossi, I. 1993. Amfore tipa Lamboglia 2 i Dressel 6 A na istočnoj obali Jadrana u svjetlu podmorskih nalaza, unpublished MA dissertation. Zagreb: Sveučilište u Zagrebu. Radić Rossi, I. 2011. Brodski tereti krovne opeke i proizvodi radionice Seksta Metilija Maksima u jadranskom podmorju, in G. Lipovac Vrkljan, I. Radić Rossi and B. Šiljeg (eds) Roman Ceramic and Glass manufactures. Production and trade in the Adriatic region, Proceedings of the 1st International Archaeological Colloquium (23-24 October 2008): 19‒30.

Crikvenica: Institut za arheologiju and Muzej Grada Crikvenice. Robinson, H. S. 1959. Pottery of the Roman period: Chronology (The Athenian Agora 5). Princeton: The American School of Classical Studies at Athens. Romanović, D. 2017. Amfore ravnog dna tipa Crikvenica iz slojeva antičke luke u Zatonu kod Zadra, in G. Lipovac Vrkljan, B. Šiljeg, I. Ožanić Roguljić and A. Konestra (eds) Roman Pottery and Glass manufactures. Production and trade in the Adriatic region, Proceedings of the 3rd International Archaeological Colloquium (4-5 November 2014, Crikvenica): 387‒406. Crikvenica: Institut za arheologiju and Muzej Grada Crikvenice. SNG Cop — Sylloge nummorum graecorum 1969. The Royal collection of Coins and Medals Danish National Museum. North Africa, Syrtica – Mauretania, G. K. Jenkins (ed.). Copenhagen: Munksgaard. SNG Ita = SNG Italia, XIV 1989. Sylloge nummorum graecorum, Italia. Milano, Civiche Raccolte Numismatiche, Vol. XIV, Cyrenaica-Mauretania, R. Martini (ed.). Milano: Comune di Milano, Settore Cultura, Civiche Raccolte Numismatiche. Starac, R. 2002. Arheološka baština Jadranova. Vinodolski zbornik 8: 189‒206. Šešelj, L., and M. Ilkić 2014. Money circulation in Liburnia in the pre-imperial period: preliminary report, in M. Alram, H. Emmerig and R. Harreither (eds) Akten des 5. Österreichischen Numismatikertages, Enns, 21.-22. Juni 2012: 43‒53. Enns–Linz: Gesellschaftfür Landeskunde und Denkmalpflege Oberösterreich. Šimić-Kanaet, Z. 2010. Tilurium II, Keramika 1997. – 2006. Zagreb: Golden marketing Tehnička knjiga. Šuta, I. 2017. Amfore kao pokazatelj ekonomskih i društevnih odnosa u okolici Kaštelanskog zaljevau 2. i 1. st. pr. Krista, in G. Lipovac Vrkljan, I. Radić Rossi and A. Konestra (eds) Adriamphorae, Amphorae as a resource for the reconstruction of economic development in the Adriatic region in Antiquity: local production. Proceedings of the workshop Zagreb, 21st April 2016: 26‒32. Zagreb: Institut za arheologiju u Zagrebu. Taras, D. 2015. Amphorae from the roman port of Aenona, in S. Demesticha (ed.) Per Terram, Per Mare: Seaborne Trade And The Distribution Of Roman Amphorae In The Mediterranean: 257‒266. Uppsala: Åströms Förlag. Tchernia, A. 1986. Le vin de l’Italie Romaine. Essai d’histoire economiqued les amphores. Roma: École française de Rome. Topić, M. 2004. Posude za svakodnevnu uporabu grublje izradbe, amfore,terakote i kultne posude iz Augusteuma Narone. Vjesnik za arheologiju i historiju dalmatinsku 96: 303‒515. Vandermersch, C. 1994. Vins et amphores de Grande Grece et de Sicile, IVe-IIIe s. avant J.-C. Naples: Centre Jean Berard. Ujčić Grudenić, J. 2014. Jadranovo-Lokvišće. Hrvatski arheološki godišnjak 11/2014: 401‒ 402.

104

An underwater archaeological survey of Havišće and Lokvišća Coves and Cape Ertak

Ujčić Grudenić, J. 2017. Keramičko posuđe s hidroarheološkog lokaliteta Mala Jana u blizini Glavotoka na otoku Krku. Miscellanea Hadriatica et Mediterranea 3: 181‒200. Will, E. L. 1982. Greco-Italic Amphoras. Hesperia 51: 338‒356.

Zabehlicky-Scheffenegger, S. and M. Kandler 1979. Burnum I. Erster Bericht über die Kleinfunde der Grabungen 1973 und 1974 auf dem Forum. Wien: Vlg. d. Österr. Akademie d. Wissenschaften.

105

The Roman shipwreck of Sv. Petar, Ilovik: old research, new possibilities Irena Radić Rossi, Bridget Buxton Abstract Today the channel that runs between the islet of Sveti Petar and the island of Ilovik in the northeastern Adriatic provides a popular anchorage for recreational sailors, offering shelter from all but the strongest Bora and Sirocco winds. The strait, which is 2.5 km long and less than 300 m wide, was also a refuge sought by ancient mariners, but not all made it safely to cover. In 1962, the well-preserved cargo of a sunken Roman ship was discovered at the northern entrance to the channel between Ilovik and Sveti Petar. By 1970, the surface of the site was already heavily looted, but rescue excavations in the early 1980s under the direction of Marijan Orlić nevertheless recovered many interesting finds. The ship’s cargo consisted of several types of amphorae, among which the most numerous were the Adriatic flat-bottomed containers usually referred to as the Forlimpopoli type, found throughout the Mediterranean in the 1st and 2nd centuries AD. In addition to the ceramics, a sestertius of the emperor Trajan pointed to an early 2nd century AD date for the shipwreck. Unusual luxurious finds, such as bronze implements and glassware, probably did not belong to the ship’s general cargo, and may point to the presence of wealthy or high-ranking passengers. At the time of the 1980s rescue excavations, however, there was little to distinguish this site from numerous other 2nd century wrecks of Roman merchantmen in the Adriatic. In 1999, the complete Hellenistic bronze statue of a Greek athlete (the celebrated ‘Croatian Apoxyomenos’ statue) was recovered from the sea off the southern cape of the island of Vele Orjule, which lies less than 3 miles northeast of the entrance to the Strait of Ilovik (Ilovačka vrata). Since no shipwreck was found near the apoxyomenos statue, Marijan Orlić raised the possibility of connecting the two sites. This hypothesis warrants investigation, as it potentially justifies further excavation of the Sv. Petar Roman shipwreck.

Keywords Shipwreck, Sveti Petar, Ilovik, amphorae, urceus, patera, apoxyomenos, Licinius Sura, Trajan.

Introduction In the early 1960s, sponge divers and coral hunters discovered an unlooted Roman shipwreck in the deep channel between the islands of Sveti Petar, Ilovik, and Lošinj (Figure 1). This information soon reached the Centre for Underwater Activities in Kostrena near Rijeka (Orlić 1986: 7).1 In 1962, divers from the Centre undertook the first campaign to document the several hundred amphorae visible at the site, but the information did not reach the appropriate institutions until sometime later. During this first investigation, Željko Korenčić took photos that show a well-preserved site (Figure 2). A new administrative division established in 1964 brought the wreck under the jurisdiction of Pula. In 1965, a team of Slovenian divers operating under the auspices of the official archaeologist Štefan Mlakar raised 17 amphorae from the site. In 1970, Radmila Matejčić from the Historical and Maritime Museum of the Croatian Littoral (Cro. Povijesni i pomorski muzej Hrvatskog Primorja) undertook another documentation campaign, raising a further seven amphorae and depositing them at Mali Lošinj. In her report, Matejčić described the surface of the Sv. Petar

wreck site as heavily looted (Figure 3), with traces of digging activity in the sandy seabed. Hoping to return the following year to continue the investigation, she left a provisional grid on the site. Unfortunately, Matejčić was unable to continue her work, and the amphorae she raised from the Sv. Petar wreck were subsequently lost. In 1978 a field survey by the Regional Institute for the Protection of Cultural Monuments (Cro. Regionalni zavod za zaštitu spomenika kulture) in Rijeka took place under the direction of Marijan Orlić. By now the wreck was heavily degraded through looting, but still showed some archaeological potential. In fact, it was the apprehension of several of the looters that helped secure the funding for two seasons of further rescue work at the site. This campaign led to the recovery of additional ship’s equipment and cargo. Wooden remains of the hull were also identified on the seabed, and their positions recorded in the site plan (Figure 4). Unfortunately, due to limited funding these remains were never excavated nor studied in detail. After 1980 the Sv. Petar wreck was not visited again until 2018, when a limited survey of the site organized by the University of Zadar in collaboration with the University

All the information in the introductory chapter is taken from Orlić 1986: 7-17. 1 

Roman Pottery and glass manufactures (Archaeopress 2022): 106–115

The Roman Shipwreck of Sv. Petar, Ilovik

Figure 1. Position of Sv. Petar, Ilovik shipwreck site, and the bronze statue found near Vele Orjule.

Figure 2. Entirely preserved site, found in 1962 (photo: Ž. Korenčić).

107

Figure 3. Looted site in 1970 (photo: Ž. Korenčić).

Irena Radić Rossi, Bridget Buxton

Figure 4. Site plan (drawing: D. Grigić).

of Rhode Island confirmed the shipwreck’s potential value for future excavation.2

The Sv. Petar shipwreck finds The inventory of finds recovered over the course of three field seasons consisted of ceramics, glass, and metal objects (Orlić 1986: 16-17). Besides the roof tiles, the rescue excavation yielded 111 well-preserved amphorae and 123 necks. Most of these were Adriatic flat-bottomed amphorae (Forlimpopoli type) in three variants (Orlić 1986: Fig. 12-14). Several amphorae of Dressel 2-4 type were also present in the ship’s cargo (Orlić 1986: Fig. 15), while one amphora of Dressel 43 type and one Cretan corn-handled amphora were likely part of the ship’s inventory (Orlić 1986: Fig. 16-17) (Figure 5). About ten amphorae lids were also recovered, some discoid (Orlić 1986: Fig. 18), and some in the form of a cup or aryballos (Orlić 1986: Fig. 19).

The Sv. Petar shipwreck site In 1970, the visible remains of the Sv. Petar shipwreck were spread across a 12 x 20 m area of sandy seafloor and Posidonia grass, at an average depth of 27 m. In 1978, the excavation zone was provisionally divided into 18 rectangles of 2 x 4 m and two rectangles of 3 x 4 m (Figure 4). After removing the fragmentary ceramics left by looters, the researchers excavated some intact amphorae from the central part of the site, but the bulk of preserved cargo lay outside their excavation grid. The presence of tegulae and imbrices (roof tiles), pottery, glass and metal ware in rectangles 5–10 indicated the stern part of the ship, where the ship’s galley would have been situated. The presence of the bilge pump collector in rectangle 11 marked the location of the central part of the hull, and the remains of an iron anchor in rectangle 20 likely came from the ship’s bow. The scanty hull remains consisted mainly of planks and frames, with some possible traces of stringers. Parts of a 20 cm wide beam uncovered in rectangles 14 and 16 were identified as the ship’s keel (Orlić 1986: 43). Based on this data, the Sv. Petar ship was a typical mediumsized Roman merchantman with an overall length of approximately 25 m (Orlić 1986: 43).

In the ship’s inventory there were also four ceramic oil lamps, one vessel of south-western Italian origin, one big plate identified as Pompeian Red Ware, a plate of Samian origin, and one cup and two jugs identified as Çandarli ware (Orlić 1986: Fig. 20-27). More unusual finds from the ship’s galley area included two square glass bottles (Figure 6), one bowl (Figure 7), and numerous glass fragments testifying to the presence of other glassware. Besides the nails, mast fittings, and the afore-mentioned bilge pump and iron anchor remains, the inventory of metal finds consisted of one sestertius of emperor Trajan (Orlić 1986: Fig. 35), and two bronze implements associated with ritual libations: a bronze patera with

The authors thank the 2018 expedition dive team of Vedran Dorušić, Jacob Sharvit, Steve Phelps and Kotaro Yamafune, and the support of Foka Dive Centre and Oceangate Foundation (Seattle, USA). 2 

108

The Roman Shipwreck of Sv. Petar, Ilovik

Figure 5. Types of amphorae recovered from the shipwreck site (drawing: V. Begović).

a long straight handle culminating in the head of a panther, and a bronze urceus (ritual pitcher) decorated with the head of Eros at the base of the handle (Figures 8-11). Finally, it appears that one lead anchor stock was recovered from the site by sponge divers, and sold to the Dominican monastery in Bol at the island of Brač (Vrsalović 2011 (1979): 152; Orlić 1986: 16). Based on the analysis of all these recovered objects, and especially the presence of Trajan’s coin (issued between 114116 AD), the shipwreck may be dated to the early 2nd century AD.3

Overall, the finds recovered in rescue excavations of the Sv. Petar shipwreck present nothing surprising compared to numerous other Roman wrecks of the period found in the Adriatic. The only unusual objects are the bronze urceus and patera, which – while ubiquitous in Roman art in contemporary depictions of state sacrifice (for example, on Trajan’s column) or allusions to the emperor’s priestly office – are not standard objects to find on a Roman shipwreck.4 In (when Trajan received his ‘optimus’ title shortly after setting out on his eastern campaigns) and 116 (when he took the title ‘Parthicus’). 4  The authors could locate no other published examples of a bronze urceus and patera excavated from an early imperial shipwreck, which at the very least suggests the objects were not standard equipment on a merchantman.

Orlić 1986: 44; see Beckmann 2000 and Woytek et al. 2007 on the dating and significance of this coin type, which belongs to the ‘Fortuna Redux’ (‘fortunate return’) series minted between 114 3 

109

Irena Radić Rossi, Bridget Buxton

Figure 8. Bronze patera (photo: R. Mosković).

Figure 6. Two square glass bottles (photo: R. Mosković).

Figure 9. Long straight handle of patera, culminating in the head of a panther (photo: R. Mosković).

Figure 7. Glass bowl (photo: R. Mosković).

fact, the urceus, patera, simpulum (ritual ladle), and lituus (priest’s wand), when they do appear in Roman imperial monuments and on coins, invariably signify the religious foundation of the emperor’s authority, a reminder that Rome’s military campaigns took place under his auspices.

the 2nd century AD, images of the urceus and patera on imperial monuments and coins were deployed exclusively to identify the emperor or the presumptive imperial heir (Boyce 1945: 45-57). A famous bronze medallion of the 2nd century AD emperor Caracalla illustrates this use of ritual implements as allusions to imperial power, both religious and military (Figure 12).5

The depiction of sacrificial implements to invoke the religious authority of Roman military leaders was well-established by the Late Republic (Stewart 1997: 170-189) and became a standard element of imperial iconography under Augustus. While in practice any senior commander in the provinces would have possessed sacred implements for making ritual offerings as part of the course of their normal duties, images of these objects invoked the ultimate source of their authority, the Pontifex Maximus himself. By

This bronze medallion of Caracalla (196 AD) comes from the collection of the American Numismatic Society (inv. 1945.32.3) and is provided open-access by ANS as a free public domain image on their website, which the authors gratefully credit. Obverse: M AVRELIVS ANTONINVS CAES – Bust draped cuirassed right / Reverse: SEVERI AVG PII FIL – Lituus, knife, patera, urceus, simpulum and aspergillum left to right on exergual line. The medallion was issued two years before Caracalla succeeded his father, the emperor Septimius Severus, and sent the message to the troops identifying his son as his successor by association with the symbols of the emperor’s power. 5 

110

The Roman Shipwreck of Sv. Petar, Ilovik

Figure 10. Bronze urceus from the shipwreck site of Sv. Petar, Ilovik (photo: R. Mosković).

Figure 11. Head of Eros at the base of the handle of the bronze urceus (photo: R. Mosković).

they were recovered from upper-class burial contexts (Apakidze and Nikolaishvili 1994: 19). Although the objects themselves were manufactured in the Roman West, the find spots are all regions on the periphery of the Empire. These are regions where Romanized communities and even Roman troops were present in the early 2nd century, especially during the campaigns of Trajan, who led Roman expeditions into Dacia in 101–102 and 105–106, and Armenia and Parthia in 113‒117AD. The bronze urceus and patera found amidst the gold and silver treasure from the 2nd-3rd century AD tomb of the ‘Roman noble’ (Tomb No. 905 at Samtavro; possibly a female burial) near Mtskheta, Georgia are nearly identical to the examples from the Sv. Petar shipwreck (Apakidze and Nikolaishvili 1994: fig. 7). What can be proposed with some confidence, then, is that these ritual objects of western origin were typically the possessions and heirlooms of wealthy and highranking families, most likely individuals whose position and privileges were connected to military service to Rome. Contemporary depictions of similar objects on Roman coins and monuments reinforced their symbolic association with the sacred responsibilities of a Roman emperor or military commander.

Figure 12. Reverse of the bronze medallion of Caracalla (196 AD); SEVERI AVG PII FIL – Lituus, knife, patera, urceus, simpulum and aspergillum left to right on exergual line (http:// numismatics.org/collection/1945.32.3?lang=en).

It was not only campaigning emperors, however, who might own a set of such fine bronze ritual implements. It appears the few comparable examples of the Sv. Petar bronze ritual objects come from the region of the Danubian frontier and Iberia along the Caucasus, where

In the context of the archaeological finds so far recovered from the Sv. Petar shipwreck, such an observation about the origin and importance of the 111

Irena Radić Rossi, Bridget Buxton

Figure 13. The bronze apoxyomenos statue discovered near the island of Vele Orjule (photo: D. Frka, courtesy of the Croatian Conservation Institute).

bronze urceus and patera at first does not appear to add much to our understanding of the site. If, however, Marijan Orlić is correct about the possible connection between the famous Vele Orjule apoxyomenos statue and the Sv. Petar shipwreck, the implications are highly significant. Before exploring them further, we need to revisit the discovery of the apoxyomenos statue and the objects found with it on the sea floor.

represents a young Greek athlete engaged in scraping himself down with a stlengis (Latin strigil) after working out or competing. Such images were a mainstay in the decoration of Greek gymnasia and the finest private homes, as well as festooning the important athletic sanctuaries of Olympia, Delphi, and elsewhere. The best Greek bronzes, or the best copies, were also highly sought-after by Roman elites for use in both private and public spaces, particularly bathing-gymnasia complexes (Gensheimer 2018: 85).

The Apoxyomenos discovered In 1999, information about the discovery of a bronze statue near Cape Glavičin on the island of Vele Orjule near Lošinj reached the Croatian Ministry of Culture (Orlić and Jurišić 1999). The statue lay at a depth of 45 m, with no traces of any shipwreck in the vicinity (Figure 13). It was raised with the help of multiple organizations in April of the same year (Sténuit et al. 2001: 198).6 After many years of conservation treatment (Domijan and Karniš 2006) and an exhibition that traveled around the world, today the statue is on display in the Museum of the Apoxyomenos in Mali Lošinj. The statue was cast in the late 2nd or 1st century BC and

From the start, the circumstances surrounding the statue’s presence on the sea floor next to a small Croatian island defied easy explanation. The apparent absence of a nearby shipwreck raised the possibility that it was deliberately cast into the sea for a ritual purpose. Nevertheless, the apoxyomenos was not found entirely alone. The 1999 summer research campaign revealed several objects nearby that appeared related — for example the statue’s bronze base, decorated with meander and swastika ornament, was found in the sand beneath it. Significantly, the base showed signs of being forcibly removed from its original location. There were also two Roman lead anchor stocks, a fragment of a Greco-italic amphora, some fragments of Lamboglia 2 amphorae, a flat-bottomed Adriatic amphora of the same Forlimpopoli type found on the Sv. Petar wreck, and a fragment of a North African amphora (Sténuit et al. 2001: 198-203).

The recovery of the statue was a joint effort of the Department for the Protection of Archaeological Heritage of the Ministry of Culture of the Republic of Croatia, Archaeological Museum in Zadar, Groupe de Recherche Archéologique Sous-Marine Post-Médiévale (GRASP) from Belgium, Oxford Maritime Trust and company for underwater works SUBMAR. 6 

112

The Roman Shipwreck of Sv. Petar, Ilovik

One anchor stock lay about 10 m southeast of the statue at a depth of 38 m. This anchor stock was 135 cm wide, and weighed around 205 kg, which points to a ship over 20 m long, and of at least 100 tons of load capacity (Sténuit et al. 2001: 199). One arm of this anchor stock was deformed, as if subjected to an enormously powerful force at the time it was lost, which seems to rule out the possibility that the anchor – or statue – were simply discarded to lighten the load.

attempted to privatize Rome’s favorite public artworks was faced with extreme censure. Pliny the Elder (Natural History 34.62) records the story of – ironically – a famous apoxyomenos statue sculpted by Lysippus that stood outside the thermae of Agrippa, which the Roman emperor Tiberius had removed to his private bedroom. Public outrage forced him to return it. There is nothing surprising, then, that a Roman merchantman the 2nd century AD should be carrying a larger-than-life Greek bronze athlete statue across the Adriatic. The fact that the statue’s base was damaged, and the statue itself contained remains of a mouse nest, indicates that this was no privately commissioned artwork, but a spoil of conquest torn from its original context in Greece or the Hellenized provinces by force or at least strong compulsion. While the notorious provincial governor Verres and the philhellenic artlover Nero provide Roman history’s best-known examples of Romans helping themselves to local statues, the fact is the practice was continuous and ubiquitous among the Italian elites over many centuries.

In the absence of a nearby shipwreck, a hypothetical scenario may be considered. The ship carrying the apoxyomenos back to Italy was being driven out of control before the Bora (north-northeast) wind, and the anchor was thrown in a last-ditch effort to secure it behind the meager shelter of Vele Orjule, but it did not hold (Sténuit et al. 2001: 200). Alternatively, the desperate sailors may have wrapped the anchor chain around the statue to give it more weight and thrown both over together. In such case, the weight of the statue could have caused the deformation of the anchor stock, when the latter hit the seabed in a place where a gap between the rocks was smaller than the span of the stock. The second smaller anchor stock, 107 cm wide, lay about 100 m southeast of the statue, at a depth of 42 m. Southwest of this stock there was a nearly complete flat-bottomed amphora of the same Forlimpopoli type as the cargo of the Sv. Petar shipwreck (Sténuit et al. 2001: 201). These objects help us to construct a picture of how the statue and the items accompanying it may have ended up so far distant from the site of the Sv. Petar shipwreck.

A more surprising thing, indeed, would have been a ship engaged in transporting looted or ‘purchased’ provincial artwork back to Rome that carried only one such priceless statue. Add the presence of sacrificial implements traditionally associated with Roman emperors and their top commanders on military campaign, and the possibility that the Sv. Petar ship was carrying spoils of war – or at least the rich fruits of culturally acquisitive imperialism – seems even more likely. This is the hypothetical scenario that encourages us to propose further investigation of an otherwise unremarkable shipwreck. Could there be other treasures of Greek art buried under the deep sand and posidonia beds at the Sv. Petar wreck site that have not yet been discovered? Could the excavation of the shipwreck yield further clues about its possible connection to the Vele Orjule apoxyomenos statue? These are the foremost questions that justify continued exploration and ideally the systematic excavation of the Sv. Petar Roman shipwreck.

A final anecdote reveals that the scenario of a ship in distress losing anchors and statue at Vele Orjule but not sinking itself until reaching Sv. Petar is not at all farfetched. During the 1999 summer fieldwork, a dangerous north-eastern wind arose (the Bora, or Croatian bura). In the brief time it took the archaeological team to wind up operations, the metal decompression frame and its buoy were physically ripped from the transom of the dive boat and carried out to sea by the powerful winds. However, the researchers eventually found it again: it had been carried several kms southwest to the entrance to the Strait of Ilovik, almost right above the Sv. Petar shipwreck. That was the first time that the Sv. Petar shipwreck research project director Marijan Orlić made the connection between the Greek statue and the Roman shipwreck.

A final possibility deserves mention, if only as pure speculation. While the Trajanic sestertius only provides a terminus post quem for the Sv. Petar shipwreck of 114 AD, the military campaigns of Trajan in Dacia and Parthia between 101-117 AD offer a logical context for the bulk transfer of appropriated provincial artworks across the Adriatic, and an explanation for ritual implements of western provincial origin finding their way across the Empire even to the northern frontiers of Armenia. Scholars have long associated Trajan’s expansionist wars with a circle of powerful advisors from his native province of Hispania, in particular his foremost friend and general Lucius Licinius Sura, who held three consulships and won triumphal honors for

Discussion Since the discovery of the Antikythera shipwreck more than a century ago, it has been a truism of classical archaeology that the best place to find Classical (or Hellenistic) Greek bronzes is on a Roman shipwreck. The Romans appreciated Greek artwork both for their private villas and public spaces such as libraries and baths (thermae and gymnasia), and any emperor who 113

Irena Radić Rossi, Bridget Buxton his role in the emperor’s Dacian campaigns (Dio Cassius 68.9.2; Jones 1970: 98-104).

summer 1999, inadvertently served as an experiment to show what happens to an object driven by the Bora wind from the position where the apoxyomenos statue was found. If we think about our hypothetical ancient ship desperately seeking safe anchorage and jettisoning everything it could to survive while being driven along before the Bora wind, all the different pieces of the puzzle fit together with the available evidence.

Possible depictions of Sura accompanying the emperor on Trajan’s column show a youthful clean-shaven individual.7 We know that Sura’s third consulship was in 107, after which he vanishes from history; Syme suggested that Sura may have died in an epidemic that swept through Italy in 108 (Syme 1980: 22). There is no evidence to support Syme’s theory other than the lack of any securely-dated mentions of Sura after 107, which might suggest that his premature demise was due to a mundane and unremarkable cause. If Sura’s third consulship was followed by a resumption of administrative duties in the provinces, he could very well have died outside Italy during Trajan’s eastern campaigns (113-117 AD). All we know for certain is that Licinius Sura received a public funeral and a statue from Trajan; no tomb is mentioned (Dio Cassius 68.15). The window of Sura’s death must fall between 108 and 117 AD.

A loss of maneuvering capability could have influenced the decision to try to stop the ship at the entrance of Orjule Channel by throwing the anchor, and possibly the statue tied to it, as it passed the southern cape of Vele Orjule. Another option would be jettisoning the statue for some other reason like sacrificing it to the sea gods, removing the possible cause of a curse, or simply making the ship lighter (Cambi 2007: 105). In any case, the statue would have sunk to the seabed during the critical last attempts to save the endangered vessel. Another attempt could have happened immediately after, by throwing the smaller anchor, accompanied by the loss of one or more amphoras from the ship’s cargo. Then when these attempts failed, the ship would have been helplessly carried to its final position next to Sv. Petar, along exactly the same route as the runaway decompression stop and buoy in 1999.

Known for his spectacular wealth and cultured personality, Licinius Sura’s influence was intellectual as well as political. He was believed to be Trajan’s speechwriter (Julian, Caes. 327 B; HA Hadr. 3). Pliny the Elder corresponded with Sura on scientific matters (Pliny Ep. 4.30; 7.2), and his fellow Spaniard, the poet Martial, referenced Sura’s wealth and good taste in several epigrams (1.49, 6.64). Most importantly for our purposes, the major legacy of Licinius Sura to Trajan’s Rome was the construction of a luxurious bath and gymnasium complex in the fashionable suburbs of the Aventine, a structure probably completed in his honor by Trajan (Jones 1970: 100). Like Trajan’s forum, the structure was probably funded ex manubiae from the spoils of the Dacian campaigns in which Sura played a leading role. It goes without saying that such a building would have been ornamented with dozens if not hundreds of statues, and the source of many of these statues would have been the Hellenized cities of the Greek East. Trajan’s continuous wars provided the perfect opportunity for a man of Sura’s authority and taste to acquire shiploads of prestigious Greek artworks for the ornamentation of his Aventine thermae. Perhaps, in the Sv. Petar shipwreck and Vele Orjule apoxyomenos, we have archaeological evidence of this campaign of artistic appropriation. And finally, in the ritual urceus and patera, evidence of the presence of a high-ranking Roman commander overseeing the shipment. Not impossibly, a clue to the final fate of Licinius Sura himself.

The Bora wind would have driven our ship through the Orjule Channel, between the islands of Mali Lošinj and Kozjak, past the island of Sv. Petar to the entrance to the Strait of Ilovik (Figure 1). Then, in order to turn into the shelter of the strait, the ship would have needed to turn beam-on to the deadly winds. Whether it attempted to do so or not, we know it was doomed. The presence in the ship’s inventory of precious objects such as the distinctive bronze patera and urceus, items invoking both wealth and military rank, as well as glass bottles, bowls, and other valuable objects that were probably used in the galley of the ship all indicate that a valuable bronze Greek statue would not have been out of place on board. Certainly, there appears to be no better candidate for the ship that carried the apoxyomenos. The question is, what else was she carrying? Romans willing to rip Greek statues from their bases, mouse nest and all, were unlikely to limit their appropriations to a single bronze apoxyomenos. In conclusion, although the site of the Sv. Petar Roman shipwreck was heavily looted in the past, and rescue excavations took much of what was left on or near the surface, there is still a good possibility that systematic excavation could help resolve many questions about its final voyage and possible relationship to the famous Croatian apoxyomenos statue.

Conclusion The strange incident with the decompression stand, lost at Vele Orjule and found again at Sv. Petar in the The identification is popular but not uncontested; see Jones 1970: 100. 7 

114

The Roman Shipwreck of Sv. Petar, Ilovik

Matejčić, R. and V. Dautova Ruševljanin 1972. Hidroarheološka istraživanja riječkog pomorskog i povijesnog muzeja na akvatoriju Kvarnera. Pomorstvo 7-8: 295‒298. Matejčić, R. and M. Orlić 1982. Rezultati prve faze hidroarheoloških istraživanja u cresko-lošinjskim vodama, in Ž. Rapanić (ed.) Arheološka istraživanja na otocima Cresu i Lošinju, (Izdanja Hrvatskog arheološkog društva 7): 161‒169. Zagreb: Hrvatsko arheološko društvo. Orlić, M. 1982. Podmorsko arheološko nalazište Ilovik, in Ž. Rapanić (ed.) Arheološka istraživanja na otocima Cresu i Lošinju, (Izdanja Hrvatskog arheološkog društva 7): 153‒159. Zagreb: Hrvatsko arheološko društvo. Orlić, M. 1986. Antički brod kod otoka Ilovika (Mala biblioteka Godišnjaka za zaštitu spomenika kulture Hrvatske 10-11). Zagreb: Republički zavod za zaštitu spomenika kulture. Orlić, M. and M. Jurišić 1999. Projekt ‘Statua-brodolom’. Obavijesti Hrvatskog arheološkog društva 31/2: 59‒65. Sténuit, M- E., R. Sténuit, M. Orlić and S. Gluščević 2001. A preliminary report on the discovery and recovery of a bronze Apoxyomenos, off Vele Orjule, Croatia. The International Journal of Nautical Archaeology 30.2: 196‒210. Stewart, R. 1997. The jug and lituus on Roman Republican coin types: ritual symbols and political power. The Phoenix 51.2: 170‒189. Syme, R. 1980. Some Arval Brethren. Oxford: Claredon Press. Vrsalović, D. 1974. Istraživanje i zaštita podmorskih arheoloških spomenika u SR Hrvatskoj. Zagreb: Republički zavod za zaštitu spomenika kulture. Vrsalović, D. 2011 (1979). Arheološka istraživanja u podmorju istočnog Jadrana; Prilog poznavanju trgovačkih plovnih putova i gospodarskih prilika na Jadranu u antici, phd dissertation. Zagreb/Split: Sveučilište u Zagrebu, Književni krug. Woytek, B. E., K. Uhlir, M. Alram, M. Schreiner and M. Griesser 2007. The Denarius under Trajan: New Metallurgical Analyses. The Numismatic Chronicle (1966-) 167: 147‒63.

References Apakidze, A. and V. Nikolaishvili 1994. An aristocratic tomb of the Roman period from Mtskheta, Georgia. The Antiquaries Journal 74: 16‒54. Beckmann, M. 2000. The early gold coinage of Trajan’s sixth consulship. American Journal of Numismatics (1989-) 12: 119‒156. Boyce, A. A. 1945. A Medallion of Caracalla. Museum Notes (American Numismatic Society) 1: 45‒57. Cambi, N. 2007. Brončani kip čistača strigila iz mora kod otočića Vele Orjule blizu Lošinja. Archaeologia Adriatica 1: 85‒109. Domijan, M. and I. Karniš (eds) 2006. Hrvatski Apoksiomen /The Croatian Apoxyomenos, (katalog izložbe/ catalogue of the exhibition). Zagreb: Ministarstvo kulture RH. Gensheimer, M. B. 2018. Decoration and Display in Rome’s Imperial Thermae: Messages of Power and Their Popular Reception at the Baths of Caracalla. Oxford: Oxford University Press. Jones, C. P. 1970. Sura and Senecio. The Journal of Roman Studies 60.1: 98‒104. Jurišić, M. 2000. Ancient Shipwrecks of the Adriatic, maritime transport during the 1st and 2nd centuries AD (BAR International Series 828). Oxford: Archaeopress. Jurišić, M. 2006. The maritime trade of the Roman province, in D. Davison, V. Gaffney and E. Marin (eds) Dalmatia: Research in the Roman Province 19702001 (BAR International Series 1576) : 175‒193. Oxford: Archaeopress. Matejčić, R. 1969. Podvodna arheološka istraživanja na području sjevernog Jadrana, in Podmorske djelatnosti s gledišta medicinskih i društvenih nauka (Mornarički glasnik, Pomorska biblioteka 22): 236‒242. Beograd. Matejčić, R. 1976. Petnaest godina hidroarheoloških istraživanja u Kvarneru. Pomorski zbornik 14: 345‒361. Matejčić, R. and V. Dautova Ruševljanin 1971. Dosadašnji načini podvodnih rekognosciranja na području Kvarnera, in Pitanja zaštite hidroarheoloških spomenika kulture na području SR Hrvatske, radni sastanak, Rijeka, 19.-20. studenoga 1971: 96‒105. Zagreb: Republički zavod za zaštitu spomenika kulture.

115

3.

Ceramic building materials production: production site location, production organisation, distribution channels, regional specificities

An overview of the CBM industries of ancient Viminacium Ljubomir Jevtović, Ilija Danković Abstract This study will analyse the CBM (Ceramic Building Material) industry of ancient Viminacium, the capital of the Roman province of Moesia Superior and the military camp of the Legio VII Claudia. Extensive archaeological excavations have unearthed 14 brick kilns in the areas surrounding the civilian settlement and the military camp. The majority of kilns were spread out across two main production areas. They were operated by the legion and were identified as the main areas of CBM production in Viminacium. Several kilns were found outside of these areas and probably represent small scale private production.

Keywords Viminacium, Moesia Superior, Ceramic building materials (CBM), roman brick kilns, brickmaking, Legio VII Claudia.

Figure 1. Plan of Viminacium with the position of the kilns (the authors).

Introduction1 The remains of ancient Viminacium are located in Eastern Serbia, next to the modern town of Kostolac, on the right bank of the Mlava river, near its confluence with the Danube (Figure 1). Its history begins with the construction of a military outpost in the second half of the 1st century AD. During the most of its existence, it was the main military camp of the Legio VII Claudia. In the late 1st - early 2nd century, a civilian settlement developed to the west of the camp. During the reign of Emperor Hadrian, in AD 117, Viminacium was granted The article results from the project: IRS - Viminacium, Roman city and military camp – research of material and non-material culture of inhabitants by using the modern technologies of remote detection, geophysics, GIS, digitalization and 3D visualization (no 47018), funded by The Ministry of Education, Science and Technological Development of the Republic of Serbia. 1 

the status of municipium, while in AD 239, it gained the status of colonia, during the reign of Emperor Gordianus III (Mirković 1986: 21-59). The city was destroyed by the Huns in AD 441 and never again regained its former glory (Поповић 1968: 39). The first excavations of Viminacium were conducted in 1882. Since then, the site was occasionally excavated throughout the 20th century and regularly since 2001 (Спасић-Ђурић 2015: 11-16). During these excavations, various parts of the site were at least partially explored and 14 brick kilns were unearthed.2

Since the presentation of this paper on the IV International Archaeological Colloquium ’Roman Ceramic and Glass Manufactures production and trade in the Adriatic region and beyond’ in Crikvenice in 2017, two new kilns were discovered. Unfortunately, due to the technical reasons, these kilns were not included in this analysis. 2 

roman pottery and glass manufactures (Archaeopress 2022): 117–124

Ljubomir Jevtović, Ilija Danković

No.

SITE

TYPE DIMENSION

HEIGHT

no. OF CROSS FLUES

GRILL OVEN TICHNESS

1

II/b

4.30 x 2.90

2.30

6

0.15

2.70 x 2.50 0.70 m

1.05 x 0.90

II/b

4.85 x 3.90

3.10

6

0.37

3.10 x 2.80 1.60 m

1.30 x 1.00

II/b

3.80* x 4.00

3.35

5*

0.47

2.60 x 2.75 1.66 m

0.90* x 0.80

4

Pirivoj Pećine Zanatski Centar Pećine Zanatski Centar Pećine Zanatski Centar

II/b

5.05 x 4.00

3.50

6

/

3.30 x 2.80 1.60 m

0.75* x 1.10

5

Pećine

II/d

7.30* x 3.80

1.95

9

/

/

/

30s of the 3rd — 20s of the 4th late 2nd - first half 3rd late 2nd - first half 3rd late 2nd - first half 3rd abandoned in first half 3RD

6

Pećine

II/b

4.30 x 2.95

2.85

4

0.40

2.70 x 2.55 1.10 m

1.00 x 0.90

2nd - first half 3rd

7

II/b

5.50 x 4.00

> 3.00

4

0.45-0.60

3.40 x 3.55 1.00 m

1.30 x 1.00

2nd - first half 3rd

II/b

4.80 x 2.75

1.60

3

/

2.20 x 2.15 0.40 m

0.85 x 0.80

/

9

Pećine Livade kod Ćuprije Livade kod Ćuprije

II/b

3.85* x 3.25

1.80

/

/

2.10 x 2.60 0.50 m

/

/

10

Provalije

II/a-b 3.10 x 3.50

/

/

/

/

/

/

2nd-3rd?

11

Provalije

II/a-b 2.60 x 2.70

/

/

/

/

/

/

2nd-3rd?

12

Provalije

II/a-b /

/

/

/

/

/

2nd-3rd?

13

Provalije

II/d

5,40* x 5,80

3.55

6*

 

/ 4.20* x 4.60

1,60

/

2nd-3rd

14

Provalije

/

/

/

/

/

/

/

/

2nd-3rd

2 3

8

OVEN HEIGHT

/

PRAEFURNIUM DATE

Table 1. Characteristics of CBM kilns of Viminacium.

In spite of relatively numerous finds, the ceramic building material (hereafter CBM) industries of Viminacium were never thoroughly studied. Even though these finds were interesting to previous researchers, only a few were published, usually not in detail, with occasional attempts at summarising the industries and their products (Јордовић 1994; Јуришић 1956). Recently, new discoveries were made and some of the older ones were reviewed and published (Jovičić and Milovanović 2017; Raičković and Redžić 2005; Redžić et al. 2018).

Site ‘Pirivoj’ - eastern necropolis A single kiln (no. 1) was discovered on the northern edge of the eastern necropolis, near the road that led east from the eastern gate of the military camp (Jovičić and Milovanović 2017). The kiln belongs to a common type of rectangular kiln (4.30 x 2.90 m) with one main canal and six side flues (type II/b) on the same level as the main canal (Figure 2: 1). It is preserved to the height of 2.30 m. The walls of the oven (2.70 x 2.50 m) are 0.70 m high. Two floor levels and a tri-layered mud plastering of the walls testify that the kiln was used multiple times. On the other hand, the vitrified walls and a collapsed part of the oven floor point to its demise. The kiln was in use between the third decade of the 3rd and the first/ second decade of the 4th century. Since no stamped CBM were found, it can be theorised that this was a privately owned workshop. To the southwest of the kiln, on the opposite side of the necropolis, a possible source of the raw material was identified. A large pit was partially excavated, filled with 2nd and 3rd century material from the cleaning of the necropolis (Raičković and Milovanović 2010: 7-56; Vuković 2010: 57-82).

Nevertheless, a more comprehensive study of the CBM industries is needed, using all of the previous work as a foundation. Thus, the aim of this paper is to give an overview and layout of brick kilns on the territory of Viminacium. A short analysis of the sites, their kilns and other objects related to the production process will be presented. This will enable the development of an overall image of the industries of Viminacium and some insight into production methods, organisation and development to be gained. Layout of the industries

based on their shape (see type II/d/1 in: Harizanov 2019: 231), which was confirmed by finds, if existed. Their types were determined using the standard typology composed by N. Cuomo di Caprio (1972), based on the shape of the kiln and the support of the perforated floor of the kiln. The kilns were given a new numeration, while their old designation, used in documentation and literature will be given in the footnotes.

Excavations have identified 4 areas of production (Figure 1), with 14 rectangular, vertical twin-chambered updraught kilns used for brickmaking3 (Figure 2, Table 1). 3 

The primary functional attribution of the kilns was carried out

118

An overview of the CBM industries of ancient Viminacium

Figure 2. Plans of Viminacium kilns (documentation of the Institute of Archaeology, Belgrade).

119

Ljubomir Jevtović, Ilija Danković

Figure 3. Plan of the Craftsmen centre with the brick kilns 2 – 4 (after Јордовић, 1994: 102, Pl. 1).

different number of arches that form the cross flues. The oven dimensions are between 2.60 x 2.75 and 3.10 x 2.80 m. The kilns were preserved to a greater extent and several layers of plastering, as well as some repair work, prove that they were used up to five times. Inside, several fragmented tiles bore the regular stamps of the Legio VII Claudia (LEGVIICL). A few were stamped with chronologically sensitive LEGVIICLANT stamps that date the kilns in the late 2nd – first half of the 3rd century.5

Site ‘Pećine’ – southern necropolis The largest ceramic production centre of ancient Viminacium was located on the same territory as the southern necropolis of the city (the site of Pećine), in the area of the modern day Kostolac B thermal power plant. Thus far, six brick kilns have been excavated (no. 2-7), while four more have been detected (Јордовић 1994: 96). All excavated kilns were in use between the end of the 2nd and the first half of the 3rd century. Several distinctive areas of production can be identified.

Next to the kilns, the remains of a porch structure were discovered (Figure 3). Its outlines can be deduced from the eight surviving pillar bases that were variously spaced and lined into four rows that are approximately at the same distance. The entire area was covered with roofing tiles, which testifies that the pillars carried a roof construction. Analogies for these types of structures are rare, but there is little doubt that this construction was used as a workshop.

The craftsmen centre was the core production area for ceramic items of Viminacium (Јордовић 1994; Raičković 2007: 11-17). Three CBM and four pottery kilns, a well, several dumping pits and a porch-like structure were spread over an area of around 700 m2 (Figure 3). The brick kilns were organised into a single production unit, with two kilns next to each other in a line, and a third next to them. They were completely dug in, except for the stokehole sides, which were oriented towards the central working area, isolated from the surrounding ground by a wall (3.6 m high). This area was used for the preparation and storage of fuel for the kilns. Each of these kilns (no. 2-4)4 has a single main canal (type II/b) and cross flues at a higher level than the main canal (Figure 2: 2-4). They vary in dimensions (4.80 x 3.90 m; 3.80 x 4.00 m; 5.05 x 4.00 m), due to the

In the vicinity of the craftsmen centre, during the laying of the foundations for a chimney of the power plant, a single kiln (no. 5) was discovered (Raičković and Redžić 2006: 85-86, 88, 94-95, 97-98, Photo 3-4; P. IV-V, VIIBased on the honorary title Antoniniana, these stamps were in use during the reign of Severus Alexander (Kurzmann 2005: 328). The use of CBM with the same stamps in the construction of the kilns was acknowledged in 2013, during their deconstruction and relocation to the Domus Scientiarium of Viminacium. 5 

In the earlier works and documentation, kilns were designated as kilns 8 – 10 (Јордовић 1994: 99-101; Raičković 2007: 13-14) 4 

120

An overview of the CBM industries of ancient Viminacium

Figure 4. Part of the clay pit with stacked CBM (documentation of the Institute of Archaeology, Belgrade).

VIII).6 It belongs to a complex type of rectangular kilns with two main canals (type II/d), larger in dimensions (7.30* x 3.80 m) and with 9 cross flues (Figure 2: 5). Only the lower part of the kiln, 1.95 m high, up to the grill, was preserved. The kiln was fired at least twice and, after its disuse, it was filled with earth and fragmented CBM, some bearing stamps of the legion (LEGVIICL). A single coin of Emperor Caracalla suggests that the kiln was abandoned in the first half of the 3rd century. Excavations of its surroundings were never conducted, but the complex type of the kiln and the vicinity of the production centre suggest that it was a part of a more organised production.

m) with a central canal and 4 cross flues (Figure 2: 6-7). They were preserved to a greater extent, with almost the entire praefurnium and the oven walls up to 1.10 m high. Kiln 6 was reused twice and then abandoned, after the collapse of its grill. On both sides of its praefurnium, piles of wasters were discovered, two of which had legionary stamps (LEGVIICL). The second kiln was used at least 3 times, and since it had no traces of damage its abandonment remains a mystery. On each side of its praefurnium opening, a pillar made out of bricks was discovered.9 After the abandonment of the kilns, the pit was filled with earth containing unfinished and cast-off CBM and other waste material from the necropolis.

In the southern part of the site, a large clay pit was partially excavated (Redžić et al 2018: 87-88). It had an irregular shape, c. 40 x 32 m in dimensions and was more than 4 m deep (Figure 4). It was used as a clay pit, a working area, storage for the finished products and as a dumping area for two kilns that were placed on its edges. Kilns no. 67 and 78 belong to the common type (type II/b) of rectangular kilns (4.30 x 2.95 m; 5.50 x 4.00

Site ‘Livade kod Ćuprije’ In the vicinity of the, by that time abandoned, southern necropolis lie the remains of a 4th century villa rustica (Jovičić and Redžić 2012). Near the villa, one pottery and two brick kilns, a well and several dumping pits were discovered. It remains uncertain if the brick kilns operated in the same period as the villa, or predate it, as is the case with the pottery kiln. If latter is the case, perhaps it should be linked to the large ceramic complex located on the site Pećine, that lies some 200-250 m

Kiln was designated as brick kiln, or brick kiln no. 3 (Raičković and Redžić 2006: 85-86). 7  Kiln no. 6, labeled kiln 1 near Church B in the field documentation, was partially publish by A. Raičković and S. Redžić (2006: 85, 91, 93, P. I, III), who designate it as Brick kiln no. 2. 8  Kiln no. 7 was named kiln 2 near Church B in the field documentation and only it’s plans and photos were previously published (Raičković and Redžić 2006: 84, 91, 96, 106, Photo 2, 4; P. I, VI). 6 

A similar structure on the entrance of the fire tunnel was found in the Roman tilery at Great Cansiron Farm in Hartfield (East Sussex). D. R. Rudling et al. (1986: 198-199, fig. 4) suggest that these pillae could have been used for blocking the opening of the praefurnium after firing was complete. 9 

121

Ljubomir Jevtović, Ilija Danković

Figure 5. Kilns 13 and 14 at site ’Provalije’ – Teko A (documentation of the Institute of Archaeology, Belgrade). .

eastwards. Both kilns (no. 8 and 9) belong to the common type of rectangular kilns (4.80 x 2.75 m; 3.85* x 3.25 m) with a central canal (type II/b). Kiln 810 (Figure 2: 8) was partially preserved, up to a height of 1.60 m with the oven (2.20 x 2.15 m) mostly destroyed and its walls preserved to a height of 0.40 m (Raičković and Redžić 2005: 84-85). Available data about the second kiln – kiln 9 (Figure 2: 9) is scarce, but based on drawings it was preserved to a height of around 1.80 m, with the oven intact (2.10 x 2.60 m) and its walls preserved to a height of 0.50 m.

discovered. Kiln 14 was located underneath the modern road, so only a part of its outer shell was uncovered (Figure 5). Next to it, remains of a another kiln were excavated.12 Kiln 13 is a complex rectangular kiln (5.40* x 5.80 m) with two main canals (type II/d) and 6 or more cross flues (Figure 5). Although most of its oven floor (4.20* x 4.60) and some parts of its walls survive up to a height of 1.60 m, it was heavily damaged. Parts of the kiln had collapsed due to vitrification and no traces of repairs were detected. In its filling, a couple of raw tiles were discovered, as well as some completely melted CBM. One tile had a stamp of the Legio IV Flavia, while four had peculiar stamps of the Legio VII Claudia with the name of the officinator of this workshop – Atilius (LEGVIICLOFATILIVIM). These stamps suggest that the workshop operated some time during the 2nd or the 3rd century AD.

Site ‘Provalije – TEKO – A’ The site occupies an area of the Kostolac A thermal power plant in Novi Kostolac, some 3.5 km west of the civilian settlement of Viminacium on the opposite bank of the river Mlava. In 1952, three brick kilns (no. 10-12) were excavated on the cite (Јуришић 1956: 129-130).11 They were arranged in a triangular formation, 28 m and 11 m apart. Unfortunately, only scarce data is available on them, but based on the given dimensions, it can be said with certainty that they belong to the type of rectangular kilns with a single main canal (type II/a-b). During rescue excavations in 2015, two new kilns were

Discussion Archaeological excavations of ancient Viminacium have identified four areas of CBM production (Figure 1). Large scale and organised production was detected on two sites – Pećine and Provalije – TEKO – A. The largest production area, with several workshops, was located on the site of Pećine, in the vicinity of the

Raičković and Redžić (2005: 84-85) designate the kiln as Brick kiln no. 1. 11  A. Jurišić (Јуришић 1956: 130) mentions that in the period 1941-1944 several other brick kilns were discovered on the territory of Novi Kostolac. Unfortunately, these finds were never published and we were unable to find any documentation or other reference concerning them. 10 

Due to technical reasons, the plan of this kiln was not available at the moment of the printing of this publication. 12 

122

An overview of the CBM industries of ancient Viminacium

civil settlement. The workshops functioned during the late 2nd and first half of the 3rd century, and occupied the same area as the Southern necropolis of the city, that developed around them. One workshop was organised around the clay pit that also served as its storage and dumping area, while the other had a more conventional layout, with kilns organised into a single production unit with other supporting structures next to them. Unfortunately, even though it is one of the most thoroughly excavated areas of ancient Viminacium, it was not systematically excavated. The result of such circumstances is that we lack the complete picture of the organization of the workshops and their spatial and temporal correlation. Finds of stamped CBM, both in the constructions of the kilns and in their backfill, testify that the entire production was operated by the Legio VII Claudia.

works have been dedicated to this subject, crucially important for the development of the city. Numerous finds of stamped CBM prove that the largest producer of this material was the Legio VII Claudia. It operated two major production areas that were organised into several smaller workshops. One of the remaining problems is that, seemingly, both centres worked in a similar time span, during the 2nd and 3rd centuries. Numerous stamps testify that military CBM production in Viminacium started in the late 1st century and lasted until the middle of the 4th century (Jevtović 2013: 62). However, at the moment the location of the earliest or latest production areas cannot be identified. On the other hand, private production was not attested by the stamps, but it seems reasonable to assume that it existed, even if only on a small scale. Three kilns (nos. 1, 8-9) can potentially be attributed to this mode of production and were probably used for a single purpose.

The second major production area was identified on the site of Provalije – TEKO – A, relatively far from the civilian settlement on the opposite bank of the Mlava river. Since very limited data regarding earlier excavations is available, and only small scale rescue excavations were conducted recently, very little can be said about the layout and organisation of this workshop. Still, the fact that five kilns were discovered, one belonging to a more complex type, point to the conclusion that this was also a larger production centre. Finds of stamped tiles led to the conclusion that this was also a workshop of the Legio VII Claudia, and even provided the name of its officinator. This workshop was in use during the 2nd and perhaps during the 3rd century.

Unfortunately, most of the elements of CBM industries were discovered during rescue excavations. Additionally, the areas surrounding the kilns were not thoroughly excavated. Even more unfortunate is the fact that only certain elements of the production were properly identified and excavated, so only a few clay pits and supporting production structures are known. Future excavations and the further study of discovered elements will bring us closer to answering some of these questions and give a better understanding of the CBM industries of ancient Viminacium. References

Two smaller production units were also discovered, one on the edge of the eastern necropolis, and another, near the remains of a 4th century Roman villa. These kilns are of a simple common type, smaller than most of the others and do not seem to be a part of any organised production. The lack of military stamped tiles inside them points to a private production facility. These kilns seem to have been built to satisfy a single purpose or market. However, this conclusion has to be taken with a degree of caution, since the immediate surroundings of the kiln on the eastern necropolis have not been explored and limited data is available on the kilns near the villa. Both the precise location and the dating of the kilns on the site Livade kod Ćuprije is unknown, but it has to be taken into account that they were located relatively near the large production complex on the site Pećine.

Cuomo di Caprio, N. 1972. Proposta di classificazione delle fornaci per ceramica e laterizi nell’area italiana, dalla preistoria e tutta i’epoca romana. Sibrium 11: 371‒464. Harizanov, A. 2019. Ceramic kilns in the territory of modern Bulgaria from the 1st to the 6th century AD. Sofia: Bulgarian Academy of Sciences, National Archaeological Institute with museum. Jevtović, Lj. 2013. Nalazi opeka sa pečatom VII Klaudijeve legije na teritoriji provincije Gornje Mezije, Unpublished master thesis, University of Belgrade. Јордовић, Ч. 1994. Грнчарски и цигларски центар у Виминацијуму. Саопштења XXVI: 95‒105. Jovičić, M. and S. Redžić 2012. Late Roman villa on the site Livade kod Ćuprije — A contribution to the study of villae rusticae in the vicinity of Viminacium. Arheologija i Prirodne nauke 7 (2011): 369‒386. Jovičić, M. and B. Milovanović 2017. Roman brick kiln from the eastern necropolis of Viminacium. Arheologija i Prirodne nauke 12 (2016): 19‒38. Јуришић, А. 1956. Археолошки радови. Саопштења I: 123‒138.

Conclusions CBM was one of the most important building materials of ancient Viminacium, discovered in various elements of its architecture. Although the earliest finds of CBM industries were discovered during the middle of the last century, they were not appropriately studied. Few 123

Ljubomir Jevtović, Ilija Danković Kurzmann, R. 2005. Soldiers or Emperors, in Z. Visy (ed.) Limes XIX proceedings of the XIX International Congress of Roman Frontier Studies: 327‒336. Pécs: University of Pécs. Mirković, M. 1986. Inscriptions de la Mésie Supérieure, Volume II, Viminacium et Margum. Belgrade: Centre d’études épigraphiques et numismatiques. Поповић, В. 1968. Увод у топографију Виминацијума. Старинар XVIII (1967): 29‒55. Raičković, A. 2007. Keramičke posude iz Zanatskog centra iz Viminacijuma. Beograd: Centar za nove tehnologije and Arheološki institut. Raičković, A. and S. Redžić 2005. Keramičke i opekarske peći Viminacijuma – lokacije ‘Pećine’ i ‘Livade kod Ćuprije’. Arheologija i prirodne nauke 1: 81‒106. Raičković, A. and Milovanović, B. 2010. Rezultati istraživanja jugozapadnog dela lokaliteta Pirivoj (Viminacijum). Arheologija i Prirodne nauke 5 (2009): 7‒56.

Redžić, S, M. Jovičić, N. Mrđić and D. Rogić 2018. Zaštitna arheološka istraživanja na lokalitetu Pećine (Viminacijum) 2016. godine, in I. Bugarski, N. Gavrilović Vitas and V. Filipović (eds) Arheologija u Srbiji, projekti Arheološkog instituta u 2016. godini: 7991. Beograd: Arheološki institut. Rudling, D. R., C. Cartwright, G. Swift, S. Foster, J. Shepherd, P. Hinton and F. Tebbutt 1986. The excavation of a roman tilery on Great Cansiron Farm, Hartfield, East Sussex. Britannia 17: 191‒230. Спасић-Ђурић, Д. 2013. Град Виминацијум. Пожаревац: Народни Музеј Пожаревац. Vuković, S. 2010. Ostaci životinja sa jugozapadnog dela lokaliteta Pirivoj (Viminacijum). Arheologija i Prirodne nauke 5 (2009): 57‒82.

124

Brick by brick, tile by tile. The ceramic building material in Labraunda, Turkey Alexandra Dolea Abstract The present study aims to provide a general overview on the ceramic building material in Labraunda, Turkey. The focus lies with two recently excavated sectors, the so-called ‘Water Pool’ and ‘East Bath’, encompassing the Roman and Late Antique materials. The research was initiated and developed throughout the 2014-2016 field campaigns. Although the present contribution is not an extended analysis on this category of finds, it underlines the importance of this find group to the general picture, as it completes archaeological and architectural features, and provides indications regarding production, use and reuse, functionality changes and adaptation.

Key words Ceramic building material, Roman and Late Antique constructions, Labraunda, Caria.

Figure 1. Location of Labraunda excavation site (© GoogleEarth).

Introduction Labraunda is located in south-western Turkey, 14 km from the modern city Milas, in the area known in ancient times as Caria (Figure 1). The Sanctuary of Zeus Labraundos was a distinguished cult centre for the Carians and it dates back to the middle of the 7th century BC. Most of the buildings from Labraunda were erected during the 4th century BC by the Hekatomnid dynasty; however, the archaeological discoveries indicate intense activities until the late 6th century AD.1 Post antique activities in Labraunda can be observed until the 13th century AD (Blid 2016: 203-206). 1 

The systematic excavations at Labraunda started in 1948 and brought to light a series of buildings such as the temple of Zeus, andrones, stoas, two roman baths and two Christian basilicas. Based on the results of the excavations and surveys, spring-houses, tombs and the fortified Acropolis were identified around the sanctuary (Figure 2). This article aims to provide a general overview on the author’s research on the ceramic building material in Labraunda. The chronology, typology, composition and the archaeological context of the finds will be presented, focusing on the Roman and Late Antique material, especially the one found in the Water Pool

roman pottery and glass manufactures (Archaeopress 2022): 125–139

Alexandra Dolea

Figure 2. General plan of the sanctuary from Labraunda (© O. Henry).

and East Bath sectors.2 The current stage of research represents the material study on-site and in the depots of Labraunda conducted during the 2014-2016 excavation seasons. The methodology of work3 was adapted to the material volume and context4, and it will be further explained.

periods. From the typological point of view, the presence of both Laconian and Corinthian types concerning the pan and cover tiles was noted. Regarding the bricks, the rectangular and square shaped types were observed, but also round shaped, typical for a hypocaust system, or trapezoidal ones were found. There are also deviations of the usual shapes, which will be mentioned within their context. These typological features are generally found in both public and private buildings, at Labraunda the public aspect being highlighted due to its role played as a sanctuary.

The ceramic construction material from Labraunda can be dated between the Archaic and the Late Antique The Water Pool sector was excavated under the direction of Dr. Ö. D. Çakmaklı and C. Bost conducted the East Baths excavations. 3  Due to the various stages of the material fragmentation, a MNI (minimum number of individual) system was applied in studying the ceramic building material. 4  The author had access to the excavation documentation exclusively through the published preliminary reports. 2 

Written sources on the brick-use in Caria Vitruvius mentions that “not only public and private buildings, but even royal structures, are built of brick” 126

Brick by brick, tile by tile. The ceramic building material in Labraunda, Turkey

(Vitr. 2, 8. 9) and also that “in the house of Mausolus, a very powerful king of Halicarnassus, though all the ornaments are of Proconnesian marble, the walls are of brick, are remarkably sound at the present day, and the plastering with which they are covered is so polished that they sparkle like glass. The prince who caused them to be thus built was not, however, restrained by economy; for, as king of Caria, he must have been exceedingly rich” (Vitr. 2, 8. 10). Vitruvius describes Mausolos’ residence from Halicarnassus and the situation after his death when Artemisia II, his sister and wife, became ruler of Caria. The revolt of the Rhodians against her is also described (Vitr. 2, 8. 14-15), ended with the victory of Artemisia and the settling of a garrison of her troops in Rhodes. Vitruvius emphasizes afterwards the importance of using bricks in construction, by asking the reader not to disregard this material: “If therefore, kings of such great power did not despise brick buildings, those who, from their great revenue and spoils in war, can afford the expense not only of squared and rough stone, but even of marble buildings, must not despise brick structures when well executed” (Vitr. 2, 8. 16).

to these structures was also unearthed sustains this hypothesis. The sanctuary has numerous springs within and around, so the water supply could have been more or less constant. These water collectors seem to have functioned in a cascade system (Çakmaklı 2015: 384). Although this type of construction is common throughout the Roman era, a more precise chronology can be proposed, namely the 1st and 2nd centuries AD. A hypothesis also supported by recent mortar analysis performed in 2014 on samples that were collected from the South terrace wall (Ergenç 2015: 306-308; 2016: 353358). Further evidence in favour of this chronological frame is provided by the pottery found in the same context with some of the ceramic construction material presented in this study. This includes a large part of the ceramic finds consisting of local/regional produced Cnidian-type bowls and kitchen ware (Çakmaklı 2014: 317 2015: 387; 2017: 257-269; Lungu 2018: 277). The discovery of a kiln (2.70 m diameter) is also highly relevant for the general context of the ceramic building material. The kiln was situated in the centre of the south basin, and was installed, most probably, after the pools system ceased to function (Figure 3). The presence of a coin issued during the reign of Constantine II could indicate the use of the kiln within or until the 4th century AD (Çakmaklı 2015: 387-388). Until now, the functionality of the kiln cannot be specified, as a large amount of ceramic construction material and pottery was found inside it, but no rejected pieces or other indicators for production (Lesguer 2016: 441-447). In general, the ceramic building material is abundant in this sector, but it lacks stabile archaeological contexts, as it was mainly found in debris, filling, levelling, and collapsed layers.

The same subject, the palace of Mausolos in Halicarnassus, is taken in consideration by Pliny the Elder. He mentions that its walls are made of brick and “covered with marble of Proconnesus” (Plin. Nat. 36. 6). The Hecatomnid dynasty has a great significance in the history of Caria and Labraunda. The fact that wellknown literary sources mention these historical figures in connection with the usage of fired bricks even in their own houses, when they could have afforded more luxurious materials, shows the high quality of the material in question (and for this reason its economic preference) and the extended usage of the bricks for all types of constructions. The ceramic building material from Labraunda also reflects these aspects as it will be shown in details below. Presumably the early pieces are characterized by (semi)fine and compact fabrics (from Archaic to Roman times), and the later productions (Late Antique to Byzantine) tend to have a rougher consistence of the fabric.

At the current stage of research,5 some observations can be disclosed concerning the typology of the ceramic construction material from the Water Pool sector (Graph 1). The pan tiles are represented through both Laconian (Brodribb 1987: 23, Fig. 12; 1987: 10-11; Wikander 1986: 43, Fig. 20, no. 15) (Figure 4.1-3) and Corinthian (Wikander 1993: 30, Fig. 4, H2; Karlsson 2010: 80, Fig. 48 (lower profile); Wulf 1999: 17, Abb. 9) (Figures 4.5, 5.6) types, which seem to belong in great majority to Fabric 9 (see Annex). The same typology was also determined for the cover tiles (Stevens 1950: 189, Plate 82, no. 2; Wikander 1986: 28, Fig. 8, no.11; Karlsson 2010: 78, Fig. 46; Henry 2014: 286, Fig. 52.) (Figure 5.7-9); in this case, the fabrics are more diverse (Fabrics 8, 9 and 10). The bricks belong to three big morphological categories: round, square (Brodribb 1987: 36-37) (Figure 6.12) and rectangular (Brodribb 1987: 41-42). As already mentioned, the round bricks tend to be connected to the presence of a hypocaust construction. This category

The Water Pool sector The archaeological situation of the Water Pool sector is characterized by a built ensemble that was identified already during the 2013 campaign (Çakmaklı 2014: 316-317). It is located in the southern area of the site, between the terrace wall and the modern road. The structures are made of small/medium sized stone blocks bind with pink hydraulic mortar and they seem to function as basins for the catchment and evacuation of the water pouring down from the sanctuary area. The fact that a part of the canalization system related

A preliminary report was already published, see Dolea 2016: 383387. 5 

127

Alexandra Dolea

Figure 3. Aerial photos of the Water pool sector. View from East (up). View from South (down) (© Ö. Çakmaklı).

128

Brick by brick, tile by tile. The ceramic building material in Labraunda, Turkey

Graphic 1. Ceramic building material overview from the Water Pool sector (© A. Dolea).

is less represented in this sector, but follows the same typological patterns like the ones documented for the East Bath sector (see below a detailed description). One particular piece is worth mentioning (Figure 6.13) as it seems to have been cut from a previous square or rather rectangular brick and (re)used for another specific purpose, like constructing or repairing a hypocaust pila (cf. similar pieces documented for the East Bath). However, the possibility of material intrusion from neighbouring constructions in the Water Pool sector should not be excluded. The water pipes (Figure 5.1011) are also representative for this sector and they usually have a semi fine fabric, very similar to the thin round and square bricks composition (Fabrics 7 and 10). Few tubulus (Brodribb 1987: 70-83; Graciani 2009: 6-7) fragments were identified within this sector and they are most probably intrusions from other buildings and contexts (Figure 4.4). Their fabric is rather (semi) coarse and dusty (Fabric 5) with common features as some types of rectangular bricks identified within both the Water Pool and East Bath sectors. During the 2017 excavation campaign further construction details were uncovered, like the presence of a herringbone-type brick floor on the bottom of the south basin and several water pipes in the north wall of the basin, out of which one seemed to be closed with a tile fragment as if it was taken out of use at a certain moment in time (Çakmaklı 2018: 316).

Figure 4. Laconian pan-tiles (1-3), tubulus (4), Corinthian pan-tile (5). (© A. Dolea).

The present overview of the Water Pool ceramic building material represents a part of the recently revealed features and originates mostly from mixed layers, 129

Alexandra Dolea

Figure 5. Corinthian pan-tile (6), Laconian covertile (7), Corinthian cover-tile (8-9), water-pipes (10-11) (© A. Dolea).

Figure 6. Square brick (12), Reworked brick (13) (© A. Dolea).

which can only partly help determining the chronology and functions. The bricks and the water pipes are very characteristic materials for the construction of this sector and they are mostly belonging to the Late Hellenistic and Roman eras. The high number of roof tiles in the graphic below has to be treated carefully, as their belonging layers were mostly mixed and they were most probably in secondary positions. Further research is needed for the in situ materials in correlation with the most recent excavated layers and various other object groups.

the honorific inscription devoted to the same person, in which the baths are also mentioned.6 This chronology is supported also by various archaeological features and finds7, as well as by some recent mortar analysis performed in 2014 on samples which were taken from walls and floors of the East Bath dated to the 1st century AD (Ergenç 2015: 306-308; 2016: 353-357). Furthermore, the type of the ceramic construction material used for the hypocaust system of the baths provides us indication for a chronological chart concerning the functionality of the building. Thus, it can be suggested that there was another construction phase of the baths at the moment of installing the suspensurae and the tubuli, which could have taken place at the end of the 1st century – first half of the 2nd century AD. This intervention might be connected to the extension of the bath marked through a

The East Bath sector This sector has been excavated partly in the early 1950s when several ceramic construction materials related to the function of the bath were noted, i.e. tubuli and tegulae mammatae (Blid 2016: 143, Fig. 149), another small scale excavation was performed in 2006 (Blid 2016: 144), while the most recent research has been initiated in 2014 (Bost 2015: 355-366).

These inscription fragments were found in 1951 in secondary positions, the first in front of the entrance of the East Church, the second in a wall of the same church; see Crampa 1972: 18-20, 133-134. For a better identification of the find spots, see Bild 2016: 132. 7  I would like to express my gratitude for the (Late) Roman pottery chronology to Dr. A. Streinu. For a recently published overview on early Roman pottery in this sector, see Streinu 2020. 6 

The date of construction is commonly placed around the mid-1st century AD due to the inscribed dedication of the baths by Tiberius Claudius Menelaus and due to 130

Brick by brick, tile by tile. The ceramic building material in Labraunda, Turkey

Graphic 2. Classification of bricks from East Baths (© A. Dolea).

dedication for a new apodyterium made by the priest Flavius Neon dated to the Hadrianic period.8 Since the archaeological situation indicates the baths were built on top of an earlier Classical/Hellenistic construction (at least partly), future research should focus on how its functionality changed and what architectural developments the building went through.9 The baths underwent a rebuilding phase in the 5th century AD, as a result of the building of the East Church, when their function changed and the baths were most probably included within the church complex (Bost 2016: 435; Blid 2016: 133).

These smoothing traces could appear only on the upper surface of the objects, which allows the identification of shapes out of small fragments and enhances the probability of differentiating between productions and/or producers. For the 2016 campaign, the study focused exclusively on bricks due to the vast variety of forms and fabrics. In addition, this material is an essential component of bath buildings; therefore, a detailed study on bricks would enhance the excavation information and the architectural features. A more complete picture will be obtained by correlating our data to the chronological frame brought by the pottery and other finds’ studies, and to the archaeological information of the documented layers.

The ceramic building material resulted from the 2015 excavations campaign at the East Bath10 was temporarily stored in the East Church, north to the baths. The piles were initially separated according to their provenance based on stratigraphical units.11 Each unit was documented by dividing the material typologically, by colour and fabric determinations,12 and nevertheless by traces of smoothing the upper surface of the objects.13

Several forms of the material can be distinguished by shape: round, square, rectangular and trapezoidal. All of these show further variations of dimensions and fabrics (Graph 2).

This dedication fragment was also found in 1951 in a secondary use, in the vicinity of the previously mentioned inscription fragments (see above fn. 25); see Crampa 1972: 21; Blid 2016: 15, 132. 9  The pre-Roman pottery discovered in the East Baths dates from the Archaic to the Hellenistic period, s. Lungu 2018: 272-273. 10  For preliminary results of the East Baths 2015 campaign, see Bost 2016: 424-435. 11  The material collection was highly selective and the excavators collected only noticeable clusters of ceramic building material. 12  According to Munsell 2000. 13  Usually the bricks and the tiles were formed and put to dry in frames. The excess of clay was removed with various tools from the upper surface, the flanges and the cutaways were shaped (for tiles), and then the visible surface was smoothened before drying and being extracted from the frames, let dry between three to four weeks and then cooked in the kilns. The underside of the tiles and especially bricks is also of high importance as it can show tool marks and the medium on which they were put to dry (sand, straws, wooden planks etc.). 8 

For instance, at least two types of round bricks were observed, representative for the hypocaust system. One form has a bigger diameter (23.5 - 26.5 cm) and a smaller thickness (3.4 - 5.2 cm), while the other one has a smaller diameter (18.6 - 21 cm), but is thicker (5.7 - 7.1 cm) (Figure 7.1). This last category includes two secondary used pieces, as they were cut probably from massive rectangular bricks (see below) belonging to Fabric 4. (Figure 7.2) Another probable secondary use was identified in two cases when circular bricks were broken in two and the middle part was covered with lime mortar, hence these half-round shaped bricks were (re)used, most probably to create seats (Brodribb 131

Alexandra Dolea

Figure 7. Round bricks (photos and drawings: S. Gür, digitization: L. Karkouki (© A. Dolea).

132

Brick by brick, tile by tile. The ceramic building material in Labraunda, Turkey

1987: 55) (Figure 7.3). Traces, which relate either to a production centre or to a producer, were identified in case of seven, round shape brick fragments, all of them having the no. 1 fabric type and irregular edges (maybe same/similar wooden frame(s) were used for modelling the shapes or even no frame). With only two exceptions mentioned above (the two pieces cut from other bricks), all the round bricks have compact and smooth fabrics (Fabrics 1, 2 and 3), being differentiated only through dimensions. Their contexts of discovery are the filling layers of the hypocaust and the collapsed suspensura. This might suggest various possibilities of interpretation, including the reparation of the hypocaust system and the provenance of material from different sources of manufacture.

From the functional point of view, the two rectangular types of bricks (the massive and the long ones) form mainly the upper elevation of the walls and the vaulted roof of the baths, both shapes being used together for a better resistance and fixation.14 The ones used for vaulting are differentiated by the mortar joints which are thicker towards the outer curve of the arch (extrados). The next shape is the square one which consists of 21 full shaped pieces. 19 bricks belong to fabric no. 1, while the other two pieces belong each to a different fabric type (nos. 1 and 10). It is very possible that the use of this shape was purposed for the construction or reparation of the hypocaust, and suspensura, but also for other building adornments, such as arches or architectural variations. The stratigraphy suggests they were installed and used mostly when the bath was functioning, as they were recovered from filling layers inside the hypocaust on top of which a general levelling was identified (Bost 2016: 426-429) (Figure 9.9).

Among the rectangular bricks, a massive form characterized mainly by coarse fabrics (no. 4 and no. 5) and smoothening traces along the shorter length could be distinguished. As particular marks are the presence of usually three parallel finger traces that go along the bricks, in ‘S’ or ‘X’ shape (Figure 8.4-5), while some going diagonally from a corner to another in a rounded shape. Furthermore, organic material and animal imprints were identified. This category of bricks was uncovered exclusively in layers with collapsed material, which can be connected to Late Antique phases of constructions, like the building of the East Church complex that affected also the former baths.

The last shape has a trapezoidal form. 37 such bricks were identified, all belonging to the no. 1 fabric type. Stratigraphically they were all found in layers, which filled the hypocaust after it ceased to function, therefore they might have been part of the suspensura and/or other collapsed architectural elements used to fill the hypocaust and level the surface (Bost 2016: 427). Almost all pieces have traces of lime mortar on both faces, which could also mean they might have been used to build an arch (Figure 9.10). One of the pieces was intentionally broken in half and it has 3.8 cm of lime mortar set on the back. Most probably it was used as a bench plate set against the wall. Another brick has a textile impression (balanced tabby – 1/1 – the spinning direction of the threads was not identified) on one of the edges. Most probable it is a part of a clothing (tunica) of the person who handled the brick after it was modelled, but before it was burnt 15 (Figure 10).

As a subgroup of the rectangular types, the longrectangular should be introduced, with predominant coarse fabrics nos. 4 and 5, but also some examples of (semi)fine fabrics like nos. 2, 6, 9 and 10. The smoothing traces go either down on the long side, or on the short length. There are 49 examples of two parallel smoothening striations on the short length. It is not known for now if these striations have a functional purpose or if it is a production mark. The same pattern was also observed in some full fragments of the massive brick form presented above. Other particular marks worth mentioning are three parallel finger traces on the upper surface, some placed in a corner, some going from a corner in a curved line to the opposite corner of the item. (Figure 8.6-7, Figure 9.8) Animal and bird prints were also registered, which should be a subject for future analyses in cooperation with archaeozoologists. Similar to the case of the above-described massive bricks, the long-rectangular ones were also uncovered from layers containing collapsed architectural features, or the fillings and levelling of the hypocaust area. As previously mentioned, the different fabrics might indicate reparations of the building complex and the variation of the sources for material supply. The chronological framework must also play an important role, especially for the fabric variations.

Same construction technique can be found in Vitr. 2. 3, 4. I would like to thank Judit Pásztókai-Szeőke for providing this information. 14  15 

133

Alexandra Dolea

Figure 8. Rectangular bricks (photos and drawings: S. Gür, digitization: L. Karkouki © A. Dolea).

134

Brick by brick, tile by tile. The ceramic building material in Labraunda, Turkey

Figure 9. Rectangular, square and trapezoidal bricks (photos and drawings: S. Gür, digitization: L. Karkouki © A. Dolea).

135

Alexandra Dolea

Figure 10. Textile impression on a trapezoidal brick (photo: C. Georgescu).

and correlated to the stratigraphy. A parallel can be traced with a recent analysis on the wall decoration materials and techniques in Labraunda (Freccero 2018: 293-297), when differences in the technology used for wall decoration were observed throughout the Hellenistic, Roman and Late Antique periods. The first era is characterized by excellent materials and technique, the second showing a decrease in attention for details, while the Late Antique decoration is defined as ‘comparatively rough and low-class standard’. Similarities can be drawn between these observations and the fabrics’ variations for the ceramic building materials along the same periods of time.

Concluding remarks Concerning the production of the ceramic construction material no rejected piece or fragments with serious defects of fabrication which could indicate a local production were observed. Despite this, one might suggest a localregional production, especially since two presumably ceramic kilns were identified within and in the territory of Labraunda (Blid 2016: 42). To prove or eventually disprove this hypothesis, petrographic analysis are required both for pottery and ceramic building material in order to gain a better understanding of the production, clay sources and the development along the centuries. Nevertheless, the intense construction activities using ceramic building material, especially during the Late Antique period, might indicate the necessity for production in the near vicinity of Labraunda. This fact is also suggested by the discovery of several lime kilns on the site (Blid 2016: 26, n. 83).

Both analysed sectors have their own particularities and show how similar material had a wide range of functions, like the variety of bricks used for floors, walls, arches, hypocaust, benches etc. In addition, it helps determining different constructive phases, reparations, variations in production, aesthetical integration and adaptation.

A hypothesis was developed based on the fabric variations of the Water Pool as well as the East Bath material. Accordingly, the (semi) fine fabrics could be connected to the Hellenistic and Roman times, while the (semi) coarse fabrics might be related to later interventions in constructions, namely Late Antique and Byzantine eras.16 Nevertheless, this material tends to be deposited, reused and/or replaced constantly, so part of its utilization should be carefully interpreted, and rather included within the general material analysis 16 

Although the study of ceramic building material attracts more and more specialists, its investigation should be included in the systematic material analysis of the ancient excavation sites as it is “the commonest of all Roman remains” (Brodribb 1987: 4), it complements the pottery studies, the general architectural features and offers important indications regarding material production, distribution and (re)use.

For similar results, see Stefanidou et al. 2015: 2257, 2259.

136

Brick by brick, tile by tile. The ceramic building material in Labraunda, Turkey

Fabric 7 – Semi fine fabric with small size calcite inclusions (~5% in the profile), small and medium size holes, silver and rare golden mica. (2.5YR, 6/6 – Light red)

Appendix: Fabric samples for ceramic construction material All the fabrics were described macroscopically, based on the selected samples for determining the main fabrics of the material. There are certain variations of colour within the chosen fragments, but the composition of the material as a main feature was taken into consideration to establish the fabric for the analysed pieces.17 In addition, microscopic photos of all the fabrics were performed (Figure 11).18

Fabric 8 – Coarse fabric with dense small and medium size calcite inclusions (~15% in the profile), small and medium size holes, rich in silver and golden mica.(5YR, 6/6 – Reddish yellow) Fabric 9 – Fine fabric with small calcite inclusions (~7% in the profile), some of them probably with organic origin (the white, long and regular ones indicated also in the microscope photos); small size holes, silver and golden mica. (5YR, 5/6 – Yellowish red)

Fabric 1 – Compact, semi fine fabric with very small calcite inclusions (~7% in the profile), small and medium size holes, silver and golden mica. (5YR, 6/4 – Light reddish brown)19 Fabric 2 – Compact, fine fabric with few, small calcite inclusions (~2% in the profile), small size holes, silver and golden mica. (10YR, 6/2 – Light brownish gray)

Fabric 10 – Fine fabric with specific composition due to the ochre iron oxides. It also contains small and rare calcite inclusions (~1% in the profile); also small size holes and rare, fine silver mica. (10YR, 6/4 – Light yellowish brown)

Fabric 3 – Compact, fine fabric with few, small black inclusions (~5% in the profile), very rare and small calcite inclusions (~2% in the profile), small size holes, very rare silver/golden mica. (5YR, 6/4 – Light reddish brown) Fabric 4 – Coarse fabric with dense inclusions: small and medium size calcite (~30% in the profile), a big black inclusion (0.9x0.6 cm) which was most probably a big concentration of aluminosilicate minerals; this was exposed to high temperatures and changed its structure to one more close to a vitrified ceramic. Small and medium size holes, very fine silver mica. (2.5 YR, 5/6 – Red) Fabric 5 – Coarse fabric with small and medium size calcite inclusions (~7% in the profile), also small fragments of grinded pottery (~1% in the profile), small and medium size holes, dense silver mica. The colour of the fabric varies because of the reduction firing: to the edges – 5YR, 7/4 – Pink, more to the core – 10YR, 7/4 – Very pale brown, the core itself – 10YR, 5/1 – Gray. Fabric 6 – Semi fine fabric with dense, fine calcite inclusions (~40% in the profile), small size holes, and silver mica. (7.5YR, 6.3 – Light brown) Archaeometric analyses were planned for refining the data regarding the fabrics, but up to this moment, no natural-scientific approach was initiated. Therefore, the current state of research presented in this article is the most advanced one presented to the public. 18  The microscope used for documenting the brick fabrics is Optika Vision Lite 2.0, while the photo dimensions are 1600 x 1200. C. Georgescu performed this work in Labraunda. 19  Munsell 2000 was used for the colour codes and inclusion percentages. 17 

137

Alexandra Dolea

Figure 11. Microscopic photos of fabrics (photos: C. Georgescu; © A. Dolea).

138

Brick by brick, tile by tile. The ceramic building material in Labraunda, Turkey

Congress on Construction History, Brandenburg University of Technology Cottbus, Germany, 20th-24th May 2009, Volume 2: 721‒728. Berlin: NEUNPLUS1. Henry, O., Bilgin Altınöz, A. G., Blid, J., Çakmaklı, Ö. D., Dufton, A., Freccero, A., Gosner, L., Hedlund, R., Lebouteiller, P., Lungu, V., Rojas, F., Tobın, F., Vergnaud, B., Waters, A. 2014. La mission Labraunda 2013 – Rapport preliminaire. Anatolia Antiqua XXII: 255‒325. Henry, O., Andersson, E., Bost, Chr., Çakmaklı, Ö. D., Commito, A., Cormier-Huguet, M., De Staebler, P., Dupont, P., Ergenç, D., Frejman, A., Kepenek, B., Lebouteiller, P., Nilsson, H., Rojas, F., Vergnaud, B. 2015. Labraunda 2014. Anatolia Antiqua XXIII: 301‒394. Henry, O., Anderson, E., Bost, Chr., Çakmaklı, Ö., Cederling, F., Commito, A., Cormier-Huguet, M., Coutelas, A., Dolea, A., Ergenç, D., Freccero, A., Frejman, A., Lebouteiller, P., Lesguer, F., Löwenborg, D., Lungu, V., Marchand-Beaulieu, Fr., Sitz, A., De Staebler, P., Vergnaud, B. 2016. Labraunda 2015. Anatolia Antiqua XXIII: 339‒457. Henry, O., Andersson, E., Blid, J., Bost, Chr., Çakmaklı, Ö., Carless-Unwin, N., Çimen, G., Eyigör, A., Freccero, A., Frejman, A., Georgescu, Cr., Goussard, E., GuimierSorbets, A.-M., Hauchart, M., Hedlung, R., Lamare, N., Lungu, V., Marchand-Beaulieu, Fr., Sitz, A., Stojanovic, I., Vergnaud, B. 2018. Labraunda 2017. Anatolia Antiqua XXVI: 209‒320. Karlsson, L. 2010. Labraunda 2009. A preliminary report on the Swedish excavations with contributions by Jesper Blid and Olivier Henry. Opuscula 3: 61‒104. Lesguer, F. 2016. 5.6.2. La fouille du four tardif, in O. Henry et al. Labraunda 2015. Anatolia Antiqua XXIII: 441‒447. Lungu, V. 2018. 5.3. Etudes ceramiques, in O. Henry, et al. 2018. Labraunda 2017. Anatolia Antiqua XXVI: 271‒282. Stefanidou, M., I. Papayianni and V. Pachta 2015. Analysis and characterization of Roman and Byzantine fired bricks from Greece. Materials and Structures 48: 2251‒2260. Stevens, G. P. 1950. A Tile Standard in the Agora of Ancient Athens. Hesperia 19/3: 174‒188. Streinu, A. 2020. Thin-walled table ware from recent excavations in the East bath complex at Labraunda (Milas, Turkey). Rei Cretariae Romanae Fautorum Acta: 551‒558. Wikander, Ö. 1986. Acquarossa. Vol. VI. The roof-tiles. Part 1. Catalogue and architectural context. Stockholm: Paul Aströms. Wikander, Ö. 1993. Acquarossa. Vol. VI. The roof-tiles. Part. 2. Typology and technical features. Stockholm: Paul Aströms. Wulf, U. 1999. Die hellenistischen und römischen Wohnhäuser von Pergamon, Altertümer von Pergamon, Band 15. Die Stadtgrabungen, Teil 3. Berlin: de Gruyter.

Ancient sources Pliny the Elder, The Natural History, John Bostock, M.D., F.R.S.H.T. Riley, Esq., B.A. London, Taylor and Francis, Red Lion Court, Fleet Street, 1855 (for the English version) Vitruvius, De Architectura, Valentin Rose, in aedibus B.G. Teubneri, 1899 (for the Latin version). http:// penelope.uchicago.edu/Thayer/E/Roman/Texts/ Vitruvius/home.html (for the English translation) References Blid, J. 2016. Labraunda 4. Remains of Late Antiquity. Stockholm: Swedish Research Institute. Bost, C. 2015. 3.4. Les Bains est, in O. Henry et al. Labraunda 2014. Anatolia Antiqua XXIII: 355‒366. Bost, C. 2016. 5.4. Les Bains est, quelques notes à propos de la campagne 2015, in O. Henry et al. Labraunda 2015. Anatolia Antiqua XXIII: 424‒435. Brodribb, G. 1987. Roman Brick and Tile. Gloucester: Alan Sutton Publishing. Çakmaklı, Ö. D. 2014. 4.5. La zone méridionale du sanctuaire, in O. Henry, et al. 2014. La mission Labraunda 2013 – Rapport preliminaire. Anatolia Antiqua XXII: 255‒325. Çakmaklı, Ö. D. 2015. 3.6. Les bassins romains, in O. Henry, et al. 2015. Labraunda 2014. Anatolia Antiqua XXIII: 384‒388. Çakmaklı, Ö. D. 2017. Labraunda Su Kompleksi Kazıları ve İnce Cidarlı Seramiği (Ön Rapor). Pamukkale University Journal of Social Sciences Institute 28: 257‒269. Çakmaklı, Ö. D. 2018. 6.4 Water Pool, in O. Henry, et al. 2018. Labraunda 2017. Anatolia Antiqua XXVI: 316. Crampa, J. 1972. Labraunda: Swedish Excavations and Researches, Vol. III, Part 2: The Greek Inscriptions, Part II (Acta Instituti Atheniensis Regni Sueciae, Series in 4° V, III: 2): 13‒133. Stokholm: Paul Åström Publishers. Dolea, A. 2016. 4.4. TCA — Tiles and bricks, in O. Henry at al. Labraunda 2015. Anatolia Antiqua XXIII: 383‒387. Ergenç, D. 2015. 2.1.3. Mortiers hellénistiques et romains, in O. Henry, et al. 2015. Labraunda 2014. Anatolia Antiqua XXIII: 306‒308 Ergenç, D. 2016. 3.3. Mortiers romains, in O. Henry, et al. 2016. Labraunda 2015. Anatolia Antiqua XXIII: 353‒358. Excoffon, P. 2009. Eléments de couverture en tuiles sciées de type corinthien à Olbia de Provence (Hyères, Var). Documents d’archéologie méridionale 32: 255‒267. Freccero, A. 2018. Wall decoration materials and technique, in O. Henry et al. Labraunda 2017. Anatolia Antiqua XXVI: 293‒297. Graciani, A. 2009. Earthenware Pieces Manufactured for Roman Thermae, in K. E. Kurrer, W. Lorenz and V. Wertzk (eds) Proceedings of the Third International 139

Examples of ceramic building materials of Roman and late antique date from Cickini near Malinska on the island of Krk (NE Adriatic, Croatia) Ranko Starac Abstract The paper presents tile and brick finds from the late Antique residential complex with church located at the site Cickini in the internal part of Krk island (NE Adriatic, Croatia). A selection of early Roman, probably reused stamped roof tiles is dealt with in the first section of the paper, while the second one is devoted to the analysis of a corpus of rectangular bricks often bearing incised motives or drawings. Their possible use as wall covers is brought forth, while their probably local production is placed within a larger regional debate.

Key words Ceramic building materials, roof tiles, bricks, late Antiquity, early Christian, stamps, incised motives, Cickini, island of Krk.

Figure 1. Location of Krk island, Cickini site and the Roman towns on the island (author: R. Starac, basemape: Google maps/Snazzymaps).

Since 2002 systematic excavations and conservation of architectural remains of a late antique – early Christian complex at the Cickini site, near the settlement Sveti Vid – Miholjice in the Municipality Malinska-Dubašnica on the island of Krk (Figure 1), are carried out. At the site remains of an early Christian church of complex layout, with vestibule and baptistery with built piscina dated to the 6th century have been unearthed (Starac 2004; 2006). In the environs of the church the walls of a residential complex, built during the 5th, reworked in the 6th and abandoned after a fire in the 7th century are being excavated (Figure 2, 3). Within the complex, ceramic building materials, like large ceramic tiles,

rounded suspensurae bricks and fragments of roof tiles – tegulae and imbrices with stamps, from nearby early Roman sites were walled. About 200 m to the north of Cickini, at the position Banošice, the densely vegetated remains of a large Roman sites are located. Towards east, roughly at the same distance and near the settlement Maršići, the site Merine is located. At the outskirts of the settlement Sveti Vid-Miholjice, near the old road Omišalj – Krk, the Roman site Ogradi is located. From one of these neighbouring sites the small altar (ara) dedicated by Quintus Fonteius Maximus to goddess Diana must also have originated. The monument, datable on the basis of palaeography to the early Principate, was found walled

Roman Pottery and glass manufactures (Archaeopress 2022): 140–148

Examples of ceramic building materials of Roman and late antique date from Cickini

Figure 2. Aerial view of Cickini site in 2018 (photo: S. Ilić, PPMHP archives).

Figure 3. Layout of the site (2018) (PPMHP archives).

141

Ranko Starac

Figure 4. Tile with stamp DE SALTV SEX MTILLI MAX and detail of the stamp (photo: R. Starac).

in the jamb of the entrance corridor located in axis with the front of the late Antique salutatorium, that is, the hall with a seven-sided apse within the residential part of the site (Figure 3).

workshop, thus it might be a product of a local north Adriatic workshop. An interesting find is that of a tile bearing the stamp DE.SALT hedera/[---]X.(vel hedera)MTLI.MX in two separate cartouches, stemming from the workshop DE SALTV SEX MTILLI MAX (Figure 4). The workshop that produced tile with the stamp De Salt(u) Sex(ti) M(e/u)tilli Max(imi) was recently discovered in the coastal area close to the road station Ad Turres, that is in the city centre of the modern town of Crikvenica (Starac 1991; Lipovac Vrkljan and Ožanić Roguljić 2007). It was active between the second half of the 1st century BC and the end of the 2nd century AD. The products of this workshop are recently being discovered on the coast and on the islands of the area between Istria and central Dalmatia (Radić Rossi 2011). On our specimen the dividing sign hedera distinguens shaped as a leaf is visible.

In the first part of the paper we will focus on stamped roof tiles. They all belong to tiles produced within the beginning of the 1st and the end of the 2nd century AD, brought to the site from its environs and installed on the roofs of the church and other structures during late Antiquity. Following, square/rectangular bricks produced in a yet unidentified local pottery workshop at the end of Antiquity, and recovered in the destruction layers of the residential complex will be dealt with. During excavations at the site 11 fragments of stamped roof tiles have so far been recovered. Among these, the vast majority belongs to stamps of the PANSIANA workshop (Matijašić 1983; 1987; 1989). That is by far the most common stamp belonging to one of the workshops located to the south of the Po river delta, and distributed along the entire coastal part of the eastern Adriatic in the time-frame of the late 1st century BC — 1st century AD (Cipriano and Mazzocchin 2007). Two stamps from Cickini belong to the period when the workshop, once belonging to Caius Vibius Pansa, had already become state property, that is to the reign of Octavian Augusts, between 27 BC and 14 AD (Juras and Jurković Pešić 2016). Fragmentary stamps lacking the initial part of the cartouche cannot be securely dated, due to the missing names of emperors present on the later stamptypes (in fact, these stamps were common during a longer period, that is until the reign of Vespasian). One fragment of stamp with preserved the letters [---]SIAN ends with the representation of the curved shepherd’s stick, the lituus (Righini 1999). Among the others, an incomplete stamp PANSIAN. was impressed on a very low quality tile, which is not typical for the tiles of this

Two stamps, QCLODIAMBR[---] and [---]AMB[---] belong to the workshop of Quintus Clodius Ambrosius. These widespread products of the workshop located in the environs of Aquileia had been distributed along the Adriatic from the mid-1st to the beginning of the 2nd century AD (Cipriano and Mazzocchin 2007). To the area of Aquileia belongs the much rarer stamp L.MINICI.PVDENTIS (Lucius Minicius Pudentis), in use from the mid-1st to the 2nd century AD also recovered at Cickini. On a tile the beginning of a cartouche with the letters MO[---] is visible, that is the stamp MODESTI, derived from the cognomen Modestus (Figure 5). A rare stamp of a small scale producer, probably working in the area of southern Liburnia (Juras and Jurković Pešić 2016: 60–61; Pedišić and Podrug 2008: 103). From the opposite coast of the Adriatic a stamp MOTANI is known from 142

Examples of ceramic building materials of Roman and late antique date from Cickini

dated to the 6th century, was fully made with tegulae. At the site in the Cickini forest, prior to the 5th century only a small rectangular building of unknown function had existed, which was then incorporated within the much larger early Christian church. Therefore, completely preserved and centuries old roof tiles must have been carefully collected, cleaned of the mortar and transported from more or less distant, but already abandoned Roman villa sites and settlements. Behind the polygonal apse of the church a built tomb vaulted with tegulae is located, while in the vestibule, underneath the mortar floor, a child grave made with tegulae was excavated. Along with tegulae, hundreds of fragments of imbrices have been found. Their quality and pastes vary substantially, that is, they belong to several different workshops. Some are smaller and thinner than the earlier examples, thus a late antique date could be proposed for their production. On several upper surfaces of imbrices wavy lines which were impressed in the wet clay are visible. One imbrex has a longitudinal threefold wavy band. Within the residential complex at Cickini (Starac 2004) hundreds of fragments of rectangular or square bricks were recovered, all of coarser production in comparison to early Roman tiles. They were exclusively found in layers of collapse within and partly outside the residential complex, that is within rooms in the northern part of the complex, located by the corridor facing the entrance to the salutatorium (Figures 3, 6). Flooring of these small rooms was made with coarsely levelled mortar. On the stone walls, preserved to a height of 130 cm, only small quantities of binder were preserved. For centuries the whole complex was covered by a dense oak forest, whose roots were growing within the walls, ‘eating’ at the mortar. The mentioned bricks were all recovered fragmentary, that is no complete example survives and their original dimensions cannot be ascertained. Nevertheless, all are of the same width, 22 or 23 cm, with a preserved length of 28 cm. Thickness spans between 4,2 and 4,6 cm. The edges are straight, the upper surfaces are smoothed, while the lower ones are uneven and coarse, bearing traces of mortar. It would seem that these bricks were rectangular, but the existence of square-shaped ones cannot be excluded. Clay is of different shades, from light ocher, to dark red or dark brown, probably due to uneven firing, but also subsequent exposure to fire, as the whole complex was destroyed in a fire. It must be again stressed that nowhere within the church, its ancillary rooms, the baptistery and the residential complex have brick floors been found. These bricks were found in layers mixed with mortar lumps, stone blocks from the collapsed walls and numerous roof tiles from the collapsed roofs. Therefore, rectangular bricks could have originally been placed as wall covering, that is, they might have covered

Figure 5. Tile with partially preserved stamp MO[desti] (photo: R. Starac).

the environs of Padua, read by Tortelli 1978-1979 as M. Otani(?) but not found in the Civic Museum of Padua (Cipriano and Mazzocchin 2007: 674, tab. 11). Finally, the tegulae fragment bearing the cartouche with the stamp [---]LLONI should be mentioned, though its workshop could not be precisely identified. Based on data acquired through archaeological excavations at other late Antique-early Christian sites on the island of Krk (Bekić and Višnjić 2008: 209–257; Jakšić and Novak 2009: 403–410; Starac 1996: 137–141; 2004: 217–221; Čaušević-Bully 2012: 135–137; Čaušević-Bully and Bully 2020: 275–279), the reuse of roof tiles (tegulae and imbrices) (on their reuse: Mills 2013: 585; Turchiano and Volpe 2016: 110; Munro 2011: 76–88) produced within the first two centuries AD to cover the roofs of churches was a wide-spread phenomenon. Early Roman tegulae with stamps such as Pansiana where reused for roofing sacral complexes of the 5th and 6th century at Baška, Jurandvor, Draga Bašćanska, Stara Baška, Punat, Krk and others, all the way to the northernmost part of Krk, that is the area of Roman Fulfinum. Very common is also the erection of early Christian complexes within earlier residential-productive architecture, thus the recycling of roof tiles must have been fairly easy. At Baška, whole roof tiles were used to cover the floor of late antique vaulted tombs, while the vault of a grave located in front of the church of St. George near Punat, 143

Ranko Starac

Figure 6. View of the small rooms in the northern part of the residential complex during excavations (photo: R. Starac).

the internal faces of the walls of the small rooms in the northern part of the residential complex. The rough shaping of the bricks, shaped within rectangular moulds and then sun dried, indicate late Antique manufacture, within a workshop that furnished the complex at Cickini during its building phase at the end of the 5th or the beginning of the 6th century.

were in use, in some form, until the end of antiquity. One is located within Soline bay, not far from Čižići, in the vicinity of the church of St. Peter (Lipovac-Vrkljan, Starac 2007). The other one (identified on the bases of large lumps of overfired kiln parts and bricks recovered from the shallows), also dated to the Roman period, is located in St. Martin bay in Dubašnica. Both sites (one on the eastern and the other on the western shores of the island) are roughly at an equal distance from Cickini, which was located close to the road communication stretching in the central part of the island, and connecting its two urban centres – Curicum (modern Krk town) and Fulfinum (Sepen near Omišalj). Future research at Cickini might bring to light a brick kiln at this very site, as these products are not present at the two mentioned workshop sites. Similarly shaped and rough are the square bricks forming the suspensurae of the thermal complex with hypocaust at Stara Baška, at the site Kraj Mira. Bricks from the hypocaust at Stara Baška bear an X sign impressed on their upper surfaces with two fingers in the wet clay before firing.

Analogies to the 4,5 cm thick wall covering bricks are not present elsewhere on the island of Krk. We might suppose that creating some sort of insulation or protection from humidity and cold of the small rooms was necessary. On the other side of the Mediterranean, in northern Africa, modern Tunisia, similar square bricks have been uncovered as covers of the walls of luxurious bishop or landlord residences of the 5th-6th century (Bonifay 2004: 439-442). The question of the location of the workshop or kilnsite for the production of these bricks remains open. Examples of bricks from the bishops’ workshop of southern Italy, dating to the 6th century, are somewhat larger (Baldasarre 2009). On the island of Krk two pottery workshops have been located thanks to reconnaissance. But, without archaeological excavations we cannot ascertain their duration, but thanks to surface finds we can suppose that both sites

Among the hundreds of fragments of wall covering bricks from Cickini several bear incised drawings (made in the wet clay before drying and firing) 144

Examples of ceramic building materials of Roman and late antique date from Cickini

Plate 1. Rectangular or square bricks with incised wavy lines; framed sherd is out of scale (photo: R. Starac).

145

Ranko Starac

Plate 2. Rectangular or square bricks with incised figural representation of birds (photo: R. Starac).

146

Examples of ceramic building materials of Roman and late antique date from Cickini

(Plate 1, 2). There are two types of incisions: one comprises rows of single or multiple wavy lines made with the fingertips; the second comprises drawings made with a sharp object, probably an iron pointy tool. So far a dozen bricks with the upper surface covered with dense, intermingling wavy lines made with a pointy tool have been discovered. The waves resemble the turbulent sea, within which silhouettes of fish can also be discerned. This certainly Christian motive, was made by an anonymous potter in an idle moment. On several other fragments motives of birds (peacock) and other feathery creatures, whose precise identification is hindered by the fragmentation of the bricks, are incised (Plate 2). In any case, the use of wavy lines on certain bricks (we have mentioned that they were also applied to imbrices) along with the naïve representations of the paradise bird, is contemporary with the application of wavy lines on pots and backing lids of the 5th and 6th centuries. Whether the Christian motives were applied intentionally, as decoration to the internal wall faces, or were created unpurposefully during manufacture currently cannot be discerned with certainty (Ceolin 2017). The bricks were not used within the church and adjacent spaces devoted to liturgy and adults’ baptism, but only in the residential, or perhaps storage area. As we still do not know who was the initiator of the building of this complex at a remote location away from Roman settlements and the sea coast. It might have been a rich individual, a Church prelate or the bishop, who built here a residence away from possible dangers.

-%20I%20laterizi%20antichi%20ed%20il%20loro%20 commercio%20nell’area%20sanvitese.pdf > Cipriano, S. and S. Mazzochin 2007. Produzione e circolazione dei laterizi nel Veneto tra I sec. a.C. e II sec. d. C. Antichita Altoadriatiche 65: 633–686. Čaušević-Bully, M. 2012. L’architecture paléochrétienne dans le Kvarner: état des connaissances et nouvelles découvertes. Hortus Artium Medievalium 18/1: 133–142. Čaušević-Bully, M. and S. Bully 2020. Najnovija istraživanja ranokršćanskih ‘crkvenih sklopova’ Kvarnera: na razmeđi arhitekture i teritorija, in: M. Sanader, D. Tončinić, I. Kaić and V. Matijević (eds) Zbornik I. skupa hrvatske ranokršćanske arheologije (HRRANA), Zagreb, 15–17. ožujak 2018: 271–291. Zagreb: Arheološki zavod Odsjeka za arheologiju Filozofskog fakulteta Sveučilišta u Zagrebu. Jakšić N. and N. Novak 2009. Il battistero paleocristiano a Bescanuova (Baška) sull’isola di Veglia. Hortus Artium Medievalium 15: 403–410. Juras, I. and F. Jurković Pešić 2016.Tegule s pečatom iz antičke zbirke Arheološkoga muzeja u Zadru. Diadora 30: 31–76. Lipovac Vrkljan, G. and I. Ožanić Roguljić 2013. Distribucija crikveničke keramike kao prilog poznavanju rimskog gospodarstva. Senjski zbornik 40: 225–270. Lipovac Vrkljan, G. and R. Starac 2007. Soline – uvala Sv. Petra (otok Krk). Annales Instituti Archaeologici III: 97‒98. Matijašić, R. 1983. Cronografia dei bolli laterizi della figulina pansiana nelle regioni adriatiche. Mélanges de l’École française de Rome. Antiquité 95/2: 961–995. Matijašić, R. 1987. La produzione ed il commercio di tegole ad Aquileia. Antichità Altoadriatiche XXIX: 495–531. Matijašić, R. 1989. Rimske krovne opeke s radioničkim žigovima. Izdanja HAD-a 13: 221–235. Mills, P. 2013. The potential of ceramic building materials in understanding Late Antique archaeology, in L. Lavan and M. Mulryan (eds) Understanding Late Antique Archaeology. Field Methods and Post-Excavation Techniques in Late Antique Archaeology (Late antique archaeology 9): 573–594. Leiden – Boston: Brill. Munro, R. 2011. Approaching Architectural Recycling in Roman and Late Roman Villas, in D. Mladenovic and B. Russell (eds) TRAC 2010: Proceedings of the Twentieth Annual Theoretical Roman Archaeology Conference, Oxford 2010: 76–88. Oxford: Oxbow Books. Pedišić, I. and E. Podrug 2008. Antički opekarski pečati iz fundusa Muzeja grada Šibeenika, Opuscula archaeologica 31 (2007): 81–141. Radić Rossi, I. 2008. Brodski tereti krovne opeke i proizvodi radionice Seksta Metilija Maksima u jadranskom podmorju, in G. Lipovac Vrkljan, I. Radić Rossi and B. Šiljeg (eds) Rimske keramičarske i staklarske radionice proizvodnja i trgovina na jadranskom prostoru Zbornik I. međunarodnog arheološkog kolokvija

Late Antique bricks (sometimes bearing incisions) used as wall covers are so far an isolated occurrence. Due to a total lack of written sources, future research at the already identified pottery workshops on Krk island and within ascertained late Antique remains dotting the island, might provide some answers. Translated by Ana Konestra (Institute of Archaeology, Zagreb, Croatia) References Baldasarre, G. 2009. Produzione ed impiego del laterizio nella Puglia centro-settentrionale e nella lucania nord-orientale fra tarda antichita e medioevo. Archeologia dell Architettura XII: 57‒75. Bekić, L. and J. Višnjić 2008. Južni dio antičke nekropole na položaju Sv. Marko – Baška. Vjesnik Arheološkog muzeja u Zagrebu 3.s./41: 209–257. Bonifay, M. 2004. Etudes sur la céramique romaine tardive d’Afrique (BAR International Series1301). Oxford: Archaeopress. Ceolin, P. 2017. I laterizi antichi ed il loro commercio nell’area Sanvitese. Star Light Editions, ww.antiqua.org < http://www.antiqva.org/Materiale/P.%20Ceolin%20 147

Ranko Starac Crikvenica, 23.-24. listopada 2008 / Roman Ceramic and Glass Manufactures Production and Trade in the Adriatic Region Proceedings of the 1st International Archaeological Colloquium, Crikvenica (Croatia), 23-24 October 2008: 19‒31. Crikvenica: Muzej Grada Crikvenice, Institut za arheologiju. Righini, V. 1999. I bolli laterizi di eta romana nella Cispadana. Le Figlinae, in V. Righini (ed.) Le Fornaci romane. Produzione di anfore e laterizi con marchi di fabbrica nella Cispadana orientale e nell’Alto Adriatico (Atti delle Giornate Internazionali di Studi, Rimini 1993): 29–68. Rimini: Musei comunali. Starac, R. 1991. Antička keramika sa lokaliteta ‘Igralište’ u Crikvenici. Vinodolski zbornik 6: 221–235.

Starac, R. 1996. Sulla scoperta di un’altra chiesa paleocristiana nell’isola di Krk (Veglia). Hortus Artium Medievalium 2: 137–141. Starac, R. 2004. Ranokršćanska crkva u šumi ‘Cickini’ kod Sršići na otoku Krku. Hortus artium medievalium 10: 217–221. Starac, R. 2006. Ranokršćanski crkveni kompleks u šumi Cickini u Malinskoj na otoku Krku. Malinska – Dubašnica: Glosa. Turchiano, M. and G. Volpe 2016. Faragola e l’eredità delle ville in italia meridionale tra tardoantico e altomedioevo. Anales de Arqueología Cordobesa 27: 77–96.

148

Stamps on ceramic and glass finds from the Velika Mrdakovica site near Šibenik Toni Brajković Abstract The paper presents a series of stamped archaeological materials barsming from the excavations of the settlement and necropolises of the hillfort of Velika Mrdakovica in Šibenik’s hinterland. Tegulae, amphorae stoppers, lamps, pottery and a glass vessel with inscriptions are presented through a catalogue and discussed in the context of their findspot and their analogies in the Adriatic and beyond.

Key words Tegulae, Pansiana, oil lamps, amphorae stoppers, stamps, Velika Mrdakovica.

Figure 1. Location of Velika Mrdakovica (position marking: T. Brajković; basemap: GoogleEarth).

Introduction The settlement at Velika Mrdakovica (Brajković 2014: 5-7; 2018: 32-33, n. 4, 35-36, 46) (Figure 1), located on the homonymous hill at an height of 141.6 m ASL, is of the hillfort type and occupies the upper third of the hill (Figure 2). The ramparts that encircle the settlement

are only partly visible and excavated or defined on its south-eastern, southern and western side, for a total length of over 100 m, although on its northern side they can be supposed with high probability. Iron Age ramparts, built with irregular stone blocks of various dimensions, placed without mortar, are visible on the southern slope, flaking from one side the SI entrance

roman pottery and glass manufactures (Archaeopress 2022): 149–171

Toni Brajković

Figure 2. Aerial image of the excavated area of the settlement on Velika Mrdakovica (photo: Arheo ko-op).

to the settlement. The other gate was probably located on the north-western side of Velika Mrdakovica, which is discernible from the detected Prehistoric and Roman path stretching below the SW and W ramparts at a distance of c. 10 m, and entering the settlement at the aforementioned position. In the area between the ramparts and the path several stone structures have been detected, which might be remains of defensive structures, though without excavations this remains just a possibility.

Velika Mrdakovica. They could have also been used for flooring or levelling of the wall’s faces, but one use does not exclude the other. Wood used for roofing – pine and evergreen oak – is present in the area today as well, although no traces were found during excavations. What was found, though, are iron nails for rafters and lookouts, and a sizable amount of tegualae and fragments of white ceiling mortar with traces of wattle on their back sides. Slightly above the walking surfaces created by flattening of the bedrock and filled with clay and debris, layers containing the most concentration of tegulae and amphorae fragments, just as other everyday objects were uncovered. These were located below a layer of charcoal – remains of a fire – which must have destroyed, at some point, at least a part of the settlement. This has created, along with erosion due to the steepness of the slopes and anthropogenic devastation of architectural remains, thick layers of rubble, whose weight crashed the archaeological material located below.

The settlement developed on terraces with a so far established elevation discrepancy of c. 20 m. Such a layout is defined by the configuration of the bedrock, which runs in parallel with an east — west direction with deviation towards north, and which was used both as source of building material and as a construction element of rooms and communications (Figure 2, 3). Particularly interesting are the vertical segments of bedrock which constitute the SE walls of the rooms. These were chiselled as vertically as possible, while their natural cracks were filled with smaller stones. In the case of two rooms – P 9 and P 12, the bedrock was shaped so to form a notch that probably housed wooden beams for the gable roof construction. Before the use of tegulae, the roofs could have been covered with thin stone slabs, as is the case with traditional Dalmatian houses. Such houses are present at Srdarići, on the SW slopes of Velika Mrdakovica, and there is no doubt that similar building techniques, conditioned by climate, vegetation and available natural resources, were in use from ancient times. This is further confirmed by finds of such stone slabs within all the rooms excavated at

Excavations yielded mostly materials of late Hellenistic and early Imperial date, although the aforementioned Iron Age ramparts and the dating of the related necropolis suggest that the beginning of settlement at Velika Mrdakovica started within the 8th - 7th centuries AD.1 The most chronologically sensitive artefacts, apart from coins, are those bearing stamps. They mostly Beginning of use of the necropolis is dated some centuries earlier, but such dates have been moved due to new comparative analyses of the materials and new insights, see Blečić Kavur and Podrug 2014: 88. 1 

150

Stamps on ceramic and glass finds from the Velika Mrdakovicasite near Šibenik

132.96

133.43 133.48

131.61

132.68

131.89 131.59

131.36

131.21

132.64

132.50

131.06 132.15

132.09

131.02 131.71

131.52 130.97

132.25 130.08

129.68

131.67

129.87

132.02

130.41

131.72

131.34 130.60

130.66 130.37

132.25

131.29

130.92 131.25

129.99

131.40 131.47

131.00

130.47 130.06

130.01

131.03

131.05

130.07

130.79

130.34

128.81 129.67

128.33

129.36

129.03 129.17

128.90

128.52

129.11

128.06 128.83

128.28 128.60

128.15

128.64

128.27 128.17

128.96

128.80 128.80 128.68

127.64

126.75 127.35 128.24

128.36 127.21

127.78

127.49 128.28

127.46

127.70

126.70

126.95

127.04

126.93 125.56

126.85

126. 126.61

127.12

125.69

126.24

126.56 126.98

126.46

125.90

PN 04

PN 05

125.47

126.58

126.77

126.20

126.20

125.79

126.46

126.32

125.56

125.63

125.99

126.14 126.61

PN 24

125.78

124.63

125.20

125.20

126.54

125.72

125.75

126.09

124.18

125.38

125.90

123.91

125.05

PN 35

125.64

125.91

125.96

123.86

PN 09

124.74

PN 20

125.69

124.91

124.20

PN 86 124.21

125.52

124.79

124.31

PN 66

125.56

PN 62 PN 65

124.39

122.95

PN 32 PN 31

PN 08 PN 11

122.59

123.44

122.80 123.97

PN 59

PN 71 PN 70

PN 51

123.00

PN 56 PN 61 PN 79

123.74 PN 48

PN 84

PN 52

PN 10

PN 69 PN 68

PN 60

PN 78

PN 85

123.69

PN 73

PN 81 PN 64

PN 82 PN 67 PN 72 PN 74 PN 80

124.66 123.31

122.85

123.99

PN 63 PN 58

124.00

122.97

123.94

122.99

123.79

PN 23 124.07

PN 34 124.60

PN 57

PN 13

PN 18 125.13

123.60

PN 75

PN 21

124.81

PN 76

PN 50 PN 49 PN 53

123.76 123.84

122.70

122.78

PN 54 PN 55 122.63

122.51

132.05

120.85

121.94

121.88

121.75

120.12

121.72

120.06 119.22

118.95 119.63

119.28

Figure 3. Plan of the excavated area of the settlement on Velika Mrdakovica (drawing: Vinko Madiraca, Arheo ko-op).

belong to roof tiles, but also oil lamps, amphorae stoppers and utilitarian vessels (glass and pottery).

Emil Podrug. Among the 119 stamps from the Šibenik – Skradin area only one was from Velika Mrdakovica (Pedišić and Podrug 2007 (2008): 92, fig. 3, 93, 128, cat. no. 106). This was due to relative paucity of research within the settlement, which was excavated in short campaigns during 1971 and 1972, and the overall lack of sensibility for the significance and value of ceramic building materials. With recent sybarsatic excavations of the settlement, started in 2014, the amounts of such materials rapidly grew, and since excavations are still ongoing it is due to further progress. Along with materials from Velika Mrdakovica settlement and necropolis, a sample from micro location Hrdanovica was also included. This toponym, which is found in cadastral books is the skewed name of Mrdakovica which developed over time, and indicated the fertile field below the hillfort, settled after the abandonment of the settlement on the hill after the pax Romana. Moreover, to the list of tile stamps two examples which, under unfortunate circumstance, ended up in the monastery collection at Visovac and in the private house of pastor Živko Maretić were added.

Discussion Out of the 33 stamped tegulae most belong to the generally most common workshop – the northern Italian Pansiana. Two stamps bars from the necropolis.2 One was used as cover of a Roman incineration grave (grave 102), while the second is mentioned without context by Krsto Stošić in his publication Selima šibenskog kotara (1941: 136). All the others bars from the hillfort, where they are present in all the rooms, belonging to a wide range of known and thus far unregistered types of Pansiana stamps. A first detailed overview of tegulae stamps from the Šibenik City Museum was published by Ivan Pedišić i For the necropolis see Brajković 2014: 8-35; 2014a: 357-367; 2018. For a full bibliography on earlier research (1969. – 1974.) see Brajković 2018: 33, n. 4. 2 

151

Toni Brajković The Pansiana workshop was established in the environs of Voghiera in the mid- 1st century BC as the private production centre of Caius Vibius Pansa. Shortly after establishment the workshop passed to the hands of the Emperors who, starting with Augustus, maintained the cognomen of its founder on the stamps, gradually adding their own markings (Matijašić 1983: 962964; Righini (ed.) 1998: 32-34, 49-51; Cipriano and Mazzocchin 2003: 50-51; Pedišić and Podrug 2007 (2008): 88-89; Pellicioni 2012: 35, 72-76, 87-88; Vukov 2016: 157). Such addition firstly consisted of signs at the end of the text, and later included additions to the text itself, containing abbreviations of the Emperors’ names. Orthographic and palaeographic differences form the bases of the typology of Pansiana stamps which, though being detailed, does not include all types and sub-types, due to a large number of known and newly established variants. Robert Matijašić proposed a subdivision in nine types and 84 variant (Matijašić 1983), Valeria Righini defined 18 types with 71 variant (Righini 1998 (ed.): 45-49), which is a reduction of a former subdivision (Righini et al: 1993: 42-44, 48-55.), Maria Teresa Pellicioni added seven variants to the typology by V. Righini (Pellicioni 2012: 42-43, 50-54, 96-99), while Mirna Vukov recently added further 13 types (Vukov 2016: 160-173). The stamps often comprise symbols placed along with the text – laurel wreath, caduceus, palmette, lituus etc. – having possible apotropaic, but certainly propagandistic meaning. All invoke glory, victory, abundance and prosperity – Roman virtues and features which are ultimately embodied in the person of the Emperor. Along with the many symbols from the Pansiana stamps, the lituus stands out – symbol of the augur, used for the first time as symbol of secular and spiritual authority on coins minted by Lucius Cornelius Sulla, and retaken by Caius Iulius Cesar, just as at the end of the Republic and during the early Principate by the Emperors (Rocco 2009: 739-740). The use of the lituus on Pansiana stamps bears the same significance as on coins. But why the lituus? Augur’s authority comprised a wider spectre of everyday life, both private and public. Bearers of the augurate, deemed by Cicero as the biggest and most sublime honour in the Republic, could forbid, postpone or allow any magistrate, just as any public gathering, Roman assembly or regular activity. All efforts of traditional Rome were directed towards the establishment of consent with the divine, defined in the form of pax deorum. To achieve this, it was necessary to recognise and interpret the will of the goods, especially Jupiter, whose consent was necessary and without which nothing could be initiated or decided. Thus, augurs, selected exclusively within members of the aristocracy, embody the values and powers evoking the antiquity and origin of the state, while their jurisdiction and right guarantee its wellbeing, just as the wellbeing of the society (Linderski 1986: 330, 332-333).

Most of the stamps from Velika Mrdakovica are not preserved whole, but thanks to orthographic and palaeographic details of the inscriptions, use of symbols, mainly the lituus, but other as well, and their dimensions, the following can be concluded – 18 stamps belong to the late Republican and Augustan periods, to Augustus or Tiberius two stamps can ascribed, to Claudius and Nero one, and to Caligula six; two more stamps are date up to 68 AD, while one is not datable (Plate 1). It is noticeable that most of the stamps belong to the early period (cat. nos 3-19).3 The list could be longer but fragmentation hinders the definition of all stamps. Stamps that can be defined within several close variants have been noted as such, while for those whose elements are insufficient for precise determination only definable dating and époque have been noted. To the variant Matijašić type III/6, Righini type 2c = Pellicioni type 2c cat. nos 3 – 54 have been ascribed; to the variant Matijašić type III/3 (without lituus), Righini type 2h = Pellicioni type 2h (without lituus), Righini type 2i = Pellicioni type 2i (with lituus) cat. no. 8 was ascribed; to the variant Matijašić type III/5, Righini type 2m = Pellicioni type 2m cat. no. 11 was ascribed, while cat. no. 6, due to damage to the stamp, can be ascribed to Matijašić type III/1 or III/6, Righini types 2a to 2c = Pellicioni types 2a to 2c. Within this type two variants so far unique for Velika Mrdakovica stand out: the previously published stamp with undefined symbol at the end of the text (cat. no. 9), to which another stamp can be added, which though being fragmentary, present the same orthographic elements and symbol (cat. no. 10). A stamp with same textual features, but with a different symbol (cat. no. 7), has recently been detected within the holdings of the Archaeological Museum in Split, and classified as type 2 h2 (Vukov 2016: 161-162, 168, no. 3). Interesting and unknown is a Pansiana variant with the text stamped within two concave rectangular cartouches, with the texts complementing each other (cat. no. 19). Within the earlier group, ending with Tiberius, two stamps can be added (cat. no. 20, 21), although due to its preserved dimension cat. no. 20 could easily belong to the Augustan era. Often, preserved parts of the stamp are not enough for closer dating, thus stamps with sufficiently preserved text or with the lituus, which was in use on the Pansiana stamps untill Caligula’s reign, where placed within a wide period ending within this Emperor’s reign (cat. no. 22-27).

To these the stamp mentioned by Krsto Stošić is added. Cat. no. 5 was ascribed to this type on the basis of palaeographic features. 3  4 

152

Stamps on ceramic and glass finds from the Velika Mrdakovicasite near Šibenik

Plate 1. Tegulae stamps from Velika Mrdakovica (author: T. Brajković).

153

Toni Brajković To the reign of Nero belongs the stamp from Hrdanovica, enclosed within the wider settlement agglomeration (cat. no. 32), while the second stamp, presenting a series of ligatures, could be earlier, i.e. belong to Claudius’ reign (cat. no. 31).

route crossing between the islands off of Šibenik, taking the Krka river towards Scardona, the main supply centre for the military camp at Burnum and the largest Liburnian community of the area – Varvaria. Along this river highway, where one of the first stops was Velika Mrdakovica, several shipwrecks and landings have been pinpointed, with numerous types of amphorae, such as Lamboglia 2, Forlimpopoli, Porto Recanati, Dressel 28, Rhodian type, Egyptian amphorae, Africana II, Africana III and Spatheion 1A, which draw a picture of intense commercial activities throughout Antiquity (Koncani Uhač 2007: 364-370). GraecoItalic, Lamboglia 2, Dressel 6a just as those of Gaulish production follow continuously the consumption of wine at Velika Mrdakovica. We should stress, however, that a detailed research on all amphorae sherds from Velika Mrdakovica is to follow, and along with it a more comprehensive picture of these materials, which will certainly bring forth new data.

Particularly interesting is a stamp discovered to the east of Room 13 (cat. no. 1). Though presenting extensive damage the stamp can without doubt be read as C‧TROS[I].5 The stamps of the gens Trosia, dated to the turn of the eras, occur in three variants – only with the gentilicum (full or abbreviated) (Gomezel 1996: 46-47, 58, 68, 71, 83, 98) or with the addition of the first letter of the praenomen (Gomezel 1995: 51-52; 1996: 58, 81, 83, 98). So far three have been ascertained – M(arcus), P(ublius) and C(aius). ‘Our’ stamps belongs to the latter – Caius and was produced, as the other tiles of his gens, in the environs of Aquileia. Stamps of C(ai) Trosi have been found in the area of Aquileia, in Carlino, Muzzana del Turgnano and Rivignano. On the eastern Adriatic a stamp is known from the Museum in Poreč, but without indication of its findspot, thus the fragment from Mrdakovica is, so far, the only one with known location data (Gomezel 1995: 51; Gomezel 1996: 58, 81, 90-91.).

A culture of wine drinking was undoubtedly present, although wine had an important role not only in consumption, but also within cults and the associated rituals (Brajković 2018: 45-46; Giunio and Alihodžić 2019: 30-32, 34, 46, 80, 82, 90). Although amphorae are present only with sherds, they suffice to define their typologies, especially fragments of diagnostic parts of the vessels, such as rims, necks, bottoms, while we must not neglect a fair amount of various types of stoppers. They could be formed by chiselling stone slabs, cutting and reshaping other ceramic artefacts or be produced along with the amphorae with the aid of one or two part moulds in the form of flat discs. The largest number of flat stoppers was discovered within an undisturbed layer in Room 9a, for which a storage/cellar function was supposed on the bases of the associated finds. Unfortunately, this kind of artefacts has seldom attracted the attention of archaeologists, which is why they are rarely published. Apart from Carl Patch, who catalogued the stoppers from Narona in 1908 (1908: 93-94), more recently only a handful of authors tackled this topic. Zrinka Buljević published the disc stoppers from the Njive site in Narona’s environs (Buljević 1999: 204-209, 212, 223-235, 287-291), Ivan Šuta those from Sikuli (Šuta 2014: 81-98) while a few authors published, or a preparing, the material from Split - Kaštela - Trogir underwater sites (Lete 2005: 9, 12-28; Šuta 2012: 82) and Hutovo Blato near Čapljina in Bosnia and Hercegovina (Zmaić and Miholjek 2013: 176, 180, 182, 184).

The second stamp from Velika Mrdakovica (cat. no. 2) is identical in fabric and texture to the previous one, and it also lacks a cartouche, while featuring debossed letters of smaller dimensions. The text – [---] Sextii is preserved. As this praenomen, rendered in the genitive form, is fairly common (Gregorutti 2011: 348, 351, 365, 373-374, 385, 397; Lipovac Vrkljan 2011: 8, 11; Radić Rossi 2011: 19, 23-24, 28), but not as the last word in a sequence, it is impossible to attribute it with certainty to any known workshop. The closest parallel can be found in Aquileia, where tegulae stamped SEXTI‧ET‧ACVTI are dated in the same timeframe as cat. no. 1 (Gomezel 1996: 48, 98, 112), but we should not dismiss the possibility of its local production at the site Rakitnica - Tri Bunara, located just 2 km from Velika Mrdakovica (see Brajković 2011). In this case we would also be inclined to an earlier data, as from Augustan times the tegulae of the Pansiana workshop have fully saturated the market of Velika Mrdakovica. Along with sherd of ceramic building materials, large quantities of different types of amphorae have been uncovered throughout the settlement: GraecoItalic, Lamboglia 2, Dressel 6a (Righini 2004: 244-250; Cipriano and Mazzocchin 2012: 241-242, 251; Cipriano 2016: 147-153), Dressel 6b, Gaulish. These ancient transport vessels have often supplied, mostly with wine, the local market of Velika Mrdakovica, which was located on a route departing from the internal

Flat disc stoppers from Velika Mrdakovica ammount to 21, of which 14 were formed in a one-part mould, and seven in a two-part mould (Plate 2, 3). They are typical for Lamboglia 2 and Dressel 6a/6b amphorae, and it is impossible to precisely determine the associated vessel type (Chinelli 1994: 464-465; Buora 2014: 2526; Žerjal and Bekljanov Zidanšek 2014: 143), while their production spans from the late Republic to the

On this occasion I would like to thank the colleague Stefania Mazzocchin from the University of Padua for indicating me the correct transcription of this stamp. 5 

154

Stamps on ceramic and glass finds from the Velika Mrdakovicasite near Šibenik

Plate 2. Amphorae stoppers from Velika Mrdakovica (author: T. Brajković).

155

Toni Brajković end of the 1st century AD, being especially common in Augustan times (Chinelli 1994: 466; Buora 2014: 25, 30; Lodi 2015: 148; Cipriano and Mazzocchin 2016: 836). Seven stoppers present a square handle, while the others have a rounded, and one a conical handle (cat. no. 50). On the bases of their preservation and superficial damage it is impossible to assess all symbols present on stoppers from Velika Mrdakovica. Two might belong to onomastic ones (cat. nos 44, 47) as the letters V and C, and I and C are visible, which might be linked to the name ICVS from a stopper from Resnik (Lete 2005: 13). Fairly widespread anepigraphic signs, interpreted as numerals (Dobreva and Luise 2014, 80) like III, V, X, are present on three examples (cat. nos 42, 43, 46), although the sign V could correspond to a letter, which is known from stopper from Šibenik,6 Spain, Austria (Magdalensberg), Italy – at sites in Friuli Venezia Giulia (including Aquileia, Rivignano and Sevegliano), Fermo (in the Marche region), Greece (Delos) and Slovenia (Chinelli 1994: 466; Buora 2014: 27-29; Dolci 2014: 56, 61; Lodi 2015: 156; Žerjal and Bekljanov Zidanšek 2014: 151-153) as stand-alone or as a part of a more complex composition. An interesting stopper with several relief triangles (cat. no. 33) which could be interpreted as the Greek letter Δ, does not have analogies. Examples with two square decorations placed at either side of the handle (cat. nos 34, 40) are similar, though not identical to stoppers from Resnik (Šuta 2012: 85, 87, T. 3, 2), and to a stopper from the underwater site Smokvica Vela near Rogoznica (MGŠ AO 5094), while the most precise analogy is a stopper from Narona (Buljević 1999: 288, T. XXVI, 46). Close analogies to a stopper with four letters Y (cat. no. 41), of which one lacks the vertical line, thus resembling a V, are found at Resnik (Šuta 2012: 91-92, T. 6,2), and with a stopper having all the Y well shaped which was found either in Šibenik channel or in st. Ante channel.7 In the case of cat. no. 45 two linear markings are currently visible, though after cleaning they might lead to some other interpretation, while in the case of cat. no. 50, two circular and one linear relief marking form a triangular shape.

though production on the eastern Adriatic is not excluded, and were marketed locally or on the eastern Mediterranean (Righini 2004: 242-244; Bezeczky 2010: 352-353; Zmaić and Miholjek 2013: 179-180; Cipriano 2016: 146). Of these five, only two stamps are fully preserved, while the others are partial, and one is fully unreadable. The stamp with the name PRVX (cat. no. 55) is known from the port of Ancona and Vela Svitnja bay on Vis island, and is datable in the first half of the 1st century BC (Cambi 1989: 313; Buora et al. 2008: 290, 297; Marengo and Paci 2008, 324). The stamp TRVPO (cat. no. 56) was also recently identified at Ancona, and it is believed to be the Latin translation of the Greek name Tryphon (Marengo and Paci 2008: 318-319, 324). Four vertical lines preserved on the stamp cat. no. 57 could be identified as IIII VIR, found at Desilo – Hutovo Blato (Zmaić and Miholjek 2013: 180-181), or as [CRISPIN]ILI, present on a Dressel 6b amphora. The last part of the stamp [---]ELIO (cat. no. 58) could belong to one of the numerous gentilitia or cognomina, and in this form cannot be attributed to any of the so far known workshops. One of the most common finds from both the settlement and the necropolis are oil lamps. The light they casted had, apart from the usual function, an important role within beliefs in the afterlife (Crnobrnja 2006: 65-66; Abramić and Colnago 2009: 175; Brajković 2018: 58; Giunio and Alihodžić 2019: 75-76, 106-108). All of this defined these products as some of the most widespread items of human craft. Within the holdings of the Šibenik City Museum 154 lamps are held, 53 or 1/3 of which come from Velika Mrdakovica. A smaller amount belong to the Firmalampen type, having their core production centres in the area around Modena, with so far recognised branches at Aquileia, Trier and Lyon, and a series of provincial variants (Schneider 1993: 131133; Auer 2012: 13; 2016: 40; Rosenthal-Heginbottom 2015: 119, 129, 134; Labate 2016: 18). Preserved and recognisable stamps on lamps from Velika Mrdakovica amount to only 12 (Plate 3, 4). On one bottom sherd only the first letter of the stamp is visible – F (cat. no. 59), which precludes its attribution. On the other hand, the partially preserved stamp ---]ILI (cat. no. 60) almost certainly belongs to the widely distributed production of the STROBILI workshop, which was deemed to be the initiator of these lamp type (Harris 1980: 140-141; Krunić 2005: 86-87; Topoleanu 2010: 165; Labate 2016: 24, 30; 2017: 224-226) and whose products are present in the Šibenik area at the necropolis of Scardona. Nine lamps could be typologically determined. Five lamps with two decorative or functional buckles and closed channel, of the Fortis and Strobili workshops, one lamp with three hollow buckles and closed channel, of the Saturnini workshop, and the lamps with F[---] stamp belong to Loeschcke IX b type (cat. num, 59-65). Two lamps, of the Fortis and Litogene workshops, with two decorative buckles on the shoulder and the channel

On five amphorae rims stamps in rectangular cartouche have been identified (Plate 3). Rim shapes, i.e. their profiles, mostly indicate their belonging to Lamboglia 2 amphorae, deemed as the first Roman Adriatic wine amphorae, datable from the end of the 2nd century BC to the beginning of the 1st century AD. They were produced in an area spanning from Apulia to Aquiliea, Signs X and V are found on stoppers from the Underwater collection of the Šibenik City Museum. Their exact number with the related stamps will be published after a full revision of the material, which is currently located in several dislocated depots. 7  The latter was found during revision of the Antiquity collection, unfortunately without inventory number. No data on the circumstance and time of recovery are known, while the only information found alongside it was ‘channel’, which might refer to Šibenik channel in a narrower sense (that is between the islands of Zalarin and Prvić and the coast) or to St. Ante channel, which is its continuation. 6 

156

Stamps on ceramic and glass finds from the Velika Mrdakovicasite near Šibenik

Plate 3. Amphorae stoppers, amphorae stamps and lamps Velika Mrdakovica (author: T. Brajković).

157

Toni Brajković

Plate 4. Lamps, pottery and glass vessels from Velika Mrdakovica (author: T. Brajković).

connected to the disk belong to Loeschcke IX c type (cat. num 66-67), while the lamp, stamped Fortis, with three decorative buckles and the channel open on both sides belongs to Loeschcke X a type.8 It can be observed that most of the Firmalampen are stamped by the most prolific workshop – Fortis, whose production is located in northern Italy, i.e. in the territory of Modena (Mutina), where numerous mould and lamp sherds have been

located at Modena, Cittanova, Magreta and Savignano sul Panaro (Mardešić 2002: 354; Krunić 2005: 85; Labate 2016: 18; 2017: 221; 2017a: 17-18). This is confirmed by excavations at Savignano sul Panaro, which supports Pliny’s mentions about the size of pottery production at Modena (Topoleanu 2010, 161; Labate 2013: 33, 37; 2017: 221). In fact, so far 52 kilns producing pottery and lamps, and spanning from the Republican times to late Antiquity have been identified, while the ones belonging to the Fortis workshop were located either at Pratoguaratto property (Labate 2013: 34) or at Mutina itself (Labate 2016: 18, 24, 30; 2017: 225). All over Europe

Unfortunately this lamp could not be located during revision of the Antiquity collection holdings. All data is derived from the inventory book and drawing. 8 

158

Stamps on ceramic and glass finds from the Velika Mrdakovicasite near Šibenik

products of the Fortis workshops have been found, which is not surprising since its production spanned from the first decades/first half of the 1st century AD to the 3rd/4th centuries (Labate 2013: 37; Labate: 224, 226). During this period, along with the main production centre, numerous provincial branches were established to supply an ever-growing market, not excluding the possibility of a Dalmatia branch as well (Harris 1980: 141; Mardešić 2002: 354). Fortis lamps from Velika Mrdakovica belong to a series of variants based on palaeographic features and symbols accompanying the text. Most bear just the name of the workshop, though several examples boast differently shaped letters or present small circles or rosettes, just as an inverted letter C. All come from the necropolis and can be dated, based on the other associated finds, between the 1st and the 2nd century AD. The Neri workshop, active from Claudius to Alexander Severus, but producing most intensely within the 2nd century, should also be located in Modena’s environs or in central Italy (Mardešić 2002: 354). Its products are particularly widespread in Pannonia and Moesia Superior (Krunić 2005: 85, 94; Topoleanu 2010: 169). The lamp from Velika Mrdakovica was found within the disturbed grave NG 11 (cat. no. 68), along with an urn of local manufacture, a thin-walled ware vessel, a belly-shaped glass bottle, a glass modiolus, nails, and as terminus post quem non an as by Vespasian. The grave is dated at the end of the 1st and the beginning of the 2nd century AD (Brajković 2014: 28). Only one other example of a Neri lamp is present within the holdings of the Šibenik City Museum, found on Bribirska glavica, of smaller dimensions and belonging to the Loeschcke X a type. The only examples of the generally rare lamps of the probably northern Italian workshops of Litogene and Saturnini from the Šibenik area come from Velika Mrdakovica, where they were used as grave goods in graves 103 and NG 10 (Brajković 2014: 29-31). The stamps of Litogene, whose eastward distribution reaches Upper Moesia (Krunić 2005: 85, 94), and Saturnini, have been noticed in the large production centres of Modena, Tries, Lyon and Rome (Schneider 1993: 134-135, T. 1-2). On the bases of the accompanying local, north Italian and south Gaulish material the Saturnini lamps has been dated within the 1st century, while the Litogene lamp has been dated to the end of the 1st and the early 2nd century on the basis of typology. An interesting sherd with the stamp [---]RPI from the upper part of the settlement, Room A, should be attributed to the Eucarpi (cat. no. 69) workshop. Finds with this stamp span from Eburacum in England to southern Italy and Dacia, though most finds come from northern Italy and in particular western Germany and northern Switzerland. They differ typologically, while most recent research and clay analyses suggest that production began in Modena’s environs, to then shift northwards, towards the limes on the Rhine (Auer 2012: 12-13, 24; 2016: 39-41; Auer and Sitz 2014: 89, 92-99; Labate 2013: 37, 39; 2016: 30; 2017: 225).

Interesting finds bars from grave NG 3. That is, a small pottery vessel with an engraved graffito on the bottom (Plate 4, cat. no. 70) and two rectangular class jugs (Plate 4, cat. no. 71) (Brajković 2014: 12-14). The jugs have extremely thin walls, attached to the shoulder and the neck of the vessel knee shaped handles with three ribs, which create a decorative loop above a wide outward turned and rounded rim (Štefanac 2012: 42-44, 182, 187, 343, cat. no. 108-109; Brajković 2014: 12-14, fig. 7). Due to the fragility of the thinly blown bluish, very transparent glass the most fragile parts of the vessels, i.e. the bodies, where found broken, precluding determination of the height, though it certainly did not exceed 11 cm. On the outer part of the bottoms stamps were impressed. On the the smaller jug stamps were impressed on the flat bottom in the form of five symmetrical relief circles located in the corners and in the middle, while on the bigger jug, on a concave bottom, the stamp is formed by three concentric circles and smaller relief circles at each of the corners. Dating of these widely distributed types, classified as Isings f 50a, spans from the second half of the 1st century AD to the first half of the 2nd century AD (Isings 1957: 63-66). In the Liburnian area a total of 20 jugs with the same type of stamp have been identified – 18 from the necropolis of Iader and two from Velika Medakovica (Štefanac 2012: kat. br. 90-109; Fadić and Štefanac 2012: 216). Stamps with smaller circles placed in roughly the same position, but on larger jugs (Isings f 50b) are present at the Đardin – Maraguša early Principate necropolis of Scardona (Štefanac 2012: 42-44, 182, 187, 343, cat. no. 110-111). They are dated to the end of the 1st - early 2nd century AD, thus being somewhat older than the Mrdakovica examples (Isings 1957: 66-67). Since such stamps, regardless of the size of the vessels, are common in the eastern Mediterranean, their provenience from Syro-Palestinian workshops has been supposed, although very similar jugs, with somewhat different shapes and stamps where produced in a supposed local workshop of the Iader area (Fadić and Štefanac 2012: 216-217; Štefanac 2012: 44). Conclusion Within several publications on Velika Mrdakovica a large number of finds dating to the last centuries BC and the early Principate have been noted. The commercial backbone was certainly the river Krka which, after the foundation of the legionary camp at Burnum and the transformation of Scardona into main supply harbour, became the ‘highway’ of not only the wider scardonitan area but of a large part of the province of Dalmatia.9 It was exactly this river, with the longest navigable About the Krka river as a navigation route in Antiquity see Brajković 2018: 33, n. 3 9 

159

Toni Brajković course within the Adriatic basin, that allowed the inhabitants of Velika Mrdakovica to connect with the rest of the known world since prehistoric times. During that period centres of power and trade shifted, while from the mid- 1st century AD the primacy was taken over by workshops from central and northern Italy, mainly Arezzo, Modena, Aquileia, Voghera, and those located on the Adriatic, some of which might have been branches of larger production centres. Tegulae, oil lamps, amphorae and smaller utilitarian vessels from Velika Mrdakovica with their quantities and span fit within a wider picture of peri-Adriatic workshops and the network of trade of this part of the Mediterranean.

Stamp dim.: L 14,2 cm; W 3,35 Letter dim.: H 2,2 – 2,3 cm; W 1– 2,1 cm; T of bars 4 – 5 mm Litus dim.: H 2,4 cm; W 1,55 cm Material: ceramic, Munsell colour – 5YR 7/8 (reddish yellow) Desc.: Fully preserved stamp on large tegulae fragment. Letters with thick bars are impressed in shallow relief within a square cartouche. The right vertical line of the N is higher that the baseline, bending in the upper part towards the arch of the letter S. A litus is present at the end of the inscription. Transcription and restitution: PANSIANA; Pansiana Date: 43 BC – 14 AD Analogies: Matijašić type III/6; Righini type 2c = Pellicioni type 2c Unpublished

Catalogue Abbreviation: dim. = dimensions; L=lenght; W = width; H = height; T = thickness; D = diameter; Desc. = Description;

Cat. no. 4 Inv. no.: MGŠ 21683 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 19 cm; H 21,4 cm; T 3,1 cm Stamp dim.: L 14 cm; W 3,2 Letter dim.: H 2,1 – 2,3 cm; W 1,1 – 1,9 cm; T of bars 4 – 5,5 mm Litus dim.: H 2,1 cm; W 1,5 cm Material: ceramic, Munsell colour – 5YR 7/8 (reddish yellow) Desc.: Almost fully preserved stamp on tegulae fragment. Letters with thick bars are impressed in shallow relief within a square cartouche. The right vertical line of the N is higher that the baseline, bending in the upper part towards the arch of the letter S. A litus is present at the end of the inscription. Transcription and restitution: PANSIANA; Pansiana Date: 43 BC – 14 AD Analogies: Matijašić type III/6; Righini type 2c = Pellicioni type 2c Unpublished

Cat. no. 1 Inv. no.: MGŠ 21523 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 8,8 cm; W 12,2 cm; T 3,13 cm Stamp dim.: L 8,5 cm; W / cm Letter dim.: H 1,6 – 2 cm; W 1,4 – 1,7 cm, T of bars 2 – 2,5 mm Material: ceramic, Munsell colour – 2.5 Y 8/3 (pale yellow) Desc.: On a smaller sherd of a tegulae part of the stamp is preserved. Letters impressed without cartouche. Regardless of damage and wear the first four letters can be read as CTRO. The last, only partially preserved letter, arch-like elements indicate it was probably an S. Transcription and restitution: C‧TROS[.]; C(ai) Tros(i) Date: end of 1st c. BC – early 1st c. AD Unpublished Cat. no. 2 Inv. no.: MGŠ 21864 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 13,5 cm; H 15,2 cm; T 3,5 cm Stamp dim.: L 3,4 cm; W / ; H / Letter dim.: H 1 – 1,1 cm; W4 – 7 cm; T of bars 1 – 1,1 mm Material: ceramic, Munsell colour – 2.5 Y 8/3 (pale yellow) Desc.: Smaller stamp located centrally on the sherd. Letters are impressed without cartouche. Transcription and restitution: [---]SEXTI; [---] Sexti Date: end of 1st c. BC – early 1st c. AD Analogies: Gomezel group F Unpublished

Cat. no. 5 Inv. no.: MGŠ 21600 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 12,4 cm; W 8,4 cm; T 1,8 cm Stamp dim.: L 8,8 cm; W 3,3 cm Letter dim.: H 2,1 – 2,2 cm; W 1,1 – 1,9 cm; T of bars 3,5 – 4,5 mm Material: ceramic, Munsell colour – 5YR 6/8 (reddish yellow) Desc.: Almost fully preserved stamp on smaller tegulae fragment. Letters with thick bars are impressed in shallow relief within a square cartouche. The right vertical line of the N is higher that the baseline, bending in the upper part towards the arch of the letter S.

Cat. no. 3 Inv. no.: MGŠ 21524 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 29,8 cm; H 34 cm; T 3 cm 160

Stamps on ceramic and glass finds from the Velika Mrdakovicasite near Šibenik

Transcription and restitution: PANSIAN[.]; Pansiaṇ[a] Date: 43 BC – 14 AD Analogies: Matijašić type III/6; Righini type 2c = Pellicioni type 2c Unpublished

Analogies: Matijašić type III/3 (without litus); Righini type 2h = Pellicioni type 2h (without litus), Righini type 2i = Pellicioni tip 2i (with litus) Unpublished Cat. no. 9 Inv. no.: MGŠ 11683 Findspot: Velika Mrdakovica – necropolis, grave 102 Tegulae dim.: L 38,1 cm; H 35,2; T 2 cm Stamp dim.: L 14,3 cm; W 2,95 cm Letter dim.: H ; W 1,05 – 2,35 cm; T of bars 5 mm Material: ceramic, Munsell colour – 5YR 6/6 (reddish yellow) Symbol dim.: H 2,55 cm; W 1,3 cm Desc.: Fully preserved stamp on two joining fragments of tegulae. Letters are impressed in shallow relief within a square cartouche. The sixth and seventh letter – AN are in ligature. The inscription ends with an unclear sign. Transcription and restitution: PANSIANA; Pansia͡na Date: 43 BC – 14 AD Analogies: text as in: Matijašić type III/3; Righini type 2h = Pellicioni type 2h Published in: Pedišić and Podrug 2008: 92, fig. 3, 128, cat. n. 106; Vukov 2016: 161, fig. 1, 162.

Cat. no. 6 Inv. no.: MGŠ 1600 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 24,8 cm; H 20,7; T 3,7 cm Stamp dim.: L 15 cm; W 3,35 cm Letter dim.: H 2,4 – 2,5 cm; W 1,3 – 2 cm; T of bars 3,5 – 6 mm Material: ceramic, Munsell colour – 5YR 7/8 (reddish yellow) Desc.: Almost fully preserved stamp on tegulae fragment, beaten off at the ends. Letters are impressed in shallow relief within a square cartouche. Transcription and restitution: PANSIANA; Pansiana Date: 43 BC – 14 AD Analogies: Matijašić type III/1 or III/6; Righini type 2a – 2c = Pellicioni type 2a – 2c Unpublished Cat. no. 7 Inv. no.: / Findspot: Velika Mrdakovica – settlement Desc.: Fully preserved stamp, with smaller damage, on tegulae fragment. Letters are impressed in shallow relief within a square cartouche. The sixth and seventh letters – AN are in ligature. At the end an unclear sign. Stamp reading: Pansia͡na Date: 43 BC – 14 AD Analogies: Righini type 2h 2 = Pellicioni type 2h 2, see: Vukov 2016, 161-162, 168, br. 3 Unpublished

Cat. no. 10 Inv. no.: MGŠ 21527 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 15,4 cm; H 10,7; T 2,6 cm Stamp dim.: L 10,8 cm; W 2,96 cm Letter dim.: H 2,1 cm; W 1,6 – 2,2 cm; T of bars 4 mm Material: ceramic, Munsell colour – 2.5YR 7/8 (light red) Symbol dim.: H 2,45 cm; W 1,3 cm Desc.: On a fragment of tegulae the end part of the stamp is preserved. Shaped as a deeper (2,6 mm) square cartouche it has well defined letters impressed in shallow relief. The sixth and seventh letter – AN are in ligature. The inscription ends with an unclear sign. Transcription and restitution: [---]IANA; [---Pans] ia͡na Date: 43 BC – 14 AD Analogies: Pedišić and Podrug 2008: 92, fig. 3, 128, cat. no. 106; Vukov 2016: 161, fig. 1, 162. Unpublished

Cat. no. 8 Inv. no.: MGŠ 21882 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 14,5 cm; H 17,7; T 2,8 cm Stamp dim.: L 9,8 cm; W ca. 3 cm Letter dim.: H 2,1 – 2,3 cm; W 1 – 2,1 cm; T of bars 4 mm Material: ceramic, Munsell colour – 5YR 6/8 (reddish yellow) Desc.: Almost fully preserved stamp on smaller tegulae fragment. Letters are impressed in shallow relief within a deeper square cartouche (2,1 mm). Based on the first vertical line of the sixth letter, A, a ligature AN is supposed. Transcription and restitution: PANSIA[..]; Pansia͡n[a] Date: 43 BC – 14 AD

Cat. no. 11 Inv. no.: MGŠ 21870 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 14,8 cm; H 21 cm; T 3,02 cm Stamp dim.: L 9,9 cm; W 3 cm Letter dim.: H 1,9 – 2; W 1,15 – 2 cm; T of bars 2 – 4 mm Material: ceramic, Munsell colour – 5YR 7/8 (reddish yellow) Desc.: On a smaller fragment an almost full stamp is preserved. Letters are impressed in shallow relief, 161

Toni Brajković with thin bars within a deeper square cartouche (2,3 mm). Two ligatures of the letters – AN are noticable. Transcription and restitution: PANSIANA; Pa͡nsia͡na Date: 43 BC – 14 AD Analogies: Matijašić type III/5; Righini type 2m = Pellicioni type 2m Unpublished

Date: 43 BC – 14 AD Analogies: typologically hard to determine Unpublished Cat. no. 15 Inv. no.: MGŠ 1602 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 9 cm; H 10,2 cm; T 2,56 cm Stamp dim.: L 4,5 cm; W / cm Letter dim.: H /; W 2,1 cm; T of bars 5 mm Material: ceramic, Munsell colour – 2,5YR 6/8 (light red) Desc.: On a small sherd of tegulae initial part of a mostly damaged stamp within a square cartouche is preserved. Letters are incomplete and barely discernible. They are rendered in shallow relief. Transcription and restitution: PAN[---]; Pan[siana] Date: 43 BC – 14 AD Analogies: typologically hard to determine Unpublished

Cat. no. 12 Inv. no.: MGŠ 21525 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 28,5 cm; H 31 cm; T 3,34 cm Stamp dim.: L 9,4 cm; W 2,9 cm Letter dim.: H 2,1 – 2,25; W 1,1 – 1,7 cm; T of bars 6 – 6 mm Material: ceramic, Munsell colour – 2,5YR 6/8 (light red) Desc.: Larger sherd of tegulae with stamp. Letters with thick bars are impressed in relief within a square cartouche. Transcription and restitution: PANSIA[---]; Pansia[na] Date: 43 BC – 14 AD Analogies: typologically hard to determine Unpublished

Cat. no. 16 Inv. no.: MGŠ 1603 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 13,5 cm; H 9,2 cm; T 2,85 cm Stamp dim.: L 6,5 cm; W 3,6 cm Letter dim.: H ca. 2,1; W 1,4 – 1,5 cm; T of bars ca. 2,5 mm Material: ceramic, Munsell colour – 5 YR 7/4 (pink) Desc.: On a small sherd of tegulae covered with sediment the initial part of a damaged stamp is preserved. Letters are rendered in shallow relief with in a concave square cartouche. Transcription and restitution: PAN[---]; Pan[siana] Date: 43 BC – 14 AD Analogies: typologically hard to determine Unpublished

Cat. no. 13 Inv. no.: MGŠ 21604 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 18,5 cm; H 16,7 cm; T 3,15 cm Stamp dim.: L 10,6 cm; W 3,43 cm Letter dim.: H 2,2 – 2,4; W 1,3 – 2,2 cm; T of bars 2,5 – 3,5 mm Material: ceramic, Munsell colour – 5YR 6/8 (reddish yellow) Desc.: On a sherd of tegulae covered with sediment a larger part of the stamp is preserved. Letters are impressed in relief within a square deeper cartouche (3,6 mm). Cartouche and letters depth (do 2,7 mm) enhance the sense of relief. Transcription and restitution: PANSIA[..]; Pansia[na] Date: 43 BC – 14 AD Analogies: typologically hard to determine Unpublished

Cat. no. 17 Inv. no.: MGŠ 1604 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 9,8 cm; H 9,3 cm; T 2,3 cm Stamp dim.: L 2,8 cm; W 3,2 cm Letter dim.: H 1,7 – 2; W 1,3 cm; T of bars 4 mm Material: ceramic, Munsell colour – 2,5 YR 6/8 (light red) Desc.: On a small sherd of tegulae the initial part of a stamp is preserved. Letters are rendered in shallow relief with in a square cartouche. Transcription and restitution: PA[---]; Pa[nsiana] Date: 43 BC – 14 AD Analogies: typologically hard to determine Unpublished

Cat. no. 14 Inv. no.: MGŠ 1601 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 15,4 cm; H 16,6 cm; T 2,61 cm Stamp dim.: L 7,3 cm; W 3 cm Letter dim.: H 2,4; W 1,6 – 2,3 cm; T of bars 3,5 mm Material: ceramic, Munsell colour – 5YR 6/8 (reddish yellow) Desc.: On a small sherd of tegulae part of the stamp is preserved. Letters are impressed in shallow relief within a square cartouche. Transcription and restitution: PANS[---]; Pans[iana?]

Cat. no. 18 Inv. no.: MGŠ 1605 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 14,1 cm; H 7,5 cm; T 3,35 cm Stamp dim.: L 3,4 cm; W 2,6 cm 162

Stamps on ceramic and glass finds from the Velika Mrdakovicasite near Šibenik

Letter dim.: H 2,2; W 1,6 cm; T of bars 5 – 5,5 mm Material: ceramic, Munsell colour – 2.5 YR 7/4 (light reddish brown) Desc.: On a small sherd of tegulae the initial part of a stamp is preserved. Letters are rendered in shallow relief with in barley concave square cartouche. Letter P has an elongated and flattened arch. Letter A is barley visible Transcription and restitution: PA[---]; Pa[nsiana] Date: 43 BC – 14 AD Analogies: typologically hard to determine Unpublished

Cat. no. 21 Inv. no.: MGŠ 1608 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 18,8 cm; H 14,9 cm; T 2,39 cm Stamp dim.: L 8,4 cm; W 3 cm Letter dim.: H 2,3; W 1,6 – 2,3 cm; T of bars 3 – 4 mm Material: ceramic, Munsell colour – 5 YR 7/6 (reddish yellow) Desc.: On a small sherd of tegulae the latter part of the stamp is preserved. Letters are rendered in shallow relief with a concave square cartouche. The last group of letter AN are in ligature Transcription and restitution: [---]IANA[---]; [--Pans]ia͡na Date: 43 BC – 37 AD Analogies: typology is hard to determine Unpublished

Cat. no. 19 Inv. no.: MGŠ 1606 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 13,1 cm; H 10,7 cm; T 2,85 cm Stamp dim.: L 6,9 + 5,9 cm; W ca. 3 cm Letter dim.: H 2,4; W 1,6 – 2 cm; T of bars 3 – 4 mm Material: ceramic, Munsell colour – 5 YR 7/8 (reddish yellow) Desc.: On a small sherd of tegulae larger part of a stamp is preserved. Letters are rendered in shallow relief within two square cartouches which are connected in obtuse angle. Within one cartouche are the first letters of the workshop (letter S was destroyed), while in the second the rest of the stamp is impressed. Transcription and restitution: PAN[.] IA[---] Pan[s]/ia[na---] Date: 43 BC – 14 AD Analogies: / Unpublished

Cat. no. 22 Inv. no.: MGŠ 1609 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 17 cm; H 11 cm; T 3,4 cm Stamp dim.: L 14,3 cm; W 3 cm Letter dim.: H 2,2 – 2,4; W 2 cm; T of bars 4 mm Material: ceramic, Munsell colour – 2.5 YR 6/8 (light red) Desc.: On a small sherd of tegulae most of a stamp is preserved, broken at the edges. Letters are rendered in shallow relief with a concave square cartouche. Transcription and restitution: [---]PANSIAN[---]; [--]Pansian[a---] Date: 43 BC – 37 AD Analogies: typology is hard to determine Cat. no. 23 Inv. no.: MGŠ 1610 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 13,5 cm; H 16,2 cm; T 3,3 cm Stamp dim.: L 13,5 cm; W 3,3 cm Letter dim.: H 2,1 – 2,4; W 1,5 – 2 cm; T of bars 2,5 – 4 mm Material: ceramic, Munsell colour – 5YR 6/8 (reddish yellow) Desc.: On a sherd of tegulae a mostly complete stamp is preserved, broken at the edges. Letters are rendered in shallow relief with a concave square cartouche. The first letter is missing, while the last letter A is preserved with only the first bars. Transcription and restitution: [---]ANSIAN[---]; [--P]ansiana[---] Date: 43 BC – 37 AD Analogies: typology is hard to determine Unpublished

Cat. no. 20 Inv. no.: MGŠ 1607 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 14,5 cm; H 17,7 cm; T 3,1 cm Stamp dim.: L 13,4 cm; W 3,1 cm Letter dim.: H 1,8 – 2,1; W 1 – 1,9 cm; T of bars 3,5 – 5 mm Litus dim.: H 2,8 cm; W 1,7 cm Material: ceramic, Munsell colour – 10R 6/8 (light red) Desc.: On a small sherd of tegulae most of a stamp is preserved. Letters are rendered in shallow relief with a concave square cartouche. After the workshop’s name there is a litus. Transcription and restitution: [---]ANSIANA; [--- P] ansiana Date: 43 BC – 14 AD Analogies: Matijašić type III/6 or Matijašić type IV/7; Rhigini type 2c = Pellicioni type 2c or Rhigini type 8b = Pellicioni tip 8b Unpublished

Cat. no. 24 Inv. no.: MGŠ 1611 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 17,7 cm; H 16,7 cm; T 2,51 cm 163

Toni Brajković Stamp dim.: L 7,5 cm; W 2,4 cm Letter dim.: H /; W 2,2 cm; T of bars 6 mm Litus dim.: / Material: ceramic, Munsell colour – 5YR 7/8 (reddish yellow) Desc.: On a sherd of tegulae a small, terminal part of the stamp, with partially visible two last letters with thick bars and regular shape is preserved. Letters are rendered within a concave square cartouche. After the letters a damage litus is present. Transcription and restitution: [---]NA; [Pansia]na Date: 43 BC – 41 AD Analogies: typology is hard to determine Unpublished

Desc.: On a small sherd of tegulae covered with sediment, the terminal part of a stamp is preserved within a concave square cartouche. The right bar of the last letter A is barely visible. After it a slighly oblique litus is present. Transcription and restitution: [---]A; [--- Pansian]a Date: 43 BC – 41 AD Analogies: typology is hard to determine Unpublished Cat. no. 28 Inv. no.: MGŠ 21867 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 8 cm; H 10,7 cm; T 3,52 cm Stamp dim.: L 6,7 cm; W 3,05 cm Letter dim.: H 2,15 ‒ 2,2 cm; W 1 ‒ 1,4 cm; T of bars 3 ‒ 5 mm Material: ceramic, Munsell colour – 5YR 7/8 (reddish yellow) Desc.: On a small sherd of tegulae the middle part of a stamp is preserved. The stamp is rendered in shallow relief within a concave square cartouche, with regular thick-barsmed letters. Transcription and restitution: [---]ANSIA[---]; [P] ansia[na] Date: 43 BC – 41 AD Analogies: typology is hard to determine Unpublished

Cat. no. 25 Inv. no.: MGŠ 1612 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 17 cm; H 18 cm; T 2,9 cm Stamp dim.: L 5,5 cm; W 2,9 cm Letter dim.: H 2,1 ‒ 2,2; W 1,6 cm; T of bars 2,5 ‒ 4,2 mm Litus dim.: H 2,6 cm; W 1,6 cm Material: ceramic, Munsell colour – 10R 7/6 (light red) Desc.: On a sherd of tegulae a small, terminal part of the stamp is preserved. Letters are rendered in shallow relief with a concave square cartouche. After the letters a litus is present. Transcription and restitution: [---]NA; [--- Pansia]na Date: 43 BC – 41 AD Analogies: typology is hard to determine Unpublished

Cat. no. 29 Inv. no.: MGŠ 21776 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 13,2 cm; H 13,4 cm; T 2,4 cm Stamp dim.: L 10 cm; W 3,3 cm Letter dim.: H 2,2 cm; W 1,4 ‒ 1,9 cm; T of bars 3 ‒ 4 mm Material: ceramic, Munsell colour – 5YR 7/8 (reddish yellow) Desc.: Two small, joining sherd of tegula are preserved, bearing a larger fragment of a damaged stamp. The stamp is rendered in shallow relief within a concave square cartouche. Transcription and restitution: [---]SIANA; [---Pan] siạna Date: 43 BC – 68 AD Analogies: typology is hard to determine Unpublished

Cat. no. 26 Inv. no.: MGŠ 1613 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 18,3 cm; H 16,8 cm; T 2,43 cm Stamp dim.: L 4,4 cm; W 3,2 cm Letter dim.: H 2,25; W 2 cm; T of bars 4 ‒ 6,5 mm Litus dim.: W 1,5 cm Material: ceramic, Munsell colour – 5YR 6/8 (reddish yellow) Desc.: On a small sherd of tegulae the terminal part of the stamp is preserved. The only preserved letter is A, rendered in shallow relief with a concave square cartouche. After the letter a litus is present. Transcription and restitution: [---]A; [--- Pansian]a Date: 43 BC – 41 AD Analogies: typology is hard to determine Unpublished

Cat. no. 30 Inv. no.: MGŠ 1615 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 12,2 cm; H 16,6 cm; T 3,52 cm Stamp dim.: L 4,7 cm; W / cm Letter dim.: H 1,6 ‒ 2 cm; W 1,5 ‒ 1,8 cm; T of bars 3 mm Material: ceramic, Munsell colour – 5 YR 6/8 (reddish yellow) Desc.: On a small sherd of tegula part of a stamp is preserved. The letters are rendered in shallow relief within a concave square cartouche.

Cat. no. 27 Inv. no.: MGŠ 1614 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 15,3 cm; H 10 cm; T 3,56 cm Stamp dim.: L 4,3 cm; W / cm Letter dim.: H / cm; W / cm; T of bars / mm Litus dim.: W 2,3 cm; W 1,5 cm Material: ceramic, Munsell colour – 2,5 YR 6/6 (light red) 164

Stamps on ceramic and glass finds from the Velika Mrdakovicasite near Šibenik

Transcription and restitution: [---]AN[---]; [---P] an[siana---] or [---Pansi]an[a---] Date: 43 BC – 79 AD Analogies: typology is hard to determine Unpublished

Material: ceramic, Munsell colour – Munsellu – 10YR 7/3 (very pale brown) Desc.: Discoid amphorae stopper shaped in a twopart mould. The rim is slightly rounded and vertical. The upper surface is flat, with a pronounced large square handle, 1,5 cm high. The middle part is characterized by a line in relief created during manufacture. On one side there are three triangles in a row, while on the other two facing triangles. Unpublished

Cat. no. 31 Inv. no.: MGŠ 1616 Findspot: Velika Mrdakovica – settlement Tegulae dim.: L 10,9 cm; H 7,1 cm; T 2,8 cm Stamp dim.: L 10,4 cm; W 3,1 cm Letter dim.: H 2 ‒ 2,3 cm; W 1,3 ‒ 2,5 cm; T of bars 3 ‒ 4 mm Material: ceramic, Munsell colour – 5YR 7/8 (reddish yellow) Desc.: On a small sherd of tegula the middle part of a stamp is preserved. The letters are rendered in shallow relief within a concave square cartouche. Before and after the gentilicium Claudi triangular interpunctions are present — triangulum disinguens facing to the right. Two ligatures are present at the letters AV and DI. Transcription and restitution: [---] ‧ CLAVDI ‧ [---]; [T͡i(beri)] ‧ Cla͡ud͡i ‧ [Pa͡ns(iana)?] ili [Ner(onis)?] ‧ Cla͡ud͡i ‧ [A͡p(n)sia(na)] Date: 41 – 79 AD Analogies: Matijašić type VI/10 (Tiberi Claudi Pansiana); Matijašić type VII/12 (Neronis Claudi Pansiana); Pellicioni type 14 b1 (Ti Claudi Pansi); Pellicioni type 16p (Neronis Claudi Pansiana). Unpublished

Cat. no. 34 Inv. no.: MGŠ 1618 Findspot: Velika Mrdakovica – settlement Dim.: D 9,78 cm; T 1,53 cm Material: ceramic, Munsell colour – 7.5YR 7/6 (reddish yellow), in places red granulation Desc.: Discoid amphorae stopper shaped in a twopart mould. The rim is slightly rounded and vertical. The upper surface is flat, in the middle there are a square handle, 1,6 cm high, and a line in relief created during manufacture. On one side a square, shallow relief decoration is present. Unpublished Cat. no. 35 Inv. no.: MGŠ 1619 Findspot: Velika Mrdakovica – settlement Dim.: D 9,82 cm; T 1,69 cm Material: ceramic, Munsell colour – 2.5Y 8/1 (white), light gray. Desc.: Discoid amphorae stopper shaped in a twopart mould. The rim is slightly rounded and in one part sloping outwards. In the middle of the upper flat surface, a relief line is present, created during manufacture. The square handle is 0,83 cm high. Unpublished

Cat. no. 32 Inv. no.: MGŠ 1617 Findspot: Hrdanovica Tegulae dim.: L 11,3 cm; H 6,9 cm; T 2,12 cm Stamp dim.: L 10,2 cm; W 2,8 cm Letter dim.: H 2,1 cm; W 1,4 ‒ 2,4 cm; T of bars 4 ‒ 4,5 mm Material: ceramic, Munsell colour – 2.5 YR 6/8 (light red) Desc.: On a small sherd of tegula part of a stamp is preserved. The letters are rendered in shallow relief within a concave square cartouche. Two ligatures are present at the letters LA and VDI. Due to the position of the vertical bars of the second letter of the cognomen Pansiana with all probability the letters AN are in ligature. The same can be said for the second group of the same letters. Transcription and restitution: [---]AVDPAN[..]AN[--]; [Ner(onis) C]l͡au͡di Pa͡n[si]a͡n(a) Date: 54 – 68 AD Analogies: Matijašić type VII/2; Righini type 16l = Pellicioni type 16l Unpublished

Cat. no. 36 Inv. no.: MGŠ 21703 Findspot: Velika Mrdakovica – settlement Dim.: D 9,58 cm; T 1,12 cm Material: ceramic, Munsell colour – 2.5Y 8/3 (pale yellow) on the bottom, 10YR 8/3 on part of the upper surface, 5YR 7/4 on part of the upper surface. Desc.: Discoid amphorae stopper shaped in a twopart mould. The rim is slightly rounded and in one part sloping outwards. In the middle of the upper flat surface, a relief line is present, created during manufacture. The handle is 0,80 cm high, and bears on the top an X sign. Unpublished Cat. no. 37 Inv. no.: MGŠ 21577 Findspot: Velika Mrdakovica – settlement Dim.: D 8,22 cm; T 1,33 cm Material: ceramic, Munsell colour – 7.5YR 8/3 (pink)

Cat. no. 33 Inv. no.: MGŠ 21589 Findspot: Velika Mrdakovica – settlement Dim.: D 10,65 cm; T 1,46 cm 165

Toni Brajković Desc.: Discoid amphorae stopper shaped in a twopart mould. The rim is slightly rounded. In the middle of the upper very damaged surface, a relief line is present, created during manufacture. The rounded handle is 0,42 cm high. Unpublished

Cat. no. 42 Inv. no.: MGŠ 21611 Findspot: Velika Mrdakovica – settlement Dim.: D 8,05 cm; T 1,38 cm Material: ceramic, Munsell colour – 2.5Y 8/3 (pale yellow) Desc.: Discoid amphorae stopper shaped in a onepart mould. The rim is rounded . The upper surface is convex, bearing the signs III and X. The centrally located rounded handle is 1,32 cm high. Unpublished

Cat. no. 38 Inv. no.: MGŠ 1620 Findspot: Velika Mrdakovica – settlement Dim.: D 9 cm; T 1,66 cm Material: ceramic, Munsell colour – 7.5YR 8/3 (pink), from beige to pinkish Desc.: Discoid amphorae stopper shaped in a twopart mould. The rim descends vertically. In the middle of the upper flat surface, a relief line is present, created during manufacture. The square handle is 1,30 cm high. Unpublished

Cat. no. 43 Inv. no.: MGŠ 1624 Findspot: Velika Mrdakovica – settlement Dim.: D 9,10 cm; T 1,55 cm Material: ceramic, Munsell colour – 2.5Y 8/4 (pale yellow) Desc.: Discoid amphorae stopper shaped in a onepart mould. The rim is slightly rounded and descends vertically. The upper surface is slightly concave at the middle, while at the edges bears three to four relief lines. The rounded handle is 1,13 cm high. Unpublished

Cat. no. 39 Inv. no.: MGŠ 1621 Findspot: Velika Mrdakovica – settlement Dim.: D 8,82 cm; T 1,66 cm Material: ceramic, Munsell colour – 5YR 7/6 (reddish yellow) Desc.: Discoid amphorae stopper shaped in a twopart mould. The rim is slighly rounded and descends vertically. The upper surface is mostly damaged though the relief line created during manufacture is visible. The rounded handle is 0,33 cm high. Unpublished

Cat. no. 44 Inv. no.: MGŠ 1625 Findspot: Velika Mrdakovica – settlement Dim.: D 9,91 cm; T 1,43 cm Material: ceramic, Munsell colour – 5YR 6/6 (reddish yellow), beige on the surface. Desc.: Discoid amphorae stopper shaped in a onepart mould. The rim is oblique. On the upper flat surface is slightly relief letters can be discerned, but due to damage they cannot be read. The handle was, it seems, square or rectangular, 1,13 cm high. Unpublished

Cat. no. 40 Inv. no.: MGŠ 1622 Findspot: Velika Mrdakovica – settlement Dim.: D 9,11 cm; T 1,24 cm Material: ceramic, Munsell colour – 2.5Y 7/4 (pale yellow) Desc.: Discoid amphorae stopper shaped in a onepart mould. The rim is sloping outwards. The upper surface is flat. The square handle is 1 cm high and it is damaged, just as are the lateral square decorations. Unpublished

Cat. no. 45 Inv. no.: MGŠ 1626 Findspot: Velika Mrdakovica – settlement Dim.: D 9,95 cm; T 1,56 cm Material: ceramic, Munsell colour – 10YR 8/4 (very pale brown). Desc.: Discoid amphorae stopper shaped in a onepart mould. The rim is oblique on one side, rounded on the other. On the concave upper surface two relief lines can be discerned. The rounded handle is 1,10 cm high. Unpublished

Cat. no. 41 Inv. no.: MGŠ 1623 Findspot: Velika Mrdakovica – settlement Dim.: D 9,67 cm; T 1,26 cm Material: ceramic, Munsell colour – 10YR 8/3 (very pale brown), and occasionally on the surface 2.5 YR 7/6 (light red) Desc.: Discoid amphorae stopper shaped in a one-part mould. The rim is mostly descending vertically. The upper surface is uneven. The centrally located square handle is 1 cm high, and from its edges signs shaped as the Greek letter Y spread, that is in one case a V, probably a badly shaped Y. Unpublished

Cat. no. 46 Inv. no.: MGŠ 1627 Findspot: Velika Mrdakovica – settlement Dim.: D 8,49 cm; T 1,65 cm Material: ceramic, Munsell colour – from 10YR 7/4 (very pale brown) to 2.5YR 5/6 (red). Desc.: Discoid amphorae stopper shaped in a onepart mould. The rim is slightly oblique. On the flat 166

Stamps on ceramic and glass finds from the Velika Mrdakovicasite near Šibenik

upper surface the sign V in relief is present. The rounded handle is 1,32 cm high. Unpublished

Material: ceramic, Munsell colour – 2.5YR 6/6 (light red). Desc.: Discoid amphorae stopper shaped in a two-part mould, as testified by the visible line of connection on both sides of the handle. The rim is rounded, while the upper surface is flat. On the upper flat surface small relief rounded and one linear marking are visible. The rounded handle is 1,42 cm high. Unpublished

Cat. no. 47 Inv. no.: MGŠ 1628 Findspot: Velika Mrdakovica – settlement Dim.: D 9,31 cm; T 1,76 cm Material: ceramic, Munsell colour – 10YR 7/4 (very pale brown). Desc.: Discoid amphorae stopper shaped in a onepart mould. The rim is oblique. On the slightly concave upper surface the sign of the letters V and C in relief could be discerned. The rounded handle is 1,07 cm high. Unpublished

Cat. no. 52 Inv. no.: MGŠ 21795 Findspot: Velika Mrdakovica – settlement Dim.: D 8,53 cm; T 2 cm Material: ceramic, Munsell colour – 5YR 7/4 (pink). Desc.: Discoid amphorae stopper shaped in a onepart mould. The rim is rounded. In the middle of the upper surface there is a rounded handle 0,5 cm high. Unpublished

Cat. no. 48 Inv. no.: MGŠ 1629 Findspot: Velika Mrdakovica – settlement Dim.: D 9,82 cm; T 1,60 cm Material: ceramic, Munsell colour – 10YR 6/4 (light yellowish brown). Desc.: Discoid amphorae stopper shaped in a onepart mould. The rim is rounded. The upper surface is flat. The rounded handle is large and 1,44 cm high. Unpublished

Cat. no. 53 Inv. no.: MGŠ 1633 Findspot: Velika Mrdakovica – settlement Dim.: D 8,53 cm; T 2 cm Material: ceramic, Munsell colour – 10YR 6/4 (light yellowish brown). Desc.: Discoid amphorae stopper shaped in a one-part mould. The rim is sligthly rounded and descends vertically. The upper surface is flat. The conically shaped handle is 1,25 cm high. Unpublished

Cat. no. 49 Inv. no.: MGŠ 1630 Findspot: Velika Mrdakovica – settlement Dim.: D 9,28 cm; T 2,15 cm Material: ceramic, Munsell colour – 2.5Y 8/3 (pale yellow). Desc.: Discoid amphorae stopper shaped in a onepart mould. The rim is rounded and on one part oblique. In the middle of the concave upper surface there is a rounded handle 1,40 cm high, with two relief linear widenings on two opposite sides. Unpublished

Cat. no. 54 Inv. no.: MGŠ 1633 Findspot: Velika Mrdakovica – settlement Stamp dim.: L 2,9 cm; W 1 cm Material: ceramic, Munsell colour – 7.5YR 6/4 (light brown); 2.5YR 5/8 (red). Desc.: A small sherd of amphora with preserved a part of an unreadable stamp within a rectangular concave square cartouche. Differences in ceramic colour might indicate uneven firing. Unpublished

Cat. no. 50 Inv. no.: MGŠ 1631 Findspot: Velika Mrdakovica – settlement Dim.: D 8,90 cm; T 2,23 cm Material: ceramic, Munsell colour – 10YR 8/2 (very pale brown) on most of the surface, 2.5YR 7/6 (light red) on part of the rim and on the lower side. Desc.: Discoid amphorae stopper shaped in a one-part mould. The rim is oblique. On the upper flat surface small relief rounded and one linear marking are visible. The conically shaped handle is 1,10 cm high. Unpublished

Cat. no. 55 Inv. no.: MGŠ 21985 Findspot: Velika Mrdakovica – settlement Stamp dim.: L 3,30 cm; W 1,51 cm Material: ceramic, Munsell colour – 7.5YR 8/3 (pink), pale ochre. Desc.: A small sherd of amphora rim with preserved stamp within a rectangular concave square cartouche. The relief stamp reads PRVX. Unpublished

Cat. no. 51 Inv. no.: MGŠ 1632 Findspot: Velika Mrdakovica – settlement Dim.: D 4,6 cm; T 1,34 cm

Cat. no. 56 Inv. no.: MGŠ 21985 Findspot: Velika Mrdakovica – settlement 167

Toni Brajković Stamp dim.: L 4,10 cm; W 1,20 cm Material: ceramic, Munsell colour – 7.5YR 8/3 (pink), pale ochre. Desc.: A small sherd of amphora rim with preserved stamp within a rectangular concave square cartouche. The relief stamp reads TRVPO. Unpublished

Cat. no. 61 Inv. no.: MGŠ AO 5322 Findspot: Velika Mrdakovica – necropolis, grave 74 Dim.: L 10,55 cm; W 7,19 cm; H 3,66 cm Material: ceramic, Munsell colour – 2.5YR 5/6 (red). Desc.: Fully preserved oil lamp with two decorative buckles on the shoulder and closed channel. On the bottom there is the stamp FORTIS within two concentric circles. Above the stamp there is a double relief circle. Type Loeschcke IX b. Unpublished

Cat. no. 57 Inv. no.: MGŠ 1634 Findspot: Velika Mrdakovica – settlement Stamp dim.: L 2,20 cm; W 1,65 cm Material: ceramic, Munsell colour – 5YR6/8 (reddish yellow). Desc.: A small sherd of amphora rim with preserved stamp within a rectangular concave square cartouche. The relief stamp reads [---]III ili III[---] or [---]ILI Unpublished

Cat. no. 62 Inv. no.: MGŠ AO 5380 Findspot: Velika Mrdakovica – necropolis, grave 79

Dim.: L 10,64 cm; W 7,13 cm; H 3,62 cm Material: ceramic, Munsell colour – 2.5YR 6/8 (light red). Desc.: Fully preserved oil lamp with two decorative buckles on the shoulder and closed channel. On the bottom there is the stamp FORTIS within two concentric circles. Type Loeschcke IX b. Published in: Brusić 2000: 14.

Cat. no. 58 Inv. no.: MGŠ 1635 Findspot: Velika Mrdakovica – settlement Stamp dim.: L 2,20 cm; W 1,65 cm Material: ceramic, Munsell colour – 5YR8/4 (pink), pale ochre. Desc.: A small sherd of amphora rim with preserved stamp within a rectangular concave square cartouche. The relief stamp reads [---]ELIO. Unpublished

Cat. no. 63 Inv. no.: MGŠ 1637 Findspot: Velika Mrdakovica – necropolis, grave NG 2 Dim.: L 8,87 cm; W 5,97 cm; H 3,00 cm Material: ceramic, Munsell colour – 2.5YR 6/8 (light red). Desc.: Fully preserved oil lamp with two decorative buckles on the shoulder and closed channel. On the bottom there is the stamp FORTIS within two concentric circles. Type Loeschcke IX b. Published in: Brusić 2000: 14.

Cat. no. 59 Inv. no.: MGŠ 1636 Findspot: Velika Mrdakovica – settlement Dim.: L 6,70 cm; W 2,5 cm; H 3,05 cm Material: ceramic, Munsell colour – 2.5YR 6/6 (light red). Desc.: On a small fragment of oil lamp the beginning of the stamp F[---] within one concentric circle is preserve. The only remaining buckle on the shoulder is pierced. The channel is closed. Type Loeschcke IX b. Unpublished

Cat. no. 64 Inv. no.: MGŠ 1638 Findspot: Velika Mrdakovica – necropolis, grave NG 12 Dim.: L 11,24 cm; W 7,55 cm; H 3,40 cm Material: ceramic, Munsell colour – 2.5YR 6/8 (light red). Desc.: Several joining fragments of an oil lamp. Two decorative buckles on the shoulder and closed channel. On the bottom there is the partially preserved stamp FORT[IS] within two concentric circles. Type Loeschcke IX b. Published in: Brajković 2014: 32.

Cat. no. 60 Inv. no.: MGŠ AO 2718 Findspot: Velika Mrdakovica – necropolis, grave 37 Dim.: L 10,00 cm; W 6,5 cm; H 3,5 cm Material: ceramic, Munsell colour – 5YR 7/6 (reddish yellow). Desc.: Mostly preserved oil lamp. The channel is closed, on the shoulder two decorative buckles. The bottom is mostly missing. On the preserved part the ending part of the stamp […..]ILI is visible within two concentric circles. The last letter I is shorter and is placed within the vertical and horizontal lines of the letter L. Type Loeschcke IX b. Unpublished

Cat. no. 65 Inv. no.: MGŠ 1639 Findspot: Velika Mrdakovica – necropolis, grave NG 10 Dim.: L 9,03 cm; W 6,16 cm; H 3,00 cm Material: ceramic, Munsell colour – 10YR 6/8 (light red). Desc.: Fully preserved oil lamp with three pierced buckles on the shoulder and closed channel. On the bottom there is the stamp SATVRNINI within

168

Stamps on ceramic and glass finds from the Velika Mrdakovicasite near Šibenik

two concentric circles. The last two letters are in ligature. Type Loeschcke IX b. Published in: Brajković 2014: 29, 31.

Desc.: Fully preserved oil lamp with three buckles on the shoulder and open channel. On the bottom it bears the stamp FORTIS. Type Loeschcke X a. Unpublished.

Cat. no. 66 Inv. no.: MGŠ 1640 Findspot: Velika Mrdakovica – necropolis, grave NG 1 Dim.: L 10,25 cm; W 7,05 cm; H 3,35 cm Material: ceramic, Munsell colour – 2.5YR 6/8 (light red). Desc.: Fully preserved oil lamp with three decorative buckles on the shoulder and channel connected to the disk. On the bottom there is the stamp FORTIS within two relief circles. Under the stamp a small rosette with central circle is impressed. Type Loeschcke IX c. Published in: Brajković 2014: 11.

Cat. no. 71 Inv. no.: MGŠ 1642 Findspot: Velika Mrdakovica – necropolis, grave NG 3 Dim. of the vessel: H 3,50; D: 9,80 cm Material: ceramic, Munsell colour – 10YR 5/8 (red) Desc.: Ceramic vessel (cup) reconstructed from several fragments. On the bottom a graffito is visible. The first letter resembles a Greek lambda, though it could be read as an A without horizontal line. The second letter is N, while the third could be I, C or S. Published in: Brajković 2014: 13-14.

Cat. no. 67 Inv. no.: MGŠ AO 5500 Findspot: Velika Mrdakovica – necropolis, grave 103 Dim.: L 10,47 cm; W 7,36 cm; H 3,69 cm Material: ceramic, Munsell colour – 2.5YR 5/6 (red). Desc.: Fully preserved oil lamp with three decorative buckles on the shoulder and channel connected to the disk. On the bottom there is the stamp LITOGENE within two concentric relief circles. The last two letters are in ligature. Type Loeschcke IX c. Unpublished.

Cat. no. 72 Inv. no.: MGŠ 1643; MGŠ 1644 Findspot: Velika Mrdakovica – necropolis, grave NG 3 Dim.: bottom MGŠ 1643 ‒ 4,45 x 4,25; bottom MGŠ 1644 ‒ 4,90 x 5,20 Material: glass Published in: Brajković 2014: 13-14; Štefanac 2012: 42-44, 182, 187, 343, kat. br. 108-109.

Cat. no. 68 Inv. no.: MGŠ 1641 Findspot: Velika Mrdakovica – necropolis, grave NG 11 Dim.: L 5,50 cm; W 4,88 cm; H 2,56 cm Material: ceramic, Munsell colour – 2.5YR 7/6 (light red). Desc.: Fragment of oil lamp with preserved bottom and stamp. The whole surface is covered with sediment. The stamp NERI is impressed within two concentric relief circles. Published in: Brajković 2014: 28.

References

Translated by Ana Konestra (Institute of Archaeology, Zagreb, Croatia)

Abramić, M. and A. Colnago 2009. Istraživanja u sjevernoj Dalmaciji. III. Iskapanja u Starigradu – Argyrvntvm. Asseria 7: 161–205. Auer, M. 2012. Die romischen Firmalampen aus dem Municipium Claudium Aguntum, 12-13 (11-24), in L. Chrzanovski (ed.) Le Luminaire antique, Lychnological Acts 3. Actes du 3e Congrès International d’études de l’ILA,Université d’Heidelberg, 21 - 26.IX.2009, (Monographies Instrumentum 44): 11–24. Montagnac: Mergoil 2012. Auer, M. 2016. Names on Lamps: Distribution /Quantity of Firmalampen and Regional Trade / Imena na svjetiljkama: distribucija/količina Firma svjetiljki i regionalna trgovina, in G. Lipovac Vrkljan, I. Ožanić Roguljić and M. Ugarković (eds) Rimske i kasnoantičke svjetiljke: proizvodnja i distribucija, kontakti na Mediteranu, Zbornik međunarodnog okruglog stola, Zagreb, 2. veljače 2015: 38–46. Zagreb: Institut za arheologiju. Auer, M. and S. Sitz 2014. Die Firmalampenproduktion des EVCARPVS – eine Studie zur römischen Keramikproduktion. Germania 92: 85–125. Bezeczky, T. 2010. Trade connections betwen Ephesus an Adriatic region. Histria Antiqua 19: 351–358. Blečić Kavur, M. and E. Podrug 2014. Nekropola gradine Velika Mrdakovica – grobovi starijega željeznoga doba / The Necropolis at the Velika Mrdakovica

Cat. no. 69 Inv. no.: MGŠ AO 5549 Findspot: Velika Mrdakovica – settlement Dim.: L 4,90 cm; W 3,20 cm; H 1,20 cm Material: ceramic, Munsell colour – 2,5YR 6/8 (light red). Desc.: Fragment of oil lamp with preserved bottom and stamp within two concentric relief circles. An oblique line of the letter A is visible, and the last three letters ‒ […]ARPI. Published in: Brajković 2014: 28. Cat. no. 70 Inv. no.: MGŠ AO 5367 Findspot: Velika Mrdakovica – necropolis, grave 77 Dim.: L 11,50 cm; W 8,00 cm; H 3,50 cm Material: ceramic, Munsell colour – /

169

Toni Brajković Hillfort – Early Iron Age Graves. Vjesnik za arheologiju i historiju dalmatinsku 107: 31–112. Brajković, T. 2011. Lokalitet Rakitnica – Tri bunara u vodičkom zaleđu / Archaeological site Rakitnica – Tri bunara in Vodice hinterland, in G. Lipovac Vrkljan, B. Šiljeg, I. Radić Rossi (eds) Rimske keramičarske i staklarske radionice. Proizvodnja i trgovina na jadranskom prostoru. Zbornik I. međunarodnog arheološkog kolokvija, Crikvenica, 23. – 24. listopada 2008: 95–101. Crikvenica: Muzej Grada Crikvenice. Brajković, T. 2014. Velika Mrdakovica, rezultati najnovijih arheoloških istraživanja (2011. – 2013.) / Velika Mrdakovica, results of the latest archaeological research (2011 – 2013) (exhibition catalogue). Vodice: POU. Brajković, T. 2014a. Stakleni pužoliki riton s liburnsko – rimske nekropole Velike Mrdakovice kod Šibenika / Glass snail-shaped riton from the Liburnian Roman necropolis in Velika Mrdakovica near Šibenik, , 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: 357–366. Crikvenica: Muzej Grada Crikvenice. Brajković, T. 2018. Ritual ukopa na nekropoli Velike Mrdakovice / Burial Rituals at the Necropolis of Velika Mrdakovica. Miscellanea Hadriatica et Mediterranea 5: 31–63. Brusić, Z. 2000. Arauzona. Velika Mrdakovica, liburnski grad i nekropola / The Liburnian City and Necropolis (exhibition catalogue). Šibenik; Muzej grada Šibenika. Buljević, Z. 1999. Njive - Podstrana: groblje iz vremena seobe naroda u Naroni (Narona II). Vjesnik za arheologiju i historiju dalmatinsku 90-91: 201–293. Buora, M., M. - B. Carre, C. Tiussi and P. Ventura 2008. Bolli su anfore Lamboglia 2 o simili dall’area Aquileiese. Rei Cretariae Romanae Fautorum Acta 40: 285‒304. Buora, M. 2014. Alcune considerazioni sui coperchi d’anfora editi dal Friuli. Quaderni Friulani di Archeologia XXII / XXIII (2012 – 2013): 25‒32. Cambi, N. 1989. Anfore Romane in Dalmazia, in: Amphores romaines et histoire économique. Dix ans de recherche, Actes du colloque de Sienne (22-24 mai 1986) (Collection de l’École française de Rome 114): 311‒337. Roma Paris: Ecole française de Rome, Diffusion de Boccard. Cipriano, S. 2016. Anfore Lamboglia 2, Dressel 6A e Dressel 6B dall piazzale della cattedrale di Ivlia Concordia: nuovi dati. Antichità altoadriatiche LXXXIII: 145‒158. Cipriano, S. and S. Mazzocchin 2003. I laterizi bollati del Museo Archeologico di Padova: una revisione dei dati materiali ad epigrafici. Bollettino del Museo Civico di Padova XCII: 29‒76. Cipriano, S. and S. Mazzocchin 2012. Produzioni anforarie dell’Italia alto e medioadriatica in età Romana, in C. S. Fioriello (ed.) Ceramica romana nella Puglia Adriatica: 241‒254. Bari: Sedit.

Cipriano, S. and S. Mazzocchin 2016. Opercula inscripta da Patavium: dati epigrafici e cronologici, in J. Bonetto, M. Salvadori, A.R. Ghiotto, P. Zanovello and M.S. Busana (eds) I mille volti del passato. (Scriti in onore di Francesca Ghedini): 825‒858. Roma: Edizioni Quasar. Chinelli, R. 1994. Coperchi d’anfora, in M. Verzár-Bass (ed.) Scavi ad Aquileia I. L’area a est del foro (Studi e ricerche sulla Gallia Cisalpina 6): 464‒490. Roma: Edizioni Quasar. Crnobrnja, A. N. 2006. Kultna namena rimskih žižaka u Gornjoj Meziji. Beograd: Filozofski fakultet, Centar za arheološka istraživanja. Dobreva, D. and B. Luise 2014. I coperchi d’anfora dagli scavi dei fondi ex Cossar ad Aquileia. dai vecchi dati ai nuovi ritrovamenti. Quaderni Friulani di Archeologia XXII / XXIII (2012 – 2013): 75‒92. Dolci, M. 2014. Coperchi d’anfora dagli scavi dell’École française de rome e dell’Università di trieste al porto fluviale di Aquileia. Quaderni Friulani di Archeologia XXII / XXIII (2012 – 2013): 55‒64. Fadić, I. and B. Štefanac 2012. Workshop stamps on square bottles from the Zadar region, in D. Ignatiadou and A. Antonaras (eds) Annales du 18e Congrès de l’associaton internationale pour l’histoire du verre, Thessaloníki, 2009: 215‒220. Thessaloniki: ZITI Publishing. Giunio, A. and T. Alihodžić 2019. Pogrebni obredi i običaji staroga Rima / The Ancient Romans’ Funerary Rituals and Customs (exhibition catalogue). Zadar: Arheološki muzej Zadar. Gomezel, C. 1995. I laterizi bollati del Museo dello Stella (Palazzolo dello Stella - Udine). Aquileia Nostra  LXVI: 9‒64. Gomezel, C. 1996. I laterizi bollati romani del FriuliVenezia Giulia: analisi, problemi e prospettive (L’Album, 4). Udine: Gruppo archeologico Veneto orientale, Fondazione Antonio Colluto. Harris, W.V. 1980. Roman Terracotta Lamps: The Organization of an Industry. The Journal of Roman Studies 70: 126‒145. Isings, C. 1957. Roman Glass from dated finds (Archaeologica Traiectina 2.). Groningen: J.B. Wolters.Koncani Uhač, I. 2007. Transport vina i ulja u šibenskom akvatoriju kroz antiku. Histria Antiqua 15: 361‒376. Krunić, S. 2005. Pregled antičkih svetiljki Singidunuma, Singindunum 4: 45‒104. Labate, D. 2013. Gli impianti produttivi della collina Modenese in età Romana: note sulla produzione di ceramica e lucerne, in D. Labate, L. Mercuri, S. Pellegrini (eds) Il mosaico ritrovato. Indagini archeologiche a Savignano sul Panaro (Quaderni di Archeologia dell’Emilia Romagna 31): 33‒40. Firenze: All’Insegna del Giglio. Labate, D. 2016. Mutina fecit. Dalle Herzblattlampen alle Firmalampen: nuovi dati sulla produzione di lucerne a matricedal territorio di Modena / Mutina fecit. Od Herzblattlampen do Firmalampen: novi podaci o svjetiljkama proizvedenima u kalupima s prostora 170

Stamps on ceramic and glass finds from the Velika Mrdakovicasite near Šibenik

Modene, in G. Lipovac Vrkljan, I. Ožanić Roguljić and M. Ugarković (eds) Rimske i kasnoantičke svjetiljke: proizvodnja i distribucija, kontakti na Mediteranu, Zbornik međunarodnog okruglog stola, Zagreb, 2. veljače 2015: 18‒37. Zagreb: Institut za arheologiju. Labate, D. 2017. Mutina Fecit: la produzione di lucerne a Modena, in L. Malnati, S. Pellegrini, F. Piccinini and C. Stefani (eds) Mutina Splendidissima. La città romana e la sua eredità: 221‒227. Roma: De Luca. Labate, D. 2017a. Il suburbio di Mutina e la stratigrafia relevata nel Parco Novi Sad: depositi antropici e alluvionali , loro correlazione e datazione, in D. Labate and L. Malnati (eds) Parco Novi Sad di Modena: dallo scavo al parco archeologico (Quaderni di Archeologia dell’Emilia Romagna 36): 15‒22. Firenze: All’Insegna del Giglio. Lete, N. 2005. Pločasti čepovi amfora. Split: Split : Podvodno istraživački klub Mornar. Linderski, J. 1986. Watching the birds: Cicero the Augur and the Augural Templa. Classical Philology 81/4: 330‒340. Lodi, G. 2015. Opercula fittili con grafemi ed iscrizioni da Ariano Ferrarese (Mesola, Ferrara). Amphora lids with symbols and inscriptions from Ariano Ferrarese (Mesola, Ferrara). Sylloge Epigraphica Barcinonensis XIII: 145‒162. Mardešić, J. 2002. Keramičke svjetiljke, in E. Marin (ed.) Longae Salonae I: 349‒364. Split: Arheološki muzej u Splitu. Marengo, S. M. and G. Paci 2008. Per la circolazione delle anfore Rodie e tardo-republicane in area Adriatica, in P. Basso (ed.) Est enim ille flos Italiae...Vita economica e sociale nella Cisalpina romana. Atti delle giornate di studi in onore di Ezio Buchi, Verona 30 nov.-1 dic. 2006: 313‒328. Verona: QuiEdit. Matijašić, R. 1983. Cronografia dei bolli laterizi della figulina pansiana nelle regioni adriatiche. Mélanges de l’École française de Rome. Antiquité 95/2: 961‒995. Patsch, C. 1908. Kleinere Untersuchungen in und um Narona. Jahrbuch für Altertumskunde 2: 87‒117. Pedišić, I. and E. Podrug 2007 (2008). Antički opekarski pečati iz fundusa Muzeja grada Šibenika / Roman Brick Workshop Stamps from the Collection of the Šibenik City Museum. Opuscula Archeologica 31: 81‒141. Pellicioni, M. T. 2012. La Pansiana in Adriatico. Tegole romane per navigare tra le sponde / Pansiana na Jadranu. Rimska strešna opeka za plovdbo med obalami. Portomaggiore: Arstudio.

Righini, V., M. Biordi and M. T. Pellicioni Golinelli 1993. I bolli laterizi romani della regione Cispadana (Emilia e Romagna), in C. Zaccaria (ed.) I laterizi di età romana nell’area nordadriatica (Cataloghi e monografie archeologiche dei Civici Musei di Udine 3): 23–91. Roma: L’Erma di Bretschneider. Righini, V. (ed.) 1998. Le fornaci romane: produzione di anfore e laterizi con marchi di fabbrica nella Cispadana orientale e nell’Alto Adriatico: atti delle giornate internazionali di studio (Rimini, 16-17 ottobre 1993). Rimini : Musei comunali. Righini, V. 2004. Fra produttori e consumatori. I materiali fittili pesanti nella Cisalpina, in S. Santoro (ed.) Artigianato e produzione nella Cisalpina, 1. Proposte di metodo e prime applicazioni: 239-264. Firenze: All’Insegna del Giglio. Rocco, G. 2009. Anelli con il lituus su statue bronzee romane, in E. Dettori and C. Braidotti (eds) Ou pan ephemeron. Scritti in memoria di Roberto Pretagostini: 735‒747. Roma: Quasar. Rosenthal-Heginbottom, R. 2015. Factory Lamps ‘Firmalampen’ in the Levant. Bulletin of the Anglo-Israel Archeological Society 33: 119‒146. Schneider, G. 1993. Studies of Roman Lamps from the Northern Provinces and from Rome, in W.V. Harris (ed.) The Inscribed Economy. Production and distribution in the Roman empire in the light of instrumentum domesticum (Journal of Roman Archaeology Supplementary Series 6): 129‒137. Ann Arbor: University of Michigan. Stošić, K. 1941. Sela šibenskoga kotara. Šibenik: Kačić. Štefanac, B. 2012. Radionički žigovi na staklenim recipijentima s područja antičke Liburnije, unpublished PhD dissertation. Zadar: University of Zadar. Šuta, I. 2014. Amphora lids from Siculi. Quaderni Friulani di Archeologia XXII / XXIII (2012 – 2013): 109‒127. Topoleanu, F. 2010. Opaiţele de tip ‘Firmalampen’ din colecţiile muzeului Judeţean di istorie şi arheologie Prahova. Arheologia Moldovei XXXIII: 151‒190. Zmaić, V. and I. Miholjek 2013. Podvodno arheološko istraživanje lokaliteta Desilo – Hutovo Blato. Godišnjak Hrvatskog restauratorskog zavoda 4: 171‒186. Žerjal, T. and I. Bekljanov Zidanšek 2014. Coperchi d’anfora con scritte, segni e grafemi dalla Slovenia. Quaderni Friulani di Archeologia XXII / XXIII (2012 – 2013): 137‒163.

171

4.

The study of material evidence of pottery, ceramics and glass production and distribution

HILARVS, from the Po valley to the Rhone valley? Armand Desbat, Valentina Mantovani, Eleni Schindler Kaudelka, Roby Stuani Abstract The discovery in Lyon and later in Vienne of workshops producing Aco beakers has raised the question of the relationship between the production within these two towns and the yet not localized Po valley workshops. The only name present both in Gaul and the Padana is Hilarvs Aco. Is this just a coincidence due to a frequent slave name, is it an imitation of products fashionable at a given time or the testimony of the transfer of Po valley craftsmen to the Rhone valley?

Keyword Aco beakers from the Po valley and the Rhône valley, Magdalensberg, Hilarvs Aco, Hilarvs Gavi.

Introduction The discovery of ceramic factories in La Muette and Loyasse in the 1960s was a small revolution in Gaul, proving the existence – in Lyon (La Muette) – of a branch in Arezzo that produced ‘Arretine’ Sigillata as well as Aco beakers until then were believed to have been manufactured exclusively in the Po Valley. Excavations carried out in Saint-Romain-en-Gal have revealed the existence in ancient Vienne, too, of a production of Aco beakers very similar to those of La Muette (Desbat 1985). Among the various signatures attested in the workshops of Lyon and Vienne some were only in these workshops such as CHRYSIPPVS or T. CAVIVS. Others like ACO and HILARVS ACO, very common in Loyasse, La Muette and Vienne/St. Romain en Gal, are also documented in northern Italy from where they were exported to Noricum, in particular the emporium on the Magdalensberg (Schindler-Kaudelka 1980; 1998; 2000). Given the almost identical names, the question arises whether the Lyon-Loyasse and Saint-Romain workshops have to be considered as subsidiaries of the Po Valley ateliers (Desbat 2000) and whether it can be assumed that the workshop run by Hilarvs Aco operated in northern Italy before it moved in the Rhone valley or just opened a branch there. In the latter case, would this be a simple technology transfer, sending mainly stamps and dies from the head office in Italy to the newly founded subsidiary in Gaul? In the former the potter Hilarvs Aco move from one region to the other? The question becomes even more complex through the presence – in northern Italy and on the Magdalensberg – of some vessels signed HILARVS GAVI S[ERVVS]. For those no direct link with the production of Hilarvs Aco is evident, while some analogies in the decoration pattern

are evident with products from La Muette signed by T. Cavivs. The beakers of HILARVS in Gaul Lyon/Loyasse The digging of a modern grave in the cemetery of Loyasse in 1967 revealed a level composed of misfired pottery waste, including a large number of Aco beakers, fragmented and deformed (Desbat et al. 1996). Potters marks are restricted to ACO and HILARVS ACO (Genin et al. 1996). HILARVS ACO appears in most cases framed by two dolphins, only rarely without accompanying sign. The production is characterized by friezes made of stylized leafs and flowers, often double-rowed, delicately and precisely manufactured, above the Kommaregen that covers the body (Figure 1). At least fifteen different friezes have been identified, most of them characteristic enough to safely attribute them even in absence of a signature. Pieces referable to this atelier are easy to distinguish from products of La Muette or Saint-Romain-en-Gal. Two more characteristics of the Loyasse production consist in the drop or crown-shaped motifs that adorn the base of the decorated zone and the presence, of a dark grey slip (Figure 2), neither attested in La Muette nor in Saint-Romain-en-Gal. Lyon/La Muette Excavations carried out in 1966 and 1975 at La Muette brought to light a large quantity of Aco beakers, 440 bearing a manufacturer’s mark, among them 7 times

roman pottery and glass manufactures (Archaeopress 2022): 173–190

Armand Desbat, Valentina Mantovani, Eleni Schindler Kaudelka, Roby Stuani

Figure 1. Atelier of Loyasse examples of friezes and potters names (photo: A. Desbat).

Figure 2. Lyon/Loyasse goblet with brownish-black slip (photo: A. Desbat).

Figure 3. Lyon/la Muette selection of signatures ACO and HILARVS (drawings: A. Desbat).

ACO, 18 times HILARVS and 114 times HILARVS ACO, with visible differences in the letters (Figure 3).

Vienne/Saint-Romain-en-Gal During the archaeological investigations carried out at the Maison des Dieux-océans, a small pit was detected. It contained almost 5000 fragments adding up to more than 130 stamped Aco beakers (Desbat 1985) as well as two moulds, one signed HILARVS ACO (Figure 6).

The lot contained only two moulds, one of HILARVS. The study of the decorations from La Muette yields 17 friezes attributed to the Hilarvs group (Desbat et al. 1996: pl. 41-42). Most of them belong to single rows of leaf or flower garlands, only two double rowed friezes are present (Figures 4 and 5).

The beakers offer a large variety of decorations and more than 130 signatures, some of them unpublished (Figure 7): CHRYSIPPVS is the most documented name, aside from ACO (4 specimens), HILARVS (4 specimens) (Figure 7 a, f), HILARVS/P (1 specimen), HILARACO (2 specimens) (Figure 7 c) and HILARVS ACO (20 specimens) (Figure 7 d, e, h).

In combining the shapes of the letters, the complete signature, the separation motifs and the patterns of the friezes, 29 different types are noticed, seven of which are also attested in Saint-Romain-en-Gal (Table 1). 174

HILARVS, from the Po valley to the Rhone valley?

Figure 4. Lyon/La Muette, examples of decoration types of Hilarvs (photo: A. Desbat).

Figure 5. Lyon/La Muette Range of shapes with signature Hilarvs (drawings: A. Desbat). 175

Armand Desbat, Valentina Mantovani, Eleni Schindler Kaudelka, Roby Stuani 1 2 3 4 5 6 7 8 9 10 11 12 13 14

HILARVS alone HILARVS alone HILARVS alone HILARVS alone HILARVS/P HILARACO HILARVS ACO HILARVS ACO HILARVS ACO HILARVS ACO HILARVS ACO HILARVS ACO HILARVS ACO HILARVS ACO

15

HILARVS ACO

16

HILARVS.ACO

17 18 19 20 21 22 23 24 25 26 27 28 29

HILARVS ACO HILARVS ACO HILARVS ACO HILARVS ACO HIL[ARVS] HI[ARVS] HILARVS ACO HILARVS ACO HILARVS ACO HILAR[VS ACO] [HIL]ARVS ACO [HILAR]VS [HILAR]VS.ACO mould fragment

without framing without framing framed by two masks framed by two bees without framing without framing without framing framed by two double leaves without framing framed by two 6 petal rosettes framed by two shells framed by rosettes. Rosette between Hilarvs and Aco,. framed by two double leaves Single komma between Hilarvs and Aco framed by two double leaves Single komma between Hilarvs and Aco without framing framed by two double leaves framed by two bees framed by two shells framed by two 8 petal rosettes framed by two 8 petal rosettes framed by two ivy leaves framed by standing Minerva and bust of Sol framed by two masks framed by two masks frieze n. 29 framed by a leaf with double rod between two vertical ivy leaves

frieze n. 1 frieze n. 31 under double frieze n. 40 under double frieze n. 46 under frieze n. 9 under frieze n. 3 under frieze n. 8 under frieze n. 12 under frieze n. 31 under frieze n. 31 under frieze n. 29 under frieze n. 29 under frieze n. 21 under frieze n. 9 under frieze n. 15 under frieze n. 15 under frieze n. 15 under frieze n. 24 under frieze n. 15 over garland under frieze n. 14 under double frieze n. 38 under double frieze n. 39 under frieze n. 21 under frieze n. 24 over garland under frieze n. 24 under frieze n. 44 no frieze preserved under frieze n. 25

Table. 1. Summary table of the discoveries of Hilarvs vessels in the Po Valley mentioned in the text.

The study revealed an identical repertoire to that defined in La Muette. Some of the moulds were used in both workshops (Desbat 1996). The group of HILARVS ACO of Vienne/St. Romain includes 11 different friezes. With the exception of no. 2, all decorations are known in La Muette. The association between frieze, manufacturer’s name and separation motifs lead to distinguish 17 different decoration patterns (Table 2). Six of the decorations are attested at La Muette (nn. 3, 6, 8, 8, 9, 9, 14, 17), three of which come from the same mould (Desbat 1996: fig. 80). The beakers of SaintRomain-en-Gal were unearthed in a pit, which probably

Figure 6. Atelier of Saint-Romain-en-Gal (Vienne), mould manufactured by HILARVS (photo: A. Desbat).

176

HILARVS, from the Po valley to the Rhone valley?

Figure 7. Atelier of Saint-Romain-en-Gal (Vienne), sampler of decorations by Hilarvs (drawings: A. Desbat).

collected the wasters of a single firing process. Given the particular situation, it is very probable that parts of the decorations made in the workshop are missing. Therefore, the materials found in this context may not represent the complete repertoire produced in Saint-Romain-en-Gal. That is why trying to distinguish

between the productions of Saint-Romain and La Muette based on the decoration patterns only remains risky. Furthermore, the absence in Saint-Romain of T. Gavius or Philarcurus may be due to a simple coincidence. In the Dangstetten camp (Roth-Rubi 2006), beakers 177

Armand Desbat, Valentina Mantovani, Eleni Schindler Kaudelka, Roby Stuani 1 2 3 4 5 6 7

HILARVS alone HILARVS alone HILARVS/P HILARACO HILARVS. ACO [HILARV]S. ACO HILA[RVS] ACO

8

HILARVS. ACO

9 10 11 12 13

[HIL]ARVS ACO HIL[ARVS]. ACO [HILARVS AC]O [HI]ARVS. ACO HILAR[US]

14 HILARVS ACO 15 HILARVS [ACO] 16 HILARVS [ACO ?] 17 [HILARVS] ACO

framed by two rosettes framed by two rosettes without framing framed by two bees framed by two rosettes framed by two double leaves framed by two masks, ivy leave between Hilarvs and Aco without framing single komma between Hilarvs and Aco framed by two masks framed by two bees framed by two rosettes framed by two rosettes without framing without framing, single komma between Hilarvs and Aco framed by two shells framed by two leaves framed by two rosettes

under frieze n. 9 double frieze unpublished under frieze n. 9 under frieze n. 15 under frieze n. 31 under frieze n. 15 under frieze n. 24

(n. 95 et 96) (n. 97) (n. 94) (n. 89 et n. 93) (n. 90 et 91) (n. 101) (n. 103)

under frieze n. 12

(n. 104, 105)

under double frieze n. 44 under frieze n. 1 under frieze unknown in La Muette under frieze n. 6 under unpublished frieze, variant of n. 1 from la Muette under frieze n. 29

(n. 106) (n. 107) (n. 108) (n. 99) (n. 109)

over a garland under frieze n. 29 under frieze n. 29

(n. 111, 112, 113, 114) (4 ex) (n. 116) (n. 115 et 117) (n. 110)

Table. 2. Summary table of the discoveries of Hilarvs vessels of the Magdalensberg mentioned in the text.

made by T. Cavius represent the majority, but most of the decorations are not attested at La Muette, while archaeometric analyses suggest an origin in Vienne (Schneider and Dasczkiewicz 2006).

The similarity of the decoration patterns with la Muette, in some cases the perfect identity and even the use of the same moulds on both sites, makes a distinction between the two workshops hazardous. Due to the high fragmentation rate, archaeometric analyses cannot always be performed, and therefore attribution is often reduced to macroscopic observations. Goblet feet at La Muette very often bear wheel-turn traces, while this is never the case at Saint-Romain.

Chronology and diffusion of the Lyon and Vienne workshops The workshop at Loyasse is the oldest and its period of activity dates between 30 and 15 BC. Loyasse’s goblets were widely distributed. Thanks to their characteristics, they have been identified at various sites in northern Gaul or on the limes: Argenton-sur-Creuse, Paris, Amiens, Arras, Bavay, Liberchies, Goeblange-Nospelt, the Titelberg, Neuss, Oberaden, Dangstetten, Basel, Zürich (Desbat 2019). They are typical features in sites with evidence of early Romanisation or early military presence. In Lyon they are abundant in the contexts of the ‘Pseudo-Sanctuaire de Cybèle’ dated between 30 and 20-15 BC. (Desbat 2012) (Figure 8).

Considering exclusively the names on the cups, it appears that the Rhône group is widely distributed, particularly towards the Limes, as already stated by M. Vegas (1970). No diffusion of the products of Lyon-Vienne neither southbound to the Narbonensis nor westbound to Aquitania can be traced. The only cup of Hilarvs Aco published from these areas, found in Ruscino (Marichal and Mayet 1980) is most certainly of italic origin. HILARVS signed beakers from the Po Valley Evidence from the Western side of the Po Valley

For the workshop at La Muette, a date between 15 BC and 15 AD had been proposed, but an exhaustive study of the finds revealed two successive periods of production. Aco beakers belong to the first, framed earlier than the Haltern horizon (Genin et al. 1996).

Only seven examples signed by Hilarvs are documented between the Pre-Alps of the eastern Piemont and Lombardy west of the Adige River, all of them devoid of coating.

Activities in the workshop of Saint-Romain-en-Gal seem to start slightly earlier than in La Muette because the composition pattern on some moulds is close to those known from Vienne.

Oleggio (NO) Tomb 79b of the necropolis contained a beaker decorated with garland of leaves oriented towards the 178

HILARVS, from the Po valley to the Rhone valley?

Figure 8. Lyon, Pseudo-Sanctuaire de Cybèle: distribution of Aco beakers in the various phases of the site, percentage in relation to thin-walled pottery.

right under which part of the name [A]CO H[I]LARVS is preserved (Figures 9, 10 n. 1, Table 1) (Deodato and Poletti Ecclesia 1999: fig. 126, n. 2; Baron 1999: fig. 361). The grave goods are dated in Augustan times (Deodato and Poletti Ecclesia 1999: 141-143, fig. 126, n. 2).

Ottobiano (PV) Tomb 36 of the necropolis yielded an important completely preserved beaker decorated with a frieze of leaves running left. A second frieze made of heartshaped elements alternating with buds holds the name HILARVS.GAVI.S, delimited on both sides by small masks facing each other (Figures 9 and 10 n. 10, Table 1) (Lavizzari Pedrazzini 1987: 84, table 19 n. 1). The final S after the genitive GAVI reveals that Hilarvs was a slave working for Gavivs. The burial is dated in the early Augustan age (Vannacci Lunazzi 1987: 82-84, fig. 15, table X n. 1). This is the oldest piece signed by Hilarvs Gavi known in the Po Valley.

Milan (MI) Excavations in the historical center brought to light three more beakers, two of them in a rescue dig carried out in 1992 in Piazza Erculea. One rim signed by HILAR.ACO (Figures 9 and 10 n. 5, Table 1) – the letters are inserted in the frieze of leaves oriented to the right – and the second, a small fragment of the wall with the fragmentary signature [HILA]RVS. ACO under a chain of left-facing vegetal elements (Figures 9 and 10 n. 6, Table 1) (Pisano 2000: 302, Figs. 5-8). Chronology and stratigraphy are still awaiting publication, therefore dating remains an open issue.

Cremona (CR) Recent excavations in the residential district of Piazza Marconi brought to light a couple of fragments of Aco beakers of the Hilarvs group. The first of these preserves the fragmentary signature HILA[RVS], inserted in a complex pattern. A double frieze consisting of a chain of leaves oriented towards the left on top and a line of heart-shaped elements alternating with buds surround the manufacturers name (Figures 9 and 10 n. 2, Table 1) (Volonté 2007: fig. 1; 2018: table II. n. 7). At least two side

Little can be said of the other Milanese fragment, found in a late antique context of the excavations of Line 3 of the Milan Metro. It preserves three letters HIL[ARVS] (Figures 9 and 10 n. 4, Table 1) placed under a chain of vegetal elements facing right (Ceresa Mori 1991: 49, table XVI n. 1).

179

Armand Desbat, Valentina Mantovani, Eleni Schindler Kaudelka, Roby Stuani

Figure 9. Friezes and signatures of Hilarvs found in the Po Valley (drawings: V. Mantovani).

180

HILARVS, from the Po valley to the Rhone valley?

Figure 10. Hilarvs from the Po Valley. 1: tomb 79b of the necropolis of Oleggio (NO), by concession of the Ministry of Cultural Heritage and Activities and Tourism — Turin, Royal Museums — Museum of Antiquities; 2 and 3: Piazza Marconi Cremona (CM), by concession of the Ministry of Cultural Heritage and Activities and Tourism – Archaeological Museum of San Lorenzo, Cremona; 4: excavations of Line 3 of the Milan Metro (MI), (after Ceresa Mori 1991, table XVI n. 1, graphic elaboration by V. Mantovani); 5-6: excavation of Piazza Erculea (MI), (after Pisano 2000, Figs. 7-8, graphic elaboration by V. Mantovani); 7-9: shipwreck of Valle Ponti, Comacchio (FE), by concession of the Ministry of Cultural Heritage and Activities and Tourism — Soprintendenza Archeologia Belle Arti e Paesaggio per le Province di Modena, Reggio Emilia e Ferrara; 10: tomb of the necropolis of Ottobiano (PV), by concession of the Ministry of Cultural Heritage and Activities and Tourism – Soprintendenza Archeologia Belle Arti e Paesaggio per le Provincie di Como, Lecco, Monza e Brianza, Pavia, Sondrio, Varese; 11: stipe del Montirone - Abano Terme (PD), by concession of the Ministry of Cultural Heritage and Activities and Tourism — Museo Nazionale Atestino — Polo Museale del Veneto.

181

Armand Desbat, Valentina Mantovani, Eleni Schindler Kaudelka, Roby Stuani stamps delimit the name, a stylized turtle in bird’s eye view and a roughly drawn female mask. In analogy to the cup of Ottobiano, the other mask probably depicted a man. The shape of the letters, as well as the mask and the heart motif point to Hilarvs Gavi rather than Hilarvs Aco.

Comacchio (FE) Three more examples of the Hilarvs group come from the shipwreck detected in 1981 in the location Valle Ponti, probably sunk between 19 and 12 BC, as evidenced by 102 lead ingots marked with the name of M. Vipsanius Agrippa in the cargo (Berti 1995: 39).

The second fragment from Cremona also belongs to Hilarvs Gavi, with part of the signature [HILARV]S.GAVI (Figures 9 and 10 n. 3, Table 1) preserved under a chain of double leaves facing left (Volonté 2018: table II. n. 6). A small trefoil flanked by two kommas divides the two parts of the name, while on other pieces only a single komma is used as punctum subdividens.

Signatures read HILARVS.GAVI, HILARVS.GAVI and HILARVS.GAVI.S (I.G. 57350, 57146 and 55916). The first one (Figures 9 and 10 n. 7, Table 1) is placed under a double frieze consisting of a chain of leaves facing left and a second band decorated with frontal theatrical masks. The name is delimited on both sides by two pairs of small masks facing each other, while between Hilarvs and Gavi there is one more stylized mask. The second one (Figures 9 and 10 n. 9, Table 1) also has a double frieze made of the superimposition of two chains of leaves facing to the left. The name divided by a single Komma appears underneath. The third piece (Figures 9 and 10 n. 8, Table 1) has a single-rowed frieze and shows the name placed below, again with a single komma as punctum subdividens. Differently to the others the name ends with the letter S which explicitly indicates that Hilarvs was a slave working in the manufacture of Gavivs (Berti 1995: 48).

We are grateful to Marina Volonté for her indications related to the archeological contexts. Both fragments are residuals. Inv. 145007 comes from a filling post 69 AD, while Inv. 161697 was found in a late antique layer. 1. 2. 3. 4. 5. 6. 7.

[A]CO H[I]LARVS Oleggio (NO), no frame preserved, Inv. 54977. HILAR.ACO Milano (MI), no frame preserved, Inv. -. [HILA]RVS.ACO Milano (MI), no frame preserved, Inv. -. HILA[RVS Cremona (CR), ] left part of the frame preserved. Stylized turtle and a mask, Inv. 145007. HIL[ARVS] Milano (MI), no frame preserved, Inv. 62033. HILARVS.GAVI.S. Ottobiano (PV), framed by a female portrait on left side and a smaller male portrait on the right, Inv. 34463. [HILARV]S.GAVI Cremona (CR), no frame preserved Inv. 161697.

The three Comacchio beakers bear no coating. It is interesting to notice that the specimen with the most complex decoration is considerably larger than the other two, only half its size. Since the samples are completely preserved, it is possible to observe the triangular lower ending of the Kommaregen. 1. [HILA]RVS.GAV[I] mask, Abano Terme (PD), Inv. 6488. 2. HILARVS.GAVI framed by two double masks Comacchio (FE), Inv. 57350. 3. HILARVS.GAVI, Comacchio (FE), Inv. 57146. 4. HILARVS.GAVI.S, Comacchio (FE), Inv. 55916.

Evidence from the Eastern side of the Po Valley In the eastern Po Valley area, between Verona in the west and Trieste in the east, two contexts bear evidence of Aco beakers signed by Hilarvs.

HILARVS cups from Magdalensberg The 16 beakers where the name of Hilarvs appears found on the Magdalensberg are very different from those found in northern Italy. Contrasts relate mainly to size, proportions and capacity, but concern also signatures and decoration patterns.

Abano Terme (PD) A fragment with the letters [HILA]RVS.GAV[I] is attested among the materials of the ‘Stipe del Montirone’. It emerged in 1951 during excavations carried out for the construction of the foundations of the Due Torri - Morosini Hotel in Abano Terme (PD) (Figures 9 and 10 n. 11, Table 1) (Lavizzari Pedrazzini 1987: table 19, n. 2; 1995: table 3, n. 4-5). The rim and the upper portion of the wall of the uncoated beaker are preserved. A very blurred stylized mask divides the two parts of the name, applied under the frieze made of a chain of heart-shaped elements alternating with buds.

A similar disparity is noted with the materials from Lyon, St. Romain en Gal and Vienne, but this is not particularly surprising, since regular trade with Gaul was restricted to very special bronze objects. Other imports from Gaul that reached the Magdalensberg are extremely scarce, to be classified in the category of articles brought privately.

182

HILARVS, from the Po valley to the Rhone valley?

The Hilarvs Aco goblets from the Magdalensberg therefore show only superficial analogies with the individual motifs of the friezes produced in the Rhone valley, decorations for which an experienced potter would have had no difficulty even without a precise reference model.

density and position in the mould shows quite often a rather sloppy, careless way in the making.

There are four, maximum five signatures attested

Hilarvs Gavi also seems to prefer Kommaregen. Nevertheless one very badly preserved sliver signed [Hilarvs] GAVI (Figure 11 and Table 2), keeps part of a round metope decoration (?), perhaps to be interpreted as a single ‘pseudo-architectonic’ arch. Unfortunately, this fragment has currently disappeared, and only the drawing can be provided.

HILARVS

Friezes and delimitation motifs

Signatures

GAVI

possible

suspected

ACO HILARVS HILARVS ACO HILARVS GAVI

Only phytomorphic friezes are testified. They mostly consist in very thin, delicate and graceful garlands, like single chains of rounded leaves, more or less stylized, oriented either to the left or to the right. Ivy and vine leaves are prevailing motifs, at times alternated with intermediate points and stylized ram’s horns, but there are also friezes with symmetrical double leaves, as well as rosettes with eight petals, tongs and heart-shaped ornaments. In one case, supplementary stamps appear inserted in the komma decoration.

For methodological reasons, the Magdalensberg team refrains from making attributions when the manufacturers name is missing. Therefore it is possible that some of the numerous fragments without remaining letters may belong to the Hilarvs group. Analysis

Triangular dots and small groups of komma can be used for dividing the parts of the name, even if some signatures do not have separators. Several elements placed on both sides of the signature are attested. Horizontal lotus flowers and eight-leaf rosettes seem limited to the Hilarvs group, while Hilarvs Gavi can be associated with a small helmet-shaped head.

Archaeometry identified two distinct clay groups in a first series of chemical analyses (Schneider 1998): Hilarvs form a first group, while Hilarvs Gavi belong to a second group. Both are clearly different from the Lyon group. However, since the fragmentation of the signed specimens did not allow extensive sampling, the number of samples is quite small and perhaps not very representative.

In the only intact piece, the Kommaregen is continuous to the bottom. There is no evidence of Hilarvs nor Hilarvs Gavi decorated in a triangular pattern with or without interspersed stamps.

Dimensions and shapes Only beakers with a maximum capacity of 2 deciliters have been identified on the site. Compared to those common in Italy they are rather small. Shapes are restricted exclusively to cylindrical vases with thin walls, relatively low necks and small rounded rims. Necks appear often slightly thickened. Unlike those of the Comacchio shipwreck or some originating in the Rhone valley, goblets are always slender, with a proportion between height and width of at least 2 to 1.

Signatures Style, position and size of the letters seem to be good features to distinguish the productions of the various ateliers. In Hilarvs Gavi’s signatures the letter A never has a broken central bar. Moreover, in these goblets, the letters slightly inclined to the left all have the same size, while in Hilarvs – alone or in connection with Aco – the letters appear straight or with a meek inclination to the right. Hilarvs Aco frequently uses a final S larger than the other letters.

No indications help to define the pattern of the Kommaregen towards the foot, and the only complete beaker keeps the Kommaregen covering the complete wall.

Chronology

Decorations and style

The Magdalensberg, as a consumer site, does not offer any conclusive ideas for the start of the production. The majority of the finds, i.e. eight out of 16 beakers, come from the enormous leveling layer OR-Hang hinter OR/18, 19, closed no later than 10 BC. (Ardis et al. 2019: fig. 6).

Aco beakers signed Hilarvs and those where the name appears combined with Aco are always decorated with Kommaregen under a single-rowed, rarely double frieze. The insertion of the komma of different sizes, 183

Armand Desbat, Valentina Mantovani, Eleni Schindler Kaudelka, Roby Stuani

Figure 11. Friezes and signatures of Hilarvs attested in the Noric emporium on the Magdalensberg (drawings: E. Schindler Kaudelka).

184

HILARVS, from the Po valley to the Rhone valley?

Figure 12. Friezes and signatures of Hilarvs attested in the Noric emporium on the Magdalensberg (drawings: E. Schindler Kaudelka).

185

Armand Desbat, Valentina Mantovani, Eleni Schindler Kaudelka, Roby Stuani Two beakers of the Hilarvs group were dumped in the ‘Mosslergrube’ excavated during the construction of the provincial road. The context is inserted in period 1, phase 2. These artefacts were already broken by 10 BC.

absent in horizon 1A, contemporary to the foundation of Lyon, Aco beakers first appear in horizon 1B where they represent 25% of the thin-walled vessels (in NMI) (Figure 8). Their number decreases rapidly in Horizon 3A, around 10 BC, and they are almost absent from horizon 3C (around 10 AD) – represented by the ‘fosses à banquet’ – where the thin-walled vessels add up to a total of 1013 fragments equal to 134 NMI.

The remaining Hilarvs and Hilarvs Gavi vessels were unearthed in the Augusteische Planierschicht des Südhanges (Schindler Kaudelka 1998) and belong the earliest group of the deposit, dated between 40/35 BC and 10 BC (Baur and Schindler Kaudelka 2015).

The first sign of ACO HILARVS and HILARVS ACO beakers on the Magdalensberg dates between 30/25 and 10 BC and this fits perfectly into the dates offered by the Loyasse and Pseudo-Sanctuaire de Cybèle materials – between 30 and 20/15 BC. The contexts of the Noricum emporium provide the same date for the products of HILARVS GAVI, in accordance with the exemplar with the HILARVS GAVI.S. beaker from Ottobiano.

The chronological evidence, unlike what seemed valid until 1998, compares and overlaps perfectly with the dating of the Comacchio shipwreck and the chronology of the Pseudo-Sanctuaire de Cybèle in Lyon. Elements for a summary A very high number of products from the Rhone valley, with a rather wide distribution, contrasts with the substantially smaller group of Aco Hilarvs/Hilarvs Aco manufactured in one or more as yet unknown places in the Po Valley.

This could support Lavizzari’s hypothesis. However, strong doubts arise when looking at the dating of the Valle Ponti/Comacchio shipwreck. The context provides a date between 19 and 12 BC, much more precise than Gavivs on the Magdalensberg and in Montegrotto/ Stipe del Montirone. (Figure 14 and Table1). Moreover, this date is not linked to a consumer but to a trading context. Is it possible to imagine the products of Hilarvs Gavi still on sale in Italy a few decades after his move to the Rhone valley? Logic and plausibility indicate that Gavivs and his slave Hilarvs never moved from northern Italy, where the workshop continued their activity until mid-Augustan times.

While ACO HILARVS and HILARVS ACO appear exclusively in the area between Oleggio and Milan, in the eastern Po Valley area only HILARVS GAVI or HILARVS GAVI.S. is evident. Products of Aco Hilarvs, Hilarvs Aco and perhaps Hilarvs are attested on the Magdalensberg, with a smaller evidence of Hilarvs Gavi. Hilarvs Gavi and Hilarvs Aco seem to have different distribution patterns, but both variations reached the Noricum emporium.

In this case, the Po Valley Hilarvs and the Gaulish Hilarvs have to be two different men.

For Gavivs and his workshop, where Hilarvs at least in the cases of Ottobiano and Comacchio declares himself to be a servvs, Fede Berti (1995: 47) and Paola Lavizzari (1997: 246) proposed an Eastern Po Valley origin, based on the distribution of the few specimens. Lavizzari, in particular, taking up previous hypotheses (Desbat 1990: 44; Schindler Kaudelka 1980: 78) suggested that Gavivs’s Po Valley workshop with its task force had moved to the Rhone valley where Hilarvs had started his own business after his manumission. Later he would have joined the branches of Aco. In order to work around the complete lack of traces of Gavivs in the Lyon area, Lavizzari (1997: 246) tries to unify the eight specimens signed T.CAVIVS of La Muette with the Gavivs formerly active in the Po Valley area.

Some further arguments may emerge from the study of the decorative style and the dies as well as the letters. Hilarvs Gavi is always written in thin letters, without apices, and the letter A always has a straight middle dash. Hilarvs in the Rhone valley productions, signs in a more stubby way, with apices, and the letter A always has a broken middle dash. Frieze decoration of Gavivs, as well as that of the Rhone valley Hilarvs, are composed in single or double rows. Motifs before and after the letters offer another indication. Lyon, Vienne/St. Romain en Gal use of rosettes, dolphins, shells and human profiles. The latter already common in other Aco workshops are used by the Gavius atelier, but the heads and turtles used, although similar, are not completely identical.

To be valid, such a hypothetical construct must set a possible transfer of the workshop of Gavivs and his slave Hilarvs to the Rhone valley before 30 BC, since the oldest workshops of the Lyon and Vienne Hilarvs started around this date. Excavations in the PseudoSanctuaire de Cybèle provide sound evidence for the dating of the Rhone valley Hilarvs (Desbat 2012). Still

Some similarities between some of the stamps used for decorating beakers present on the Magdalensberg and Hilarvs made in the Rhone, which would point to imports. The frieze of Magdalensberg M66 OR-Hang 186

HILARVS, from the Po valley to the Rhone valley?

Figure 13. Hilarvs in the Noric emporium in the Magdalensberg. The numbering follows the order of Tab. 2 (photo: V. Mantovani).

187

Armand Desbat, Valentina Mantovani, Eleni Schindler Kaudelka, Roby Stuani

Figure 14. Timeline for Hilarvs.

(Figure 13.3 and Table 2) shows a die similar to one used in Loyasse and on a second vase of T. Cavivs (frieze 5 of La Muette). The same on M67 OR/19 (Figure 13.1 and Table 2), analog to frieze 14 of La Muette, and M81 SH/7 (Figure 12 and Table 2) recording frieze 34 of Philarcvrvs from la Muette. However, two of them have been analysed (Schneider 1998: samples number F482; Schindler Kaudelka 1980: Taf. 28/32, F477 Taf. 35/95) and their Italic origin is confirmed.

where two groups are present. The earlier one with rather thin walls and thin translucent slip appears albeit timidly in context 2. Those with a thicker wall and a thicker coating show up in the contexts 4 and 5 where they are in all probability residual (Figure 13.1). The Magdalensberg material produced one sample with a honey coloured lead glaze (Figure 13.2), normally known from Acastus wares exclusively (Schindler Kaudelka et al. 2017). Beakers with a partly dark smoky surface are quite common, but this does not look like a coating. It has to be considered as a result of staking and uneven oxygen atmosphere in the kiln (Figure 13: 3, 4, 7, 8 10) In the Po Valley, on a general level, the presence of red slip seems once again to belong to the latest examples (Conspectus 1990: 182; Mantovani 2015: 67). This seems confirmed by the absence of any slip in all the specimens signed by Hilarvs Gavi, Hilarvs and Aco Hilarvs from Italy.

From the material attested in the Po Valley at least two of the specimens from Milan (Figure 10. n. 5 and 6, and Table 1) show affinities to the Rhone Valley repertoire, in particular concerning the style of the letters where one recalls Vienne’s type n. 7. The Rhone Valley products signed by T. Cavivs, suggested by Paola Lavizzari to have been the immigrant Gavivs from the Po Valley, show simple, very thin letters, without apices as well as small heads as delimiting elements. On the other hand, T. Gavivs’ letter A corresponds to those used by the Rhone Valley Hilarvs with apices and central broken dash.

To sum up this latest panorama rising after the review of analogies, similarities, differences and discrepancies between the materials examined, the results are ambivalent. There are more open issues than firm achievements.

With the specimens available, a clear-cut definition in the styles is missing. Therefore unequivocal elements to distinguish Gavivs active in the Po Valley from T. Cavivs of La Muette cannot be provided.

The comparison between the Rhone Valley production and the few examples manufactured in Northern Italy and used there as well as those exported to Noricum did not produce comprehensive data for the nature of their connection. The links between the two production areas remain vague, blurred and indistinct. A more or less consistent chronology starting in early Augustan

Aco beakers with red Glanzton slip appear later in Lyon and Vienne St. Romain, as they do on the Magdalensberg 188

HILARVS, from the Po valley to the Rhone valley?

times was confirmed. Superficial stylistic homogeneity is not a conclusive argument, since any expert potter was able to reproduce the delicate, nevertheless rather simple decorations easily. No compulsory connections between different ateliers have to be created. The look of fashion and the spirit of the age can be found all over the Mediterranean, even Asia Minor where similar looking pottery was made in Augustan times.

Ceresa Mori, A. 1991. Ceramica a pareti sottili, in D. Caporusso (ed.) SCAVI MM3. Ricerche di archeologia urbana a Milano durante la costruzione della Linea 3 della Metropolitana (1982-1990), Vol. 3.1: 41‒56. Milano: Edizioni Et. Deodato, A. and L. Poletti Ecclesia 1999. Le tombe e i loro corredi, in G. Spagnolo Garzoli (ed.) Conubia gentium: la necropoli di Oleggio e la romanizzazione dei Vertamocori (Catalogo della Mostra, Palazzo Bellini, Oleggio — 23 gennaio-30 aprile 1999): 57‒286. Torino: Omega. Desbat, A. 1985. L’atelier de gobelets d’Aco de SaintRomain-en-Gal (Rhône) Etude préliminaire. SFECAG. Actes du Congrès de Reims: 10‒14. Desbat, A. 1990. Les ateliers gaulois de gobelets d’Aco. Archéologia 262: 42‒47. Desbat, A. 1995. Les productions précoces de céramiques à glaçure plombifère de la vallée du Rhône. Rei Cretariae Romanae Favtores Acta XXXIV: 39‒47. Desbat, A. 1996. Les gobelets d’Aco, in A. Desbat, M. Genin and J. Lasfargues (eds): 68‒99 et Pl. 40‒77. Desbat, A. 2000. Les ateliers lyonnais et leurs rapports avec les ateliers padans, in G.P. Brogiolo and G. Olcese (eds) Produzione ceramica in area padana tra il II secolo a.C.e il VII secolo d.C.: nuovi dati e prospettive di ricerca (Documenti di Archeologia 21): 79‒92. Mantova: Società Archeologica Padana. Desbat, A. 2012. Lyon, Lugdunum: Les contextes précoces des fouilles du ‘Sanctuaire de Cybèle’, in P. Barral et S. Fichtl (eds) Actes de la table ronde tenue à Bibracte, Glux-en-Glenne, 15-17 octobre 2007 (Bibracte 22): 65‒94. Glux-en-Glenne: BIBRACTE, Centre archéologique Européen. Desbat, A. 2019. La chronologie des gobelets d’Aco de Lyon et de Vienne, in H. Dolenz et K. Strobel (eds) Chronologie und vergleichende Chronologien zum Ausgang der Römischen Republik und zur Frühen Kaiserzeit, Klagenfurt: 119-134. Desbat, A., M. Genin and J. Lasfargues (eds) 1996. Les productions des ateliers de potiers antiques de Lyon, I. Les ateliers précoces. Gallia 53: 1‒250. Desbat, A. and M. Genin 1996. Les ateliers précoces et leurs productions, in A. Desbat, M. Genin and J. Lasfargues (eds): 219‒241. Lasfargues, J., A. Lasfargues, J. Poncet and H. Vertet 1970. Découverte de deux fragments de gobelets à parois fines décorés à Roanne. RAE XXI/1-2: 221‒222. Lasfargues, J. and H. Vertet 1968. Observations sur les gobelets d’Aco de l’atelier de la Muette à Lyon. R.A.C. VII/1: 35‒44. Lavizzari Pedrazzini, M. P. 1983. Due biccheri ‘tipo Aco’ dalla necropoli di Giubiasco. Archéologie Suisse 14: 54‒63.  Lavizzari Pedrazzini, M. P. 1987. Ceramica romana di tradizione ellenistica in Italia settentrionale. Il vasellame ‘tipo Aco’. Firenze: La Nuova Italia. Lavizzari Pedrazzini, M. P. 1995. Il deposito del Montirone. Quaderni di Archeologia del Veneto XI: 109‒166.

The ground remains shifting when looking at the names. Hilarvs, a quite common slave name in the Roman world offers no indication of a possible univocal nature. Right in Augustan times, the name Hilarvs appears on a number of Sarius cups distributed in the regio decima italica and beyond. The vessels reached Liburnia and Dalmatia, Pannonia as well as the Magdalensberg (Mantovani 2015: 89; Lavizzari Pedrazzini 2010: 91). Cups, plates and platters in plain sigillata made in Arezzo, in the Gulf of Naples and in the Po Valley are stamped Hilarvs. There are single peregrinus (?) names (OCK 949-953), and also slaves and freedmen who sign combined with the name of a dominus or a patronus, (OCK passim). There is not even a remote possibility to understand whether any of these persons were connected to each other. In the context covered by this discussion, Hilarvs could be considered a single person working in the Po Valley and then responsible for a branch set up in the Rhone Valley. Nevertheless, the possibility remains that we are in front of an unknown number of different people named Hilarvs, who were active independently in the last quarter of the 1st century BC in different territories. Partly supported by the study of stylistic elements and decorative differences, which is always a very hazardous and unsafe ground, it can be suggested we are in presence of three main actors, two Italians Hilarvs Gavi servvs and a Hilarvs signing either Aco Hilarvs or Hilarvs Aco, as by the way does Acastvs, and one (or two?) Hilarvs living in Lugdunum and/or Vienne. Whether this Hilarvs Aco who worked in the Rhone Valley was the manager of a subsidiary, an independent immigrant or a native with a Latin name, remains open for discussion. References Barone, A. M. 1999. Funzionalità e raffinatezza. La ceramica a pareti sottili, in G. Spagnolo Garzoli (ed.) Conubia gentium: la necropoli di Oleggio e la romanizzazione dei Vertamocori (Catalogo della Mostra, Palazzo Bellini, Oleggio — 23 gennaio-30 aprile 1999): 321‒325. Torino: Omega. Berti, F. 1995. Osservazioni a margine di alcune ceramiche del relitto navale di Valle Ponti (Comacchio) e sul loro commercio, in B.M. Giannattasio (ed.) Viaggi e Commerci nell’Antichità, Atti VII Giornata Archeologica: 39‒155. Genova: D.AR.FI.CLET. 189

Armand Desbat, Valentina Mantovani, Eleni Schindler Kaudelka, Roby Stuani Lavizzari Pedrazzini, M. P. 1997. Nuove osservazioni sul vasellame ‘tipo Aco’. Athenaeum 85: 233‒251. Lavizzari Pedrazzini, M. P. 2010. Il vasellame ‘tipo Sarius’: ceramica romana di tradizione ellenistica in Italia settentrionale. Rivista Archeologica dell’antica provincia e Diocesi di Como 190: 67‒278. Mantovani, V. 2015. Ceramiche fini da mensa di Adria romana. Le indagini di via Retratto (1982 e 1987) (Collezioni e Musi Archeologici del Veneto 48). Roma: Giorgio Bretschneider editore. Marichal, R. and F. Mayet 1980. Céramiques à parois fines et gobelets d’Aco à Ruscino. Ruscino I (RAN Supplément 7): 264‒270. OCK — Comfort, H., A. Oxé and PH. Kenrick 2000. Corpus Vasorum Arretinorum. Bonn: Dr. Rudolf Habelt GMBH. Pisano, A. 2000. Milano Piazza Erculea: la ceramica fine da mensa, in Milano tra l’età repubblicana e l’età augustea: atti del convegno di studi, 26-27 marzo 1999, Milano: 301‒306. Milano: Grafiche Serenissima.  Roth-Rubi, K. 2006. Dangstetten III. Das Tafelgeschirr aus dem Miltärlager vom Dangstetten. Stuttgart: Theiss, Konrad. Schindler Kaudelka, E. 1980. Die römische Modelkeramik vom Magdalensberg (Kärntner Museumsschriften 62). Klagenfurt: Verlag des Geschichtsvereines fur Kärnten. Schindler-Kaudelka, E. 2000. Ceramica norditalica decorata del Magdalensberg: problemi aperti, in G.P. Brogiolo and G. Olcese (eds) Produzione ceramica in area padana tra il II secolo a.C.e il VII secolo d.C.: nuovi dati e prospettive di ricerca: 53‒78. Mantova: Societa Archeologica Padana.

Schindler-Kaudelka, E. 2002. La datation des premiers contextes du Magdalensberg. SFECAG. Actes du congrès de Bayeux: 263‒274. Schindler-Kaudelka, E. 2006. La ceramica norditalica decorata. A che punto siamo arrivati ? in S. Menchelli and M. Pasquinucci (eds) Territorio e produzioni ceramiche. Atti del Convegno Internazionale Pisa 20-22 ottobre 2005: 239‒244. Pisa: Plus. Schindler-Kaudelka, E., G. Schneider and S. ZabehlickyScheffenegger 1997. Les sigillées padanes et tardopadanes. Nouvelles recherches en laboratoire. SFECAG. Actes du Congrès du Mans 1997: 481‒493. Schindler Kaudelka, E. unter Mithilfe von G. Schneider 1998. Die römische Modelkeramik vom Magdalensberg 2, in G. Piccottini (ed.) Die Ausgrabungen auf dem Magdalensberg 1980 bis 1986 (MagdalensbergGrabungsbericht 16): 289‒378. Klagenfurt: Verlag des Geschichtsvereines fur Kärnten. Schneider, G. and M. Daszciewicz 2006. Chemische Analysen zum Tafelgeschirr aus dem Militärlager von Dangstetten, in Roth-Rubi 2006: 169‒193. Vannacci Lunazzi, G. 1987. La necropoli romana di Ottobiano. RAC 168: 47‒103. Volonté, M. 2007. Nuovi rinvenimenti di Acobecher a Cremona, in F. Butti Ronchetti (ed.) Produzioni e commerci in Transpadana in età romana, Atti del Convegno (Como, 18 novembre 2006) (CD Rom). Como. Volonté M. 2018. Ceramica decorata a matrice, in L. Arslan Pitcher, E.A. Arslan, P. Blockley and M. Volonté (eds) Amoenissimis aedificiis Lo scavo di piazza Marconi a Cremona vol. 2 (I materiali): 203‒214. Quingentole (MN): Società Archeologica Padana.

190

Terre sigillate galliche. La collezione “storica” del Museo Archeologico Nazionale di Aquileia (Friuli Venezia Giulia – Italia) Patrizia Donat Abstract The paper discusses the Gaulish Terra Sigillata of the "historical" collections of Museo Archeologico Nazionale of Aquileia. The collection contains 194 vessels, 149 of which are made in Southern Gaul, 30 in Central Gaul and 5 in Eastern Gaul (atelier of Saturninus and Satto; Rheinzabern); for six vessels it was not possible to identify the origin, while four other with a surface without figurative decoration may have been produced in Central or in Eastern Gaul (Figura 1). The repertoire consists mainly of cups Drag. 37 (176 cups) and Drag. 29 (9 cups); undecorated vessels or vessels with barbotine and carving decoration are documented sporadically (Figura 2). The terra sigillata of La Graufesenque definitely prevails within the production southern gaulisch production (62 %); less present is the pottery of Banassac (15 %); the remaining 22 % of the vessels has been generically attributed to production of southern Gaul (Figura 4). Most pottery with stamps has been known since the initial studies of the class, dating back to the early twentieth century; exceptions are a cups Drag. 37 with a stamps of the potter Rosette I (South Gaulish Terra Sigillata; Tavola 1: 4) and another stamps of Albucianus (Central Gaulish Terra Sigillata; Tavola 3: 3). The last potter had already been recognized within the museum collection, while the stamp Rosette I is known thanks to the findings made in the excavations in the area of Canale Anfora (Aquileia). A first shy start of imports to Aquileia was fixed, in the Tiberian age, with a modest intensification in the Neronian-Flavian age, when even the oldest pottery belonging to the collection was given. The peak of presences, however, is reached between 80 AD and the middle of the second century AD. This is testified both by the materials in the Museum's collection and by more recent excavations. In the Trajanic-Adrian age, Aquileia reached, although in smaller quantities, along with the pottery of La Graufesensenque, also that of Banassac and the first ceramists of central and eastern Gaul. The collection, as well as those from the excavations, show that imports from central Gaul continued, never more rarefied, even during the second half of the 2nd century AD; to these, however, are added, in sporadic form, those from Rheinzabern. This analysis of the "historical" collections of Gaulish Terra Sigillata, preserved in the National Archaeological Museum of Aquileia, highlighting the large number of finds kept there, makes it possible to definitively archive the hypothesis that imports from Gaul are sporadic evidence. Comparison with the evidence from archaeological excavations gives a better calibration of this consideration, leading to the conclusion that the influx of goods from the Gaulish provinces was limited (3 % - 5 % of all fine ware), but not sporadic. It would seem that the ceramics arrived in Aquileia through the Alps, the Po and the Upper Adriatic endolagunar route. The concentrations of material recovered at Canale Anfora and at the mouth of the TImavo seem to confirm this hypothesis and at the same time indicate a possible existence of privileged terminals for materials coming from Gaul. The research will focus on ceramists in the future, to try to understand whether certain potters, unlike others, may have chosen to expand their network of trade also to the market of northern Italy and in particular Aquileia.

Key words “Historical” collection of Museo Archeologico Nazionale di Aquileia (Friuli Venezia Giulia - Italia); Gaulish Samian Ware; gaulish potters; distribution in Aquileia and her ager.

Lo stato della questione Da poco più di una decina d’anni gli studi sulle terre sigillate galliche da Aquileia e dal suo territorio si sono intensificati dopo decenni di silenzio. La presenza di tali produzioni nella città altoadriatica è nota fin dall'inizio del XX secolo grazie agli studi di Robert Knorr (Knorr 1912: 43-44, 50, Taf. XXV: 6, Taf. XXX: 4-7); si tratta di reperti conservati all'epoca nell'ImperialRegio Museo della località. Molti di questi oggetti erano già precedentemente esposti nella raccolta del Museo comunale, istituzione attiva dal 1873 al 1882; nel 1883 sono confluiti in deposito all'Imperial-Regio Museo,

che con il passaggio all'Italia divenne l'attuale Museo Archeologico Nazionale di Aquileia.1 Dopo una lunga stasi nelle ricerche, all'inizio degli anni Ottanta, Franca Maselli Scotti analizzò un campione di meno di 20 esemplari, conservato nel Museo Archeologico Nazionale e comprensivo anche di quanto già edito da Knorr.2 Sulla base di tale lavoro si è ritenuto Per la storia della raccolta del Museo Archeologico Nazionale di Aquileia: Giovannini 2004; 2007. Ancora oggi alcune delle coppe (T. 1: 2-3; T. 2: 3-4), tra cui quelle pubblicate da Knorr, conservano i supporti in filo di stagno con cui erano esposte nelle vetrine fin dai primordi del percorso del Museo (Maionica 1911: 94, 104; Brusin 1929: 205-210; entrambi citati in: Ventura e Zulini 2012-2013: 371, nota 3). 2  Maselli Scotti 1981; 1984; 1987; 1988. Pochi altri esemplari sono stati 1 

roman pottery and glass manufactures (Archaeopress 2022): 191–202

Patrizia Donat a lungo che la presenza della importazioni galliche ad Aquileia fosse assai poco significativa, anche perché la studiosa, pur sottolineando come la sua pubblicazione non fosse esaustiva, non ha indicato il quantitativo totale del vasellame custodito nel Museo (Maselli Scotti 1981: 243, nota 1). I materiali pubblicati da Knorr e Maselli Scotti sono stati negli ultimi anni nuovamente ripresi in alcuni recenti studi (Donat 2015; Gabucci 2017).

operare dei confronti basati su dati omogenei e perché, inoltre, nell’ultimo decennio il panorama relativo alle importazioni dalla Gallia è in continua evoluzione.7 Si possono tuttavia proporre alcune riflessioni. Appare innanzitutto verosimile supporre che nelle collezioni storiche del Museo le ceramiche galliche siano documentate in numeri così elevati rispetto alle altre classi, perché all'epoca si selezionavano i reperti sulla base di criteri estetici, ed è indubbio che la decorazione a matrice delle coppe tipo Drag. 29, Drag. 30 e Drag. 37, che costituiscono il 96 % della collezione, abbia la capacità di attirare interessi di tipo collezionistico. Questa ipotesi è suffragata anche dalle quasi inesistenti presenze di orli e fondi non decorati riferibili a questi tipi. La scarsità del vasellame non decorato, non può invece essere ricondotta al medesimo motivo, perché questa mancanza è una costante nel panorama aquileiese. Un criterio estetico potrebbe essere stato seguito, almeno parzialmente, anche negli scavi del 1988 a Canale Anfora, dove le terre sigillate galliche costituiscono il 15% di tutte le ceramiche fini, raggiungendo quantitativi simili alle sigillate italiche e orientali.8 Nello scavo del 2004-2005, operato con criteri stratigrafici in un'area molto più ristretta del Canale Anfora, le terre sigillate galliche costituiscono il 5% di tutte le ceramiche fini, una percentuale molto più contenuta, ma che non può tuttavia essere considerata insignificante. Si potrebbe, di conseguenza, ipotizzare che perlomeno alcuni settori dell'area portuale di Canale Anfora abbiano rivestito un ruolo particolare nello stoccaggio, nella commercializzazione e nella distribuzione di questo vasellame, visto anche che gli scavi nell'area settentrionale del Porto Fluviale hanno restituito un quantitativo nettamente inferiore di ceramiche fini galliche; tuttavia non va dimenticato che alcuni reperti da Canale Anfora sembrerebbero provenire da vicini contesti abitativi (Donat et al. 2017: 65-66; Maggi e Merlatti 2015: 447-448, nota 49, fig. 13).

La revisione, avviata di recente, delle collezioni museali “storiche” di ceramiche fini,3 in cui questo lavoro si inserisce, ha permesso di individuare poco meno di 200 recipienti in terre sigillate galliche, il 90% dei quali è rimasto fino a oggi inedito. Vista l'antichità delle collezioni, benché vi sia la certezza, che i reperti siano di provenienza aquileiese, non è stato possibile risalire al preciso contesto di rinvenimento.4 Nel caso delle terre sigillate galliche si tratta di una raccolta molto numerosa, se si pensa che le sigillate africane costituenti la raccolta “storica” del Museo di Aquileia sono composte da appena un centinaio di esemplari (Ventura e Zulini 2012-2013: 374-376; Zulini 2015: 183; Zulini in questo volume) e che le stesse considerazioni si possono fare sui rapporti esistenti tra il vasellame gallico e quello orientale (Maggi in questo volume). Questo dato non corrisponde a quanto documentato nelle indagini di scavo (Donat et al. 2017); in alcune zone della città, infatti, le produzioni galliche sono del tutto assenti,5 mentre in altre la loro presenza è molto contenuta.6 Un'eccezione è costituita dagli scavi effettuati prima (1988) da Luisa Bertacchi (Donat et al. c.s.) e poi (2004/2005) da Franca Maselli Scotti (Maggi et al. 2017) nell'areale di Canale Anfora, che spicca come la zona di Aquileia, da cui proviene il quantitativo maggiore di terre sigillate galliche (Donat 2015: fig. 5-9: 11, 12-14; Donat et al. c.s.; 2017: 66, nota 7-10). Non è possibile attualmente trovare una causa per le differenze nelle attestazioni, perché non si possono poi recentemente editi da Ada Gabucci, che in merito al quantitativo complessivo dei materiali conservati nel Museo aquileiese parla genericamente di un paio di casse: Gabucci 2017: 9.2.3, fig. 71, nota 79. 3  Per la terra sigillata orientale: Maggi e Ventura in questo volume. Per la terra sigillata africana: Ventura e Zulini 2012-2013; Zulini 2015; Zulini in questo volume. 4  I reperti, una sessantina dei quali conserva ancora il numero d'inventario, indicazioni sulla collezione di appartenenza o la collocazione museale, sono in parte descritti nell'inventario, compilato immediatamente dopo il passaggio del Museo dall'amministrazione austriaca a quella italiana. Nell’inventario non viene, però, mai riportata la località di rinvenimento; spesso, inoltre, è risultato impossibile riferire la descrizione fatta nell'inventario ad uno specifico reperto: Ventura e Zulini 2012-2013: 374-376. 5  Area a est del foro (edificio ad impianto basilicale) Verzár Bass (ed.) 1991; 1994; insula tra foro e porto fluviale: Fontana (ed.) 2017. 6  Area a est del foro (domus): Buora et al. 1995: 142, tav. XVI; area di via Bolivia: Ceazzi e Del Brusco 2007: 122; Fondi ex Cossar; Bonetto et al. 2013: 158, tab. 6, tav. 2: 2; area a nord del Porto Fluviale: Maggi e Merlatti 2015: 447, nota 49; necropoli della Beligna: Donat 2015: 45, fig. 3: 10.7; Donat et al. 2017: 66, nota 7-10.

Nel 2015 un articolo di Paola Maggi e Renata Merlatti documentava nell'intera Aquileia 50 frammenti di sigillate galliche (Maggi e Merlatti 2015: 447-448, nota 49, fig. 13), mentre nello stesso anno un lavoro della scrivente riferiva di 331 pezzi (Donat 2015: 39), oggi il numero di attestazioni si è quasi raddoppiato (Donat 2017). 8  Da questa indagine, che ha interessato un'area piuttosto estesa della zona di Canale Anfora, vengono 102 frammenti di terre sigillate galliche; si tratta del quantitativo più importante mai recuperato ad Aquileia in un'unica area: Donat et al. c.s. 7 

192

Terre sigillate galliche. La collezione “storica” del Museo Archeologico Nazionale di Aquileia

Figura 1. Rapporto percentuale delle diverse produzioni rappresentate.

Terre sigillate galliche dalle collezioni “storiche” del Museo Archeologico Nazionale di Aquileia: quantità, stato di conservazione dei reperti, produzioni, tipologia

dire se tale caratteristica possa essere causata dalla giacitura o se invece sia la spia di luoghi di fabbricazione diversi, rispetto a quelli determinabili attraverso la punzonatura (Gabucci 2017: 9.2.3).

Lo studio, ancora in corso, presenta un primo tentativo di sintesi sulla collezione “storica” di terre sigillate galliche conservate nel Museo Archeologico Nazionale di Aquileia. Esso prende in considerazione sia il materiale edito in precedenza da Knorr, Maselli Scotti e Gabucci sia quello inedito, che costituisce la maggior parte del complesso ceramico; si tratta, infatti, di 194 esemplari, di cui è edita finora poco più di una ventina di vasi.

Come già evidenziato da precedenti lavori su Aquileia, Trieste e i loro territori (Donat 2015; 2017), anche in questo caso le produzioni del Sud della Gallia (Figura 1) sono predominanti (149 esemplari; 77 %); solo discretamente attestato è il vasellame dalla Gallia centrale (30 esemplari; 15 %). Una novità, rispetto a quanto già edito (Maselli Scotti 1981: 251-252, Tav. IV: 1-4, 6) è costituita dall'individuazione di due coppe frammentarie tipo Drag. 37, realizzate nelle officine della Gallia dell'Est di Saturninus e Satto (1 %); non è stato possibile, invece, determinare l'esistenza di nuovi recipienti importati da Rheinzabern (2 %). Il rimanente 5 % è stato dubitativamente attribuito da un lato alle produzioni della Gallia meridionale o centrale (3 %) e dall'altro a quelle della Gallia centrale o orientale (2 %).

Il vasellame integro si conta sulle dita di una mano; tuttavia la maggior parte dei reperti, pur essendo frammentata, non ha quasi mai dimensioni minime e sono pochissime le porzioni di vaso non decorate o che conservano decorazioni poco leggibili; è verosimile pensare, di conseguenza, che i manufatti siano stati recuperati in giacitura primaria e che quelli integri fossero parte di un corredo funerario.

Il vasellame non decorato a matrice è solo sporadicamente rappresentato (5 %). Ciò probabilmente anche a causa di una selezione operata a favore dei vasi decorati o bollati nel momento in cui la collezione museale si è formata; tutta la ceramica non lavorata a matrice ha infatti una decorazione incisa, excissa, alla barbotina o presenta marmorizzazione o marchi (Figura 2). I recuperi fatti nei contesti di scavo, come ad esempio le indagini effettuate a Canale Anfora tra il 2004 e i 2005, mostrano, comunque, che il mercato aquileiese ha sempre preferito le decorazioni a matrice. Tale fenomeno, che trova riscontro in generale ad

La classificazione del materiale è stata fatta in base alla tipologia e allo studio dei punzoni decorativi e dei bolli; per la terra sigillata liscia non è stato perciò possibile proporre un’unica area di fabbricazione. Quasi tutti i reperti, indipendentemente dalle produzioni, presentano la caratteristica vernice compatta e brillante; tuttavia alcuni frammenti si discostano dagli altri a causa di una vernice, che si scrosta facilmente. In assenza di analisi archeometriche non è possibile 193

Patrizia Donat

Forma e tipo

Quantità

le importazioni delle terre sigillate africane e orientali (Donat 2017: 101-102).

Coppa Dragendorff 29

9

Le terre sigillate sudgalliche

Coppa Dragendorff 37

176

Come già accennato, analogamente a quanto noto dalle indagini di scavo e dagli altri contesti aquileiesi (Donat 2015; 2017), le ceramiche sudgalliche costituiscono la produzione assolutamente più testimoniata: con 149 frammenti rappresentano quasi 77 % di tutte le terre sigillate galliche (Figura 1).

Tipi decorati a matrice Coppa Dragendorff 30

1

Tipi non decorati a matrice Forma e tipo

Quantità

Coppa a listello Ritterling 12 marmorizzata

1

Coppa Dragendorff 33

1

Piatto da portata Dragendorff 39

1

Coppa Dragendorff 27

Coppa a listello Dragendorff 38 Bicchiere/olla Dragendorff 54

1

Le prime importazioni, documentate da una coppa a listello marmorizzata Ritterling 12 (40-70 d.C.)9 e da quattro coppe Drag. 29 (Maselli Scotti 1981: Tav. I: 3; Tav. 2: 4), con decorazioni a girali e/o festoni, sono tutte riconducibili all'età claudio-neroniana o alla prima età flavia. Tra i ceramisti più antichi sono stati riconosciuti (Figura 3), sulla base della decorazione, Masclus (50-70 d.C.; T. 1: 1) (Gabucci 2017: Fig. 71: 2) e MeΘillus (60-80 d.C.).10 Dai precedenti studi sappiamo che i primi sporadici arrivi di ceramiche galliche in area nordadriatica si datano in età tiberiana; tuttavia è in età neroniana e nella prima età flavia, che sembra essere cominciato un approvvigionamento contenuto, ma abbastanza costante (Donat 2008: 162-165; 2015: 3943, 2017: 108). Per quanto riguarda il materiale oggetto di questo studio il rapporto tra le presenze (Figura 2) delle coppe Drag. 29 (9 esemplari; 8 %) e delle più tarde coppe Drag. 37 (137 esemplari; 92 %), tutto a favore di queste ultime, così come la quasi totale assenza delle produzione marmorizzata (1 unico frammento) sono un forte elemento a sostegno di una consistente intensificazione degli scambi con la Gallia meridionale negli ultimi due decenni del I sec. d.C. Questo tipo di situazione trova pieno riscontro nel panorama di Aquileia, Tergeste e dei loro territori (Donat 2015: 45); l'unica eccezione è costituita dai ritrovamenti fatti nell'area delle foci del Timavo, dove il tipo Drag. 29 costituisce ben il 36 % dei ritrovamenti, indiziando per questa zona un inizio delle importazioni in linea con quello del restante territorio, ma un più consistente afflusso di merci già in età neroniano-flavia. Il recupero, purtroppo fuori contesto, potrebbe essere riferibile sia ad un deposito di merci, da ricondurre al complesso sistema portuale, insistente nell'area, sia a una delle lussuose ville, che sorgevano nel territorio circostante (Donat 2008: 162-165; 2015: 42). Come già evidenziato per Canale Anfora, dunque, sulla base anche di questi ritrovamenti, sembrerebbe che determinate aree

1 3

Figura 2. Le forme e i tipi rappresentati. Terra sigillata sudgallica La Graufesenque Ceramista AMANDVS?

Datazione 115-130 d.C.

Quantità 1

CENSOR/CENSORINVS

80-100 d.C.

2

BIRAGILLVS

L. CINGIVS SENOVIR L. COSIVS M. CRESTIO FRONTNVS

GERMANVS III

GERMANVS IV? MASCLVS

MEΘILLVS

MERCATOR MOMO/IVSTVS

90-120 d.C.

115-130 d.C. 100-130 d.C. 80-110 d.C. 80-100 d.C. 75-110 d.C. 90-120 d.C. 50-70 d.C. 60-80 d.C.

90-120 d.C. 70-100 d.C.

3 1 4 7 1 2 1 1 1

14 1

Somma Terra sigillata sudgallica Banassac Ceramista

 

Datazione

Quantità

Somma

 

2

GERMANVS V LENTINVS

130-150 d.C. 120-150 d.C.

39

1 1

Figura 3. Terre sigillate sudgalliche: vasellame ornato a matrice ascrivibile ad un ceramista sulla base della decorazione.

Aquileia, Tergeste e nell'entroterra delle due città, va verosimilmente messo in relazione con il fatto che in tutta l'area continuavano ad essere in uso i vasi norditalici, meno belli, ma anche meno costosi. A ciò si aggiunge che cominciavano contemporaneamente

Gabucci pubblica il frammento con una doppia foto di fronte e di retro, identificandolo con due pezzi diversi: Gabucci 2017: 9.2.3., Fig. 71: 1. Il tipo è già noto ad Aquileia in più esemplari nell'area di Canale Anfora a dimostrazione del fatto che non si tratta di una presenza casuale. Donat 2017: 106. 10  Per le datazioni della terra sigillata sudgallica si fa riferimento a: Mees 1995. 9 

194

Terre sigillate galliche. La collezione “storica” del Museo Archeologico Nazionale di Aquileia

Figura 4. Terre sigillate sudgalliche: le produzioni.

portuali potessero aver avuto la funzione di “terminal” per le merci in arrivo dalla Gallia.

2015: 45; Gabucci 2017: 9.2.3). Non era stato, invece, in precedenza individuato il marchio, Rosette I, utilizzato a La Graufesenque tra il 100 ed il 130 d.C. (Tavola 1: 4) e già noto ad Aquileia negli scavi condotti a Canale Anfora da Luisa Bertacchi (Donat 2015: 45, Fig. 8). La coppa Drag. 30 è testimoniata da un unico frammento, realizzato probabilmente nell'officina di Germanus III (75-110 d.C.; Maselli Scotti 1981: 246, Tav. II: 5; Donat 2015: 42).

Ritornando ora alla collezione del Museo di Aquileia lo studio dei punzoni ha finora permesso di attribuire il 78% del materiale a un centro di fabbricazione (Figura 4). Nettamente prevalente è il vasellame di La Graufesenque (93 fr.; 62 %), mentre quello di Banassac raggiunge il 15 % (23 fr.); le rimanenti coppe Drag. 37 (33 frammenti) sono state genericamente assegnate al vasellame sudgallico (22 %).

Il repertorio figurativo offerto da questo complesso di materiale è molto ricco. Sono stati infatti classificati finora, secondo Oswald (Oswald 1936-1937) e/o Hofmann (Hofmann 1988), più di novanta punzoni. Sulle Drag. 37 prevale, in accordo con quello che è il periodo di picco delle importazioni, lo schema decorativo a metope (65 coppe; Tavola 1: 3, 6-7), seguito da quello a stile libero (25 coppe; Tavola 1: 5); in quest'ultimo si riconoscono fregi con figure umane e divine, animali in corsa tra gli alberi e la caccia flabellata, già testimoniata nell’area delle foci del Timavo e a Concordia (Donat 2008: 165, nota 553, tav. VI, 96, tav. XVI, 15). Meno frequenti sono gli schemi a festoni semicircolari, lisci o con foglie, che occupano l'intera superficie decorata o solamente una fascia12 (9 coppe) e quelli a girali (6 coppe; Tavola 1: 4). Nelle coppe con schema a metope ricorrono con una certa frequenza le scene erotiche (Tavola 1: 3), presenti sia sui prodotti di La Graufesenque che di Banassac (Maselli Scotti 1981: Fig. 8) (7 coppe; punzoni: Osw. XC, A, C, F, G, O, W), Artemide con cerbiatto (4 coppe; punzoni: Osw. 104; Tavola 1: 6) e i satiri (8 coppe; punzoni: Osw. 596, Osw. 722 ; Tavola 1: 7). Mentre le scene erotiche compaiono solo nella collezione del

Al momento, sempre attraverso la classificazione della decorazione, sono stati individuati 39 recipienti inquadrabili nel repertorio usato da precisi ceramisti di La Graufesenque e due in quello usato da altrettanti ceramisti di Banassac (Figura 3). Per quanto riguarda La Graufesenque, in accordo con quanto già discusso a proposito della datazione, solo sei ceramisti si possono datare tra il 50 e il 100 d.C., mentre i rimanenti 34 vanno collocati tra la media età flavia e l'età adrianea; tuttavia i vasai, i cui prodotti sono numericamente meglio rappresentati (Biragillus, M. Crestio, Germanus III11, Germanus IV, Mercator), si datano in età traianea. La forte presenza dei manufatti di Mercator nell'emporio aquileiese è già stata messa in luce in precedenza (Donat 2015: 43). L'officina di L. Cosius (Tavola 1: 2), che ha operato in età traiano-adrianea (100-130 d.C.), è stata riconosciuta grazie alla presenza di cinque frammenti, attribuiti al ceramista, quattro sulla base della decorazione e uno grazie a un bollo frammentario intradecorativo (Maselli Scotti 1981: 246: Tav. I: 4; Donat 11 

Lo schema decorativo è già testimoniato nell’area delle foci del Timavo (Donat 2008: Tav. XVI, 10, 12, Tav. XVII, 22). 12 

Già noto in area altoadriatica: Donat 2008: 165.

195

Patrizia Donat

Tavola 1 .Terre sigillate sudgalliche: vasellame ornato a matrice (foto P. Donat su autorizzazione del Polo Museale del Friuli Venezia Giulia).

196

Terre sigillate galliche. La collezione “storica” del Museo Archeologico Nazionale di Aquileia

Tavola 2. 1, 3 Terra sigillata da Banassac: vasellame ornato a matrice. 2, 4 Terra sigillata centro gallica: vasellame ornato a matrice (foto P. Donat su autorizzazione del Polo Museale del Friuli Venezia Giulia).

197

Patrizia Donat museo, i satiri13 e Artemide14 con cerbiatto sono spesso rappresentati nel panorama aquileiese e regionale. Tra le coppe provenienti dal centro di Banassac (Tavola 2: 1, 3) ricorre anche lo schema di busti alternati a foglie o elementi decorativi vegetali (Maselli Scotti 1981: Tav. III: 1; Gabucci 2017: Fig. 71: 4), riconosciuto anche in altri contesti della città altoadriatica.15

190 d.C.).20 Tardo è anche il marchio ALBVCIANI (155180 d.C.), noto in precedenza, come la maggior parte dei marchi aquileiesi (Maselli Scotti 1981: 243-244, Tav. I: 2); il riesame dell'intera collezione ha permesso, però, di individuare su un fondo di coppa un ulteriore bollo di Albucianus (Tavola 3: 3), identico al primo, che però a differenza di quello già conosciuto è stato stampigliato sopra la solcatura concentrica del fondo. Benché non si sia ancora completato il riconoscimento di ogni singolo punzone, il periodo di punta delle importazioni dalla Gallia centrale sembrerebbe attestarsi in età antonina, momento nel quale, oltre alle ceramiche già menzionate si possono datare 18 coppe Drag. 37.

La terra sigillata liscia, infine, è rappresentata, oltre che dalla coppa a listello Ritterling 12, già nota ad Aquileia, anche da una coppa Drag. 33, bollata OF.L.COS.VIRIL (75-110 d.C.; Maselli Scotti 1981: 243-244, Tav. I: 1; Donat 2015: 45; Gabucci 2017: 9.2.3.). La terra sigillata centrogallica e gallica orientale

La percentuale di presenza del vasellame centrogallico tra il materiale dalla vecchia collezione del museo risulta essere decisamente più alta rispetto a quanto genericamente documentato ad Aquileia e nel territorio (Donat 2017: 45); va però segnalato che negli scavi svolti con metodo scientifico nell'area di Canale Anfora essa si attesta su quantitativi abbastanza simili a quelli riscontrati per la collezione museale (Donat 2017: 101, 104, 106).

Benché le importazioni dalla Gallia centrale siano meno numerose rispetto a quelle dall'area meridionale (15 %), non possono essere considerate sporadiche né limitate ad un periodo cronologico determinato (Figura 1).16 I contatti con la Gallia centrale incominciano, infatti, nel corso della prima metà del II sec. d.C., e, per quanto non siano molto frequenti, sono contemporanei a quelli dalla Gallia meridionale, anche se questi ultimi mantengono la supremazia nelle importazioni.17 A questo periodo si possono datare, oltre alla già edita coppa Drag. 37 (Tavola 2: 2), con marchio intradecorativo Buttrio (120145 d.C.),18 altre tre coppe Drag. 37, ascritte, sulla base della decorazione, ai ceramisti di età traiano-adrianea, tra i quali è stato riconosciuto il Potier à la Rosette. Alla prima età antonina si fa risalire il marchio retrogrado, subdecorativo, in cartiglio rettangolare di Mapillo (Tavola 2: 2; 140-165 d.C.).19 Sono comunque ben presenti anche i ceramisti più tardi, come Cinnamus (Tavola 3: 1; 135-180 d.C.) e soprattutto Paternus (Tavola 3: 2; 160-160-

Oltre alle coppe decorate a matrice si ricordano cinque vasi in terra sigillata liscia, per i quali non è possibile proporre con certezza un'attribuzione alla sigillata centrogallica o di Rheinzabern (Tavola 3: 5). Sono testimoniati: tre bicchieri/olle Drag. 54/Lud. Vd (150-250 d.C.) decorati alla barbotina, ad intaglio e a sfaccettature, due dei quali già noti (Maselli Scotti 1981: Tav. IV, 2, 6); una coppa a listello Drag. 38 (100-200 d.C.); un piatto da portata Drag. 39 (120-200/250 d.C.), decorato alla barbotina e prodotto prevalentemente nelle officine di Rheinzabern (Maselli Scotti 1981: Tav. IV: 3).21 Mentre il piatto da portata Drag. 39 (Tavola 3: 6) è al momento un unicum nel panorama regionale, gli altri tipi sono invece già documentati nelle due indagini di scavo svolte a Canale Anfora (Donat 2015: 46, Fig. 14: 7, 9; 2017: 106-108: Fig. 2: 5-6).

Aquileia, Canale Anfora – Scavi Bertacchi (autopsia personale); Castions di Strada, Cividini 2002: 176-177, T.s. 38; è possibile però che quest'ultima coppa sia di provenienza aquileiese (comunicazione di Tiziana Cividini); area delle foci del Timavo: Donat 2008: Tav. XVII: 21. 14  Per l'ampia diffusione su territorio regionale del punzone: Donat 2017: 103, nota 31. Oltre che nel punzone Osw. 104, la divinità compare anche nel punzone Osw. 118 (inedito, collezione del museo e Aquileia - Piazza Capitolo: Donat 2015: Fig. 6 e Osw. 110 (Maselli Scotti 1981: Fig. II: 3). 15  Donat 2015: 43: 9. La revisione del materiale ha permesso di individuare altri due frammenti della coppa già pubblicata da Maselli Scotti (Maselli Scotti 1981: T. II: 1), Tavola 2: 3. 16  Le datazioni per la terra sigillata gallica sono state determinate secondo Stanfield e Simpson 1990. 17  Sulla precocità delle importazioni dalla Gallia centrale ad Aquileia e su un suo possibile arrivo nella città altoadriatica assieme ai prodotti di Banassac: Gabucci 2017: 9.2.3. 18  Knorr 1912: Taf. XXX: 4; Maselli Scotti 1981: Tav. III: 2; Gabucci 2017: 9.2.1, 9.2.3., Fig. 71: 5 (con ulteriore bibliografia e con la pubblicazione dello stesso marchio da Altino). La revisione delle terre sigillate ha permesso di individuare un altro frammento appartenente alla stessa coppa. Per altre presenze dello stesso ceramista ad Aquileia: Donat 2015: 46. 19  Come la maggior parte dei marchi anche questo è ampiamente edito, anche se non in tutte le pubblicazioni l'interpretazione è corretta: Knorr 1912: Taf. XXX: 5; Maselli Scotti 1981: Tav. IV: 5; Stanfield e Simpson 1990: 321-322, pl. 166: 8; Donat 2015: 46; Gabucci 2017: 9.2.3., Fig. 71: 6. 13 

Le officine della Gallia dell'est sono scarsamente rappresentate. La produzione di Rheinzabern, oltre alle due coppe Drag. 37 con marchi di Belsus I (180250 d.C.) e di Secundinus Avitus (160-200 d.C.) e al bollo di età antonina su ceramica liscia, [A]PRO FECIT,22 già noti dall'inizio del Novecento, non è altrimenti testimoniata (Knorr 19012: Taf. XXX: 6-7; Maselli Scotti 2012: 251-252: Tav. IV: 1, 4). Una novità è stata invece Per il momento non è stata operata la distinzione proposta da Rogers nell'ambito delle sintassi decorative ascritte a questi ceramisti: Rogers 1999. Le produzioni di Cinnamus sono già state riconosciute in: Maselli Scotti 1981: 250-251, Tav. III: 1. Sulla diffusione dei prodotti di questo ceramista ad Aquileia: Donat 2015: 46, Fig. 12; Donat 2017: 106, nota 34. 21  Le datazioni seguono quanto proposto da Dénes Gabler: Groh e Gabler 2017: 136-151. 22  Non è stato reperito tra il materiale esaminato (Maselli Scotti 1981: 251-252). 20 

198

Terre sigillate galliche. La collezione “storica” del Museo Archeologico Nazionale di Aquileia

Tavola 3. 1-2 Terre sigillate centro galliche: vasellame ornato a matrice. 4 Terra sigillata dalla Gallia orientale: vasellame ornato a matrice. 2, 5-6 terra sigillata liscia dalla Gallia centrale e orientale (foto P. Donat su autorizzazione del Polo Museale del Friuli Venezia Giulia).

199

Patrizia Donat l'individuazione di due coppe Drag. 37 (Tavola 3: 4), ascritte alla produzione dell'atelier di Saturninus e Satto (90-150 d.C.).

fatti nell’areale di Canale Anfora e in quello delle foci del Timavo rafforzano tale ipotesi, essendo entrambi i siti ricollegabili al sistema portuale insistente sulle direttrici endolagunari.

Le coppe decorate a matrice Drag. 37 continuano a dominare il mercato, mentre la ceramica liscia rimane documentata in quantitativi poco significativi; tuttavia i confronti noti dagli scavi fanno capire, che anche questa aveva un certo smercio, soprattutto se decorata (Donat 2015: 46; 2017: 106-108, Fig. 2).

In età traiano – adrianea ad Aquileia giunge, anche se in quantitativi più ridotti, assieme al vasellame di La Graufesensenque, anche quello di Banassac e dei primi ceramisti del centro e dell’est della Gallia. I recipienti dalla collezione, così come quelli dagli scavi, dimostrano che le importazioni dalla Gallia centrale continuano, mai più rarefatte, anche nel corso della seconda metà del II sec. d.C.; ad esse si aggiungono, in forma, però, sporadica quelle da Rheinzabern. Che la continuità dei contatti con le province galliche sia rimasta attiva fino almeno alla fine del II sec. d.C. sembrerebbe essere dimostrato anche dagli arrivi, al momento poco significativi, di ceramica metallescente di Treviri (Dobreva e Mantovani 2017: 239-242). Nonostante la sporadicità di questi rinvenimenti, la presenza dei prodotti della Gallia orientale ad Aquileia costituisce un caso quasi unico in tutta la Cisalpina (Gabucci 2017: 2).

Alcune considerazioni conclusive Quest’analisi delle collezioni “storiche” di terre sigillate galliche, conservate nel Museo Archeologico Nazionale di Aquileia, mettendo in evidenza l'ingente numero di reperti ivi custoditi, permette di archiviare definitivamente l'ipotesi, che le importazioni dalla Gallia siano evidenze sporadiche da collegare al particolare gusto di singoli personaggi particolarmente abbienti (Maselli Scotti 1981). Il raffronto con le evidenze dagli scavi archeologici, solo alcuni dei quali sono stati svolti, però, con metodi scientifici, dà il modo di calibrare meglio questa considerazione, portando a ritenere che l’afflusso di merci dalle province galliche sia stato contenuto (3 - 5 % di tutta la ceramica fine), ma non sporadico.

La distribuzione dei recipienti gallici nel territorio di Aquileia evidenzia contenute presenze da localizzarsi lungo la via endolagunare e nelle ville più lussuose (Donat 2015; 2017: 102); qui come nell’emporio altoadriatico la richiesta si è concentrata sulle coppe decorate a matrice, non presenti nel repertorio della ceramica fine norditalica, dopo l’esaurirsi delle forme caratterizzanti il I sec. d.C. fino all’età flavia. Rispetto a queste ultime, inoltre, la qualità del vasellame transalpino è nettamente superiore; si tratta, infatti, di merce raffinata e costosa. Il costo elevato, unito al fatto che le coppe Drag. 29 e soprattutto le coppe Drag. 37 più capienti potessero costituire un pezzo unico nei servizi da mensa, ha sicuramente contribuito a favorire un’importazione moderata delle produzioni galliche, che tuttavia dovevano essere molto apprezzate, visto che i fabbricanti norditalici di media età imperiale ne hanno imitato le forme e le decorazioni (Donat e Maggi 2017).23

Un primo timido inizio delle importazioni ad Aquileia è stato fissato, in età tiberiana, con una modesta intensificazione in età neroniano-flavia, momento al quale si data anche il vasellame più antico appartenente alla collezione. Il picco delle presenze si raggiunge però tra l’80 d.C. e la metà del II sec. d.C. Ciò è testimoniato sia dai materiali della collezione del Museo che da quelli da scavi più recenti; tale evidenza trova riscontro in tutta l’area padana, mentre nel resto del Mediterraneo si nota una contrazione nelle esportazioni (Donat 2015: 43; Gabucci 2017: 2.3.1, Tav. 1). Le ceramiche galliche, in particolare quelle sudgalliche di La Graufesenque, sono infatti documentate in diversi contesti dell’emporio altoadriatico (Donat 2015: Fig. 3); particolarmente interessante è la concentrazione di ritrovamenti nell’area portuale di Canale Anfora, che porterebbe a ipotizzare per la zona una destinazione privilegiata come “terminal” degli arrivi dalle province galliche. Un altro “terminal” va forse localizzato nell’area delle foci del Timavo, dove è stato recuperato un ingente quantitativo di terre sigillate sudgalliche (Donat 2008). In considerazione della costante distribuzione delle ceramiche galliche lungo la direttrice fluviale del Po e le vie endolagunari altoadriatiche ad essa collegate e, al contrario, del suo rarefarsi man mano che si procede verso l’Adriatico meridionale (De Mitri 2017), si ritiene che questi prodotti siano stati trasportati, dopo aver valicato le Alpi, attraverso tali arterie (Donat 2015: 42; Gabucci 2017). Gli abbondanti ritrovamenti

L’analisi della collezione “storica” del Museo di Aquileia ha confermato nel complesso quanto già emerso dallo studio dei materiali da scavo; la possibilità di analizzare un quantitativo di materiale così cospicuo ha però dato l’opportunità di notare nell’ambito del vasellame sudgallico, pur nella varietà dei motivi decorativi presenti, il ripetersi di determinati schemi, punzoni e marchi, riferibili a determinati ceramisti. Altrettanto interessante appare il ricorrere degli stessi elementi tra il materiale proveniente dalle indagini di scavo. Ci si ripromette, in futuro, di approfondire questa linea In un recente scambio di opinioni Allard Mees mi ha suggerito anche di prendere in considerazione la possibilità di provare a mettere in relazione gli arrivi di terra sigillata gallica nell’area aquileiese con la presenza coeva di truppe sulle coste adriatiche orientali. 23 

200

Terre sigillate galliche. La collezione “storica” del Museo Archeologico Nazionale di Aquileia

di ricerca, per comprendere se determinati ceramisti, a differenza di altri, possano aver scelto di ampliare la loro rete di traffici anche al mercato norditalico e in particolare aquileiese.

recenti, in S. Magnani e P. Ventura (eds) Scavare nei musei. Elementi di novità e questioni di metodo. Atti del Convegno, Aquileia 7 giugno 2013 (Quaderni Friulani di Archeologia 25): 31‒43. Donat P. 2017, Terra sigillata gallica, in P. Maggi, F. Maselli Scotti, S. Pesavento Mattioi e E. Zulini (eds) 2017: 101-108. Donat, P. e P. Maggi 2017. Terre sigillate italiche, in P. Maggi, F. Maselli Scotti, S. Pesavento Mattioli e E. Zulini (eds) 2017: 71‒98. Donat, P., P. Maggi, P. Ventura and E. Zulini c.s., Aquileia, Canale Anfora, recupero 1988: materiali ceramici come indicatori commerciali, in I. Borzić, E. Cirelli, E., K. Jelinčić Vučković, A. Konestra e I. Ožanić Rogulić (eds) Trade. Transformations of Adriatic Europe (2nd - 9th c.). Proceedings of the conference, Zadar, 11-13 Febbraio 2016, c.s. Donat, P., P. Maggi e E. Zulini 2017. Ceramica fine: quadro introduttivo, in P. Maggi, F. Maselli Scotti, S. Pesavento Mattioli e E. Zulini (eds) 2017: 65‒68. Fontana, F. (ed.) 2017. Aquileia, l’insula tra foro e porto fluviale. Lo scavo dell’Università degli Studi di Trieste. 1. La strada. (Scavi ad Aquileia III). Trieste: Edizioni Università di Trieste. Gabucci, A. 2017. Attraverso le Alpi e lungo il Po: importazione e distribuzione di sigillate galliche in Cisalpina (Collection de l’École française de Rome 390). Roma: École française de Rome (edizione on line). Giovannini, A. 2004. Le istituzioni museali pubbliche di Aquileia: spunto per uno studio delle fasi storiche. I. Dal Museo eugeniano all’I.R. Museo e agli allestimenti di Enrico Maionica. Aquileia Nostra 75: 457‒518. Giovannini, A. 2007. Il patrimonio archeologico di Aquileia prima del 1882. Le raccolte private e il “Museo patrio della città”, in G. Cuscito e C. Zaccaria (eds) La ricerca antiquaria nell'Italia nordorientale dalla Repubblica Veneta all'Unità (Antichità altoadriatiche 64): 233‒316. Trieste: Editreg srl. Groh, S. e D. Gabler 2017. Terra sigillata aus den Zivilstäden von Carnuntum und Aquincum. Eine Analyse des Sigillata-Importes der Provinz-Hauptstädte von Pannonia Superior et Inferior (Archäologische Forschungen in Niederösterreich 4). Krems: Landessammlungen Niederösterreich und der Donau-Universität. Homann, B. 1988, L'atelier de Banassac, (Revue archéologique SITES Hors-Séerie 33) 1988. Knorr, R. 1912. Südgallische Terra Sigillata-Gefäße von Rottweil. Stuttgart: Kohlhammer. Lipovac Vrkljan, G., B. Šiljeg, I. Ožanić Roguljić e A. Konestra (eds) 2017. Roman Ceramic and Glass Manufactures. Production and Trade in the Adriatic region. Proceedings of the 3rd international archaeological colloquy, Crikvenica 23.-24 october 2014. Crikvenica/ Zagreb: Muzej Grada Crikvenice, Institut za arheologiju. Maggi, P., F. Maselli Scotti, S. Pesavento Mattioli e E. Zulini (eds) 2017. Scavi di Aquileia IV. Materiali per

Bibliografia Bonetto, J., D. Bragagnolo, V. Centola, D. Dobreva, G. Furlan, E. Madrigali, A. Menin, A. e C. Previato 2013. Aquileia (UD). Fondi ex Cossar. Relazione delle ricerche 2009. Notiziario della Soprintendenza per i Beni Archeologici del Friuli Venezia Giulia 4: 134‒168. Brusin, G. 1929. Aquileia. Guida storica e artistica con prefazione di Roberto Paribeni. Udine: La Panarie. Buora, M., G. Cassani, M. Fasano e A.R. Termini 1995. Saggi di scavo ad Aquileia (1989-1990). Quaderni Friulani di Archeologia 5: 91‒162. Callegher, B. (ed.) 2015. Studia archaeologica Monika Verzár Bass dicata (West & East Monografie 1). Trieste: Edizioni Università di Trieste. Ceazzi, A. e A. Del Brusco 2007. Lo scavo della Scuola di Specializzazione in Archeologia dell’Università di Trieste. Rapporto preliminare sulla ceramica (Campagne 2002-2004). Archeografo Triestino 67: 119‒148. Cividini, T. 2002. Presenze romane nel territorio del Medio Friuli. 9. Castions di Strada. Tavagnacco: Provincia di Udine. De Mitri, C. 2017. Ceramiche fini da mensa di produzione occidentale nell’area ionico-adriatica (II sec. a.C. II sec. d.C.): presenza e circolazione nella penisola salentina, in G. Lipovac Vrkljan, B. Šiljeg, I. Ožanić Roguljić e A. Konestra (eds) Roman Pottery and Glass manufactures. Production and Trade in the Adriatic region. Proceedings of the 3th international archaeologica colloqui, Crikvenica, 4th - 5th november 2014: 351 - 362. Crikvenica: Muzej Grada Crikvenice, Institut za arheologiju. Dobreva, D. e D. Mantovani 2017. Ceramiche fini di produzione dalmata, pannonica, retica, e di Treviri ad Aquileia, in G. Lipovac Vrkljan, B. Šiljeg, I. Ožanić Roguljić e A. Konestra (eds) Roman Pottery and Glass manufactures. Production and Trade in the Adriatic region. Proceedings of the 3th international archaeologica colloqui, Crikvenica, 4th - 5th november 2014: 225 - 250. Crikvenica: Muzej Grada Crikvenice, Institut za arheologiju. Donat, P. 2008. Le produzioni sudgalliche, in R. Auriemma, V. Degrassi, P. Donat, D. Gaddi, S. Mauro, F. Oriolo e D. Riccobono, Terre di mare: paesaggi costieri dal Timavo alla penisola muggesana, in R. Auriemma e S. Karinja (eds) Terre di mare. L'archeologia dei paesaggi costieri e le variazioni climatiche. Atti del Convegno Internazionale di Studi, Trieste, 8-10 novembre 2007. Trieste: Università degli Studi di Trieste Museo del mare “Sergej Mašera” Piran: 162‒167. Donat, P. 2015. Terra sigillata gallica in Italia nordorientale. Dalle collezioni museali alle scoperte 201

Patrizia Donat Aquileia. Lo scavo di Canale Anfora (2004-2005). Trieste: Editreg. Maggi, P. e R. Merlatti 2015. Ceramiche fini nell’alto Adriatico. Produzione e flussi commerciali tra II sec. a. C. e II sec. d. C., in Y. Marion e F. Tassaux (eds) AdriAtlas et l’histoire de l’espace adriatique du VIe s. a.C. au VIIIe s. p.C. Actes du colloque international de Rome, 4-6 novembre 2013(Scripta Antiqua 79): 165‒182. Bordeaux: Ausonius Éditions. Maionica, E. 1911. Guida dell’I.R. Museo dello Stato di Aquileia. Wien: Österreichisches Archäologisches Institut. Maselli Scotti, F. 1981. Terra sigillata della Gallia ad Aquileia, in M. Mirabella Roberti (ed.) Aquileia e l'Occidente (Antichità Altoadriatiche 19): 239‒257. Udine: Arti Grafiche Friulane. Maselli Scotti, F.1984. La ceramica di Aquileia. Il vasellame da mensa, in M. Mirabella Roberti (ed.) I musei di Aquileia II Arti applicate - ceramica - epigrafia – numismatica (Antichità altoadriatiche 24): 39‒69. Udine: Arti Grafiche Friulane. Maselli Scotti, F. 1987. Terre sigillate di Aquileia e Tergeste. Produzioni italiche ed importazioni galliche ed orientali. Rei Cretariae Romanae Fautorum Acta 25-26: 207‒224. Maselli Scotti, F. 1988. La ceramica nelle fortificazioni di età romana in Friuli, in T. Miotti (ed.) Castelli del Friuli 7: 261‒294. Udine: Del Bianco.

Mees, A. 1995. Modelsignierte Dekorationen auf südgallischer Terra Sigillata (Fundeberichte aus Baden Württemberg 54). Stuttgart: Theiss. Oswald, F. 1936-1937.  Index of Figure-Types on Terra Sigillata «Samian Ware». Liverpool: University Press. Rogers, G. B. 1999.  Poteries sigillées de la Gaule centrale.  II.  Les potiers (Revue Archéologique. Sites. Hors série 40). Lezoux: Lezoux Centre archéologique Duchasseint. Stanfield, J. A. e G. Simpson 1990. Les Potiers de la Gaule Centrale (Revue Archéologique. Sites. Hors série 37). Gonfaron: Association Française d'Archéologie Métropolitaine. Ventura, P. e E. Zulini 2012-2013. Attestazioni di terra sigillata africana ad Aquileia. Conoscenze pregresse e materiali inediti dai magazzini del Museo Archeologico Nazionale. Aquileia Nostra 8384: 371‒399. Verzár Bass, M. (ed.)1991. Scavi ad Aquileia. I. L’area ad est del Foro, Rapporto degli scavi 1988. Roma: Quasar. Verzár Bass M., (ed.)1994. Scavi ad Aquileia. I. L'area ad est del Foro. Rapporto degli scavi 1989-1991. Roma: Quasar. Zulini, E. 2015. Nuove presenze di terra sigillata africana ad Aquileia, in Callegher (ed.) 2015: 183‒193.

202

La ceramica metallescente di Treviri: nuovi dati da Verona (Italia) Valentina Mantovani Abstract Due nuove attestazioni di ceramica metallescente di Treviri rinvenuta a Verona, in via Cantore 2a e in Vicolo Calcirelli stimolano interessanti riflessioni in merito alla presenza di questa classe di materiali a sud delle Alpi. L’esemplare integro di Vicolo Calcirelli è stato rinvenuto in una sepoltura ad incinerazione che, sebbene disturbata e con un corredo parziale, reca al suo interno oggetti di rilevante interesse, come frammenti vitrei di decorazione a gabbia riferibili ad un vaso pseudo-diatretum, un candeliere in bronzo e uno specchio con presa posteriore in argento. Questi oggetti, insieme alla presenza del bicchiere di Treviri, danno una connotazione particolare alla sepoltura, la quale può essere forse riferita ad un individuo che può aver avuto contatti diretti con i territori gallici e germanici, o che trova, proprio in quelle province, i natali.

Key words Verona, Vicolo Calcirelli, Cantore street, Trier black-slipped ware, pseudo-diatretum, candlestick, mirror.

La ceramica metallescente di Treviri La ceramica metallescente di Treviri, denominata anche “Trier black-slipped ware” o “Moselkeramik”, è tra le produzioni della Gallia Belgica ad aver goduto di una maggiore distribuzione nei territori dell’impero romano: oltre ad una cospicua presenza nella valle della Mosella, l’esportazione, la diffusione e il rinvenimento di questi manufatti si concentra negli insediamenti civili adiacenti ai principali siti dell’asse reno-danubiano dove erano di stanza le legioni e la sua diffusione avvenne con un commercio organizzato, almeno in direzione della Pannonia, attraverso Mosella, Reno e Danubio. Più recenti e ancora estremamente circoscritte sono, invece, le sue attestazioni a sud delle Alpi, testimoniate fino ad oggi dai reperti di Camporosso in Valcanale, antica Statio Bilachiniensis (Harsanyi 2013a: fig. 6-7; 2013b: fig. 3), di Aquileia (Mantovani 2016) e Milano (Airoldi 2011: 110-111, scheda 26).1 Gli ateliers, localizzati nell’area sud-ovest dell’antica Augusta Treverorvm, furono attivi a partire almeno dalla metà del II sec. d.C., ma materiali di qualità elevata sono stati prodotti ed esportati a partire dalla metà del Sono note in ambito padano anche ceramici che parrebbero ispirarsi alle produzioni della Gallia centrale e nord-orientale. Ad Aqui Terme è attestata un’olla in ceramica comune con labili tracce di ingobbio grigiastro la cui forma riprende i tipici bicchieri di Treviri, con una decorazione a losanghe excise a ricreare dei grappoli d’uva stilizzati (Robino 2008: 44-45). Un secondo tentativo di imitazione è attestato da un bicchiere del Capitolium di Brescia decorato con un’iscrizione graffita post cottura le cui lettere capitali rimaste [---]BIBE[---] imitano le frasi conviviali di incitamento al bere dipinte a barbottina bianca sulla ceramica metallescente di Treviri (Portulano 2002: 405, tav. III, 6) Per un più ampio inquadramento sulla questione si rimanda da ultimo a Mantovani in press. 1 

III sec. d. C. fino al terzo quarto del III d.C., momento identificato come periodo di apogeo della produzione. Si assiste successivamente ad un decadimento estetico dei manufatti, oltre che ad una netta contrazione delle esportazioni, da mettere in relazione con le devastazioni germaniche che negli anni 276 – 277 d.C. interessarono la Gallia (Symonds 1992; Künzl 1997; Vilvorder 2010; Harsanyi 2013a; 2013b). Tratti salienti di questa classe di vasellame potorio sono l’impasto rosso, ben cotto, che presenta frattura a sandwich di colore rosso-grigio. Il rivestimento, il cui colore varia dal bruno-grigio al nero, è grasso, brillante e metallescente. Il repertorio morfologico è costituito fondamentalmente da tazze e bicchieri con forma ovoide e pancia più o meno espansa modellata con depressioni ovoidali o circolari. La decorazione è costituita da rotellature e sovradipinture ottenute con la barbottina bianca che definiscono motivi pseudo geometrici, floreali o iscrizioni in latino che invitano a godere del banchetto e della vita (“Riempimi”, “Beviamo”, “Mescimi”, “Ai miei amici”). Gli esemplari di Verona Importanti per la ricostruzione delle dinamiche commerciali, economiche e culturali che hanno caratterizzato la diffusione di questa particolare classe di manufatti sono, pertanto, i due esemplari documentati a Verona.2 Ringrazio la Dott.ssa Brunella Bruno, funzionario archeologo della Soprintendenza archeologia, Belle Arti e Paesaggio di VR, RO e VI, per avermi concesso lo studio e la pubblicazione del presente gruppo di materiali, nonché Margherita Bolla, Elisabetta, Roffia, Marianne Stern, Constanze Höpken, Helga Sedlymeyer, Eleni Schindler 2 

roman pottery and glass manufactures (Archaeopress 2022): 203–211

Valentina Mantovani

Figura 1. Verona, ubicazione dei rinvenimenti di ceramica metallescente di Treviri (autore: V. Mantovani).

strato più basso, adagiato sugli embrici (US 182) e composto dalla terra di rogo nella quale erano adagiati e immersi reperti integri e frammentati. Tale livello si presentava superiormente sigillato da un potente strato di maceriume visibile almeno fino al bordo superiore del lato lungo conservato (US 177). Da questa unità stratigrafica sono stati raccolti pochi materiali, tra i quali spiccano frammenti di ceramica invetriata rinascimentale – che ci aiutano a definire il momento dell’intervento di scasso – e almeno due elementi pertinenti, con ogni probabilità al corredo, e che in seguito analizzeremo. La loro presenza nello stato soprastante a quanto rimane intatto della sepoltura può indicare che almeno in parte l’attività di disturbo abbia intaccato in maniera meno pesante la parte inferiore della sepoltura, lasciando nella porzione inferiore del riempimento US 177 traccia del corredo.

Vicolo Calcirelli Il primo di essi fa parte del corredo di una sepoltura a cremazione deposta all’interno di cassa con alzato in muratura di laterizi e fondo in embrici rinvenuta nel 2011 in Vicolo Calcirelli, all’angolo con Vicolo Giacciaia, nel corso dello scavo per la realizzazione di un parcheggio interrato.3 La tomba faceva parte della necropoli lungo la via Postumia, in una zona interna rispetto alla via principale con sepolture piuttosto distanziate tra di loro (Arzone et al. 2015: 373-374) (Figura 1). Al momento del rinvenimento l’incinerazione si presentava priva di copertura, completamente scassata su due lati (uno lungo e uno corto) e rasata fino allo Kaudelka, Roby Stuani, Barbara Lepri , Andrea Stella, Stefano Buson e Giulia Bison per i loro preziosi consigli e l’aiuto offertomi. 3 Contesto inedito. Materiali parzialmente privi di numero di Inventario Generale. Ringrazio sentitamente la Dott.ssa Paola Fresco, titolare della Società Archeologica Multiart, per avermi dato modo di consultare la documentazione del cantiere relativa alla sepoltura in oggetto.

La sepoltura non è da ritenersi, pertanto, integra, ma fortemente disturbata, dato confermato anche dallo 204

La ceramica metallescente di Treviri: nuovi dati da Verona (Italia)

2 1

5

8

4

3

7

6

9

10

12

11

12 13

14 Tavola 1. Verona, Vicolo Calcirelli, corredo della sepoltura contenente un bicchiere in ceramica metallescente di Treviri (disegni: V. Mantovani).

stato di conservazione di alcuni oggetti rinvenuti nello strato di terra di rogo, in particolare i vetri, estremamente frammentati.4

specchio in bronzo; US 182/3 frr. di ossa combuste; US 182/4 elemento in ferro non determinabile; US 182/5 stilo bronzo; US 182/9 bicchiere di Treviri; US 182/17 candeliere in bronzo; US 182/18 ansa in vetro + vetro combusto e deformato; US 182/20 frr. subcentimetrici di vetro non fuso; US 182/21 frr. subcentimetrici di ceramica comune, forma e produzione non determinabile; US 182/22 fondo di brocca + collo di balsamario + numerosi frr. sub centimetrici di vetro non fuso; US 182/23 chiocciola; US 182/24 frr. di ossa combuste; US 182/25 frr.

Quanto raccolto pertinente alla sepoltura (33 elementi) può essere cosi riassunto: US 177/ 1 orlo di balsamario + numerosi frr. sub centimetrici di vetro non fuso; US 177/4 presa in argento; US 182/2 4 

205

Valentina Mantovani Tra gli elementi integri o più completi spicca il bicchiere in ceramica (US 182/9), con corpo globulare leggermente schiacciato e collo tronco-conico ascrivibile al tipo Symonds 1-gruppo 39/Niederbieber 33/Trier 1, forma prodotta a Treviri e databile attorno alla metà del III sec. d.C. (Tavola 1: 1). L’esemplare presenta tre fasce orizzontali e parallele di rotellatura disposte tra la spalla e la parte inferiore della pancia. Del tutto assente è la tipica decorazione sovradipinta con barbottina bianca.

apposte le decorazioni. Ipotizzando possa trattarsi di un calice o di un bicchiere, il frammento A (diametro cm 12) potrebbe essere considerato il “bordo” superiore della gabbia, mentre il secondo frammento (B) potrebbe essere pertinente alla porzione di parete verso il fondo (diametro massimo cm 8). Tali misure potrebbero essere compatibili con il cantaro di Colonia, ove il diametro massimo della gabbia misura cm 11. Tuttavia, ruotando l’orientamento dei due pezzi veronesi di 180° (frammento A sul fondo della gabbia, frammento B vicino al collo) le misure potrebbero essere a loro volta compatibili con una bottiglia, avvicinandosi, quindi, in tal caso al confronto padovano.

Arricchivano il corredo della sepoltura alcuni oggetti di estremo interesse. Il primo di essi è un vaso in vetro decolorato, del quale si conservano solo i filamenti lavorati a giorno con andamento a zig-zag che dovevano formarne la gabbia decorativa a “rete di serpenti” di un contenitore classificabile nell’ambito degli “pseudo-diatretum” (US 182/30). Il frammento “A” sembra finito su un lato, mentre il frammento “B” conserva ancora adesa una porzione di parete non rettilinea in vetro soffiato incolore (Tavola 1: 2).

Al momento non pare possibile trovare una soluzione in merito alla forma ed origine del vaso. Il Dish Cantharus è attributo a produzione di Colonia, così come gli esemplari provenienti dalle sepolture di Colonia Agrippina; altri esemplari con analoga lavorazione rinvenuti nei territori renani vengono più propriamente attribuiti agli ateliers attivi nella regione di Magonza nel III sec. d.C. (soprattutto per gli esemplari di brocche con ansa decorata con tale tipo di filamenti) (Foy et al. (eds) 2018: 284, 305, 334). Scarti di alcuni elementi pertinenti ad una lavorazione a giorno a rete di colore blu cobalto provengono, inoltre, da una fornace per il vetro identificata a Pègue (Drome), datata tra la fine del II e gli inizi del III sec. d.C. (Guérit and Ferber 2015: 47, fig. 5 n. 19).

Ciò che rimane del manufatto richiama, in ambito padano, la decorazione applicata su una bottiglia pseudo-diatretum rinvenuta a Padova e attribuita, sulla base delle somiglianze con gli analoghi materiali del Tesoro di Begram, ad una produzione di area alessandrina, o forse siro-palestinese, inquadrabile tra la fine del I sec. d.C. o, al più tardi, nei primi decenni del II sed. d.C. (Bonomi e Tagliaferro 2006). Spostandoci invece oltralpe, altri confronti per la decorazione a rete di serpenti presente nella sepoltura di Vicolo Calcirelli provengono dal famoso Dish Cantharus, conservato al Corning Museum, e il suo parallelo già a Schloss Goluchow, in Polonia, i quali si ritengono rinvenuti a Colonia nel 1866 e datati dal Fremersdorff dapprima attorno al 200 d.C. e successivamente alla seconda metà del III sec. d.C. (Whitehouse 1988: 253254). Attualmente, la scheda del reperto conservato al Corning Museum of Glass riporta la datazione 200-325 d.C..5 Esemplari in vetro decolorato con decorazione a gabbia sono attestati anche in alcune sepolture di Colonia (Höpken and Liesen 2013: 413, abb. 33 5/6) e più genericamente in Gallia del Nord (Foy et al. (eds) 2018: 261, 334).

Si sottolinea, invece, per quanto riguarda i pezzi del tesoro di Begram attribuiti ad una produzione di area alessandrina o siro-palestinese, che la maggior parte delle lavorazioni a “rete di serpenti” attestate nel sito, così come gli stessi supporti sulla quale è applicata, presentano una colorazione blu. Allo stato attuale dei fatti non sembra possibile definire, senza il supporto di analisi archeometriche, l’area di origine dei frammenti veronesi. Tuttavia – tenendo conto anche dell’origine degli altri materiali presenti nel corredo veronese – le attestazioni di filamenti incolori tipici di area renana così come la diffusione della maggior parte dei confronti noti sembrerebbero rimandare più ad una produzione di III sec. d.C. collocabile a nord delle Alpi, invece che ad area alessandrina o siro-palestinese, con conseguente slittamento della cronologia.

L’esiguità dei frammenti veronesi non permette l’identificazione della forma originaria sulla quale erano

Tra i numerosi piccoli frammenti vitrei è possibile riconoscere, poi, almeno due orli di brocche in vetro soffiato azzurro-verde chiaro Isings 56a (US 182/27) (Tavola 1: 3-4), forma assai frequente nei corredi veronesi lungo la Postumia e per la quale è stata proposta, alla luce del tipo di fattura (pareti estremamente sottili) e analogie dei pezzi, una produzione locale veronese (Roffia 1998: 99). Potrebbe essere pertinente ad una

di vetro subcentimetrici combusti e deformati; US 182/26 fondo balsamario/bottiglia + numerosi frr. sub centimetrici di vetro non fuso; US 182/27 2 orli di brocche Isings 56a; US 182/ 28 fr. di selce; US 182/29 terra di rogo; US 182/30 frr. di pseudotiatreta. Seguono le 13 monete edite ed inventariate: VR 88774, VR 88773, VR 88775, VR 88763, VR 88771, VR 88764, VR 88765, VR 88767, VR 88766, VR 88772, VR 88768, VR 88769, VR 88770. 5  https://www.cmog.org/artwork/disch-cantharus?image=0&search =collection%3Ae472320258ff81d9cf60b81270a27444&page=1

206

La ceramica metallescente di Treviri: nuovi dati da Verona (Italia)

di queste forme anche il fondo in vetro azzurro-verde chiaro con pareti molto sottili (US 182/26) (Tavola 1: 9)

maniglia a fascia semplice) poi attestarsi con maggiore frequenza nel corso del III sec. d.C. nei territori dell’Alto e Basso Reno, in Gran Bretagna, Francia, Austria e Jugoslavia. A partire dal IV secolo, i centri di produzione si spostano verso la parte orientale dell’Impero romano e verso le aree del futuro regno bizantino (regione del Mar Nero, Caucaso, Nord Africa e Siria) (Lerner 1996: 30). La produzione di questi particolari specchi va esaurendosi dopo il VII secolo nella parte occidentale dell’impero e il IX o X secolo, al più tardi , in quella orientale (Lerner 1996: 31).

È documentato anche un orlo di balsamario in vetro soffiato azzurro-verde De Tommaso 32 (US 177/1) (Tavola 1: 5) e un fondo particolarmente umbonato, forse pertinente al medesimo esemplare ma proveniente da US 182/22 (Tavola 1: 8). Tra gli altri elementi vitrei il cui stato di conservazione ne permette una vaga identificazione morfologica compaiono poi un secondo collo di balsamario (US 182/22) (Tavola 1: 6), il cui colore e lo spessore delle pareti non paiono riconducibili all’esemplare di Tavola 1 n. 5, ed un’ansa a gomito e sezione circolare (US 182/18) (Tavola 1: 7), il cui vetro, molto alterato e opaco, presenta una colorazione cangiante che varia dal blu al rosso cupo. L’inquadramento morfologico di questi ultimi reperti risulta alquanto difficoltoso. Per restanti frammenti vitrei, di dimensioni sub-centimetriche, solo un intervento di restauro potrebbe, forse, fornire qualche ulteriore evidenza. Tra di essi si segnalano, ad ogni modo, alcuni piccoli elementi fortemente deformati come conseguenza dell’esposizione al fuoco della pira funebre (US 182/18 e US 182/25).

In ambito italico i casi più rappresentativi di questa forma sono gli specchi di Pompei, Napoli ed Aquileia, inquadrabili nel I sec. d.C. (Lerner 1996: 32). L’esemplare aquileiese, in particolare (diametro cm 10 ca.), rinvenuto in una sepoltura ad incinerazione e decorato nella parte non riflettente con scena delle Tre Grazie (Lerner 1996: 34, nota 7; Brusin 1929: 170, fig. 116), è dotato di una maniglia a fascia semplice che reca piedini del tutto identici a quello di Vicolo Calcirelli. In merito alla decorazione superiore della maniglia del pezzo veronese, i confronti più stringenti paiono essere l’esemplare di Vienne, in Francia (diametro cm 20), e lo specchio conservato al Metropolitan Museum of Arts e attribuito al Tesoro di Antiochia (diametro cm 23,3). Il primo esemplare, datato al III sec. d.C., presenta una elegante decorazione a squame che corre sia lungo il bordo della parte posteriore dello specchio, sia sulla maniglia conformata “a dita contrapposte” della presa, nella quale compare anche un elemento tortile nel punto di sutura delle due dita, in posizione analoga a quella del nostro esemplare. Lo specchio del Metropolitan Museum of Arts presenta, invece, una fattura più semplice: solo la presa “a dita contrapposte” è decorata con una fine spina di pesce. Questo esemplare, ricongiunto in un secondo momento al gruppo del Tesoro di Antiochia, viene datato così come il resto del tesoro al VI-VII sec. d.C., ma va sottolineato che, proprio per le vicissitudini che ne hanno portato all’identificazione e all’inserimento nel gruppo di Antiochia, la Lerner ne propone una datazione al IV sec. d.C. (Lerner 1996: 32).

Di rilievo, nel corredo, è anche uno specchio modellato con una lega di bronzo specifica per gli specchi di forma circolare, con piattello a bordo diritto e fondo leggermente convesso del diametro di cm 9,4 (US 182/2) (Tavola 1: 10). Lo specchio è dotato, al centro della parte posteriore concava, di una presa in argento (US 177/4) nella cui parte sommitale è presente una decorazione con due dita contrapposte ed unite al centro mediante una serie di quattro borchiette. La parte inferiore della maniglia si collega allo specchio mediante due supporti conformati, nella parte terminale, a zampa di animale. La saldatura fra le due parti è stata operata, con ogni probabilità, con brasatura dolce con lega di piombo stagno e disossidante a resina di pino.6 Al momento del rinvenimento lo specchio era adagiato sullo strato carbonioso (US 182), mentre la maniglia, staccatasi probabilmente contestualmente alle attività di disturbo della sepoltura, è stata rinvenuta poco al di sopra nello strato che ricopriva quanto rimaneva ancora di indisturbato della sepoltura (US 177).

In merito alla provenienza ed alla datazione del pezzo veronese, pare ragionevole pensare possa trattarsi, alla luce delle somiglianze dimensionali, della più consistente diffusione in ambito renano e della decorazione, di un manufatto di III sec. d.C. ascrivibile ad ambito romano occidentale. Tale ipotesi concorderebbe con l’area di provenienza di alcuni altri manufatti facenti parte del corredo, rimandando anche in questo caso ad una produzione forse di area gallica o germanica.

Specchi con prese posteriori e piattelli non perfettamente piatti (la parte riflettente risulta essere quella convessa, e grazie alla forma ha potere ingrandente), iniziano a comparire, seppure con esigue attestazioni, nella parte occidentale dell’impero non prima del I sec. d.C. (diametri inferiori ai 10 cm e

Lo specchio di Vicolo Calcirelli è senza dubbio di dimensioni più contenute e di fattura meno elaborata

Ringrazio sentitamente Stefano Buson, funzionario restauratore conservatore del Museo Archeologico Nazionale di Altino, per la preziosa consulenza. 6 

207

Valentina Mantovani rispetto agli altri esemplari noti ma si sottolinea che, così come gli altri reperti metallici della sepoltura e qui analizzati, non sono ancora stati sottoposti a restauro, e la rimozione delle incrostazioni e della patina superiore che avvolge il manufatto potrebbero forse rivelare la presenza di eventuali ulteriori decorazioni. Si auspica, in questo senso, un intervento di pulizia dei reperti metallici, al fine di poterne definire con maggior cura i dettagli.

1

Tra il materiale metallico è stato rinvenuto anche uno stilus in bronzo integro munito di spatola e punta dritta, la cui asticciola è finemente decorata con modanature tortili (US 182/5) (Tavola 1: 11). Il manufatto corrisponde al gruppo tipologico P 57 identificato da V. Schaltenbrand Obrecht (Schaltenbrand Obrecht 2012) per i materiali di Augusta Raurica che ne colloca la produzione tra la seconda metà del II sec. d.C. e il III sec. d.C. Analoghi stili scrittori si segnalano a Conimbriga (Alarcão 2000: 123, scheda 367 n. 4), Aquileia, Lanciano, Napoli, Ostia, Roma, Virunum, Augst, Munigua, Intercisa e Brigetio (Schaltenbrand Obrecht: 170).

Tavola 2. Verona, via Cantore 2a: frammento di bicchiere in ceramica metallescente di Treviri (disegno: V. Mantovani).

cavaliere sopra a cavallo). In esergo tre globetti (Arzone et al. 2015: 373, nota 1). Le altre monete, tutta di zecca romana, sembrano seguire una certa progressione cronologica che va da l’asse di Domiziano (82 d.C.) (VR 88773) fino ad un asse di Commodo (184-186 d.C.) (VR 88768). Quest’ultima moneta risulta essere di circa 60 anni più antica del reperto più recente facente parte del corredo, ovvero il bicchiere di Treviri.

Nel corredo si segnala anche un candeliere in bronzo a doppia coppa emisferica dotato del supporto cilindrico concavo per la candela (US 182/17) (Tavola 1: 12), corrispondente al gruppo 1/1 identificato dalla Ciugudean (Ciugudean 2005: 139). Il manufatto trova piena concordanza con analoghi candelieri rinvenuti ad Alba Iulia (Ciugudean 2005: fig. 2), Brumath (Higelin 2015a: 154), Horbourg-Wihr (Higelin 2015b: 329) e Colonia (Boeselager 1993: 284, fig.1, n. 6) ma attestati in numerosi altri siti della Gallia, della Germania e del Noricum (Goethert 1994: 358-366; Sakl-Oberthaler 1997; Piccottini 1987: 21-22, Abb. 18).7 Si tratta di manufatti la cui diffusione si colloca tra l’età flavia e la metà del III sec. d.C. (Ciugudean 2005: 139), anche se il nostro esemplare potrebbe essere riferibile, più nello specifico, al periodo compreso tra Marco Aurelio e Commodo.8

Via Cantore 2a Il secondo esemplare di bicchiere di Treviri proviene dalle indagini condotte in via Cantore 2a nel 2014. Si tratta di un piccolo frammento di spalla, pertinente ad un bicchiere, decorato con sovradipintura bianca. Il manufatto è stato recuperato in un livello superficiale del sito che ha restituito un contesto esteso con materiale compreso tra la media età imperiale e la prima metà del IV d.C. (Bruno in press) (Tavola 2: 1).9 La decorazione è composta da un giro di rotella sotto il quale si sviluppano una catena (Künzl 1997, Typentafel 9.14) e sottostanti elementi sinuosi sovradipinti con barbottina bianca di difficile lettura. In base alla porzione di profilo conservatosi, il frammento potrebbe essere riferibile alla serie dei bicchieri tronco-conici Symonds 1- gruppi 36-40/Niederbieber 33/Trier 1, tipici della metà del III sec. d.C..

In stato di conservazione più degradato versa, infine, un elemento allungato in ferro (lunghezza cm 10, larghezza cm 2,6 ca.) (US 182/4) (Tavola 1: 13), il cui stato di conservazione, in assenza di restauro, non permette di indentificarne con precisione la forma.

Considerazioni di carattere economico e sociale

Delle 13 le monete attestate (Tavola 1: 14) — indicate nel catalogo della pubblicazione come “ripostiglio”, e al quale si rimanda per ulteriori dettagli numismatici (Arzone et al. 2015) — sei sono giunte a noi combuste e deformate. L’esemplare più antico è una moneta celtica di zecca gallica in bronzo (VR 88774), datata al I sec. a.C. raffigurante, al dritto, testa maschile a d.(?) tra due globetti, mentre al rovescio un cavallo a d. (o

Il rinvenimento di due esemplari di bicchieri di Treviri a Verona riapre il dibattito in merito alla loro presenza a sud delle Alpi. I materiali noti in area italica sono per la maggior parte privi di contesto di rinvenimento, o provenienti da stratigrafie in cui avevano carattere residuale (Mantovani 2016: 462-463). Le ipotesi relative al loro

Si rimanda alla pagina http://artefacts.mom.fr/result. php?id=CHL-4001 per un elenco più completo dei siti. 8  Si ringrazia Helga Sedlmayer per la preziosa consulenza. 7 

9 

208

“Apertura Area Nord, 27-08-2014” (Bruno in press).

La ceramica metallescente di Treviri: nuovi dati da Verona (Italia)

rinvenimento vertono attorno alla mobilità di persone e all’eventuale testimonianza diretta di commercio tra la Gallia Belgica e la Pianura Padana, in relazione allo sfruttamento dei canali di comunicazione già esistenti (via dell’Ambra e il valico alpino di Iulium Carnicum per Aquileia, ad esempio) (Mantovani 2016: 460-461). Per il caso specifico dell’emporio nord-adriatico non si possono escludere nemmeno legami più o meno diretti con figure militari attive nel territorio, la cui presenza è documentata da due iscrizioni cittadine in cui si menziona un Valerius Valens, signifer della Legio XIII Gemina, con la coppia consolare del 244 d.C. (Tiberio Pollenio Armenio Peregrino e Fulvio Emiliano) (CIL V, 8237 = Inscr. Aq. 297; CIL V, 808 = Inscr. Aq. 309).

Considerata la diffusione nelle medesime aree di coevi specchi con presa posteriore, e i confronti con gli esemplari di Aquileia e di Vienne, pare verosimile che anche lo specchio della sepoltura veronese possa essere inquadrato all’interno della produzione occidentale dell’Impero. I restanti materiali, considerate anche le condizioni frammentarie in cui sono giunti a noi, non aggiungono ulteriori informazioni al contesto. In assenza di analisi dei reperti osteologici, e consci del fatto che la sepoltura era fortemente disturbata, e pertanto con ogni probabilità non completa, le uniche osservazioni che possono essere qui mosse in merito al sesso del defunto, sono relative alla composizione del corredo conservatosi. La presenza dello specchio rimanda alla sfera femminile, anche se va ricordato che in ambito renano sono documentate sepolture maschili con specchi spesso in associazione con rasoi (Bridger 1996: 261). Un ulteriore indizio che porta a presupporre che possa trattarsi di un defunto di sesso femminile proviene dal confronto più specifico con gli specchi con maniglia a fascia rinvenuti in sepolture, per le quali le analisi antropologiche hanno dimostrato trattarsi di tombe femminili (Lerner 1996: 31). La presenza del bicchiere di Treviri non è, come già detto, elemento discriminante per determinare la tipologia della sepoltura, mentre lo stilo sarebbe più propriamente un oggetto riferibile alla sfera maschile. Si ricorda, infine, che la sepoltura padovana corredata dalla bottiglia pseudo-diatretum è stata attribuita, anch’essa ad una donna.

In ogni caso, l’attestazione di questi manufatti nell’importante porto adriatico sono un chiaro segnale della presenza, nella Aquileia di III sec. d.C., di un ceto agiato con una capacità di acquisto tale da permettersi stoviglie di pregio e pienamente amalgamato alla koiné economica e culturale di quel periodo storico. Un substrato sociale, forse identificabile con personalità militari presenti, proprio in quegli anni, nel territorio aquileiese. Analoghe osservazioni paiono essere calzanti ai due rinvenimenti veronesi. A tal proposito merita un approfondimento l’esemplare rinvenuto nella sepoltura di Vicolo Calcirelli. In ambito renano, la ceramica metallescente di Treviri è discretamente documentata all’interno delle sepolture, ove compare sovente in associazione a corredi di un certo pregio caratterizzati da vetri finemente lavorati, gioielli d’oro e di ambra e vasellame di bronzo (Höpken and Liesen 2013: Abb. 53, 62, 65, 116; Harsányi 2013b: 348). In particolare, le indagini antropologiche effettuate sui resti dei defunti pertinenti a sepolture di Colonia mettono in luce la presenza di ceramica metallescente di Treviri sia in tombe maschili che femminili (Höpken and Liesen 2013).

La parzialità del dato archeologico raccolto, lo stato di conservazione di alcuni manufatti, probabilmente esito di attività di disturbo, e l’assenza di un restauro sui materiali metallici non permettono di spingersi oltre. Alcune altre osservazioni possono però esserci di aiuto a meglio definire, forse, le origini del defunto. La presenza di vetri di produzione transalpina a Verona era già nota in tre contesti sepolcrali a cremazione inquadrabili tra la fine del II ed il III sec. d.C. (Roffia 1998). Il fenomeno delle importazioni vitree da area transalpina interessa, già dalla fine dei II sec. a.C. e per tutto il III e IV sec. d.C., l’intera area nord-italica, con una concertazione nei territori centrali della Pianura Padana (Roffia 1998: 101), anche in relazione alle vie commerciali di più facile percorrenza. Per Verona, d’altra parte, passava l’arteria della via Claudia Augusta padana che univa il Po con la città, fino a Trento e ai passi Resia e Fern, giungendo ad Submuntorium, nei pressi di Donauwörth, godendo quindi di un collegamento privilegiato con l’asse reno-danubiano (Roffia 1998: 101).

Al di fuori della Gallia Belgica, i rinvenimenti di bicchieri di Treviri in ambito funerario sono rari e ad oggi le uniche attestazioni note paiono essere i a quattro bicchieri facenti parte del corredo di una tomba di Aquincum (Harsányi 2013b: 348, nota 14) e l’esemplare milanese (Airoldi 2011). L’esemplare veronese di Vicolo Calcirelli rappresenta, pertanto, insieme al corredo funebre, un interessante e nuovo dato. I materiali in associazione, ed in particolare il candeliere e i frammenti di decorazione a gabbia in vetro bianco, richiamano in maniera più o meno concreta, non solo per tipologia, ma anche per area di distribuzione, i territori della Gallia e della Germania. 209

Valentina Mantovani Non va dimenticato, poi, che proprio il commercio di vetri d’oltralpe, già nel pieno II d.C., è documentato da iscrizioni sia in Cisalpina che nelle regioni renane in cui si cita un corpo di Negotiatores Cisalpini et Transalpini, i quali avevano il compito di seguire le attività commerciali tra questi territori (ad esempio, Como: ILS 7527; Milano: CIL, V 5911 = ILS 7279). Ancor più interessante è il dato che tale corporazione era in mano, soprattutto, a cittadini di Colonia e di Augusta Treverorum, che svolgevano le attività in diverse città dei territori interessati (Roffia 1998: 102).

Alarcão, A. M. 2000. Museu Monográfico de Conimbriga. Catalogue. Lisbon: Instituto Portugues de Museus. Arzone, A., F. Biondani e D. Calomino 2015. Ritrovamenti monetali di eta’ romana nel Veneto. Provincia III: Verona. Vol. 1 Verona. Padova: Esedra Editrice. Boeselager, D. 1993. Die Beigaben kombination reicher Brandgräber in Köln, in M. Struck (ed.), Römerzeitliche Gräber als Quellen zu Religion, Bevölkerungsstruktur und Sozialgeschichte (Arch. Schr. Des Inst. für Vorund Frühgeschichte der J.G. Universität Mainz 3): 283– 292. Mainz: Institut für Vor- und Frühgeschichte. Bonomi, S. and C. Tagliaferro 2006. Uno pseudo-diatretum da Padova. Quaderni di Archeologia del veneto. Serie Speciale 2: 105–111. Bolla, M. 2000. Militari e militaria nel territorio veronese e gardesano (III-inizi V sec. d.C., in M. Buora (ed.), Miles Romanus dal Po al Danubio nel Tardoantico. Atti del Convegno Internazionale (Pordenone-Concordia 17-19 marzo 2000): 99–138. Pordenone: Consorzio universitario Pordenone. Bridger, C. J. 1996. Das römerzeitliche Gräberfeld „An Hinkes Weisshof“, Tönisvorst-Vorst, Kreis Viersen (Rhein. Ausgr. 40). Köln ‒ Bonn: Rheinland Verlag, Habelt. Bruno, B. La fascia esterna alle mura occidentali di Verona: riletture e nuovi dati. Archeologia Veneta, in press. Brusin, G. 1929. Aquileia. Guida storica e artistica. Udine: La Panarie. Callegher, B. 2007. L’Antichità, in B. Callegher, L. Passera e A. Saccocci (eds) Limiti e potenzialità degli studi sulla circolazione monetaria in ambito locale: il caso delle Valli del Natisone, in Le valli del Natisone e dell’Isonzo tra centroeuropa e Adriatico, Atti Conv. Intern. di Studi, San Pietro al Natisone (UD) 15-16/9/2006: 246‒255. Roma: Quasar Edizioni. Ciugudean, D. 2005. Chandeliers en bronze d’Apulum (II). Apulum 42: 137‒144. Foy, D., F. Labaune-Jean, C. Leblond, C. Martin Pruvot, M-T. Marty, C. Massart, C. Munier, L. Robin e J. Roussel-Ode 2018 (eds). Verres incolores de l’antiquité romaine en Gaule et aux marges de la Gaule. Volume 2. Typologie et analyses (Archaeopress Roman Archaeology 42). Oxford: Archaeopress. Giussani, A.1904. Un sepolcreto romano del tempo di Tiberio a Como. Rivista Archeologica della Provincia e antica Diocesi di Como: 35–67. Goethert, K. 1994. Die römischen Lampen und Kerzenhalter aus Metall im Rheinischen Landesmuseum Trier. Trierer Zeitschrift 57: 315‒374 Guérit, M. and E. Ferber 2015. Découverte de deux ateliers de verriers de la fin du II siécle et du début du III siécle au Pègue (Drome). Bullettin de L’associàtion pour l’Archeologie du Verre 2015: 44–49. Harsányi, H. 2013a. Die Trierer schwarz engobierte Ware und ihre Imitationen in Noricum und Pannonien (Austria Antiqua 4). Wien: UPG.

Il materiale della sepoltura di Vicolo Calcirelli sembrerebbe ben rispondere a questa presenza: si potrebbe trattare, forse, di una donna, facente parte di un ceto sociale agiato che nel III secolo d.C. animava la città di Verona, la quale può essere venuta in contatto diretto o indiretto, con manufatti culturalmente gravitanti attorno ai territori della Gallia, durante un soggiorno in quei luoghi. O potrebbe trattarsi, addirittura, di una di quelle personalità di origine transalpina presenti in ambito padano, alle quali era affidata la gestione del commercio con le aree a nord delle Alpi. Questo scenario ben si accorderebbe con la connotazione tipica di alcuni materiali che ne compongono il corredo, a differenza delle altre sepolture veronesi nelle quali compaiono sporadici vetri di importazione, ma senza altre connotazioni etniche particolari. Un dato che potrebbe forse rafforzare tale ipotesi – pur specificando che quella che si propone è solo una suggestione – è la presenza della moneta celtica di zecca gallica di I sec. a.C. rinvenuta nella sepoltura. Monete in bronzo celtiche del I sec. a.C. si trovano abbastanza capillarmente (ma sempre con pochi esemplari per volta) nei vari siti della pianura padana, così come è nota anche la presenza di monete celtiche padane in sepolture coeve o di pochi anni più recenti (solo per citarne alcune, Callegher 2007: 252, nota 68; Salzani 1987: 277-278; Giussani 1904: 35). Non si è però a conoscenza di monete celtiche di zecca gallica di I sec. a.C. in sepolture di III d.C. Infine, non si esclude che la sua presenza a Verona potrebbe poi essere forse giustificata dal legame con qualche personalità militare, la cui presenza è ben documentata a Verona dai ripostigli del Ceraino, quello della Venéra e da numerose altre testimonianze materiali ed epigrafiche edite (Bolla 2000). Bibliografia Airoldi, F. 2011. Scheda 26. Vasetto di produzione rennana, in S. Lusuardi Siena, M. P. Rossignani, M. Sannazaro (eds) L’abitato, la necropoli, il monastero. Evoluzione di un comparto del suburbio milanese alla luce degli scavi nei cortili dell’Università Cattolica: 110–111. Milano: Vita e Pensiero. 210

La ceramica metallescente di Treviri: nuovi dati da Verona (Italia)

Harsányi, H. 2013b. La ceramica ad ingobbiatura nera di Treviri. Una merce costosa in Pannonia durante l’epoca severiana. American Journal of Ancient History 6‒8: 341‒360. Higelin, M. 2015a. Des objets racontent la vie quotidienne à Brumath à l’époque romaine, in B. Schnitzler (ed.) Brumath Brocomagus: capitale de la cité des Triboques: 154–163. Strasbourg: Mam Strasbourg. Higelin, M. 2015b. Le métal, in G. Alberti, C. Bébien, M. Higelin, S. Vauthier (eds) Horbourg-Wihr, HautRhin, lotissement Kreuzfeld 2. Un quartier périphérique de l’agglomération antique, rapport de fouilles préventives, vol. 3.1.. Analyses, études spécialisées : aspects de la vie matérielle, DRAC: 233–348. Strasbourg: Archéologie Alsace. Höpken, K. and B. Liesen 2013. Römische Gräber im Kölner Süden II Von der Nekropole um ST. Severin bis zum Zugweg. Kölner Jahrbuch 46: 369–571. Künzl, S. 1997. Die Trierer Spruchbecherkeramik. Dekorierte Schwarzfirniskeramik des 3. Und 4. Jahrhunderts n. Chr. (Trierer Zeitschrift Beiheft 21). Trier: Druckerei Ensch GmbH. Lerner, J. 1996. Horizontal-Handled Mirrors: East and West. Metropolitan Museum Journal 31: 11‒40. Loesch, C. 2012. Exquisite Roman Glass at the Frontiers of the Roman Empire. Early 3rd century glass from Tomb 7 of the Rheinbach-Flerzheim necropolis in Germany, in V. Arveiller e H. Cabart (eds) Le Verre en Lorrane et dans les regions voisine (Acte du ccolloque de l’AFAV, Metz, 18 ett 19 novembre 2011). Monogr. Instrumentum 42): 63–74. Montagnac: Editions Mergoil. Mantovani, V. 2016. Ceramiche fini da mensa di età medio imperiale ad Aquileia: la ceramica metallescente di Treviri, in M. Buora e S. Magnani (eds) Le iscrizioni con funzione didascalico-esplicativa. Committente, destinatario, contenuto e descrizione dell’oggetto nell’instrvmentvm inscriptvm. Atti del VI Incontro Instrvmenta Inscripta, Aquileia (26-28 marzo 2015) (Antichità Alto Adriatiche LXXXIII): 453–464. Trieste: EUT. Mantovani, V. in press. Identità culturale, mode, imitazioni e ispirazioni. La ceramica come fine, la ceramica come mezzo, in Pottery from Ancient to Contemporary World. Materials, Economies, Societies, (and Forgeries), atti della International Spring School.

Piccottini, G. 1978. Archäologische Mitteilungen aus Kärnten III. Carinthia I 168: 5‒33. Portulano, B. 2002. Le iscrizioni graffite sui manufatti ceramici, in F. Rossi (ed.) Nuove ricerche sul Capitolium di Brescia. Scavi, studi e restauri: 395– 406. Milano: Edizioni Et. Salzani, L. 1987. La tomba 4 della necropoli di Valeggio sul Mincio (Verona), in D. Vitali (ed.) Celti ed Etruschi nell’Italia Settentrionale dal V secolo a.C. alla romanizzazione. Atti del colloquio internazionale (Bologna 12-14 aprile 1985): 277‒278. Bologna: University Press. Sakl-Oberthaler, S. 1997. Beleuchtungsgerät I-II, in H. Ubl (ed.) Katalog der Schausammlung „Römerzeit“ des Museums Lauriacum - Enns (Forschungen in Lauriacum 12/2): 266 Nr. V/L-1. Enns-Wien: Klosterneuburg. Schaltenbrand Obrecht V. 2012. Stilus. Kulturhistorische, typologisch-chronologische und technologische Untersuchungen an römischen Schreibgriffeln von Augusta Raurica und weiteren Fundorten (Forschungen in Augst 45/1). Augts: Augusta Raurica. Symonds, R. P. 1992. Rhenish wares. Fine dark coloured pottery from Gaul and Germany (Oxford University Committee for Archaeology. Monograph 23). Oxford: Oxbow Books. Vilvorder, V. 2010. La céramique métallescente de Trèves, in T. Brulet, F. Vivorder e R. Delage (eds) La céramique romaine ed Gaule du Nord: 351‒355. Turnhout: Brepols. Whitehouse, D. 1988. Kantharos Dish, in D. B. Harden (ed.) Vetri dei Cesari, catalogo della Mostra (Roma, Musei capitolini, Novembre 1988): 253–254. Milano: Arnoldo Mondadori S.p.A. Siti web https://www.cmog.org/artwork/disch-cantharus?ima ge=0&search=collection%3Ae472320258ff81d9cf60b 81270a27444&page=1 http://artefacts.mom.fr/result.php?id=CHL-4001

211

Le anfore galliche nella Regio X: dati preliminari Andrea Cipolato, Giuseppe Indino Abstract This article is a preliminary contribution of a wider work which goal is carried out on the presence of Gallic amphorae in the Augustan Regio X. These containers, unlike other geographical areas of the Peninsula and the Mediterranean basin, are very rare in Regio X (Venetia et Histria) between the 1st and 4th centuries, judging by the data in our possession. The amphoric typology most present is the Gauloise 4, used for the transport of wine but probably also of other foodstuffs, coming from ateliers in Narbonne Gaul of which the one in Sallèles d’Aude is the most investigated and best known. At the moment, only some important centers such as Oderzo, Altino, Aquileia, Trieste, Emona have returned fragments of these containers. Other fragments are reported along roads and rivers, endo-lagoon routes such as the Postumia, the Annia, the Po-Adige-Brenta-Piave-Tagliamento system and the lagoons of Venice and Grado that characterize the region’s trade. This is the case of Monfalcone on the Adriatic Sea or sites close to the lagoon belt such as Mirano, Oderzo and Canale San Felice and Lio Piccolo in the same lagoon of Venice. Others have been found along the Istrian coast, in Skolarice and Rovinj. Compared to the rarity of these containers in Regio X, we can make some valid arguments in this regard. A very plausible argument is the following: in the Adriatic, food products such as oil, wine, fish and derivatives travel in other amphoric containers of local and regional production (Illyrian, southern, Greek, Hispanic, African and eastern Italy). Nevertheless, another assumption can be opposed to this: Gallic wine, from the middle of the first century, travelled in wooden barrels whose remains have been found in a number along the road axes of the Rhine and Danube Limes, especially in Roman military camps. A single fragment of a barrel has been found in our study region, precisely on the island of Grado, near Aquileia. What are the causes of such a shortage of Gallic amphorae in Regio X? What are the transit and supply routes for these products? Who are the buyers and consumers of Gallic wines? If so, can we talk about an actual trade or a sporadic consumption of Gallic wine in Regio X? Is there a link between the circulation of Gallic sealed wines, particularly from southern Gaul, and these containers? Were the diffusion flows used the same? A series of reflections are proposed here with a view to wider and deeper work. Key words Gallic amphorae, Regio X, Gallia Narbonensis, trade, wine, Adriatic Sea, Altino, Venice Lagoon, Aquileia, Grado Lagoon.

Introduzione del fenomeno delle anfore galliche La produzione di contenitori da trasporto di origine gallica conosce a partire dal I secolo d.C. una notevole distribuzione in molti mercati dell’Impero, dall’Europa Centrale fino al Medio Oriente e in Asia. La conquista romana della Gallia meridionale e la creazione della provincia della Narbonese aprì la possibilità a molti veterani dell’esercito di Roma di sfruttare i nuovi e vasti territori a disposizione per fini agricoli. Molteplici flussi commerciali si aprirono in breve tempo tra l’area di recente annessione ed il mondo Mediterraneo. Infatti, se durante l’ultima fase della Repubblica (II-I secolo a.C.) la Narbonese importa massivamente vino dalla penisola italica (Tchernia 1986), verso la fine del I secolo a.C. si verifica una graduale inversione di tendenza grazie ad un importante sviluppo dei vitigni locali (Nat. Hist. XIV, 26, 43, 132; XVII, 2, 21, 116) in grado di garantire il fabbisogno vinario della popolazione e di esportarne il surplus sia su scala regionale, sia – successivamente – verso mercati internazionali.

All’inizio della fase augustea nascono i primi ateliers di produzione d’anfore che ricalcano modelli morfologici italici ed iberici; i contenitori subiscono una distribuzione modesta lungo l’asse Rhône-SaôneRhin (Desbat e Martin Kilcher 1989), verso Bordeaux (Laubenheimer e Watier 1991) e ad Ostia (Hesnard 1989). Così a Marsiglia (Bertucchi 1992: 130) e nel Vaucluse (Carru et al. 1997) si ricalcano le forme Dressel 1, le Dressel 2-4 nelle officine di Corneilhan nel dipartimento de l’Hérault (Laubenheimer 1985: 316), dove vengono riprodotte anche le Pascaul 1, quest’ultime fabbricate anche ad Aspiran e a Montans nel dipartimento del Tarn (Laubenheimer 1985). Altre forme invece vengono ideate localmente, come la Gauloise 2 a fondo piatto prodotte a Marsiglia e a Sainte-Cécile-les-Vignes (Meffre 1992). È solo con l’Alto Impero (I secolo d.C.) che i prodotti gallici subiscono una distribuzione caratterizzata da tratti più massivi, anche su direttrici a lungo raggio per oltre quattro secoli. Nascono le forme Gauloise 1 nella bassa valle del Rhône e le Gauloise 5, esportate fino in Britannia, mentre l’emblema di tale fenomeno commerciale rimane senza dubbio la Gauloise 4, di gran

Roman Pottery and glass manufactures (Archaeopress 2022): 212–221

Le anfore galliche nella Regio X: dati preliminari

Figura 1. Mappa dell’incidenza dei traffici anforici di origine gallica nella penisola italica e nelle regioni limitrofe (mapped by authors; basemap: Google Earth).

lunga il contenitore gallico più attestato nei mercati romani, maneggevole e facilmente trasportabile (10 kg di tara ed un massimo di 30 l di derrate) e che arriva a servire anche alcuni siti del Golfo Persico e dell’India (Begley et al. 1996). La sua diffusione si accresce progressivamente nel corso del Medio Impero anche grazie alle varietà di vino che trasportava, segnalato dai tituli picti rinvenuti sulle superfici dei contenitori, come il picatum (a fermentazione sospesa durante la vinificazione) e l’amineum, vitigno di possibile origine greca, ma diffuso in Italia e nelle province (Laubenheimer 2001: 55; Pecci et al. 2013).

dell’edito si ravvisa una scarsa attenzione ai dati quantitativi, evidenziati solamente in alcuni principali contributi, aspetto che comporta inevitabilmente una difficile e parziale lettura analitica del fenomeno. Il presente lavoro costituisce un primo tentativo di colmare tale lacuna nella ricerca e mira a comprendere maggiormente le modalità e le tempistiche attraverso le quali si è coniugato il fenomeno in Italia e nelle regioni limitrofe ad essa (Raetia, Noricum, Slovenia, Croazia), ponendo particolare attenzione agli indici di presenza di anfore galliche nella Regio X, caratterizzati da numeri alquanto inferiori rispetto ad altre realtà (Figura 1).

Gli atelier più attivi sembrano essersi raggruppati in una vasta area corrispondente alla Narbonese, in cui quello di Sallèles d’Aude è quello più indagato e conosciuto. Ulteriori figlinae sono state rintracciate nel nord della Gallia (Centro-Valle della Loira e GrandeEst), nella Tarraconense e nella regione di Alicante, a Denia (Laubenheimer 2001: 33-50), la cui produzione era prevalentemente legata all’esportazione del vino prodotto in quelle regioni verso aree della Britannia, dell’Europa continentale, dell’Egitto e dell’India (Laubenheimer 2001: 51-66).

Nel corso del Medio Impero le anfore ed i prodotti alimentari gallici assumo una certa rilevanza nei centri costieri tirrenici e dell’entroterra. La loro ricezione è ben testimoniata ad Ostia, Roma, Settefinestre, Toscana e Liguria, nonché in vari relitti rinvenuti lungo le coste mediterranee francesi e nel Mar Tirreno in Sardegna e Corsica (Tito 2013). In questa fase gran parte della considerevole produzione vinaria della Narbonese veniva assorbita dall’Urbe; infatti, nonostante una timida presenza di galliche testimoniata nel contesto romano di via Nova/Clivo Palatino tra il 64-68 d.C. (0,2%), la presenza aumenta considerevolmente in svariati siti urbani nel corso del secolo successivo, fino alle attestazioni nei livelli archeologici compresi tra il 90-110 d.C. di via Sacra/ via Nova di percentuali che si aggirano ad ¼ del totale

Analisi della distribuzione nella Penisola italica e nelle regioni limitrofe La Penisola italica è risultata finora carente di studi sistematici circa la presenza di contenitori anforici provenienti dalla Gallia. Da un preliminare spoglio 213

Andrea Cipolato, Giuseppe Indino

Figura 2. Mappa di distribuzione dei contenitori gallici nella Regio X (mapped by authors; basemap: Google Earth).

(Carandini et al. 2006; Panella 1992). Quantità analoghe sono ravvisabili nel contesto delle Terme del Nuotatore ad Ostia nei livelli compresi tra il 160 ed il 180 d.C. (Panella 1992: 185-206), mentre a Settefinestre questi contenitori raggiungono verso la metà del II secolo d.C. oltre il 50% delle presenze totali (Panella 1989: 139-178).

Il commercio gallico ritorna su numeri quantitativamente significativi lungo le regioni d’oltralpe, come a Cambodunum nell’antica provincia della Raetia con circa il 23% dei contenitori totali, mentre a Curia scende al 12%, fino a toccare il minimo dell’1% a Favianis nel Noricum (Schimmer 2015) e nel Magdalensberg (Bezecsky 1998).

Una rete di porti e approdi sicuri doveva garantire il flusso commerciale dei prodotti provenienti dalla Narbonese lungo la costa ligure, mentre l’arteria fluviale rappresentata dal Tanaro ne favoriva la penetrazione in discreta quantità anche nei mercati dell’entroterra. Pertanto, nella Regio IX si registrano percentuali che si aggirano attorno al 10% ad Alba Pompeia (Alba) ed Augusta Bagiennorum (Bene Vagienna, CN), mentre a Genua le percentuali sembrano ridotte (rispettivamente Bruno 1997; Quiri 2014; Milanese 1993). Anche per i contesti studiati ad Augusta Taurinorum (Torino), Eporedia (Ivrea), Biella (necropoli di Cerrione) ed i materiali d’Industria a Monteu da Po gli indici di attestazione della direttrice gallica non raggiungono l’1% (rispettivamente Brecciaroli Taborelli and Gabucci 2007: 255-256; Gabucci e Quiri 2008: 45-78; Brecciaroli Taborelli 2000: 31‒35; Quiri 2011: 107-116; Zanda 2011: 31-32). La marginalità del dato in tali mercati potrebbe dipendere dalla presenza molto più incisiva di merci come olio istriano e vino pugliese provenienti dal versante adriatico, i quali provvedono da soli ad almeno la metà del fabbisogno locale, mentre in misura inferiore si registrano prodotti dall’Egeo e dall’area siro-palestinese.

Anche lungo entrambi i versanti del mare Adriatico i dati offerti risultano omogeneamente attestati su percentuali che non raggiungono l’1%, come a Brindisi negli scavi di via S. Chiara e del cortile del Vescovado (Auriemma et al. 2012), ma anche a Durazzo (Albania) nello scavo del bastione meridionale (Bolzoni 2012) e nell’Adriatico centrale lungo le coste italiche a Suasa (Gamberini 2014: 535, 538) e quelle croate a Zaton, antico porto di Enona (Gluščević 2007: 13). Alla luce dei dati finora collazionati sembrerebbe che la vicinanza ai principali canali di afflusso commerciale dei prodotti alimentari largamente diffusi in Adriatico coincida con un decremento della presenza dei contenitori gallici nel quadro degli equilibri complessivi delle merci in un sito. Infatti, i principali hub commerciali dell’Alto Adriatico (come Ravenna, Altino ed Aquileia) dovettero giocare il ruolo di capolinea per le grandi vie di penetrazione nell’entroterra (sia lungo la Pianura Padana verso ovest sfruttando la rete fluviale del Po e/o gli assi stradali esistenti in un sistema di comunicazione integrato, sia verso nord fino al limes danubiano) delle merci che venivano dal mare, prevalentemente d’importazione dall’Adriatico e dall’Egeo. 214

Le anfore galliche nella Regio X: dati preliminari

Un diretto riscontro è offerto dalla recente attività di ri-analisi di gran parte dei materiali anforici depositati nei magazzini della Soprintendenza Archeologica di Venezia e provenienti da alcuni siti archeologici lagunari indagati a partire dagli anni ‘70-‘80 del secolo scorso (Cipolato 2018b). L’area era segnalata come priva di provenienze anforiche di origine gallica, ma è stato possibile individuare ben 11 frammenti corrispondenti a 4 esemplari di Gauloise 4 ed 1 di Gauloise 5 della metà del II secolo d.C. (Tavola 1.3-4; Tavola 2.5), dato assolutamente innovativo per il panorama locale, dal momento che gli unici frammenti di produzione della Narbonese erano attestati ad Altino (Cipolato 2018a; Toniolo 1991). Gli esemplari erano stati usati come materiale di riempimento in alcune strutture arginali sommerse di età romana (Fozzati and Toniolo 1998), ma sono testimoni della circolazione di tale direttrice anforica, che non doveva servire solamente il centro altinate, ma anche il relativo comprensorio occupato da ville marittime ed impianti produttivi (Calaon and Cipolato 2019).

La situazione delle anfore galliche nella Regio X A fronte di un modello di distribuzione dei contenitori gallici alquanto eterogeneo con picchi di presenza distribuiti lungo i principali centri della costa tirrenica e le regioni d’oltralpe, i dati che provengono dalla Regio X (Venetia et Histria) appaiono alquanto omogenei nel restituire una situazione di assoluta penuria di anfore provenienti da questa direttrice commerciale (in nessun sito mai eccedenti l’1%). Una presenza esigua, ma attestata abbastanza capillarmente in molti dei principali siti e territori della zona (Figura 2), è rappresentata principalmente dalla forma Gauloise 4 nelle sue varianti della seconda metà del I secolo d.C. e del secolo successivo (Tabella 1). Nei principali centri dell’area come Opitergium (Tavola 2.7), Altinum (Tavola 1.1-2), Aquileia (Tavola 2.6) e Tergeste (Tavola 3.8) la percentuale d’incidenza delle anfore della Narbonese è di circa l’1% del totale delle forme anforiche (rispettivamente Cipriano and Ferrarini 2001; Cipolato 2018a e Toniolo 1991; Bueno et al. 2012; Auriemma et al. 2012). Nella laguna di Venezia la forma Gauloise 5 è stata rinvenuta in associazione con quella Gauloise 4, di cui altri frammenti provengono da siti costieri, come Monfalcone, o prossimi la fascia lagunare, come Mirano (Pascale 1998). Lungo l’opposta sponda croata a Skolarice viene segnalata la presenza di anfore Gauloise 5 (Tavola 3.9), datate tra la fine del I e la prima metà del II secolo (Žerjal 2008: 467), mentre nel porto di Vestar – all’interno della baia di Rovigno – sono state ritrovate anfore Gauloise 1 e Gauloise 3 (Bekić and Pešić 2014: 123124). Probabili attestazioni sarebbero presenti anche ad Emona (materiali inediti), dove le merci potevano giungere con facilità grazie al ruolo di collegamento con l’entroterra ed i siti costieri svolto dal fiume Sava.

Un primo bilancio della ricerca Alla luce dei dati desunti è possibile avanzare un primo bilancio relativo alla portata del fenomeno gallico nell’area alto adriatica. L’esigua quantità di esemplari anforici nei contesti porterebbe a considerare il fenomeno non tanto come un vero e proprio commercio, quanto piuttosto come l’esito di limitati contatti internazionali tra i due comprensori o addirittura residuo di semplici dotazioni di bordo delle imbarcazioni che raggiungevano i porti adriatici. Le tipologie anforiche testimoniano l’importazione di vino gallico avvenuta in un preciso momento cronologico, ovvero tra la metà del I e la successiva metà del II secolo d.C.

Nel complesso, è possibile inquadrare cronologicamente le morfologie presenti nei contesti tra la metà del I e la metà del II secolo d.C., ma sono limitati i casi in cui sia risultato possibile associare i materiali ai livelli stratigrafici di riferimento, con una conseguente difficoltà a delineare con sistematicità e su larga scala le tempistiche e le modalità di afflusso di questi materiali nei siti in esame. Infatti, tranne per le indagini stratigrafiche di Aquileia, Oderzo, Trieste e Chiunsano (rispettivamente Bueno et al. 2012; Cipriano e Ferrarini 2001; Auriemma et al. 2012; De Zuccato 2016), le restanti attestazioni sono frutto di recuperi subacquei, di segnalazioni avvenute negli ultimi anni del secolo scorso e provenienti da depositi museali.

In questa fase storica l’arco adriatico è servito principalmente da vini di produzione “locale” (apuli, centro-italici, emiliano-romagnoli, padani, croati), ma è attestata anche la direttrice di traffico orientale rappresentata in particolar modo dall’Egeo (Cipriano e Mazzocchin 2018 e 2011; Auriemma et al. 2016). Questo fattore dovette rappresentare un limite alla proliferazione di prodotti vinari gallici, esitanti ad inserirsi in un mercato già saturo. Dal quadro finora emerso sembrerebbe che le rotte marittime e fluviali fossero preferite a quelle terrestri per la veicolazione dei prodotti dalla Narbonese. Ma quali possono essere stati nello specifico i canali preferenziali di transito che consentivano non solo alle anfore, ma anche alla produzione di sigillate galliche l’arrivo nella Regio X?

Dall’attività di ricognizione della documentazione edita è emerso con evidenza l’annosità ormai marcata dei contributi, che necessiterebbero di una revisione relativa perlomeno all’interpretazione tipologica dei frammenti anforici, alla luce dei nuovi apporti scientifici che ha saputo dare la ricerca negli ultimi anni.

La rotta marittima che collegava i porti mediterranei del Golfo del Leone come Narbonne e Arles gli scali italiani 215

Andrea Cipolato, Giuseppe Indino Sito

Tipologia Contesto Rinvenimento

Tipologia Anforica

Frammenti / Nmi

Chiunsano (RO)

villa rustica

scavo archeologico

gallica

1 fr. / 1 NMI

Mirano (VE), loc. Castelliviero di Zianigo

funerario

rinvenimento fortuito

gallica?

1 fr. / 1 NMI

II d.C.

Altino (VE), loc. Ghiacciaia

quartiere residenziale della città antica

survey

Gauloise 4

2 fr. / 2 NMI

II d.C.

deposito

Gauloise 4

1 fr. / 1 NMI (bollo APM)

II d.C.

Laguna di Venezia (VE), Canale S. arginatura Felice (Palude Bombaio)

indagini subacquee

Gauloise 5

4 fr. / 1 NMI

½ II d.C.

Laguna di Venezia (VE), Canale S. Felice (Palude Bombaio) arginatura

indagini subacquee

Gauloise 4

6 fr. / 3 NMI

½ II d.C.

Laguna di Venezia (VE), Lio Piccolo

villa rustica

indagini subacquee

Gauloise 4

1 fr. / 1 NMI

½ II d.C.

Meolo (VE), loc. I Forni

agricolo/produttivo rinvenimento fortuito gallica?

1 fr. / 1 NMI

I d.C.

Oderzo (TV), via Spinè

bonifica (drenaggio scavo archeologico V e XIII)

Gauloise 4

2 NMI

I d.C.

Oderzo (TV), via Spinè

bonifica (drenaggio scavo archeologico VII)

Dressel 2-4

1 NMI

II ½ I d.C.

Aquileia (UD), ex fondo Sandrigo

porto fluviale, sponda orientale

Gauloise 4

1 fr. / 1 NMI

II d.C.

Gauloise 4

2 fr.

residuale in livelli tardo antichi

Museo Archeologico Nazionale di Altino (VE)

Aquileia (UD), domus delle Bestie Ferite

domus

scavo archeologico scavo archeologico scavo archeologico

Aquileia (UD), Canale Anfora

canale portuale

Monfalcone (GO), fiume Timavo

Datazione

in livelli di ½ I – fine III d.C.

gallica recupero subacqueo Gauloise 4

1 fr. / 1 NMI

Duino (TS), III ramo del Timavo

discarica in canale fluviale

recupero subacqueo Gauloise 4

1 fr. / 1 NMI

I – ½ II d.C.

Trieste (TS), Park S. Giusto

villa suburbana

scavo archeologico

Gauloise 4

1 NMI

residuale in livelli di III-IV d.C.

Skolarice (Croazia)

villa

scavo archeologico

Gauloise 5

Rovigno (Croazia), porto di Vestar

recupero subacqueo Gauloise 1, Gauloise 3

Ljublijana (Slovenia)

scavo archeologico

fine I – ½ II d.C.

Gauloise 4

Tabella 1. Siti di rinvenimento di anfore galliche nella Regio X.

della costa tirrenica (dalla Liguria, Toscana, Lazio e Campania) è testimoniata dalle evidenze di numerosi relitti ritrovati lungo le coste sarde e toscane (Cibecchini 2017: 119-130; Beltrame and Cipolato 2019: 25-32; Galasso 1997: 121-134; Leoncini 2007: 6-15; Alberti et al. 2015: 7475). Le anfore galliche, una volta sbarcate in queste zone, avrebbero raggiunto la Cisalpina e poi i territori della Regio X attraverso la fitta rete integrata del sistema viario terrestre e fluviale (la massiccia presenza di sigillate galliche a Ventimiglia potrebbe supportare questo canale di ricerca). Verosimilmente venivano utilizzate navi da carico di medio-grandi dimensioni, mentre per le vie interne alla Penisola si passava ad imbarcazioni fluviali

più piccole. Nei tratti terrestri si può ipotizzare l’uso, oltre ai carri cisterna, delle botti, ben attestate dalle fonti (De bello civili II, 11, 2; De bello gallico VIII, 42, 1) e da diversi resti lignei rinvenuti lungo il limes renano-danubiano ed a Grado (Marlière 2001: 80-81). Diversamente, appare poco plausibile ed economicamente svantaggiosa la rotta marittima che prevedeva il periplo dello stretto di Messina e la risalita dell’Adriatico fino ai porti lagunari di Altino, Aquileia e dell’Histria. Ad esclusione della buona presenza di sigillate galliche nel meridione italico (il cui centro di distribuzione poteva essere rappresentato da Ostia 216

Le anfore galliche nella Regio X: dati preliminari

Tavola 1. 1. AL 9172, Gauloise 4 dal Museo Archeologico Nazionale di Altino (Toniolo 1991); 2. AL 61060, Gauloise 4 da Altino, loc. Ghiacciaia (Cipolato 2018a); 3. IG 70910, Gauloise 5 da Laguna nord di Venezia, Canale S. Felice (Cipolato 2018b); 4. IG 70909, Gauloise 4 da Laguna nord di Venezia, Canale S. Felice (Cipolato 2018b) (elaborazione grafica A. Cipolato).

217

Andrea Cipolato, Giuseppe Indino

Tavola 2. 5. IG 341583, Gauloise 4 da Laguna nord di Venezia, Lio Piccolo (Cipolato 2018b); 6. 140.064, Gauloise 4 da Aquileia, porto romano – sponda orientale (disegno di A. Cipolato); 7. Gauloise 4 da Oderzo, via Spinè (Cipriano e Ferrarini 2001) (elaborazione grafica A. Cipolato).

218

Le anfore galliche nella Regio X: dati preliminari

Tavola 3: 8. Gauloise 3 da Trieste, Park S. Giusto (Auriemma et al. 2012); 9. Gauloise 5 da Skolarice (according to: Auriemma et al. 2008).

o Roma), i siti di accesso all’Adriatico non hanno restituito – come si è visto – significativi quantitativi ceramici che possano supportare questa tesi.

Un altro asse si sarebbe potuto sviluppare attraverso il bacino reno-danubiano fino al Norico, in cui da Virunum le merci sarebbero scese verso il centro di Aquileia, servendo principalmente i mercati della Venetia nord-orientale. Tale rotta ricalcherebbe quella di distribuzione delle sigillate prodotte nel sud della Gallia ed indicherebbe la scelta da parte dei mercatores di sentieri preferenziali in cui i convogli trasportavano sia vettovaglie utili per l’esercito disposto lungo il limes, sia vino di produzione narbonese. Non è esclusa pertanto una domanda vinaria legata ad esigenze militari o alla semplice richiesta da parte di veterani che abitavano i territori della Regio X.

La morfologia stessa delle anfore galliche rinvenute nella Regio X (prevalentemente a fondo piatto) ben si sposerebbe con un trasporto su carri, chiatte o imbarcazioni a basso pescaggio. Rimane pertanto aperta l’ipotesi di un tragitto commerciale interamente ricoperto da percorsi terrestri lungo il versante alpino occidentale, ovvero le stesse rotte ricalcate per il trasporto di sigillate dalla Gallia centrale alla Cisalpina. I passi alpini del Monginevro, del Moncenisio o del Piccolo San Bernardo rappresentavano un percorso preferenziale (Arnaud 2008: 5), almeno nei mesi estivi, via Lugdunum e Vienne, per poi scendere nella Val di Susa verso Augusta Taurinorum e nelle valli minori, congiungendosi poi direttamente all’arteria stradale della Postumia.

I dati qui presentati sono frutto delle ricerche preliminari condotte dagli scriventi tra il 2017 e il 2019. La ricerca è in fieri e ulteriori dati provenienti da nuovi contesti e stratigrafie studiate sono in corso di elaborazione e saranno oggetto di una futura pubblicazione.

219

Andrea Cipolato, Giuseppe Indino antichità (Quaderni della Soprintendenza Archeologica del Piemonte. Monografie 6): 516‒532. Alba (IT). Bueno, M., M. Novello and V. Mantovani 2012. Progetto Aquileia: casa delle Bestie Ferite. Commercio e consumo ad Aquileia. Analisi delle anfore tardoantiche alla luce di alcuni contesti. Rei Cretariae Romanae Fautorum Acta 42: 159‒168. Calaon, D. and A. Cipolato 2018. La laguna nord di Venezia in età romana e tardoantica, in M. Bressan, D. Calaon and D. Cottica (eds) Vivere d’acqua. Archeologie tra Lio Piccolo e Altino: 27‒40. Treviso: Antiga. Carandini, A., M.T. D’Alessio and H. Di Giuseppe 2006. La fattoria e la villa dell’Auditorium nel quartiere Flaminio di Roma. Roma: L’Erma di Bretschneider. Carru, D., M. Gonzales and G. Gurbiel 1997. Mazan. Le Jonquier-Plein-Panier. Bilan Scientifique Régional de Provence-Côte d’Azur: 140‒143. Cibecchini, F. 2017. Les routes de commercialisation du vin de Narbonnaise: l’apport des épaves profondes au large de la Corse. Gallia 74-2: 119‒130. Cipolato, A. 2018a. Anfore italiche, egeo-orientali e galliche, in L. Sperti, M. Tirelli and S. Cipriano (eds) Prima dello scavo. Il survey 2012 ad Altino: 142‒153. Venezia: Edizioni Ca’ Foscari. Cipolato, A. 2018b. Anfore in contesto. Contenitori da trasporto da scavi stratigrafici ed interventi di recupero nella Laguna Nord di Venezia: nuove prospettive di ricerca. Scuola Interateneo di Specializzazione in Beni Archeologici (Università di Venezia, Udine, Trieste), in corso di stampa. Cipriano, S. and F. Ferrarini 2001. Le anfore romane di Opitergium. Oderzo: Lions club di Oderzo, Veneto Banca. Cipriano, S. and S. Mazzocchin 2011. Bonifiche con anfore a Padova: note di aggiornamento alla cronologia e alla distribuzione topografica. Antenor Quaderni 20: 331‒367. Cipriano, S. and S. Mazzocchin 2018. Sulla cronologia delle anfore Dressel 6A: novità dai contesti di bonifica della Venetia. Rei Cretariae Romanae Fautorum Acta 45: 261‒271. Desbat, A. and S. Martin Kilcher 1989. Les amphores sur l’axe Rhône-Rhin à l’époque d’Auguste, in Amphores romaines et histoire économique. Dix ans de recherche. Actes du colloque de Sienne (22-24 mai 1986)(Collection de l’École française de Rome 114): 339‒365. Roma Paris: Ecole française de Rome, Diffusion de Boccard. De Min, M. 1986. L’antico Polesine: testimonianze archeologiche e paleoambientali. Catalogo delle esposizioni di Adria e di Rovigo. Adria: Museo Nazionale Archeologico. De Zuccato, G. 2016. L’insediamento romano di Chiunsano. Gli scavi dell’Università di Bochum (1992-2000). Firenze: All’Insegna del Giglio. Fozzati, L. and A. Toniolo 1998. Argini–strade nella laguna di Venezia, in S. Pesavento Mattioli (ed.) Bonifiche e drenaggi con anfore in epoca romana: aspetti tecnici e topografici. Atti del seminario di studi (Padova 19-20 ottobre 1995): 197‒208. Modena: F. C. Panini.

Bibliografia Alberti, A., M. L. Gualandi, F. Logiudice, O. Raffo, C. Rizzitelli, P. Siclari and G. Sorrentino 2015. Pisa, Piazza dei Miracoli. Materiali dagli scavi del 1998, in ΑΓΩΓΗ. Atti della Scuola di Specializzazione in Beni Archeologici (2011-2012) VIII-IX: 57‒97. Arnaud, P. (ed.) 2008. Via Iulia Augusta : un itinéraire romain exceptionnel emblématique de notre historie et de notre identité. Menton: Communauté de la Riviera française. Auriemma, R., V. Degrassi, D. Gaddi and P. Maggi 2016. Canale anfora: uno spaccato sulle importazioni di alimenti ad Aquileia tra I e III secolo d.C. Antichità Altoadriatiche 84: 379‒403. Auriemma, R., V. Degrassi and E. Quiri 2012. Produzione e circolazione di anfore in Adriatico tra III e IV secolo: dati da contesti emblematici, in S. Fioriello (ed.) Ceramica  romana  nella Puglia  adriatica. Indagini archeologiche a Egnazia. Dallo scavo alla valorizzazione: 255‒298. Bari: Sedit. Begley, V. 1996. The ancient port of Arikamedu. New excavations and researches 1989-1992. Ecole française d’Extrême Orient. Pondichéry: Centre d’histoire et d’archéologie, École française d’Extrême-Orient. Bekić, L. and M. Pešić 2014. Amfore i njhovi čepovi / Die Amphoren aus Veštar und ihre Pfropfen , in L. Bekić (ed.) Luka Veštar / Der Hafen von Veštar: 121‒136. Zadar: Međunarodni centar za podvodnu arheologiju. Beltrame, C. and A. Cipolato 2019. Un relitto romano con carico di marmo a Punta del Francese (StintinoSassari). L’Archeologo Subacqueo 68: 25‒32. Bertucchi, G. 1992. Les amphores et le vin de Marseille. Revue Archeologie de Narbonnaise (Supplement 25). Paris: Editions du CNRS. Bezeczky, T. 1998. Amphora types of Magdalensberg. Arheološki vestnik 49: 225–242. Bolzoni, G. 2012. Produzioni e commerci di ceramiche fini, comuni e da cucina di età romana (III sec. a. C.-III sec. d.C.) a Durazzo: lo stato degli studi, in S. Fiorello (ed.) Ceramica romana nella Puglia adriatica. Indagini archeologiche a Egnazia. Dallo scavo alla valorizzazione: 301‒312. Bari: Sedit. Brecciaroli Taborelli, L. 2000. La necropoli: aspetti rituali, sociali, economici, in L. Brecciaroli Taborelli (ed.) Alle origini di Biella. La necropoli romana: 27‒70. Biella: Umberto Allemandi & C. Brecciaroli Taborelli, L. and A. Gabucci 2007. Le mura e il teatro di Augusta Taurinorum: sequenzestratigrafiche e dati cronologici. Materiali per la datazione, in L. Brecciaroli Taborelli, L. (ed.) Forme e tempi dell’urbanizzazione nella Cisalpina (II s. a.C. – I s. p. C.): 243‒259. Borgo S. Lorenzo: All’Insegna del Giglio. Bruno, B. 1997. Contenitori da trasporto: i consumi di olio, vino e di altre derrate, in F. Filippi (ed.) Alba Pompeia. Archeologia della città dalla fondazione alla tarda

220

Le anfore galliche nella Regio X: dati preliminari

Gabucci, A. 2007. Attraverso le Alpi e lungo il Po: importazione e distribuzione di sigillate galliche nella Cisalpina (Collection de l’École Française de Rome 532). Roma: École Française de Rome. Gabucci, A. and E. Quiri 2008. Eporedia: appunti su terre sigillate e anfore tra tarda repubblica ed età imperiale. Quaderni Archeologici del Piemonte 23: 45‒78. Galasso, M. 1997. Rinvenimenti archeologici subacquei in Sardegna, in Atti del Convegno nazionale di archeologia subacquea : Anzio, 30-31 maggio e 1. giugno 1996: 121‒134. Bari: Edipuglia. Gamberini, A. 2014. Anfore, in L. Mazzeo Saracino (ed.) Scavi di Suasa I. I reperti ceramici e vitrei dalla Domus dei Coiedii: 533‒586. Bologna: Ante Quem. Gluščević, S. 2007. Antička luka u Zaton / Ancient harbour in Zaton. Zadar: Arheološki muzej Zadar. Hesnard, H. 1989. Un dépôt augustéen d’amphores à La Longarina, Ostie, in J.H. D’Arms and E.C. Kopfféds (eds) The Seaborne Commerce of Ancient Rome (Studies in Archaeology and History, Memoirs of the American Academy in Rome 36): 141‒156. Roma: American Academy in Rome. Laubenheimer, F. 1985. La production des amphores en Gaule Narbonnaise sous le haut-Empire (Annales Littéraires de l’Université de Besançon 327). Besançon: Université de Franche-Comté. Laubenheimer, F. 2001. 20 ans de recherches à Sallèles d’Aude. Besançon: Presses Universitaires Franc-Comtoises. Laubenheimer, F. and B. Watier 1991. Les amphores des allées de Tourny à Bordeaux. Aquitania 9: 5‒39. Leoncini, E. 2007. Cantiere delle navi antiche di Pisa: materiali ceramici dal carico della Nave A (US1010). Gradus 2.1: 6‒15. Marlière, E. 2001. Le tonneau en Gaule romaine. Gallia 58: 181‒200. Meffre, J.C. 1992. Ste-Cécile-les-Vignes (Vaucluse), in F. Laubenheimer (ed.) Les amphores en Gaule, production et circulation. Table ronde internationale (Metz 4-6 octobre 1990) (Annales Littéraires de l’Université de Besançon 474): 25‒35. Paris: Les Belles Lettres. Milanese, M. 1993. Genova Romana: Mercato E Citta Dalla Tarda Eta Repubblicana a Diocleziano Dagli Scavi Del Colle Di Castello (Genova - S.silvestro 2) (Studia Archaeologica). Roma: L’Erma di Bretschneider. Panella, C. 1989. Le anfore italiche del II secolo d.C., in in Amphores romaines et histoire économique. Dix ans de recherche. Actes du colloque de Sienne (22-24 mai 1986)(Collection de l’École française de Rome 114): 139‒178. Roma - Paris: Ecole française de Rome, Diffusion de Boccard.

Panella, C. 1992. Mercato di Roma e anfore galliche nella prima età imperiale, in F. Laubenheimer (ed.) Les amphores en Gaule, production et circulation. Table ronde internationale (Metz 4-6 octobre 1990) (Annales Littéraires de l’Université de Besançon 474): 185‒206. Paris: Les Belles Lettres. Pascale, S. 1998. Rinvenimenti anforacei nel Trevigiano: tipologie e commerci. Archeologia e territorio 17: 59‒77. Pecci, A., G. Giorgi, L. Salvini and M.A. Cau Ontiveros 2013. Identifying wine markers in ceramics and plasters using gas chromatograpgy-mass spectrometry. Experimental and archaeological materials. Journal of Archaeological Science 40: 109‒115. Quiri, E. 2011. Le anfore, in L. Brecciaroli Taborelli (ed.) Oro, pane e scrittura. Memorie di una comunità inter Vercellas et Eporediam: 107‒116. Roma: Quasar. Quiri, E. 2014. Le anfore, in M.C. Preacco (ed.) Augusta Bagiennorum. Storia e archeologia di una città augustea: 201‒211. Torino: Celid. Ricci, A. 1985. Settefinestre.Una villa schiavistica nell’Etruria romana. La villa e i suoi reperti 2. Modena: Panini. Schimmer, F. 2015. Flussi commerciali in Raetia. Le anfore di Cambodunum (Kempten, Baviera). Quaderni friulani di archeologia 25: 173‒178. Tchernia, A. 1986. Le vin de l’Italie romaine. Essai d’histoire économique d’après les amphores (Bibliothèque des Écoles françaises d’Athènes et de Rome 261). Roma: Écoles françaises de Rome. Tito, V. 2013. Il Relitto di Cala Cupa, Isola del Giglio (GR): revisione critica e confronti nell’ambito del commercio del vino gallico. Notiziario della Soprintendenza per i Beni Archeologici della Toscana 8/2012: 166‒172. Toniolo, A. 1991. Le anfore di Altino. Archeologia Veneta 14‒36. Zanda, E. 2011. Industria. Città romana sacra a Iside. Torino: Umberto Allemandi & C. Žerjal, T. 2008. Trade and consumption of foodstuffs in Northern Adriatic: the case of the villa Školarice, in R. Auriemma and S. Karinja (eds) Terre di mare. L’archeologia dei paesaggi costieri e le variazioni climatiche. Atti del Convegno Internazionale di Studi (Trieste 8-10 novembre 2007): 462‒469. Trieste - Piran: Università degli studi di Trieste, Pomorski muzej Museo del mare “Sergej Mašera”.

221

Terra sigillata norditalica decorata a matrice intenzionalmente nera: alcune considerazioni Valentina Mantovani, Eleni Schindler Kaudelka, Roby Stuani, Émilie Mannocci, Carla Corti Abstract Progress in our understanding of the process of manufacture in Aco beakers and Sarius cups is mainly focused on decoration patterns, contextual studies and chronology. The paper attempts a first case study of north Italic moulded pottery that shows a black coating on commonly red artifacts, and tries to distinguish intentional from accidental colouring, as it is present in a large number of fine wares that belong to different Roman pottery classes. Previously all black-coloured fragments pertinent to “normally red” tableware were considered ruined by fire. The question is still open since some of the black wares do not appear damaged but look deliberately fired to obtain a dark surface.

Key words Nord-italica decorata, Terra Sigillata, rouge et noir.

Introduzione La nostra conoscenza della ceramica nord-italica decorata a matrice negli ultimi decenni non ha fatto grandi passi in avanti. Ciò nonostante, il confronto fra i punzoni che caratterizzano i frammenti di questo vasellame apre sempre nuove prospettive di ricerca, che permettono un lento ma progressivo avanzamento nello studio della classe. In occasione del III Colloquio internazionale sulla ceramica ellenistica tenutosi a Kaštela nel giugno 2017 è stato presentato un aggiornamento sulle origini e i modelli d’ispirazione della prima produzione di terra sigillata nord-italica decorata a matrice. In quell’occasione fu rimarcata l’impossibilità di giungere, al momento, ad una chiara definizione della genesi di questa classe, sottolineando come, nonostante gli aggiornamenti ed i nuovi rinvenimenti, numerose siano le problematiche che rimangono tuttora inevitabilmente aperte (Schindler Kaudelka et al. 2020). Durante il continuo lavoro di disamina dei frammenti editi è stato identificato un interessante argomento di ricerca che ha come focus l’indagine delle cause, intenzionale o accidentali, che hanno portato alla realizzazione di vasellame in terra sigillata norditalica decorata a matrice con vernice di colore nero/ brunastro. Il rinvenimento di questi vasi “anomali”, caratterizzati a volte da impasti grigi, merita a nostro avviso un approfondimento, tramite il quale cercheremo di inquadrare le peculiarità ed i limiti di questo fenomeno. Alterazioni complete o parziali dei rivestimenti sono documentati anche in altre produzioni di terra sigillata, ma vista la complessità dell’argomento, e la difficoltà

di tracciare i caratteri comuni di questo fenomeno, abbiamo scelto di iniziare la disamina partendo dall’analisi della produzione nord-italica decorata a matrice. Le caratteristiche formali e tecnologiche di questa classe permettono, infatti, una facile identificazione ed isolamento dei pezzi e, sebbene il fenomeno sia ancora lungi dall’essere definito su vasta scala, tramite la revisione di questi manufatti si cercherà di evidenziarne le peculiarità e i limiti. Al contempo non possiamo prescindere da un costante confronto con le altre produzioni di sigillata nord-italica che presentano analoghe caratteristiche tecnologiche, funzionali a definire il contesto socio-economico in cui si colloca il consumo di questi particolari prodotti. Terra sigillata nord-italica con rivestimento nero: stato della questione e problemi di definizione Vari sono i manufatti in terra sigillata che presentano superfici nere o brunastre. Un caso abbastanza emblematico, e utile all’introduzione del presente lavoro, è rappresentato da due coppette in terra sigillata liscia rinvenute nel sito del Magdalensberg, in un primo momento edite come ceramica a vernice nera1 e successivamente attribuite alla produzione di terra sigillata, definendo il colore della vernice come esito di combustione (Schindler 1986: 368). Una terza coppetta dell’emporio norico con bollo ARRET pubblicata inizialmente come vernice nera (Schindler 1967: tav. 6 n. 20) è stata edita una seconda volta come terra sigillata bruciata (Schindler 1986) (Figura 1). ‟Obwohl Ton und Überzug kein Anzeichen zeigen, dass die Tasse Taf. 6,22 ursprünglich nicht schwarz war, besteht kein Zusammenhang mit schwarzer Glanz-Ware” (Schindler 1967: 63, Tav. 6 n. 22). 1 

Roman Pottery and glass manufactures (Archaeopress 2022): 222–238

Terra sigillata norditalica decorata a matrice intenzionalmente nera: alcune considerazioni

Figura 1. Coppette in terra sigillata liscia del Magdalensberg (foto: E. Schindler Kaudelka).

Va ricordato, per dare un’idea della complessità della questione, che il bollo ARRE/TINVM attestato anche a Bologna (scavo di via Rizzoli) su due frammenti di coppe con rivestimento rispettivamente nero e rosso. Dal medesimo contesto provengono inoltre altre sigillate con rivestimento nero uniforme, oltre ad alcuni frammenti decorati a matrice. Sono qui documentati frammenti di coppe Consp. 22, Consp. 26 e Consp. 38, prodotta anche con corpo ceramico grigio vernice nera, oltre a piatti Consp. 18 (Corti 2016: 90).

non hanno dato sin ora un’interpretazione univoca a questo fenomeno. C’è chi ritiene questi manufatti frutto di un’azione intenzionale e chi, al contrario, li considera effetti di cause secondarie, come un’errata esposizione al fuoco primario (ovvero pezzi di seconde scelte o scarti di produzione dovuti ad uno scarso controllo della camera di cottura) o come alterazioni causate da incendi o roghi (Zabehlicky-Scheffenegger 1997). Partendo da questi presupposti, e limitando per ora l’analisi ai soli esemplari nord-italici decorati a matrice, com’è possibile definire la natura dei vasi che presentano queste caratteristiche? Per provare a rendere la questione più semplice abbiamo isolato in tre principali gruppi alcuni materiali noti in bibliografia che presentano tra loro caratteristiche omogenee.

La questione viene ulteriormente complicata dall’impossibilità di stabilire, tramite analisi archeometriche, i motivi di questa anomala colorazione della vernice poiché l’archeometria non permette di chiarire se l’esposizione al fuoco o ad una cottura riducente sia frutto di un’azione primaria o secondaria. Più facilmente identificabili a livello archeometrico sono, invece, le alterazioni post-deposizionali, risultato della giacitura in ambienti particolarmente acidi, i quali restituiscono ceramiche con evidenti variazioni del colore. In questi specifici casi, recenti studi hanno dimostrato come, il terreno di giacitura di un manufatto possa mutare, attraverso alterazioni chimiche e fisiche, la composizione dei rivestimenti, facendo loro assumere colorazioni che variano dal grigio al rosa, fino al bianco, come per il caso delle terre sigillate di via Retratto, ad Adria (Mantovani 2015: 106-111).

Terra sigillata nord-italica decorata a matrice con rivestimento nero uniforme Al primo gruppo possono afferire esemplari in sigillata nord-italica decorata a matrice caratterizzati da un rivestimento di colore nero uniforme, più o meno brillante, e da un impasto che varia dal colore bruno al grigio, come i seguenti frammenti di bicchiere tipo Aco: • quattro esemplari dal Magdalensberg (Schindler Kaudelka 1980: tav. 13 n. 1, tav. 17 n. 32, tav. 18 n. 40) di cui uno inedito (Inv. M82 SH/9A) (Figura 2); • un esemplare da Verona, in via Redentore 9 (Stuani 2017: p. 112, Tav. 21.13); • un esemplare firmato TYMS, dall’isola di Krk (Figura 3);2

Terra sigillata nord-italica decorata a matrice di colore nero (scuro) La prima domanda che ci siamo posti è: esiste realmente una produzione di terra sigillata nord-italica a matrice intenzionalmente nera? Certa è l’esistenza di manufatti con un rivestimento nero e impasti grigi o brunastri, ma i vari studiosi che si sono occupati dell’argomento

Inedito, comunicazione di Ana Konestra, che ringraziamo per la segnalazione. 2 

223

Valentina Mantovani, Eleni Schindler Kaudelka, Roby Stuani, Émilie Mannocci, Carla Corti

Figura 2. Quattro esemplari di bicchieri tipo Aco del Magdalensberg con rivestimento nero (foto: V. Mantovani).

• due esemplari rinvenuti nella stipe del Montirone, ad Abano Terme (Lavizzari Pedrazzini 1995: 146, fig. 8 nn. 5–8) (Figura 4);3 Il fenomeno è documentato anche sulle coppe tipo Sarius e occasionalmente anche su altre forme di coppe decorate a matrice: • due esemplari rinvenute ad Aquileia (UD), una delle quali con impasto grigio (Figura 5); 4 • un frammento di parete di piccola coppa Sarius da Adria (RO), con impasto grigio (Figura 6); 5 • un frammento di orlo di coppa Consp. R 4 e per le coppe tipo Sarius, un frammento di orlo da Bologna, via Rizzoli (Fava 1962, p. 57, n. 380; Corti 2016, p. 90) (Figura 7); • 10 frammenti (riconducibili ad almeno quattro esemplari) dal sito del Fréjus, rinvenuti in un contesto chiuso, datato tra il 30 ed 25 a.C. (Rivet 2019) (Figura 8).6

Figura 3. Frammento di bicchiere tipo Aco firmato TYMS, da Krk (foto: A. Konestra, su cortese concessione di PPMPH Rijeka).

Tali manufatti sono stati, a nostro avviso, prodotti volutamente con queste caratteristiche e, così, l’identificazione dei pezzi sembrerebbe semplice. Tuttavia, l’esiguità dei frammenti non permette mai di giungere ad una definizione certa del manufatto, come ci dimostrano un esemplare veronese rinvenuto in via Redentore 9 (Stuani 2017: p. 113, Tav 22.6) in cui frammenti combacianti presentano colorazioni estremamente diverse, esito di differenti luoghi di giacitura: il piccolo frammento di color nero, rinvenuto in una strato diverso da quello che conteneva i due frammenti di color rosso, è stato certamente esposto ad un fuoco secondario. Come avremmo interpretato Il materiale edito viene descritto con argilla rosata, ma l’analisi autoptica dei frammenti, esposti presso il museo di Este, conferma la presenza di un rivestimento coprente, uniforme e brillante di colore nero sull’esemplare più completo, mentre il secondo, di cui si conserva solo una piccola porzione di parete, presenta un rivestimento quasi completamene abraso (entrambi i pezzi sono identificati da I.G. 6352). Ringraziamo qui la Dott. ssa Gonzato per averci permesso di visionare il materiale. 4  Visione autoptica: un esemplare rinvenuto presso il sito della Casa delle Bestie ferite (I.G. 142472) ed un secondo presso i magazzini del Museo Archeologico Nazionale di Aquileia. 5  I.G. 16780. 6  Comunicazione personale di Lucien Rivet e Sylvie Saulnier. 3 

Figura 4. Due esemplari di bicchiere tipo Aco della stipe del Montirone-Abano Terme (PD) (foto: V. Mantovani).

224

Terra sigillata norditalica decorata a matrice intenzionalmente nera: alcune considerazioni

Figura 5. Frammento di coppa tipo Sarius con rivestimento nero molto abraso rinvenuta nel sito della Domus delle bestie Ferite, ad Aquileia (UD) (foto: V. Mantovani).

Figura 6. Frammento di coppa tipo Sarius rinvenuta in via Retratto, ad Adria (RO) (foto: V. Mantovani).

Figura 7. Frammenti di orli e pareti di coppe tipo Sarius e calice Consp. R 4 da via Rizzoli, Bologna (foto: C. Corti).

Figura 8. A sinistra, un frammento di coppa decorata a matrice con rivestimento nero dal sito del Fréjus (Rivet 2019), mentre a destra un frammento con rivestimento rosso e analoga scena figurata dal Magdalensberg (foto: E. Schindler Kaudelka).

225

Valentina Mantovani, Eleni Schindler Kaudelka, Roby Stuani, Émilie Mannocci, Carla Corti

Figura 9. Due frammenti combacianti pertinenti allo stesso esemplare di coppa tipo Sarius, da Verona (foto: R. Stuani).

Figura 10. Coppa tipo Sarius di Spinimbecco (foto: V. Mantovani).

il singolo frammento di colore nero, privo del suo corrispettivo rosso (Figura 9)? Terra sigillata liscia e decorata a matrice con rivestimento nero non uniforme Al secondo gruppo afferiscono gli esemplari in sigillata liscia o decorata a matrice in cui il rivestimento non è di colore omogeneo (che può variare dal nero, al bruno, al rossiccio, a volte verso il porpora), e per i quali non è facile comprendere se siano il risultato di cotture poco controllate, di esposizioni ad un fuoco secondario, o di alterazioni dovute ai terreni di giacitura:7 Figura 11. Coppa con scena di commedia del Magdalensberg (foto: E. Schindler Kaudelka).

• Portiamo ad esempio la coppa tipo Sarius di Spinimbecco (Stenico 1973: 118, figg. 8 a–b), proveniente da un sepolcreto ad incinerazione rivenuto tra gli anni 1868 e 1873, che presenta un rivestimento brunastro estremamente eterogeneo, e un impasto di colore beige aranciato. In questo caso, la colorazione esterna potrebbe essere l’esito di alterazioni postdeposizionali, di esposizione al fuoco secondario (quello della pira funebre) o di una cottura errata. Comunque, anche in quest’ultimo caso ci risulta impossibile stabilire se l’intenzione del vasaio fosse quella di realizzare una coppa di colore rosso o nero (Figura 10). • Sempre di produzione nord-italica è l’esemplare di coppa con scena di commedia del Magdalensberg, il cui rivestimento, virante dal rosso al brunastro/grigio, non ci permette di

Figura 12. Coppa decorata a matrice conservata al Metropolitan Museum di New York, firmata SEX AVLIENI. Fonte del 30/08/2019

Si segnala la presenza di rivestimenti di colore non uniforme anche tra il materiale di Bologna come sottolineato già in Mazzeo Saracino 1983: 476; questo materiale necessita tuttavia di una complessiva revisione, anche per comprendere meglio la natura di queste alterazioni (che si isolano nettamente dal materiale a rivestimento nero uniforme), con un’attenta analisi estesa all’intero contesto ceramico, non solo alle sigillate, che sarà oggetto di studio specifico a breve. 7 

226

Terra sigillata norditalica decorata a matrice intenzionalmente nera: alcune considerazioni

capire se possa trattarsi di un esemplare difettato o esposto ad un fuoco secondario (Figura 11).8 • Un altro esempio è il frammento di coppa decorata a matrice conservata al Metropolitan Museum di New York, firmata SEX AVLIENI, di produzione puteolana o flegrea, in cui la copertura brunastra è stata interpretata dagli studiosi come un difetto di cottura (Soricelli 2017: 101, fig. 24) (Figura 12). • Sono attestati anche nella produzione dei Plotidii di Scoppieto materiali con rivestimento non uniforme, che vira dal rosso, al bruno-nerastro (Bergamini et al. 2017: tav. 3 n. 6). Terra sigillata bruciata Fenomeni analoghi sono presenti anche sulle numerose coppe in terra sigillata liscia, di produzione aretina e orientale B, rinvenute nell’Agora Tetragonos di Efeso e firmate da C. Sentius, caratterizzate da una colorazione che vira tra il rosso corallo ed il bruno-nerastro, e che gli studiosi ritengono essere tutte esito di esposizione a fuoco secondario (Zabehlicky-Scheffenegger, 1996: 253271; 1997).

Figura 13. Grande calice con scena di caccia del Magdalensberg, firmato da Clemens (foto: V. Mantovani).

Per sottolineare questo fenomeno torniamo ancora alla disamina dei frammenti di terra sigillata liscia. Ne è un chiaro esempio il caso della firma C. ACO del Magdalensberg apposta su un fondo pertinente, con ogni probabilità, ad una coppa Consp. 22 (produzione Padana C) con rivestimento corallino (ZabehlickyScheffenegger e Sauer 2000: 71, fig. 3), marchio attestato anche su un piatto rosso Consp. 20 (produzione Padana C) da Massongex (Paccolat e Wiblé 2007: fig. 4 n. 8). Il medesimo bollo, ma con colorazione che vira dal rosso al bruno-nerastro, è presente su tre esemplari di piatti Consp. 18.2 (produzione Padana C) facenti parte della sepoltura n. 40 di Porta Palio, a Verona (Cavalieri Manasse e Bolla 1998: 103-11). In quest’ultimo caso, essendo la sepoltura un’incinerazione, l’alterazione cromatica può ritenersi successiva alla cottura, e imputabile all’esposizione del fuoco della pira funebre. Esistono, tuttavia, dati che ci permettono di approfondire ulteriormente la tematica e di delineare la natura di questo fenomeno.

Con una discreta casistica nella terra sigillata, così come in altri classi di materiali, il materiale può giungere a noi effettivamente bruciato ed avere un aspetto bruno, nero, in alcuni casi semivetrificato e il corpo parzialmente collassato. In particolare, per questa tipologia di reperti si è già sottolineato come, al di là della problematica oggetto del presente lavoro, tali evidenze rappresentino un altro soggetto di studio estremamente delicato nel quale non solo confluisce la comprensione dell’oggetto in se, falsata da colorazioni “non consuete”, ma anche la sua provenienza ed il luogo di produzione (Schindler-Kaudelka, ZabehlickyScheffenegger 2000: 571-574). Il fenomeno è attestato anche in terra sigillata decorata a matrice come si può notare sul grande calice con scena di caccia del Magdalensberg, firmato da Clemens (Schindler Kaudelka 1998, 134) (Figura 13).

Sono noti in bibliografia esemplari integri o frammentari di terre sigillate padane lisce con rivestimento nero e impasto brunastro o, a volte, grigio. Tra le forme meglio documentate vi è la coppa Consp. 22, attestata al Magdalensberg da tre esemplari in Padana C sui quali compare un bollo mal interpretabile (Schindler 1967: Taf. 6, 22, Foto Taf. 4b) e a Zevio, nel veronese (Rodegher 2012–2013: t. n. 109, n. 3, tav. VIII n.3), oltre ad almeno un esemplare di Consp. 21 sempre in Padana C da Legnano (MI) (Gorla e Polidoro in press). In via Rizzoli, a Bologna, sono documentati diversi esemplari di coppe Consp. 22 e Consp. 26, di fondi bollati CHRY/C.VOLV (OCK 2505), ARRE/TINVM (OCK 242-244), AMICVS (OCK 86) e SARI (OCK 1782), di colore nero e rosso (Corti 2016); tutti questi bolli sono già attestati in numerosi altri siti in

Intenzionalità sì, intenzionalità no: ulteriori dati Dopo questa breve disamina, nella quale sono state indagate le argomentazioni relative ad alcuni casi esemplari, cercheremo ora di analizzare quali elementi archeologici siano a supporto della produzione intenzionale di vasellame in terra sigillata nord-italica a matrice con rivestimento nero. L’argomento assai complesso, e di difficile definizione, corre sul sottile limite che divide la certezza dalla mera ipotesi.

Si rimanda, da ultimo, a Schindler Kaudelka et al. 2017, con precedente bibliografia. 8 

227

Valentina Mantovani, Eleni Schindler Kaudelka, Roby Stuani, Émilie Mannocci, Carla Corti

Figura 14. A sinistra, il fondo di coppetta con rivestimento nero bollata da Passius Telamo proveniente da Adria e, a destra, la versione con rivestimento rosso del Magdalensberg (foto: V. Mantovani e S. Zabehlicky-Scheffenegger).

variante rossa.9 Le coppe Consp. 22, così come i piatti Consp. 21, appaiono in età augustea nell’orizzonte Oberaden, mentre più recenti sono le coppe Consp. 26, prodotte tra l’età tiberiana e quella neroniana (e forse oltre).

a vernice nera e la terra sigillata (Schindler 1986: 361363; Conspectus 1990: 52, 68, 114). A proposito dell’Italia settentrionale, Jean Paul Morel d’altronde osservava come la “tradition ethnique” e il gusto, che coinvolge direttamente la domanda da parte dei consumatori, sia un fattore sottostimato in materia di produzione ceramica (Morel 1987: 130). Un’osservazione legata a quelli che definisce fenomeni di “attardamento” che caratterizzano l’intera regione con la comparsa di marchi in planta pedis su ceramiche a vernice nera prodotte ancora in età tiberiana e con la presenza di forme di transizione alla terra sigillata caratterizzate da un rivestimento “rossastro” (Morel 1987: 125-126).

Sempre in via Rizzoli (Bologna) sono documentate le firme MAE/PATES (OCK 1083),10 insieme alla variante MAE/PATIS, e LAVRI (OCK 1023) su coppe Consp. 38, anche con rivestimento nero e argilla grigia, i cui punzoni sono del tutto uguali a quelli delle coppe in sigillata nord-italica datate tra l’età augustea e la primissima etá tiberiana (Corti 2015; 2016). Analoga situazione si riscontra per la firma SATVR, anch’essa presente su coppe Consp. 38 prodotte sia con vernice rossa che, come per il marchio MAE/PATES, anche in ceramica con rivestimento nero e pasta grigia.11

Si inserisce quindi appieno in questo conteso di coesistenza di produzioni ceramiche diverse, come espressione di un gusto locale fortemente sentito e legato alla tradizione ma al contempo pienamente inserito in un contesto regionale aperto e molto ricettivo; esemplificativo a tal proposito è il caso delle coppe Consp. 38 realizzate in nord-Italica con corpi ceramici sia grigi che rosati, rivestiti con vernici rosse e nere (terra sigillata, terra sigillata a rivestimento nero, terra sigillata con impasto grigio)(Corti 2016: 91-93). Una coesistenza di forme, che gli archeologi tendono a classificare in produzioni ceramiche diverse (ma che in antico nascevano slegate da queste costruzioni mentali), inoltre evidenziata dalla comparsa su coppe prodotte con rivestimenti rossi e neri degli stessi marchi (SATVR e MAE/PATES), almeno in un caso ottenuti dallo stesso punzone (MAE/PATES).

Per questa forma, così come per le patere Consp. 1/ Morel 2248/Lamboglia 7, il piatto Consp. 10/ Morel 1420–1469 e la coppa Consp. 36/ Morel 2983-2987 non é mai stata contestata la coesistenza di materiali intenzionalmente neri e rossi (Schindler 1967: 32, fig. 3, 28; Schindler e Scheffenegger 1977: 25;12 Conspectus 1990: 118; Morel 1998: 251)13 e, anzi, tale vasellame è sempre stato considerato come caso esemplificativo del fenomeno di coesistenza e di transizione tra la ceramica Per la distribuzione si rimanda alla relativa scheda in OCK di ciascun bollo qui enunciato. 10  Analoghi esemplari sono documentati anche nell’insediamento urbanorustico di Domagnano (Repubblica di San Marino) (Bottazzi e Bigi 2001). 11  Per le attestazioni in terra sigillata vedi Jorio 1991: tav. 20,1 e Gabucci 1995: tav. 7, 44–45; per l’esemplare in ceramica a pasta grigia e rivestimento nero si rimanda a Giordani 1990: fig. 3, fig. 8, 5; vedi Corti 2016: 91, fig. 4. 12  “Allen diesen Formen 1/1 bis 1/7 lagen also schwarze Vorbilder oder Ausgangsformen zugrunde. Dass sie zu der Übergangsware gehören, zeigen außerdem die Komplexsignaturen”. 13  Morel riferendosi al rapporto tra la vernice nera e la vernice rossa ricorda come già gli Etruschi ”…hanno mostrato una certa indifferenza, o un certo eclettismo, per quanto riguarda il colore dei vasi, che li ha spinti a produrre questa o quella forma sia con vernice nera sia con vernice rossa...”. 9 

Potrebbe trattarsi, in questo specifico caso, di un’ulteriore testimonianza della grande variabilità all’interno della produzione della sigillata nord-italica, prodotte con rivestimenti rossi e neri fenomeno questo osservato non solo nella ESB, ma anche nella produzione della ceramica Campana C, già nota a Nino Lamboglia. Le motivazioni in merito alla presenza di manufatti 228

Terra sigillata norditalica decorata a matrice intenzionalmente nera: alcune considerazioni

con argille grigie, possono essere ricercate nell’ambito di precise scelte tecnologiche, nella quantità di tempo e di legname necessario per la cottura, e non sempre esclusivamente attribuiti a materiali combusti.

Un’evoluzione simile è documentata anche nella Eastern Sigillata B sia nel primissimo periodo della produzione, con le forme Atlante 1–3 (Lang-Auinger e Mitsopoulos-Leon 2007: n. H11, pl. 118), che in quello finale, soprattutto nelle forme 77 e 78. Non va dimenticata, inoltre, la fabbricazione in questi ateliers di manufatti realizzati esclusivamente con rivestimento nero, come le brocchette con una o due anse.14

Altri dati per approfondire la comprensione dell’esistenza di manufatti con argille grigie provengono dai materiali firmati da Passius Telamo, figulo padano che bolla sia esemplari neri, come nel caso del fondo di Adria (RO), località dalla quale provengono altre due coppe con analoga finitura ma prive di bollo, (Mantovani 2015: tab. 5 n. 94), sia esemplari rossi, come al Magdalensberg (Schindler e Scheffenegger 1977: tav. 109) e in numerosi altri siti della Pianura Padana (OCK 1380) (Figura 14).

Forse ispirati dalla stessa produzione locale di ESB sono, inoltre, i Graue Platten di Efeso, i grandi vassoi da portata con rivestimento nero e impasto grigiobrunastro esportati in tutto l’impero (ZabehlickyScheffenegger et al. 1996; Mantovani 2015: 115), a cui sembrerebbero ispirarsi i manufatti della fornace della Celsa (Roma), che produceva, anche, piatti da portata circolari caratterizzati sia da argille bruno-grigiastre e rivestimenti nerastri che argille rossastre con vernice di colore rosso, spesso di ottima qualità (Carrara 2013).

Nei casi menzionati siamo sempre in presenza di vasi di buona fattura, con rivestimenti coprenti, uniformi, ed abbastanza brillanti. Sembra quindi naturale pensare si tratti di manufatti intenzionalmente neri, soprattutto sulla scorta della casistica, assai più nutrita rispetto a quella della terra sigillata padana decorata a matrice. E proprio questo gruppo di esemplari in Padana C nera, la cui tecnologia di produzione si diversifica dalla canonica sigillata con vernice sinterizzata (ZabehlickyScheffenegger e Sauer 2000: 71, fig. 3) può essere di aiuto nella definizione del fenomeno, che sembra seguire una specifica tendenza, e almeno per la produzione liscia padana, concentrarsi su una produzione tipica di età augustea, rappresentando un supporto cronologico e produttivo per la coeva produzione nord-italica a matrice con rivestimento nero.

La moda del rivestimento nero è documentata anche nella produzione di Céramique belge (Deru 1995) come quello della Sigillata Helvetica (Martin-Kilcher 1999: 33) e in tutte le manifatture regionali di terra sigillata che realizzavano ceramica a vernice non sinterizzata. La PGW pannonica nera, soprattutto nella sua prima fase, quella di I sec. d.C., era caratterizzata da un impasto grigio e un rivestimento nero, fenomeno riscontrato con minor frequenza anche negli esemplari più tardi, fino al III d.C. In particolare, per la Pannonische Glanztonware fabbricata nei centri del sud-est della Pannonia il bilinguismo dei colori sembra, secondo Tino Leleković, presente soprattutto nelle forme che mostrano una certa affinità con i repertori tipologici orientali e africani, e sembra perdere importanza nelle forme più tipiche del repertorio occidentale.15

Pertanto, se il bilinguismo cromatico del “rouge et noir” fu in uso, sebbene in maniera ridotta rispetto al volume dell’intera produzione padana, sulla sigillata liscia, in un momento storico in cui i vasa samia erano per definizione rossi, quali elementi ci impediscono di escludere che tale fenomeno non abbia interessato anche la decorata a matrice?

Comunicazione personale Kathleen Slane, che ringraziamo per la preziosa consulenza: ‟Deliberately firing black of sigillata vessels. Yes, of course. It is very well known in the East, in ESA, ESB and in Corinthian (and other) imitations as well. There is first of all the experimental black-firing stage at the very beginning of ESA and Italian sigillata production; for this see Tel Anafa II.1, pp. 273–274. Then there is deliberate black-firing later on (although in fact most examples are gray ware rather than black-glaze with a red or white biscuit). I think it is restricted in time (Augustan and Hadrianic) rather than continuous and have always assumed that it has nothing to do with nostalgia for black-glaze pottery but perhaps comes from a desire to imitate silver vessels? In ESA, the gray-ware is restricted to pitchers (and we tested it with NAA to prove it is the same clay as ESA). There is also Cypriot gray-ware, lagynoi rather than pitchers, which are found ca. 100 B.C. and could also be an experimental phase of Cypriot sigillata. In ESB there are two periods when gray-ware versions were produced alongside red ones: at the very beginning, Forms 1–3 (ex. is FiE IX, 2/2, 1 no. H11, pl. 118), and towards the end, Forms 77 and 78 are usually black and Form 80 is sometimes gray-ware. Forms 77 and 78 are relatively common east of the Theater and I’ve only ever seen one red one, all the rest are black slipped gray ware. At Corinth I have identical pitchers in the early 1st century, one of which is red slipped with an orange biscuit and the other black-slipped over buff. There are also some pieces that are not quite ESB but related that are gray-ware in the Augustan period. And, of course, there are the trays and platters from Ephesos as well as the Type XIX lamps. On the other hand, I can’t think of any examples in Pergamene sigillata but there might be some in Meyer-Schlichtmann.” 15  Comunicazione personale Tino Leleković a seguito di un contributo presentato ma non pubblicato al convegno RCRF di Lisbona, nel 2016. 14 

Rouge et noir. Fenomeno noto in tutte le classi di ceramica con rivestimento a “vernice” Esempi famosi di manufatti in sigillata con rivestimento nero sono comuni nell’intera produzione di sigillate del mondo romano. Abbastanza noto è il fenomeno in oriente, documentato non solo nella fase iniziale di sperimentazione della Eastern Sigillata A – come ad esempio a Tel Anafa e a Beirut (Warner Slane 1997: 273–274) – denominata “black-slip predecessor” e inquadrabile nel primo trentennio del II sec. a.C., ma anche nella fase più matura, in un periodo compreso tra l’epoca augustea e quella adrianea, così come suggeritoci da Kathleen Slane. 229

Valentina Mantovani, Eleni Schindler Kaudelka, Roby Stuani, Émilie Mannocci, Carla Corti

Figura 15. Bicchiere decorato a matrice fabbricato nell’atelier di Thonon-lesBains, nei pressi del lago di Lemano, in Francia (Da Berman 2015: 249, 251 fig. 9).

Per quanto riguarda la sigillata propriamente detta va citata la produzione decorata a matrice con rivestimento nero di Lezoux — come ad esempio il bicchiere Bet 84/Déchelette 64 = Drag. 30R — almeno per quanto riguarda la fase 4 di età Flavia, che ha prodotto Drag. 30 e 37, e la fase 5 databile alla prima metà II sec. d.C., nella quale vennero fabbricate anche Drag. 64.16 Tuttavia, in particolare per la produzione di Lezoux, gli studiosi propendono per inquadrare questo specifico fenomeno, all’interno della famiglia delle ceramiche metallescenti.17

a titolo esemplificativo ai piatti Hisp. 6a3, Drag. 15/17 e alle coppe Drag. 29 e 37 (Berman 2015: 249, 251, fig. 9).18 Gianluca Soricelli ci notifica, inoltre, dell’esistenza di materiali intenzionalmente neri anche all’interno della sigillata a rilievo tardo-puteolana (Soricelli 2001: 259-264). Spostandoci oltralpe, verso il versante occidentale, menzioniamo poi l’esemplare di coppa Consp. 14 di produzione lionese rinvenuta ad Avenches e firmata da Ateius (OCK 269), considerata intenzionalmente nera (Blanc e Meylan Krause 1997: fig. 10b, no cat. 1) (Figura 16), ed un secondo esemplare di Consp. 22 nera, sempre bollata ATEI, dal Pseudo-sanctuaire de Cybèle (Desbat 2008: 527-533) (Figura 16). Secondo Armand Desbat questi pezzi, rinvenuti al di fuori dei territori delle officine, sono da considerarsi come manufatti di seconda scelta ugualmente commercializzati.19 Si tratterebbe, in questo specifico caso, di un ulteriore gruppo rispetto a quelli sopra enunciati, ma per questi materiali, ancor più che per quelli analizzati fino ad ora, il discrimine tra bellezza, difetto, prima e seconda scelta, è ancora più labile.

Sempre in terra sigillata gallica va citato l’esemplare di bicchiere decorato a matrice fabbricato nell’atelier di Thonon-les-Bains, nei pressi del lago di Lemano, in Francia (Berman 2015: 249, 251 fig. 9) (Figura 15). Il fenomeno è attestato anche in area spagnola, non solo nella fase di transizione tra produzione a vernice nera e terra sigillata tipo Penaflor, ma anche in età imperiale. Come evidenziato dagli esemplari con rivestimento nero e pasta aranciata in sigillata Hispanica, si rimanda

 Ringraziamo Philip Bet per la consulenza, della quale citiamo in seguito lo scambio epistolare: “Lezoux a produit de la sigillée noire durant les phases 4 (=période flavienne, argile siliceuse) et 5 (=première moitié IIe s.). Il s’agit essentiellement de sigillée moulée : Drag 37 et Drag 30 pour la phase 4, Drag 37, Drag 30 et Drag 64 pour la phase 5. Pour la sigillée lisse, forme L008“. 17  Secondo Armand Desbat. 16

Ringraziamo Macarena Bustamante per la segnalazione. Armand Desbat, che ringraziamo, ci scrive: “[…] Pour ce qui est des sigillées noires il y en a quelques unes à Lyon, provenant de la Muette, mais je pense que ce sont des ratés vendus comme 2e choix. […]”. 18  19 

230

Terra sigillata norditalica decorata a matrice intenzionalmente nera: alcune considerazioni

Figura 16. Consp. 14 firmata da Ateius (OCK. 269) di produzione lionese rinvenuta ad Avenche (da Blanc and Meylan Krause 1997: fig.10b, no cat. 1). Consp. 22 firmata da Ateius (OCK 269) di produzione lionese rinvenuta a Lione Pseudo-Sanctuaire de Cybèle (da Desbat 2008: 527-533).

sembra si tratti più di una imitazione di vasellame metallico, meno di un retaggio della ceramica a vernice nera, così come suggerisce Kathleen Slane in un suo lungo resoconto epistolare.20

Note conclusive In conclusione, la moda del “rouge et noir” benché “timida” e poco frequente è trasversale alla produzione romana di terra sigillata e alle analoghe manifatture verniciate ad essa apparentate che vanno dalle coste dell’Asia minore, passando per la Pannonia, la Gallia e la Spagna, in un periodo compreso tra il II a.C. ed il III d.C. Tale fenomeno è in parte condizionato e influenzato dal substrato culturale, dalle tempistiche in cui è avvenuta la romanizzazione — soprattutto per quanto riguarda il materiale eseguito in cottura A — senza dimenticare i fattori legati alla mode e ai gusti.

In area italica, ed in particolar modo in ambito norditalico, il percorso forse è diverso. Sia le sigillate lisce che decorate a matrice padane con vernice di colore nero sono attestate soprattutto in epoca augustea e tiberiana, momento in cui in Italia settentrionale la produzione di ceramica a vernice nera è ancora attiva, come dimostrano le caratteristiche patere Lamboglia 7/16 con bollo in planta pedis (Dolenz et al. 2009: 258– 260; Mantovani 2015: 20-23).

All’interno delle varie produzioni citate diversi paiono essere i modelli ispiratori: per la Eastern Sigillata A,

20 

231

Si veda nota 14.

Valentina Mantovani, Eleni Schindler Kaudelka, Roby Stuani, Émilie Mannocci, Carla Corti

Figura 17. Alcuni esempi di dischi di impilamento su fondi di ceramica a vernice nera rinvenuta al Magdalensberg (foto: E. Schindler Kaudelka).

Il legame e lo scambio di modelli tra queste due produzioni deve essere stato in quel momento storico abbastanza forte, come testimoniano i marchi del ceramista Calenus (OCK 486) sulle Lamboglia 7/16 di Adria (RO) (Mantovani 2015: 22, tab. 2 n. 5, tav. 2f.), similmente a quanto attestato anche nella necropoli di Pellegrina/San Carlo (Isola della Scala -VR) (Biondani 2008: 70), oltre che su un fondo di piatto in sigillata padana Consp. B 1.8, ancora una volta adriese (RO) (Mantovani 2015: 63, tab. 5 n. 27). Può trattarsi, in questo specifico caso, del perdurare di una sorta di “romantico” attaccamento alla moda del nero?

finale, di età tiberiana, moltissimi esemplari di piatti, patere e coppe a vernice nera presentano il tipico disco di impilamento di colore rossastro, più o meno intenso, posto al centro del fondo interno del vaso (Figura 17). Il fenomeno è documentato anche nelle stoviglie a vernice nera rinvenute a Pompei (McKenzie Clark 2013: 16, PC 20), ad Arezzo (Morel 1963: 56), nella produzione calena terminale (Pedroni 1990: 170) e nella produzione di Rhode, in Spagna occidentale (Principal e Ribera i Lacomba 2013: 132). Questa abitudine, così ampiamente diffusa, non può essere esclusivamente imputabile ad una realizzazione più cursoria dei manufatti, ma la presenza di un elemento circolare di colore rosso può aver avuto anche una valenza estetica e essere stato realizzato volontariamente. Questo rappresenterebbe un ulteriore tassello utile alla definizione del fenomeno della terra sigillata nera.

E ancora va citato il caso della firma di Amphio Sari presente sia su una patera adriese (RO) Morel 2284/ Lamboglia 7 (Mantovani 2015: 63, tab. 2 n. 1, tav. IIa) che su una coppa Consp. 14 da Iulia Concordia (VE) di colore rosso (Sandrini 2007: 206, scheda n. 144), caso questo esemplificativo di molti altri vasi per i quali non è mai stata affrontata o messa in discussione la questione del bilinguismo cromatico. Analogo gioco di scambi si riscontra su due esemplari in terra sigillata e su una coppetta a vernice nera della necropoli della Ravisa (Brescello – RE), prodotti bollati dall’officina aretina di Murrius, i quali paiono essere testimonianza di una produzione diversificata anche in ambito aretino (Bresi 2013: 86). E sempre da Bologna, via Rizzoli, oltre al marchio ARRE/TINVM, sopra menzionato, sono attestati con rivestimenti rossi e neri anche prodotti firmati dall’aretino Volusenus e dai nord-italici Amicus e Sarius (in cartiglio rettangolare o in planta pedis) (Corti 2016: 90).

In numerose altre classi, come la ceramica comune da cucina e da dispensa, la presenza di manufatti e forme coevi prodotti con colorazioni bruno/nere o beige/ rossastre è sempre stato considerato un atto volontario. Per questi manufatti si sono sempre indagati fattori di carattere cronologico, tecnologico, estetico e culturale. Nell’ambito delle ceramiche fini da mensa le produzioni a pareti sottili a pasta grigia, prodotte in area nord-italica, ticinese e in area medio adriatica, affiancandosi alle canoniche forme con impasto rosato, vengono commercializzate con discreto successo, e subito inserite nel repertorio dei atelier locali anche in Pannonia, Grecia, Britannia, e penisola iberica (Tassinari 2014-2015: 146; Menozzi 1995).

Un ultimo aspetto, legato proprio all’estetica della ceramica a vernice nera della fase terminale, presenta interessanti spunti di riflessione. In ambito nord-italico, a partire soprattutto dall’età augustea e per tutta la fase

Ricordiamo, infine, quanto affermato da Andreas Grüner (2017: 25-36), secondo cui nello studio della ceramica romana si ha l’abitudine di non considerare l’importanza delle valutazioni di carattere estetico 232

Terra sigillata norditalica decorata a matrice intenzionalmente nera: alcune considerazioni

che, nel caso della moda del “rouge et noir”, crea effetti cromatici differenti su analoghi oggetti. Tale dimenticanza è forse figlia della formazione archeologica degli ultimi decenni, incentrata più sugli aspetti economici, tecnologici e tipo-cronologici dei manufatti e meno su quella storico-artistica, con la conseguente perdita di importanza del fattore estetico e del simbolo che l’oggetto rappresenta.

non si faceva menzione del particolare colore della vernice, così come per i sopra citati frammenti della Stipe del Montirone. E’ possibile quindi che anche altri manufatti già pubblicati, editi solo tramite disegni e privi d’informazioni sul colore dell’impasto e della vernice, presentino un rivestimento nero, che può essere individuato solo tramite la visione del pezzo. Il rinvenimento dei frammenti di Comacchio rappresenta un elemento importante per la comprensione della produzione intenzionale di terra sigillata nord-italica decorata a matrice con rivestimento nero. In effetti, il fatto che questi manufatti siano stati trovati in un contesto chiuso, insieme ad altri bicchieri e coppe con vernice rossa, conferma la coesistenza di queste due diverse produzioni. Inoltre la cronologia del relitto, databile tra il 19 ed il 12 a.C. grazie alla presenza a bordo di lingotti in piombo stampigliati (Domergue et al. 2012: 94), fornisce un precisa indicazione sul periodo di fabbricazione di questa particolare produzione.

In conclusione, tornando alle ceramiche nord-italiche decorate a matrice, potremmo dire che, accanto ai vasi non intenzionalmente ma solo casualmente neri, esistono manufatti prodotti con la chiara volontà di ottenere un aspetto nero, anziché rosso. La moda del “rouge et noir” è presente con chiari esempi nella produzione di terra sigillata nord-italica e non esistono elementi sufficientemente forti per escludere la volontà di realizzare anche vasi a matrice con le medesime caratteristiche. Tuttavia, lo stato di frammentarietà dei pezzi e le scarse testimonianze di queste varianti produttive non ne rendono sempre facile ed univoca l’identificazione. Alla luce di questa variabilità ogni singolo esemplare deve pertanto essere valutato singolarmente, in base alle sue caratteristiche formali, quali la qualità del rivestimento, l’omogeneità del colore, la presenza di un impasto rosato o grigio. Alle informazioni dedotte dall’osservazione fisiche vanno, poi, aggiunte quelle relative al contesto di provenienza e alla presenza di indicatori di un’esposizione al fuoco secondario, come ad esempio il rinvenimento in corredi di sepolture ad incinerazione, la possibilità di commercializzazioni di seconda qualità, oltre che, last not least, la deposizione in ambienti acidi che ne possono alterare la colorazione. Addendum

Figura 18. Bicchiere tipo Aco con rivestimento nero del relitto di valle di Ponti (Comacchio) (Rielaborazione grafica da Berti 1995, Tav. XXX,1).

Émilie Manocci Nel carico del relitto rinvenuto a Comacchio nel 1981 (Berti 1990) è stato rinvenuto un lotto importante di coppe tipo Sarius e di bicchieri tipo Aco in buona parte ancora inediti.21 Tra questo materiale sono stati individuati quindici frammenti di terra sigillata nord-italica decorata a matrice con rivestimento nero uniforme. Mentre la maggior parte dei manufatti inediti, tra cui sei orli e due pareti, non mostrano tracce di decorazione, la visione autoptica dei frammenti ha permesso di individuare un bicchiere troncoconico già noto in bibliografia (I.G. 59717), del quale si conservano tre frammenti di orlo e quattro di pareti (Figura 18), caratterizzato da impasto grigio e rivestimento nero. Nelle precedenti pubblicazioni (Berti 1995: 54, n. 43, Tav. XXX, 1; Lavizzari Pedrazzini 1997: 245, tav. 5, 2) 21 

Materiale in corso di studio da parte di É. Mannocci.

233

Valentina Mantovani, Eleni Schindler Kaudelka, Roby Stuani, Émilie Mannocci, Carla Corti Tabella 1. Tabella riepilogativa del materiale censito e citato nel testo. Inv.

Sito

Forma

Firma

Decorazione

Cronologia

TS NORD-ITALICA DECORATA INTENZIONALMENTE NERA (rivestimento uniforme) M65 17 M65 32

Magdalensberg (Kärnten Bicchiere - Österreich) Magdalensberg (Kärnten Bicchiere - Österreich)

NORBAN[I] -

M65 44

Magdalensberg (Kärnten Bicchiere - Österreich)

-

M65 29

Magdalensberg (Kärnten Bicchiere - Österreich)

-

M82 SH/11 M82 SH/12 M82 SH/9A

Magdalensberg (Kärnten Bicchiere - Österreich) Magdalensberg (Kärnten Bicchiere - Österreich) Magdalensberg (Kärnten Bicchiere - Österreich)

-

-

kommaregen

Bicchiere

-

non inventariato

Krk (Hrvatska)

Bicchiere

TYMS

6352

Stipe del Montirone (Abano terme, PD - Italia)

Bicchiere

-

6352

Stipe del Montirone (Abano terme, PD - Italia)

Bicchiere

-

142472 16780 nr. 380 nr. 497 inv. 5421 non inventariato non inventariato non inventariato non inventariato

Adria (RO -Italia)

Coppa

Bologna (Italia), via Rizzoli Bologna (Italia), via Rizzoli Bologna (Italia), via Rizzoli

Coppa tipo Sarius Coppa tipo Sarius Calice Consp. R4

Frejus (France)

Catene di perle a losanghe e tiberiano cerchietti kommaregen augusteo con ovolo kommaregen

Verona (Italia)

Aquileia, magazzini MAN Coppa (UD - Italia) Aquileia, Casa delle Bestie Coppa ferite (UD - Italia)

catena di perle tiberiano e rosette

-

non inventariato

non reperibile

geometricotiberiano floreale catena di perle tiberiano e rosette

-

augusteo

augusteotiberiano prima e kommaregen media età augustea prima età kommaregen augustea kommaregen a augusteolosanghe tiberiano kommaregen a augusteolosanghe tiberiano geometricofloreale geometricofloreale

augusteotiberiano augusteotiberiano augusteotiberiano età augusteotiberiana età augusteotiberiana età augusteotiberiana

Bibliografia Schindler Kaudelka 1980: tav. 13 n. 1 Schindler Kaudelka 1980: tav. 18 n. 40

Illustrazione   Fig. 2 Fig. 2

Schindler Kaudelka 1980: tav. 18 n. 40

-

Schindler Kaudelka 1980: tav. 17 n. 32

-

Schindler Kaudelka 1998: n. 1 Schindler Kaudelka 1998: n.3

Fig. 2 Fig. 2

inedito

-

Stuani 2017: p. 112, Tav. 21.13

-

Inedito

Fig. 3

Lavizzari Pedrazzini1995: 146, fig. 8 nn. 5–8. Lavizzari Pedrazzini1995: 146, fig. 8 nn. 5–8.

Fig. 4 Fig. 4

Inedito

-

Inedito

Fig. 5

inedito

Fig. 6

Fava 1962: 57, n. 380

Fig. 7

-

floreale

-

floreale

 

figurato

-

 

Coppa

-

figurato

20-15 a.C.

Rivet 2019

Fig. 8

Frejus (France)

Coppa

-

vegetale?

20-15 a.C.

Rivet 2019

-

Frejus (France)

Coppa

-

puntinato

20-15 a.C.

Rivet 2019

-

Frejus (France)

Coppa

-

puntinato

20-15 a.C.

Rivet 2019

-

menzione in Corti 2016: 90 menzione in Corti 2016: 90

TS decorata NON INTENZIONALMENTE NERA (rivestimento non uniforme) 39269, 39306 e Verona (Italia) 39323

Fig. 7

  prima e media età augustea prima metà I d.C.

Coppa

-

floreale

35885

Spinimbecco (VR - Italia) Coppa

-

floreale

non reperibili

Agora Tetragonos di Efeso (Türkiye)

Coppa

C. Sentius

-

20 a.C.-10 d.C.

17194

Metropolitan Museum (New York - USA)

Coppa

SEX AVLIENI

figurato

metà I d.C.?

M80 SH/5

Magdalensberg (Kärnten Calice – Österreich)

Clemens

scena di caccia tardotibericane-cinghiale ano

M50 101,102

Magdalensberg (Kärnten Coppa – Österreich)

-

figurato

234

Fig. 7

20 .-10 .a.C

Stuani 2017: p. 113, Tav 22.6 Stenico 1973: 118, figg. 8 a–b Zabehlicky–Scheffenegger, 1996: 253-271; Zabehlicky–Scheffenegger 1997 Soricelli 2017: 101, fig. 24 Schindler Kaudelka 1998: 184, figg. 371, 373, 388 Schindler-Kaudelka 2017, 197, Fig. 10 a, 10 b

Fig. 9 Fig. 10   Fig.12 Fig. 13 Fig. 11

Terra sigillata norditalica decorata a matrice intenzionalmente nera: alcune considerazioni

Inv.

Sito

Forma

Firma

Decorazione

Cronologia

Bibliografia

Illustrazione

239219

Scoppieto (TR -Italia)

Coppa

-

figurato

inizi I d.C.età flavia

Bergamini et al. 2017, tav. 3 n. 6

-

prima e media età augustea

Schindler 1967, Tav. Schindler and Scheffenegger 1977 tav. XXX, Schindler 1986, Abb.

Fig. 1

Rodegher 2012–2013: t. n. 109, n. 3, tav. VIII n.3

-

Gorla and Polidoro in press

-

TS NORD-ITALICA LISCIA INTENZIONALMENTE NERA non inventar- Magdalensberg (Kärnten Consp. 22 iato – Österreich) 3 esemplari

illeggibile

-

non reperibile Zevio (VR - Italia)

Consp. 22

-

-

non reperibile Legnano (MI - Italia)

Consp. 21

 

 

Consp. 18

 

 

Consp. 22

 

 

Consp. 26

 

 

inv. 5400 nr. 5369, s.n. nr. 326

Bologna (Italia), via Rizzoli Bologna (Italia), via Rizzoli Bologna (Italia), via Rizzoli

nr. 78, nr. 74

Bologna (Italia), via Rizzoli

Consp. 38

MAE/ PATIS e MAE/ PATES

 

nr. 17

Bologna (Italia), via Rizzoli

Consp. 38

LAVRI

 

TS LIONESE LISCIA con difetto di cottura non reperibili Avenches (Switzerland)

Consp.14

ATEIUS

-

prima e media età augustea augusteotiberiano età augusteotiberiana

menzione in Corti 2016: 90 menzione in Corti età augustea 2016: 90 prima metà menzione in Corti I d.C. 2016: 90 augusteaprimissima etá tiberiana (per la produzione Corti 2016, Fig. 2, 15-16 della forma) e Fig. 3, 1-2 augusteaprimissima etá tiberiana (per la produzione Corti 2016, Fig. 2, 12 e della forma) Fig. 3, 5-6 augusteo

Lyon (France) Pseudo Consp. 22 ATEIUS augusteo Sanctuaire de Cybèle Vernice nera e TS NORD-ITALICA LISCIA CON MEDESIMO BOLLO rettangolare o ovale non reperibili

 

-

-

-

 

Blanc and Meylan Krause Fig. 16 1997: fig.10b, no cat. 1 Desbat 2008: 527–533  

Consp. 36 TS nera

PASSI/ TELAM

-

20-1 a.C.

Mantovani 2015: tab. 5 n. 94

Consp. 11 TS TS rossa

PASSI/ TELAM

-

20-1 a.C.

Zabehlicky-Scheffenegger 2000, fig. 3

Fig. 14

Magdalensberg (Kärnten Consp. 5 TS - Österreich) rossa

ARRETI

-

Schindler and Scheffenegger 1977 tav. 17,1, Taf. 91

-

non inventar- Magdalensberg (Kärnten Consp. B3.12 iato - Österreich) TS nera

ARRET

-

Schindler 1967, Tav. 6,20, e 4C’’

-

non inventar- Magdalensberg (Kärnten Consp. B 3.12 C. ACO iato - Österreich) TS rossa

-

Augusteo

Zabehlicky-Scheffenegger 2000 fig. 3

-

123919, 123920 e 123937

Verona (Italia)

Consp. 18.2 TS rossa

C. ACO

-

Augusteo

Inedito

-

16542

Adria (RO -Italia)

Lamboglia 7

-

Consp. 14 TS TS rossa

prima età augustea

Mantovani 2015: 63, tab. 2 n. 1, tav. IIa.)

C32061

Concordia Sagittaria (VE - italia)

AMPHIO SARI

82875

Adria (RO -Italia)

80390

Adria (RO -Italia)

80237

Adria (RO -Italia)

Magdalensberg (Kärntnon inventaren – Österreich) 10 iato esemplari M66

prima e media età augustea prima e media età augustea

AMPHIO SARI

-

Consp. B 1.8 TS rossa

CALENI

-

15 d.C. +

Mantovani 2015: 63, tab. 5 n. 27

Lamboglia 7/16

CALENI

-

15 d.C. +

Mantovani 2015: 22, tab. 2 n. 5, tav. 2f.)

non reperibile Isola della Scala (VR - Italia) Lamboglia 7/16 CALENI

-

15 d.C. +

Biondani 2008: 70

Vernice nera e TS NORD-ITALICA LISCIA CON MEDESIMO BOLLO IN PL. PEDIS

235

prima età augustea

Sandrini 2007: 206, scheda n. 144

Fig. 14

 

-

Valentina Mantovani, Eleni Schindler Kaudelka, Roby Stuani, Émilie Mannocci, Carla Corti productions manufacturées de l’Antiquité à l’époque moderne 42:17–19. Corti, C. 2016. Le coppe Haltern 14/Conspectus 38 in terra sigillata nord-italica. Produzione e diffusione. Rei Cretariae Romanae Fautorum Acta 44: 85–94. Desbat, A. 2008. La sigillée des fosses à banquets du “Sanctuaire de Cybèle” à Lyon. SFECAG. Actes du congrès de L’Escala/Empuries: 527‒533. Deru, X. 1995. La céramique belge dans le nord de la Gaule: caractérisation, chronologie, phénomènes culturels et économiques (Publ. d’Hist. de l’Art et d’Archéo. de l’Université catholique de Louvain 89). Louvain: Louvain-la-Neuve. Di Meo, P. 2000-2001. Proposta di classificazione della ceramica comune da Legnano romana e dal suo territorio, tesi di laurea inedita, Università statale di Milano. Dolenz, H., S. Krmnicek, E. Schindler-Kaudelka, H. Sedlmayer e S. Zabehlicky-Scheffenegger 2009. Zur voranexionszeitlichen Siedlung auf dem Magdalensberg. Fundberichte aus «Österreich» 47: 235–266. Domergue, C., P. Quarati, A. Nesta e P. R. Trincherini 2012. Retour sur les lingots de plomb de Comacchio (Ferrara, Italie) en passant par l’archéométrie et l’épigraphie, in A. Orejas e C. Rico (eds) Minería y metalurgia antiguas. Visiones y revisiones. Homenaje a Claude Domergue (Collection de la Casa de Velázquez 128): 81‒104. Madrid : Casa de Velázquez. Gabucci, A. 1995. Marchi di fabbrica da Tortona. Terra sigillata italica, nord-italica e sud-gallica. Quaderni di Archeologia del Piemonte 13: 29–58. Giordani, N.1990. La bassa pianura modenese in età romana: la documentazione archeologica, in M. Calzolari e N. Giordani (eds) Archeologia a Mirandola e nella Bassa modenese: 85–109. Mirandola: Banca Popolare di San Felice sul Panaro edizione. Gorla, D. e L. Polidoro in press. Una scoperta dai magazzini del Museo di Legnano. Bollettino della Società Arte e Storia di Legnano. Grüner, A. 2017. Schönheit und Massenproduktion. Die Ästhetik der Terra Sigillata, in M. Flecker (ed.) Neue Bilderwelten. Zu Ikonographie und Hermeneutik Italischer Sigillata. (Tübinger Archäologische Forschungen 23): 25–37. Rahden/Westf: Verlag Marie Leidorf GmbH. Jorio, S. 1991. Terra sigillata, in D. Caporusso (ed.) Scavi MM3. Ricerche di archeologia urbana a Milano durante la costruzione della linea 3 della metropolitana, 1982-1990: 57–88. Milano: Edizioni ET. Lang-Auinger C. e V. Mitsopoulos-Leon 2007. Die Basilika am Staatsmarkt in Ephesos. 2. Teil: Funde  der klassischen bis römischen Zeit (FiE 9, 2, 3). Wien: VÖAW. Lavizzari Pedrazzini, M. P. 1995. Il deposito del Montirone. Quaderni di Archeologia del Veneto 11: 109‒166. Lavizzari Pedrazzini, M. P. 1997. Nuove osservazioni sul Vasellame “Tipo Aco”. Athenaeum n.s. 85/1: 233‒251. Mantovani, V. 2015. Ceramiche fini da mensa di Adria romana. Le indagini di via Retratto (1982 e 1987)

Bibliografia Bergamini, M., C. Troso e V. Dezza 2017. La manifattura romana di Scoppieto (Italia -Umbria). La produzione di terra sigillata, in M. Flecker (ed.) Neue Bilderwelten. Zu Ikonographie und Hermeneutik Italischer Sigillata. Tübinger Archäologische Forschungen 23: 73–90. Rahden/Westf: Verlag Marie Leidorf GmbH. Berman, L. 2015. Céramiques gallo-romaines fabriquées a Thonon-les-Bains. S.F.E.C.A.G. Actes du congrès de Nyon: 245‒252. Berti, F. 1990. Fortuna Maris: la nave romana di Comacchio. Bologna: Nuova Alfa. Berti, F. 1995. Osservazioni a margine di alcune ceramiche del relitto navale di valle Ponti (Comacchio) e sul loro commercio, in B. M. Giannattasio (ed.) Viaggi e commerci nell’antichità. Atti VII giornata archeologica (Genova, 25 novembre 1994): 39‒65. Genova: Dipartimento di archeologia, filologia classica e loro tradizioni “Francesco Della Corte”. Biondani, F. 2008. Tracciati della via “Claudia Augusta padana” e siti di età romana nel comune di Isola della Scala, in Da Ostiglia a Verona. Archeologia e storia di una strada romana. Guida alla Mostra (Isola della Scala, Chiesa di Santa Maria Maddalena, 13 settembre-8 dicembre 2008): 65–71. Isola della Scala (VR): Tipografia Gi. Bi. Blanc, P. e M. F. Meylan Krause 1997. Nouvelles données sur les origines d’Aventicum. Les fouilles de l’insula 20 en 1996. Bulletin de l’Association Pro Aventico 39: 29–100. Bottazzi, G. L. e P. Bigi 2001. Domagnano. Dal Tesoro alla storia di una comunità in età romana e gota. San Marino: Museo di stato. Bresi, I. 2013. Storia di Bresciello. L’età romana. Parma: Monte Università Parma. Ça s’est passé sous vos pieds, catalogo della mostra, Museo di Chablais (28 marzo-8 novembre 2015) Cavalieri Manasse, G. e M. Bolla 1998. Osservazioni sulle necropoli veronesi, in P. Fasold, T. Fischer, H. von Hesberg e M. Witteyer (eds) Bestattungssitte und kulturelle Identität. Grabanlagen und Grabbeigaben der frühen römischen Kaiserzeit in Italien und den NordwestProvinzen (Atti Conv. “Römische Gräber des 1. Jhs. n. Chr. in Italien und in den Nordwestprovinzen”, Xanten, 1995) (Xantener Berichte 7): 116–139. Bonn : RheinlandVerlag, R. Habelt. Carrara, M. 2013. Patinarum Paludes. Scarti di “Graue Platten” e relativa fornace presso la Celsa (Roma). Bollettino di Archeologia On Line 3: 1‒27. Conspectus 1990. Ettlinger, E., B. Hedinger, B. Hoffmann, Ph. Kenrick, G. Pucci, K. Roth-Rubi, G. Schneider, S. Von Schnurbein, C. M. Wells e S. Zabehlicky-Scheffenegger 1990. Conspectus formarum terrae sigillatae italico modo confectae. Bonn: Dr. Rudolf Habelt GmbH. Corti, C. 2015. Il bollo MAE/PATES su coppe Consp. 38 di produzione nord-italica. Instrumentum. Bulletin du Groupe de travail européen sur l’artisanat et les 236

Terra sigillata norditalica decorata a matrice intenzionalmente nera: alcune considerazioni

(Collezioni e Musei Archeologici del Veneto 48). Roma: Giorgio Bretschneider editore. Martin-Kilcher, St. 1999. Ts-Imitation, in C. Schucany, S. Martin-Kilcher, L. Brtger e D. Paunier (eds) Römische Keramik in der Schweiz (Antiqua 31): 33‒36. Basel: Archeology Switzerland. Mazzeo Saracino, L. 1983. Problemi della terra sigillata italica nella Regione VIII, in G.A. Mansuelli (ed.) Studi sulla città antica. L’Emilia Romagna: 465-495. Roma: L’Erma di Bretschneider. Menozzi, O. 1995. La Ceramica a Pareti Sottili Grigie, in N. Christie (ed.) Settlement and Economy in Italy 1500 BC - AD 1500. Papers of the Fifth Conference of Italian Archaeology (Oxbow Monograph 41): 579‒590. Oxford: Oxbow Books Ltd. Morel, J. - P. 1963. Notes sur la céramique étruscocampanienne. Vases à vernis noir de Sardaigne et d’Arezzo. Mélanges d’archéologie et d’histoire 75/1: 7‒58. Morel, J. - P. 1987. La céramique à vernis noir en Italie septentrionale, in D. Vitali (ed.) Celti ed Etruschi nell’Italia centro-settentrionale dal V sec. a.C. alla romanizzazione. Atti del colloquio internazionale (Bologna 1985): 111‒133. Bologna: Bologna University Press. Morel, J. - P. 1998. La ceramiche a vernice nera del Piemonte: tipologia, storia, cultura, in L. Mercando (ed.) Archeologia in Piemonte. L’età romana, Vol. II: 235‒252. Torino: U. Allemandi. Paccolat, O. e F. Wiblé 2007. Quelques considerations sur les relations commerciales entre le Valais et l’Italie du nord a l’époque romaine, in F. Butti Ronchetti (ed.) Produzioni e commerci in Traspadana in età romana, Atti del Convegno (Como, Villa Olmo, 18 novembre 2006) (CD Rom). Como. Pedroni, L. 1990. Due bolli in “planta pedis” da Cales e la ceramica a vernice nera terminale. Samnium 63: 170‒181. Principal J., A. Ribera e Lacomba, A. 2013. El material más apreciado por los arqueólogos. La cerámica fina.  La cerámica de barniz negro, in A. Ribera i Lacomba  (ed.) Manual de cerámica romana. Del mundo Helenístico al Imperio romano: 41‒146. Madrid: Comunidad de Madrid, Museo Arqueológico Regional, Colegio de Doctores y Licenciados en Filosofía y Letras y en Ciencias de la Comunidad de Madrid. Rivet, L. 2019. Datations céramiques entre 45 et 15 av. n.è.: l’apport des paliers chronologiques issus des fouilles sur la Butte Saint-Antoine à Fréjus aux sigillées italiques et aux types Aco et Sarius, in Chronologie und vergleichende Chronologie zum ausgang der römischen Republik und zur frühen Kaiserzeit (Klagenfurt 19–21 September 2018): 437–448. Rodegher G. 2012-2013. La necropoli di S. Martino di Zevio – Piazza Rivalunga (Verona): inquadramento preliminare, tesi di laurea magistrale inedita, Università degli Studi di Padova.

Sandrini, M.G. 2007. Schede, in E. Pettenò (ed.) Vasa Rubra. Marchi di fabbrica sulla terra sigillata da Iulia Concordia: 254‒255, 304. Padova: Esedra Editrice. Schindler, M. 1967. Die “Schwarze Sigillata” des Magdalensberges (Kärntner Museumsschriften 43). Klagenfurt: Verlag des Geschichtsvereines fur Kärnten. Schindler, M. 1986. Die “Schwarze Sigillata” des Magdalensberges 2; Neufunde seit 1965. Die Ausgrabungen auf dem Magdalensberg 1975 bis 1976 (Magdalensberg - Grabungsbericht 15): 345‒ 390. Klagenfurt: Verlag des Geschichtsvereines fur Kärnten. Schindler, M. e S. Scheffenegger 1977. Die glatte rote Terra Sigillata vom Magdalensberg (Kärntner Museumsschriften 62). Klagenfurt: Verlag des Geschichtsvereines fur Kärnten. Schindler Kaudelka, E. 1980. Die römische Modelkeramik vom Magdalensberg (Kärntner Museumsschriften 66). Klagenfurt: Verlag des Geschichtsvereines fur Kärnten. Schindler Kaudelka, E. 1998. Die römische Modelkeramik vom Magdalensber 2, in G. Piccottini (ed.) Die Ausgrabungen auf dem Magdalensberg 1980 bis 1986 (MagdalensbergGrabungsbericht 16): 289‒378. Klagenfurt: Verlag des Geschichtsvereines fur Kärnten. Schindler-Kaudelka, E. e S. Zabehlicky-Scheffenegger 2000. Es war der Schadensbrand und nicht der Fehlbrand. Rei Cretariae Romanae Fautorum Acta 36: 571‒574. Schindler Kaudelka, E., V. Mantovani e J. Krajsek 2017. Norditalica decorata: die rein dekorative Seite der  italiche  Reliefkeramik. Fallbeispiele aus Adria, vom Magdalensberg und aus Celje, in M. Flecker (ed.) Neue Bilderwelten. Zu Ikonographie und Hermeneutik Italischer Sigillata (Tübinger Archäologische Forschungen 23): 177–198. Rahden/Westf: Verlag Marie Leidorf GmbH. Schindler Kaudelka, E., S. Magnani, V. Mantovani, L. Maritan, A. Roumégous, and R. Stuani 2020. Il vasaio Νικoστρατος: le testimonianze di Altinum, Verona e Aurasio, in I. Kamenjarin and M. Ugarković (eds) Exploring the Neighborhood. The Role of Ceramics in Understanding Place in the Hellenistic World (IARPotTH 3): 163‒177. Wien: Phoibos Verlag. Soricelli, G. 2001. Terra sigillata decorata a rilievo da Pompei: la serie “tardo-puteolana”, in P.G. Guzzo (ed.) Pompei. Scienza e Storia: 259‒264. Milano: Mondadori Electa. Soricelli, G. 2017. La sigillata puteolana decorata a rilievo. Alcune osservazioni, in M. Flecker (ed.) Neue Bilderwelten. Zu Ikonographie und Hermeneutik Italischer Sigillata (Tübinger Archäologische Forschungen 23): 91–112. Rahden/Westf: Verlag Marie Leidorf GmbH. Stenico, A. 1973. Terra Sigillata nord-italica dal territorio di Verona romana, in Il territorio veronese in età romana, Atti del Convegno (22-23-24 ottobre 1971): 109‒136.Verona: Accademia di agricoltura, scienze e lettere di Verona. 237

Valentina Mantovani, Eleni Schindler Kaudelka, Roby Stuani, Émilie Mannocci, Carla Corti Stuani, R. 2017. Análisis de la cerámica de barniz negro y de su contexto de hallazgo como instrumento de interpretación histórico arqueológica del territorio: el caso de Verona, tesi di dottorato inedita, Università di Barcelona.http:// diposit.ub.edu/dspace/bitstream/2445/113006/1/ ROBY%20STUANI_TESIS.pdf> Tassinari, G. 2014-2015. La ceramica a pareti sottili in Lombardia (e non solo). Conoscenze, studi, questioni. Rivista di Studi Liguri 80-81: 125–174. Warner Slane, K. 1997. The Fine Wares, in S. C. Herbert (ed.) Excavations at Tel Anafa, II, i. The Persian, Hellenistic, and Roman Plain Wares (Journal of Roman Archaeology supplementary series vol. 10.2): 247– 393. Ann Arbor MI: Portsmouth RI. Zabehlicky-Scheffenegger, S. 1997. Rouge et noir. Forum Archaeologiae - Zeitschrift für klassische Archäologie 4 /VIII.

Zabehlicky-Scheffenegger, S. e R. Sauer 2000. Metodi di distinzione dei due gruppi di sigillata padana augustea trovati sul Magdalensberg, in G. P. Brogiolo e G. Olcese (eds) Produzione ceramica in area padana tra il II secolo a.C. e il VII secolo d.C.: nuovi dati e prospettive di ricerca (Documenti di Archeologia 21): 69–78. Modena: SAP. Zabehlicky-Scheffenegger, S. 2003. C. Sentius and his commercial connections, in C. Abadie-Reynal (ed.) Les céramiques en Anatolie aux époques hellénistique et romaine. Actes de la Table Ronde d’Istanbul, 22-24 mai 1996 (Varia Anatolica 15): 117‒119. Istanbul: Institut français d’études anatoliennes Georges Dumézil.

238

Canale Anfora, recupero 1988: contributo alla conoscenza della ceramica fine orientale ad Aquileia Paola Maggi, Paola Ventura Abstract The paper deals with the Eastern Mediterranean fine pottery from the rescue excavation conducted by Luisa Bertacchi in 1988 in a nowadays filled stretch of the so-called “Canale Anfora” in Aquileia: the extraordinarily abundant, still unpublished, findings recovered from it are highly representative of the Middle Imperial age. The analysed vessels find quite punctual comparison with the complex coming from the stratigraphic investigations that were conducted in a different sector of the same channel in 2004-2005. Together with the recently published Eastern sigillata and Roman Corinthian ware from the historical collections of the Museum, the pottery from Canale Anfora increases our knowledge of these productions, scarcely testified in our region up to 15 years ago. The nearly 150 finds mostly belong to the Eastern Sigillata: alongside a few specimens of ES A (both closed and open forms), about 90% of the total amount can be ascribed to ES B, including 14 forms – two of them classified as ES B1, the remaining ones in the later ES B2 production; the common types (the most frequently occurring Hayes 60, together with Hayes 62 and Hayes 63 plates, and Hayes 70, Hayes 71, Hayes 74, Hayes 75, Hayes 76, Hayes 80 bowls) are associated to more sporadically or so far never attested ones, e.g. Hayes bowl 79. Only one specimen bears a stamp in Greek letters, of a rather frequent type (MAP / KOY), while about ten specimens have anepigraphic marks on the bottom (palmettes, rosettes). On a plate only hypothetically assigned to Eastern Sigillata the name CHRYSIIROS can be read. Among moulded vessels we count three fragments of Roman Corinthian ware, all of them belonging to cups / pyxides with a cylindrical body, depicting scenes of Amazonomachy or battle. The Knidian ware is testified by fragments of two zoomorphic vases (in the shape of a lion and a ram) and of one jug.

Key words Aquileia, Anfora Channel, Eastern Sigillata, Roman Corinthian ware, Knidian relief ware.

Premessa1 La presentazione nel 2017, al Convegno di cui qui si concretizzano gli atti, del vasellame da mensa orientale proveniente dallo scavo di un tratto del Canale Anfora ad Aquileia effettuato nel 1988 da Luisa Bertacchi (Bertacchi 1988; Bertacchi 2000)2 era intesa come un apporto utile a ricostruire un’immagine attendibile dell’incidenza di queste produzioni, fino allora sottostimate, nella documentazione archeologica della colonia altoadriatica (Figura 1). Il poster presentato si poneva in continuità con un contributo a più voci del 2016, in cui si erano considerate alcune classi ceramiche significative per la definizione dei traffici in età medio/tardoimperiale (Donat et al. 2022)3 estrapolandole da quell’imponente complesso di Ringraziamo Marta Novello, Direttore del Museo Archeologico Nazionale di Aquileia, per aver agevolato lo studio del materiale; la nostra riconoscenza va anche al personale del museo, in particolare a Adriana Comar, autrice dei disegni. Le fotografie sono di Paola Maggi. 2  Come è noto, il Canale Anfora costituiva una delle principali vie di traffico commerciale di Aquileia, ben integrata nel suo sistema portuale; con un percorso di circa 6 km a partire dal limite urbano occidentale raggiungeva la Laguna di Marano. Nel 1988 fu esplorato con metodo non stratigrafico, su una fascia di circa 300 metri, un tratto ormai interrato del canale a est del Fiume Terzo, nell’immediata periferia della città. 3  Accanto al vasellame fine orientale, venivano in tale sede esaminate tutte le altre ceramiche da tavola e veniva inoltre fornita un’analisi quantitativa delle ceramiche comuni di importazione. 1 

materiali (databile fra la fine del I secolo a.C. e il IV/V secolo d.C. con una fase prevalente tra fine I e III d.C.), ad oggi rimasto sostanzialmente inedito.4 L’esame di questi reperti era stato stimolato dal progetto – giunto a conclusione proprio nel 2017 – dell’edizione integrale dello scavo condotto successivamente, nel 2004-2005, da Franca Maselli Scotti in un contiguo tratto del canale (Maggi et al. (eds) 2017). Come accennato, la decisione di dedicare in occasione del Convegno di Crikvenica una trattazione specifica alle ceramiche fini orientali derivava dalla constatazione delle scarsissime attestazioni allora note ad Aquileia e più in generale nella Venetia orientale,5 per motivi che sono stati illustrati altrove (Ventura e Maggi 2020: 157). Da allora, tuttavia, i dati noti hanno beneficiato di un significativo incremento: infatti, non solo sono uscite alcune edizioni di recenti scavi aquileiesi includenti nuovi lotti di sigillate orientali, ma sono stati anche Dallo scavo proviene una quantità molto ingente di ceramica, con più di 3000 esemplari classificati, oltre ad un consistente nucleo di anfore. 5  Il poster segnalava l’incremento dei materiali noti rispetto ai 24 frammenti registrati nella precedente sintesi in Maggi 2006, con gli ulteriori 107 frammenti da vari contesti editi nel lasso di tempo fra 2006 e 2017. I 134 pezzi allora presentati andavano quasi a raddoppiare il corpus conosciuto, che risultava triplicato conteggiando anche i 144 frammenti dallo scavo di Canale Anfora 2004-2005, poco dopo effettivamente editi: 135 frammenti di Eastern Sigillata (Maselli Scotti 2017), tre di ceramica corinzia (Mandruzzato 2017) e sei di cnidia (Giovannini e Mandruzzato 2017). 4 

roman pottery and glass manufactures (Archaeopress 2022): 239–253

Paola Maggi, Paola Ventura

Figura 1. Aquileia, Canale Anfora — scavo 1988. Localizzazione del tratto del canale indagato (rielaborazione da Maggi et al. (eds) 2017).

affrontati, a cura delle scriventi, la ricognizione e lo studio delle ceramiche fini importate dall’Oriente conservate nei depositi del Museo Archeologico Nazionale; ne è derivata la pubblicazione della sigillata e delle coppe corinzie,6 in cui si è fornito un quadro

aggiornato al 2020 delle attestazioni complessive nell’emporio altoadriatico (Ventura e Maggi 2020: 157158, 14 anni dopo Maggi 2006). Rimandiamo dunque a quella sede per le premesse, in particolare per quanto riguarda gli studi pregressi ad Aquileia e in ambito regionale; in aggiunta va però segnalato l’importante contributo alle conoscenze nel frattempo scaturito dal lavoro per l’edizione, nel 2021, del materiale proveniente

L’analisi ha riguardato la sigillata orientale A (un esemplare) e B (32 pezzi) e la ceramica corinzia (34 frammenti, di cui almeno 13 nuove presenze), mentre è stato rinviato a un secondo momento l’esame della ceramica cnidia. 6 

240

Canale Anfora, recupero 1988: contributo alla conoscenza della ceramica fine orientale ad Aquileia

Figura 2. Aquileia, Canale Anfora — scavo 1988. Le ceramiche fini orientali nel momento dello studio (photo: P. Maggi).

dalle indagini 2009-2013 ai fondi Cossar.7 In questo contesto le ceramiche fini orientali assommano a ben 208 esemplari, venendo ora a rappresentare a livello regionale l’insieme coerente più cospicuo e articolato, visto l’ampio spettro di produzioni documentate (Dobreva e Griggio 2021).8 P. M., P. V.

rilevato che la minore incidenza, sul totale, rispetto a quanto riscontrato per la stessa classe nell’ambito del materiale proveniente dallo scavo del 2004-2005, ove essa predomina con oltre un terzo fra tutte le ceramiche fini, discende probabilmente dall’esclusione dall’analisi, in quel caso, dei livelli più superficiali, per concentrarla sulle fasi di vita del canale databili all’età medio-imperiale, periodo che corrisponde all’apogeo delle importazioni dal Mediterrano orientale (Donat et al. 2022; Donat et al. 2017: 65 e fig. 1 – 38,12% su 525 esemplari – e in particolare 67-68; Ventura e Maggi 2020: 158).

Il quadro delle attestazioni Passando a esaminare le presenze (ed assenze), ci troviamo davanti ad un campione di 147 frammenti, quasi tutti diagnostici (138), a testimonianza della modalità selettiva messa in atto nel recupero, abbastanza scontata trattandosi di un intervento di scavo non stratigrafico (Figura 2 e Tabella 1). Il lotto costituisce approssimativamente il 19% di tutte le ceramiche fini provenienti dal contesto (circa 700 pezzi inquadrabili per produzione e forma, di cui quasi la metà è pertinente alla sigillata africana (Donat et al. 2022). È già stato

È invece abbastanza sovrapponibile l’incidenza delle diverse classi di vasellame fine orientale (anche se nel nostro caso non si sono prese in considerazione le ceramiche comuni orientali da mensa), con un’assoluta prevalenza della ES B (125 diagnostici, pari a oltre il 90% del totale) rispetto alla ES A (5), alla ceramica corinzia (3) e alla cnidia (3).9 Eastern Sigillata

In precedenza erano state diffuse alcune limitate anticipazioni riferite a contesti rappresentativi in Bonetto et al. 2009 e in Dobreva 2012, di cui si era tenuto conto in Ventura e Maggi 2020. Resta autonomamente valido il censimento delle sigillate fini dai vecchi scavi condotti ai fondi Cossar da L. Bertacchi pubblicate in Dobreva e Luise 2017, parimenti già riportato in Ventura e Maggi 2020. 8  Siamo riconoscenti alle autrici per l’anticipazione dei risultati in fase di bozze. Ne abbiamo tenuto conto soprattutto riguardo alle attestazioni, senza invece confrontarci con gli aggiornamenti alle carte di distribuzione e con le più ampie considerazioni di carattere storico-commerciale; a tale contributo si rinvia per bibliografia esaustiva sulle forme, che per la maggior parte trovano corrispondenza in entrambi i contesti. 7 

In dettaglio, la ES A è presente in quantità ridotte, annoverando comunque tanto forme aperte che forme chiuse, come attestato sia dai pochi (cinque) elementi Nello scavo di Canale Anfora 2004-2005 la ES B rappresentava il 69,63% delle orientali da mensa (cfr. Donat et al. 2017: 67, fig. 4), ma, se si scorpora la ceramica comune da tavola, ammonta anche in questo caso a oltre il 90% (con 132 pezzi su 144). 9 

241

Paola Maggi, Paola Ventura Tabella 1. Aquileia, Canale Anfora — scavo 1988: tabella delle presenze relative alle sigillate orientali. TAVOLA / FIGURA

FORMA

BOLLO/GRAFFITO

INVENTARIO ALTRI ESEMPLARI

Tavola 1,2 Tavola 1,3  /

Hayes 51 Hayes 101 o 104 Hayes 105 ?

     

572828 572830 572833

Figura 3 Eastern Sigillata B Tavola 1,4 Tavola 1,5 Tavola 1,6 Tavola 1,7 Tavola 1,8 Tavola 2,1  /  /

n.d. (brocca)   Hayes 19 Hayes 30 Hayes 30 Hayes 55 Hayes 55 - formato grande Hayes 60 - variante antica Hayes 60 - passaggio variante tarda Hayes 60 passaggio - variante tarda piccolo formato Hayes 60 - variante tarda di grande formato

                   

Hayes 60 - variante tarda piccolo formato Hayes 60 - variante tarda formato n.d. Hayes 60 - variante tarda grande formato Hayes 60 - variante n.d.

 

Tavola 2,5 Tavola 3,1 e Figura 4,4

Hayes 62A Hayes 63

Tavola 3,2   Tavola 3,3

Hayes 63 Hayes 62 o Hayes 63 (fondo) Hayes 70

Tavola 3,4 Figura 4,2

Hayes 70 ? Hayes 70 ?

Figura 4,7

Hayes 70 ?

Tavola 3,5 e Figura 4,5 Figura 4,6

Hayes 71

Tavola 3,6 e Figura 4,9

Hayes 74A

Tavola 3,7 Tavola 4,1 e Figura 4,3  /

Hayes 74A - formato piccolo Hayes 75

  bollo a rosetta entro cartiglio circolare diam. 1 cm     bollo mal stampigliato e illeggibile (rosetta?)   bollo a palmetta schematizzata - lungh. 1,2; largh. 0,9 cm bollo a rosetta a otto petali entro cartiglio circolare - diam. 0,8 cm; all’esterno del fondo graffito X bollo a rosetta a sette petali - diam. 0,8 cm bollo evanido (rosetta?) entro cartiglio circolare - diam. 0,7 cm bollo a lettere greche entro cartiglio rettangolare MAPKOY - cartiglio: 0,9 x 0,8 cm bollo illeggibile bollo a palmetta lungh. 1,5; largh. 1 cm bollo anepigrafe non riconoscibile

Eastern Sigillata A   Tavola 1,1 Hayes 51

Tavola 2,2 Tavola 2,3 Tavola 2,4  / Figura 4,1  /

   

bollo anepigrafe frammentario (palmetta?) Hayes 60 - variante tarda piccolo formato  

Hayes 71 ?

Hayes 75

  572829

   

572826   572884 572827 572883 572865 572940 572885 572911 572907

          572864   572923   572908, 572912, 572914

572892

572887, 572889, 572890, 572891, 572893, 572897, 572898, 572899, 572901, 572902, 572904, 572906, 572909, 572915, 572916, 572918  

572886 572895

    bollo a palmetta 572919 lungh. 1,8; largh. 1,1 cm    

242

572841 572851 572863 572931 572924 572938 572925 572926

     

572888, 572894, 572896, 572903, 572905, 572910, 572913 572900, 572917  

572921, 572920, 573254 + (probabili) 572922, 572935, 572936, 574613 572840   572852, 572853, 572854        

572881

  572882

572927

572929

572835

572837

572839 572845

572836, 572838  

572847

572842, 572843, 572844, 572846, 572848, 572849, 572850, 572880, 575975

Canale Anfora, recupero 1988: contributo alla conoscenza della ceramica fine orientale ad Aquileia

TAVOLA / FIGURA  / Tavola 4,2 Tavola 4,3

FORMA

Hayes 74 o Hayes 75 (fondo) Hayes 76A Hayes 76B

BOLLO/GRAFFITO      

INVENTARIO ALTRI ESEMPLARI 572930 572855 572866

Tavola 4,4

Hayes 77

 

572867

Tavola 4,5 Tavola 4,6

Hayes 79 ? Hayes 80

   

572939 572871

Figura 4,8

n.d. (fondo di ciotola)

 /

n.d. (fondi piani)

bollo a rosetta 572928 stilizzata o asterisco entro cartiglio circolare - diam. 0,7 cm    

Çandarli ware Tavola 5,1 Tavola 5,2 Eastern Sigillata ? Figura 4,10

  L 19 L 19   n.d. (piatto)

        bollo CHRYSIIROS entro cartiglio rettangolare con lati brevi “a coda di rondine” cartiglio: 2,8 x 0,4 cm

  572945 572946   573268

  572861, 572862, 572856, 572857, 572858, 572859, 572860     572868, 572869, 572870, 572872, 572873, 572874, 572875, 572876, 572877, 572878, 572879 + (probabili) 572932, 572933, 572934, 574612

  572831 (con banda di doppia immersione), 572937, 572941, 572942          

Tav. VI, 20); al momento non se ne conoscerebbero altri esempi ad Aquileia. Desta maggiori incertezze la classificazione del fondo di una bottiglia o brocca, più probabilmente la seconda (Tavola 1: 3), nel tipo Hayes 104B, per il quale viene indicato il parallelismo alla serie antica di epoca romana (Atlante II: 43-44 e tav. IX, 5), mentre rinvia forse a una brocca Hayes 105, ancora assegnata alla stessa serie ellenistica e sue derivazioni, una spalla con attacco di ansa (Atlante II: 44 e tav. IX, 6). Non pare trovare un riferimento diretto nella tipologia codificata – e viene quindi inserito dubitativamente nella ES A – l’unico orlo conservato, pertinente a una brocca con orlo everso esternamente modanato, collo cilindrico e stretto ed ansa a gomito con fitte solcature esterne, impostata sulla spalla carenata (Figura 3).

diagnostici sia dai frammenti indistinti (tre pareti, rispettivamente di forma chiusa, di coppa e di piatto).

Benché assenti nello scavo di Canale Anfora del 20042005 (Maselli Scotti 2017: 109), bottiglie o brocche, di forme diverse da quelle sopra proposte o spesso non identificate, erano comunque già state rinvenute in alcuni contesti aquileiesi (Maggi 2006: 186, Appendice 1 con bibliografia precedente).

Due coppe di cui si conserva solo il fondo (Tavola 1: 1-2 – il secondo con tre ditate sull’esterno del piede) sono tuttavia inquadrabili con una certa attendibilità nella forma Hayes 51, che si colloca alla fine della serie romana del primo periodo (70-120 d.C.) (Atlante II: 37 e

Rientrano tutti agevolmente nelle forme standardizzate i 125 frammenti classificabili della ES B, che si distribuiscono fra 14 tipi, di cui due relativi alla produzione B1 (assente nello scavo del 2004-2005, cfr. Maselli Scotti 2017: 109) e i restanti alla B2.

Figura 3. Aquileia, Canale Anfora — scavo 1988. Brocca verosimilmente pertinente alla Eastern Sigillata A (photo: P. Maggi).

243

Paola Maggi, Paola Ventura

Tavola 1. Aquileia, Canale Anfora — scavo 1988: Eastern Sigillata A (1-3); Eastern Sigillata B1 (4-6); Eastern Sigillata B2 (7-8). Disegni di A. Comar (scala 1:2).

244

Canale Anfora, recupero 1988: contributo alla conoscenza della ceramica fine orientale ad Aquileia

Tavola 2. Aquileia, Canale Anfora — scavo 1988: Eastern Sigillata B2. Disegni di A. Comar (scala 1:2).

245

Paola Maggi, Paola Ventura Nella serie più antica si è individuato un esemplare (Tavola 1: 4) della forma Hayes 19, prototipo risalente alla metà del I secolo d.C. della poi preponderante Hayes 60 (Atlante II: 57 e tav. XII, 13), al momento apparentemente non riscontrata (ma non si può escludere un inquadramento di esemplari analoghi nella più tarda evoluzione in ES B2). A questo piatto si affiancano due coppe frammentarie (Tavola 1: 5-6) del tipo Hayes 30, che ancora riflettono nella carenatura e nella decorazione a rotella il prototipo aretino, collocandosi pertanto agli inizi del I secolo d.C. (Atlante II: 59 e tav. XII, 28). Novità per Canale Anfora, gli si affianca ora ad Aquileia un esemplare analogo dai fondi Cossar (Dobreva e Griggio 2021: 204 e tav. III, 1).

con bollo), entrambi databili in età flavio/traianea.12 È già stata sottolineata la coincidenza anche con le presenze nelle collezioni del Museo (Ventura e Maggi 2020: 165 e fig. 3, 6-7), mentre ai fondi Cossar il primo è riconosciuto dubitativamente (a fronte dell’attestazione della più rara variante Hayes 62B), accanto al relativamente comune Hayes 63 (Dobreva e Griggio 2021: 213-214 e tav. V, 3-6). Era pure noto dallo scavo più recente (Maselli Scotti 2017: 169 e fig. 5) un esemplare della ciotola Hayes 70 (Atlante II: 66-67, tavv. XIV, 19 e XV, 1), ora incrementato da quattro pezzi, privi dell’orlo ma assegnabili con buona verosimiglianza alla forma (Tavola 3: 3-4 e Figura 4: 2 e 7 – tre con marchio sul fondo); in assenza della parte superiore non è possibile specificarne – nell’ambito dell’arco cronologico compreso fra 50/75 e 125 d.C. – la pertinenza alla fase precoce o tarda, diversamente dal contesto dei fondi Cossar, ove la prima è stata riconosciuta come lievemente prevalente sulla seconda (Dobreva e Griggio 2021: 211-212 e tav. V, 8-11).

Passando alla ES B2, si sono riconosciuti tre esemplari (Tavola 1: 7-8 – il secondo di grande formato) del piatto Hayes 55, databile fra 70/75 e 150 d.C. (Atlante II: 62 e tav. XII, 18-19): già presente ad Aquileia, anche questa forma viene incrementata da un frammento dai fondi Cossar (Dobreva e Griggio 2021: 201-211 e tav. III, 8). Ben 39 esemplari (e, dubitativamente, ulteriori quattro) appartengono al piatto Hayes 60 (Atlante II: 64 e tav. XIV, 5-8): si sono distinti la variante più antica (50/60 - 80/90 d.C.), rientrante in un formato medio (Tavola 2: 1), alcuni frammenti inquadrabili nella forma di passaggio alla variante tarda, nel duplice formato grande e piccolo, ed infine i più numerosi pezzi nella variante tarda (80/90 - 150 d.C.), con la suddivisione fra piatti di dimensioni grandi (Tavola 2: 2 – con bollo anepigrafe, v. infra)10 e piccole (Tavola 2: 3-4).11 Tale discrimine è stato possibile anche per un fondo con bollo (Figura 4: 1), di sicura assegnazione alla variante tarda di grande formato, mentre i restanti sono attribuiti genericamente alla forma, in alcuni casi in maniera incerta. Questo piatto era il secondo, per attestazioni, nello scavo del 20042005 (Maselli Scotti 2017: 110 e fig. 3 – 35 esemplari, di poco sopravanzato dalla ciotola Hayes 75), ma resta il più frequente in tutta l’area friulana ed alto-adriatica, come confermato dalle collezioni del Museo di Aquileia (Ventura e Maggi 2020: 165 e fig. 3, 3-5) e, da ultimo, dai rinvenimenti al fondo Cossar (13 esemplari, sia nella variante antica che in quella tarda – cfr. Dobreva e Griggio 2021: 212-213 e tavv. IV, 3-9, V,1-2).

La ciotola Hayes 71, con due attestazioni sicure, una delle quali – con bollo – certamente del formato maggiore (Tavola 3: 5 e Figura 4: 5), e due fondi di probabile assegnazione, uno parimenti bollato (Figura 4: 6), non aveva invece precedenti nel sito; compariva però nelle collezioni del Museo (Ventura e Maggi 2020: 165 e fig. 3, 8-9) e torna in maggiori quantità ai fondi Cossar (Dobreva e Griggio 2021: 210, tavv. V, 12-13 e VI, 1-2). I marchi con rosetta fanno propendere per una datazione flavio-traianea (Atlante II: 67 e tav. XV, 2). L’unico bollo a lettere greche ben inquadrabile (cfr. infra) compare su una delle due ciotole di maggior formato del tipo Hayes 74 (Tavola 3: 6 e Figura 4: 9), nella variante A con orlo liscio – come già visto per i corrispondenti piatti Hayes 62 – (cfr. Atlante II: 68 e tav. XV, 6); a questa si aggiungono tre frammenti del formato più piccolo (uno, con vernice nera e marchio illeggibile – Tavola 3: 7), che ci riportano ancora ad epoca flavio-traianea. Sono costanti i riferimenti, con quantità comparabili, allo scavo 2004-2005 (Maselli Scotti 2017: 110 e fig. 6, 1-3), alle collezioni del Museo (Ventura e Maggi 2020: 165 e fig. 3, 10) ed ai fondi Cossar (Dobreva e Griggio 2021: 213-215 e tav. VI, 3-7).

Rappresentano delle conferme, rispetto a quanto già attestato nel contesto indagato nel 2004-2005 (Maselli Scotti 2017: 110 e fig. 4), il piatto con orlo ripiegato Hayes 62A (che identifica la variante con orlo non decorato: Atlante II: 65 e tav. XIV, 10), presente con due esemplari (Tavola 2: 5), e il piatto con orlo incurvato all’esterno Hayes 63 (Atlante II: 65 e tav. XIV, 13), con cinque ricorrenze (Tavola 3: 1-2 e Figura 4: 4 – il primo

Riflette tendenze distributive già note la relativamente maggiore incidenza della ciotola Hayes 75, con un totale di 11 esemplari, in due casi (Tavola 4: 1 e Figura 4: 3) recanti un marchio, mentre un altro fondo resta di dubbia assegnazione fra i tipi Hayes 74 o Hayes 75. Questa seconda forma (Atlante II: 68 e tav. XV, 9), correlabile al già incontrato piatto Hayes 63, ne replica la cronologia fra 70/75 e 120 d.C. Con 38 individui si tratta del vaso maggiormente documentato nello scavo

Si sono considerati i diametri uguali o maggiori di 20 cm. Diametri da 12 a 19 cm. L’estrema frammentarietà non consente di definire il formato degli ulteriori esemplari. 10  11 

12 

246

Un fondo resta di dubbia attribuzione all’uno o all’altro tipo.

Canale Anfora, recupero 1988: contributo alla conoscenza della ceramica fine orientale ad Aquileia

Tavola 3. Aquileia, Canale Anfora — scavo 1988: Eastern Sigillata B2. Disegni di A. Comar (scala 1:2).

247

Paola Maggi, Paola Ventura

Tavola 4. Aquileia, Canale Anfora — scavo 1988: Eastern Sigillata B2. Disegni di A. Comar (scala 1:2).

248

Canale Anfora, recupero 1988: contributo alla conoscenza della ceramica fine orientale ad Aquileia

Figura 4. Aquileia, Canale Anfora — scavo 1988. Alcuni dei marchi documentati sulla Eastern Sigillata B (1-9) e su vasellame di produzione orientale non identificata (10) (photo: P. Maggi).

del Canale Anfora 2004-2005 (Maselli Scotti 2017: 110 e fig. 6, 4-5) e della forma più attestata nel campione (per quanto limitato) del Museo Archeologico (Ventura e Maggi 2020: 165 e fig. 3, 11-12), mentre non spicca in maniera così marcata ai fondi Cossar (Dobreva e Griggio 2021: 2013-2016 e tav. VI, 8-11).

Un solo esemplare (Tavola 4: 4) documenta rispettivamente il grande piatto Hayes 77 (Atlante II: 69 e tav. XV, 12), con cronologia simile alla appena menzionata forma 76B, e – se il riconoscimento è corretto (Tavola 4: 5) – la ciotola Hayes 79 (Atlante II: 69 e tav. XV, 14), di poco antecedente (fine I - inizio II secolo d.C.). Solo la seconda era già segnalata, con tre esemplari, nel contesto scavato nel 2004-2005 (Maselli Scotti 2017: 112), mentre nessuna delle due forme è presente nei complessi del Museo Archeologico e dei fondi Cossar ed anzi la prima (assente anche in Maggi 2006: Appendice 1, che censiva tutto il pregresso) parrebbe riconosciuta per la prima volta in regione.

Una ancor più stretta coerenza con il materiale dallo scavo 2004-2005 è rilevabile riguardo alla ciotola Hayes 76 (Atlante II: 68-69 e tav. XV, 10-11), presente sia nella variante minore 76A, con un esemplare (Tavola 4: 2) che non sembra però possibile discriminare con maggior precisione per cronologia, sia in quella di dimensioni più grandi 76B, a cui appartengono altri sette pezzi (Tavola 4: 3). Nel sito erano infatti già in precedenza segnalate entrambe e con proporzioni simili (tre a fronte di sette – Maselli Scotti 2017: 112 e fig. 7, 1-3): qui pure si rilevava la seriorità della seconda variante (prima metà II secolo d.C.), che resta invece l’unica emersa ai fondi Cossar (Dobreva e Griggio 2021: 216 e tav. VI, 12-15).

A conclusione della rassegna dei tipi, la ciotola Hayes 80 (Atlante II: 69-70 e tav. XV, 15), riconoscibile (ove sufficientemente conservata) solo nel formato maggiore (Tavola 4: 6), annovera 12 esemplari, accanto a quattro fondi di quasi certa attribuzione, ponendosi nella fase finale della produzione e diffusione della ES B (80-150 d.C.). Si tratta della terza forma in assoluto più 249

Paola Maggi, Paola Ventura

Tavola 5. Aquileia, Canale Anfora — scavo 1988: Eastern Sigillata C (1-2); ceramica corinzia (3); ceramica cnidia a matrice (4-6). Disegni di A. Comar (scala 1:2).

250

Canale Anfora, recupero 1988: contributo alla conoscenza della ceramica fine orientale ad Aquileia

attestata nello scavo 2004-2005 (Maselli Scotti 2017: 110 e fig. 8, 1-3), presente anche nelle collezioni storiche del Museo (Ventura e Maggi 2020: 165) e in misura limitata ai fondi Cossar, qui in controtendenza rispetto alla capillare distribuzione che la contraddistingue (Dobreva e Griggio 2021: 216 e tav. VII, 1-2).

Dubbia è l’attribuzione alla ES B – e più in generale a una delle sigillate prodotte nel Mediterraneo orientale – di un fondo su alto piede ad anello di un probabile piatto che sfugge a un preciso inquadramento morfologico: entro un cartiglio rettangolare allungato, con lati brevi “a coda di rondine”, vi è stampigliato il nome greco Chryseros (Figura 4: 10). Il marchio, con E resa a due tratti verticali, sembra privo di confronti (Donat et al. 2022).

Dall’analisi morfologica restano esclusi solamente cinque fondi: uno, caratterizzato dalla presenza di un bollo (Figura 4: 8: v. infra) appartiene a una ciotola, mentre almeno due sono sicuramente relativi a piatti, uno dei quali reca in evidenza la banda di doppia immersione.

Ceramiche decorate a matrice Le importazioni di ceramica fine dal Mediterraneo orientale sono rappresentate nel contesto studiato anche da vasellame decorato a matrice, seppure in misura nettamente minoritaria rispetto alle stoviglie lisce in Eastern sigillata.14

La ES C (Çandarli) compare con due esemplari della coppa L19 (Tavola 5: 1-2), il primo dei quali con superficie bruna (forse post-deposizionale): entrambi privi del piede – parte più caratterizzante a fini cronologici – e con diametri (11-12 cm) compatibili con tutta la durata della produzione (metà I – inizio II secolo d.C.). Essi sembrano comunque da attribuire alla variante più recente, dell’inizio del II secolo d.C. (Atlante II: 76 e tav. XVII, 6). La classe non era attestata nello scavo 20042005 (Maselli Scotti 2017: 112), né fra i materiali antichi del Museo (Ventura, Maggi 2020: 169);13 compare però fra i reperti del porto fluviale (Maggi e Merlatti 2007, p. 566 e fig. 11c) ed ora anche fra quelli del fondo Cossar, dove conferma il suo ruolo di forma più frequente di questa produzione pergamena in area nord-adriatica (Dobreva e Griggio 2021: 224, tab. 3 e tav. VIII, 1-3).

Lo scavo ha restituito tre frammenti di ceramica corinzia, tutti pertinenti alla coppa/pisside a corpo cilindrico che costituisce la forma preferenziale della produzione (forma Malfitana 1). Il pezzo più completo (Mandruzzato 2008: 307, fig. 2. 1) si riferisce alla parte superiore del vaso e mostra una scena figurata dai contorni poco marcati che denota l’uso di una matrice eccessivamente utilizzata (Figura 5: 1). Nonostante la poca leggibilità dei dettagli, è possibile riconoscere una serie di punzoni caratteristici delle decorazioni con Amazzonomachia o altri episodi di combattimento.15 In basso a sinistra si distinguono le gambe distese verso destra di un guerriero ferito riverso a terra, riferibile a una scena in cui viene soccorso da un compagno (Malfitana 2007: 102, scena A); sopra è raffigurato, in piccole dimensioni, un guerriero nudo colto nell’alto di cadere appoggiandosi a uno scudo rotondo (Malfitana 2007: 103, scena H)16; il resto dello spazio figurato è occupato da un cavaliere dal mantello svolazzante che avanza con impeto verso sinistra, pronto a colpire con la lancia (Malfitana 2007: 103, scena D).17 L’associazione di questi tre punzoni riporta alla sequenza compositiva più frequentemente documentata, che si apre proprio con l’accostamento dei primi due motivi per proseguire con il terzo (Malfitana 2007: 116, 3a sequenza).

P. V.

I bolli Dodici esemplari in ES B sono caratterizzati da un marchio al centro del fondo. Nei casi in cui questo risulta riconoscibile, il repertorio comprende soprattutto bolli anepigrafi conformati a palmetta (Tavole 2: 2 e 4: 1; Figura 4: 1-3), talora schematizzata, oppure a rosetta con numero di petali variabile (Tavola 3: 1, 3, 5; Figura 4: 4-7) o ancora a rosetta talmente stilizzata da somigliare a una stella o a un asterisco (Figura 4: 8). Tali bolli compaiono esclusivamente sui servizi pertinenti alla produzione ES B2 (forme Hayes 60, 63, 70, 71, 75), secondo una prassi comune (Atlante II: 52). Su una ciotola Hayes 74 compare invece il marchio MAP/KOY, di cui si leggono in entrambe le righe solo le prime due lettere iniziali (Tavola 3: 6; Figura 4: 9); esso risulta già attestato ad Aquileia su tre esemplari facenti parte della collezione storica del museo, uno dei quali è riconducibile alla forma Hayes 75 (Ventura e Maggi 2020: 165, con bibliografia precedente).

Il secondo frammento è relativo alla parte inferiore del fregio e alla carenatura modanata che raccorda la parete In gran parte le testimonianze riferibili alle produzione di ceramica orientale a matrice restituite dallo scavo di Canale Anfora 1988 sono state già oggetto di trattazione in Mandruzzato 2008, dove sono state considerate insieme a quelle individuate durante le indagini del 20042005. 15  Per l’identificazione delle scene ci si è basati sulla classificazione elaborata da Malfitana 2007; il tema dell’Amazzonomachia corrisponde al gruppo II del repertorio iconografico identificato dallo studioso. 16  Il piccolo formato deriva dal fatto che si tratta di un motivo con funzione di riempitivo: mai utilizzato da solo, è sempre presente sopra o sotto un altro soggetto, spesso proprio quello A come sul pezzo del Canale Anfora. 17  Va notata, in questo caso, l’assenza dei cipressi che di norma incorniciano il soggetto e fanno parte integrante della raffigurazione. 14 

In quella sede si è espresso anche il dubbio che coppe già identificate come L19 talora possano essere in realtà riconducibili a produzioni tardopadane o padane di età medioimperiale. 13 

251

Paola Maggi, Paola Ventura

Figura 5. Aquileia, Canale Anfora — scavo 1988. Coppe/pissidi corinzie decorate con scene di battaglia (1-2) e frammento di brocca in ceramica cnidia (3). Scala 1:1 (photo: P. Maggi).

al fondo (Figura 5: 2). Della decorazione, anch’essa riconducibile a scene di battaglia, resta solo una piccola porzione. A sinistra si intravvede la zampa piegata di un cavallo rivolto verso destra;18 al centro è presente una figura femminile stante, priva delle gambe, con tunica panneggiata e con il braccio destro piegato all’altezza del petto: sebbene i dettagli non siano nitidi, essa sembra attribuibile al soggetto iconografico costituito da una probabile Amazzone colta di prospetto mentre tiene in braccio un bimbo per proteggerlo (Malfitana 2007: 103, scena G).19 Alla fine della sequenza conservata si osservano le gambe rivolte a destra e la veste a pieghe di un’altra figura che ha un braccio proteso verso il basso per soccorrere un guerriero a terra, del quale si scorge una mano: sembra trattarsi di parte della medesima scena individuata nell’ornato della coppa corinzia precedentemente descritta (Malfitana 2007: 102, scena A).

Nel Canale Anfora sono stati recuperati nel 1988 anche tre reperti riconducibili alle officine ceramiche attive a Cnido tra il tardo II e il III secolo d.C. Due si riferiscono a vasi zoomorfi (Tavola 5: 4-5), uno dei quali è conformato ad ariete (Mandruzzato 2008: 305-306, fig. 1.1 e Mandruzzato 2017: 139, fig. 1.3, con erronea attribuzione al deposito archeologico indagato nel 20042005), mentre l’altro verosimilmente riprende la forma di un leone o altro felino (Mandruzzato 2008: 305-306, fig. 1.1). Entrambi gli esemplari presentano un’ansa ad anello in corrispondenza del dorso dell’animale, dove si trova anche l’imboccatura del vaso, e un piccolo foro aperto nel muso con funzione di versatoio. I soggetti riprodotti risultano frequenti nel repertorio della ceramica cnidia, attestati sia su askoi che su brocche,21 e più in generale nel vasellame a matrice orientale anche di epoca più tarda (Erlich e Foerster 2012: 1 e 7, in particolare figg. 1 e 3). In particolare per i contenitori conformati ad ariete si registrano diversi confronti tra il vasellame cnidio documentato in ambito regionale (Maselli Scotti et al. 2003: 46, tav. V.7 – da Tergeste; Maselli Scotti 2008: 68, tav. IV. 4 – dalla villa di Ronchi nell’agro aquileiese; Ventura e Degrassi 2005: 97 – dalla villa di Lucinico, anch’essa in territorio di Aquileia; Auriemma et al. 2008: 162, tav. VII.109 – dal deposito di Punta dei Cocci presso il Villaggio del Pescatore, legato al grande complesso residenziale di Palazzo d’Attila - Casa Pahor sulla costa tergestina).

Il terzo esemplare è pertinente al fondo e alla parte bassa della parete del vaso ed è caratterizzato, all’altezza della carenatura che segna il passaggio tra vasca e fondo, da una stretta fascia con una fila di bottoncini (Tavola 5: 3). Tale elemento decorativo accessorio risulta utilizzato alla base del fregio solo su alcuni dei sottotipi riconosciuti da D. Malfitana per la forma della coppa/pisside; è attestato in associazione a determinati temi iconografici, corrispondenti a scene di battaglia o dionisiache (Malfitana 2007: 119, tabella 3).20

È infine attribuibile alla ceramica a matrice importata da Cnido un piccolo frammento di parete probabilmente pertinente a una brocca, del cui ornato restano una

L’elemento potrebbe essere pertinente alla scena E (Malfitana 2007: 103) che rappresenta un guerriero nel momento in cui cade dal suo cavallo abbattuto: l’animale è ritratto con le zampe piegate mentre si sta accasciando al suolo. 19  A fianco della figura compare un grande scudo rotondo che in effetti sembrerebbe presente anche sull’esemplare dal Canale Anfora. 20  La decorazione secondaria a ovoli o bottoncini nel punto di carena è attestata sui sottotipi 3, 6, 7 e 10 individuati da Malfitana 2007 per 18 

la forma 1. 21  Numerose testimonianze si annoverano nella collezione del British Museum, consultabile al link https://www.britishmuseum.org/collection.

252

Canale Anfora, recupero 1988: contributo alla conoscenza della ceramica fine orientale ad Aquileia

foglia di vite e due braccia alzate (Tavola 5: 6 e Figura 5: 3).

and Christian Archaeology in Memory of Michele Piccirillo ofm (Studium Biblicum Franciscanum. Collectio Maior 52): 1‒10. Milano: Edizioni Terra Santa. Giovannini, A. e L. Mandruzzato 2017. Produzioni cnidie, in P. Maggi et al. (eds) 2017: 139‒143. Maggi, P. 2006. La distribuzione delle sigillate di produzione orientale nell’arco Adriatico nordorientale: problemi di metodo, in S. Čače, A. Kurilić e F. Tassaux (eds) Le routes de l’Adriatique antique. Géographie et économie. Actes de la Table ronde du 18 au 22 septembre 2001, Zadar (Mémoires 17): 179‒194. Bordeaux: Ausonius. Maggi, P., F. Maselli Scotti, S. Pesavento Mattioli e E. Zulini (eds) 2017. Materiali per Aquileia. Lo scavo di Canale Anfora (2004-2005) (Scavi di Aquileia IV). Trieste: Editreg. Maggi, P. e R. Merlatti. 2007. L’evoluzione delle importazioni ad Aquileia: II. Produzioni italiche e orientali: la ceramica, in G. Cuscito e C. Zaccaria (eds) Aquileia dalle origini alla costituzione del ducato longobardo. Territorio – economia – società (Antichità altoadriatiche 65): 547‒581. Trieste: Editreg. Malfitana, D. 2007. La ceramica “corinzia” decorata a matrice. Tipologia, cronologia ed iconografia di una produzione ceramica greca di età imperiale (Rei Cretariae Romanae Fautorum Acta suppl. 10). Bonn: Rei Cretariae Romanae Fautores. Mandruzzato, L. 2008. Some fine wares from the eastern Mediterranean in Aquileia between the 2nd and the 3rd century AD. Rei Cretariae Romanae Fautorum Acta 40: 305‒308. Mandruzzato L. 2017. Ceramica corinzia, in P. Maggi et al. (eds) 2017: 137-138. Maselli Scotti, F. 2008. La ceramica fine di produzione orientale, in F. Maselli Scotti (ed.) Luoghi di vita rurale. Un percorso che attraversa i secoli: 65‒68. Ronchi dei Legionari: Comune di Ronchi dei Legionari. Maselli Scotti F. 2017. Terra sigillata orientale, in P. Maggi et al. (eds) 2017: 109‒114. Maselli Scotti, F., V. Degrassi e G. Mian 2003. Gli scarichi della domus di Piazza Barbacan a Trieste: un contesto di II - inizi III secolo d.C. Atti e Memorie della Società Istriana di Archeologia e Storia Patria 103: 19‒105. Maselli Scotti, F. e S. Pesavento Mattioli 2017. Canale Anfora nel contesto idrografico e topografico aquileiese, in P. Maggi et al. (eds) 2017: 15‒20. Ventura P. e V. Degrassi 2005. Riesame del materiale ceramico della villa di Lucinico (Gorizia, Italia). Rei Cretariae Fautorum Acta 39: 91‒101. Ventura P. e P. Maggi 2020. Importazioni di ceramiche fini orientali ad Aquileia. Nuovi dati dalle collezioni storiche del Museo Archeologico Nazionale. Rei Cretariae Fautorum Acta 46: 157‒171.

P. M. Bibliografia Atlante II = Atlante delle forme ceramiche II. Ceramica fine romana nel bacino mediterrano (tardo ellenismo e primo impero). Roma: Istituto della Enciclopedia Italiana, 1985. Auriemma, R., V. Degrassi, D. Gaddi, S. Mauro, F. Oriolo e D. Riccobono 2008. Terre di mare: paesaggi costieri dal Timavo alla penisola muggesana, in R. Auriemma e S. Karinja (eds) Terre di mare. L’archeologia dei paesaggi costieri e le variazioni climatiche. Atti del Convegno Internazionale di Studi (Trieste, 8-10 novembre 2007): 75‒211. Trieste - Pirano: Università degli Studi di Trieste - Pomorski muzej “Sergej Mašera” Piran. Bertacchi, L. 1988. Aquileia – Marignane Basse. Aquileia Nostra 59: 371. Bertacchi, L. 2000. Il canale Anfora, in M. Buora e F. Prenc (eds) Canale Anfora. Realtà e prospettive tra storia, archeologia e ambiente. Atti del Convegno (Aquileia – Terzo d’Aquileia, 29 aprile 2000) (Quaderni aquileiesi 6-7): 31‒32. Trieste: Editreg. Bonetto, J., D. Bragagnolo, V. Centola, D. Dobreva, G. Furlan, E. Madrigali, A. Menin e C. Previato 2009. Aquileia (UD). Fondi ex Cossar. Relazione delle ricerche 2009. Notiziario della Soprintendenza per i Beni Archeologici del Friuli Venezia Giulia 4: 134‒168. Dobreva, D. 2012. Studio e analisi di alcuni contesti della domus centrale presso i fondi ex Cossar, in J. Bonetto e M. Salvadori (eds) L’architettura privata ad Aquileia in età romana. Atti del Convegno di Studio (Padova, 21-22 febbraio 2011) (Antenor Quaderni 24): 369‒405. Padova: Padova University Press. Dobreva D. e A. M. Griggio 2021. Le ceramiche fini orientali, in J. Bonetto, S. Mazzocchin e D. Dobreva (eds) Aquileia. Fondi Cossar. 3.3. Tomo 1. I materiali ceramici (Scavi di Aquileia II): 185‒284. Roma: Edizioni Quasar. Dobreva, D. e B. Luise 2017. I materiali dagli scavi di Luisa Bertacchi ai Fondi Cossar, in J. Bonetto e V. Centola (eds) Fondi Cossar. 1. Scavi, ricerche e studi del passato (Scavi di Aquileia II): 165‒231. Roma: Edizioni Quasar. Donat, P., P. Maggi, P. Ventura e E. Zulini 2022. Aquileia, Canale Anfora — 1988’s finds: pottery as a trade marker, in I. Borzić, E. Cirelli, K. Jelinčić Vuković, A. Konestra e I. Ožanić Roguljić (eds) TRADE. Transformations of Adriatic Europe 2nd - 9th century. Proceedings of the conference (Zader 11th-13th February 2016). (in stampa) Erlich, A. e G. Foerster 2012. Zoomorphic vases  of the fourth-sixth centuries from the north of Palestine, in L. D. Chrupcała (ed.) Christ is Here! Studies in Biblical 253

One man’s trash is another man’s treasure. Household waste deposit from Porta Pisana and trends of fine ware consumption within the settlement of Krk (in the last centuries BCE) Marina Ugarković, Ranko Starac Abstract The article offers preliminary insights to ceramic fine wares from a rare urban context of Late Iron Age Kvarner (north-eastern Adriatic) that came to light through archaeological excavations conducted outside the Krk’s city walls (Porta Pisana locality), on the island of Krk. The discovered layers dated between the advanced/late 3rd century BCE and the first decades of the 1st century AD, have been interpreted as refuse layers of household debris, which makes them, for now, a unique regional find. Herein we shall focus on types of discovered fine tableware remains typically described within the well-known classes of Hellenistic (and earlier) pottery (red-figure ware, Gnathia ware, black coated ware, grey slipped ware, red slipped ware and relief ware). The analysis of style, morphology, fabric and context of sherds that once belonged to vessels used for setting the table in Krk’s community of the advanced 3rd-1st c. BCE, with some earlier residues, reveals trends of tableware supply defined by the consumers’ context, the local cultural taste and demand, as well as the associated seaborne connectivity. Through this lens, one can reconstruct continuity and change in the aspects of daily life, i.e. drinking and dining practices from Protohistory into the early Roman times, embedded in micro-regional socio-political circumstances.

Key words Kvarner (NE Adriatic), island/city of Krk, Porta Pisana, household waste, Hellenistic pottery, fine tableware, drinking and dining gear, insularity and connectivity.

Where it all started The island of Krk, the second biggest Adriatic island, is situated within the Kvarner bay, the northernmost inlet of the Eastern Adriatic (Figure 1). Due to its favourable geographic location and morphology, the island has been entangled in a wide local/regional communication network that facilitated the development of first protoand urban centres, while continuously serving as a bridge to the mainland (e.g. Blečić 2007; Blečić Kavur 2014: 19; on ancient communications in Kvarner, Konestra 2016: 13-46 and cited bibliography). The city of Krk (Roman, and probably pre-Roman Curicta, in the Middle Ages known as Veglia), is situated in the south-western part of the island, on a small coastal hill overlooking a protected bay (Figure 1, 2). The settlement belonged to a prehistoric community developed as a hillfort (gradina), probably already in the (Late) Bronze Age, with continuity throughout the Iron Age and protohistory as well as into the Roman period, experiencing its first urban organisation already prior to its formal inclusion within the Roman state and continuing to witness significant urban changes until today, preserving its mostly late medieval / early Modern urban layout (Fiorentin 2001; Bolonić and Žic Rokov 2002; Blečić Kavur 2014: 37 and cited bibliography; Konestra 2016: 94; Makarun 2016a; 2016b). Traditionally, the island of Krk and Kvarner region have in the Iron Age been associated with the

Liburnian cultural circle (Batović 1987a), while recent studies, due to observed differences in material culture in comparison to the ‘classical’ (south) Liburnian area, suggest a separate, Kvarner cultural group (Blečić 2007:110; 2014: 165). In the Roman period, Krk island had two municipia – Curicta and Fulfinum, connected to a network of rural ones (Zelić 1993), while the whole of the Kvarner area was a part of the Roman province of Dalmatia, more precisely of its northernmost region — Liburnia. Numerous archaeological finds, earliest from the Late Bronze Age, originate from excavations outside the city walls, but also within the city, and testify to the continuity of life on this locality during the last three millennia (e.g. on Late Bronze Age finds, Buršić Matijašić 2011; Blečić Kavur 2014; on luxurious find of an Iron Age warrior equipment, Balen Letunić 1992; on different types of Iron Age attire dress remains and adornment, as well as other finds, Blečić Kavur 2015; on finds from Samblich collection, Buršić-Matijašić and Matijašić 2011). Possibly already in the Early Iron Age but latest in protohistory (4th-3rd c. BCE) the gradina was encircled by a fortification wall, a monumental pre-Roman rampart built with massive square shaped blocks laid without mortar (Faber 1963-1965; 1978; 1979; Lipovac 1991), a feature known from different sites along the Eastern Adriatic (Faber 2000 with earlier bibliography; Blečić Kavur 2015: 51-56). A part of the defence system was built in the late Republican period,

Roman Pottery and glass manufactures (Archaeopress 2022): 254–284

One man’s trash is another man’s treasure

Figure 1. The Kvarner region (author: Lujana Paraman).

perhaps as a consequence of the battle at Krk that took place in 49 BCE as a part of the Roman civil war between Cezar and Pompeius (Lipovac 1991: 38 and n. 11 for earlier bibliography; on historical circumstance of this battle, Bilić Dujmušić 2014, for the connection between the battle and fortification see p. 121). Another part was built during late antiquity, by reusing stone blocks of ruined ancient monumental buildings. Finally, the walls were significantly restored between the 12th and 14th c. and completely reconstructed in the late 15th c. (e.g. Bradanović 2016: 28, and cadastre plan from 1821). The remains of the longue dureé necropolis have been preserved on the west side of the city, near the ruined church of Sv. Lovre, in the so-called ‘villa Šinigoj’ area, with evidence spanning from the Iron Age through antiquity and medieval times (Cubich 1874-75; Sticotti 1893, Sticotti and Nowotny 1896; Marchesetti 1924: 126; Polonijo and Bulić 1931; Žic Rokov 1962; Matejčić 1963; Glogović 1989, 6-7, 46-47; Bradanović and Ciković 2013; Makarun 2016c).

located, giving access to one of the main streets of the city (ancient decumanus, Zelić 1993: 12 with bibliography) and the previously reconstructed wall section under the ‘old building of the former kindergarten’ (Figure 2). Such a venture that comprised of four excavated trenches (A-D) was undertaken in an attempt to capture and document the foundations of the fortification, for the purpose of defining the condition of the city walls and consequent planning of their conservation (Starac 2012). As it turned out, the documented remains of foundations of the two trapeze-shaped towers connected to the fortification, same as the upper wall parts, have been constructed in the early Middle Ages and built on loose sandy soil full of organic residues (Starac 2012: 457). Since the excavation has not discovered any earlier constructions, which is opposed to the situation previously uncovered in the western part of the town walls, it has been concluded that in this area the visible and investigated ramparts have been formed as a later attachment to the outside face of an older prehistoric and (early) Roman city wall. Due to the steepness of the slope under the outer face of the eastern wall and the danger of collapsing, only a part of the planed surface was excavated, including a very small surface beneath the bottom of the rampart that nicely illustrated the non-existence of the stone foundation, and that the wall lies, in fact, directly on the soil sediments. Moreover, the almost 4 meters deep documented layers (from the bottom of the ramparts below) were formed by successive deposits of soil, ash, organic waste (animal bones and remains of sea shells and snails) and fragments of pottery (both fine and

In 2011 the Maritime and History Museum of the Croatian Littoral in Rijeka conducted rescue archaeological excavations in the area of the eastern fortification wall of the city of Krk.1 The investigation focused on the outer face of the wall section situated above the Porta Pisana gate — an area where Roman eastern town gates were also The excavation was directed by Ranko Starac, curator of the Prehistoric and Roman collections of the aforementioned museum, see Starac 2012. 1 

255

Marina Ugarković, Ranko Starac

Figure 2. Krk city and position of Porta Pisana (photo: Ranko Starac; elaboration: Ana Konestra).

the inner face of today’s visible early Medieval city wall. Consequently, the evidence of Krk’s household waste deposits formed in a time range between the advanced 3rd century BCE and the first decades of the 1st century CE now represent not only an important but indeed a unique urban context in the Kvarner Gulf, but also in the broader eastern Adriatic region. Krk’s Household Waste: The Evidence of Ceramic Tablewares In pursue of aspects of consumption patterns and localness of Krk’s community in the last centuries BCE, this article will attempt to present the evidence of the fine ceramic wares that have been recovered in Krk’s household waste deposit (Porta Pisana), focusing on a preliminary analysis of the wares that belong to different classes of the so-called Hellenistic pottery, with some earlier residues. This, together with previously analysed early Roman wares (thin-walled ware and terra sigillata) (Konestra 2016; Konestra and Ugarković forthcoming), is a first step towards gaining insights to the full picture illustrated by the material remains recovered within this important and indeed regionally for now unique context. Figure 3. Trench B1, west profile, SU 013, 012, 109, 016, 117 (photo: Ranko Starac).

The ceramic fragments are analysed according to morphological and stylistic traits, as well as macroscopic insights into fabric and surface treatment, and are presented according to wares, shapes and chronology.

coarse wares) and amphorae, including more than 90 amphorae stoppers (Figure 3). These have formed as the result of centuries-long continuous deposition of urban household waste or hearth waste, that was discarded outside the earlier city walls, that enclosed a smaller perimeter and were obviously located in proximity of

South Italian Geometric/Banded ware? Only one wall fragment with a convex profile, that belongs to a closed rounded shape, has preserved 256

One man’s trash is another man’s treasure

Pl. 1

2

3

1

5

4

6

8 7

9

11

10 Over-painted decoration:

red

preserved in negative

white

black

Plate 1. 1. South Italian Geometric/Banded ware; 2. Red figure glaux skyphos; 3-12. Gnathia and other overpainted wares (drawings: Martina Korić; plates: Ana Konestra).

257

12

Marina Ugarković, Ranko Starac elements of its decoration applied directly, without the addition of a slip (Cat. no. 1; Plate 1). The fine fabric is yellow-light brown, without visible inclusions, while several brown-black painted concentric lines of different thickness are visible on its exterior surface. Considering the state of preservation of the fragment, it is difficult to assign it with any certainty to a specific ware and further speculate on its typologicalchronological traits. That said, we can cautiously note that a similar decoration can appear on South Italian Geometric ware, whose Eastern Adriatic occurrences have been dated from the 9-8 to 6-5th centuries BCE (Škoro 2020 and earlier bibliography) and have been discovered in the Kvarner region as well (Blečić Kavur 2015: 197-199). Even though lines of different thickness are also typical of the so-called Banded Ware, alternatively known as ceramica a fasce, and ceramica di stile misto, where they appear with other motives, produced in Greek (from the 7th century BCE) and/or indigenous south Italian centres (from the 6th century BCE) until the 4th-3rd centuries BCE (Riccardi 2007; for eastern Adriatic, Čelhar and Borzić 2016: 76; Ugarković 2019, I: 11-115, and earlier bibliography), the thickness, number, and the depiction of lines on our example seem to be more suggestive of the former ware, which would indicate this sherd as a residual one. However, due to still many uncertainties, with regard to the production of different wares, of which more examples have been found in Krk as indicated below, this suggestion has to remain a tentative one.

in the contexts of Greek and indigenous settlements, necropolises and sanctuaries (Ugarković 2019, I: 85, 86 and bibliography; Kirigin 2020 and bibliography). Among eastern Adriatic sites, more examples of owl skyphoi have been found in Pharos, where also a possibility of their local production has been hypothesized (Katić 1999-2000), while other examples/fragments are known from Nesactium, in the area of Temple B (Mihovilić 2002: 503-504, pl. 2,4; 2004: 107d) as well as sanctuaries discovered on the island of Palagruža (Kirigin 2020: 6365) and Vilina cave (Perkić 2010: 159-161, 206 no. 404).3 Motives depicted on these skyphoi have been linked to religious significance and therefore their distinct use in that context was hypothesised, particularly within the original Greek context (Watson 1999 with cited bibliography). Notwithstanding the variety of contexts where such vessels were discovered that would favour their multifaceted meanings, the overall survey seemed to indicate that glaux skyphoi were in most cases used for domestic purposes (Cook 1940: 787). Although sherd from Krk is too fragmentary to permit safe detailed attribution, its style and fabric, along with the contextual evidence, allow us to hypothesise its possible South Italian provenance and a date of the late 5th or first half of the 4th century BCE. That being said we may further observe some stylistic resemblance in the rendering of the preserved upper part of the owl’s head to examples interpreted as 5th and 4th centuries BCE South Italian products (Johnson 1955: 121, pl. 38, fig. 38, close to Farwell group; 122, 123, pl. 38, fig. 47, from Athens, but defined as of Italian origin; De Juliis 1996, from San Severo; Kušan Špalj and RendićMiočević 2006: 43, from the Archaeological Museum of Zagreb, unknown provenance; Becker et al. 2016: 98, 99 from tomb 8-01 in Sant’Angelo Vecchio dated to late 5th century and attributed to the Pisticci Painter of Metaponto). Similar style was noticed on several examples found in connection to the Greek sanctuary dedicated to Diomedes on the island of Palagruža, for some of which a provenance in Canosa, Northern Apulia has been suggested (Kirigin 2020: 64, fig. 7, 9; 65). Although this appears to be the first example of an owl skyphos found in the Kvarner region (but not in the northern Adriatic, e.g. Nesactium), and the only red-figure fragment discovered at Porta Pisana, several other sherds of tableware conceivably associated also with South Italian red-figure production(s) have previously been noted among the Kvarner insular communities such as at Osor on Cres island (Faber 1980: 298, sl. 11, 4, sl. 12: 2-5; Mihovilić 2002: 504-505; 2004: 108, Fig. 13; Blečić Kavur 2015: 200-202), Kaštelina at Lopar on Rab island (Batović 1987b: 163, Sl. 11: 2),

Red-figure ware The earliest securely identified evidence of imported tableware is a small rim fragment found in trench B (SU 006/012) attributed to red-figure owl skyphos (Cat. no. 2, Plate 1). It depicts a part of the owl’s head, facing front, whereas eyes are indicated with a thick irregular circle, with a central dot, along with a circle of dots on the face, without the eyebrows. There seems to be no evidence of milthos on the reserved area, while fine fabric without visible inclusions is pink in colour. The owl skyphoi, normally depicting this animal standing between two sprays of olives, originated in the Attic workshops, where they reached their heyday in the 5th century BCE (Moore 1997: 64; Watson 1999).2 Local imitations have so far been noted in Corinth, South Italy and Etruria (Johnson 1953; 1955; McPhee 1981: 282; Watson 1999, with cited bibliography; Becker et al. 2016: 98). Glaux skyphoi, typically dated between the 5th and first half of the 4th century BCE, have been recorded on a number of Mediterranean sites, inclusive of both Adriatic coasts, where these were discovered

One whole skyphos of type B is known from grave 218 of the Vlaška Njiva necropolis in Issa on the island of Vis (Ugarković 2019, I, 85, 86; II, 10, 218.2/CP 1). However, due to the state of the preservation of its surface it remains unclear if the skyphos was just black glazed or had a red-figure decoration as well. 3 

In Attic production they appear on both A and B type of skypos, but mostly on the latter, popularity of which resulted in a looser use of term glaux (owl) for such skyphos shape (Johnston 1953). 2 

258

One man’s trash is another man’s treasure

including the island of Krk (Mihovilić 2002: 505). The presence of such tableware in the Kvarner region is hardly a surprise since South Italian red-figure pottery, based on the current state of evidence mostly (but not exclusively) of Late Apulian style, has been noted on a number of sites along the eastern Adriatic coast, including its northern tip (for the latest overview of the current evidence see Ugarković 2019: 58-70 with bibliography).

The number and depiction of these lines, however, differ, from three lines presented in a horizontal manner (two white lines with a red line in between set partially obliquely, Cat. no. 6, Plate 1) to five lines that give impression of being concentric, this time with convexity towards the rim (central thicker red line, flanked with double white lines, Cat. no. 5, Plate 1). Another rim, Cat. no. 8, Plate 1, with a slightly less concave transition to the body, has five concentric lines presented in a somewhat close manner as the latter, but now applied only in white colour. Close in shape, but with two small groves set on the outer lip above the two white painted horizontal lines (Cat. no. 9, Plate 1), another fragment morphologically resembles black coated kantharoi. Both of the last described sherds no longer qualify as Gnathia ware and are described as other (white) overpainted ware.

Gnathia and other overpainted wares Fragments of black coated tableware with traces of overpainted decoration have been documented in several layers (trench B, SU 006/02, SU 016/017; trench B1-D, SU 17; trench D, SU 17). Some of these can clearly be attributed to Gnathia ware, while for others this remains doubtful. Moreover, for many sherds, due to their fragmentary state, discussion regarding their production centre/workshop, remains a speculative one, while in other cases it was difficult even to determine shapes with a high degree of certainty. That being said, the morphology and visible decoration of some of the recovered fragments do offer a possibility for informative insights.

Numerous belly fragments of Gnathia ware are documented as well, some of which show traces of decoration in negative, in one case a combination of painted and incised lines (Cat. no. 11, Plate 1). Other fragments exhibit vertical grooves covering the body, typical of Late Gnathia production, among which a better preserved one shows a red line encircling the double handle while two wavy white lines are visible on the left (Cat. no. 10, Plate 1).

The preservation of black coating on most sherds can be characterized as good/very good, while fine fabric, without visible inclusions, is typically pink to light brown. Several mildly out-turned rims with white and red painted lines below the rim (red one being distinctively thicker), might have belonged to skyphoi, but are too small for estimation of rim diameter (Cat. no. 3, Plate 1; Cat. no. 7, Plate 1). Most frequent, however, are out-turned rims, 9 – 10 cm in width, majority of which seem to belong to (small) jugs. The best preserved jug fragment has a smooth concave transition from the rim to the rounded shape of the body (Cat. no. 4, Plate 1). The visible decoration shows alternation of white and red lines below the rim and on the belly, with a palm branch in between. The set of painted lines (or thin zones), especially those on the body, appear to be concentric, with the convex side towards the bottom of the vessel. Another wall fragment of a jug, bears similar decoration of palm branch (with somewhat diverse depiction of stylized leaves) and simple lines (Cat. no. 12, Plate 1). Due to firing conditions, that is a longer reducing atmosphere in the kiln to which this vessel has been exposed, its fabric and coating turned grey and the decoration black. Such alterations in the colour of the coating have already been noted on some examples of Gnathia ware found in Issa on the island of Vis (Miše 2017: 242; Ugarković 2019: II, 26, 65.2/GN 1).

Although discussing productions based on such small fragments is methodologically challenging at best, there are clear indications some of these might belong to imports from South Italy, more specifically to the latest production phase of this ware conceivably of North Apulian (Daunian, Late Canosan) provenance. A thick red line, bordered with one or two often very thin white lines, like on Cat. no. 3 and 4, is typical for the 3rd century BCE Late Canosan group (Green 2001: 63; Lanza Catti 2008: 157). Some of the wall fragments (e.g. Cat. no. 10) might have belonged to Apulian jugs, but this cannot be said with any certainty. Imported skyphoi, oinochoai and plates of Late Apulian Gnathia, some of which clearly of Late Canosan group, have been previously noted in Krk (Makarun and Novak 2016), as well as elsewhere in the Kvarner region, e.g. Osor or Kastav necropolis (Mihovilić 2002: 507-509; Matejčić 1974: 56, 69, tab. VII; Blečić 2002: 120, 146, Tab. 12; Makarun 2014; Blečić Kavur 2015: 202-209). This comes as no surprise since Late Canosan products seem to be the most common type of imported Gnathia in the eastern Adriatic, with skyphos being the most common shape (Miše 2015: 28, 29). Skyphoi (with horizontal white and red lines) from i.e. Palagruža have been attributed to Canosan or Peucetian workshops (Miše 2017: 241), while analogies in shape and (partial) decoration have been found in tombs of Issa (Cambi et al. 1981: Pl. X. 26; Kirigin and Marin 1985: Pl. VIII. 1, Pl. IX. 9, 11, 16; Miše 2015: no. 48, 50, 53-57, 90-94; Čargo

Other (jug?) fragments with similar out-turned rims and concave transition to the shoulder/body, are also decorated with red and white lines below the rim. 259

Marina Ugarković, Ranko Starac 2010: 117-118, nos. 16-19; Ugarković 2019, I: 79; II: 117). That being said, what merits further observation is a clear absence of certain motives, such as the complete lack of the so-called ‘toothed saw’ which in combination with the horizontal lines otherwise commonly appears on Gnathia skyphoi from i.e. Issa, with examples known from the north Adriatic as well, as indicated by e.g. Kastav finds.

island (Borzić 2017: 75, tomb 0, no. 5). The branch of the Kopila example, also in combination with red and white lines, attributed to Late Canosan production of the late 3rd century BCE, same as on our Cat. no. 4, is indicated with a double line (Borzić 2017: 150, 151; for ramo secco of Apulian production, with a double line as branch Redavid 2010, 133, 134, fig. 31, nos. 4 and 7). Gnathia fragments with palm branches have also been noted on Palagruža island (Miše 2017: 243, fig. 6). In terms of morphology, however, even the best preserved jug fragment from Krk (Cat. no. 4) does not have all the elements of shape to allow us a full comparative analysis. Jugs are certainly known from Daunian and Peucetian Gnathia productions, but these are typically characterized by somewhat different morphological traits and motives, frequently also with bodies covered in vertical ribs (Lanza 2005: 69, 70; Lanza Catti 2011: 266, fig. 1). Small jugs without ribs, also decorated with some kind of branch on the neck, are known from EtruscoLazian production, but these too do not correspond in shape (e. g. Lanza 2005: 118, olpetta; Morel 1981, series 5223a, 5224a, 5226a, 5227a, Pl. 157-159). Besides the above-mentioned jug (small pelike?) from Issa (Vlaška njiva necropolis), that has more emphasized shoulder and a longer neck, the only other Gnathia jug from Issa (that is not oinochoe or pelika), was found in Martvilo necropolis and attributed to 3rd century BCE probable Daunian or Etruscan workshops (Miše 2015: 110). It has a similar outline of the body, at least with regard to preserved part of the Krk example, and while here too red and white lines appear below the rim and on the belly, an ivy branch decorates the area in between. However, the closest analogies to our Krk example, in terms of both form and decoration, come from several other eastern Adriatic sites, including its hinterland. Small jugs with everted rims and rounded shoulders, decorated with a simple motif of a palm branch appear in Southern Liburnia, namely in Asseria (Govorčin and Borzić 2018: tab. 2, 1, 3), while another very close analogy was found in Trogir (personal inspection of the material). Quite similar jug, at least with regard to preserved parts, has been deposited in a tomb of Prozor necropolis in Lika, dated to the 2nd century BCE (Balen Letunić 1996: 28, Tab. 2, 4).

For now, there are no clear elements that would allow to discuss the potential presence of Gnathia from South Apulian workshops or Gnathia that could be associated earlier, Middle Apulian production, in spite of the fact that shreds attributed to late 4th c. BCE Dunedin and Sidewinder groups of the Middle Gnathian Taras production, as well as a possible late 4th/early 3rd c. BCE Messapian production have been recorded in nearby Osor (Faber 1980: 309, fig. 13, 1, 3; Mihovilić 2002: 507, fig. 5, 2; Blečić Kavur 2015: 202-205; Miše 2015: 23, 24, 26 and earlier bibliography) and other localities outside Kvarner (Miše 2015: 23-24 and earlier bibliography). The presented picture could be to a great extent a result of chronology of this particular context that would favour a more advanced and late 3rd century BCE as the earliest date of most finds (even though some earlier examples are clearly documented as well), although one has to keep in mind the fragmentarity of the material record which does not allow us to make definite conclusive statements. On the other hand, some sherds speak in favour of a probable regional provenance of Gnathia ware with simpler decoration made under the influence of the Late Canosan group, whose typical traits, along with horizontal red and white lines include i.e. different varieties of the ramo secco motif (Green 2001: 62, 63; Redavid 2010: 133, fig. 31; Miše 2017: 244). Some of these motifs have also been linked to a slightly earlier north Apulian production, the Knudsen group (middle production phase, late 4th-early 3rd century BCE) or the transition from this group to the Late Apulian production (Miše 2017: 250). The motif of a palm branch noticed on our jugs (Cat. nos. 4 and 12) has been previously attested, in combination with other motives, on south Italian Gnathia jugs found in the eastern Adriatic, i.e. in tombs of Issa (for oinochoai from Martvilo necropolis: Cambi et al. 1980: T. X, 1, tomb 3/1976; Čargo 2010: 76; Miše 2015: 68, no. 6; 69, no. 8, both with horizontal lines and alternating grapes and tendrils below, attributed to Canosan production of the late 4th-first half of the 3rd century BCE, and to a transitional group from Knudsen to Late Canosan group; Miše 2015: 79, no. 26, end of 3rd century BCE, probably Apulian with West slope influence; for small jug (pelike?) from Vlaška Njiva necropolis: Ugarković 2019, I: 80; II: 27, contextually dated to the late 4th and early 3rd centuries BCE) or Kopila hillfort on Korčula

In a similar light we can observe some of the other fragments with only red and white (Cat. nos. 5 and 6) or just white lines (Cat. nos. 8 and 9), aligned in a horizontal or slightly concentric way. Examples similar in shape, size and decoration can be found in Asseria (Govorčin and Borzić 2018: 42, T 2:4-7; III: 1-4), Cape Ploča (Šešelj 2010: 68, 69, Cat. no. 18, with two white horizontal lines) or Trogir (Kalebota 2015: 64, 65, Cat. no. 9; 82-84, Cat. nos. 30-32). Therefore, as recently noted in the case study of pottery found in Asseria (Govorčin and Borzić 2018: 42), 260

One man’s trash is another man’s treasure

the material from Krk further illustrates the emerging evidence of a group of jugs of the advanced 3rd-2nd century BCE, made in the tradition of Gnathia ware and decorated under the influence of the Late Canosan style, but yet with somewhat different morphological and decoration traits, that could be considered suggestive of a regional provenance. That is followed by examples of morphologically similar 2nd (early 1st?) century BCE tableware, now with white overpainted decoration. Issa as currently the only known eastern Adriatic workshop that manufactured Gnathia ware and ware made in the Gnathia tradition, has so far not provided evidence of similar tableware. That could be taken as an indicator that a search for provenance should focus elsewhere, however, one also has to keep in mind that Issaean Gnathia is best known from the tombs of Issa, where shapes must have been defined by the local burial rite, while the evidence from the settlement is lacking. Weather Issaean potters also produced a different type of merchandise for other everyday activities or, perhaps, export, following the demand of other markets, is not something that should be dismissed as an unlikely possibility, but remains difficult to discuss based on the current state of evidence. While definitive answer to the question of provenance has to, for now, remain an open one, one has to note that during the advanced 2nd and 1st centuries BCE the local ceramic workshop in Issa did also produce ceramic shapes (oinochoai, pelikai and skyphoi, again known mainly from tombs) decorated with simple white painted decoration (Miše 2015: 37; Ugarković 2019, I: 98). In such a way it was following some of the global trends, since overpainted white decoration, i.e. in form of single or double lines below the rim, is known in other productions of black coated ware, for example in the Campana A production at the beginning of the 2nd century BCE, where painted circles on the floor of bowls can also appear form the mid- 2nd until the early 1st century BCE (Reynolds Scott 2007: 7).

Corinthian c. 225-170/160 BCE production; Gamberini 2016: 249, tav. 4, 49, for Phoinike, 2nd-first half of the 1st century BCE tomb; Morel 1981: 263, pl. 95, series 3453 for 2nd century BCE central Italian production in Nursia), it seems to have been especially esteemed in the eastern Adriatic, most notably its southern and central regions, where productions, in Apollonia-Dyrrachium and Issa-Resnik areas have been hypothesised (e.g. Šegvić et al. 2012). This has been corroborated by numerous examples discovered in settlements, necropolises and sanctuaries, and now further north in Krk, also with slightly later grey slipped and red slipped variants (see below) (generally on this shape in Dalmatia, Šešelj 2005: 56-59; 2008: 105, 108, 111; Jelušić et al. 2007: 19, n. 13; Šešelj 2010: 137-146; Tonc 2013: 116-120; Govorčin and Borzić 2017: 78, 152, 153; Borzić 2017: 49-51, tab. IV, 4-7 n. 22, V, 7; Ugarković and Konestra 2018: 75, pl. 1. 18; Ugarković 2019, I: 89-90; for much fewer occurrences of the shape in the opposite side of the Adriatic see e.g. De Mitri 2016: 30, 31, fig. 9). The black to brown coated articulated kantharoi should be generally dated in the 2nd century BCE (perhaps from the advanced/late 3rd century BCE) and its presence and numbers in the Krk deposits now confirm the popularity of this type of a drinking cup also in the Kvarner region, or at least at this locality. Another conceivable type of kantharos has a straight but slightly inward inclining upper wall, ending with a short diagonally out-turned rim (Cat. no. 17, Plate 2). As such it bears resemblance to 3rd century BCE cyma kantharoi from e.g. Phonikie (Gamberini 2016: 55-57, especially nos. 8 and 16, tav. 1 and 2), where a similar rim can at times also appear on the 2nd century BCE articulated kantharoi (Gamberini 179, no. 31, tav. 3, from tomb of the 2nd half of the 2 century BCE). The other common type of drinking vessel is the mastos cup, characterized by a hemispherical to slightly conical shape of the body, mostly with mildly convex walls and typically with an outer groove below the rim (Cat. nos. 20-25, Plate 2). The rim diameter in substantially preserved examples is ca. 11-12 cm. Previous occurrences of this shape in the eastern Adriatic have been noticed mostly in central Dalmatia, e.g. Stobreč (Ugarković and Konestra 2020: 50, Pl. 1, 5), Cape Ploča (Šešelj 2010: 112-114, in grey ware), Pakleni islands (Ugarković and Konestra 2018: 74, Pl. 1, 4, direct analogy to our Cat. no. 21), while a similar fragment is known from Asseria in south Liburnia too (Govorčin and Borzić 2018: tab. V, 1). Some morphological differences among these examples can be observed, but for now, it remains unclear if these should be considered workshop indicative and if mastoi were also produced regionally, or imported from e.g. the opposite shore (e.g. Aesis, The Marches, Brecciaroli Taborelli 1996-1997: 90, F 2570; 150, no. 272) or Apulia. The latter manufacture of black

Black coated ware A number of layers have produced fragments of black coated tableware that bear no traces of decoration and refer to productions dated to the advanced/late 3rd - early 1st centuries BCE. Among drinking cups, kantharoi and mastoid cups seem to be the choice of preference. The former are mostly characterised by an out-turned rounded rim (Cat. nos. 13-16, Plate 2), 7-9 cm in diameter, sometimes slightly thickened (Cat. no. 16), one or two grooves below the lip and concave upper wall. Two body fragments (Cat. nos. 18-19, Plate 2) show that at least some of these belonged to articulated kantharoi, with grooves below and/or above their widest part. Although the articulated kantharos is a form of drinking cup known from several central Mediterranean productions (e.g. James 2018: 86-87 for 261

Marina Ugarković, Ranko Starac

Plate 2. 13-26 Black-coated ware (drawings: Martina Korić; plates: Ana Konestra).

262

One man’s trash is another man’s treasure

coated variety of mastoid cups started in the late 3rd and flourished throughout the 2nd century BCE, while the form itself, in grey slipped grey ware productions, commonly appeared until the beginning of the 1st century BCE (Yntema 2005: 64, 65; De Mitri 2016: 30; De Mitri and Mauro 2019: 286, 287).

2nd centuries BCE (Brecciaroli Taborelli 1996-1997: 125127, especially no. 107). One of the most common bowls of Hellenistic pottery, the open plain bowl with incurving rim, so-called echinus bowl, is represented in Krk deposit with only one small example (Cat. no. 35, Plate 3, rim diameter 12 cm). Although the shape was continuously used in the Greek circle since the Classical Period, the deep body as seen on Cat. no. 35 is a typical trait of its Hellenistic variant (e.g. Prag 1992a, 106-107, especially no. 829 dated to late 3nd and 2nd centuries BCE). At that point, such bowl type was produced in a number of Mediterranean workshops and commonly used in the Ionian/Adriatic area (for different versions of incurving rim bowls Morel series 2700 general, closest 2812d, 2nd century BCE, pl. 75 because of the straight lower part of the vessel; for a late, Augustan, local variant of a central Italian production, Aesis worksop, Brecciaroli Taborelli 1996-1997: 156, F 2760; De Mitri 2016: 30; De Mitri and Mauro 2019: 286). In the eastern Adriatic, until recently several examples have been documented only in Greek (or Greek affiliated) settlements (for Pharos: Jeličić Radonić 1995: 107, no. 1; Issa: Ugarković 2019, I: 94, classical type; Trogir: Kovačić 2002: 395, fig. 18, right, first three from above; for Epetion: Ugarković and Konestra 2020: 50, Pl. 1, 6,) and sanctuaries (Šešelj 2010: 85-86), but such situation is slowly changing as shown by recent publications that indicate its wider spread among different local communities (e.g. Pakleni islands: Ugarković and Konestra 2018: 75, 76; Asseria: Govorčin and Borzić 2018: 48, 49, tab. V, 2).

Bowls are represented with several main types and their varieties of different sizes. The best preserved fragment belongs to a deep hemispherical bowl with a rounded and slightly thickened rim (diameter c. 18 cm) with a triple outer groove on the exterior (Cat. no. 26, Plate 2). Similar profile appears briefly in early Hellenistic Attic workshops, on the so-called bolster cups, while outer triple grooves are also present in different 3rd and 2nd centuries BCE productions of hemispherical cups and bowls (e.g. Athens: Rotroff 1997: 96, 97, 107-109; Phoinike: Gamberini 2016: 180, tav. 5, 58, first half of 3rd C. BCE; Canosa: Morel 1981: 142, Pl. 32, 2155a1, late 3rd century BCE; Etruria: Morel 1981: 186, pl. 57, 2573a 1, first half of the 2nd century BCE; central Italy, Adriatic, Aesis: Brecciaroli Taborelli 1996-1997: 145, 147 F 2560, 2570, later 3rd and 2nd centuries BCE). Other fragments of deep hemispherical bowls share a similar plain rim, but without the grooves (Cat. nos. 28, 29, Plate 3, the former slightly thickened on the inner side) (for examples in Apulian production Prag 1992a: 87, 88; for Cat. no. 29; 1992a: 108-110, especially no. 839 dated to the 2nd-1st centuries BCE; for different productions Morel 1981, pl. 79-83, series 2900; Dobreva and Griggio 2011: 83, Lamboglia 27 form).

Another common shape in the repertoire of Hellenistic pottery was a bowl with out-turned rim (Rotroff 1997: 156-161). Besides having a long manufacture tradition, during Hellenistic times it was produced in various Mediterranean workshops (De Mitri 2016: 30). Based on morphological features, three varieties can be noticed in Krk’s deposit. Two examples belong to small hemispherical bowls with a more convex wall profile and short diagonally outturned rim of 11-12 cm in diameter (Cat. nos. 30 and 31, Plate 3). Somewhat close variant of the 3rd century BCE central and southern Italian production is documented in Morel series 1551b (Morel 1981: 138, pl. 31).

Conical bowls with rounded rim and grooved inner walls, like our Cat. no. 27 (Plate 3, rim diameter 18 cm) are known from the end of the 4th century BCE and appear commonly throughout the Hellenistic period, both black and grey coated (for e.g. Apulian black coated production, Prag 1992a: 82-84; see also Morel general 2150 series, with closest analogy to 2151, Morel 1981: 141, pl. 32; and series 2972, Morel 1981: 241, 242, pl. 82, late 3rd and 2nd centuries BCE). In Attic ceramic production bowls with grooved inner walls appear on parabolic mastoid cups of the first quarter of the 2nd century BCE as well as in type 1 of the cup with interior decoration of the 3rd and less commonly 2nd centuries BCE (Rotroff 1997: 109-113). A relatively short production, ca. 225– 150 BCE, of this shape is also known in Corinth (James 2018: 110-112). Manufacture of a similar shape, known as coppa troncoconicha dated mostly in the 3rd century BCE, appears in Phoinike (Gamberini 2016: 66-69), while such a form, but without the inner grooves is known from a local 2nd century BCE production in northern Italy at Aquileia (Mandruzzato and Masselli Scotti 2003: 383, tav. 1, 5). The most direct analogy to our example can be found in the central Italian products from Aesis, dated to the second half of the 3rd and first half of the

The remaining two fragments belong to different varieties, known also as Form 28 of Campana A (Lamboglia 1952: 177, 178) and Morel series 2640 and 2650. The main difference is in the morphology of their rims. The first example (Cat. no. 33, Plate 3) has a flat, horizontally projecting rim with parallel upper and lower surfaces, same as the plate commonly produced in Apulia during the 3rd century BCE (but also later) (Prag 1992a: 81, 82). The second variety, present with one fragment, exhibits a slightly downturned and shortly out-turned rounded rim, with the flat upper wall (Cat. no. 34, Plate 3). 263

Marina Ugarković, Ranko Starac

Pl. 3

27

28

29

30

31

32

33

34

35

36

37

Plate 3. 27-37 Black-coated ware (drawings: Martina Korić; plates: Ana Konestra).

264

One man’s trash is another man’s treasure

Pl. 4 38

39

40 41

43 42

44 46

45

Plate 4. 38-40 Black-coated ware; 41-46 Red-slipped ware (drawings: Martina Korić; plates: Ana Konestra).

265

Marina Ugarković, Ranko Starac Analogous shape has been recorded in central Dalmatia, e.g. Pakleni islands (Ugarković and Konestra 2018: 76, pl. 1, 7), Issa (Ugarković 2019, I: 94, II: 49, 90.1, CP 55) and Trogir (Kovačić 2002: 395, fig. 18, left up). Close examples are attested in a number of different 3rd and 2nd centuries BCE black coated productions (Morel 1981: 198, pl, 62, close to series 2642, late 2nd century BCE, Northern Italy; Rotroff 1997: fig. 59-60; De Mitri and Mauro 2019: 287, fig. 2, 13; Gamberini 2016: 109111, especially tav. 34, 328). Both varieties have been produced and frequently attested also in northern Italy in late 2nd and 1st century BCE (Dobreva and Griggio 2011: 83, tav. 1, 6; 2.1 and cited bibliography), while a central Italian production, ranging from 150 BCE to Augustan times, is known too (Brecciaroli Taborelli 1996-1997: 91, 95, F 2273).

and during the first half of the 1st centuries BCE (Rotroff 1997: 155) it was much more popular in different Italic productions (Lamboglia 1952: form 5, Camapana A; Morel series 2250-2280, Morel 1981: 152.163, pls. 3946; Mandruzatto and Masseli Scotti 2003: tav. 2, 5 and 7, for possible Alto-Adriatico production; for Phoinike: Gamberini 2016: 107, 108, 2nd and 1st century BCE, and therein cited analogies; Apulia: Prag 1992a, 88-89, so-called lid-plates; north Italian production: Dobreva and Griggio 2011: 85-877, Lamboglia 5 form, and cited bibliography; for a central Italian production, Aesis worksop, dated between 120/110 - 50/40 BCE: Brecciaroli Taborelli 1996-1997: 94 F F 2283) where it was also manufactured in grey ware and later Roman wares (Prag 1992a: 88). Both of the Krk examples are not preserved enough to assert their production centre, although the rim of Cat. no. 38 seems to have most analogies with the 1st century BCE northern Italian and Etruscan productions. For a geographically closest black coated version of the same type found in Jadranovo, northern Italian origin has been assumed (Konestra 2015: 118, fig. 3, 1-3). Other occurrences are known from Dalmatia too, e.g. Trogir (Kovačić 2002: 395, left, middle), although in the eastern Adriatic it is more represented in its grey slipped variant (see below).

One fragment of a very deep vessel, with a slightly thickened rim and a groove on its exterior, has a fairly vertical although not straight upper wall (Cat. no. 32, Plate 3). Even if such walls if not its size are more typical of skyphoi, they can appear on bowls as well (e.g. Prag 1992a: 107, 292, Cat. no. 836), including the central Italian (Adriatic) production of c. 270/150 BCE (Brecciaroli Taborelli 1996-1997: 150-153, closest analogy no. 272 on fig. 79).

Several handles, including vertical handle with an elongated profile and a pronounced middle rib (Cat. no. 40, Plate 4) belong to jugs.

Three types are represented among the serving dishes/ plates, one of which (Cat. no. 36, Plate 3) is part of a deep dish with an upturned and inward rolled rim. Some shape analogies could be found on dishes from Phoinike, e.g. piattelli con basso orlo (Gamberini 2016: 100-101) dated to 3rd and early 2nd centuries BCE, as well as plates with inward rolled rim popular in Apulia from the later 4th until 2nd centuries BCE (with some appearing also in 1st century BCE) (Prag 1992a: 84-87), and central Italy and Etruria (some examples of Morel series 2233, 2234, 2236 and bowls 2760, 2770 Morel 1981: 150, 151, pl. 36, 37, 3rd and 2nd centuries BCE; series 2811a, 227, pl. 74). In central Dalmatia similar type of plate has so far been documented only on Pakleni islands (Ugarković and Konestra 2018: 76, pl. 1, 8).

Grey slipped ware Advanced 2nd and 1st centuries BCE productions are represented to some extent with black to brown and red coated wares, but the majority of pottery from that period belongs to grey slipped grey ware, including relief ware, a large amount of which conceivably manufactured in regional eastern Adriatic workshop(s). Same as for the black coated ware, numerous grey ware fragments have been recovered at different eastern Adriatic sites, and yet the current state of systematic and contextual research of this pottery class is still not far ahead, especially with regard to regional production, that has for grey ware, for now, been assumed only at Issa, but this again has been based exclusively on the available finds from mortuary context (Ugarković and Šegvić 2017; Ugarković 2019, I: 100-105 and earlier bibliography).

Another type of shallow dish has a mildly emphasised rim, slightly flaring on both ends with a concave middle (Cat. no. 37, Plate 3). Although some analogies could be found in some examples of piattelli con basso orlo from Phoinike (Gamberini 2016: 192, 272, first half of the 3rd century BCE) the closest examples belong to central Italian production of piatti con piccolo orlo, that seem to be produced in the Aesis workshop between 170/160 150/140 BCE (Brecciaroli Taborelli 1996-1997: 88-90, 93, 94, F 1640).

The most frequent shapes among the grey slipped tableware are drinking cups (Cat. nos. 47-55, Plate 5). Among these, kantharoi with an out-turned rim (Cat. no. 49) and/or articulated kantharoi (Cat. nos. 51 and 52) could be clearly noted. The long-lasting popularity of the latter has thus been confirmed with occurrences in different pottery classes, from black coated to grey and red slipped (see below). Besides Krk, grey slipped

The last example is a plate with upturned rim (Cat. no. 38 and 39, Plate 4). Although such a shape appears also in black coated Attic production at the end of the 2nd 266

One man’s trash is another man’s treasure

Pl. 5

48

47

49

50

51

53

52

55

54

56

57

Plate 5. 47-57 Grey-slipped ware (drawings: Martina Korić; plates: Ana Konestra).

267

Marina Ugarković, Ranko Starac grey ware version of this beloved eastern Adriatic shape has been recovered from a number of sites, such as necropolises in Dragišić (Miše 2019: 179, fig. 5, 4; 180) and Kopila (Borzić 2017: 78, 182, 183, 4/4), settlements of Vis (Šešelj 2008: 109), Trogir (personal inspection; Ugarković and Šegvić 2018, tab. 1, T 7) and Resnik (Brusić 2004: 20, fig. 27; Šešelj 2008: 109; Kamenjarin 2016, fig. 4, 7) as well as sanctuaries at Cape Ploča (Šešelj 2008; 2010: 140-143, br. 1, 4-7) and Nakovana (for a distribution map see Šešelj 2008: fig. 12). Even though it appears not to have been that commonly used in Apulia (De Mitri and Mauro 2019: 291), notable finds of both grey and red slipped articulated kantharoi were recently uncovered in the shipwreck of Torre Santa Sabina (De Mitri 2016: 30), while the shape is present in the repertoire of the Apulian production of grey gloss ware (AGG) (Yntema 2005: 76, form 37).

later variant of the black coated Cat. nos. 30 and 31, showing a clear analogy to an early variant of the AGG form 12 (especially 12, b) attested in the MetapontoTaranto area where it has been dated to the later 2 and early 1st centuries BCE (Yntema 2005: 36 and earlier analogies; De Mitri, Maoro 2019: 290, fig. 4, 10). With only 12,5 cm in diameter our fragment is smaller than the typical AGG variant which is between 16 and 20 cm in diameter. Among the two main types of dishes, first is a medium sized shallow dish (c. 26 rim diameter) ending with a horizontally articulated rim (Cat. no. 59). It bears some, although not precise, analogies to certain examples gathered in AGG form 3 that appear to be rather common in large parts of southeast Italy in the period of the late 2nd century BCE until Augustan times (Yntema 2005: 24, 25). The appearance of this shape in the northern Adriatic, e.g. Aquileia, is also dated to the late 2nd and 1st centuries BCE (Mandruzatto and Massieli Scotti 2003: 391, 392, tav. 3, 3), while a similar example is known from Pakleni islands (Ugarković and Konestra 2018: 77, pl. 2, 28).

For some fragments (Cat. nos. 47, 50), one of which shows remains of three horizontal white painted lines below the rim (Cat. no. 47), remains uncertain (even when probable) if they belonged to kantaharoi (or perhaps skyphoi). Both kantharos and skyphos are known in Issaean grey ware (Ugarković 2019, I: 101, fig. 128), although so far without traces of white painted decoration that can otherwise appear on red slipped and other late Hellenistic painted wares, also of assumed local manufacture.

The other dish type has a plain vertical or slightly incurving rim, and a broad gently sloping floor. It is represented in several varieties that differ in size, depth of the vessel and morphological traits of the rim. For example, Cat. no. 60 has a medium sized rim (25 cm in diameter), with a slight concave profile (Plate 6). The vessel of Cat. no. 61, on the other hand, is deeper and the rim (c. 19 cm in diameter), set diagonally and slightly outwards at its end, is longer (Plate 6). The smallest example (Cat. no. 62) has a short vertical rim (15.5 cm in diameter), slightly convex in its profile (Plate 6). The general shape shows analogies to very common forms of the AGG production, form 4 and 5 that differ mostly in the transition to the wall (more gentle in the former) and date (second half of the 2nd and early 1st centuries BCE is suggested for the former and a date between 90/70 BCE and the Augustan period for the latter) (Yntema 2005: 26-30). Our examples are not preserved to such an extent that these morphological differences could be observed. However, Cat. no. 62 seems to be more analogous to AGG form 4, while Cat. no. 60 finds clear similarities with its c variant. The same shape is known as lid plates that appear in late 2nd century BCE in Gravina (Prag 1992b:141-143, with analogies). Close examples can be found in Aquileia (for Cat. no. 60: Dobreva and Griggio 2011: 85, 86, tav. 3,1; for Cat. no. 62: Dobreva and Griggio 2011: 85, 85, tav. 3, 2), while shapes similar to Cat. no. 61 are known from central Italy (Brecciaroli Taborelli 1996-1997: 136, 137, specie 2270 dated to advanced 1st century BCE and Augustan times). Judging from the number of sites where such a dish type has been recovered, e.g. Zadar (Čondić and Vuković 2017: 97), Cape Ploča (Šešelj 2010: 111), Trogir

Another vessel type used for drinking is the hemispherical cup, with a straight (Cat. no. 54) or mildly convex upper body wall (Cat. nos. 53 and 55) that continues into different varieties of short and outcurving rims (8-9 cm in diameter). Typically, these cups have several horizontal grooves below the rim, like on nos. 54 and 55, frequently with relief decoration on the lower body (here not preserved), but can also appear without it (Cat. no. 53). Corresponding cups for which regional production is assumed are known from other sites too, mainly from central Dalmatia, e.g. Trogir (Kovačić 2002: fig. 15; Ugarković and Paraman 2020: pl. 1, 6-11), Cape Ploča (Šešelj 2010: 160, no. 19) and Pakleni islands (Ugarković and Konestra 2018: 77, pl. 2, 37). In the AGG production a similar type of rounded cup with grooves is known as form 33 dated to the second half of the 2nd c. (Yntema 2005: 68, 69). Among the two types of bowls, first is a hemispherical bowl with plain rounded rim (16-17 cm in diameter) and calotte-shaped vessel (Cat. nos. 57 and 58, Plates 5 and 6). As such it resembles form 13 of the AGG production, dated to the middle and second half of the 2nd century BCE (Yntema 2005: 37). Another fragment of a hemispherical bowl has a short out-turned rim, slightly concave on its upper surface (Cat. no. 56, Plate 5). It could be viewed as a somewhat 268

One man’s trash is another man’s treasure

Pl. 6

58

59

60

61

62

64

65

63

Plate 6. 58-65 Grey-slipped ware (drawings: Martina Korić; plates: Ana Konestra).

269

Marina Ugarković, Ranko Starac (personal inspection), Pakleni islands (Ugarković and Konestra 2018: 76, 77, pl. 2, 23-27), Vis (Ugarković and Šegvić 2017: 171-172, fig. 9; Ugarković 2019, I: 103; II: 221, SP 40, 104.8), Salona (Von Gonzenbach 1975: 198, 199, 50a), Narona (Topić 2017: 158, kat. br. 48, 119), it seems to have been one of the most common shapes for serving food of the late 2nd and 1st centuries BCE in the eastern Adriatic too.

the fragments show a simple out-turned rim (Cat. no. 41), sometimes with a double groove below it (Cat. no. 43), or a lower body of the articulated kantharos ending in a ring foot (Cat. no. 44). The preserved upper part of Cat. no. 41 is similar to the red-slipped kantharos with a rounded body emphasized in its lower part, found in tomb 150 of Vlaška njiva necropolis in Issa and described as a local product of the late 2nd - 1st centuries BCE (Ugarković 2019, I: 107; II: CRP 4, 150.10). One fragment found on Pakleni islands (Ugarković, Konestra 2018:76, pl. 1, 20) resembles Cat. no. 43, showing analogies of rim and grooves on examples of the Issaean grey ware kantharoi from Martvilo necropolis (Čargo 2010: 139) and Cape Ploča (Šešelj 2010: 117). Another fragment of a red slipped articulated kantharos is known from Cape Ploča (Šešelj 2010: 146, no. 14), while for such vessels discovered in southern Italy, a Dyrrachion production has been assumed (De Mitri and Mauro 2019: 293, fig. 6, 4).

One handle, with elongated profile with an emphasised middle rib, belongs to a jug (Cat. no. 63), while some of the illustrated ring feet with concave bottoms could have also been a part of small jugs (Cat. no. 64, 65). Similar handles can be found on jugs from e.g. Resnik (Brusić 2004: 20, fig. 30) or Issa (closest on oinochoae, e.g. Ugarković 2019, II: 205, SP 12, 92.1 or p. 229, 142.3, SP 15). As an addition to what was emphasised in the introduction to this chapter, we may add that chemical analysis was conducted on a number of grey ware samples from several eastern Adriatic sites, including fragments from Porta Pisana that belonged to the following shapes: dishes with articulated and vertical rim, kantharos, including the articulated kantharos, jug, mold made bowl and crater (indicated in Table 1). The results were suggestive of a single provenance/ workshop whose products were also consumed in Trogir and at Cape Ploča (Ugarković and Šegvić 2018). This, however, should be confirmed on a larger number of sample and comparative material.

Hemispherical cup, with a short rounded out-tuned rim and several deep grooves below it belongs to a type of a mould-made bowls (Cat. no. 45, Plate 4) that is regionally well known in grey ware and typically associated with relief decoration (for close shape analogy in grey ware, Ugarković and Paraman 2020: pl. 1, 7). One handle, with an elongated convex-concave-convex profile, belongs to a jug (Cat. no. 46, Plate 4). The same profile is known from Issa, in association with presumed locally produced trefoil oinochoae from the 2nd-1st centuries BCE tomb 93 (Vlaška njiva necropolis) (Ugarković 2019: II: 209, 73.2, SP 13).

Red slipped ware The application of red coating on tableware has gained popularity from the second half of the 2nd century BCE onwards (Van Der Enden et al. 2016: 255) and reached its peak in Roman times, although the technique itself has been known and sporadically used since the 6th century BCE (Ugarković 2019, I: 107 I, fn. 913). The fragments presented here belong to the so-called Late Hellenistic red slipped ware, whose shape repertoire is similar to those of black and grey slipped ware, but with the application of a thin, matt red coating (Šešelj 2010:186). In the eastern Adriatic the advanced 2nd and 1st century BCE production of red slipped ware has been hypothesized at Dyrrachion (Shehi 2008: 9-13), while another was plausibly in Issa, as indicated by morphological traits of the shapes in tombs (e.g. kantharos, table amphora) (Ugarković 2019, 1: 107). As a result of firing conditions, the slip can range from brown to red in colour, as in Cat. no. 42.

Relief ware Relief ware was a common type of advanced 2nd and 1st century BCE fine tableware, spread along the eastern Adriatic as inferred by its occurrence in different communities and contexts (for a summary of current state of evidence, Ugarković and Paraman 2020: 302304). For most of these finds a regional manufacture has been assumed. Notwithstanding that mould fragments for production of relief bowls and craters have been found in several sites, namely at Issa, St. Clement-Pakleni islands, Resnik and Zadar (Ugarković and Paraman 2020 and bibliography indicated in footnote 10-13 on p. 303) and that preliminary geochemical analyses indicated a possibility for more than one workshop (Ugarković and Šegvić 2018), the location of such workshop(s) has, for now, been connected more arguably only with Issa (Čargo and Kamenjarin 2020 and earlier bibliography) and hypothesized for Resnik (Šegvić et al. 2012). Regional relief ware is typically made in grey fabric, with grey (rarely red) slip, while occasionally pink clay with red or brown to black slip occurs as well (Miše and Šešelj 2008: 118; Šešelj 2010: 150; Ugarković and Paraman

The majority of red slipped ware from Krk are fragments of drinking vessels (Cat. nos. 41-45, Plate 4). These indicate that kantharoi remained the most beloved shape. Although, none were preserved whole, some of 270

One man’s trash is another man’s treasure

Pl. 7

66

67

69

68

72 70

73

71

Plate 7. 66-73 Relief ware (drawings: Martina Korić; plates: Ana Konestra).

271

Marina Ugarković, Ranko Starac 2020). That said, the regional production of relief ware in terms of morphology, decoration and characteristic of workshops has not been studied in depth so far.

Dragišić (Brusić 1999: 69, 144, fig. 12, no. 58). A number of fragments are decorated with different floral motives, sometimes highly stylized (e.g. Cat. no. 76-81, Pl. 8), while one other wall sherd is adorned with another combination that includes dolphins (Cat. no. 73, Plate 7). Among these, the majority is grey ware with grey slip, while some fragments exhibit different fabrics such a brown clay with grey (Cat. no. 80) or red (Cat. no. 79) slip.

Two shapes are associated with Krk’s relief ware fragments, a bowl and a crater, which are at the same time the most common relief ware forms. The handle-less mould made hemispherical bowl could be attributed to both eastern Adriatic production and imports, the majority (if not all) of the latter are to be connected to Ionian workshop(s). The preserved rims of the former show two varieties of out-turned rims and associated walls with horizontal grooves. A slightly rounded rim, with a convex inner side continues into a vertical and then more convex wall (Cat. no. 66, Plate 7), while a shorter (Cat. no. 67, Plate 7) or longer (Cat. no. 68, Plate 7) rim with an almost flat inner surface is connected to a more vertical wall of, probably, a deeper cup. Both shapes have been previously noted in central Dalmatia. Morphologically, the first finds analogies to examples found in e.g. Trogir (for the closest, Ugarković and Paraman 2020: pl. 1, no. 2, 3), Resnik (Kamenjarin 2014: 144, nos. 8-12), and Pakleni islands (Ugarković and Konestra 2018: Pl. 3, 53), while the second shows similarities to finds from Cape Ploča (Šešelj 2010: 157, no. 13), Resnik (Brusić 1999: 79, 156, fig. 24, A133; Kamenjarin 2014: 142, no. 6) and Salona (von Gonzenbach 1973: 196, 197, no. 25).

Several wall fragments, based mostly on their thickness, are conceivably associated with regionally produced craters, exhibiting traces of different floral motives (e.g. Cat. nos. 74-75, Plate 8). As briefly mentioned in the previous chapter, the chemical analysis that included limited samples of presumed regional both mould made bowl and crater was suggestive of a shared provenance with other grey ware without relief decoration (Ugarković and Šegvić 2018). Contemporary with regional products, imports not only appear as well but their numerical presence, at least if we judge on the preserved diagnostic fragments, seems to at least match, if not outnumber, the regional products. These are evidenced with fragments of bowls with different morphological traits, fabric and decoration. The majority, if not all, can be attributed to eastern Aegean production, known also as the so-called Ionian bowls, with a light brown distinctively micaceous fabric and reddish-brown slip that turns brownish-black on the top of the rim. Morphologically, two different rim types are attested: plain rim with a slight incurving wall (Cat. nos. 88, 89, Plate 9) and mildly oblique and outturned rim, with a sharp transition to the body (Cat. nos. 84-87, Plate 9). Different motives, mainly floral but also geometric, decorate the exterior in bands. The adorning upper band, below the rim, includes motives such as the most common guilloche (Cat. nos. 84, 88, 89, Plate 9; for such a motif with the same rim as on Cat. no. 88 see example from Budva: Marković 2012: 17, 289, Tab. 1, no. 4, different fabric), once preserved with a part of the second row with pointed leaves (Cat. no. 84, Plate 9). Other motives that appear in the upper rim section are e.g. rosettes (Cat. no. 87, Plate 9, for analogies from e.g. Valesio: Yntema 2006: 114, fig. 18, 3, 4) or ionian kymation (Cat. no. 86, Plate 9), while additional motives such as the bead and reel are attested on wall fragments (Cat. nos. 91 and 92, Plate 9). The only preserved base has a big rosette at its centre (Cat. no. 90, Plate 9). The fabric, morphology and decoration find direct analogies in mould-made bowls produced in Ephesos, associated with the ΠΑΡ workshop as well as other smaller succeeding workshops, even though some examples, e.g. with incurving rim, could perhaps be connected to the early production (Rogl 2014). Same could be said for the preserved base fragment whose rosette with two petal types in one layer resembles type 1 (especially 1 b) of bottom medallions appearing in Ephesian production, also attributed to ΠΑΡ workshop (see Rogl 2014: 122-124).

The typical slightly convex base of such bowls, with three moulded appliqués (usually called ‘figural masks’, but also shells, human heads or simple buttons) serving as a stand, is preserved in Cat. nos. 70-72 (Plate 7), while one flat base (Cat. no. 69, Plate 7), is more commonly associated with contemporary classes without relief decoration (e.g. grey ware, Šešelj 2010: 123, no. 25). The motives that adorn the vessels’ exterior, presented sometimes but not always in horizontal bands separated by relief lines, are mostly floral, but also geometric and zoomorphic (dolphins) and yet offer merely a glimpse of the once visible repertory. The majority of single motives, sometimes even the same combinations, have been attested on other eastern Adriatic sites (for Issa and Resnik: Čargo and Kamenjarin 2020: Pl. 1-6). Only one fragment, Cat. no. 66 (Plate 7), shows a part of a recognizable decoration of the upper wall, below the rim, that is a row with two alternating floral motives. Few sherds exhibit different motives applied to the base and the lower wall, including floral motives surrounding the moulded feet (Cat. nos. 70 and 71, Plate 7) or covering the lower body ending with a flat undecorated base (Cat. no. 69, Plate 7). A combination of dolphins that alternate with palmettes (?) and a row of horizontally laid double S spirals can be seen on the lower body of Cat. no. 72 (Plate 7). A similar decoration with dolphins has previously been found in Resnik (Brusić 1999: 92, 167, fig. 35, A215; Čargo and Kamenjarin 2020: Pl. 4 crater) and 272

One man’s trash is another man’s treasure

Pl. 8

74

75 76

77

78 79

80

81

82

83

Plate 8. 74-83 Relief ware (drawings: Martina Korić; plates: Ana Konestra).

273

Marina Ugarković, Ranko Starac

Pl. 9

84

85

86

87

88

89

90

92 91

Plate 9. 84-92 Relief-ware (drawings: Martina Korić; plates: Ana Konestra).

274

One man’s trash is another man’s treasure

The bowls made in ΠΑΡ and succeeding ateliers, produced in the second half of the 2nd and first decades of the 1st centuries BCE have been found in the wider Mediterranean and Black Sea areas (Rogl 2014: 132-135, with early production starting in the course of the first half of the 2nd century BCE, and the production end hypothesized for the mid-1st century BCE). The appearance of such merchandise in Krk is of no surprise, due to early Roman involvement in the region, encouraging tendencies in broader trade and connectivity that resulted also in whole vessels or fragments of Ionian bowls, among which also confirmed products of Ephesian workshop (e.g. Menemachou signature from Resnik), being found in different sites such as Trogir (Ugarković and Paraman 2020: 309, pl. 2, 15, 24), Resnik (Brusić 1999: 15, 93. 169 fig. 37, A222, A224; Rogl 2014: 133f), Cape Ploča (Miše and Šešelj 2008: 116 fig. 5, 3; 118; Šešelj 2010: 170-173 Cat. no. 38-39), Pakleni otoci (Ugarković and Konestra 2018: 78f. Cat. no. 66), Palagruža (Miše and Šešelj 2008: 114 fig. 2, 2), Salona (von Gonzenbach 1978: 186. 196 f. cat 37. 38) and Budva (personal inspection). One bowl with a similar profile and black-brown slip might have been an imitation of Ionian bowls (Cat. no. 83, Plate 9).

cups seem to be skyphoi, the most common shape of drinking vessel throughout the whole examined timeframe, is the kantharos, especially its regionally popular variant, the articulated kantharos, as observed by both quantity and longevity (different pottery classes, black coated, grey- and red slipped). First, kantharoi seem to have been used side by side with mastoid cups, while in the advanced/late 2nd century BCE, even though kantharoi still remained in use at least in the early/first half of 1st centuries BCE, peoples’ tastes changed and they turned to new types of drinking vessels, as they now mostly drank from mould made bowls, which remained in use until the Augustan times. Thin-walled ware dotdecorated beakers are also present from the onset of their production, that is the first decades of the 2nd century BCE, in a variety of fabrics that might suggest different provenances, including conceivably northern Italy, while later in the 1st century BCE other types of TWW beakers appear (Konestra and Ugarković in print). The growing amphorae evidence suggests start of importation of foreign beverages, i.e. wine, the latest in the 4th century BCE, although broader appropriation of wine consumption practices seem to be more related to the 3rd - 2nd centuries BCE, somewhat later in comparison to the central and southern Adriatic (Blečić Kavur 2015: 233; Glavaš et al. 2020; Konestra and Ugarković in print). Similarly, changes and the variety of choices of drinking cup shapes might point to a broader consumption of beverages in everyday domestic contexts taking place in the advanced 3rd century BCE, while different drinking equipment, some of which appear to have been used simultaneously (e.g. kantharos and mastoid cups), might suggest either multipurpose drinking vessel or that particular cup-shape was used for different drinking occasions. Besides cups, the drinking gear included tableware jugs, as evidenced in our context by fragments of Gnathia and other later jugs that are present in most pottery classes, however, the apparent lack of craters, at least in earlier phases, indicates different habits of wine consumption, when compared to e.g. Greek cultural context, where crater is central to the act of social wine drinking. If we were to judge by the available evidence, craters start to appear in both domestic and funeral contexts of Krk’s community in the late 2nd-early 1st century BCE, as seen by examples of relief ware found at both Porta Pisana and Šinigoj necropolis (Brusić 1999: 62, Cat. no. 11.). Before that, one might consider if other types of domestic vessels, coarse wares or containers not made in clay might have served a similar or other role in drinking customs (for more detailed analysis on changes in drinking customs in Kvarner region see Konestra and Ugarković, in print).

Concluding remarks The Kvarner Gulf has thus far rarely yielded evidence of urban contexts prior to Roman consolidation of power in the area, while even less material has been published. This alone makes the studied site an important local and (micro-) regional context where the documented material evidence enables us to gain insight into aspects of continuity and change in daily life from Protohistory into early Roman times. That said, the presented material bears an additional notable significance as it represents household debris of the last few centuries BCE, a not only rare but indeed so far unique regional context. While the analysis of the whole assemblages is still ongoing, the preliminary study of fine wares offers insights on how tables were set in this community and allows us to pursue trends of tableware supply defined by the consumer’s context, local cultural taste and demand, as well as the associated connectivity. Different parts of the functional gear attributed to various pottery classes and diverse provenances point also to some aspects of continuity and change with regard to drinking and dining practices, allowing us a glimpse into processes driven by local agency and stimulated by their involvement in multicultural economic and cultural networks. The earliest fine ware imports belong to the (advanced) 4th and 3rd centuries BCE (e.g. glaux skyphos and possibly some Gnathia sherds), while the preserved layers from Porta Pisana clearly point to a steady supply and broad everyday use of fine tableware in Krk’s community at the latest from the (advanced/late) 3rd centuries BCE. With regard to drinking habits, while first imported drinking

At the table, people now dined from fine ware plates, a shape that is a kind of a novelty since it does not seem to be familiar in regional earlier Iron Age assemblages (Konestra and Ugarković in print). Earliest imports of 275

Marina Ugarković, Ranko Starac such plates appear in the 3rd century BCE (south Italian Gnathia plates), though a steadier supply is visible from later on, especially in the last two centuries BCE. Several types of plates have been detected, though a dish with several varieties of short vertical rim that appears both in black coated and grey slipped variant seems to be mostly used. To this, we may add a complete lack of particular shapes, e.g. fish plates that were otherwise quite popular in several eastern Adriatic sites, most notably in the central and southern areas (e.g. Pharos, Issa, Trogir, Epetion, Budva, etc.; Ugarković 2013: 87, 88), and that could speak of cultural preferences but could partially also reflect chronology of investigated context. Same could be said for the echinus bowl, otherwise one of the commonest shapes of black coated pottery of the 4th-2nd centuries BCE in the central Mediterranean (De Mitri 2016: 30) that is represented here with only one example, in comparison to varieties of hemispherical bowls and bowls with out-turned rims that could have played multipurpose roles, i.e. for both dining and drinking.

ΠΑΡ- and other closely related workshops of Ephesos, otherwise previously recorded at several other eastern Adriatic markets (Rogl 2014: 133-134). Notwithstanding Krk as the currently northernmost occurrence of Ionian mould made bowls in the eastern Adriatic, their quantity and at least equal representation with regional products of the same shape in the Porta Pisana layers is indeed remarkable, especially when compared to other sites e.g. Trogir where 12% of mould made bowls from the settlement have been attributed to overall imports (Ugarković and Paraman 2020: 312). Novelties in the acquisition and appropriation of imported tableware, coupled with changes in drinking and dining practices were undoubtedly greatly impacted by growing upsurge in the mobility of agency and goods that in the observed period culminated with the establishment of Roman (in)formal rule in the region and their more intense military and commercial relations with the eastern Mediterranean, followed by insertion of local communities into wider trading network. In addition to Asia Minor tableware, this is nicely illustrated by importation of other goods, departing even from the far-flung Levantine markets, as indicated by the discovery of Phoenician amphoriskos in Krk (Ugarković and Šegvić 2019). In that vein, Krk was an integral part of Kvarner that in the last centuries BCE seemingly belonged to a broader regional consumption koine, with local agency (re) defining its micro-regional nuances (Konestra and Ugarković in print, with bibliography). That said, the noted is merely a glimpse into dynamics of cultural practices and localisms maintained by Krk’s insularity and connectivity, and yet the recovered potsherds offer a rare opportunity to unlock aspects of everyday domestic life, whose full potential is yet to be uncovered through contextual studies of the whole assemblage.

If we turn to how they lit their homes, the preliminary material analysis did not find evidence of the use of oil-lamps prior to the early Imperial period, when fragments of Roman lamps of Loeschke I/IV type appear. We do know, however, that lamps were imported to the island and broader Kvarner region already in the late 2nd and 1st centuries BCE, as indicated by the eastern Mediterranean import with the signature of ΑΡΙΣΤΟΝΟΣ (Aristonos) from Baška collection as well as Esquilino 2 and a decorazione radiale types from Baška and Jadranovo, of probable North and South Italian productions (Ožanić Roguljić and Konestra 2019: 349340; Konestra and Ugarković, in print). The observed provenances of drinking cups point to association of the earliest specimens with south Italian production(s), that were soon after replaced by more regional products that continue to appear until the end of the old era and the beginning of the new. In that light, we may assume contacts with the central Adriatic, notably central Italy and central Dalmatia at least throughout the last centuries BCE by assigning particular products to workshops of e.g. Aesis or Issa, even though this should not be taken for granted since many aspects of tableware studies, particularly with regard to its Adriatic production, especially its eastern part, are still in their infancy, which makes attributions unsecure at best. The products of Northern Italian workshops, however, seem to be present from the onset of their manufacture. To these we can now add the newly discovered but not at all unexpected links with Asia Minor, as clearly the locals accepted contemporary Mediterranean ‘fashion’ through the importation and use of widely dispersed and popular Ionian bowls of the advanced 2nd and 1st centuries BCE. As noted earlier, many of these have traits that can be associated with

Acknowledgements This work was supported in part by the Croatian Science Foundation under the project RED - Roman economy in Dalmatia: production, distribution and demand in the light of pottery workshops (IP-11-2013-3973). For that, we thank Goranaka Lipovac Vrkljan (Institute of Archaeology, Zagreb), PI of HRZZ RED project, and the Croatian Science Foundation. We would like to express our gratitude to Ana Konestra (Institute of Archaeology in Zagreb), for her kind and continuous help provided in different aspects of the preparation of this manuscript. Also, we wish to thank Matija Makarun (Arheo Kvarner d. o. o.) for his help in gaining insights into the latest excavations of Krk and kind permission to use the information relevant to fine wares in this article, as well as Lujana Paraman (Museum of the Town Trogir), for making the map with the position of Kvarner region. 276

One man’s trash is another man’s treasure

Table 1. Catalogue. Cat. No.

Context

Ware

Shape, preserved part

Drawing

South Italian Geometric/Banded ware? 1

Trench B1-D South Italian Geometric/Band- Closed shape, wall fragment SU 17 ed ware? Red-figure ware 2 Trench B Red-figure ware Glaux skyphos, rim SU 006/012 South Italian? SF 026 Gnathia and other overpainted wares

Pl. 1, 1

3

Trench B North Tower SU 006/012

Skyphos, rim

Pl. 1, 3

4 5 6

SU 17 Gnathia ware SU 17 Gnathia ware Trench B1- D Gnathia ware SU 17 Trench D Gnathia ware SU 17 Trench B Other (white) overpainted North Tower ware SU 006/012 SU 171.-3. 6. Other (white) overpainted ware Trench B1-D Gnathia ware SU 17 Trench B Gnathia ware North Tower SU 016/017 Trench B1-D Gnathia ware, grey fabric SU 17 Black-coated ware

Jug, rim/wall Possible a jug, rim fragment Possible a jug, rim fragments

Pl. 1, 4 Pl. 1, 5 Pl. 1, 6

Skyphos, rim

Pl. 1, 7

(Jug?), rim and wall fragment

Pl. 1, 8

Kantharos (?), rim and wall fragment Closed shape (jug) probably oinochoe, wall fragment and part of the handle Closed shape (jug), wall fragment

Pl. 1, 9 Pl. 1, 10

Closed shape (jug), wall fragment

Pl. 1, 12

Black coated ware Black coated ware

Kantharos, rim/wall Kantharos, rim/wall

Pl. 2, 13 Pl. 2, 14

Black coated ware Black coated ware

Kantharos, rim/wall Kantharos, rim/wall

Pl. 2, 15 Pl. 2, 16

Black coated ware Black coated ware Black coated ware

Kantharos, rim/wall Articulated kantharos, wall Articulated kantharos, wall

Pl. 2, 17 Pl. 2, 18 Pl. 2, 19

Black coated ware Black coated ware

Mastos, rim/wall Mastos, rim/wall

Pl. 2, 20 Pl. 2, 21

Black coated ware

Mastos, rim/ wall

Pl. 2, 22

Black coated ware

Mastos, rim/ wall

Pl. 2, 23

Black coated ware Black coated ware

Mastos, rim/ wall Mastos, rim/ wall

Pl. 2, 24 Pl. 2, 25

Black coated ware

Pl. 2, 26

Black coated ware

Bowl, hemispherical, triple outer groove, rim/wall Bowl, conical, rim/wall, inner grooves

Black coated ware Black coated ware

Bowl, hemispherical, rim/wall Bowl, hemispherical, rim/wall

Pl. 3, 28 Pl. 3, 29

7 8 9 10 11 12

13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29

SJ 17 Trench B North tower SU 006/016 SU 17 Trench B North tower SU 016 Trench B1- D SU 17 Trench B1- D SU 17 Trench D SU 17 Trench B1- D SU 17 Trench D SU 17 Trench D SU 17 Trench C SU 003 SU 17 Trench B North tower SU 006/016 Trench B between two towers Trench D SU 017 SJ 17 Trench B North tower SU 016

Gnathia ware

277

Pl. 1, 2

Pl. 1, 11

Pl. 3, 27

Marina Ugarković, Ranko Starac Cat. No.

Context

Ware

Shape, preserved part

Drawing

30 31

Black coated ware Black coated ware

Small bowl, out-turned rim Small bowl, out-turned rim

Pl. 3, 30 Pl. 3, 31

32

Trench B1- D SU 17 Trench B SU 006/012 SU 17

Black coated ware

Pl. 3, 32

33

SJ 17

Black coated ware

Trench B SU 006/012 35 SU 017 36 SJ 17 37 SU 16/17 38 Trench B between the two towers below SU 013 39 Trench B SU 006/012 40 Trench D SU 17 Red slipped ware

Black coated ware

Bowl (?) very deep, out-turned rim, rim/ wall Bowl, rim/body, out-turned projecting flat rim Bowl, rim/body, out-turned rounded rim

Black coated ware Black coated ware Black coated ware Black coated ware

Small echinus bowl, rim/wall Shallow dish, incurving rim, rim/wall Shallow dish, slightly flaring rim, rim Plate with vertical rim, rim/wall

Pl. 3, 35 Pl. 3, 36 Pl. 3, 37 Pl. 4, 38

Black coated ware

Plate with vertical rim, rim

Pl. 4, 39

Black coated ware

Jug, handle

Pl. 4, 40

41

Trench C SU 3 42 SU 17 43 Trench B North tower SU 016/017 44 Trench B between two towers SU 17 45 Trench B North tower SU 016/017 46 Trench B SU 006, 012, 016 (south from SU 009, 013) Grey slipped grey ware

Red slipped ware

Kantharos, rim/wall

Pl. 4, 41

Red slipped ware Red slipped ware

Kantharos, wall/handle Skyphos/kantharos, rim/wall

Pl. 4, 42 Pl. 4, 43

Red slipped ware

Articulated kantharos, base/lower wall

Pl. 4, 44

Red slipped ware

Cup, rim/wall

Pl. 4, 45

Red slipped ware

Jug, handle

Pl. 4, 46

47

Trench B North tower SU 016 Trench B SU 006/012

Grey slipped grey ware

Kantharos/skyphos (?), rim/wall

Pl. 5, 47

Grey slipped grey ware

Kantharos, rim/wall

Trench B North tower SU 016/017 Trench B North tower 016/017 Trench B North tower SU 006/012 Trench B North tower SU 016 Trench B North Tower SU 006/016 Trench B1- D SU 17

Grey slipped grey ware

Kantharos, rim/wall and handle

Pl. 5, 48 (sample K 5;Ugarković and Šegvić 2018) Pl. 5, 49

Grey slipped grey ware

Kantharos/skyphos/rim

Pl. 5, 50

Grey slipped grey ware

Articulated kantharos/lower part of the vessel

Grey slipped grey ware

Articulated kantharos/body

Grey slipped grey ware

Hemisherical cup, rim/wall

Pl. 5, 51 (sample K 6, Ugarković and Šegvić 2018) Pl. 5, 52 (sample K 10, Ugarković and Šegvić 2018) Pl. 5, 53

Grey slipped grey ware

Cup, rim/wall

Trench B Grey slipped grey ware SU 16 (east and north from 009, 013) Trench B Grey slipped grey ware North tower SU 016

Cup, rim/wall

Pl. 5, 54 (sample K 8, Ugarković and Šegvić 2018) Pl. 5, 55

Bowl with the out-turned rim, rim/wall

Pl. 5, 56

34

48 49 50 51 52 53 54 55 56

278

Pl. 3, 33 Pl. 3, 34

One man’s trash is another man’s treasure

Cat. No.

Context

Ware

Shape, preserved part

Drawing

57

Trench B North tower SU 016 Trench B North tower SU 016 Trench B SJ 009, 013, 006, 012, 016

Grey slipped grey ware

Bowl, with the plain rim, rim/wall

Pl. 5, 57

Grey slipped grey ware

Bowl, with the plain rim, rim/wall

Pl. 6, 58

Grey slipped grey ware

Plate with a horizontal rim, rim/wall

60

Trench B SJ 009, 013, 006, 012, 016

Grey slipped grey ware

Plate, with the vertical rim, rim/wall

61

Trench C SU 003 Trench B SU 006, 012, 016 (south from SU 009, 013) SU 16/17

Grey slipped grey ware

Plate, with the vertical rim, rim/wall

Pl. 6, 59 (sample K 1, Ugarković and Šegvić 2018) Pl. 6, 60 (sample K 2, Ugarković and Šegvić 2018) Pl. 6, 61

Grey slipped grey ware

Shallow plate with a vertical rim, rim/ wall

Pl. 6, 62

Grey slipped grey ware

Jug, handle

Pl. 6, 63

base

Pl. 6, 64

base

Pl. 6, 65

58 59

62 63 64 65

66 67 68 69

Trench B between north Grey slipped grey ware and south towers SU 006/012 Trench B Grey slipped grey ware North tower SU 016/017 Relief ware Trench B Between two towers Trench B1- D SU 17 Trench B North tower SU 016/017 SU 17

70

SU 003 next to SU 002, SF 011

71

Trench B SU 006/012 Trench D SU 17 SU 017

72 73 74

Trench B1-D SU 16

75

SU 17

76

79

Trench D SU 17 Trench B chance find Trench B North tower SU 016 SU 017

80

Deponium find

81

Trench C SU 3 SU 17

77 78

82

Relief ware / grey slipped grey Bowl, rim/wall ware Relief ware / grey slipped grey Bowl, rim/wall ware Relief ware / grey slipped grey Bowl, rim/wall ware

Pl. 7, 66

Relief ware / grey slipped grey Bowl, base/wall ware Relief ware / grey slipped grey Bowl, base/wall ware

Pl. 7, 69

Pl. 7, 67 Pl. 7, 68

Relief ware / grey slipped grey ware Relief ware / grey slipped grey ware Relief ware / grey slipped grey ware Relief ware / grey slipped grey ware

Bowl, base

Pl. 7, 70 (sample K 3, Ugarković and Šegvić 2018) Pl. 7, 71

Bowl, base/wall

Pl. 7, 72

Bowl, wall

Pl. 7, 73

Crater, wall

Relief ware / grey slipped grey ware Relief ware / grey slipped grey ware Relief ware / grey slipped grey ware Relief ware / grey slipped grey ware

Crater ?, wall

Pl. 8, 74 (sample K 9, Ugarković and Šegvić 2018) Pl. 8, 75

Crater?, wall

Pl. 8, 76

Bowl, wall

Pl. 8, 77

Bowl, wall

Pl. 8, 78

Relief ware / grey slipped grey ware Relief ware / grey slipped brownware Relief ware / grey slipped grey ware Relief ware / red slipped ware

Bowl, wall

Pl. 8, 79

Bowl, wall

Pl. 8, 80

Bowl, wall

Pl. 8, 81

Bowl, wall

Pl. 8, 82

279

Marina Ugarković, Ranko Starac Cat. No.

Context

Ware

Shape, preserved part

Drawing

83

Trench B1- D SU 17

Bowl, rim/wall

Pl. 8, 83

84

Trench D SU 006/017 Trench B North tower SU 006/012 Trench B SU 006, 012, 016 (south from SU 009, 013) Trench B SU 006, 012, 016 (south from SU 009, 013) Trench B1- D SU 17 SU 17 Trench B1- D SU 17 Trench B North tower SU 016 Trench B1-D SU 17

Relief ware / brown-black slipped ware Relief ware / Ionian

Bowl, rim/wall

Pl. 9, 84

Relief ware / Ionian

Bowl, rim/wall

Pl. 9, 85

Relief ware / Ionian

Bowl, rim/wall

Pl. 9, 86

Relief ware / Ionian

Bowl, rim

Pl. 9, 87

Relief ware / Ionian Relief ware / Ionian Relief ware / Ionian Relief ware / Ionian

Bowl, rim Bowl, rim Bowl, base Bowl, wall

Pl. 9, 88 Pl. 9, 89 Pl. 9, 90 Pl. 9, 91

Relief ware / Ionian

Bowl, wall

Pl. 9, 92

85 86 87 88 89 90 91 92

Blečić, M. 2007. Reflections of Picens Impact in the Kvarner Bay, in M. Guštin, P. Ettel and M. Buora (eds) Piceni ed Europa. Atti del convegno: 109‒122. Trieste: Editreg. Blečić Kavur, M. 2014. At the crossroads of worlds at the turn of the millennium. The Late Bronze Age in the Kvarner region (Musei Archaeologici Zagrabiensis Catalogi et Monographiae XI). Zagreb: Arheološki muzej u Zagrebu. Blečić Kavur, M. 2015. Povezanost perspektive. Osor u kulturnim kontaktima mlađeg željeznog doba / A coherence of perspective. Osor in cultural contacts during the Late Iron Age. Koper-Mali Lošinj: Založba Univerze na Primorskem, Lošinjski muzej. Bolonić M. and I. Žic Rokov 2002. Otok Krk kroz vjekove. Zagreb: Kršćanska sadašnjost. Borzić, I. 2017. Ceramic Finds, in I. Fadić and A. Eterović Borzić (eds) The City of the Dead on the Field of Life: The Necropolis of the Kopila Hillfort on the Island of Korčula: 61‒85. Zadar: Muzej antičkog stakla. Bradanović, M. 2016. Grad Krk u Srednjem vijeku. Split: Muzej hrvatskih arheoloških spomenika Split. Bradanović, M. and D. Ciković 2013. The Church of St. Lawrence Outside the Town Walls of Krk. Hortus artium medievalium - journal of the Interantional Research Center for Late Antiquiti and Middle Ages 19: 183‒196. Brecciaroli Taborelli, L. 1996-1997. Marche. Jesi (Ancona) — L’officina ceramic di Aesis (III sec. a. C. – I sec. d.C). Notizie degli Scavi di Antichità s. IX/VII-VIII: 5‒250. Brusić, Z. 1999. Hellenistic and Roman Relief Pottery in Liburnia (BAR International Series 817). Oxford: Archaeopress. Brusić, Z. 2004. Resnik: hidroarheološka istraživanja. Kaštela: Muzej grada Kaštela. Buršić-Matijašić, K. 2011. Nalaz brončanih predmeta s otoka Krka. Prilozi Instituta za arheologiju u Zagrebu 27( 2010): 73‒82.

References Balen Letunić D. 1992. Nalaz ratničke opreme iz Krka. Vjesnik Arheološkog muzeja u Zagrebu 3. s./XXIV-XXV: 21‒34. Balen Letunić, D. 1996. Figuralno ukrašene trapezoidne pojasne kopče tipa Prozor. Vjesnik arheološkog muzeja u Zagrebu 3.2.: 28‒29. Batović, Š. 1987a. Liburnska grupa, in A. Benac and S. Gabrovec (eds) Praistorija Jugoslavenskih Zemalja V – željezno doba: 339‒390. Sarajevo: Akademija nauka i umjetnosti Bosne i Hercegovine, Centar za balkanološka ispitivanja. Batović, Š., 1987b. Prapovijesni ostaci na otoku Rabu. Rapski zbornik, Zbornik radova sa znanstvenog skupa o otoku Rabu održanog od 25. do 27. listopada 1984. godine: 147‒169. Zagreb: JAZU, Skupština općine Rab. Becker, M. J., I. E. M. Edlund-Berry, M. Limoncelli, E. G. D. Robinson, F. Silvestrelli, S. Viva, C. D’Annibale, M. Mazzoli and F. Perugin 2016. The Tombs at Sant’Angelo Vecchio, in: F. Silvestrelli, I. E. M. Edlund-Berry and J. Coleman Carter (eds) The Chora of Metaponto 6. A Greek Settlement at Sant’Angelo Vecchio: 85‒128. Austin: University of Texas Press. Bilić Dujmušić, S. 2014. Bitka kod Krka godine 49. pr. Kr., in G. Lipovac Vrkljan, B. Šiljeg, I. Ožanić Roguljić and A. Konestra (eds) Roman pottery and glass manufactures. Production and trade in the Adriatic region (Proceedings of the 2nd International colloqium, Crikvenica, 28th-29th October 2011) (ZIA 2): 107‒121. Crikvenica-Zagreb: Muzej Grada Crikvenice, Institut za arheologiju. Blečić, M. 2002. Kastav u posljednjem tisućljeću prije Krista. Vjesnik arheološkog muzeja u Zagrebu, s. 3/ XXXV: 67‒142.

280

One man’s trash is another man’s treasure

Buršić-Matijašić, K. and R. Matijašić 2011. Novi arheološki nalazi s otoka Krka (iz zbirke Samblich) / New archaeological findings from the island of Krk (from the Samblich collection). Krk: Centar za kulturu grada Krka. Cambi, N., B. Kirigin and E. Marin 1980. Zaštitna arheološka istraživanja helenističke nekropole Isse (1976. i 1979.) — preliminarni izvještaj. Vjesnik za arheologiju i historiju dalmatinsku 75: 63‒83. Cambi, N., B. Kirigin and E. Marin 1981. Excavations at Issa, Island of Vis, Yugoslavia 1976,1979 — A Preliminary Report. Rivista di archeologia 4: 83‒86. Cook, A. A. B. 1940. Zeus: A Study in Ancient Religion, Vol. 3, part 1. Cambridge: Cambridge University Press. Cubich, G. B. 1874-1875. Notizie naturali e storiche sull’isola di Veglia. Trieste: Stabilimento tipografico Appolonio & Caprin. Čargo, B. 2010. Martvilo. Istraživanja jugozapadne isejske nekropole do 1970 / Martvilo. Research into the southwest Issa necropolis up to 1970 (Katalozi i monografije Arheološkog muzeja u Splitu 3). Split: Arheološki muzej u Splitu. Čargo, B. and I. Kamenjarin 2020. Hellenistic Mouldmade Pottery from Issa (Vis) and Siculi (Resnik–Kaštela), Croatia (A Preliminary report), in I. Kamenjarin and M. Ugarković (eds) Exploring the neighbourhood. The role of ceramics in understanding place in the Hellenistic world. Proceedings of the 3rd Conference of IARPotHP Kaštela, June 2017, 1st–4th (IARPotHP 3): 327‒342. Wien: Phoibos. Čelhar, M. and I. Borzić 2016. Gradina u Zemuniku Donjem – nalazi željeznog i rimskog doba, in: J. Faričić and Z. Dundović (eds) Zemunik u prostoru i vremenu: 69‒117. Zadar-Zemunik: Sveučilište u Zadru. Čondić, N. and M. Vuković 2017. U temeljima antičkog grada. Iz arheoloških slojeva liburnskog Zadra. Zadar: Arheološki muzej Zadar. De Juliis, E. M. 1996. San Severo. La necropoli di Masseria Casone. Bari: Edipuglia. De Mitri, C. 2016. Trade in Pottery within the Lower Adriatic in the 2nd c. BCE. Journal of Hellenistic Pottery and Material Culture 1: 21‒36. De Mitri, C. and F. Mauro 2019. Le Rouge et le Noir. La ceramica fine da mensa: continuita` e cambiamenti nell’area ionico-adriatica nel corso del II sec. a. C., in A. Peignard Giros (ed.) Daily Life in a Cosmopolitan World: Pottery and Culture During the Hellenistic Period. Proceedings of the 2nd Conference of IARPotHP Lyon, November 2015, 5th– 8th (IARPotHP 2): 285‒299. Wien: Phoibos. Dobreva, D. and A. M. Griggio 2011. La ceramica a vernice nera dai fondi ex Cossar ad Aquileia: problematiche e prospettive di ricerca. Quaderni Friulani di Archeologia XXI: 77‒100. Faber, A. 1963-65. Antički bedemi grada Krka. Vjesnik za arheologiju i historiju dalmatinsku LXV-LXVII: 45‒54.

Faber, A., 1978. Arheološka topografija Krka. Arheološki pregled 19: 74‒77. Faber, A., 1979. Istraživanja prethistorijsko-antičke topografije u Krku. Arheološki pregled 20: 183‒186. Faber, A. 1980. Osor–Apsorus iz aspekta antičkog pomorstva. Diadora 9: 289–311. Faber, A., 2000. Gradnja protohistorijskih i ranoantičkih bedema u priobalju Ilirika. Histria Antiqua 6: 145‒170. Fiorentin A. 2001. Krk-Splendissima civitas Curictarum (Krčki zbornik-posebno izdanje sv. 38). Krk: Adamić. Gamberini, A. 2016. Ceramiche fini ellenistiche da Phoinike: forme, produzioni, commerci (Scavi di Phoinike, Serie Monografica 2). Bologna: Ante Quem. 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 Amphora Finds, in I. Kamenjarin and M. Ugarković (eds) Exploring the neighborhood. The role of ceramics in understanding place in the Hellenistic world. Proceedings of the 3rd Conference of IARPotHP Kaštela, June 2017, 1st–4th (IARPotHP 3): 267‒282. Wien: Phoibos. Glogović, D. 1989. Prilozi poznavanju željeznog doba na sjevernom Jadran: Hrvatsko Primorje i Kvarnerski otoci (Monografije JAZU 1). Zagreb: JAZU. Govorčin, P., and I. Borzić 2018. Prilog poznavanju helenističke keramike iz liburnske Aserije. Asseria 14: 29‒66. Green, J. R. 2001. Gnathia and other overpainted wares of Italy and Sicily — a survey, in P. Lévêque and J.-P. Morel (eds) Céramiques hellénistiques et romaines III: 57103. Paris: Presses Universitaires Franc-comtoises. James, S. 2018. Hellenistic pottery. The fine wares (Corinth VII). New Jersey: American School of Classical Studies at Athens. Jeličić Radonić, J. 1995. Pharos- antički grad, in J. Jeličić Radonić and B. Rauter Plančić (eds) Pharos. Antički Stari Grad (Exhibition catalogue): 96- 115. Zagreb: Muzejsko galerijski centar. Jelušić, J., V. Krstić, D. Ratković and M. Vrzić 2007. Blago budvanske nekropole. Budva: Muzej grada Budve. Johnson, F. P. 1953. An owl skyphos, in G.E. Mylonas and D. Raymnond (eds) Studies Presented to David Moore Robinson on his Seventieth Birthday 2: 96‒105, pl. 32b. Saint Louis: Washington University. Johnson, F. P. 1955. A note on owl skyphoi. American Journal of Archaeology 59: 119‒124, pl. 35-38. Kalebota, M. 2015. Grčko-helenistička keramika iz Trogira, unpublished MA thesis. Zadar: University of Zadar. Kamenjarin, I. 2014. Hellenistic Moldmade Relief Pottery from Siculi (Resnik). Vjesnik za aheologiju historiju dalmatinsku 107: 129‒160. Kamenjarin, I. 2016. Hellenistic Pottery from Siculi (Resnik), in S. Japp and P. Kögler (eds) Tradition and Innovations. Tracking the Development of Pottery from the Late Classical to the Early Imperial Periods, Proceedings of the

281

Marina Ugarković, Ranko Starac 1st Conference of IARPotHP Berlin, November 2013, 7th10th (IARPotHP 1): 177-185. Vienna: Phoibos. Kirigin, B. 2020. Owl skyphoi around the Adriatic, in N. Sekunda (ed.) Wonders Lost and Found. A celebration of the archaeological work of Professor Michael Vickers: 58‒69. Oxford: Archaeopress. Kirigin, B. and E. Marin 1985. Issa ‘80 — preliminarni izvještaj sa zaštitnih arheoloških iskopavanja helenističke nekropole Martvilo u Visu: Novi i neobjelodanjeni natpisi iz Visa. Vjesnik za arheologiju i historiju damatinsku 78: 45‒72. Konestra, A. 2015. Ricerche nei musei della Liburnia settentrionale (Quarnero, Croazia): potenzialità, nuove attestazioni e aggiornamenti sulla diffusione di alcune tipologie ceramiche. Quaderni Friulani di Archeologia XXV: 117‒122. Konestra, A. 2016. Italska terra sigillata i keramika tankih stijenki na području sjeverne Liburnije: tipologija, kronologija i distribucija, unpublished PhD thesis. Zadar: University of Zadar. Konestra, A. and M. Ugarković in press. To each their own: fine wares, appropriation and changing social practices in the last centuries BCE Kvarner (NE Adriatic, Croatia), in F. Diosono and D. Maschek (eds) Beyond Hybridity and Code-Switching: Material Approaches to Cultural Concurrences in the Late Hellenistic Mediterranean, in print Kovačić, V. 2002. Nuove scoperte nella Tragurion ellenistica, in N. Cambi, S. Čače and B. Kirigin (eds) Grčki utjecaj na istočnoj obali Jadrana, Zbornik radova sa znanstvenog skupa održanog od 24.-26. rujna 1998. godine u Splitu: 375‒395. Split: Književni krug. Kušan Špalj, D. and A. Rendić Miočević 2006. Najljepše grčke vaze iz riznice Arheološkog muzeja u Zagrebu. Zagreb: Arheološki muzej u Zagrebu. Lamboglia, N. 1952. Per una classsificazione preliminare della ceramica campana. Genova: Bordighera. Lanza, E. 2005. Ceramica di gnathia al Museo di antichità di Torino. Mantova: SAP. Lanza Catti, E. 2008. La ceramica ‘di Gnathia’ al Museo nazionale Jatta di Ruvo di Puglia. Ipotesi di ricontestualizzazione (Antenor Quaderni 11). Roma: Quasar. Lanza Catti, E. 2011. Diversificazione funzionale dei manufatti ceramic in rapporto al contest: il casa delle ceramic ‘di Gnathia’ nella Peucezia Preromana, in Tra protoistoria e storia. Studi in onore di Loredana Capuis (Antenor Quaderni 20): 265‒279. Roma: Quasar. Lipovac, G. 1991, Razmatranja o problemima antičkog bedema grada Krka-povodom novih nalaza, Prilozi Instituta za arheologiju u Zagrebu 8: 37-42. Makarun, M. 2014. Izvještaj sondažnih arheoloških istraživanja u svrhu legalizacije objekta na lokaciji k.č. 1015/1/15 u naselju Osor na otoku Cresu, unpublished excavation report. Makarun, M. 2016a. Krk – Galija 6. Hrvatski arheološki godišnjak 12/2015: 466‒468.

Makarun, M. 2016b. Krk – Galija 14. Hrvatski arheološki godišnjak 12/2015: 468-469. Makarun, M. 2016c. Krk – vila Schinigoi (crkva sv. Lovre). Hrvatski arheološki godišnjak 12/2015: 470‒472. Makarun, M. and N. Novak 2016. Izvještaj sondažnih arheoloških istraživanja u svrhu legalizacije objekta na lokaciji Jurja Križanića 19 u gradu Krku, unpublished report. Mandruzatto, L., F. Massieli Scotti 2003. Provenienza della ceramica a vernice nera di Aquielia. Aquileia nostra 74: 378‒394. Marchesetti, C. 1924. Isole del Quarnaro. Ricerche paletnologiche, Regio X (Venetia et Histria). Notizie degli Scavi di Antichità XXI: 121‒148. Marković, Č. 2012. Antička Budva. Nekropole – istraživanja 1980–1981. Podgorica: Matica crnogorska. Matejčić R. 1963. Izvještaj o arheološkom iskapanju u gradu Krku. Vijesti muzealaca i konzervatora Hrvatske 12: 113–115. Matejčić, R. 1974. Prethistorijska nekropola Mišinac u Kastvu. Liburnijske teme 1: 52‒77. McpHee, I. 1981. Red-figured pottery from Corinth sacred spring and elsewhere. Hesperia 50/3: 264‒284. Mihovilić, K. 2002. Grčki i helenistički nalazi u Istri i Kvarneru, in: N. Cambi, S. Čače and B. Kirigin (eds) Grčki utjecaj na istočnoj obali Jadrana / Greek influence along the East Adriatic coast. Zbornik radova sa znanstvenog skupa održanog 24. do 26. rujna 1998. god. u Splitu: 499‒519. Split: Književni krug. Mihovilić, K. 2004. Ceramica greca in Istria. I Greci in Adriatico. Hesperia 18/2004: 101‒121. Miše, M. 2015. Gnathia and related Hellenistic ware on the East Adriatic coast. Oxford: Archaeopress. Miše, M. 2017. Banquet among the stars: Gnathia and West Slope ware from the Diomedes sanctuary on the island of Palagruža. Vjesnik za arheologiju i historiju dalmatinsku 110-1: 233‒255. Miše, M. 2019. Drinking wine in Liburina: Hellenistic ware in indigenous necropolis Dragišić in Dalmatia, Croatia, in A. Peignard Giros (ed.) Daily Life in a Cosmopolitan World: Pottery and Culture During the Hellenistic Period. Proceedings of the 2nd Conference of IARPotHP Lyon, November 2015, 5th– 8th (IARPotHP 2): 175‒185. Vienna: Phoibos. Miše, M. and L. Šešelj 2008. The late Hellenistic and early Roman fine pottery from the sanctuaries of Diomedes in Dalmatia. Rei Cretariae Romanae Fautores Acta 40: 113‒119. Moore, M. B. 1997. Attic red-figured and white-ground pottery (The Athenian Agora. Results of Excavations Conducted by the American School of Classical Studies at Athens 30). Priceton: American School of Classical Studies at Athens. Morel, J.-P. 1981. Céramique campanienne: les formes. Rome: Ecole française de Rome. Ožanić Roguljić, I. and A. Konestra 2019. Rassegna topografica e cronologica delle lucerne fittili romane 282

One man’s trash is another man’s treasure

dalla regione del Quarnero (Croazia), in L. Chrzanovski, A. Nestorović and V. Vidrih Perko (eds) Ancient lamps from Balkans and beyond, Acts of the 4th International Lychnological Congress (‘Ex Oriente Lux’, Ptuj, 15th-19th of May, 2012): 339‒352. Montagnac: Mergoil. Perkić, D. 2010. Svetište u Vilinoj špilji iznad izvora rijeke Omble, in J. Poklečki Stošić (eds) Antički Grci na tlu Hrvatske: 159‒161. Zagreb: Galerija Klovićevi dvori. Polonijo, M. and F. Bulić 1931. Starinski nalaz u Krku. Povodom pedesetogodignjice jubileja Vjesnika za arheologiju i historiju dalmatinsku 1878-1928: 73‒76 + T I, II. Prag, A. J. N. W. 1992a. Black glaze ware: a catalogue, in Small, A. M. (ed.) An Iron Age and Roman Republican Settlement on Botromagno, Gravina di Puglia. Excavations of 1965-1974, Vol. II, The Artifacts (Archaeological monographs of the British school in Rome 5): 70‒131. London: British school in Rome. Prag, A. J. N. W. 1992b. Grey glaze ware, in A. M. (ed.) An Iron Age and Roman Republican Settlement on Botromagno, Gravina di Puglia. Excavations of 1965-1974, Vol. II, The Artifacts (Archaeological monographs of the British school in Rome 5): 134‒154. London: British school in Rome. Redavid, V. 2010. La ceramica ‘di gnathia’ nella ‘necropoli occidentale’ di Egnazia. Fasano: Schena Editore. Reynolds Scott, A. 2007. Cosa. The Black-Glaze Pottery 2 (Supplements To The Memoirs Of The American Academy In Rome). Ann Arbor: University of Michigan Press. Riccardi, A. 2007. Ceramica a facse e di stile misto, in E. M. De Juliis (ed.) Rutigliano I. Catalogo del Museo Nazionale archeologico di Taras II,2, Rutigliano I Le necropoli di contrata Purgatorio Scavo 1978: 351‒385. Taranto:  Scorpione editrice. Rogl, C.2014. Mouldmade Relief Bowls from Ephesos– The Current State of Research, in P. Guldager Bilde and M. L. Lawall (eds) Pottery, Peoples and Places. Study and Interpretation of Late Hellenistic Pottery (Black Sea Studies 16): 13‒139. Aarhus: Aarhus University Press. Rotroff, S. 1997. Hellenistic pottery. Athenian and imported wheelmade table ware and related material (The Athenian Agora XXIX). New Jersey: The American School of Classical Studies at Athens. Shehi, E. 2008. La ceramica a vernice rossa da Dyrrachium. Rei Cretariae Romanae Fautores Acta 40: 9‒17. Starac, R. 2012. Lokalitet: Krk gradske zidine. Hrvatski arheološki godišnjak 8/2011: 456‒460. Sticotti, P. 1893. Bericht uber eine ausflug nach Liburnien und Dalmatien. Archäologisch − epigraphische Mittheilungen 16: 34‒43. Sticotti, P. and E. Nowotny 1896. Aus Liburnien und Istrien. Archäologisch − epigraphische Mittheilungen XIX/2: 159‒180. Šegvić, B., L. Šešelj, D. Slovenc, B. Lugović and R. Ferreiro Mahlmann 2012. Composition, techonology of manufacture

and Circulation of Hellenistic Pottery from the Eastern Adriatic. A case study of three archaeological sites along the Dalmatian coast, Croatia. Geoarchaeology 27: 63‒87. Šešelj, L. 2005. Utjecaj dirahijskih radionica na području srednje Dalmacije. Vjesnik za arheologiju i historiju dalmatinsku 98: 49‒61. Šešelj, L. 2008. The influence of the Dyrrachium pottery workshops in central and south Dalmatia. Rei Cretariae Romanae Fautores Acta 40: 105‒112. Šešelj, L. 2010. Promunturium Diomedes: Svetište na rtu Ploča i jadranska pomorska trgovina u helenističkom razdoblju, unpublished PhD thesis. Zadar: University of Zadar. Škoro, I. 2020. Južnoitalska mat slikana keramika iz Zadra: arheološka istraživanja do 1989. godine. Diadora 33-34: 69‒100. Tonc, A. 2013. Rezultati probnih iskopavanja na gradini Svete Trojice 2012. godine. Annales Instituti Archaeologici IX: 116‒120. Topić, M. 2017. Keramika, in M. Topić (ed.) Forum Naronitanum: 148‒271. Vid: Arheološki Muzej Narona. Ugarković, M. 2013. Red-figure fish plate. Vjesnik za arheologiju i povijest dalmatinsku 106: 75‒98. Ugarković, M. 2019. Geometrija smrti: isejski pogrebni obredi, identiteti i kulturna interakcija. Antička nekropola na Vlaškoj njivi, na otoku Visu / The Geometry of Death: Issa’s burial rites, identities and cultural interaction The ancient necropolis at Vlaška njiva on the island of Vis, I/1-2 (Katalozi i monografije Arheološkog muzeja u Splitu 6). Split-Zagreb: Arheološki muzej u Splitu, Institut za arheologiju. Ugarković, M. and B. Šegvić 2017. Tableware of the living or pots for the dead? An introduction to the 2nd and 1st c. BC greyware from the tombs of Issa (Eastern Adriatic), in G. Lipovac Vrkljan, B. Šiljeg, I. Ožanić Roguljić and A. Konestra (eds) Proceedings of the 3rd International Archaeological Colloquy ‘Roman Ceramic and Glass Manufactures; Production and Trade in the Adriatic region’, Crikvenica 4th – 5th November 2014 (ZIA 12): 161‒183. Crikvenica-Zagreb: Muzej Grada Crikvenice, Institut za arheologiju. Ugarković, M. and A. Konestra 2018. Hellenistic and Early Roman pottery from Soline bay of the St. Clement island, in M. Ugarković (ed.) Praetoria longe lateque lucentia, Zbornik radova posvećen Vlasti Begović povodom 65. obljetnice života (ZIA 8): 71‒107. Zagreb: Institut za arheologiju. Ugarković, M. and A. Konestra 2020. Stobreč — bedemi: preliminarni pregled keramičkih nalaza s istraživanja 2012. godine / Stobreč — fortifications: preliminary assessment of pottery finds from the 2012 excavations, in Okolica Kaštelanskog zaljeva u prošlosti (Izdanja HAD-a, Vol. 33): 47‒64. Zagreb: Hrvatsko arheološko društvo. Ugarković, M. and B. Šegvić 2018. The role of archaeometry in the study of production and 283

Marina Ugarković, Ranko Starac dispersal of Hellenistic pottery in Dalmatia, with new evidence on the chemistry of grey-ware tableware, in G. Lipovac Vrkljan and A. Konestra (eds) Pottery Production, Landscape and Economy of Roman Dalmatia. Interdisciplinary approaches: 89-106. Oxford: Archaeopress. Ugarković, M. and B. Šegvić 2019. New insights into the far-flung markets of Levantine products: The Phoenician amphoriskoi from Issa (Eastern Adriatic, Croatia), in A. Peignard Giros (ed.) Daily Life in a Cosmopolitan World: Pottery and Culture During the Hellenistic Period. Proceedings of the 2nd Conference of IARPotHP Lyon, November 2015, 5th– 8th (IARPotHP 2): 67‒79. Wien: Phoibos. Ugarković, M. and L. Paraman 2020. Appropriation of the Hellenistic Relief Ware in Ancient Trogir (Central Dalmatia, Eastern Adriatic): Preliminary Observations, in: I. Kamenjarin and M. Ugarković (eds) Exploring the neighbourhood. The role of ceramics in understanding place in the Hellenistic world. Proceedings of the 3rd Conference of IARPotHP Kaštela, June 2017, 1st– 4th (IARPotHP 3): 301‒326. Phoibos: Wien. Van De Enden, M., P. Bes, and E. Conoci 2016. Red Ware, in F. Silvestrelli, I.E.M. Edlund–Berry and J.C. Carter (eds) The Chora of Metaponto 6. A Greek settlement at Sant´ Angelo Vecchio: 255‒265. Austin: University of Texas Press.

von Gonzenbach, V. 1975. Pottery from closed deposits, in C. W. Clairmont, S. Handler Auth, V. von Gonzenbach (eds) Excavations at Salona, Yugoslavia (1969-1972): 181‒209. New Jersey: Noyes Press. Watson M. 1999. The owls of Athena: some comments on owl-skyphoi and their iconography. Art Bulletin of Victoria 39: 35‒44. Yntema, D. G. 2005. Conspectus forumarum of Apulian Grey Gloss Ware (ceramica a pasta grigia). Amsterdam : Institute of Archaeology, Vrije Universiteit. Yntema, D. G. 2006. The Birth of a Roman Southern Italy, a case study: Ancient written sources and archaeological evidence on the early-Roman phase in the Salento district, southern Italy (3rd-1st century BC). Babesch. Bulletin antieke beschaving 61: 95‒137. Zelić, D. 1993. Nastanak urbanih naselja na otoku Krku. Radovi Instituta za povijest umjetnosti 17/2: 7‒17. Žic Rokov, I. 1962. Ubikacija rimskog groblja i neki drugi problemi u Krku. Bulletin Zavoda za likovne umjetnosti Jugoslavenske akademije znanosti i umjetnosti 10/1-2: 33‒41.

284

Pots for storage of food from Crikvenica Ivana Ožanić Roguljić Abstract Among around 100 types of vessels that were produced in the pottery workshop of Sextus Metilius Maximus one type can be pointed out as a storage vessel used for transportation and trade in smaller range of distribution. Double- or single-handled vessels appear in several variants and may have played a number of roles, from tableware to storageware. In view of their shape, they are mostly considered as pots, although certain forms may have been used as cups or small table kraters for mixing wine. Analyses of the contents revealed that certain types were used for storing and perhaps also transporting garum, (allec), fruits and olives, while some are considered as honey-storage vessels. They differ by the shape of the rim (JW 1 and JW 2). Differences can be found in the size of the vessels. It is difficult to say what were their contents, but since we can suspect that garum amphorae were produced at the site, some of those pots could also be used in the same way. Distribution of this type of pots produced in Crikvenica is attested at Senj, Bakar and Žuta lokva and we can assume that these pots were used for the goods needed in the regional trade

Key words Roman Dalmatia, Roman pottery, Crikvenica, Sextus Metillius Maximus, Roman food, garum, honey, fruit, figs.

ʽThe site of the pottery workshop is located near the mouth of the Dubračina Stream and a gorge connecting it with the Vinodol valley, a rare flysch area within the dominantly karst landscape of the Kvarner (Benac et al. 2006: Figs 1, 2), the area with thick clay sediments. Data derived from finds excavated within the site stresses a link to this resource-rich hinterland. (...) Tile stamps bearing the name of the owner of the ceramic workshop – Sextus Metilius Maximus – indicate also a type of property that it belonged to a saltus, seen as an estate devoted to farming, but also to pasture and forest exploitation. (...) Timber for fuel and also clay needed for ceramic production must have been likewise brought from the Vindol valley.ʼ (Welc et al. 2017: 119). Crikvenica’s hinterland provided land transportation infrastructure necessary to market ceramics, pottery and commodities, by means of the road running through it. The workshop produced amphorae for wine, olive oil and garum, multifunctional pottery (fine tableware, coarse ware for cooking, storage and transportation, special forms) and construction material (Lipovac Vrkljan and Ožanić Roguljić 2013). Double- or single-handled pots Among around 100 types of multifunctional vessels that were produced in the pottery workshop of Sextus Metilius Maximus, one type can also be pointed out as a storage vessel used for transportation and trade in smaller range of distribution. Double- or single-handled pots appear in several variants and may have played a number of roles, from tableware to storage ware. In view of their shape, they are mostly considered as pots, although certain forms may have been used as cups or small table kraters for mixing wine (Ožanić Roguljić 2012).

At this moment residue analysis are not made on the two handled pots made in Crikvenica. Samples from excavations in Crikvenica are waste material from the workshop and were never in use, Bakar and Žuta Lokva samples are probably contaminated due to the conditions of their storage in the museums. From analogies we know that double handled pots played a number of roles, from tableware to storage ware. In view of their shape, they are mostly considered as pots or jugs, although certain smaller forms may have been used as cups or small table kraters for mixing wine. Analyses of the contents revealed that certain types like JW 1 were used for storing and perhaps also transporting garum, and other processed fish products (allec, salsamenta, small fish, parts of fish), fruits (figs) and olives, while some types are considered as honeystorage vessels (Ostia II: 95, T LVIII; Crane 1983: 267; Schindler-Kaudelka 1998; Djaoui et al. 2014; Djaoui and Capelli 2017; Picchi and Menchelli 2018: 403). Three types of wide-necked pots with two handles were produced in Crikvenica, which differ by the shape of the rim (JW 1 and JW 2) and by height (JW 3). Vessels were mostly preserved in fragments, so the difference between Types JW 1 – 3 can be determined based on the height of the handle, which in the case of JW 3 was around 5 – 6 cm, while in JW 1 and JW 2 it measured between 6,6 and 9,5 cm. This difference is important because JW 1-2 and JW 3 might have different usage. JW 1 (Figure 1) Double-handled oval vessel with strap rim decorated with flutes. The volume of JW 1 vessels is around 2,735 l, or five sextarii in Roman measurements. It is made

roman pottery and glass manufactures (Archaeopress 2022): 285–288

Ivana Ožanić Roguljić in fabric CRI 3. In analogies variations of JW 1 Type continued long in use, from the end of the 1st century until the beginning of the 3rd century (Schindler Kaudelka 1989: T 26: 26; Gasperetti 1996: 30; Plesničar Gec (ed.) 2006: T 38:2; De Stefano 2008: 112, T 2O: 19. 1; Ožanić Roguljić 2018: 37-43). JW 2 (Figure 1, Figure 3) Single- or double-handled oval vessels. Differ from the previous type in the shape of the strap rim, which is smooth. They are made in fabrics CRI 3, CRI 4 and CRI 5. A jug found in SU 103 (small kiln) stands out in particular (Figure 2). Although it is preserved almost complete, it is very distorted due to faulty manufacture. A vessel of this kind must have been discarded the moment it was retrieved from the kiln. The site of Bedriacum yielded both versions (with one or two handles). The ovoid shape of the jugs has a very long tradition in Italian ceramics. The examples from Bedriacum were found in the layers from the 1st century BC until the mid-1st century AD, and the dates from Padua are the same (Rossi 2014: T LIII: 2.1,1). They appear at a number of 1st century sites, while in Ostia they were found also in a 3rd century context (Schindler Kaudelka 1989:

Figure 1. Pots with two handles JW 1; JW 2; JW 3 (author: M. Gregl ), (JW 1: rim diameter: 10 cm; bottom diameter: 6,5 cm; wall thickness: 0,3 cm; place of find: PN 1276, Su 103; JW 2: h: 15,8 cm; rim diameter: 12 cm; bottom diameter: 7,7 cm; wall thickness: 0,4-0,6 cm; place of find: PN 814/817, Su 103; JW 3: h: 12,1 cm; rim diameter: 12,5 cm; bottom diameter: 6,1 cm; wall thickness: 0,3 cm; place of find: PN 1528, F 16, 303).

Figure 2. Type JW 2- distorted rim, damage done during drying or firing (author: M. Galić).

286

Pots for storage of food from Crikvenica

(necropolis) and Žuta Lokva (roadside station) and at this point we can assume that these pots especially types JW 1 and JW 2 were used for the goods needed in the regional trade of the northern part of Roman Dalmatia used as for transport but also convenient for storage and as ritual offering. They can be dated from the second half of the 1st CE to beginning of 2nd CE. Types from Crikvenica would belong to first group of Picchi and Menchelli (Picchi and Menchelli 2018: 410) which they describe as ovoid and pear-shaped form with high rim and vertical handles. Pot type JW 3 could be rather used as table ware (for serving food or mixing vine). It is impossible to know what kind of goods they were meant to hold. Finds from necropolis can indicate that the Bakar vessels contained fruits as offering rather than fish product like allec which are probably very rare as offering. It is also possible to consider them as container for honey as funeral offering (Ransome 2004: 122). The find from Žuta Lokva is important because it is a proof of distribution along the Senia - Siscia road and it is found as the inventory of a roadside station. In context of a small road station which served food and beverages for travelers these containers could storage all of the proposed products.

Figure 3 Type JW 2 with one handle; h: 19 cm; rim diameter: 14 cm; bottom diameter: 9,1 cm; wall thickness: 0,5 cm; place of find: PN 1257, E 18, Sj 25 (author: M. Gregl).

Amphorae are usually considered as containers for goods and most of the known trade of agricultural products and economy is described by the analysis of their typology and distribution on global and regional scale. Amphora for fish products was produced in Crikvenica (Lipovac Vrkljan and Konestra 2017), but, smaller vessels like the two-handled pots can also be an excellent document for patterns of regional trade since they are more suitable for traveling on the roads or for loading on small boats. They are also more convenient for personal shopping in the markets. Analysis of such vessels can enhance the knowledge of diverse agricultural products that were important part of a Roman provincial economy. Presence of pots with two handles produced in Crikvenica is the sign of much wider range of products that circulated the markets of northern Dalmatia and in the future those pots should be more carefully analyzed.

T 26; Ostia III: 141, T XXIX: 186; Della Porta and Sfredda 1996: 142, Fig. 182-184; Gasperetti 1996: 30). Jugs of this type were produced also in large dimensions. Such an unpublished specimen was found at the necropolis in Bakar, while in Bedriacum and Padua such forms were found also in contexts from the Augustan period, although their fabric was completely different (Della Porta and Sfredda 1996: 177, Fig. 192). JW 3 (Figure 1) At the middle of the belly of double-handled vessels with strap rim there is sometimes a knee of sorts, flutes or (most commonly) smooth surface. They were made in fabric CRI 3. Volume of a whole vessel is volume is 1300 ml. A completely preserved specimen allows us to observe that the height of the vessel and the rim diameter differ in only a few millimeters. A vessel from Herdonia, dated from the Augustan period to the mid-1st century, resembles the Type JW 3 (De Stefano 2008: 83, T 20: 18.1). Pot JW 3 from Crikvenica is also of very fine making and it is possible that it was used as tableware as a krater for mixing vine but also for serving food (Ožanić Roguljić 2018: 37-43).

Description of fabrics mentioned in the text:

Conclusion

Fabric CRI 3: The following shapes were made in this fabric: jugs, bowls, wide-necked jugs with two handles, large bowls, pots, lids, censers, vessels with button-shaped receptacles on handles, strainers, vessels with a perforated mesh, small amphorae, ceramoplastics, gaming tokens, amphorae. This is the most common fabric. It consists of levigated, porous clay that can be considered as characteristic for the Crikvenica workshop.

Distribution of this group of vessels produced in Crikvenica is attested at Senj (Senia, urban contest), Bakar

C: soft, tiny white and glistening inclusions (1mm, 20 - 30%), small black spots, reddish yellow 287

Ivana Ožanić Roguljić 5 YR 6/8, yellowish red 5 YR 5/8, light red 2.5 YR 6/8, S: smooth surface

Lipovac Vrkljan, G. and I. Ožanić Roguljić 2013. Distribucija crikveničke keramike kao prilog poznavanju rimskog gospodarstva. Senjski zbornik 40: 255–270. Lipovac Vrkljan, G. and A. Konestra 2017. Crikvenička riblja amfora — indikator proizvodnje/trgovine ribljim prerađevinama?, in G., Lipovac Vrkljan, I. Radić Rossi and A. Konestra (eds) Adriamphorae. Amfore kao izvor za rekonstrukciju gospodarskoga razvoja jadranske regije u antici: lokalna proizvodnja. Radovi s okruglog stola, Zagreb, 21. travnja 2016.: 48‒62. Zagreb: Institut za arheologiju. Ostia II = A. Carandini and C. Panella (eds) 1970. Le terme del nuotatore: Scavo dell’ambiente I. Roma: De Luca. Ostia III. = A. Carandini and C. Panella (eds) 1972. Le terme del nuotatore, scavo degli ambiente III, VI, VII, V e di un saggio nell’area SO. Roma: De Luca. Ožanić Roguljić, I. 2012. Klasifikacija i tipologija rimske keramike s lokaliteta Crikvenica Igralište — proizvodi keramičarske radionice Seksta Metilija Maksima, Unpublished PhD dissertation, Filozofski fakultet Sveučilišta u Zagrebu, Zagreb. Ožanić Roguljić, I. 2018. Overview of the capacity of pottery vessels from the workshop of Sextus Metillius Maximus (Crikvenica), in G. Lipovac Vrkljan and A. Konestra (eds) Pottery Production, Landscape and Economy of Roman Dalmatia. Interdisciplinary approaches (Archaeopress Roman archaeology 47): 37‒43. Oxford: Archaeopress. Picchi G. and S. Menchelli 2018. Pots for food. Some regional instances in Italy. Rei Cretariae Romanae Fautorum Acta: 403‒412. Plesničar Gec, Lj. (ed.) 2006. Emonski forum (Annales Mediterranea 10). Koper: Založba Annales. Ransome, H. M. 2004. The Sacred Bee in Ancient Times and Folklore (Dover Books on Anthropology and Folklore). Mineola (NY): Dover Publications. Rossi, C. 2014. Le necropoli urbane di Padova romana (Antenor Quaderni). Padova: Padova University Press. Schindler-Kaudelka E. 1998. Die gewöhnliche Gebrauchskeramik vom Magdalensberg (Archäologische Forschungen zu den Grabungen auf dem Magdalensberg 10, Kärntner Museumsschriften 72). Klagenfurt: Verlag des Landesmuseums für Kärnten. Welc, F., G. Lipovac Vrkljan, A. Konestra and T. Rosić 2017. Remote sensing of a Roman pottery workshop. Report on a geophysical survey carried out in Crikvenica (ancient Ad Turres, Croatia). Studia Quaternaria 34/2: 119–130.

Fabric CRI 4: Forms made in this fabric included jugs, large bowls (basins), pots and even a fragment of a thin-walled cup of Type TWP 6.1. It differs from the previous one by rough surface and larger quantity of tiny white inclusions. C: soft, large quantity of tiny round white inclusions (1mm, 30 - 40%), small black spots, red 2.5 YR 5/8; S: rough surface Fabric CRI 5: In this fabric were made: censers, bowls, jugs, wide-necked jugs with two handles, large bowls, pots, lids, vessels with button-shaped receptacles on handles. Fabric CRI 5 appears of good quality, although some fragments are very loose, which is due to poor firing. C: hard, some tiny white inclusions (1mm, 20%), and small black spots visible with naked eyes, appears well-levigated, red 2.5 YR 5/8 red 2.5 YR 5/6, S: rough surface References Benac, Č., I. Ružić and E. Žic 2006. Ranjivost obala u području Kvarnera. Pomorski zbornik 44/1: 201–214. Crane, E. 1983. The Archaeology of Beekeeping. Ithaca (NY): Cornell University Press. Della Porta, C. and N. Sfredda 1996, La ceramica comune, in L. Passi Pitcher (ed.) Bedriacum. Ricerche archeologiche a Calvatone, 1.2. Milano: Edizioni ET. De Stefano, A. 2008., Un contesto ceramico di età repubblicana e primo/imperiale dall’area delle due domus, in G. Volpe and D. Leone (eds) Ordona XI: ricerche archeologiche a Herdonia: 45–144. Bari: Edipuglia. Djaoui, D., G. Piquès and E. Botte 2014. Nouvelles donnees sur les pots dits ‘á garum’ du Latium, d’apres les dècouvertes subaquatiques du Rhône (Arles), in E. Botte and V. Leitch (eds) Fish & Ships. Production et commerce des salsamenta durant l’Antiquitè. Actes de l’atelier doctoral, Rome 18-22 juin 2012. (Bibliothèque d’archéologie méditerranéenne et africaine 17): 175‒197. Aix-en-Provence: Éditions errance, Centre Camille Jullian. Djaoui, D. and C. Capelli 2017. Objets d’importation ou objets personnels ? La dotation de bord des marins au regard du grand commerce. SFECAG. Actes du Congrès de Narbonne: 115‒132. Gasperetti, G. 1996. La ceramica comune da mensa e dispensa nella Campania romana, in Michel Bats (ed.) Les Céramiques communes de Campanie et de Narbonnaise (Ier s. av. J.-C. – II e s. ap. J. – C.). La vaisselle de cuisine et de table, (Collection du Centre Jean Bérard 14): 19–64. Naples: Centre Jean Bérard. 288

Roman pottery from Siculi – a short overview Ivanka Kamenjarin Abstract The study of Roman pottery from Siculi is only just beginning. This article presents a short overview of completely restored vessels on display in the permanent exhibition at the Museum of the Town of Kaštela, Kaštel Lukšić, Croatia. So we can only give a summary conclusion concerning the type of vessels found in Siculi. The finding of a carrot amphora, however, is especially prominent.

Key words Siculi, Roman pottery, carrot amphora.

Figure 1. Location of Siculi within Kaštela bay (author: I. Kamenjarin).

On the western side of Kaštela Bay (Figure 1), at the Resnik site, in the vicinity of the hotel complex, an ancient settlement was discovered and has been excavated since 1991, although with an occasional hiatus. The settlement has two basic development phases: the first one dates to the period from the end of the 2nd to the end of the 1st century BC, and its architectural remains are much better preserved, along with numerous pottery finds, while the second phase dates from the 1st to the end of the 4th/ beginning of the 5th century AD and is connected to the colonization of Claudius' veterans and the settlement's reconstruction. The architectural remains from this phase are not that well preserved, both due to the hotel's building and due to intense agricultural activities over the last few centuries. Pottery is a great indicator of a settlement's life – it can show us the settlement's commercial relationships,

economy, and sometimes even its customs. Along with numismatic finds, it helps date the site more accurate, which is also the case with Siculi (Ilkić 2019). The examination of Roman pottery from Siculi has only just started, so this article will present a short overview of it. In total, 11 graves have been found, all of which belong to the Roman period. Five of them are in amphorae, whereas the rest belong to a type of walled tombs with multiple burials (Kamenjarin 2013). Given that shattered amphorae covered the deceased, it is difficult to say to which exact type those amphorae belonged, but for now, they can be attributed to the North African production of the 4th and 5th centuries (Kamenjarin and Šuta 2011: 77-85). Apart from amphorae, larger vessels found here include some sherds of dolia. Unfortunately,

roman pottery and glass manufactures (Archaeopress 2022): 289–296

Ivanka Kamenjarin

Figure 2. Site plan of Resnik from 1991 excavations (drawing: M. Popović).

they haven't been completely preserved, so we can only determine their appearance through analogy: an elongated spherical shape with a flat bottom on a small leg with a minor central depression. The closest analogy to them would be Trstenik in Kaštel Sućurac (Radić Rossi 2006).

1988: Nr. 1017). The other oil lamp (Plate 1: 7) also has a flower covering its whole disc, but it also has a stylized pine branch on its shoulder. It is dated slightly later, to the fourth/beginning of the fifth century (Hayes 1980: T 34, 281; Atlante I: T. CLVII 2). During excavations in 1991, in the area marked as Sonda 1's area, north of the preserved plastered flooring (Figure 2: PF), a carrot amphora was found. The area in question exhibits a higher concentration of amphorae than usual, but this kind of amphora is not common on our coast, nor is it common elsewhere in the Mediterranean (Ožanić 2005). It is a smaller amphora (two or three liters in volume), with a plain or rounded rim without a neck, with two handles placed on the shoulder. It got its name due to the shape of its body, which is elongated and narrows conically at the end. It has a ribbed surface, with a sandy texture of the clay and reddish-brown to red color. Due to the presence of sand in the composition of its ceramic, it was originally assumed that it comes from Egypt (Martin-Kilcher 1994). However, archaeometric tests showed that it originates from Palestine (Carreras Monfort and Wiliams 2002). They are considered to have been used for transporting dates overseas. The lack of these amphorae in the aforementioned countries is explained by the fact that dates were easily accessible there, so there was no need to store them. Most of the published examples come from the western part of the Roman Empire, especially from the military sites of England and the Rhine and date back to the Flavian period (Carreras Monfort and Wiliams 2002).

The fine pottery is comprised of dishes used for dining. This includes plates and bowls of different sizes, as well as jugs, all of which belong to the standard repertoire of that particular period found at all Roman sites on the eastern coast of the Adriatic (Konestra 2019). The superficial analysis of the Roman pottery remains in Siculi was conducted on complete or mostly complete vessels which are a part of the Museum of the Town of Kaštela's permanent collection. It was noticed that African red slip ware in its various forms, mostly plates and smaller bowls (especially Hayes 51, 52, 58B), is prevalent. Upon closer inspection during the excavations, a distinct lack of terra sigillata was noticed, something that demands special attention in the future. Oil lamps are often found throughout Antiquity, and the same goes for Siculi. Complete examples have been found in previously mentioned graves, and they date back to the period between the 3rd and 5th century AD. They belong to different types of Athenian and Corinthian workshops, some of them unadorned or with stylized pine branches on their shoulders (Bubić 2012). Two oil lamps with a stylized flower on their disc stand out. The first one comes from Grave 2 (Plate 1: 6) and has the motif of a rosette with elongated petals on its disc, and based on that, using analogies, it can be dated to the third or fourth century (Di Filippo Balestrazzi

In his list of towns on the eastern coast of the Adriatic, Pliny the Elder places Siculi, the place where Claudius 290

Roman pottery from Siculi – a short overview

Plate 1. Oil lamps (drawings: 1-4 — D. Žanić, 5-7 — M. Golubić).

settled his veterans, between Trogir and Solin. This finding could potentially support the claim that the veterans, even after retiring from service, still continued to consume dates on a regular basis, dates that were imported from distant eastern lands.

1. Inv. no. 3869 An oil lamp with an elongated nozzle, oval discus has a circular filling hole in the centre and raised rim around the discus. An open nozzle channel leads to a semi-circular nozzel. Two shoulder lugs with six imprinted circles each are on the bevelled shoulder. The shoulders and the edge of the nozzle are adorned with relief circles and multiple oval cymatium. Oval handle with a small hole has double

Catalogue All measurments are in centimeters. Abbreviations used are: H = height; W = width; L = length; D = diameter 291

Ivanka Kamenjarin incised lines in the middle, whereas on the back, towards the lamp's bottom, there is a single incised line with imprinted dots on the side. At the bottom is a ring connected to the nozzel through a laddershaped motif. Fine pottery of a yellowish-orange colour with a reddish coating. There are visible burning marks on the nozzle.

5. Inv. no. A1/21 A small oil lamp of an oval shape with a raised rim around the plane circular discus. The filling hole is in the middle. The handle is plate-like, with no ornamentation. Slithgly dented in base is defined by a groove. Completely unadorned, made of ochre ceramic.

Dimensions: L - 12; H - 5,5; W - 8. Datation: 4th – 5th century AD

Dimensions: L - 6,6; W - 5,2. Datation: 3rd - 4th century AD Analogy: Di Filippo Balestrazzi 1988: Nr 1158.

2. Inv. no. 3871 A pear-shaped oil lamp, reconstructed. The plate-like handle is solid with a incised line in the middle. The concave oval discus has a filling hole in the middle. The motif of stylized pine branches is on the bevelled shoulders. The bottom has a ring and an incised line that goes towards the handle. The ceramic is of low quality, worn-out, with light-brown color.

6. Inv. no. A1/20 A complete round oil lamp with a triangular nozzle. The discus is concave and adorned with a rosette of elongated petals. The filling hole is in the discus's centre surrounded by a circle. An air hole is located on the double concentric grooves that separates the nozzle from the discus. Flat plain base. The platelike handle is adorned with two incised lines. The lamp is red, with burning marks on the nozzle.

Dimensions: L - 10,9; H - 5,3; W - 7,1. Datation: 4th – 5th century AD Analogy: Bubić 2012: No. 35.

Dimensions: L -7,6; W - 5,8. Datation: 3rd - 4th century AD Analogy: Di Filippo Balestrazzi 1988: Nr. 1017.

3. Inv. no. 3872 A pear-shaped oil lamp, with a hole in the central part of the discus, the air hole halfway through the channel towards the nozzle. The area around the filling hole is adorned with concentric lines, as are shoulders, too. The same motif is on the outer side of the reinforcement. Annular bottom has a relief line and leads towards the handle. Ceramic is of bad quality, the surface is worn out, and the colour is light-brown.

7. Inv. no. A1/24 A pear-shaped oil lamp with a concave discus. In it's centre there is a flower motif around the filling hole. The bevelled shoulder is adorned with a stylized pine branch which is poorly visible. A large part of the nozzle's tip is missing. In the middle of its platelike handle is an incised line. The bottom is slightly indented. The colour is orange-red.

Dimensions: L - 10; H - 4,5; W- 7. Datation: 4th – 5th century AD Analogy: Bubić 2012: No. 94.

Dimensions: L - 9; W - 5,9. Datation: 4th – beginning of the 5th century Analogy: Hayes 1980: T. 34, 281; Atlante I: T. CLVII 2.

4. Inv. no. 3870

8. Inv.no. 2497

The body of this oil lamp is pear-shaped, an elongated nozzle. The discus is slightly concave and damaged at the place where the filling hole should be (traces). Unclear relief is worn-out. Bevelled shoulders are adorned with transverse lines. The handle is platelike and solid, with an incised line on the ridge. Slightly depressed bottom is surrounded by a incised line which open ends continue towards the handle. A wide open nozzle channel, circular wick hole nozzle has burn marks. The coating is missing. It could belong to North-African red-glaze pottery.

A plate with flat base and a triangular overhanging rim. On the inside of the floor: in the middle two grooves enclosing a ring of rouletting. About one quarter of the vessel has been preserved. Hard red ceramic. Could belong to Eastern Sigillata B. Dimensions: D - 18, H - 3,5. Datation: 2nd century Analogy: Hayes 2008: No. 354; 1983: No. 105. 9. Inv. no. A1/6

Dimensions: L - 11; H - 4,6; W - 7,1. Datation: 4th – 5th century AD Analogy: Bubić 2012: no. 20.

A flat-base plate with a round rim. No decoration in the centre. The coating is not well preserved. Hard 292

Roman pottery from Siculi – a short overview

Plate 2. Fine ware (photo: T. Bartulović).

293

Ivanka Kamenjarin red ceramic. Almost completely preserved. Probably belongs to Eastern Sigillata B.

three grooves. It is possible that the stamp was in the middle of the vessel, but it is poorly preserved and unidentifiable. The bowl belongs to Eastern Sigilata B2 production (form Hayes 60). Incidental spillover system 2005, quadrant 14

Dimensions: D - 16,5, H – 3. Datation: late 1st – early 2nd century AD Analogy: Hayes 2008, No. 328; Atlante II, form 63.

Dimensions: D - 18,5; H - 4,5. Datation: / Analogy: Hayes 1983: No. 32; 1972: form 60.

10. Inv. no. 3845 A shallow and wide vessel with an indented base. Belongs to north-African pottery. Orange in colour with a slightly darker glaze. Originates from a devastated grave (Kamenjarin 2013: 133-134).

15. Inv. no. 2500 A bowl with flat base. The rim is thickened and rounded, its edge on the inside has a groove. On the inside of the floor a rosetta stamp. The bowl belongs to Eastern Sigilata B2 production. Incidental spillover system 2005, quadrant 14

Dimensions: D - 25; H - 5.2. Datation: 4th century AD Analogy: Hayes 1972: form 50 A/B. 11. Inv. no. 3793

Dimensions: D - 24; H - 6,5. Datation: / Analogy: Hayes 1972: form 76 b.

Bowl with low ring-foot, vertical wall with sharply transitions towards the base. The rim is bevelled, and the fold is on the outer side profiled with four grooves. North-African production, probably a variant of Hayes 14 b. Incidental spillover system quadrants 12 and 14.

16. Inv. no. 999 A carrot-shaped amphora type Bb (Carreras Monfort and Williams 2002: 134). Rounded body which narrows below handles and becomes conical spike bottom. The rim is nearly vertical, slightly outcurved (Vipard 1995: type 3b1). The handles have ring cross section and are placed on the shoulder. Ribbed surface, the clay's texture is sandy, while the colour varies from reddish-brown to light orange.

Dimensions: D – 20; H – 7,5. Datation: 2nd century Analogy: Hayes 1972, form 14 b?. 12. Inv. no. 2499 Bowl Dimensions: diameter height

Dimensions: H - 58; W - 20. Datation: 1st - 2nd century AD? Analogy: Carreras Monfort and Williams 2002: 134; Vipard 1995: type 3b1.

13. Inv. no. 3798 Smaller bowl with a flat bottom which is slightly elevated at the edge. Outwardly drawn rim is thickened, and the edge of the rim is made prominent on the inside with a carved line. At the base, on the outside, five concentric lines have been carved in. On the inside of the bottom, in the central part, a stamp in the shape of a palmette has been impressed. The bowl belongs to Eastern Sigilata B2 production.

17. Inv. br 2828 A jug, made of brownish-grey ceramic, with a gently everted lip, short and narrow neck, almost globular body which narrows down towards the low ringfoot. Vertical strap handle with grooves on spine goes from shoulder to neck. Made on a pottery wheel with a defect – while the ceramic was still soft, an indentation was created in the middle of the body.

Dimensions: D - 15.5; H - 5,4. Datation: 3rd century AD Analogy: Hayes 1972: form 44; Robinson 1959: No. M32.

Dimensions: H - 20; W - 15. Datation: 3rd century AD Analogy: Nikolić-Đorđević 2000: type VII/7.

14. Inv. no. 2572

18. Inv. no. 2496

Deep plate with a flat base. It has a biconical rim. On the inside, the transition from the body to the rim is marked with two grooves. There is also some decoration on the inner side of the bottom, with

A pear-shaped mug small flat base, slightly curved high rim with inturned lip-band. Handle with an elyptical cross section starts from the middle of 294

Roman pottery from Siculi – a short overview

Plate 3. 16 carrot amphora, 17 jug, 18 pitcher, 19 Tripolitanian 1 amphora (photos: T. Bartulović, drawings: S. Bubalo).

295

Ivanka Kamenjarin the rim and ends on the shoulder. The inside is ribbed, while most of the outside, apart from the lower part, was dipped into reddish-brown colour before baking. The mug belongs to Cnidian pottery production. Incidental spillover system 2005, quadrant 15

Di Filippo Balestrazzi, E. 1988. Lucerne del Museo di Aquileia, vol. II, 2. Aquileia: Associazione Nazionale per Aquileia. Hayes, J. 1972. Late Roman Pottery. London: The British School at Rome. Hayes, J. 1980. Ancient Lamps in the Royal Ontario Museum. Vol. I. Greek and Roman Clay Lamps. Toronto: Royal Ontario Museum. Hayes, J. 1983. The Villa Dionysos Excavations, Knosssos: The Pottery. The Annual of the British School at Athens 78: 97‒169. Hayes, J. 2008. Roman Pottery: Fine-Ware Imports (The Athenian Agora vol. 32). Princeton: The American School of Classical Studies at Athens. Ilkić, M. 2019. Ancient coins from Siculi. Kaštela: Museum of the Town of Kaštela and University of Zadar. Kamenjarin, I. 2013. Antički grobovi iz Sikula. Kaštelanski zbornik 10: 119‒140. Kamenjarin, I. and I. Šuta (eds) 2011. Ancient Siculi (Exhibition catalogue). Kaštela: Museum of the Town of Kaštela. Konestra, A. 2019. Antičko i kasnoantičko keramičko posuđe, in H. Anić (ed.) Salona iza Porta Andetria. Arheološka istraživanja u ulici Stjepana radića u Solinu: 41‒62. Split: Arheološki muzej Split. Martin-Kilcher S. 1994. Die romischen Amphoren aus August und Kaiseraugst 2-3 (Forschungen in Augst 7/2, 7/3). Augst: Römermuseum Augst. Nikolić-Đorđević, S. 2000. Antička keramika Singidunuma – oblici posuda. Singidunum 2: 11‒206. Ožanić, I. 2005. Tipovi amfora iz Cibala. Vjesnik za arheologiju i povijest dalmatinsku 98: 129‒145. Radić Rossi, I. 2006. Prošupljeni doliji – osebujni nalazi iz hrvatskog primorja. Histria Antiqua 14: 83‒93. Robinson, H. 1959. Pottery of the Roman Period: Chronology (The Athenian Agora vol. 5). Princeton: The American School of Classical Studies at Athens.

Dimensions: H - 20; W rim - 9. Datation: 2nd century AD Analogy: Hayes 2008: No. 1632. 19 Amphora An amphora with a gently pointed ending, slighlty prominent shoulder and profiled rim with no neck, and annularly curved handles typical for the end of 3rd/beginning of the 4th century. The deceased (a child) was placed on its back with arms stretched along body. The amphora was surrounded by smaller rocks. Dimensions: H - 104; W body - 50 Datation: end of 3rd/beginning of the 4th century Analogy: / Sources Pliny the Elder, Passini Refrences Atlante I = Enciclopedia dell’arte antica, classica e orientale. Atlante delle forme ceramiche I (medio e tardo Impero), Pugliese Carratelli, G. (ed.). Roma: Istituto della Enciclopedia Italiana fondata da Giovanni Treccani. Bubić, V. 2012. Kasnoantičke svjetiljke iz arheološkog muzeja u Splitu. Vjesnik za arheologiju i povijest dalmatinsku 105: 117‒178. Carreras Monfort C. and D. F. Williams 2002. ‘Carrot’ amphoras: a Syrian or Palestinian connection? The Roman and Byzantine Near East 3: 133‒144.

296

First results of the excavations of the Roman Villa of Bunje (Brač Island, Croatia) Emmanuel Botte, Kristina Jelinčić Vučković, Ana Konestra, Ivana Ožanić Roguljić Abstract An international team has been excavating a Roman villa on the island of Brač at the Novo Selo Bunje site since 2015. Different pottery categories from different production centres were found. Amphorae came from Northern African, Eastern Mediterranean, possibly Eastern Adriatic and Italian workshops along with transported goods. Sigillata finds vary from early Italian, Eastern Mediterranean and later North African types. From the 3rd century AD ARS is dominant along with, later, Phocaean red slipped ware. Pannonian fine wares, such as dark gray PSW bowls and plates, and marmorized bowls, have also been found, just as thin-walled ware from Italic workshops. A certain amount of table ware was not identified in terms of origin, and given the quantity we assume it was produced in yet unknown Eastern Adriatic workshops. Dolia are presumed to have been produced locally too. Cooking ware is present with different vessels, imported from the Eastern Mediterranean, South-Eastern Adriatic (Illyirian ware), North African workshops while some were probably produced locally. Tegulae mostly carry PANSIANA stamps. Key words Brač island, Dalmatia, Novo Selo Bunje, Amphorae, Terra Sigillata, ARS, LRC, Thin-walled ware, Dolia, Pannonian pottery, Tegulae, stamps.

The excavations on the Roman Villa of Bunje (island of Brač, Central Dalmatia) are lead by a Croatian and French team working since 2015 on this project (Botte et al. 2016; Turkalj et al. 2016; Botte et al. 2017; Jelinčić Vučković et al. 2017; Jelinčić Vučković and Botte 2018). It is part of a wider project called ‘Eastern Adriatic Roman villae: form and chronology of the exploitation of shore and islands during Antiquity’, directed by E. Botte (Centre Camille Jullian, CNRS), A. Bertrand (Paris-Est Marne University) and K. Jelinčić Vučković (Institute of Archaeology of Zagreb), with financial support of the École française de Rome and the French Ministère de l’Europe et des Affaires étrangères. Financial support was provided by Ministry of culture (Croatia) and Municipality of Selca. Its aim is to work on the economic situation in Central and Southern Dalmatia between 2nd century BC and 3rd century AD, more precisely before and after Rome started to found colonies in this territory from the mid 1st century BC. Its approach is basically economic, based on the study of villas with food production equipments (wine, oil, fish products) or / and evidence of containers production (amphorae). General presentation of the site The Bunje villa is situated on the southern side of a hill named Velo Gračišće culminating at 218 m height, on which a prehistoric site is also attested (Figure 1). Both sites are known since more than one century for the villa and since the 1960’s for the prehistoric site (Bulić

1914; Fisković 1981; Vrsalović 1969; Zaninović 1968; Cicarelli 1982; Jelinčić 2005; Jelinčić Vučković 2013). The villa was known for the presence of two sarcophagi, one bearing a Christian monogram, one column with a representation of a goat and one monolithic limestone block being part of the olive oil production facility, called mola olearia, which served to crush the olives before bringing them to the press. We started to work on the site in 2015, first making important cleaning and topographic works. It resulted that the villa had a hold of c. 2500 m2, measuring 120 by 160 roman feet (35 by 50 m) (Botte et al. 2016; Turkalj et al. 2016; Botte et al. 2017; Jelinčić Vučković et al. 2017; Jelinčić Vučković and Botte 2018). The general plan is typical of the classical Roman villas in the countryside (Figure 2), with both a residential section (pars urbana) and a production area (pars rustica) separated by a courtyard. Since 2016 we started to excavate the pars urbana and discovered the baths area, with a caldarium and a tepidarium both equipped with apsidal receiving vats, and installed on a hypocaust heating system working with rectangular shaped tubes for the hot air circulation. The baths had a decoration using marble plating (Figure 3) from Carrara quarries and Greek green porphyry, and white mosaics were on the floor. In 2017 we discovered the furnace at north of the caldarium, and explored the rest of the residential section north of the baths, with rooms opened on a

roman pottery and glass manufactures (Archaeopress 2022): 297–303

Emmanuel Botte, Kristina Jelinčić Vučković, Ana Konestra, Ivana Ožanić Roguljić

Figure 1. Location of Novo Selo Bunje (map: K. Jelinčić Vučković). portico (Botte and Jelinčić 2017; Jelinčić Vučković and Botte 2018: 128, Plan 1).

we can say that the site was first occupied by a small installation, maybe closer to a farm than a villa. During the 2nd century AD, the villa as we see it now has been built. After a period of inoccupation that we can only put during the 3rd century, a reoccupation of the site from the mid 4th century until the end of the 5th or beginning of the 6th century is well attested by the structures and the artefacts. In this contribution a preliminary overview of ceramic and pottery finds is presented, aiming at illustrating the wares present at the site, and their contribution to the knowledge on its chronology.

The site was for a long period a cultivated area, and many of the layers have been mixed and / or destroyed. That made the interpretation of the site difficult and challenging. Hopefully, some of the deeper levels are partially preserved but they do not represent the major part. In this context, we have a general overview of the occupation of the site but for now we are not able to propose an assured phasing with a precise chronology of the development of the villa. Anyway, 298

First results of the excavations of the Roman Villa of Bunje

Figure 2. Novo Selo Bunje site, 2017. (photo: E. Botte). (Matijašić 1983; Pellicioni 2012). Some of the tegulae had a secondary purpose as loom weights. The beginnings of the villa are characterised also by the presence of fine wares such as Italian and Eastern Mediterranean sigillata, and characteristic thin-walled pottery (Figure 4). Only a few fragments of Italian sigillata have been recognised, while thin-walled ware of the same provenience is more substantially present with several examples of bowls of the Shindler Kaudelka D/E fabric both plain and with barbottine decoration. Their dating is very wide and can be placed from the second quarter of the 1st century AD to well into the 2nd century (Shindler Kaudelka 1975: 106-111; Ricci 1985: 283-285; Zampori Vanoni 1987: 176; Martin 1995: 185-186).

Figure 3. Fragments of marble plating from baths (photo: E. Botte).

Later thin-walled ware is present with the so-called ollette a collarino, probably of Eastern Mediterranean production, datable from the end of the 1st to the 3rd century AD (Hayes 1997: 69-70; 2008: 102-103; Parica 2008; Pellegrino 2009). To the same region of provenience can be ascribed a more substantial number of ES B plates and bowls. Three recurring shapes have been recognised: Hayes 60 (AD 50/60-150), Hayes 7475 (AD 70/75-120) and Hayes 80 (AD 80-150 or later)

The pottery and ceramic evidence To the beginnings of the occupation of the site can be attributed several PANSIANA stamps on tiles (Figure 6), whose production is attested since the Augustan period until the Flavian period, and whose distribution is confirmed all along the Adriatic coast. At least one production area is attested in Voghiera in North Italy 299

Emmanuel Botte, Kristina Jelinčić Vučković, Ana Konestra, Ivana Ožanić Roguljić

BU.2016.5020/16

BU.2016.5020/15 P. fine

P. fine

TWW (1st century AD)

BU.2016.5020/5 Sigillée

BU.2016.5020/7 Sigillée

BU.2016.5012/2 Sigillée

ESB 74b (AD 70/75-120)

ESB 60 (AD 50/60-150)

BU.2017.5045.10

ESB 80 (80-150 AD)

BU.2016.HS5/1 Sigillée, diamètre incertain

PRSW 3E? (5th/6th century AD)

PRSW 3 small var./4 similis (5th/6th century AD)

BU.2017.9012.01 BU.2017.9005.01

Hayes 23B (150 - 3rd century AD)

Hayes 9A (100-160AD)

BU.2017.5045.03

BU.2017.9012.03

Hayes 68 (370-425 AD)

Hayes 196 (2nd-3rd century)

BU.2017.5051

Hayes 52B (4th-5th century) Figure 4. Pottery finds (drawing: A. Boisson, E. Botte). (Hayes 1985: 63, 64, 68, 70; cf. Lund 2003: 133 for a later dating of mid-2nd century shapes and in general for the ware production). A partially preserved stamp of rounded rectangular shape on a vessel base bears the letters [..] P? / KOY (Figure 5), possibly belonging to the production of ES B signed by MAP/KOY (cf. Borzić, Eterović Borzić 2015: 37-39 for recent overview of Adriatic distribution).

Dalmatian coast seem to cease or decrease substantially within the second half of the 1st century AD giving way to those from the eastern Mediterranean and north Africa. Later eastern Mediterranean fine wares are present with Phocaean Red Slip ware (LR C), with so far recognised forms Hayes 3 (probably variant E) and a possibly small variant of Hayes 3/4 similis of late 5th/6th AD century (Hayes 1972: 338-339).1

This preliminary outline places the most numerous shapes recognised within a typical repertoire of the wares, but also of the Eastern Adriatic market in each given timeframe (cf. Borzić, Eterović Borzić 2015; Ugarković, Konestra 2018). In fact, Italic imports on the

1 

300

For difficulties in assessing PRSW form 3 types see Vaag 2003.

First results of the excavations of the Roman Villa of Bunje

Figure 6. Sherd of ARS (photo: E. Botte).

Figure 5. Base sherd with MAP/KOY stamp (photo: E. Botte).

Figure 8. Amphorae (drawing: A. Boisson, E. Botte).

Figure 7. Tegula with a stamp (photo: E. Botte).

Figure 9. Coin of Constantine II (337-340) (photo: E. Botte).

African cooking ware is represented with Hayes 23B, 185, 196. African Red Slip ware are the most frequent potsherds on the site (forms Hayes 9A, 9B, 10B, 14A, 16A, 24, 31, 50B, 52B, 58B, 61 A/B4, 61 B, 61C, 68, 91A) (Figure 4 and 6).

Roguljić, Konestra 2017: Fig.1, 453-459). From the same period is Illirian cooking ware. Different dolia were found too, with analogies on Dalmatian islands2 without known origin of production. Coarse ware is present too with pots and lids. This class is poorly studied in Dalmatia and only analogies on the island are observed so far.

Their forms are associated with African and oriental amphorae which attest an intense occupation from the end of the 4th until the 5th or first half of the 6th century AD (forms Keay 27A, LR 1, 2 and 3) (Figure 8) (Hayes 1972; Bonifay 2004). All these forms are already familiar in the Eastern Adriatic, a brief summary of forms and distribution of African imports (ARS, coarse ware and amphorae) is given by Jelinčić Vučković (2019: Fig. 3-5, 525-534).

Tableware of unknown but possibly Adriatic origin is present too. Deeper study of the context and comparison with few well published and excavated sites will hopefully give us some better insight. This panorama is completed by several coins (Figure 9) discovered during the excavations. They are all late coins. Two are clearly identifiable. The first one

Pannonian production is present on the site too with Pannonian Slip ware but also with other decorated tableware (marmorized bowls), dating to the 2nd and 3rd century AD (Ožanić Roguljić 2016: 22; Ožanić

We are grateful to Igor Borzić who kindly showed us sherds of dolia from the island of Korčula for comparison. 2 

301

Emmanuel Botte, Kristina Jelinčić Vučković, Ana Konestra, Ivana Ožanić Roguljić is a bronze of Constantine II (337-340) whose very good conservation indicates that it had a very short circulation period. The second one was on the floor of a room of the residential quarter, and represents Justinian, probably of the Salona mint, and that can be dated from the second third of the 6th century.

Fisković, I. 1981. Ranokršćanski spomenici s otoka Brača. Vjesnik za arheologiju i historiju dalmatinsku 78: 105‒137. Hayes, J. W. 1972. Late Roman Pottery. London: The British School at Rome. Hayes, J.W. 1985. Sigillata orientale B, in Atlante delle forme ceramiche II. Ceramica fine romana nel bacino Mediterraneo (tardo Ellenismo e primo Impero), Enciclopedia dell’arte antica classica e orientale: 49‒70. Roma: Treccani. Hayes, J.W. 1997. Handbook of Mediterranean roman pottery. London: British Museum Press. Hayes, J.W. 2008. The Athenian Agora 32. Fine-ware imports. Princeton: American School of Classical Studies at Athens. Jelinčić, K. 2005. Rustična vila na Bunjama kod Novog Sela na otoku Braču. Vjesnik za arheologiju i povijest Dalmatinsku 98: 121–132. Jelinčić Vučković, K. 2013. Terenski pregled lokaliteta Novo Selo Bunje na otoku Braču. Annales Instituti archaeologici IX: 167–174. Jelinčić Vučković, K., Botte, E. and A. Bertrand, 2017. Archaeological excavation on the Novo Selo Bunje site on the island of Brač, 2016. Annales Instituti archaeologici XIII: 111‒116. Jelinčić Vučković, K. and E. Botte 2018. Archaeological excavation on the Novo Selo Bunje site on the island of Brač, 2017. Annales Instituti archaeologici XIV: 127‒135. Jelinčić Vučković, K. 2019. Le importazioni Africane trovate sull’isola di Brač/Brattia, Dalmazia in: E. Cirelli, E. Giorgi and G. Lepore (eds) Economia e Territorio. L’Adriatico centrale tra tarda Antichità e alto Medioevo (BAR S2926). Oxford: British Archaeological Reports. Lund, J. 2003. Eastern sigillata B: a ceramic fine ware industry in the political and commercial landscape of the Eastern Mediterranean, in C. Abadie-Reynal (ed.) Les céramiques en Anatolie aux époques hellénistiques et romaines. Actes de la Table Ronde d’Istanbul, 23-24 mai 1996, (Varia Anatolica 15): 125-136. Istanbul: Institut Français d’Études Anatoliennes-Georges Dumézil. Martin, A. 1995. Trento — palazzo Tabarelli. Ceramica a pareti sottili, in E. Cavada (ed.) Materiali per la storia urbana di Tridentum: 177‒194. Trento: Archeoalp, Provincia autonoma di Trento. Matijašić, R. 1983. Cronografia dei bolli laterizi della figulina pansiana nelle regioni adriatiche. Mélanges de l’École française de Rome. Antiquité 95/2: 961‒995. Ožanić Roguljić, I. 2016. Tipologija rimske keramike iz Vinkovaca. Zagreb: Institut za arheologiju. Ožanić Roguljić, I. and A. Konestra 2017. Pannonian slipped ware in Dalmatia in: G. Lipovac Vrkljan, B. Šiljeg, I. Ožanić Roguljić and A. Konestra (eds) Zbornik III. međunarodnog arheološkog kolokvija. Rimske keramičarske i staklarske radionice. Proizvodnja i trgovina na jadranskom prostoru. 4. - 5. studenoga 2014 : 453‒460. Zagreb/Crikvenica: Institut za arheologiju, Muzej Grada Crikvenice.

Synthesis It is now clear that the villa, maybe not the one that we can see now but more probably a modest installation, was built during the second half of 1st century AD, attested by the use of Pansiana tiles, the fine ware and the amphorae. The site continued to be occupied until the 3rd century AD, with large villa rustica and its large scale production areas, which is confirmed by the presence of African amphorae and red slip ware. We are not sure yet if there is a continuing occupation of the site during the 4th century until the 5th century AD, or if it was abandoned for some period as it can be observed on other sites of the Dalmatian coast. But a major occupation of the site is confirmed during the 5th and 6h centuries by the important quantity of ceramics, including eastern Mediterranean (LR 1, 2 and 3 amphorae) and African (several ARS and amphorae forms) imports. Recently, we found a grave using two Keay 61 amphorae as container for a child grave, so we can now assure that the site was occupied at least until the 7th century. Given the vicinity of Salona, and the fact that major trade routes did not pass close to the northern shore of the island of Brač, we assume that all imported goods came through this major centre. References Bonifay, M. 2004. Etudes sur la céramique romaine tardive d’Afrique (BAR International Series 1301). Oxford: Archaeopress. Borzić, I., Eterović Borzić, A. 2015. Terra sigillata from Roman Asseria — a contribution to understanding pottery import in Liburnia. Asseria 13: 11‒88. Botte, E., A. Bertrand and K. Jelinčić 2016. Bunje (Novo Selo, Croatie), Campagne de fouilles 2015. Cronique des activités archéologiques de l’École française de Rome  [En ligne], Balkans, mis en ligne le 10 mars 2016. DOI : https://doi.org/10.4000/cefr.1519 Botte, E. and K. Jelinčić 2017. Bunje (Novo Selo, Croatie). Chronique des activites archeologiques de l’Ecole francaise de Rome [En ligne], Balkans, mis en ligne le 16 mai 2017. DOI: 10.4000/cefr.1723 Bulić, F. 1914. Trovamenti antichi a Selca. Bullettino di archeologia e storia Dalmata 37: 105‒106. Ciccarelli, A. 1982. Zapažanja o otoku Braču. Beograd. 302

First results of the excavations of the Roman Villa of Bunje

Parica, M. 2008. Istočnomediteranska keramika iz antičke luke u Pakoštanima. Prilozi Instituta za arheologiju u Zagrebu 25: 81‒96. Pellegrino, E. 2009. Les céramiques communes d’origine orientale dans le Sud de la Gaule au Haut-Empire. Le gobelet Marabini LXVIII, in M. Pasqualini (ed.) Les céramiques communes d’Italie et de Narbonnaise: structures de production, typologies et contextes inédits (IIe s. av. J.-C.-IIIe s. ap. J.-C.), Actes de la table ronde de Naples, 2006: 251‒281. Naples: Centre Jean Bérard. Pellicioni, M. T. 2012. La Pansiana in Adriatico. Tegole romane per navigare tra le sponde. Portomaggiore: Arstudio. Ricci, A. 1985. La ceramica a pareti sottili, in Atlante delle forme ceramiche II. Ceramica fine romana nel bacino Mediterraneo (tardo Ellenismo e primo Impero), Enciclopedia dell’arte antica classica e orientale: 231‒358. Roma: Treccani. Shindler Kaudelka, E. 1975. Die dünnwandige Gebrauchskeramik vom Magdalensberg, (Archäologische Forschungen zu den Grabungen auf dem Magdalensberg 3, Kärntner Museumsschriften 58). Klagenfurt: Verlag des Landesmuseums für Kärnten. Turkalj, K., N. Šegvić, E. Botte and A. Bertrand 2016. Brač Novo Selo Bunje 2015. Annales Instituti archaeologici, XII: 155‒159.

Ugarković, M. and A. Konestra 2018. Hellenistic and Roman fine-wares from Soline bay on st. Clement island near Hvar, in M. Ugarković (ed.) Praetoria longe lateque lucentia. Zbornik radova posvećen Vlasti Begović povodom 65. obljetnice života (ZIA 8): 71–106. Zagreb: Institut za arheologiju. Vaag, L. E. 2003. A closer look at the making of Phocaean Red Slip Ware bowls, in C. Abadie-Reynal (ed.) Les céramiques en Anatolie aux époques hellénistiques et romaines. Actes de la Table Ronde d’Istanbul, 23-24 mai 1996 (Varia Anatolica 15): 203‒207. Istanbul: Institut Français d’Études Anatoliennes-Georges Dumézil. Vrsalović, D. 1960. Kulturno-povijesni spomenici otoka Brača. Brački zbornik 4: 33–161. Zampori Vanoni, M. L. 1987. Ceramica a pareti sottili. in L. Passi Pitcher (ed.) Sub ascia. Una necropoli romana a Nave: 172‒178. Modena: Panini. Zaninović, M. 1968. Neki primjeri smještaja antičkih gospodarskih zgrada. Arheološki radovi i rasprave 4-5: 357–373.

303

Production and distribution of late antique glass and fine wares in Romagna (2nd-7th cent. CE) Tania Chinni, Enrico Cirelli Abstract During the last season of archaeological excavations and surveys, local glass and fine wares production centres exhibiting a long production tradition have been identified in Romagna (Classe, Ravenna, Rimini, Cesenatico, Faenza). These products were distributed on mid- and north Adriatic sites, together with other more successful products, coming from Mediterranean trade routes, from Eastern territories and North Africa. In this paper, we will try to underline the major typologies of products and forms identified in this region and to identify production centres and distribution patterns during Late Antiquity and the early middle Ages.

Key words Romagna, archaeological studies, glassware, fineware, Roman productions, trade routes.

Introduction The south-eastern part of the region Emilia-Romagna is one of the most dynamic territories in the history of the Italian Peninsula: from the 2nd to the 7th century CE this territory received and redistributed a large quantity of products from the whole Mediterranean basin. Among these products, glass objects and fine wares were two of the most valuable artefacts in cultural and social terms. Coinciding with the current administrative districts of Ravenna, Forlì-Cesena and Rimini, Romagna has enjoyed important commercial trade connections since ancient times: Villanovan civilization created here one of its most flourishing cities (Verucchio) which was in direct communication with Northern Europe for amber trade. With the arrival of the Romans and the foundation of Rimini as a Roman colony (268 BCE), Romagna became an important commercial outpost, with ports, roads and channels, which allowed a direct connection with the Po river, taking advantage of the natural aptitude of the territory to the marshland (Dall’Aglio 1990: 35‒36; Gaucci 2013: 79‒82; Malnati 2006:75‒76). To these elements the role attributed to Ravenna during the last period of the Roman Empire must be added. In the 5th century, the city became sedis imperii and the final destination of many important trade routes, with uninterrupted flows of goods (Augenti et al. 2009; Baldassari and Cirelli 2006; Cirelli 2014; Cirelli and Cannavicci 2014). This situation was unchanged until at least the 8th century, when the Mediterranean commercial crisis reduced the volume of trade along these routes, without altering the economic role of this territory. Recent archaeological discoveries show that rural landscape was also active with a renewal

of production in several aspects highlighted by the presence of some ceramic and glass manufactures in various sites. From a historical point of view, this aptitude of Romagna is now reflected in the chrono-typological investigations of the numerous archaeological finds from this area. T. C., E. C. The glass markets The importance of maritime routes in ancient trade has been highlighted by several authors (Hodges 2012; McCormick 2001; Mundell Mango 1996), and finished and semi-finished glass objects travelled along them (Whitehouse 2003). At the commercial port of Classe (Ravenna) the importance of this trade between the 5th and the 8th centuries has been highlighted, thanks to the discovery of a small kiln dedicated to local secondary glass production. The archaeological and archaeometrical studies confirmed that the kiln was used for melting raw glass chunks and wastes imported from the SyroPalestinian area and from the coasts of the Eastern Mediterranean (Augenti and Cirelli 2010: 607‒609; Chinni et al. 2019: 134‒135; Cirelli and Tontini 2010; Maltoni et al. 2015: 22‒23; 2018: 197‒209; Tontini 2006). However, this secondary production could certainly not satisfy the demands and needs of an imperial capital, which Ravenna was between the 5th and 8th centuries CE. Some products had to necessarily come from other production centres. To try to understand the capacity of the local glass markets, in this study, the glass assemblages from two contexts from Romagna were compared: Ad Novas and the Basilica Petriana.

Roman Pottery and glass manufactures (Archaeopress 2022): 304–320

Production and distribution of late antique glass and fine wares in Romagna

79 79

Beakers 16

Bottles Goblets

27

10

2

27

Bowls Lamps

11

4 0

Plates

3 3

Unguentaria 0

Jewellery 0

70

10 6 20

40

60

Basilica Petriana

80

100

Ad Novas

Graph 1. Glass objects in Ad Novas and Basilica Petriana: comparisons of the most attested categories (elab. T. Chinni).

Both mentioned by the 9th century historian Andrea Agnello, the two contexts are just 30 km away from each other, but they are different in type. Ad Novas was probably the early town centre of Cesenatico (Forlì-Cesena district), a small settlement dated to the 1st-7th century CE, identified as a consumption centre and local market for goods (Sami 2010; Sami et al. 2014; 2019). The Basilica Petriana of Classe was the biggest and the most important church in Classe, built in the 5th century CE and severely damaged in the 8th century by an earthquake (Augenti et al. 2009). The glass assemblages from these two contexts have also been compared with the glass objects found in Ravenna and Rimini, the main coastal commercial hubs in this area between the Roman period and Late Antiquity.

fragments of globular beakers Isings 96c with faceted decorations, recovered in two variants: one ‘with rice grain and oval decoration’ and one ‘with vertical lines and double circles’ (Paolucci 1997: 99‒103, 177). The first pattern seems more common along the Mediterranean, with comparisons extending from Egypt to the Rhineland (Chinni 2017: 66‒68). In Italy, some comparisons were identified, for instance, in Bordighera, Aquileia, Luni and Voghenza (Paolucci 1997: 99). The second type appears distributed in the same area, but is less common. The faceted decorations are not extraneous in the previous Roman production, but they appear associated to the form Isings 96c only from the middle of the 3rd century. The origin of these decorations is still uncertain: the production of faceted decorations during the Roman period is often related to Alexandria of Egypt or to the eastern part of the Empire (where they were widespread even after the 8th century CE) (Oliver 1984), but many specimens have also been found in the Rhineland, Pannonia and in the Black Sea region, therefore some authors believe that specialized workshops could be located also in these areas (Rütti 1991: 335).

Comparing the glass assemblages, the two contexts show a certain uniformity, with the prevalence of common shapes for domestic use (Graph1) (Chinni 2017: 43‒44, 61‒62). The most attested category is the beaker (Plate 1), within which the same typologies are recognized: the cylindrical beaker on ring-foot (type Isings 85b), the globular beaker with cut and unfinished rim (type Isings 96), the truncated beaker with bottom raised on a foot (type Isings 109) and the truncated-conical beaker with flat bottom (type Isings 106b-c). These four typologies together can cover a chronological distribution between the end of the 2nd and the end of the 6th century CE.

The same chronological distribution established for the group of beakers is confirmed by bowls (Plate 2), where the majority of finds falls into two types: bowls with more or less deep basin and externally refolded rims (type Isings 44-45 and their Late Antique variant, type Isings 115), and bowls with vertical depressions (type Isings 117). Both were generally used from the 2nd to the 6th century CE.

Although the found typologies are the same, among the finds of the Basilica Petriana it is possible to find some interesting specimens (Plate 1). As in the case of 305

Tania Chinni, Enrico Cirelli

Plate 1. Main forms discussed. From Ad Novas: rims of beaker Isings 85b and beaker Isings 106c; from Basilica Petriana: decorated fragments of beakers Isings 96c, rim of beaker Isings 106b, bottoms of beaker Isings 85b and beaker Isings 109 (author: T. Chinni).

306

Production and distribution of late antique glass and fine wares in Romagna

Plate 2. Main forms discussed. From Ad Novas: rim of bowl Isings 44/45, decorated bottom of bottle Isings 50/51, melon bead, hexagonal bead, blu diamond faceted bead and trilobitenperlen; from Basilica Petriana: rim of bowl Isings 117, decorated bottom of bowl Isings 116, rim of bottle Isings 120/121, rim of bottle Isings 126/127, stem of goblet Isings 111. (author: T. Chinni).

307

Tania Chinni, Enrico Cirelli In this group, at the Basilica Petriana, it is also possible to find a more valuable specimen: a fragment of bottom decorated with scratches, belonging to a bowl type Isings 116 (Plate 2). The bowl, made in colourless glass shows a decoration representing a Greek cross, surrounded by small circles (presumably four) and contained in a larger circle. Around it, some Greek letters can refer the object to the so-called ‘ΠΙΕ ΖHΣHΣ’ vessels. In fact, the inscription recalls the Roman traditional motto ‘Drink! May you live’, which can be found on different types of objects (bottles, bowls and beakers), from the pagan world to the Christian and Jewish traditions (Auth 1996; Avery 1921: 172‒173). The decorative models associate with the inscription are extremely different, synonymous of a non-standardized production, made on specific request by the clients. The most similar model to the fragment recovered in the Basilica Petriana is the bowl Isings 117 found in Dichin (Bulgaria) (Rehren and Cholakova 2014: 87‒88), however, without further supporting data, it seems very difficult to hypothesize a common origin.

confirm this situation, with a small but well attested group of fragments referable to goblets.

The two contexts begin to show differences with the group of bottles (Plate 2). At Ad Novas the number of fragments of square bottles (type Isings 50/51) is more consistent. They are particularly widespread for storage oil and wine between the 1st and the 4th century CE and, sometimes, they are characterized by the presence of a stamps on the bottom, identifying the original workshop. Among the specimens found at Ad Novas a bottom with a flower stamp was identified, which finds comparisons with vessels recovered at Zadar and Starigrad (Fadić and Štefanac 2009: 207) and in the surroundings of Grado (shipwreck of Iulia Felix, first half of the 3rd century CE) (Toniolo 2007).

Blue diamond faceted beads and hexagonal beads often appear together, sometimes also combined with melon beads. However, Ellen Swift hypothesizes a Danubian origin of the blue diamond faceted beads, with imitation workshops distributed from Eastern to Central Europe (Swift 2000: 56). Finally, the case of the trilobitenperlen appears more peculiar. This type of beads are large, in black glass and with two lateral holes for the passage of the wire. Thea Haevernick recognized them in many contexts, but in very limited numbers (Haevernick 1975). According to her studies, the production of trilobitenperlen would be located in the historical regions of Noricum and Rethia (eastern Alpine arc), from where they would have spread in the Mediterranean basin thanks to the commercial mediation of Aquileia. In the Romagna area, trilobitenperlen have recently been reported in Ravenna (Novara and Montevecchi 2000) and Cesena (Fadini and Montevecchi 2001), where an entire bracelet of trilobitenperlen with masks has been recovered from a 4th century burial, in the Monte Garampo excavation (near Cesena). Although limited and occasional, these findings can attest the circulation of specific typologies of products probably external to the local markets.

Considering the forms for daily use, it is clear that Ad Novas and the Basilica Petriana had good access to ordinary quality products, which match perfectly with the dating of the sites and the peculiarities of the two contexts. If at the Basilica Petriana it is the decorations of common types that show the variability of the glass market, at Ad Novas this is observable by shifting the focus on less popular products, such as jewellery (Plate 2). At Ad Novas one mellon bead, one polygonal bead with hexagonal base in green glass, one blue diamond faceted bead and one trilobitenperlen were recovered. The melon beads are very common from Roman period onwards: they show a characteristic opaque light blue colour, with vertical ribs. Considering their greater frequency in Italian contexts, and in particular along the Adriatic coast, Danièle Foy hypothesizes their production in this area (Foy 2010: 466‒468).

Compared to Ad Novas, at the Basilica Petriana we have an opposite situation: the most attested types of bottles are the Isings 120/121 and Isings 126/127 (Isings 1957). These are cylindrical bottles with a funnel rim very common in the Late Antique period, which are mainly distinguished by the type of raised foot (ring-foot in the form 120, stripfood in the 121) and the number of handles (1 handle in the form 126, two handles in the 127). From the 5h century CE, the most produced type in the Mediterranean area is the goblet (Plate 2). However, their reduced presence at Ad Novas suggests that this type of vessels was not particularly requested in the small settlement or that the few recovered individuals could play a different role in the local economy, for instance as lamps. As these objects are deemed more valuable than the more common beakers, the goblets are also often found in funerary and ecclesiastical contexts; this situation has allowed the archaeologists to identify this alternative use (Antonaras 2008: 2324; Foy 2004). The Basilica Petriana contexts seems to

T. C. The fine wares Between the mid-Roman time and the early Middle Ages the main fine wares that circulated at Midwestern Adriatic sites are notably ARS and later Phocaean Red Slip Ware, with limited evidence of some Greek, Cypriot and other sporadic Eastern Mediterranean finds, overall coming from excavations conducted within the Pescara valley. In all these regions (Abruzzo, Marche, 308

Production and distribution of late antique glass and fine wares in Romagna

Figure 1. Mid Adriatic Red Slip Ware, distribution map: 1) S. Basilio (RO); 2) Bondeno (FE); 3) Finale Emilia (MO); 4) Mirandola (MO); 5) Carpi (MO); 6) Reggio Emilia; 7) Congenio (MO); 8) S. Damaso (MO); 9) Spilamberto (MO); 10) Savignano sul Panaro (MO); 11) Budrio (BO); 12) Monterenzio (BO); 13) Villa Clelia (BO); 14) Riolo Terme e grotta di Re Tiberio (RA); 15) Brisighella (RA); 16) Faenza; 17) Russi (RA); 18) Lugo (RA); 19) Ravenna; 20) Classe (RA); 21) S. Zaccaria (Decimano, RA); 22) Forlì; 23) Ad Novas - Cesenatico (FC); 24) Cesena; 25) S. Giovanni in Compito (FC); 26) Galeata (FC); 27) Rimini; 28) Sarsina (FC); 29) Cattolica (RN); 30) S. Pietro in Cotto (RN); 31) Fiesole (FI); 32) Firenze; 33) Colombarone (PU); 34) Pesaro; 35) Fano (PU); 36) Urbino; 37) Tifernum Mataurense (S. Angelo in Vado, PU); 38) Arezzo; 39) Fossombrone (PU); 40) Senigallia (AN); 41) Corinaldo (AN); 42) Suasa (AN); 43) Ostra (AN); 44) Arcevia (AN); 45) Sentino (Sassoferrato, MC); 46) Gubbio (PG); 47) Cortona/Ossaia (AR); 48) Roselle (GR); 49) Settefinestre (GR); 50) Perugia, 51) Osimo (AN); 52) Porto Recanati (MC); 53) Villa Potenza (MC) 54) S. Severino Marche (MC); 55) Urbisaglia (MC); 56) Assisi (PG); 57) Bevagna (PG); 58) Poggio Gramignano (TR); 59) Pennavecchia (TR); 60) Ancona; 61) Novafeltria (RN); 62) Pennabilli (RN); 63) Montelabbate (PU); 64) Colbordolo (PU); 65) Montecalvo in Foglia (PU); 66) Sassocorvaro (PU); 67) Macerata Feltria (PU); 68) Belforte all’Isauro; 69) Fermignano (PU); 70) Peglio (PU); 71) Cantiano (PU); 72) Jesi (AN); 73) Monte Roberto (AN); 74) Cingoli (MC); 75) Potenza Picena (MC); 76) Treia (MC); 77) Fermo; 78) Ager Firmanus (survey); 79) Camerino (MC); 80) Chienti, RIMEM (MC); 81) Domagnano (San Marino); 82) Sassatello (BO); 83) Sestino (AR); 84) Salto del Lupo (FE); 85) Voghenza (FE); 86) Cassana (FE); 87) Mantova; 88) Brescia; 89) Verona; 90) Trento; 91) Altino (VE); 92) Cittanova (VE); 93) Aquileia (UD); 94) Trieste; 95) Parenzo; 96) Pola; 97) Cincelli (AR); 98) Fiorano (MO); 99) Riccione (RN); 100) Misano Adriatico (RN); 101) San Giovanni in Marignano; 102) Santarcangelo di Romagna (RN); 103) Ville Unite, San Pietro in Vincoli (RA) (author: E. Cirelli).

309

Tania Chinni, Enrico Cirelli Romagna), quite at the same time, a local production of fine ware characterized overall by open forms with a trans-regional distribution area started, known as Mid-Adriatic Red Slipped Ware (MARSW), and other classes of materials, with mostly closed forms, such as ‘Crecchio Ware’, ‘Val Pescara Ware’ and Glazed Ware, such as ‘Classe Ware’.

the site of Corinaldo, where a kiln has been identified, but without wasters (Figure 1.41). The floruit of the mid Adriatic distribution, possibly filled the regional fine ware market with 2nd and 3rd century forms, relief decorated, with fine fabric and red slipped with an overpainted black strip. The main forms are bowls similar to Italic and North Italic Sigillata products, such as the forms Brecciaroli Taborelli 1, 4 (Brecciaroli Taborelli 1978) and Maioli 3 (Maioli 1976) (Plates 3.1; 3.4; 7.15). There should also be a close connection with mid Roman Padana production, as testified by Suasa’s finds with rouletted decoration (Biondani 2015). Rouletted decoration has also been found over unpainted late Antique dishes from Monte Torto, possibly linked with the Lamboglia 42 form (Pignocchi 2001: 121‒122). In this period a 2nd-3rd century imitation, or ¸rather we might say ‘inspiration’, of A/D Hayes 33 form started, but also of Hayes 16 and Hayes 27, coming from the ‘A’ repertoire, identified in various mid- Adriatic sites. MARSW is the best represented group of fine wares in the 5th century contexts in sites such as Madonna del Piano di Corinaldo, Suasa, Sentinum and Monte Torto di Osimo. At Madonna del Piano excavations the form Brecciaroli Taborelli 10/17 (Plate 4.6-9) is associated with a Hayes 61C dish (Assenti 2015), and several 5th century African Red Slip forms found at Coiedii’s domus at Suasa belong to the same contexts (Biondani 2014). The same is true for Sentinum where various MARSW forms are in use contemporarily with some Hayes 61B and Hayes 73 forms (Brecciaroli Taborelli 1978: 6-7). Well attested are the forms Brecciaroli Taborelli 10/17 (Monte Torto di Osimo) and 19/22 (Plate 5.14; 7.1-3), now produced without the characteristic brown lines and again associated with various late Antique African Red Slip fragments (Giuliodori 2001: 69).

We do not have at the moment a good closed context with a quantitative analysis for the 3rd century fine wares. But as a general trend we observed that during this period local productions covered almost half of the need for red slip vessels, with a high percentage of imports from both Tunisia and the Eastern Mediterranean. I take as example of a general trend the contexts from the beginning of the 4th century filling of a bath area of a villa, transformed with the construction of a small oratorium (Cirelli et al. 2017), known as the Mausoleum of bishop Severus at Classe. The vessels found there were mostly MARSW, probably produced in Ravenna’s territory (Maioli 1976). The only productive evidence, a mould, comes from its territory in Ville Unite in the southern suburbs of Ravenna (Maioli 2008: 69) (Figure 1.103). But some scholars also think they might have been produced in the nearby Umbrian region (Tortorella 1996: 324) or within the territory of the Roman city of Sarsina (Figure 1.28), where a lot of this fine ware and some rare glazed pottery have been found (Stoppioni 2000: 563; 2015). From these productive centres, vessels reached several rural and urban sites along the Via Aemilia and various settlements in the Po river valley (Bologna, Modena, Reggio Emilia etc.; Labate 2006: 52), distributed all around its ancient delta (e.g. Voghenza, Salto del Lupo and Cassana; Corti 2007: 275, fig. 2). Some products also crossed the main trans-Apennines roads, such as the one that started from Faventia towards Tuscany (Corchia and Zaccagnino 2005: 571). Another possible productive zone for (MARSW) is the city of Tifernum Mataurens (Figure 1.37), where various pottery kilns have been identified, without wasters clearly connecPlate with this fine ware production (Palermo 2006: 108). Technical differences and many profile variants within the vessel repertoires indicate multiple centres of production, with regional and interregional distributions (Mazzeo Saracino 1991: 56). This class of pottery has a proper morphological repertory and mostly an original profile variation and from the 5th to the 7th century contexts we also find some new forms with the same fabrics without overpainted black stripes (Plate 3-7). On the contrary instead, petrographic analysis conducted on some MARSW fragments found at Suasa demonstrated a homogeneous production zone and a chemical composition (Mazzeo Saracino 1996; Nannetti et al. 1996) similar to the older Italic Red Slip Ware, identified for instance in Assisi (Monacchi 1992: 292‒293). One of the possible production centres is on

Another italic production is still active, imitating Hayes 61 dishes (Fontana 1998; 2005: 268, tav. 6.) (Plate 7.4), in sites such as Forum Sempronii (Gori and Luni 1978: 23, n. 39, fig. 13), Suasa (Assenti and Roversi 2010: 285, fig. 6,1; Biondani 2014), Madonna del Piano (Assenti 2010: 462, fig. 4,7), Castelfidardo (Mercando 1979: 150, fig. 62a) and Monte Torto (Pignocchi 2001: 122-13), but also Urbania and Peglio (Biondani 2015), Monte Roberto (Frapiccini et al. 2006: 245-246; fig. 17,2), Cone d’Arcevia (Mercando 1979: 96-97), Potentia (Verreyke 2005: 104-105; fig. 2,3) and Colombarone (Dall’Aglio et al. 1997: 87). Various are also the imitations of Hayes 91, but characterized by a more distinctive italic profile. At Suasa Biondani also identified some Hayes 60 and 67 imitations (Biondani 2014), decorated with very similar stamps, an attempt to imitate the Tunisian originals. This production is also distributed at Colombarone, Fermo and Classe (Biondani 2014; Pupilli 1994; 1996: 47, fig. 42‒43).

310

Production and distribution of late antique glass and fine wares in Romagna

Plate 3. Mid Adriatic Red Slip Ware (MARSW) forms: 1) Brecciaroli Taborelli (BT)1, Sentinum; 2) Maioli (M)2, Ravenna; 3) BT3, Sentinum; 4) BT4, Sentinum; 5) M12, Ravenna; 6) M13, Ravenna; 7) BT5, Sentinum; 8) BT6a, Sentinum; 9) BT6b, Sentinum (redrawn: E. Cirelli).

311

Tania Chinni, Enrico Cirelli

Plate 4. MARSW forms: 1-2) Brecciaroli Taborelli (BT)7, Sentinum; 3) Maioli 5, Sarsina (Stoppioni 2008); 4) M4, Ravenna; 5) BT9, Sentinum; 6-9) BT10, Sentinum; 10) BT11a, Sentinum; 11) BT11b, Sentinum. (redrawn: E. Cirelli).

312

Production and distribution of late antique glass and fine wares in Romagna

Plate 5. MARSW forms: 1) Brecciaroli Taborelli (BT)15, Sentinum; 2-3) BT16, Sentinum; 4) BT16, Sarsina; 5-9) BT12, Sentinum; 10-11) BT14, Sentinum; 12) Maioli (M)14, Sarsina; 13) Stoppioni 5, Sarsina; 14) BT19/22, Sarsina (redrawn: E. Cirelli).

313

Tania Chinni, Enrico Cirelli

Plate 6. MARSW forms: 1-3, 5) BT17, Sentinum; 4) BT15, Sarsina; 6) Stoppioni 7, Sarsina; 7) Maioli (M)1, Ravenna; M14, Ravenna (redrawn: E. Cirelli).

314

Production and distribution of late antique glass and fine wares in Romagna

Plate 7. MARSW forms: 1-3) Brecciaroli Taborelli (BT)19, Sentinum; 4) Imit. Hayes 61A, Ravenna; 5-7) BT20, Sentinum; 8-9) BT22, Sentinum; 10-14) BT23, Sentinum and Ravenna; 15) Maioli 3, Ravenna (redrawn: E. Cirelli).

315

Tania Chinni, Enrico Cirelli Closed forms, particularly pitchers, imported from Abruzzo’s rural settlements or the eastern Mediterranean, have been found at Urbs Salvia, in 5th7th centtury contexts (Fabrini 2000: 143‒144.) and at Suasa (Biondani 2014; Gamberini and Morsiani 2019).

local/regional fine wares, in the first site in respect to the second, where the quantity is high in comparison with minor centres due to the distinction in that case within products decorated with black lines and other local slip wares. This group in Forum Sempronii is much more numerous in respect to Monte Torto di Osimo (Giuliodori 2001: 67) and Urbs Salvia (Fabrini et al. 2006: 331, 335). In the ager Firmanus on a sample of 210 individuals African Red Slip Ware and Mid Adriatic Red Slip Ware are in a ratio of 6:1 (Menchelli and Cerbone 2012: 13, 16). Other fine wares are also distributed, mostly in Monte Torto di Osimo (Giuliodori 2001: 69; Pignocchi 2001: 115) and Urbs Salvia, where ‘painted ware’ is the major fine ware between the 5th and the 7th century (Fabrini et al. 2006: 361). MARSW products reached sites in south Umbria, towards the Tiber river course and some Tyrrhenian sites. On the opposite side, we have the last evidence of rare MARSW finds from the Mantua (Scalari 2004: 91), Brescia (Massa 1999: 120) Verona (Morandini 2008: 342, tav. XXVII.10-13), from Altino’s excavations (Cipriano 2010: 164-164, fig. 6.4), from Cittanova (Borghero, Maring 1989: 150), Aquileia (Cividini 1994: 338, tav. 52), probably Trieste (Maggi 2007: 29, tav 4.43) and from some Istrian sites. Sporadic is also the distribution of those vessels that between the 4th and 5th century reached also Trento (Oberosler 2000), Retia and south Bavaria (e.g. Augsburg), still to verify with archaeometric analyses.

A contemporary production of glazed pottery, pitchers with or without spout and mortars, is attested in Ravenna’s territory (Cirelli 2008). Another important ceramic production identified close to the city walls of Ravenna is the production of lamps, as recently demonstrated by a 6th century mould imitating Atlante X/Hayes II form, found in via Galilei, the same type identified also in Santarcangelo di Romagna, lottizzazione Spina (Stoppioni 2015: tav. 5.1‒2) and a 7th century mould found in the harbour area (Cirelli 2019). The main context of finds is Classe (Maioli and Gelichi 1992) and the favourite distributive pattern is north Italy, but small quantities are also exported towards mid-Adriatic settlements (Gelichi 1985: 97), with a lower intensity after the Conca valley (Maccabruni 1987: 176). Small quantities have been identified at Forum Sempronii (Gori and Luni 1982: 129), in the theatre of Fano (Profumo 2004: 166), at Senigallia (Cirelli et al. 2013), Ancona (Profumo 2004), Sentinum (Gelichi 1985: 100), Helvia Recina’s territory (Gnesi et al. 2007: 132) and in the ager firmanus, with other repertories and possibly from other production centres (Pupilli 1996: 44). During the 6th and the 7th century a growing regional fine wares production, in respect to African Red Slip Ware and Phocean Red Slip vessels, and particularly in the Marche territory are the local red slip wares identified mainly in Pesaro (Ermeti 1998: 611-612) and Classe (Cirelli 2015: 16, fig. 2), with a repertoire of open forms different from the African types and some flanged bowls. In Helvia Recina’s territory a dish with 6th-7th century stamps, similar to other North-Italian vessels has been found (Gnesi et al. 2007: 132). Pitchers and small domestic amphorae with the same fabrics are also testified, such as those from Castel Trosino Lombard cemetery (Paroli and Ricci 2005) and Fossombrone/ piazza Mazzini (Profumo 1997: 59, fig. 8).

E. C. Conclusions The local glass market shows a uniform situation between the 2nd and 7th centuries CE. Comparing the two case studies with the other centres in Romagna, it is possible to confirm the presence of a market strongly oriented towards the same typologies. This is the case of Rimini, where beakers Isings 85b and Isings 96c, bowls Isings 44 and bottles types Isings 120/121 and 126/127 are very common (Biondani 2005; Maioli 1992; Negrelli 2008). In the urban contexts of Rimini, few fragments attest to the limited presence of goblet type Isings 111, which instead are very numerous in the excavation of the so-called Theoderic Palace, in Galeata (Forlì-Cesena), where they were probably used both as drinking vessels and as lamps (Mazzeo Saracino 2004). More decorated objects, such as vessels with faceted decorations recovered both in Ravenna, for instance in the Late Antique contexts in the Basilica and Monastery of Saint Severus excavations (Chinni 2017:136-137), and Rimini, at Piazza Ferrari excavations (Negrelli 2008), and beads are clearly referable to aristocratic or ecclesiastical elites, who also recalled these products from distant areas. A clear example is also the ‘ΠΙΕ ΖHΣHΣ’ bowl: the Greek inscription, the decorative model, as well as the characteristics of the glass, such as colour, imperfections and bubbles, clearly identify

The pattern of distribution of these fine wares is interregional and has its centre between Romagna and Marche (Tortorella 1996: 325, fig. 1), but we do not have clear evidence of production centres or kilns. Difference of fabrics and slips between the finds show the possibility that there were several centre of production, such as Rimini or Santarcangelo, with a ceramic tradition that started in the second half of the 2nd century and lasting until the 7th century (Cirelli 2017: 300, fig. 4). Sporadic, if not absent, is the distribution in the southern Marche. It has been observed that MARSW attestations in comparison with African Red Slip Ware counted in two main settlements in northern Marche — Forum Sempronii and Suasa — indicate a majority of 316

Production and distribution of late antique glass and fine wares in Romagna

it as a product made on request. Although there are still no supporting data regarding the origins of this object, as well as of any other decorated vessel from these contexts, the small number of fragments suggests that they are probably imported from the Rhineland or eastern limes.

Archaeology and Archaeometry. Comparison between western and eastern Mediterranean: 605–615. Oxford: Archaeopress. Augenti, A. and E. Cirelli 2012. From suburb to port: the rise (and fall) of Classe as a centre of trade and redistribution, in S.J. Keay (ed.) Rome, Portus and the Mediterranean: 205‒221. London: British School at Rome. Augenti, A., E. Cirelli and D. Marino 2009. Case e magazzini a Classe tra VII e VIII secolo (scavi 20022005), in G. Volpe and P. Favia (eds) V Congresso Nazionale di Archeologia Medievale (Foggia-Manfredonia, 2009): 138–144. Firenze: All’Insegna del Giglio. Auth, S.H. 1996. Drink may you live! Roman motto glass in the context of Roman life and death, in Annales du 13e Congrès de l’Association Internationale pour l’Histoire du Verre (Pays Bas, 1995): 103–112. Lochem: AIHV. Avery, L.C. 1921. Early Christian gold class. The Metropolitan Museum of Art Bulletin 16(8): 170–175. Baldassari, R. and E. Cirelli 2006. Amphorae and trade in Classe and Ravenna from 5th to 7th century, in J. Zozaya Stabel-Hansen, M. Retuerce, M. A. Hervas and A. De Juan (eds) Actas del VIII Congreso Internacional de cerámica medieval en el mediterráneo (Ciudad Real-Almirago, 2006): 923–927. Ciudad Real: Asociación Española de Arqueología Medieval. Biondani, F. 2005. I vetri, in L. Mazzeo Saracino, M. Zaccaria and F. Biondani (eds) Il complesso edilizio di età romana nell’area dell’ex Vescovado a Rimini: 293–96. Firenze: All’insegna del Giglio. Biondani, F. 2014. Terra sigillata medioadriatica ed altre ceramiche da mensa medio e tardoimperiali, in L. Mazzeo (ed.) Scavi di Suasa I. I reperti ceramici e vitrei dalla domus dei Coiedii: 251‒291. Bologna: Ante Quem. Biondani, F. 2015. Ceramiche fini da mensa in territorio marchigiano fra III e VI secolo: produzioni regionali e importazioni, in E. Cirelli, F. Diosono and H. Patterson (eds) 2015: 253‒275. Borghero, I. and T. Maring 1989. Prime valutazioni cronologico-funzionali sulla presenza romana nell’area di Cittanova. Venezia Arti 3: 148‒152. Brecciaroli Taborelli, L., 1978. Contributo alla classificazione di una terra sigillata chiara italica. Rivista di Studi Marchigiani 1: 1‒38. Chinni, T. 2017. Produzione e circolazione dei manufatti in vetro in Romagna nel Medioevo (V-XV sec.), [Dissertation thesis], Alma Mater Studiorum Università di Bologna. Dottorato di ricerca in Storia culture civilta’, 29 Ciclo. DOI 10.6092/unibo/ amsdottorato/8243. Chinni T., E. Cirelli, S. Maltoni, M. Vandini, A. Silvestri and G. Molin 2019. La diffusione del vetro nell’Adriatico centrale: studi tipologici e dati chimici dal porto di Classe, in E. Cirelli, E. Giorgi and G. Lepore (eds) 2019: 131–138. Cipriano, S. 2010. L’edificio termale di Altino. Quaderni di Archeologia del Veneto 26: 159‒167.

As a result, the inclusion of Romagna in two distinct economic circuits does not appear to be impossible: one maritime, connecting this area to the eastern and Egyptian coasts, and one terrestrial (or mixed), connecting it to central and eastern Europe. The same trend has been demonstrated by mid-Adriatic ceramic production. Different percentages of fine ware imports show us the relevance of Tunisian trade between the 3d and the 4th century with Ravenna as centre of consumption, but from the beginnings of the 5h and the half of the 6th century as a main centre of collection and distribution (Augenti and Cirelli 2012) and again in the second half of the 6th and beginnings of the 7th as a centre of consumption. Romagna’s territory grew up with this new economic/commercial increase. The towns were probably able to receive cargos coming from north Africa and the eastern Mediterranean and play the role of economical intermediary with the countryside, the hinterland and the Po river valley. T. C., E. C. References Antonaras, A. C. 2008. Glass lamps of the Roman and Early Christian periods. Evidence from the Thessaloniki area, in C.-A. Roman and N. Gudea (eds) Trade and local production of lamps from the prehistory until the Middle Age. Lychnological Acts 2. Acts of 2nd International Congress on Ancient and Middle Age Lighting Devices (Zalau-Cluj-Napoca, 2006): 23–30. Cluj-Napoca: Editura Mega. Assenti, G. 2010. Prime note sui materiali dello scavo: il saggio F, in E. Giorgi and G. Lepore (eds) 2010: 459‒472. Assenti, G. 2015. Ceramica comune, vasellame fine da mensa e anfore dalle Marche fra IV e VIII secolo: il caso di Madonna del Piano-Corinaldo (AN), in E. Cirelli, F. Diosono and H. Patterson (eds) 2015: 283‒290. Assenti, G. and G. Roversi 2010. Considerazioni cronologiche sui reperti ceramici da alcuni contesti del Foro, in E. Giorgi and G. Lepore (eds): 275‒286. Augenti, A., F. Boschi and E. Cirelli 2008. Il sito della Basilica Petriana a Classe: dalla diagnostica archeologica allo scavo. Ocnus 18: 103–116. Augenti, A. and E. Cirelli 2010. Classe: un osservatorio privilegiato per il commercio della tarda Antichità, in S. Menchelli, S. Santoro, M. Pasquinucci and G. Guiducci (eds) LRCW 3. Late Roman Coarse Wares, Cooking Wares and Amphorae in the Mediterranean: 317

Tania Chinni, Enrico Cirelli Cirelli E. 2008. Ravenna. Archeologia di una città. Firenze: All’Insegna del Giglio. Cirelli, E. 2014. Typology and diffusion of amphorae in Ravenna and Classe between the 5th and 8th centuries AD, in N. Poulou-Papadimitriou, E. Nodarou and V. Kilikoglou (eds) 2014: 541–552. Cirelli, E. 2015. Dall’alba al tramonto. Il vasellame di uso comune a Ravenna e nel suo territorio tra la tarda Antichità e l’alto Medioevo (III-VII sec.), in E. Cirelli, F. Diosono and H. Patterson (eds) 2015: 13‒19. Bologna: Ante Quem. Cirelli, E. 2017. Settlements and pottery: local versus long-distance trade in Mid-Adriatic Italy (4th-8th c.), in G. Lipovac Vrkljan, B. Šiljeg, I. Ožanić Roguljić and A. Konestra (eds) Rimske keramičarske i staklarske radionice. Proizvodnja i trgovina na jadranskom prostoru (Crikvenica, 2014): 295‒315. Crikvenica-Zagreb: Muzej Grada Crikvenice, Institut za Arheologiju. Cirelli, E. 2019. Ravenna e il mare, in G. Sassatelli and F. Corbara (eds) Classis. Ravenna. Museo della città e del territorio: 88‒99. Milano: Skira. Cirelli, E. and A. Cannavicci 2014. A 6th century dump from Classe (Ravenna), in N. Poulou-Papadimitriou, E. Nodarou and V. Kilikoglou (eds) 2014: 963–974. Cirelli, E., F. Diosono and H. Patterson (eds) 2015. Le forme della crisi. Produzioni ceramiche e commerci nell’Italia centrale tra Romani e Longobardi (III-VIII sec. d.C.). Atti del Convegno, Spoleto-Campiello sul Clitunno, 2012). Bologna: Ante Quem. Cirelli, E., E. Fabbri, A. Juárez and L. Turell 2017. Settore 5 (5000) – Mausoleo Nord, in A. Augenti, N. Christie, J. Laszlowski and G. Ripoll (eds) La basilica di San Severo a Classe. Scavi 2006: 99-108. Bologna: Ante Quem. Cirelli, E., E. Giorgi and G. Lepore (eds) 2019. Economia e territorio. L’Adriatico centrale tra tarda Antichità e alto Medioevo (British Archaeological Report International Series 2926). Oxford: BAR Publishing. Cirelli, E. and S. Tontini 2010. Produzione vetraria a Classe, in M. Vandini (ed.) Riflessioni e trasparenze: diagnosi e conservazione di opere e manufatti vetrosi. Atti del Convegno dell’Associazione Italiana di Archeometria (Ravenna, 2009): 125–133. Bologna: Pàtron Editore. Cividini, T. 1994. Ceramica africana da cucina, in M. Verzár-Bass (ed.), Scavi ad Aquileia, I, L’area ad est del Foro. Rapporto degli scavi 1989-91: 328‒344. Roma: Quasar. Corchia, R. and C. Zaccagnino 2005. Una villa romana a Cincelli (Arezzo). Archeologia Classica 56: 557‒579. Corti, C. 2007. La villa di Salto del Lupo. Un insediamento nell’area del Delta padano tra Età romana e Alto Medioevo, in F. Berti, M. Bollini, S. Gelichi and J. Ortalli (eds) Genti del Delta. Da Spina a Comacchio: 257‒272. Ferrara: Corbo. Dall’Aglio, P. L. 1990. La viabilità romana in Emilia Romagna e nelle Marche settentrionali, in G. Adani (ed.) Vie del commercio in Emilia Romagna Marche: 35–49. Cinisello Balsamo (Milano): Amilcare Pizzi Editore.

Dall’Aglio, P. L., G. Trovabene, M. Destro and M. Vergari 1997. Colombarone (Pesaro), in M.T. Guaitoli (ed.) Scavi e ricerche del Dipartimento di archeologia. Mostra fotografica: 79‒88. Bologna-Imola: Ante Quem. Ermeti, A. L. 1998. Pesaro: ceramica tra VI e VII secolo da sterri urbani, in Saguì 1998: 611‒616. Fabrini, G. M. 2000. L’area del Tempio-Criptoportico ad Urbs Salvia. Risultati preliminari delle campagne di scavo 1995-1999. Picus 20: 113‒158. Fabrini, G. M., M. Giuliodori, S. Forti and C. Capponi 2006. Produzioni ceramiche tarde da contesti del foro di Urbs Salvia. Studi maceratesi 40: 311‒391. Fadić, I. and B. Stefanac 2009. Workshop stamps on square bottles from the Zadar region, in D. Ignatiadou and A. C. Antonaras (eds) Annales du 18e Congres del’Association Internationale pour l’Histoire du verre: 206–210. Thessaloniki: AIHV. Fadini, N. and G. Montevecchi 2001. Un bracciale con ‘Trilobitenperlen’ da una necropoli cesenate, in D. Ferrari (ed.) Il vetro di ogni tempo. Scoperte, produzioni, commerci, iconografie. Atti della V Giornata di Studio (Massa Martana, Perugia, 1999): 51–55. Milano: Tipografia B.T.Z. di Bernareggi & C. Ferreri, D. 2019. Tracce di un quartiere artigianale presso la basilica Petriana a Classe, in E. Cirelli, E. Giorgi and G. Lepore (eds) 2019: 138–144. Fontana, S. 1998. Le ‘imitazioni’ della sigillata africana e le ceramiche da mensa italiche tardo-antiche, in Saguì 1998: 83‒100. Fontana, S. 2005. Le ceramiche da mensa italiche medioimperiali e tardo-antiche: imitazioni di prodotti importati e tradizione manifatturiera locale, in D. Gandolfi (ed.) La ceramica e i materiali di età romana. Classi, produzioni, commerci e consumi: 259‒278. Bordighera: Istituto di Studi Liguri. Foy, D. 2004. Les verres, in A. Ben Abed-Ben Khader, M. Fixot, M. Bonifay and S. Roucole (eds) Sidi Jdidi I: la basilique Sud: 317–329. Roma: Ecole Française de Rome. Foy, D. 2010. Les verres antiques d’Arles: la collection du Musée départemental Arles antique. Paris: Errance. Frapiccini, N., M. Mancini, E. Percossi, G. Pignocchi and H. Verreyke 2006. Nuove acquisizioni su alcuni insediamenti rurali tardoantichi nelle Marche centromeridionali. Studi Maceratesi 40: 233-310. Gamberini, A. and S. Morsiani 2019. Suasa in età tardoantica: nuovi dati dalla via del Foro, in E. Cirelli, E. Giorgi and G. Lepore (eds) 2019: 329‒338. Gaucci, A. 2013. I porti del delta padano nel IV secolo a.C., in F. Boschi (ed.) Ravenna e l’Adriatico dalle origini all’età romana: 71–90. Bologna: Ante Quem. Gelichi, S. 1985. Ceramica invetriata anteriore al Mille dalle Marche. Il gruppo fanese, in La ceramica invetriata tardoromana e alto medievale. Atti del Convegno (Como, 1981): 97‒104. Como: Edizioni New Press. 318

Production and distribution of late antique glass and fine wares in Romagna

Giorgi, E. and G. Lepore (eds) 2010. Archeologia nella Valle del Cesano da Suasa a Santa Maria in Portuno, Atti del Convegno per i venti anni di ricerche dell’Università di Bologna (Castelleone di Suasa-Corinaldo-San Lorenzo in Campo, 2008). Bologna: Ante Quem. Giuliodori, M. 2001. Terra sigillata tarda medio adriatica, in Monte Torto di Osimo. L’impianto produttivo: 67‒70. Falconara: Servizio Editoria della Soprintendenza Archeologica per le Marche. Gnesi, D., S. Minguzzi, U. Moscatelli and S. Virgili 2007. Ricerche sugli insediamenti medievali nell’entroterra marchigiano. Archeologia Medievale 34: 113‒140. Gori, M. and G. Luni 1978. Impianto termale di Forum Sempronii (San Martino del Piano di Fossombrone). Notizie da Palazzo Albani 7/2: 7‒26. Gori, M. and G. Luni 1982. Edificio termale a Forum Sempronii. Nota preliminare. Picus 2: 119‒129. Haevernick, T. E. 1975. Trilobitenperlen. Folia archaeologica 25: 105–129. Hodges, R. 2012. Adriatic Sea trade in an European prospective, in S. Gelichi and R. Hodges (eds) From one sea to another. Trading places in the European and Mediterranean Early Middle Ages. Proceedings of the International Conference (Comacchio, 2009): 209–234. Turnhout: Brepols. Isings, C. 1957. Roman glass from dated finds. Groningen: Wolters. Maccabruni, C. 1987. Ceramica romana con invetriatura a piombo, in Céramiques hellénistiques et romaines, II: 167‒189. Paris: Les Belles Lettres. Labate, D. 2006. Fiorano e la valle del torrente Spezzano. Archeologia di un territorio. Firenze: All’Insegna del Giglio. Maggi, P. 2007. Terre sigillate italiche, in C. Morselli (ed.) Trieste Antica. Lo scavo di Crosada. I materiali: 16‒32. Trieste: Editreg. Maioli, M. G. 1976. Terra sigillata tarda del Ravennate. Rei Cretariae Romanae Favtorvm Acta 16: 160‒173. Maioli, M. G. 1992. Materiali dei vivi e dei morti. III.2 Rimini tardoantica e bizantina: i materiali, in A. Turchini (ed.) Rimini medievale. Contributi per la storia della città: 237–304. Rimini: Ghigi. Maioli, M. G. 2008. La ceramica invetriata a Sarsina, in A. Donati (ed.) Storia di Sarsina. I. L’età antica: 693‒712. Cesena: Stilgraf. Maioli, M. G. and S. Gelichi 1992. La ceramica invetriata tardo-antica e altomedievale dall’Emilia Romagna, in L. Paroli (ed.) La ceramica invetriata tardoantica e altomedievale in Italia, Atti del Seminario (Certosa di Pontignano, 1990): 215‒278. Firenze: All’Insegna del Giglio. Malnati, L. 2006. Rimini prima di Rimini, in F. Lenzi (ed.) Rimini e l’Adriatico nell’età delle guerre puniche. Atti del Convegno (Rimini, 2004): 75–95. Bologna: Ante Quem. Maltoni, S., T. Chinni, M. Vandini, E. Cirelli, A. Silvestri and G. Molin G. 2015. Archaeological and archaeometric study of the glass finds from the

ancient harbour of Classe (Ravenna- Italy): new evidence. Heritage Science 3.1(1): 1–29. Maltoni, S., F. Gallo, A. Silvestri, M. Vandini, T. Chinni, A. Marcante, G. Molin, and E. Cirelli 2018. Consumption, working and trade of Late Antique glass from north Adriatic Italy: An archaeometric perspective, in D. Rosenow, M. Phelps, A. Meek and I. Freestone (eds) Things that Travelled: Mediterranean Glass in the First Millennium CE: 191–214. London: University of City of London. Massa, S. 1999. Le imitazioni di ceramiche fini da mensa tra tarda età romana e alto medioevo, in G.P. Brogiolo (ed.) S. Giulia di Brescia: gli scavi dal 1980 al 1992. Reperti preromani, romani e alto medievali: 119‒124. All’Insegna del Giglio: Firenze. Mazzeo Saracino, L. 1991. Lo studio dei materiali ceramici da Suasa: stato attuale e prospettive, in Archeologia nelle valli marchigiane. Misa, Nevola e Cesano: 125‒128. Perugia: Electa Editori Umbri. Mazzeo Saracino, L. 2004. Circolazione dei manufatti e dati cronologici dallo studio dei materiali, in S. De Maria (ed.) Nuove ricerche e scavi nell’area della villa di Teodorico a Galeata, Atti della Giornata di Studi (Ravenna, 2002): 135–156. Bologna. Mazzeo Saracino, L., F. Biondani and M.C. Nannetti 1996. Primi risultati di una ricerca su terre sigillate tarde di area medioadriatica. Ocnus 4: 137‒158. McCormick, M. 2001. Origins of the European economy. Communications and commerce AD 300-900. Cambridge: Cambridge University Press. Menchelli, S. and O. Cerbone 2012. Ceramiche fini nell’ager Firmanus (Fermo, Marche meridionali). FOLD&R 268:1‒26. Mercando, L. 1979. Marche. Rinvenimenti di insediamenti rurali. Notizie di Scavi 1979: 89‒296. Monacchi, D. 1992. Terra sigillata chiara italica da un contesto urbano di Assisi. Bollettino di Foligno 16: 289‒310. Morandini, F. 2008. Le ceramiche fini e le terre sigillate di produzione non africana, in G. Cavalieri Manasse (ed.) L’area del capitolium di Verona. Ricerche storiche e archeologiche: 331‒342. Verona: Soprintendenza per i Beni Archeologici del Veneto. Mundell Mango, M. 1996. Byzantine Maritime Trade with the East (4th–7th Centuries). ARAM Periodical 8.1: 139‒163. Nannetti, M. C., V. Minguzzi, N. Morandi, and L. Mazzeo Saracino 1996. Parametri mineralogici e geochimici utili per la caratterizzazione e la distinzione in gruppi di provenienza di ‘terre sigillate’ di Suasa (AN). Museologia scientifica 13, Suppl.: 399‒417. Negrelli, C. 2008. Rimini Capitale. Strutture insediative, sociali ed economiche tra V e VIII secolo. Firenze: All’Insegna del Giglio. Poulou-Papadimitriou, N., E. Nodarou and V. Kilikoglou (eds) 2014. LRCW 4: Late Roman coarse wares, cooking wares and amphorae in the Mediterranean: archaeology 319

Tania Chinni, Enrico Cirelli and archaeometry: the Mediterrean: a market without frontiers (British Archaeological Reports International Series 2616). Oxford: Archaeopress. Novara, P. and G. Montevecchi 2000. ... in agro decimano... : per un catalogo del patrimonio storico archeologico del territorio a sud di Ravenna. Ravenna: Montanari Edizioni. Oberosler, R. 2000. Attestazioni di Terra Sigillata tarda Medio-Adriatica a Trento, in G.P. Brogiolo and G. Olcese (eds) Produzione ceramica in area padana tra il II secolo a.C. e il VII secolo d.C.: nuovi dati e prospettive di ricerca. Convegno internazionale (Desenzano, 1999). Mantova: SAP. Oliver, A. J. 1984. Early Roman faceted glass. Journal of glass studies 27: 35–58. Palermo, L. 2006. I reperti mobili, in M. Tornatore (ed.) Una domus con mosaici a Tifernum Mataurense (Sant’Angelo in Vado): 97‒112. Urbania: Arti grafiche Stibu. Paolucci, F. 1997. I vetri incisi dall’Italia settentrionale e dalla Rezia nel periodo medio e tardo imperiale. Firenze: All’Insegna del Giglio. Paroli, L. and M. Ricci 2005. La necropoli altomedievale di Castel Trosino. Firenze: All’Insegna del Giglio. Pignocchi, G. 2001. Ceramica comune e ceramica da fuoco, in Monte Torto di Osimo. L’impianto produttivo: 111‒130. Falconara: Servizio Editoria della Soprintendenza Archeologica per le Marche. Profumo, M. C. 1997. Archeologia funeraria e insediamento nelle Marche nei secoli VI e VII, in L. Paroli (ed.) L’Italia centro-settentrionale in età longobarda, Atti del Convegno (Ascoli Piceno, 1995): 55‒79. Firenze: All’Insegna del Giglio. Profumo, M. C. 2004. La ceramica altomedievale nelle Marche, in S. Patitucci Uggeri (ed.) La ceramica altomedievale in Italia, Atti del V Congresso di Archeologia Medievale (Roma, 2001): 163‒176. Firenze: All’Insegna del Giglio. Pupilli, L. 1994. Il territorio del Piceno centrale in età romana. Impianti di produzione. Villae rusticae. Villae di otium. Ripatransone (AP): Maroni. Pupilli, L. 1996. Il territorio del Piceno centrale dal Tardoantico al Medioevo: dall’otium al negotium. Ripatransone (AP): Maroni. Rehren, T. and A. Cholakova 2014. Glass supply and consumption in the late Roman and early Byzantine site Dichin, northen Bulgaria, in D. Keller, J. Price and C. Jackson (eds) Neighbours and successors of Rome: traditions of glass production and use in Europe and the Middle East in the later 1st millennium AD: 83– 94. Oxford: Oxbow. Rütti, B. 1991. Die romischen Glaser aus Augst und Kaiseraugst. Augst: Romermuseum Augst. von Saldern, A. 1980. Ancient and Byzantine glass from Sardis. Cambridge (MA): Harvard University Press.

Saguì, L. (ed.) 1998. Ceramica in Italia: VI-VII secolo, Atti del Convegno in onore di John W. Hayes (Roma, 1995). Firenze: All’Insegna del Giglio. Sami, D. 2010. Archeologia a Cesenatico: lo stato delle ricerche e prospettive per il futuro. Studi Romagnoli LXI: 11–33. Sami, D., E. Baldi, A. Booth, T. Chinni and L. Toniolo 2014. Ad Novas-Cesenatico. From Roman road to Late Antique wooden structures. An interim report on the evaluation test pits and excavation at Ca’ Bufalini, 2006. Fold&r 318: 1‒20. Sami, D. and Christie, N. 2019. The Roman Road and the mansio of Ad Novas at Cà Bufalini (Cesenatico, FC), in E. Cirelli, E. Giorgi and G. Lepore (eds) 2019: 172–181. Scalari, C. 2004. Le terre sigillate, in G.P. Brogiolo (ed.) Gli scavi al Battistero di Mantova (1984-1987): 89‒94. Mantova: SAP. Stoppioni, M. L. 2000. Ceramiche medio-adriatiche, in M. Marini Calvani (ed.) Aemilia. La cultura romana in Emilia Romagna dal III secolo a.C. all’età costantiniana: 563‒567. Venezia: Marsilio. Stoppioni, M. L. 2008. La Sigillata tarda di Sarsina, in A. Donati (ed.) Storia di Sarsina, I. L’età antica: 713‒762. Cesena: Stilgraf. Stoppioni, M. L. 2015. Romagna sudorientale e appenninica: imitazioni fittili di ceramiche da mensa e di stoviglie metalliche da portata nella tarda Antichità, in E. Cirelli, F. Diosono and H. Patterson (eds) 2015: 63‒74. Swift, E. 2000. The end of the Western Roman Empire. An archaeological investigation. Stroud: Tempus. Toniolo, A. 2007. ‘... pallentia solphurata fractis permutat vitreis...’. Il carico di rottami di vetro del relitto di Grado, in A. M. Visser Travagli and D. Ferrari (eds) Il vetro nell’alto Adriatico. Atti delle IX Giornate Nazionali di Studio (Ferrara, 2003): 57–69. Imola: La mandragora. Tontini, S. 2006. Vetri e produzione vetraria a Classe (Ravenna) in età tardoantica e altomedievale. Bulletin de l’association pour l’Antiquité tardive 15: 45–55. Tortorella, S. 1996. Considerazioni sulla sigillata tarda dell’Italia centro-settentrionale. Studi Miscellanei XXX (1991‒1992): 323‒335. Verreyke, H. 2005. Late Roman pottery in the Potenza Valley. First results regarding late Roman Potentia (Porto Recanati, Marche, Italy). Rei Cretariae Romanae Favtorvm Acta 39: 103‒107. Whitehouse, D. 2003. ‘Things that travelled’: the surprising case of raw glass. Early Medieval Europe 12: 301–305.

320

The imports of high quality early Imperial glass in Romula (Pannonia) Irena Lazar Abstract The article presents and gives preliminary overview about the rich and interesting glass assemblage of early imperial coloured mould made fine ware discovered in Roman road and custom station Romula (Slovenia) in the period from 2001 to 2005. The complete study of the material is in progress but due to large amount of glass finds we would like to point out some results and observations. Monochrome translucent coloured fine ware and monochrome opaque coloured fine ware, characteristic for the first three quarters of the 1st century AD, have distinguished angular and carinated profiles and are further characterized by strong colours and high standards of craftsmanship. Among the Romula material the emerald green glass is predominant for all forms of fine wares (plates, cups, pyksides) while the opaque vessels are made of light blue glass (cups). There is also a significant group of black appearing mould-made glass vessels (plates, cups, bowls, pyksides).

Key words Roman glass, mould-made glass, coloured fine ware, opaque glass, black glass, Pannonia, Romula, Slovenia.

During the 2nd and 1st century BC the Helenistic states were incorporated into the Roman Empire. As stated by D. Grose (1989: 241) this furnished the impetus for the creation of the Roman Glass industry that flourished from the Augustan time to the end of the Empire. The glass craft developed quickly and spread all over the Roman Empire. It became a strong and independent Roman enterprise which achieved the maturity within less than a century. This was possible on one side because of the expansion of trade during the early imperial period, the rapid spread of the glass factories and of the knowledge how to make raw glass and glass products and the combination of the old and newly discovered production techniques (Grose 1986: 66; 1989: 242). But in the years of development and growth the Roman glass making and glass industry faced several significant and important changes in the course of the 1st century BC and AD. During the Roman imperial period glass objects were manufactured in secondary glass workshops all over the Empire (Stern 1999) by the use of imported raw glass. Archaeological findings identified the Syro-Palestinian coast as the main area for primary glass production (Nenna 2007; Gorin-Rosen 2012). Free blowing slowly replaced the time consuming moulding techniques, glass ware suddenly became a commodity available to all classes in their everyday life. Numerous manufacturing centres operated in the Mediterranean and in the western part of the Empire, some specialized in mass production while the others developed in more specialized workshops for high quality products (Stern 1999).

We would focus on the products of the later, recently excavated in Slovenia, since not many archaeological excavations and settlement research in the area give opportunity to study high quality early imperial glass. The article presents and gives preliminary information about the very rich glass assemblage of early imperial coloured mould made fine ware discovered in Roman road and custom station Romula in the period from 2001 to 2005. The complete study of the material is in progress but due to the interesting and large amount of glass finds we would like to point out some preliminary results and observations. The extensive rescue excavations started during the motorway construction on the route Ljubljana – Novo Mesto – Brežice, which corresponds also with the former Roman via publica (in early development also via militaris) Emona – Neviodunum – Siscia. Ribnica near Brežice in Lower Carniola (Slovenia), a Roman post and customs station known as Romula, is one of the most extensive sites investigated along the new motorway route in the recent period (Breščak 2004; Lazar 2020). The name of the station is known from Tabula Peutingeriana. With its control and supply role, the station’s location on the route of the main Roman road (via publica Aquileia – Emona – Neviodunum – Siscia) was an act of careful strategic planning. Due to its position on the route from the valley of the Krka and Sava rivers to the Pannonian plain, it also supervised the river traffic along the Sava (Savus). The settlement grew up along the Roman road on the narrowest part of the terrace between two small hills that certainly served as observation points (Figure 1). Furthermore, crammed between the Sava River and the northernmost slopes of

roman pottery and glass manufactures (Archaeopress 2022): 321–330

Irena Lazar

Figure 1. Map with the site of Ribnica - Romula (Pannonia) and Roman communication routes (prepared by A. Preložnik).

the Gorjanci Hills, it occupied the narrowest point of transition between the hilly area of Dolenjska (Lower Carniola) and the flatland of Pannonia. The ideally positioned road and custom station Romula, which was marked also on Itinerarium Antonini, was a link between Italy and the Balkan Peninsula (Lazar 2020).

One such is a partly preserved one-handled beaker signed by Ennion (Lazar 2004: 53, fig. 17; 2005: 40), made from yellowish glass. The decoration consists of pillars, palmettos, concentric circles and a part of a star (?) and can be defined as Near Eastern style of decoration (Lightfoot 2015: 36, fig. 26; Lehrer 1979: Pl. VI: 2). In the centre, inside a square frame, there is the inscription in Greek MNHΘH OAΓO PAZΩN (mnesthe ho agorazon – Let the buyer be remembered!). The second inscription panel ENNI / ΩNEΠ / OIHCE N (Ennion made me), which we can presuppose on the basis of the analogies, is not preserved. The preserved inscription is of a type J, and together with the decoration (Lightfoot 2015: 28) compares this beaker with a beaker of greenish glass found in Soluntum (Solanto) Italy (now Palermo Museum in Sicily) (De Bellis 2004: 129–133, figs. 6, 8) and with a beaker from The Shlomo Moussaieff Collection (Lightfoot 2015: 92). These completely preserved onehandled beakers were blown into a mould with the same pattern as the one from Ribnica (De Bellis 2004: 133).

New excavations uncovered the extensive settlement and necropolis area. The structures date from the late 1st century BC to the end of the 4th or even the first years of the 5th century AD. The heart of the settlement was located on a terrace of the Sava river to the east of a brook and archaeologists researched official and production buildings, several segments of the Roman road and western and eastern necropolis with 170 graves (Lazar 2020; Figure 2). It is, therefore, not surprising that the site yielded rich and variegated glass material from the settlement and necropolis, confirming thus the lively trade connections and routes leading through the station.

The recent finds prove that Ennion products were widely distributed also on the west, including Spain, Greece, France (McClellan 1983: 76), Slovenia and Croatia (Lazar 2005; 2022; Buljević 2012; 2015).

During the excavation we have already singled out some important discoveries from the group of glass material. It is worth to mention some individual finds which are of particular importance for the site and the area, before we present the group of early imperial mould-made glass.

Another find is a fragment with mould-blown decoration similar to Ennion products (Lazar 2022). 322

The imports of high quality early Imperial glass in Romula (Pannonia)

Figure 2. The excavated area of Romula (prepared by A. Ogorelec; after Lazar 2020).

This is a fragment of a conical beaker made of bluish, naturally coloured glass. There is no inscription field on it, but the decoration shows similarities with glasses signed by Ennion, and with other signed beakers. We can recognize four registers of decoration divided with horizontal lines of pearls. Two registers with floral decoration show in lower line elements of a Dionysiac style decoration known on two-handled Ennion cups (Lightfoot 2015: 36, fig. 27), the upper line decoration consists of columns with garlands. The fragment ends with vertical flutes on upper part and a net pattern on the lower part, decoration typical for mould-blown products made by Ennion and/or Aristeas.

monochrome opaque coloured fine ware were defined as family III – Roman cast monochrome vessels (Grose 1989: 254). Due to the angular and carinated ceramiclike profiles the group was also known as so-called glass with ceramic profiles, as it was defined by Berger (1960: 24–30) in his study of the glass from Vindonissa. This group of fine wares is characterized by angular lathe-turned forms, strong colours and very high standards of craftsmanship. The group has no stylistic ties with the production of glass in the late Hellenistic period. It was observed, that several of the colours, like emerald green, are the inventions of the newly established Roman glass industry (Grose 1991: 2). The angular and carinated forms reveal the influence of the contemporary metal ware, bronze or even the silver ware of the Julio-Claudian era (Grose 1989: 254). The distribution of these products is western, since the majority of the examples were discovered on Italian and other European archaeological sites; that’s why this group is understand as a distinctive product of the early Roman glass industry on the west and the Italian production centre is assumed or more precisely the factory or a cluster of workshops of the Roman Italy.

The largest group of the material we are now working on is the early imperial mould-made glass. We are talking about two distinctive groups of fine wares in early Imperial contexts (Grose 1991: 2). The Group one consists of coloured fine wares, made by mouldpressing and the Group two of colourless fine wares, made by casting or mould-pressing (Lierke 2009) . The group of colourless wares was presented at the AIHV conference in Istanbul in 2018 and will be published at the conference acta (Lazar 2021), so here we would concentrate on coloured fine wares.

These glasses first appeared in the first quarter of the 1st century AD, perhaps ongoing to about AD 60 (Grose 1989; 1991; 2017). Only a generation or two after its appearance the coloured fine ware obviously fell from fashion. The group of monochrome coloured vessels is very homogeneous and Grose defined ten principal

When publishing the impressive collection of early ancient glass from the Toledo Museum of Art, D. F. Grose defined six families of the mould-made vessels within the group of the Early Roman Glass (1989: 244– 261). Monochrome translucent coloured fine ware and 323

Irena Lazar

Figure 3. Forms of mould-made coloured translucent glasses – plates (drawings: J. Kobe).

324

The imports of high quality early Imperial glass in Romula (Pannonia)

Figure 4. Forms of mould-made coloured translucent and opaque glasses – cups and pyksides (drawings: J. Kobe).

325

Irena Lazar forms – plates, cups or bowls (hemispherical, concave and carinated), pyksides and less common large plates or trays (Grose 1989: fig. 135; 1991: 2), but today some additional individual forms can be added. In Romula the finds of this group of glass are surprisingly numerous and forms do not appear only as individual finds but in numbers. We have recognized eight of ten principal forms: cups and plates in various forms and pyxis with subtypes (Figures 3, 4) and we would single out the main characteristics of the assemblage.

and some of them include high quality mould-blown glass and amber objects. The comparisons to the material from well dated sites or strata are known from Magdalensberg (Czurda-Ruth 1979: 65–91), where all fragments presumably date before AD 45, when the site was abandoned, with few examples attributed to late Augustan levels (1979: 70– 71). In Vindonissa dated examples are concentrated in Tiberian and early Claudian contexts, some continuing into Neronian period (Berger 1960: 24–30). Some fragments are also known from Salona (Buljević 2016: 64; dark green and blue glass vessels).

Monochrome translucent coloured fine ware For the group of the translucent monochrome fine wares the colours of translucent emerald green, cobalt and dark blue, peacock blue are very distinctive. Emerald green and peacock blue are the colours developed by the Roman glass industry and used almost exclusively for this class of glass vessels. Among the Romula material the emerald green glass is predominant for all forms of fine wares.

At Cosa, dated examples of translucent and opaque fines wares come from five deposits (Grose 2017: 81). The earliest is deposit IV (before 25–15 BC) associated with the reoccupation of the town in Augustan times. It is significant that the opaque ware in light blue and white is represented in this deposit. As Grose argues, they can represent some of the earliest products of the Roman glass industry or they were products imported from the eastern Mediterranean (Grose 2017: 81–82).

Plates (15 vessel fragments; Figure 3: 1-7) with constricted, straight or convex walls are made predominantly of translucent dark green (various shadows of emerald green) glass and only two of them are made of dark blue or cobalt blue glass. Some of the plates fragments have shallow grooves on the base and several of them have a small circle in the centre where they were affixed to the lathe.

Deposits V and VI are dated to the period of c. 25–15 BC to c. 40–45 AD, forms of carinated bowls, pyxis and a bowl with out-turned rim are represented in these strata (Grose 2017: 81, pl. 8: 134, 136, 137, 139; pl. 9: 154), some of them come from a shop storeroom destroyed in c. 40–45 AD. Vessels of this group presented in deposits VII and VIII (prior to c. 50–55 AD) prove the production of this group of glass vessels until the mid-1st century or even later, but not after the last quarter of the 1st century (Grose 2017: 82).

Cups (17 vessels fragments; Figure 4: 1- 6; 6) vary in size and forms; they are hemispherical, with convex walls or have carinated walls. Some of the fragments have shallow grooves on the walls or on the inner side of the base, but none of the bases shows attachment remains on the base on the outside. They are again made predominantly of translucent deep or emerald green glass, only two fragments belong to a dark blue cup.

Monochrome opaque coloured fine wares Glasses of this group are made of opaque light blue glass. There is also one very deteriorated base fragment of red glass, but it is questionable if it can be included in this group. The opaque glass vessels are represented only with one form, a bowl, with two variants: they have hemispherical or carinated walls (Figure 4).

All fragments of pyksides (5 vessels fragments; Figure 4) in the group of translucent coloured ware are made of deep green glass. Their diameter and height vary, from 5.6 to 11 cm in diameter and from 4 to 4.5 cm in their height. Some examples have a distinct circle on the base (S 1050, list 2), where the vessel was attached to the lathe.

The date of the opaque vessels is first half of the 1st century AD. Judging from the glass products found in Cosa (Grose 2017: 81) in the deposit dated before 25/15 BC their production may have begun already at the last decades of the 1st century BC. Prototypes for the Roman monochrome opaque vessels may have been the Hellenistic products of the 1st century BC also made in white and light blue opaque glass (Grose 2017: 82).

The presented group of glass has some individual comparisons on Slovenian sites from Emona (insula 32; Plesničar 1983: pl. 23: 12, 16, 19, 22, 25), Celeia and Poetovio (unpublished) but so far nowhere these glasses appeared in such quantity as in Romula. It is also significant that in Romula these are all settlement finds and those vessels were not used or added as grave goods, although graves from the mid-1st century and on were excavated on the western necropolis of the site

Black (appearing) glass vessels (Figure 5) Although black glass vessels, the term is used as a terminus technicus accepted by glass specialists (Cosyns 326

The imports of high quality early Imperial glass in Romula (Pannonia)

Figure 5. Forms of mould-made black glass from Romula (after Lazar 2019).

327

Irena Lazar The vessels from Romula are made of deep violet or deep brown glass that looks like black. All the vessels belong to the mould-made products and several of them have distinctive angular lathe-turned forms (Lazar 2019). The production of the black appearing glass vessels in the 1st century AD reflects the popularity of the coloured mouldmade fine wares produced in the period of Julio-Claudian dynasty (Grose 1991: 3; Cosyns and Fontaine 2009: 83). Among the forms we have recorded plates, bowls any pyksides. Plates are presented in various types, they have straight or convex walls, slightly profiled rim, their base is often decorated with shallow grooves of concentric circles on the inner side (Figure 5: 1–5). The bowls have out-turned walls and slightly profiled or out-turned rim (Figure 5: 6, 7). The group of pyksides or vessels with cylindrical walls is represented with several individual products that vary mostly in size. Their diameter measures from 5 to 10 cm and the outside of their base is often decorated with shallow grooves of concentric circles (Figure 5: 8–10). Cosyns and Fontaine (2009: 80, tab. 1, 2) have divided the forms of the mould-made black appearing glass vessels from the 1st century to several groups and forms; further on Cosyns defined 14 principal forms with subgroups for the mould-made glass (2011: 48, fig. 33), and 19 forms for the free blown vessels (2011: 52, fig. 34). It is possible that this group will develop with new finds and publications. Within the glass material from Slovenia we can already define new types of the so far defined individual forms.

Figure 6. Photo of the deep green cup with constricted walls, Romula (photo: A. Ogorelec).

2015), are often studied separately from other glasses, I have included it in this article, since several fragments of mould-made vessels were recorded among the glass assemblage. But what is black glass and how we define it? Thanks to the research of P. Cosyns in the past decades the Roman black glass from Britannia, Gallia Belgica and Germania Inferior was studies within several projects (Cosyns and Hanut 2005; Cosyns and Fontaine 2009; Cosyns 2011). It is known that black glass vessels are present in the 1st century AD and again from the middle of the 2nd to the middle of the 3rd century (Cosyns 2015: 190). Black glass jewellery is introduced later and is in use well into the 5th century AD (Cosyns et al. 2006; Jelinčić 2009). In addition to the chronological, geographical and typological distribution analysis of these products, also the archaeometry and study of the glass compositions is extremely valuable (Cagno et al. 2014).

Conclusion remarks The presented group of glass has some individual comparisons on Slovenian sites from Emona (insula 32; Plesničar 1983: pl. 23), Celeia and Poetovio, but so far nowhere these glasses appeared in such quantity as in Romula. It is also significant that these are all settlement finds and those vessels were not used or added as grave goods, although graves from the mid-1st century and on were excavated on the western necropolis of the site and some of them include high quality mould-blown glass and amber objects. Although the majority of the presented coloured wares comes from destroyed strata, their use in Romula is confirmed with coin finds, two thirds of them ranging from the Augustan period to the mid-2nd century. But comparisons with the individual finds from dated deposits in Roman Empire confirm the use of the coloured translucent and opaque ware similar to the finds from Romula in the period of the first three quarters of the 1st century.

The black glass vessels were produced with various techniques. In the 1st century (from Tiberian to the Flavian period) they were mould-made and free blown, but in the second period (mid-2nd to the last quarter of the 3rd century) these glasses were free blown, with very few exceptions (Cosyns 2015: 191). As written by J. Bayley (1999: 90, 92), the black glass is the glass that is so densely coloured that it is opaque and appears black. Only with the assistance of strong transmitted illumination in this black glass one colour will become visible (Cosyns 2015: 191).

Our preliminary observation could be that the whole group of the fine wares in Romula (coloured and 328

The imports of high quality early Imperial glass in Romula (Pannonia)

colourless) reflects the Roman glass market taste of the 1st century. The individual demand and taste for luxury glasses in the Empire soon offered and produced a new group of fine glass tableware – colourless instead of deep, strong colours. In Romula the demand for luxury ware continues and colourless glass overtook the households and tables at the second half of the 1st and 2nd centuries (Lazar 2021).

survey. Annales du 16e Congrès de l’AIHV: 113–118. London: Association internationale pour l’histoire du verre. Cosyns, P. And S. Fontaine 2009. La vaisselle en verre d’apparence noire dans les provinces occidentales au 1er siècle ap. J. C., in K. Jansens, P. Degryse, P. Cosyns, J. Caen and L. Van’t dack (eds) Annales du 17e Congrès de l’AIHV: 80–87. Antwerp: Association internationale pour l’histoire du verre. Cosyns, P. et al. 2006. Black glass vessels and jewellery in the Roman Empire: a work in progress, in G. Creemers, B. Demarsin and P. Cosyns (eds) Roman Glass in Germania Inferior. Interregional Comparisons and Recent Results: 30–41. Hasselt: Provinciaal GalloRomeins Museum Tongeren. Czurda-Ruth, B. 1979. Die römischen Gläser vom Magdalensberg. Kärntner Museumsschriften 65. Klagenfurt: Verlag des Landesmuseums fur Kärnten. Gorin-Rosen, Y. 2012. Remains of a Glass Industry and Glass Finds from Horbat Biz’a.Atiqot 70: 49–62. Grose, D. F. 1984. Glass Forming Methods in Classical Antiquity: Some Considerations. Journal of Glass Studies 26: 25–34. Grose, D. F. 1986. Inovation and Change in Ancient Techologies: The Anomalous Case of the Roman Glass Industry, High technology Ceramics: Past, Present and Future,Ceramics and Civilisation, vol. 3: 65–79. Westerville, Ohio. Grose, D. F. 1989. Early Ancient Glass. Core-formed, Rodformed, and Cast Vessels and Objects from the Late Bronze Age to the Early Roman Empire, 1600 B.C. to A.D. 50. New York, Toledo. Grose, D. F. 1991. Early Imperial Roman Cast Glass: The Translucent Coloured and Colourless Fine Wares, in K. Painter and M. Newby (eds) Roman Glass. Two Centuries of Art and Invention: 1–18. London: The Society of Antiquaries of London. Grose, D. F. 2017. The Hellenistic, Roman and Medieval glass from Cosa. Rome: Univesity of Michigan Press, American Academy in Rome. Jelinčić, K. 2009. Rimske staklene narukvice kao prilog poznavanju antičkog Orešca / Roman glass bracelets as a means to understanding Orešac in antiquity. Opuscula Archaeologica 33: 105–124. Lazar, I. 2003. Rimsko steklo Slovenije / The Roman Glass of Slovenia. Opera Instituti archaeologici Sloveniae 7. Ljubljana: Slovenska akademija znanosti in umetnosti, Znanstvenoraziskovalni center. Lazar, I. 2004. Spiegelungen der Vorzeit. Antikes Glass in Slowenien / Odsevi davnine. Antično steklo v Sloveniji, in I. Lazar, Die Römer – Glas, Ton, Stein (Exhibition catalogue): 11–82. Celje – Ptuj – Maribor: Regional Museum Celje. Lazar, I. 2005. Ennion Beaker. Instrumentum 21: 40–41. Lazar, I. 2019. Črno steklo – da ali ne : rimski vrč iz groba 4/1988 na ptujski obvoznici in nove najdbe posod iz

The study of the material is in progress, together with other small finds, necropolis and the site architecture. We hope to add significant new information about the presented glass material and its use by the end of the project. References Bailey, J. 1999. Notes on the composition of coloured glasses, in M. Guido (ed.) The Glass beads of AngloSaxon England c. AD 400–700: 89–93. London: Woodbridge, Boydell Press. Berger, L. 1960. Römische Gläser aus Vindonissa. Basel: Gesselschaft pro Vindonissa. Breščak, D. 2004. Romula – Ribnica na Dolenjskem, in D. Prešeren (ed.) The Earth Beneath Your Feet. Archaeology on the Motorways in Slovenia. Guide to Sites: 237–240. Ljubljana: Zavod za varstvo kulturne dediščine Slovenije. Buljević, Z. 2012. Traces of Ennion and Aristeas in the Roman Province of Dalmatia (Exhibition Catalogue). Split: Archaeological Museum Split. Buljević, Z. 2015. Ennion and Aristeas Glasware Found in the Roman Province of Dalmatia, in C. Lightfoot (ed.), Ennion: Master of Roman Glass (Exhibition Catalogue): 61–67. New York: Metropolitan Museum. Buljević, Z. 2016. Ranorimsko salonitansko i tilurijsko stakleno posuđe u kontekstu nalaza staklenog posuđa s područja provincije Dalmacije. Unpublished PhD dissertation, Zagreb: University of Zagreb. De Bellis, M. 2004. Le coppe da berre di Ennione: Un aggiornamento. Aquileia nostra 75: 121–190. Cagno, S. et al. 2014. Deeply colored and black-appearing Roman glass: a continued research. Journal of archaeological science 42: 128–139. Cosyns, P. 2011. The Production, Distribution and Consumption of Black Glass in the Roman Empire during 1st to 5th Century AD. An Archaeological, Archaeometric and Historical Approach. Unpublished PhD dissertation, Free University Brussels. Cosyns, P. 2015. Beyond the Channel! That’s quite a different matter. A comparison of Roman black glass from Britannia, Gallia Belgica and Germania Inferior, in J. Bayley, I. Freestone and C. Jackson (eds) Glass of the Roman World: 190–204. Oxford & Philadelphia: Oxbow Books. Cosyns, P. and F. Hanut 2005. Black glass of second to third-century date in northern Gaul: a preliminary 329

Irena Lazar na pogled črnega stekla. Zbornik Pokrajinskega muzeja Ptuj: 29–43. Ptuj: Pokrajinski muzej Ptuj – Ormož. Lazar, I. 2020. Romula – Ribnica pri Brežicah / Ribnica near Brežice, in J. Horvat, I. Lazar and A. Gaspari (eds), Mala naselja rimske dobe v Sloveniji / Small Settlements of the Roman Period in Slovenia. Opera Instituti archaeologici Sloveniae 40: 387402, Ljubljana: Slovenska akademija znanosti in umetnosti, Znanstvenoraziskovalni center. Lazar, I. 2021. New finds of Roman high quality glass from Romula (Pannonia), in Annales du 21e Congrès de l’AIHV: 149-160. Istanbul. Lazar, I. 2022. Rimska čaša s podpisom mojstra ENNION-a - prva najdba odkrita v Sloveniji. Studia Univeristatis Hereditati 10/1: 15-26. Lehrer, G. 1979. Ennion. A First Century Glassmaker (Exhibition catalogue). Tel Aviv. Lierke, R. 2009. Die nicht-geblasenen antiken Glasgefässe / The non-blown ancient glass vessels. Offenbach/Main: Deutsche Glastechnische Gesellschaft.

Lightfoot, C. 2015. Ennion: Master of Roman Glass (Exhibition Catalogue). New York: Metropolitan Museum. Mandruzatto, L. 2008. Vetri antichi del Museo Archeologico Nazionale di Aquileia. Corpus delle Collezioni del Vetro in Friuli Venezia Giulia 4. Udine. McClellan, M. C. 1983. Recent Finds from Greece of First-Century A.D. Mold-Blown Glass. Journal of Glass Studies 25: 71–78. Nenna, M. D. 2007. Production et commerce du verre à l’époque impériale: nouvellesdécouvertes et problématiques. Facta 1: 125–147. Plesničar, Lj. 1983. Starokrščanski center v Emoni. Katalogi in monografije 21. Ljubljana: Narodni muzej. Stern, E. M. 1999. Roman Glassblowing in a Cultural Context. American Journal of Archaeology 103: 441–484.

330

Glass beads from the archaeological site of Burnum (Amfiteatar, Campus) Ivana Jadrić-Kučan Abstract Diverse glass finds including beads have been found in the systematic archaeological research of the Roman military camp Burnum (the area of the amphitheater (Amfiteatar) and army training ground (Campus)). Recovered examples exhibit great diversity of forms and ornaments and they mostly date to the first half of the 1st century originating from all more important workshop centers at the time: Egyptian circle (Alexandria), Syro-Palestinian circle (Sidon), northern Italy region (Aquileia).

Key words Burnum, Dalmatia, beads, Medusa, Roman army.

Figure 1. Topography of roman military camp Burnum (photo: S. Ferić, microlocation tags by I. Jadrić - Kučan).

Roman military camp Burnum is located in the area of the village of Ivoševci near Kistanje, on a location of exceptional strategic importance controlling the crossing over the Krka river. Ancient sources (Pliny. HN, 3, 138) and recovered archaeological finds (epigraphic monuments, architectural monuments, various small finds) suggest that the agglomeration of Burnum consisted of different components: Roman legionary camp with the accompanying territory (in rectangular shape, dimensions 330 x 295 m), amphitheater and

campus (SW of the legionary camp), auxiliary castella (located east and north of the legionary camp) and later municipium (from the time of Hadrian) that developed from canabae and the settlement of the Liburnian Burnistae (Gradina near Puljani) (Patsch 1895: 379–393; 1900: 71–85; Reisch 1913: 112–135; Zaninović 1968; Zabehlicky-Scheffenegger and Kandler 1979) (Figure 1). Systematic archaeological research in Burnum started from 2003 and it is ongoing, conducted by the Drniš

roman pottery and glass manufactures (Archaeopress 2022): 331–339

Ivana Jadrić-Kučan City Museum and the Department of Archaeology of the University of Zadar in cooperation with the University of Bologna (Cambi et al. 2006; 2007; Campedelli 2007a: 57–78; 2007b: 52–64; 2007c: 131–138; Borzić 2008: 91–101; Miletić 2010: 113–142; Miletić 2011: 263–279; Borzić 2011: 279–291; Jadrić 2011: 361–377; Glavaš 2012: 89–103; Borzić and Jadrić-Kučan 2013a: 517–522; Borzić and Jadrić-Kučan 2013b: 523–535; Glavičić and Miletić 2013: 157–172; Borzić et al. 2013; 2014; 2016; Cambi et al. 2014; Jadrić-Kučan 2014: 351–357; Jadrić-Kučan and Zaninović 2015). Diverse glass finds including beads have been found in the systematic archaeological research of the Roman military camp Burnum (Borzić 2011: 279–291; Jadrić 2011: 361–377; Borzić and Jadrić-Kučan 2013a; JadrićKučan 2014: 351–357). Presented beads were found in the excavations of the amphitheater in Burnum and the army training ground, in the layers that were probably formed by filling the cracks in the terrain in order to build the mentioned objects in the mid-1st century. They are related to Claudius’ expansion of the camp principium when distribution of the existing structures was changed or new ones were built, the terrain was cleared and new foundations were dug which had to result in huge amount of waste usable in building the amphitheater and campus that were formed in the process. Other possibility should also be considered, that material might have been transported from the dump used in the first half of the 1st century that was probably located in immediate vicinity.

Figure 2. Tombstone of the cavalryman T. Flavius Bassus from Hofheim (from M. C. Bishop & J. C. N. Coulston, 2006).

Egyptian lotus bud. They were very widespread and used over a long period of time. They were made of different materials: faience, frit, glass, or even gold. They were usually strung in necklaces, but other possibilities of their use are not a priori excluded, such as bracelets or decorations of pinheads. In case of necklaces we know of examples of strings of beads of different lengths that consisted of melon beads only in different colours and sizes, or they could have been mixed with other kinds of beads. On the tombstone of the cavalryman T. Flavius Bassus from Hofheim they decorate horse equipment (Bishop and Coulston 2006: 13, fig. 4) (Figure 2). A bronze chain with melon beads was found around the neck of a horse skeleton in Krefeld-Gellep in Asciburgium (Siepen 2007) suggesting that they were used on horse harness due to their apotropaic meaning.

Beads belong to the oldest objects made of glass since as early as the 2nd millennium BC glass was an important material for their manufacture, perhaps even the most important. Some beads had symbolic and apotropaic meaning, and that is exactly why they were made of glass, material ascribed with magical powers. They were used in a number of ways and had different functions, the decorative one being the most important. Necklaces of beads were worn around the neck, hands and ankles. They were also used as hair adornments, decorations on clothes and very frequently to decorate fibulae. Although beads are primarily associated with the jewelry for ladies, they are also found in children’s graves where they functioned as amulets, to protect them during their lifetime and in the afterlife. Melonshaped beads were used to decorate horse equipment, or even arms. Examples from Burnum exhibit great diversity of forms and decoration, so they were classified into several types.

1.1. Melon beads made of frit (Plate 1: 1) These beads were made of material called frit,1 a mixture of silica grains, lime, soda and coloured metal copper oxides fused at relatively low temperatures of 600-800°C (Spaer et al. 2001: 35).2 This grainy material would retain its natural pastel colour (bluish-green, turquoise or Egyptian blue, light green), and it was particularly suitable for making small objects, decorations or amulets (Riha 1990: 80). Due to their intensive light blue colour they were worn as amulets.

1. Ribbed beads/melon beads (Plate 1: 1, 2)

In the scholarly literature we find different versions: German term ‘Kieselkeramik’ (‘gravely ceramic’), English term ‘blu frit’, Croatian term ‘frit staklo’. 2  Ancient faience or frit is not alwyas easy to distinguish from glass that had lost its translucency either because other elements were added or due to atmospheric agents. That is why ancient glass ornaments are often incorrectly described as made of faience or frit. 1 

In the scholarly literature they are referred to as melon beads (Melonenperlen) as their shape resembles a melon, or Lotosperlen, since they were modelled after a blue 332

Glass beads from the archaeological site of Burnum (Amfiteatar, Campus)

1

2

3

Pl. 1 Glass beads from Burnum (photo by: I. Jadrić-Kučan, J. Zaninović)

Plate 1. Glass beads from Burnum (photo: I. Jadrić-Kučan, J. Zaninović).

333

Ivana Jadrić-Kučan Custom of wearing blue beads to ward of the ‘evil eye’ can still be found in Egypt.

Seventeen specimens of this bead type have been found in the Burnum region, five of which show traces of burning of different intensity. Parallel finds in the territory of Croatia can be found in Tilurium and Narona (Buljević 2014: 264, cat. no. 215–217; Glučina 2016: 220, cat. no. 328).

Method od manufacture of this type of beads is still unclear. Emilie Riha cites the thesis by Van Der Sleen who believed that they were made in a sort of mold similar to modern ice cream spoon in which ribs were modelled (Riha 1990: 80). However analysis of the beads from the Roman colony of Augusta Raurica in Switzerland and Roman legionary camp Kaiseraugst (castrum Rauracense) in Austria indicates that they have quite irregular grooves, and ribs that have curved semicircular shape. Their tapering suggests that they were not made in a mold, but the ribs were modelled subsequently on the bead body (Riha 1990: 758–1145, Taf. 33–35). On the basis of an example from Magdalensberg in Austria Barbara Czurda-Ruth believes that a metal rod was inserted into the finished bead body and then this ball was rolled over a ribbed surface, and in that way ribs were impressed into the bead body (CzurdaRuth 1979: 201).

2. Spherical/oval, annular and globular beads with applied horizontal trails (Plate 1: 3) Beads from this group have identical decoration as the previous group, melon beads with applied trails, but they are different as their body can be spherical/ oval (type 2.1), annular (type 2.2) or globular (type 2.3). Although there are similarities between them, it is difficult to say if they originate from the same workshop center on the basis of such small number of specimens (Czurda-Ruth 1979: 188, Type E: 2049). All these beads date to the early period of the Roman Empire and they were made by wrapping a glass trail around a metal rod, whose width determined the opening of the bead, then molten glass trail in lighter color was applied on the bead. After cooling the metal rod was simply pulled out.

Original place of their production was Egypt (Riha 1990: 80). They were so common in the western part of the Roman Empire in the 1st century that it is reasonable to assume that there were workshops for their production in western Europe as well (Czurda-Ruth 1979: 201; Riha 1990: 80). They are most common in the second half of the 1st century and in the early 2nd century (Riha 1990: 80–82), and they were used throughout the Roman imperial period until the 5th century (Guido 1978: 100; Arveiller-Dulong and Nenna 2011: 212, cat. no. 295). 3

Five examples of this type have been found in Burnum, two of which burnt. They were made of opaque black or dark blue glass with an applied white or yellow wavy line that encircles the bead body. Parallel find in the territory of Croatia was found in Tilurium (Buljević 2014: 264, cat. no. 214). 3. Cylindrical beads with featherlike ornament (Plate 2: 1)

Eight examples of this type of beads were found in Burnum. In Croatia comparable finds were unearthed at the following sites: Aenona, Tilurium (Borzić 2013: 57, cat. no. B72; Buljević 2014: 265, cat. no. 218–220; 2017: 213–214. T. 11., cat. no. 180–185; Glučina 2016: 221–222, cat. no. 329–342).

These are rather big beads decorated with an ornament that is referred to in the scholarly literature as ‘a piuma di uccello’ (‘bird feather’), ‘a pettine ‘ (‘like a comb’), ‘fenicio’ (‘Phoenician’) (Mandruzzato 2008: 159). Bead body is dark (black, dark brown, light green), and the decoration is in some lighter colour (white, light blue, yellow).

1.2. Melon beads with applied trails (Plate 1: 2)

Beads of this type were made in two ways. Soft glass trails in different colours were wrapped alternately one next to another around the metal rod, they were fused after heating and the resulting decoration was visible throughout the cross-section of the bead. The other way was to wrap the molten glass trail in lighter colour around the finished cylindrical body of a dark bead (whose ends where rounded in fire), and then the bead was rolled on some smooth and hard base connecting the lighter trail with the bead body (Czurda-Ruth 1979: 190–191; Spaer et al. 2001: 52, 99–100). In both ways of manufacture wrapped glass trail was pulled sharply in shape of the letter V while hot by some metal object, three to four times on the bead body, resulting in a featherlike decoration.

They resemble beads made of faience or frit in shape, but they are made of transparent or opaque glass and have only few thick ribs4 decorated with applied trail in lighter colour than the rest of the body. Trails spread orthogonally on the bead ribs or in waves around the bead body. Place of their manufacture is unknown, and they date to the period from the 3rd century BC to the 2nd century (Riha 1990: 82, 1146–1151, Taf. 36; Mandruzzato 2008: 157). Finds of this type of beads in Great Britain date to the period from the Flavians to the Antonines, and then they are missing only to reappear in Late Antiquity. 4  Thicker ribs suggest different production techniques: it is possible that the finished ball was rolled over ribbed slate or ribs were made using some glassmaking instrument. 3 

334

Glass beads from the archaeological site of Burnum (Amfiteatar, Campus)

 

 

1

2

3 Plate 2. Glass beads from Burnum (photo: I. Jadrić-Kučan, J. Zaninović).

335

Ivana Jadrić-Kučan They were found in the area of Carthage where they are dated to the mid-3rd and 2nd century BC (ArveillerDulong and Nenna 2011: 106, 166 a, b). Their production center was in Alexandria, and in Aquileia for the area across the Alps (Czurda-Ruth 1979: 190; Mandruzzato 2008: 159). Roman beads are dated to the period from the 1st to the 5th century but they were also present later in the Byzantine period. Seven specimens of this type of beads were found in Burnum, three of which are burnt. They are made of opaque black, dark brown or light green glass with featherlike decoration made of white glass trail. In Croatia they were also found in Tilurium and Sisak (Buljević 2014: 263–264, cat. no. 212–213; 2017: 214, T. 11, Kat. br. 187; Baćani 2017: cat. no. 97).

Figure 3. Necklace made of bronze wire with in situ black bead (photo: I. Jadrić-Kučan).

Four examples of this type of beads were found in the area of Burnum, they were made of opaque blue, black or dark red glass with white dots containing even smaller blue or red dots.

4. Beads with ‘specks’ or ‘coloured dots’ (Plate 2: 2) These are beads with spherical/oval or annular form decorated with ‘specks’ or ‘coloured dots’ that were more or less regular and applied randomly.

6. Monochrome, undecorated beads (Plate 3: 1) The oldest beads were monochrome. Later with more massive glass production around the mid-2nd millennium BC bead producers from Mesopotamia and Egypt started making bichrome and polychrome beads (Spaer et al. 2001: 57).

Place of manufacture cannot be determined. They are dated from the 1st to the 4th century (Czurda-Ruth 1979: 188; Templemann-Maczyńska 1985: 48, Taf. 4; Riha 1990: 84, 1175, Taf. 37; Mandruzzato 2008: 158). Twenty three examples of this type of beads were found in Burnum. They differ in combinations of colours: there are white beads with blue dots and dark blue beads with colourful dots (white, yellow, light blue). Comparable find in the territory of Croatia was found in Tilurium (Buljević 2014: 265–266, cat. no. 221, 223–226; 2017, 214, T. 11, cat. no. 186).

Monochrome undecorated beads with smooth surface from Burnum can be further subdivided on the basis of their form into globular/biconical, spindle-shaped and annular beads. 6.1. Globular/biconical beads These beads were popular with all age groups and they were very widespread. Place of their production is difficult to determine, they are dated to the period from the 1st to the 5th century (Riha 1990: 86, Taf. 38, 1219, 1222; Mandruzzato 2008: 160). Three examples of this type were found in Burnum, made of opaque glass in dark blue, light blue and black colour. Particularly interesting among them is the black bead found in situ on the fragment of a necklace made of bronze wire with round cross-section (Figure 3).

5. ‘Eye’ beads (Plate 2: 3) These are beads whose bodies are decorated with ‘eye’ motifs that symbolically represent human face. Pronounced apotropaic qualities are ascribed to this kind of beads and from prehistory they played an important role in magical beliefs as they were supposed to ward off ‘evil eyes’ since it was believed that the eyes reflect envy and jealousy, and therefore could even be harmful. At the beginning of the 1st millennium BC Phoenician city of Tyre was famous for production of this type of beads, and they were exported across the Mediterranean and the neighbouring Near East (Panini 2007: 41).

6.2. Spindle-shaped beads These are very small elongated spindle-shaped beads, about 10 mm long and 3 to 4 mm wide. They were made of different materials: gold, ivory, opaque and transparent glass. Place of their production is difficult to determine, they are dated from the 1st to the 4th century (Riha 1990: 87, 1227–1231, Taf. 38).

These beads were very popular which resulted in a number of their variants. Pupil of the eye usually consists of a darker part which was encircled by one or more circles in contrasting colour in relation to the bead body, but there are different variants in its manufacture, just like with the rest of the eye (Spaer et al. 2001: 77–98; Mandruzzato 2008: 156–157).

An example of a spindle-shaped bead made of dark blue translucent glass was found in the region of Burnum.

336

Glass beads from photo by: I. J the archaeological site, J. Zaninović of Burnum Pl. 2 Glasss beads from m Burnum (p Jadrić‐Kučan ć) (Amfiteatar, Campus)

 

1 Pl. 3 Glasss beads from m Burnum (p photo by: I. JJadrić‐Kučan, J. Zaninovićć) 

Plate 3. Glass beads from Burnum (photo: I. Jadrić-Kučan, J. Zaninović).

these beads is a miniature human face usually depicted frontally with dark eyebrows, square nose, dark eyes, brownish-red lips and white complexion. Such face can be dated quite precisely between the 1st century BC and the 1st century which is relatively short period of their production. Some beads are modelled as heads with bare shoulders and a necklace around the neck (Goldstein 1979: 274, cat. no. 821; Stern and SchlickNolte 1994: 410, cat. no. 149; Genett 2014: 14) suggesting that probably a female person was depicted, while sex on the other face beads cannot be determined.

6.3. Annular beads Beads of this type come in different variants of annular form, from narrow thin rings with a wide opening, to more massive rings with small opening that come close to globular form and ‘globular beads’. Annular beads were popular in the era of the Celts, and later with the Romans who produced them in a different way – by wrapping a molten glass cane around a metal rod (Riha 1990: 85). They were very popular throughout the Roman imperial period. Majority was made of translucent glass, and only a small number of opaque glass or other materials such as bronze, ivory. Four examples of this type were found in Burnum.

7.1. Medusa bead (Figure 4) Bead finds from Burnum comprise an example of polychrome flat discoid mosaic bead depicting a female face. The bead was perforated on the sides, and the ends are flatly cut. Face of a woman is depicted with white complexion and square nose, dark eyes and eyebrows, red lips and stylized hair sections. It is lined with three wreaths, the first in red, the second in white and the final and the thickest in green. This face, unnatural and ugly, is believed to represent Greek mythological creature Medusa or Gorgon, with ‘black’ fields between the hair sections depicting stylized snake heads growing out of her hair.

7. Beads of mosaic glass These beads come in different forms (globular, spherical/oval, tubular, flat, etc.) and with different decorations made in the mosaic technique that can be subdivided on the basis of decoration (beads with chessboard motif, beads with chessboard motif and a face, flower(s) beads, face beads, beads with depiction of Medusa’s head, beads with an orant in the central field, etc.). They were probably manufactured in Egypt or Syria, and they are dated from the 1st century BC to the mid-1st century.

According to Greek mythology Medusa’s appearance would render any onlooker into stone. Therefore Medusa beads have special apotropaic importance as they ward off any evil which is why they were very popular motif on the Roman amulets and beads. Similar examples of beads in Croatia are kept in the Archaeological

The best known type of the mosaic beads are beads with a face depiction. They are either globular made of several plaques of the mosaic cane placed around the middle of the pellet or flat as a simple mosaic plaque that had a tiny perforation on the top. Main characteristic of 337

Ivana Jadrić-Kučan and colours of beads suggests how popular they were among the Roman soldiers. Beads with elaborate mosaic decoration were particularly expensive and therefore rare, in other words they were inaccessible to ‘regular’ soldiers as attested by the polychrome flat discoid mosaic bead with a depiction of the face of Medusa or Gorgon, mythological creature from Greek mythology warding off evil. Therefore it was a popular motif on Roman amulets and beads giving them special apotropaic qualities. Although there are many open questions regarding beads, mentioned conclusions indicate that the study of this interesting category of archaeological finds can result in a completely different image of everyday life of a Roman soldier in the Roman province of Dalmatia. References Arveiller-Dulong, V. and M. D. Nenna 2011. Les verres antiques du Musée du Louvre III. Louvre – Paris. Baćani, I. 2017. Staklo Siscije. Sisak. Bishop, M., C. and J. C. N. Coulston 2006. Roman Military Equipment, From the Punic Wars to the Fall of Rome. Oxford. Borzić, I. 2008. Ennion čaše iz Burnuma. Archaelogia Adriatica 2: 91–101. Borzić, I. 2011. Sarius šalice iz Burnuma / Sarius cups from Burnum, in G. Lipovac Vrkljan, I. Radić Rossi and B. Šiljeg (eds), Rimske keramičarske i staklarske radionice, Proizvodnja i trgovina na jadranskom prostoru, Zbornik 1. međunarodnog arheološkog kolokvija, Crikvenica, 23. – 24. listopada 2008: 279–291. Zagreb/Crikvenica: Institut za arheologiju, Muzej Grada Crikvenice. Borzić, I. 2013. Antika, Arheološka istraživanja okoliša crkve Sv. Križa u Ninu (Katalog izložbe): 27–59. Nin. Borzić I. and I. Jadrić-Kučan 2013a. Staklo puhano u kalup iz Burnuma, in M. Sanader, A. Rendić-Miočević, D. Tončinić, I. Radman-Livaja (eds) Proceedings of the XVIIth Roman Military Equipment Conference: Weapons and Military Equipment in a Funerary Context (XVII Roman Military Equipment Conference, Zagreb, 24th – 27th May, 2010): 517–522. Zagreb: Arheološki muzej u Zagrebu. Borzić I. and I. Jadrić-Kučan 2013b. Opskrba hranom u rimskom legijskom logoru u Burnumu – analiza amfora, in M. Sanader, A. Rendić-Miočević, D. Tončinić, I. Radman-Livaja (eds) Proceedings of the XVIIth Roman Military Equipment Conference: Weapons and Military Equipment in a Funerary Context (XVII Roman Military Equipment Conference, Zagreb, 24th – 27th May, 2010): 523–535. Zagreb: Arheološki muzej u Zagrebu. Borzić, I., M. Glavičić, Ž. Miletić and J. Zaninović 2013. Burnum, Kuhinja rimskog vojnika. Šibenik. Borzić, I., N. Cambi, M. Glavičić, Ž. Miletić, I. JadrićKučan and J. Zaninović. 2014. Arheološka zbirka

Figure 4. Medusa bead (photo: J. Zaninović).

Museum in Zagreb,5 Archaeological Museum in Split, and worldwide in the Corning Museum of Glass, in the collection of Ernest Wolf, and in the Dobkin Collection (Goldstein 1979: 274, cat. no. 820; Stern and SchlickNolte 1994: 414–415, cat. no. 153–154; Spaer et al. 2001: 124, cat. no. 207–208; Buljević 2002a: 318, cat. no. 7, 8; Marin 2002: 33). Conclusion Recovered examples from Burnum exhibit great diversity of forms and ornaments. On the basis of the context of finds they are dated mostly to the first half of the 1st century and they originate from all important workshop centers at the time: the Egyptian circle (Alexandria), Syro-Palestinian circle (Sidon), northern Italy (Aquileia). Their significant number in the area of Burnum (74 pieces) suggests that beads had an important decorative or apotropaic function even at the military sites. This refers in particular to the melon beads made of frit that were worn as amulets because of their intensive light blue colour. A number of these beads were found in the Roman legionary camps indicating they were a popular decorative/status symbol of Roman soldiers. Further on they were found in necklaces around horse’s necks suggesting that some of them might have been used as ornaments on horse harness. Although glass was getting cheaper from the Hellenistic period, and thereby more accessible to different social classes, glass beads were still fashionable artifacts among well-to-do people. Wide range of sizes Two examples from Sisak are in permanent display of the collection of Greek and Roman antiquities of the Archaeological Museum in Zagreb. 5 

338

Glass beads from the archaeological site of Burnum (Amfiteatar, Campus)

Burnum. Javna ustanova ‘Nacionalni park Krka’, Drniš – Šibenik. Borzić, I., N. Cambi, M. Glavičić, Ž. Miletić, I. Jadrić-Kučan and J. Zaninović. 2016. The Burnum, Archaeological collection. Javna ustanova ‘Nacionalni park Krka’, Drniš – Šibenik. Buljević, Z. 2002. Toaletni i nakitni oblici, Longae Salonae I, II: 313–326; 154–156. Split. Buljević, Z. 2014. Stakleni inventar, Tilurium III., Istraživanja od 2002.–2006. godine: 225–292. Zagreb. Buljević, Z. 2017. Stakleni inventar, Tilurium IV., Arheološka istraživanja od 2007. do 2010. godine: 175–238. Zagreb. Cambi, N., M. Glavičić, D. Maršić, Ž. Miletić and J. Zaninović 2006. Amfiteatar u Burnumu, stanje istraživanja 2003. – 2005. Drniš – Šibenik – Zadar. Cambi, N., M. Glavičić, D. Maršić, Ž. Miletić and J. Zaninović 2007. Rimska vojska u Burnumu / L’esercito romano a Burnum. Drniš – Šibenik – Zadar. Cambi, N., M. Glavičić, D. Maršić, Ž. Miletić and J. Zaninović 2014. Burnum, Imperatores militesque. Šibenik. Campedelli, A. 2007a. Il progetto Burnum (Croazia), Ocnus 15: 57–78. Bologna. Campedelli, A. 2007b. Relazione della campagna di rilievo e scavo nel castrum romano di Burnum, Drniš (18 – 30 Agosto 2006), N. Cambi et al., 2007, Rimska vojska u Burnumu: 52–64. Drniš – Šibenik – Zadar. Campedelli, A. 2007c. Sul campo: il municipium romano di Burnum (Drniš, Croazia), Groma1: 131–138. Bologna. Czurda-Ruth, B. 1979. Die römischen Gläser vom Magdalensberg, Archäologische forschungen zu den Grabungen auf dem Magdalensberg 6. Klagenfurt. Gennett, A. V. 2014. Glass Beads, Selections from The Corning Museum of Glass. Corning, New York. Glavaš, V. 2012. Antefixes with representation of tragic mask from Burnum and Asseria, Archeologia Adriatica 5: 89–103. Zadar. Glavičić, M. and Ž. Miletić 2013. Arhitektura amfiteatra u Burnumu, Histria antiqua 22: 157–172. Pula. Glučina, T. 2016. Arheološki muzej Narona, Katalog stalnog postava, Katalozi i monografije 1. Vid. Goldstein, S. M. 1979. Pre-Roman and Early Roman Glass in The Corning Museum of Glass. Corning, New York. Guido, E. 1978. The Glass Beads of the Prehistoric and Roman periods in Britain and Ireland. London. Jadrić, I. 2011. Staklene rebraste zdjelice iz Burnuma / Small glass ribbed bowls from Burnum, in G. Lipovac Vrkljan, I. Radić Rossi and B. Šiljeg (eds), Rimskekeramičarskeistaklarske radionice, Proizvodnja i trgovina na jadranskom prostoru, Zbornik 1. međunarodnog arheološkog kolokvija, Crikvenica, 23. – 24. listopada 2008: 361–377. Zagreb/Crikvenica: Institut za arheologiju, Muzej Grada Crikvenice. Jadrić-Kučan, I. 2014. Ulomak staklene bočice s reljefnim prikazom glave Meduze, 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: 351–357. Zagreb/Crikvenica: Institut za arheologiju, Muzej Grada Crikvenice. Jadrić-Kučan, I. and J. Zaninović 2015. Fibule iz Arheološke zbirke Burnum. Drniš – Šibenik. Mandruzzato, L. 2008. Vetri Antichi del Museo Archeologico Nazionale di Aquileia, Ornamenti e oggettistica e vetro pre- e post-romano, Corpus delle Collezioni del Vetro in Friuli Venezia Giulia 4. Trieste. Marin, E. 2002. Arheološki muzej Split, vodič. Split. Miletić, Ž. 2010. Burnum – vojničko središte provincije Dalmacije, in I. Radman-Livaja (ed.) Nalazi rimske vojne opreme u Hrvtaskoj: 113–142. Zagreb. Miletić, Ž. 2011. Proizvodnja tegula u Burnumu u kontektsu građevinskih aktivnosti / Production of tegulae in Burnum in the context of building activities, in G. Lipovac Vrkljan, I. Radić Rossi and B. Šiljeg (eds), Rimske keramičarske i staklarske radionice, Proizvodnja i trgovina na jadranskom prostoru, Zbornik 1. međunarodnog arheološkog kolokvija, Crikvenica, 23. – 24. listopada 2008: 263–279. Zagreb/Crikvenica: Institut za arheologiju, Muzej Grada Crikvenice. Panini, A. 2007. Middle Eastern and Venetian Glass Beads, Eighth to Twentieth Centures. Milano. Patsch, K. 1895. Rimski kameniti spomenici Kninskog muzeja, Glasnik Zemaljskog muzeja u Bosni i Hercegovini Sv. VII: 379–422. Patsch, K. 1900. Archäologisch-epigraphische Untersuchungen zur Geschichte der römischen Provinz Dalmatien. Wissenschaflitche Mittheilungen aus Bosnien und der Hercegovina: 3–115. Wien. Reisch, E. 1913. Die Grabungen des Österreichischen archäologischen Institutes während der Jahre 1912 und 1913. Das Standlager von Burnum. Jahreshefte des Österreichischen archäologischen Institutes in Wien 16: 113–135. Riha, E. 1990. Der römische Schmuck aus Augst und Kaiseraugst. Forschungen in Augst 10. Augst. Siepen, M. 2007, Außergewöhnliches Pferdegeschirr aus Gellep, in S. Krefeld (ed.) Archäologie im Rheinland 2007, Landschaftsverband Rheinland, Rheinisches Amt für Bodendenkmalpflege (viewed 09 November 2019). Spaer, M., D. Barag, T. Ornan, and T. Neuhaus 2001. Ancient Glass in the Israel Museum: Beads and other small objects. Jerusalem. Stern, E., M. and B. Schlick-Nolte 1994. Early Glass of the Ancient World: 1600 B.C. – A.D. 50, Ernesto Wolf Collection. Ostfildern. Tempelmann-Maczyńska, M. 1985. Die Perlen der römischen Kaiserzeit und der frühen Phase der Völkerwanderungszeit im mitteleuropäischen Barbaricum. Mainz am Rhein. Zabehlicky-Scheffenegger, S. and M. Kandler 1979. Burnum I. Erster Bericht über die Kleinfunde der Grabungen 1973 und 1974 auf dem Forum. Wien. Zaninović, M. 1968. Burnum, castellum-municipium. Diadora 4: 119–129. 339

Glass askos from the ancient necropolis of Zadar Timka Alihodžić Abstract During the rescue archaeological research conducted in 2008 by the Archaeological Museum in Zadar in Polačišće Street ninety graves were unearthed belonging to a large necropolis of ancient Iader where over 2100 graves have been excavated in the last 30 years. Although both burial rites are represented, cremation graves yielded more grave goods, in particular glass artifacts, as was the case with the previous research. An askos-type vessel made of cobalt blue glass found in a cylindrical stone urn (grave 45) represents a unique find. Not a single such artifact was found after reviewing over 7000 complete glass objects and fragments of different vessels that were inventoried in the Collection of Ancient Glass of the Archaeological Museum in Zadar. Comparable finds have not been evidenced in the wider region of Liburnia.

Key words Askos, grave, necropolis, glass, grave goods.

Figure 1. Positions of Roman tombs in Zadar (left) and position of Polačišće street site (map: I. Čondić Begov).

Zadar is situated on a natural peninsula that is connected with the mainland on its southeastern end. Good strategic position, abundant water sources and mild climate boosted the creation and development of the settlement in the 9th and 8th centuries BC. The Liburnian settlement became a colony (Iulia – colonia Iader) after the Roman conquest. Builders of ancient Zadar in the 1st century gave this settlement its urban appearance respecting its natural position, laws of construction and the natural location (Alihodžić 2019: 10). Pre-Roman settlers of Zadar, the Liburnians, buried their dead outside the settlement practicing inhumation. When the Roman power was established the tradition of burials outside the city was continued, but inhumation was accompanied by cremation ritual characteristic of the Roman civilization. Explicit prohibition of burials within the cities Hominem mortuum in Urbe ne sepelito ne urito1 (The dead may not be buried or cremated within the city) was recorded in the laws written around 451 BC. Therefore the necropoles of ancient Zadar were formed outside the urban area, along the roads (Figure 1). Information on discoveries of 1 

Leges duodecim tabularum (X, I).

ancient graves outside the cities, in particular in Relja, were recorded as early as the 19th century (Gluščević 2005).2 Extensive archaeological research had been conducted in the wider area of Relja by the end of the 20th century and at the beginning of the 21st century, before the shopping mall, residential buildings, bank, hospital etc. were built.3 Over 2100 graves were uncovered mostly dating to the Roman and Late Roman period. There were also few Iron Age graves dating to the 7th and 6th centuries BC and about three hundred graves that might be attributed to the Venetian siege of Zadar. Anyhow it is an area with over a millennium-long continuity of burials.

This comprehensive dissertation provides all data concerning the research history of the necropoles of ancient Zadar in the period from the end of the 19th century until the end of the 20th century. 3  Most excavations were conducted by the Archaeological Museum in Zadar (Nedved 1980; Gluščević 1990, 2005; Fadić 2007; Alihodžić 2009; 2010; 2014; Perović and Fadić 2009; Vučić 2010), then the University of Zadar (Gusar and Vujević 2011) and private archaeological companies in cooperation with the Archaeological Museum in Zadar. 2 

Roman Pottery and glass manufactures (Archaeopress 2022): 340–343

Glass askos from the ancient necropolis of Zadar

Figure 2. Grave 45 (photo: T. Alihodžić). Ninety new graves were unearthed in the rescue archaeological research conducted in 2008 in the part of Polačišće Street in Relja when storm sewer system was built (Alihodžić 2009: 508–509). Both cremation and inhumation were represented in the excavated graves. As in the previous research, grave goods in cremation graves exceed their counterparts in inhumation graves in number and diversity. Objects made of glass are more numerous than the finds made of clay, bone or other materials.

in the upper part (A20091) and the lid (A20092) with a hollow cylindrical handle ending in a folded hornshaped form. This type of finds was dated to the time span of the 1st and 2nd centuries by a number of authors (Ravagnan 1994; Fadić 2001; Gluščević 2005 etc.). Identical examples found in the Roman-era necropoles in the large area of the Roman Empire were dated to the 1st and 2nd centuries on the basis of coin finds. Three glass flasks (A20094, A20095, A20096) were also dated to the 1st and early 2nd century on the basis of parallels with other graves (Gluščević 1990; Gregl 1997; Fadić 2001 etc.).

Grave 45 After the corpse was burnt on the pyre the bone remains were deposited in a glass urn which was then placed in a cylindrical stone urn (Figure 2). All goods were within a glass and a stone urn: three glass flasks, glass askos, small thin-walled bowl, plate fragments (terra sigillata), bronze mirror with a handle, cosmetic/pharmaceutical plaque, bronze spoon and a bone pin with a head shaped like a fist holding a flat object (Figure 3). In the layer of soot around the stone urn were burnt balsamaria and several iron nails.

Production of small thin-walled bowls (A20101) started in the late Augustan and Tiberian period and lasted until the mid-2nd century. The earliest examples are unornamented hemispherical forms as the one from grave 45 (Brukner 1981; Atlante I 1981; Atlante II, 1985; Brusić 1990; Borzić 2014). Fragmentary plate (terra sigillata italica) on a ring base (A20102) has a barbotine relief decoration in form of a human face and an animal (dog, lion) running that were modelled by trailing diluted clay. Central part is missing, but a part of the stamp with the initial letter C or G can be discerned. It is difficult to say which letter

Glass urn with squat body tapers to slightly emphasized base and profiled opening, with H-shaped handles 341

Timka Alihodžić

Figure 3. Cataloged objects from the grave 45 (photo: T. Alihodžić). exactly it is due to damage. The first examples of these forms date to the Augustan era, year 10 BC, and they were common until the beginning of the 2nd century AD. Hairpin with the head shaped like a fist holding a plaque (A20105) is associated with the iconography of the cult of Sabasius (Kovač 2017). Many authors hold an opinion that these hairpins had distinctly magical and cult character. They were characteristic of the Early Imperial period (Petković 1995), but they were also used later on.

(A20093) of dark blue glass, made by using free blowing technique is a unique find at the Zadar necropolis. Askos is a Greek term for a vessel used as a jug placed horizontally when not in use, often without a distinct base, foot or other standing surface. The example from Zadar has ovally shaped body, tapering on one side and growing into a neck that ends with an expansion and an annular rim bent inwards (Figure 4). Lower part is flattened. The handle that connects the rim with the body was applied subsequently and modelled as a strip.

Cosmetic/pharmaceutical plaque of stone (A20104), a small bronze spoon (A20106) and a cane with a silver ring (A20103) are common grave goods in the 1st and 2nd centuries belonging to the cremation ritual, as well as the bronze mirror with a handle (A20107). This mirror type in certain necropoles was dated on the basis of coins and in most cases it belongs to grave goods recovered from the 1st century graves.

The recovered askos has enriched the repertory of the Collection of Ancient Glass of the Archaeological Museum in Zadar that holds over 7000 objects. Parallels for the Zadar example have not been found among the explored necropoles in Croatia. Parallel examples dating to the 1st century are kept in Museo Archeologico Nazionale di Adria (IGAD 21657, prov. Adria – Collezione Bocchi) and two examples from Pompeii in Museo  Nazionale  Archeologico di Napoli (Isings 1957: form 59).

Exact form of the burnt balsamaria (A20096, A20097, A2099, A20100) could not be determined since they were deformed, but parallels with other cremation graves suggest dating to the 1st and 2nd centuries.

Having in mind this preliminary dating of grave goods to the 1st and early 2nd century, and attested dating of askoi found in the area of Italy, in particular the example from Pompeii from the 1st century, the askos from grave 45 from Polačišće Street is dated to the 1st century.

All the mentioned grave goods are common not only at the Zadar necropolis but also at necropoles across the Roman Empire. However the small askos-type vessel 342

Glass askos from the ancient necropolis of Zadar

Atlante II, 1985. Enciclopedia dell›arte antica. Atlante delle forme ceramiche II. Istituto della Enciclopedia italiana. Roma. Borzić, I. 2014. Radionički pečati na italskoj tera sigilati iz Burnuma. Archaeologica Adriatica 7/2013: 133–150. Zadar. Brukner, O. 1981. Rimska keramika u jugoslovenskom dijelu provincije Panonije. Savez arheoloških društava Jugoslavije. Beograd. Brusić, Z. 1990. Italska terra sigillata u Liburniji. Diadora 12: 79–105. Zadar. Fadić, I. 2001. Antičko staklo u Liburniji. Neobjavljena doktorska disertacija. Zadar: Sveučilište u Zadru. Gluščević, S. 1990. Rimska nekropola u Kaljskoj ulici. Diadora 12: 107–159. Zadar. Gluščević, S. 2005. Zadarske nekropole od 1. do 4. stoljeća. Organizacija groblja, pogrebni obredi, podrijetlo, kultura, status i standard pokojnika. Neobjavljena doktorska disertacija. Zadar: Sveučilište u Zadru. Gregl, Z. 1997. Rimske nekropole sjeverne Hrvatske. Zagreb: Arheološki muzej u Zagrebu. Gusar, K. and D. Vujević 2011. Numizmatički nalazi s lokaliteta Jazine – Lignum u Zadru. Zbornik radova 6. Međunarodnog numizmatičkog kongresa u Hrvatskoj: 113–126. Rijeka: Hrvatsko numizmatičko društvo. Isings, C. 1957. Roman glass from dated finds. Archaeologica Traiectina 2. Kovač, M. 2017. Tipologija i tehnologija izrade rimskih koštanih predmeta na području donje Panonije na primjeru nalaza iz Murse I. Neobjavljena doktorska disertacija. Zagreb: Filozofski fakultet. Nedved, B. 1980. Zaštitno istraživanje rimskih grobova u Zadru. Diadora 9: 341–361. Zadar. Perović, Š. and I. Fadić, 2009. Zaštitno arheološko istraživanje dijela antičke nekropole Zadra u Zrinsko-frankopanskoj ulici. Diadora 23: 45–132. Zadar. Ravagnan, G. L. 1994. Vetri antichi del Museo Vetrario di Murano. Corpus delle collezione archeologiche del vetro Veneto 1. Comitato nazionale italiano. Association internationale pour l›histoire du verre. Venezia. Štefanac, B. 2009. Staklene kapaljke s antičke nekropole Relja u Zadru. Vjesnik za arheologiju i povijest dalmatinsku 102: 109-127. Vučić, J. 2010. Zadar – Glagoljaška ulica. Hrvatski arheološki godišnjak 6/2009: 539‒540. Zagreb: Ministarstvo kulture, Uprava za zaštitu Kulturne baštine.

Figure 4. Glass askos – A20093 (photo: I. Čondić Begov). The workshop center where the askos was made cannot be determined for now, although it is reasonable to assume that this was import from the eastern Mediterranean region where glass dropper flasks were produced as well (Štefanac 2009: 109–127). Diverse objects placed in graves in the belief that the deceased person might use them in the afterlife suggest that population of different economic profiles and religious affiliations lived in ancient Zadar. Well-to-do class definitely used benefits of the developed trade all over the Roman Empire. References Alihodžić, T. 2009. Ulica Polačišće. Hrvatski arheološki godišnjak 5/2008: 508–509. Zagreb: Ministarstvo kulture, Uprava za zaštitu Kulturne baštine. Alihodžić, T. 2010. Zadar – Relja (parkiralište). Hrvatski arheološki godišnjak 6/2009: 540–543. Zagreb: Ministarstvo kulture, Uprava za zaštitu Kulturne baštine. Alihodžić, T. 2014. Zadar Relja (Ulica Katarine Zrinski i Put Murvice). Hrvatski arheološki godišnjak 10/2013: 492–493. Zagreb: Ministarstvo kulture, Uprava za zaštitu Kulturne baštine. Alihodžić, T. 2019. Ispričat ću ti priču. Katalozi i monografije 33, 2. dopunjeno izdanje. Zadar: Arheološki muzeja Zadar. Atlante I. 1981. Enciclopedia dell›arte antica. Atlante delle forme ceramiche I. Istituto della Enciclopedia italiana. Roma.

343

Glass finds from Tomb 7 from the Late Iron Age necropolis of the Kopila hillfort on the island of Korčula Anamarija Eterović Borzić, Igor Borzić Abstract The systematic archaeological investigations which have been taking place since 2012 at the site of the hillfort settlement of Kopila on the island of Korčula, are focused primarily on the necropolis i.e. on one of at least two of its nuclei, made up of mutually connected monumental tomb plots, they have enriched the present very modest knowledge about the local protohistoric community. During six seasons seven tombs were investigated, mostly dating from the 4th to 1st centuries BC. The tombs were used repeatedly, and the exceptionally large amount of archaeological material, primarily Hellenistic ceramic vessels, metallic parts of attire and offensive weapons plus glass and amber decorated objects, points to the fact that this was an economically powerful and commercially very active indigenous community. The present overall analysis of the material remains of the tombs have enabled the clear interpretation of the funeral customs and their specificities which are particularly recognised in two tombs (Tomb 1 and Tomb 7), in which large number of children of perinatal and natal age were buried, and which are for now a unique example within the knowledge of Illyrian funeral customs. The glass material from Tomb 7 which consists of a large number of beads and pendants and date from the 3rd to 1st centuries BC, and which are a product of Hellenistic workshops throughout the Mediterranean will be analysed. A glass balsamary found in the context of a cremation burial inside the tomb, within the necropolis unique in every sense will also be analysed separately. This kind of burial and the connected rite represent a cultural shift in relation to earlier customs and are undoubtedly the consequence of the interaction with the Roman civilisation.

Key words Korčula, Kopila hillfort, Late Iron Age, necropolis, glass.

Work on the archaeological topography of the southern Adriatic island of Korčula has resulted, inter alia, in locating some twenty Iron Age hillfort settlements (Radić and Bass 1998: 361-403; 2001: 133-146; Radić 2001: 25-50). Such dense indigenous population, in existential terms based on optimal economic conditions involving agriculture, livestock breeding and fishing, in a certain way justifies dynamic history of the island in the last millennium BC, additionally interpreted from written and epigraphic sources (Radić and Borzić 2017a: 1934). Except for indigenous population (Plerei) (Strabo, Geographica VII, 7,5,7), in the 6th century BC Korčula was potentially visited by Greek colonizers from Cnidus in Asia Minor (Pseudo-Scymnus, Periegesis 425-427; Strabo, Geographica VII, 7,5,7; Plinius Secundus, Naturalis Historia III, 152; on interpretation of the sources in Lisičar 1951; Rendić-Miočević 1980: 229-250; Mastrocinque 1998; Suić 2001: 161-168; Zaninović 2015: 95-117; Radić and Borzić 2017a: 20-23), and in the 3rd century BC by the colonizers from Adriatic Issa (Brunšmid 1898: 2-14, no. 1; Rendić-Miočević 1966: 133-141; Kirigin 1985: 98; Lombardo 2002: 130; Kuntić-Makvić and Marohnić 2010: 75-77). By the end of the same century the Illyrian Kingdom became an active participant in the events on the island (Livius 40. 42,5; Papazoglu 1967: 123-145; Zaninović 2015: 191-317), and finally also the Roman Republic (Appianus, Ill. 16; Šašel-Kos 2005; Zaninović 2015: 405-469). In certain cases mentioned written sources are concise and clear, but they can

also be ambiguous and that is where science steps in trying to find material evidence that might corroborate described conditions and relations of the highlighted participants. Unfortunately, the state of systematic exploration of the Iron Age on the island of Korčula is poor and except for the mentioned topography, at the moment it comes down to the project of research of Kopila hillfort (Stražišće). It is a settlement strategically located on a dominant position north of Blatsko Polje, offering a view of the wide maritime zone towards the Pelješac Channel, Bay of Neretva and the islands of Hvar and Vis. It has been recognized as an archaeological site back in the 19th century (Ostoich 1858: 111-112; Radić and Vuletić-Vukasović 1890: 104-105), and until recently it has been mentioned only as a findspot of six Corinthian vessels and an abundance of metal parts of attire and jewellery, typical of the Iron Age communities of the southern Dalmatian cultural circle (Lisičar 1973: 7; Batović 1986; Radić 2001; 2003: 76-77; Marijan 2000). The project of systematic excavations at the site started in 2012,1 and most of the field research activities have been focused on the research of newly found necropolis that is unique in terms of spatial organization (Radić and Borzić 2017b: 49-58; 2017d: 105-120; 2017e: 310316) (Figure 1). It is located on a plateau of protruding The Municipality of Blato, Cultural Center in Vela Luka, Department of Archaeology of the University of Zadar and the Museum of Ancent Glass in Zadar participated in the project funded by the Ministry of Culture of the Republic of Croatia. 1 

Roman Pottery and glass manufactures (Archaeopress 2022): 344–353

Glass finds from Tomb 7 from the Late Iron Age necropolis of the Kopila hillfort

Figure 1. Necropolis of the Kopila hillfort (Nucelus 1 and 2) (photo: M. Vuković; microlocation marks by I. Borzić).

lingula at the foot of the hillfort, over Blatsko Polje. The necropolis consists of two nuclei, about 50 meters apart, each consisting of a number of combined monumental complex built grave plots, roughly round in shape. A total of seven tombs have been excavated so far, six in the western nucleus (Nucleus 1) and one in the eastern one (Nucleus 2), all indicating repeated use for periods of different duration, generally corresponding to the time span from the mid-4th to the late 1st century BC.

and pendants), amber beads and pendants (Eterović Borzić 2017a: 99-116; 2017b: 117-124), metal jewellery (earrings, necklaces, rings and buttons) and clothing items i.e. fibulae as well as pins (Radić 2017: 85-95). Such abundant and diverse inventory of the tombs suggests a general conclusion that it was a community that had an opportunity to participate actively in the regional trade relations at least in the last four centuries BC, which is not surprising given its strategic position and resource potentials (Radić and Borzić 2017c: 139-145).

Systematic study of five tombs from Nucleus 1 excavated from 2012 to 2016 (Tomb 1 – 5/6) has been presented monographically in the catalogue of the exhibition The city of the dead above the field of life – necropolis of the Iron Age settlement of Kopila on the island of Korčula (Fadić and Eterović Borzić 2017). Analysis of the osteological material from these tombs has indicated that in case of Tombs 2, 3, 4 and 5/6 we have family resting place in which the deceased of both sexes and different ages were buried. As opposed to that, central Tomb 1 was intended for inhumations of a number of children, mostly neonates (Radovčić 2017: 125-138). Funerary ritual was manifested in both cases in the same way – through depositing abundant Hellenistic ware, predominantly of symposiastic character (southern Italic and Issaean Gnathia, as well as black, brown, red and grey-glazed pottery) (Borzić 2017: 61-84; 2020: 363-376), and both offensive and defensive weapons with adult men (spears, daggers and a fragment of Illyrian helmet) (Radić 2017: 95-97). All tombs yielded a considerable amount of glass jewellery (necklaces

Results of the newly excavated Tomb 7 support these insights. It is located between Tombs 2 and 3 west/south of it and Tomb 8 located in the peripheral eastern part of Nucleus 1 (Figure 2). The tomb has truncated circular shape defined by monumentally built ring made of often coarse trapezoidal stone blocks arranged in at least three rows in height. In the inner space, filled with stones, was a rectangular burying area in the centre (150 x 65 x 80 cm), with usual orientation of 120–300°. Its bottom part consists of one row of coarse stone blocks leaning on bedrock or sea pebbles (Figure 3). It is interesting that once again we have a child’s tomb used repeatedly in the period from the 3rd to 1st century BC, meaning simultaneously with Tomb 1. Assumption based on the situation from Tomb 1, where the number of deposited vessels corresponds to the number of the deceased, suggests that at least 150 children were buried in Tomb 7. In stratigraphic terms the tomb is quite complicated exactly because of its character and often impossible precise definition of separate burial phases (Figure 4). 345

Anamarija Eterović Borzić, Igor Borzić

Figure 2. Excavated tombs of the Nucleus 1 on the necropolis of the Kopila hillfort (photo: M. Vuković ; microlocation marks by I. Borzić).

Figure 3. Tomb 7 (photo: M. Vuković).

Figure 4. Tomb 7 (photo: I. Borzić).

A burial from the very top is completely different from all other examples recorded in this necropolis. It is a cremation burial in a ceramic urn surrounded by three ceramic balsamaria and one glass specimen. Such practice corresponds to the Roman funerary ritual in all aspects that is recorded on the eastern Adriatic coast on the Issaean necropolises from the mid-1st century BC (Ugarković 2019), which is roughly the dating framework of the burial from Kopila. Under this burial level, there are a number of smaller detached units

within the tomb. Although the archaeological image was quite confusing, it seems that these units relate to some recognizable/preserved skeletal remains of the deceased, framed or laid on a stone slab, and completely preserved ceramic vessels of Hellenistic origin (greyglazed ware, Canosan and Issaean Gnathia pottery and black-glazed ware) associated with the bones, scarce examples of metal attire items as well as a number of jewellery objects (necklaces made by stringing glass and ceramic beads, pierced coin and one earring). 346

Glass finds from Tomb 7 from the Late Iron Age necropolis of the Kopila hillfort

Figure 5. Tomb 7, glass beads in situ (photo: I. Borzić).

In that way we have reached the central topic of the paper, which is glass inventory of Tomb 7. The settlement in Kopila has become very indicative with regard to presence and use of glass objects by the Iron Age communities of central and southern Dalmatian area owing to excavations of the necropolis. One of the exceptional glass finds from Kopila is an archaic plaque bearing a depiction of a siren (Radić 2003: 73-74; Eterović Borzić 2017a: 101-104). On the other hand, all excavated tombs offered an exceptional amount of glass artefacts, thousands of pieces. Most of these items belong to glass beads used for making necklaces, but certain examples were used for additional decoration of earrings (Eterović Borzić 2017a: 106-107, 187 fig. 55, cat. no. 4/23). Work on typology based on technological and morphological criterion resulted in their classification into the following categories (Eterović Borzić 2017a: 104-116): translucent monochrome and opaque annular, cylindrical, globular, ovoid, biconical, whorl-shaped beads as well as the examples with granules including the ones known as the Adria type; and technologically diverse polychrome beads represented by cylindrical, globular, globular-spherical beads and specific eye beads. In addition to the mentioned types, glass repertory from Kopila was enriched by small number of pendants (pointed polychrome pendants, pendants in shapes of vessel-aryballos, bunch of grapes and tooth), and gems and inlays.

typology, with some novelties. Here as well most glass artefacts comprise beads whose main division is based on their colour (monochrome or polychrome). The first ones, monochrome, are most numerous, coming in annular, cylindrical, globular and globular-spherical form (Figure 6). Most of these beads are very small (2–4 mm), in blue, greenish, yellow, amber, white, or colour of decolorized glass. It is certain that all of them belonged to jewellery but it is also possible they were sewn on clothes. In terms of technology it is easy to differentiate between the examples made by winding and by drawing technique. In both cases it is a technologically undemanding process that resulted in production of these beads in many workshop centres, causing their wide distribution in different epochs. Focusing on Tomb 7 only, we can notice that they are present in all stratigraphic units associated with separate phases of its use under the previously mentioned cremation burial, which definitely dates them from the 3rd to mid-1st century BC. Somewhat more precise dating, at least in the context of this tomb, can be offered for 27 globular yellowish beads recovered in situ (Figure 5). It was evidently a necklace consisting of some other forms of beads that we will mention in continuation. As for its context, it was found with the bones of a child, skyphoi of the Issaean Gnathia pottery dated to the first half of the 2nd century BC, and pseudo-Cypriot pin from the same period. Their continuous duration into the second half of the 2nd century, and possibly in the first half of the 1st century BC is attested by their presence in

As expected, glass inventory of Tomb 7 mostly corresponds to the finds from other tombs in terms of 347

Anamarija Eterović Borzić, Igor Borzić

Figure 6. Monochrome glass beads (photo: T. Lučić).

Tomb 4 that is dated exactly to the mentioned timespan (Radić et al. 2017: 180-197).

in the lowest layers of the tomb that is chronologically determined to the 3rd century BC by Gnathia pottery of late Canosan variant. The closest parallels for these beads, and the remaining glass finds from Kopila, can be found in the Iron Age graves from Vičja Luka on the island of Brač (Blečić Kavur and Kavur 2017: 97, fig. 3) where they formed a necklace together with yellowish pendants in shape of an amphoriskos. Identical combination has been recorded in Greece which seems to be the most likely region of origin of both forms. We refer primarily to the island of Rhodes, as one of the most important glassmaking centres in the Mediteranean, where primary and secondary production is assumed from the 6th century BC, and perhaps even earlier (Triantafyllidis 2003: 131).

Out of the mentioned abundance of very small monochrome beads we can only tentatively single out few bigger examples, identical in terms of typology and technology. Specifically this refers to blue annular beads or yellowish-amber globular examples (Figure 6/8, 9). The former, only three pieces, were found in context with Gnathia type pottery dating to the turn of the 2nd century BC, but identical specimens in Tomb 6 suggest their dating throughout the 3rd century, and possibly also to the end of the 4th century BC (Radić et al. 2017: 197-203). Blue or rarely yellowish biconical beads constitute a special category of monochrome beads from Tomb 7. Different variants of biconical beads have already been found among the glass inventory from tombs from Kopila (e.g. in Tombs 1, 3 and 4), and these forms of reverse hourglass are known as decoration of a silver earring from Tomb 4 from the second half of the 2nd and first half of the 1st century BC (Radić et al. 2017: 195, cat. no. 4/40). Examples from Tomb 7 (Figure 6/10) are almost certainly earlier as they were found with the previously mentioned necklace, meaning in the context of the first half of the 2nd century BC, but also

A rather large globular-ovoid example in yellowishgreenish colour also belongs to monochrome beads from Tomb 7 (Figure 6/11). It was found on the very bottom of the tomb, meaning in the context of the mid3rd century BC. Polychrome beads in Tomb 7 are not as numerous, and their typological repertory corresponds to beads from other tombs in this necropolis. The most important representatives of this category of finds are definitely eye beads produced back from the mid-2nd millennium 348

Glass finds from Tomb 7 from the Late Iron Age necropolis of the Kopila hillfort

Figure 7. Polychrome glass beads (photo: T. Lučić).

in the region of the Near East, Mesopotamia and Egypt wherefrom production gradually spread to several areas in the European mainland. Greek island of Rhodes can be mentioned in this context once more where big production of these beads was recorded in the Hellenistic period that is also the period of our interest (Nenna 1995: 11-12, 19-21). All specimens from Kopila, including the ones from the analyzed Tomb 7 (Figure 7/3-9) belong to the variant with layered eyes on the criterion of form and manufacture details. Dark blue beads are dominant, with eyes inserted from white to central dark blue. As for the technological characteristics, we might distinguish the beads whose eyes were made by alternate applying molten glass rods in different colours and the ones whose body was decorated by previously made and cut eyes, subsequently finished when the bead was made.

mid-4th to mid-1st century BC, which is supported by the fact that in Tomb 7 they were found equally in the lowest (mid-3rd century BC) and highest burial layers (first half of the 1st century BC). Two globular-spherical beads found in situ with several more various beads also belong to polychrome beads (Figure 7/1 and 2). The first example has white trails applied on the edges of the blue body whose centre is decorated by another, in this case, spiral white-blue trail. The second example differs only in the colour of edge trails that are yellow-ocher in this case. A very similar bead in technological terms, but with a cylindrical body with an embossed central spiral trail was recovered from Tomb 3 (Radić et al. 2017: 177, cat. no 3/38). Its closest parallels were recorded in the Hellenistic layers of the Greek island of Lemnos (site of Myrina) (Arveiller-Dulong and Nenna 2011: 181, 141a) so it should not be surprising if these and newly found beads from Tomb 7, similar in terms of technology and concept originated from workshops in this part of the Mediterranean. As already mentioned, they were found in situ, and the details of their position in the tomb provide the chronological framework of their use from the mid-2nd to mid-1st century BC. Namely, they constitute a part of the inventory of a burial clearly defined by lower and upper limestone slab. While pottery recovered mostly from the same burial level is dated throughout the second half of the 2nd century BC (Issaean Gnathia pottery), burial above the mentioned,

Presence of this type of beads on Kopila corresponds to their big popularity that has to do, at least to a certain extent, with their apotropaic character. In the neighbourhood they were found in a LiburnianRoman necropolis in Nadin, in indigenous tombs on the island of Brač (Vičja Luka, Žaganj Dolac) (Marović 1971: 14-16, fig. 8, 23; Marović and Nikolanci 1977: 48, Tab. VIII), Konavli (Š. Batović, 2003, 233, Sl. 19, 5) and Bosnia and Herzegovina (Gorica, Vir (Posušje), Gruda) (Marović 1971: 23), but also in the tombs of Hellenistic Issa (Ugarković 2019). Their presence in all tombs of the Kopila necropolis in this context dates them from the 349

Anamarija Eterović Borzić, Igor Borzić

Figure 8. Necklace made of three coins and glass beads (photo: I. Borzić).

yielded a necklace made of three coins, inter alia, one of which belongs to the Late Republican denarius from the year 60 BC (RRC = Crawford 420/1: P. Plautius Hypsaeus, Av. P YPSAE SC. Head of Neptune, right; Rv. C YPSAE COS / PRIV – CEPIT Jupiter in quadriga, left) (Figure 8). The second category of glass artefacts consists of differently modelled pendants. Tomb 7 yielded two such finds one of which is a heart-shaped pendant with perforated vertical axis made of bluish-greenish glass (Figure 9/1). It was found very close to the previously mentioned polychrome beads dating it to the second half of the 2nd century BC. This dating is confirmed by the Hellenistic chronological determination of the closest eastern Mediterranean parallels from an unfortunately unknown site that are kept in Louvre (Arveiller-Dulong and Nenna 2011: 226, cat. 303). The second pendant in flattened oval shape is made of bluish translucent glass (Figure 9/2). It was found at the very bottom of the tomb which dates it to the 3rd century BC. Its form is similar to so-called glass signet rings that were very popular in the Classical and Hellenistic periods, but this example lacks engraved decoration.

Figure 9. Glass pendants (photo: T. Lučić).

Figure 10. Glass balsamarium (photo: T. Lučić).

350

Glass finds from Tomb 7 from the Late Iron Age necropolis of the Kopila hillfort

For the end we have left possibly the most interesting glass object from Tomb 7. It is a glass balsamarium (Figure 10) documented in a unique cremation burial whose grave goods consisted of three ceramic balsamaria placed around the ceramic urn of local provenance. Glass cobalt blue balsamarium has cylindrical body with an emphasized transition to high cylindrical neck that ends in an everted rim. In technological terms it is a freely blown object, which is also most interesting. Namely we know that this technique developed in the mid-1st century BC (Grose 1977: 11) so that the balsamarium from Kopila represents one of the earliest glass objects produced in this way in the region of southern Dalmatia. Parallels can be found across the world that became Roman by this time so dating is not questionable (Isings 1957: 24, type 8; De Tommaso 1990: 81-85, type 60, 67, 70, 71, 72 ; Buljević 2002: 401).

found after wet sieving of the entire deposit of earth and small stones from the tomb. Another illustrative indicator is almost complete lack of beads from the devastated ‘Tomb 0’, contents of which was emptied onto the drystone wall built at the turn of the 20th century (Radić et al. 2017: 148-151). It is difficult to say if that is the reason why we emphasize Kopila as an Iron Age site with the biggest amount of glass objects on the eastern Adriatic, but we should not neglect the lack of more systematic research of necropolises that might provide some comparative material such as the Liburnian necropolises V. Mrdakovica and Dragišić, then the ones in Herzegovina (Neum, Gorica, Mrljanovac, etc.) or individual graves on the Dalmatian islands. There are some illogical inconsistencies in publications of material from these sites. It is worth mentioning almost unbelievable fact about virtually complete lack of glass finds in quite numerous Hellenistic graves on the eastern Adriatic. Among the published Issaean graves, glass beads have been found in graves 46b, 91, 93 and 141 at the necropolis Vlaška Njiva (Ugarković 2019: 171). Situation is not much better in Hellenistic Budva where glass beads were recorded in only two tombs, Gr. 6 (2 beads) and XLV (19 smaller and one bigger bead in a necklace) (Marković 2012: 26 and 74). Presented image makes Kopila seem like a prominent site with regard to glass finds, but it is debatable if this image is realistic concerning the frequency of glass use in the funerary ritual. If we refer to the case of Tomb 7, four numerical indicators can be mentioned: tomb duration (200 years), number of burials (c. 150), number of beads (c. 400) and number of beads necessary to make a certain jewellery item, necklace or bracelet. Unfortunately in most cases we do not know the last number, especially because it depends on the size of used beads and their combinations on jewellery. However there are also examples that can be used as good illustrations, such as the necklace from Vičja Luka (necklace made mostly of eye beads = 40 beads; the one made of 55 biconical beads and 16 amphoriskoi). Since this is a child’s grave, probably a smaller number of beads was necessary for one necklace, but still they do not reach the number that would determine use of glass jewellery as a common practice, but more a momentary practice of the participants (family) associated with the burial act, availability or some other aspect that we cannot perceive at present. Situation in Tomb 4 clearly shows that glass jewellery belonged exclusively to female part of the adult population so there is a possibility that the same practice was reflected in case of children. In case of bigger or more elaborate beads, we should not exclude the possibility of intentional depositing individual beads next to a child.

Presented repertory of glass objects from Tomb 7 of the hillfort settlement of Kopila on the island of Korčula has not changed previous insights about this type of finds in the local community. Still, even corroborating certain things has its scientific value. On the basis of the entire inventory, the tomb has been dated to the period from the mid-3rd century to roughly mid-1st century BC and the assumption is that it was used over a hundred times in that interval. This fact is a starting point of two short discussions, or more precisely considerations, first of which is related to unfortunately poor possibility of more precise dating of certain glass forms. The fact is that glass beads were found in all vertical and horizontal zones not only in Tomb 7 but also in all other repeatedly used tombs of the Kopila necropolis. However, one needs to keep in mind the nature of archaeology, and the fact that the development of situation in tombs with regard to distribution of beads is often purely accidental, resulting from their ‘free circulation in the tomb’ caused by reasons that can be natural ‘gravitational’, but also artificial historical (period of tomb use) and recent (our excavations). Therefore there is not any other possibility for a number of these finds but to date them to the mentioned period that roughly corresponds to the Hellenistic dating for most recorded bead types, as common in professional literature. Even the examples from more or less intact context are in accordance with this dating, and its narrower chronological framework. The second discussion refers to the question of amount of glass finds in the tombs that is general frequency of their use and presence in indigenous communities of southern Dalmatian region and its surroundings. We believe that with our excavation procedures and postexcavation work we managed to collect majority if not all glass beads that are usually so small that they are almost invisible to the naked eye. A good illustration of this statement is the fact that most beads were

Finally we will refer shortly to origin areas of the glass material. So far we have determined several possible origin areas of glass objects from Kopila on the basis of 351

Anamarija Eterović Borzić, Igor Borzić parallels and well-grounded assumptions ranging from the eastern Mediterranean, Egypt, Black Sea region, over north Africa, north Italy or Celtic area. Analysis of the artefacts from Tomb 7 may have brought us closer to the ‘most serious’ centre that most of our objects originate from – Rhodes, where most comparative material can be found, and whose workshops produced an exceptionally high number of glass beads in the Hellenistic period. Furthermore it is a very important strategic and trade centre that definitely played an important role in general distribution of products of other workshop centres that were found in Kopila.

Kirigin, B. 1985. Zapažanja o helenističkoj nekropoli Isse. Materijali 20: 91–110. Kuntić-Makvić, B. and J. Marohnić 2010. Natpisi, in Antički Grci na tlu Hrvatske, exhibition catalogue: 73– 89. Zagreb: Klovićevi dvori. Lisičar, P. 1951. Crna Korkira i kolonije antičkih Grka na Jadranu. Skopje: Filozofskiot fakultet. Lisičar, P. 1963. O jednom tipu lučne fibule iz naših arheoloških zbirki. Radovi Filozofskog fakulteta u Zadru, Razdio povijesnih znanosti 2/2: 25–36. Lisičar, P. 1973. Cenni sulla ceramica antica. Arheologica Iugoslavica XIV: 3–27. Lombardo, M. 2002. I Greci a Kerkyra Melaina (Syll³ 141): pratiche coloniali e ruolo degli indigeni, in N. Cambi, S. Čače and B. Kirigin (eds) Grčki utjecaj na istočnoj obali Jadrana: 121–140. Split: Književni krug. Marijan, B. 2000. Željezno doba na južnojadranskom području. Vjesnik za arheologiju i historiju dalmatinsku 93: 7–221. Marković, Č. 2012. Antička Budva: nekropole: istraživanja 1980–1981. Podgorica: Matica Crnogorska. Marović, I. 1971. Željeznodobni grobovi u Žaganj docu kod Sumartina (o. Brač). Vjesnik za arheologiju i historiju dalmatinsku 65: 1963–1965. Marović, I. and M. Nikolanci 1977. Četiri groba iz nekropole u Vičoj Luci (o. Brač) pronađena u 1908. god. Vjesnik za arheologiju i historiju dalmatinsku LXX– LXXI (1968–1969(1977)): 5–55. Mastrocinque, A. 1998. Da Cnido a Corcira Melaina – Uno studio sulle fondazioni greche in Adriatico. Trento: Università degli studi-Dipartimento di storia della civiltà europea-Facoltà di lettere e filosofia. Ostoich, N. 1858. Compendio storico dell’isola di Curzola. Zara: G. Woditzka. Papazoglu, F. 1967. Poreklo i razvoj Ilirske države. Godišnjak Centra za balkanološka ispitivanja III: 123– 145. Radić, D. 2001. Topografija otoka Korčule, in B. Čečuk, Ž. Rapanić and Ž. Tomičić (eds) Arheološka istraživanja na području otoka Korčule i Lastova, Izdanja Hrvatskog arheološkog društva 20: 25–50. Zagreb: Hrvatsko arheološko društvo. Radić, D. 2002. Uvod u arheološku topografiju Općine Blato. Blatski ljetopis 3: 231–246. Radić, D. 2003. Kopila, in T. Šeparović (ed.) Blato do kraja 18. stoljeća. Zbornik radova znanstvenog skupa održanog u Blatu u kolovozu 2003. godine, vol. 1: 63–88. Blato: Općina Blato. Radić, D. 2017. Metal finds, in I. Fadić and A. Eterović Borzić (eds) The city of the dead above the field of life. Necropolis of Kopila hillfort on the island of Korčula: 85– 97. Zadar: Muzej antičkog stakla u Zadru. Radić, D. and B. Bass 1998. Current Archaeological Research on the Island of Korčula, Croatia. Vjesnik za arheologiju i historiju dalmatinsku 90–91: 361–403. Radić, D. and B. Bass 2001. Gradine otoka Korčule (pretpovijesne strukture na vrhovima korčulanskih

References Arveiller-Dulong, V. and M. D. Nenna 2011. Les Verres antiques du Musee du Louvre III: Parure, instruments et elements d’incrustation. Paris: Louvre éditions. Batović, Š. 1986. Dalmatinska kultura željeznog doba. Radovi Filozofskog fakulteta u Zadru, Razdio povijesnih znanosti 25 (12): 5–60. Batović, Š. 2003. Boka Kotorska u prapovijesti. Zbornik Pomorskog muzeja Orebić: 191–238. Blečić Kavur, M. and B. Kavur 2017. Mnogo nijansi prozirnog: staklene perle u obliku amforiska iz Vičje luke. Vjesnik za arheologiju i historiju dalmatinsku 110/1: 93–112. Borzić, I. 2017. Ceramic finds, in I. Fadić and A. Eterović Borzić (eds) The city of the dead above the field of life. Necropolis of Kopila hillfort on the island of Korčula: 61– 83. Zadar: Muzej antičkog stakla u Zadru. Brunšmid, J. 1898. Die Inschriften und Münzen der griechischen Städte Dalmatiens. Wien: A. Hölder. Buljević, Z. 2002. Stakleni balzamariji. Longae Salonae: 383–454. Split: Arheološki muzej u Splitu. De Tommaso  G. 1990. Ampullae vitreae. Contenitori in vetro di unguenti e sostanze aromatiche dell’Italia romana (I sec. a.C.–III sec. d.C.) (Archaeologica 94). Roma: Bretschneider. Eterović Borzić, A. 2017a. Glass finds, in I. Fadić and A. Eterović Borzić (eds) The city of the dead above the field of life. Necropolis of Kopila hillfort on the island of Korčula: 99–116. Zadar: Muzej antičkog stakla u Zadru. Eterović Borzić, A. 2017b. Amber finds, in I. Fadić and A. Eterović Borzić (eds) The city of the dead above the field of life. Necropolis of Kopila hillfort on the island of Korčula: 117–124. Zadar: Muzej antičkog stakla u Zadru. Fadić, I. and A. Eterović Borzić (eds) The city of the dead above the field of life. Necropolis of Kopila hillfort on the island of Korčula. Zadar: Muzej antičkog stakla u Zadru. Grose, D. F. 1977. Early blown glass: the western evidence. Journal of Glass Studies 19: 9–29. Isings, C. 1957. Roman glass from dated finds. Groningen – Djakarta: J.B. Wolters. 352

Glass finds from Tomb 7 from the Late Iron Age necropolis of the Kopila hillfort

brda), in B. Čečuk, Ž. Rapanić and Ž. Tomičić (eds) Arheološka istraživanja na području otoka Korčule i Lastova, Izdanja Hrvatskog arheološkog društva 20: 133– 146. Zagreb: Hrvatsko arheološko društvo. Radić, D., I. Borzić and A. Eterović Borzić 2017. Catalogue, in I. Fadić and A. Eterović Borzić (eds) The city of the dead above the field of life. Necropolis of Kopila hillfort on the island of Korčula: 148–203. Zadar: Muzej antičkog stakla u Zadru. Radić, D. and I. Borzić 2017a. The island of Korčula on the crossroad from Praehistory to Antiquity, in I. Fadić and A. Eterović Borzić (eds) The city of the dead above the field of life. Necropolis of Kopila hillfort on the island of Korčula: 19–34. Zadar: Muzej antičkog stakla u Zadru. Radić, D. and I. Borzić 2017b. Kopila hillfort, in I. Fadić and A. Eterović Borzić (eds) The city of the dead above the field of life. Necropolis of Kopila hillfort on the island of Korčula: 35–60. Zadar: Muzej antičkog stakla u Zadru. Radić, D. and I. Borzić 2017c. Conclusions, in I. Fadić and A. Eterović Borzić (eds) The city of the dead above the field of life. Necropolis of Kopila hillfort on the island of Korčula: 139–145. Zadar: Muzej antičkog stakla u Zadru. Radić, D. and I. Borzić 2017d. Excavation of the necropolis of the Kopila hillfort near Blato on the island of Korčula – preliminary report, in D. Demicheli (ed.) Illyrica Antiqua in honorem Duje Rendić Miočević. Proceedings of the International conference, Šibenik 12th – 15th September 2013: 105–134. Zagreb: Odsjek za arheologiju Filozofskog fakulteta Sveučilišta u Zagrebu. Radić, D. and I. Borzić 2017e. The island of Korčula: Illyrians and Greeks. Vjesnik za arheologiju i historiju dalmatinsku 110/1: 303–325.

Radić, F. and V. Vuletić-Vukasović 1890. Predhistorički predmeti s otoka Korčule i poluotoka Pelješca u Dalmaciji. Viestnik hrvatskoga arkeologičkoga družtva XII/3: 73–78. Radovčić, D. 2017. Antropological analysis, in I. Fadić and A. Eterović Borzić (eds) The city of the dead above the field of life. Necropolis of Kopila hillfort on the island of Korčula: 125–138. Zadar: Muzej antičkog stakla u Zadru. Rendić-Miočević, D. 1966. Isejska naseobina u Lumbardi (Korčula) u svjetlu novih istraživanja. Vjesnik za arheologiju i historiju dalmatinsku LXVIII: 133–141. Rendić-Miočević, D. 1980. O knidskoj kolonizaciji Korčule. Diadora 9: 229–250. Suić, M. 2001. Korkyra h Melaina. Arheološka istraživanja na području otoka Korčule i Lastova, in B. Čečuk, Ž. Rapanić and Ž. Tomičić (eds) Arheološka istraživanja na području otoka Korčule i Lastova, Izdanja Hrvatskog arheološkog društva 20: 161-168. Zagreb: Hrvatsko arheološko društvo. Šašel-Kos, M. 2005. Appian and Illyricum. Ljubljana: Narodni muzej Slovenije. Triantafyllidis, P. 2003. Classical and Hellenistic glass workshop from Rhodes, in D. Foy and M. D. Nenna (eds) Echanges et commerce du verre dans le monde antique, Actes du colloque de l’AFAV Aix-en-Provence et Marseille 7–9 juin 2001 (Monographies Instrumentum 24): 131–138. Montagnac: Éditions Mergoil. Ugarković, M. 2019. Geometrija smrti: isejski pogrebni obredi, identiteti i kulturna interakcija. Antička nekropola na Vlaškoj njivi, na otoku Visu, I/1-2. Split, Zagreb: Arheološki muzej u Splitu, Institut za arheologiju Zaninović, M. 2015. Ilirski ratovi. Zagreb: Školska knjiga.

353

Roman glass from Lokvišća - Jadranovo site Bartul Šiljeg, Kristina Turkalj Abstract Since the beginning of the 20th century, various data on archaeological remains have been recorded in this area, indicating a continuous life during Roman rule in the area since the mid-1st century BC to the 5th century. The glass objects confirm the continuity of life at Lokvišća. Most of the items are dated to the 1st century so this indicates a higher intensity of life on the site. The oldest item is part of the rim and wall of a ribbed bowl made of light blue glass dating from the 1st century.

Keywords Lokvišća, Jadranovo, Roman glass, Antiquity, late Antiquity.

Figure 1. Orthophoto map of Croatia with indicated the wider region of Jadranovo and the position of Lokvišća site (basemap: DGU geoportal, digital orthophoto 2017/2018; image: K. Turkalj)

The advantageous position of Lokvišća Bay in Jadranovo was used as a port since prehistoric times (Figure 1). In addition to being a convenient port, it is also the closest point for crossing to the Island of Krk to the position suggestively named Voz. A naval battle that took place in the year 49 BC in this isthmus between Havišće / Lokvišća, Krk and the Islet of St Mark, between the opposing warlords in the civil war of Pompey and Julius Caesar has been recorded since the Roman times. Since the beginning of the 20th century, various data on archaeological remains in this area has been recorded (Brunšmid 1901: 118-120; Starac, Kružić 2014). At the end of the same century, archaeological excavations were carried out, which, in addition to the Bronze Age finds, also revealed valuable finds from Roman times in this area (Starac 2013; Starac, Kružić 2014; Lipovac Vrkljan, Starac 2014; BAZA). The findings indicate a continuous

life during the Roman rule at Havišće / Lokvišća, which can be traced back since the mid-1st century BC to the 5th century AD, i.e. the period of late Antiquity. The analysis of glass objects found during these researches provides almost the same impression. The glass dates from the 1st to the 5th century AD (Šiljeg 2016). A part of the rim and the wall of a ribbed bowl (Plate 1: 3) made of light blue glass (Isings 3c; Lazar 2.1.4.) dated to the 1st century is chronologically the oldest find, while the best preserved one is a two-handled footed beaker (Plate 1: 1) from which we have a preserved rim with a neck and a tubular thickening, a handle and a foot (Isings 38; Lazar 3.7.5., Transparentna 2013: 68: No. 93). This type of beaker is usually 9 to 12 cm high. They can be dated to the second half of the 1st century and are assumedly produced in Italy (Lazar 2003: 110; Gregl

Roman Pottery and glass manufactures (Archaeopress 2022): 354–361

Roman glass from Lokvišća - Jadranovo site

2009). A part of the ribbed decoration (Plate 3: 6; Plate 4: 6), which appears on bowls / Plates from the end of the 1st and in the 2nd century (Isings 43) can, also, be assigned to the same period. A Plate, similar to this one in colour, was found in Bakar (Gregl, Lazar 2008: 105). It is attributed to Cypriot workshops (Lazar 2003: 67, 1.3.2). The part of a bowl with the wall sloping outwards with a curved rim (Plate 2: 5) dates to the 1st century (Isings 41b; Lazar 2.4.1.). A similar type of bowl (Isings 42; Lazar 2.4.2) can be found with a horizontally curved rim, as well (Plate 2: 8). The rims of cups (Plate 2: 10) belong to simple shapes which we roughly date from the 1st to the 2nd century. The rim with a part of the wall (Plate 2: 11) with brushed decorations occurs from the 1st to the 4th century. Concavely drawn bases (Plate 2: 7; 3: 3) may belong to cups or bottles like the one of which a part of the neck with a rim is preserved (Plate 2: 6). Half of the leg of a footed cup (Plate 2: 4; Plate 4: 3) made of greenish glass dates from the mid- 1st to the 3rd century. The base of a tubular balsamarium (Plate 3: 5) is dated to the 1st - 2nd century. A part of the wall with a larger application (Plate 3: 7; T. 4: 5) precedes the favourite medieval decoration on glasses (krautstrunk). A fragment of bluish-green glass with an adhered handle (Plate 3: 9) can be dated to the 1st and the 2nd century and is a rare find in Roman glassmaking.

rim is straight and vertical. Feebly accented ribs continuing towards the base can be visible on the walls of the bowl. Dimensions: height 3.7 cm; width 5.1 cm, wall thickness 0.3 cm Dating: 1st century Analogies: Ising 3 c, Lazar 2.1.4. Bibliography: Šiljeg 2016: 75, 142. 2. Footed beaker with handles (Plate 1: 1) Footed beaker with two handles made of achromatic glass, free-blown. The rim with a neck, handle and the foot are preserved. The rim of the beaker is curved outward while there is a double folding on the neck. The handle is band-shaped with a widening on its upper part. The base is concavely drawn and the rim is folded inward to form a tubular widening. Dimensions: height 12 cm; rim diameter 9.5 cm; base diameter 3 cm, wall thickness 0.08 cm Dating: mid - 1st – mid - 2nd century Analogies: Ising 39; Lazar 3.7.5; Transparentna 2013: 68, 93. 3. Base of a conical goblet (Plate 3: 2; Plate 1: 2) The base of a cup formed into a foot is preserved as well as a part of the wall. The wall narrows towards the concavely drawn base so that the standing surface is the edge of the foot itself. The glass is greenish, and free-blown.

A small fragment of folded and fastened blue glass is apparently a part of a handle (Plate 3:11; Plate 4:4) of an aryballos (Isings 61; Lazar 2003: 8.3.3.). A foot-formed base of a cup (Plate 1:2; Plate 3:2) with a body that narrows towards the base and widens towards its rim (Isings 109a, Lazar 2003: 216, 3.8.6.) can, probably, be dated to a somewhat later period i.e. to the 3rd and the 4th century. Its assumed form can be seen on a cup from Solin (Transparetna 2013: 49, br. 59). Fragment of the rim and the body of a lamp with a handle (Plate 2:1; Plate 4:2) (Isings 134; Lazar 9.2.1.) can be dated from the 4th century up to the Middle Ages (Uboldi 1995). Olive-green hemispherical cups (Plate 2:2; Plate 4:1) with an outward curved rim are dated to the 4th and the 5th century. Other items are either too fragmented or too untypical to be precisely dated so all we can say is that they originate from the Roman times.

4. Balsamarium (Plate 3: 5)

Glass findings from Lokvišće fit into the frame determined by similar finds from Bakar and Crikvenica - Igralište site. Most of the items can be dated to the 1st century, which points to a higher intensity of life on the site with continuance from the prehistoric times to the Middle Ages.

5. Foot (Plate 2: 4; Plate 4: 3)

Dimensions: height 3.2 cm; base diameter 5.2 cm, wall thickness 0.15 cm Dating: 3rd - 4th century Analogies: Ising 109a; Lazar 3.8.6.

The base of a tubular balsamarium. Greenish, freeblown glass. Rounded base. Probably a tubular balsamaria type. Dimensions: height 2.1 cm, wall thickness 0.3 cm Dating: 1st - 2nd century Analogies: Lazar 8.6.1.; Baćani 2017: cat. n. 4, 8.

The foot of a cup or a beaker, concavely drawn, with an inward folded rim that makes a tubular widening. The glass is light-green.

Catalogue

Dimensions: base diameter 4.6 cm Dating: mid - 1st ct - 3rd century Analogies: Transparentna 2013: 65, 91; Ising 36, 38a, 40; Lazar 3.8.3.

1. Ribbed bowl (Plate 1: 3) A part of the rim and the wall of a blue glass bowl. Made by folding of a glass disc over the mould. The 355

Bartul Šiljeg, Kristina Turkalj

Plate 1. Glass finds from Lokvišća (drawings: S. Čule; photos: H. Jambrek).

356

Roman glass from Lokvišća - Jadranovo site

Plate 2. Glass finds from Lokvišća (drawings: S. Čule).

357

Bartul Šiljeg, Kristina Turkalj

Plate 3. Glass finds from Lokvišća (drawings: S. Čule).

358

Roman glass from Lokvišća - Jadranovo site

6. Base of a cup or a bottle (Plate 3: 3)

Dimensions: rim diameter, 11.5 cm, wall thickness 0.08 cm Dating: 1st - 2nd ct. Analogies: Ising 41b; Lazar 2.4.1.

A concavely drawn base with traces of a metal holder on its underside. The glass is greenish, free-blown.

13. Bottle rim (Plate 2: 6)

Dimensions: width 3.8 cm, wall thickness 0.3 cm

Slightly outward curved rim, drown-out and folded inward. Greenish, free-blown glass.

7. Base of a cup or a bottle (Plate 2: 7) A concavely drawn base with traces of a metal holder on its underside. The glass is greenish, free-blown.

Dimensions: rim diameter 5.3 cm 14. Cup rim (Plate 3: 4)

Dimensions: width 3.8 cm, wall thickness 0.2 cm

The rim is slightly curved outward and thickened. Achromatic, free-blown glass.

8. Base of a cup or a beaker (Plate 2: 3) A slightly concavely drawn base. The glass is greenish, free-blown.

Dimensions: pr. rube 6.6 cm, wall thickness 0.1 cm Dating: 1st - 2nd century Analogies: Transparentna 2013: 47, 55.

Dimensions: width 6 cm, wall thickness 0.2 cm

15. Bowl rim (Plate 2: 5)

9. Lamp with a vertical handle (Plate 2: 1; Plate 4: 2)

The rim is slightly curved outward and thickened. Greenish, free-blown glass.

The rim and the body of a lamp with the handle is preserved. The glass is greenish, transparent, freeblown. The rim is curved and the handle is attached to the very edge of the lamp, positioned vertically downward.

Dimensions: rim diameter 11.5 cm, wall thickness 0.08 cm Dating: 1st - 2nd ct. Analogies: Ising 41b; Lazar 2.4.1.

Dimensions: height 4 cm, wall thickness 0.1 cm Dating: 4th - 7th century Analogies: Ising 134; Lazar 9.2.1.

16. Fragment of the rim with a part of the wall (Plate 2: 11)

10. Cup rim (Plate 2: 2; Plate 4: 1)

Thickened rim, walls decorated with several zones of shallow, dense incisions. Greenish, free-blown glass.

The outward curved, straight cut rim of a hemispherical cup. Olive-green, transparent, freeblown glass.

Dimensions: height 3.4 cm, wall thickness 0.08 cm Dating: 1st - 4th century

Dimensions: height 3.3 cm, wall thickness, 0.1 cm Dating: 3rd - 5th century Analogies: Ising 96; Lazar 3.10.1; Transparentna 2013: 50, 63, 64. 11. Cup rim (Plate 2: 10)

17. Fragment of the vessel wall (Plate 2: 9) The wall is decorated with a strap of shallow, dense incisions. Greenish, transparent glass.

Slightly outward curved and rounded rim, Greenish, free-blown glass.

Dimensions: height 2.9 cm, wall thickness 0.1 cm 18. Ribbed band (Plate 3: 6, Plate 4: 6)

Dimensions: rim diameter 7.6 cm Dating: 1st - 2nd century Analogies: Transparentna 2013: 48, 57.

A fragment of a ribbed decoration adhered on the vessel wall. Yellowish glass.

12. Bowl rim (Plate 2: 5)

Dimensions: width 2.3 cm, wall thickness 0.15 cm Dating: 1st - 2nd ct. Analogies: Ising 43; Lazar 1.3.2.

Outward curved, thickened rim, Greenish, freeblown glass.

359

Bartul Šiljeg, Kristina Turkalj

Plate 4. Glass finds from Lokvišća (photos: H. Jambrek).

360

Roman glass from Lokvišća - Jadranovo site

19. Fragment of the cup wall (Plate 3: 7; Plate 4: 5)

References Baćani, I. 2017. Staklo Siscije/Glass of Siscia. Sisak: Gradski muzej. Brunšmid, J. 1901. Arheološke bilješke iz Dalmacije i Panonije, IV. Viestnik Hrvatskoga arkeologičkoga družtva V: 87‒169. Isings, C. 1957. Roman Glass from Dated Finds (Archaeologica Traiectina 2). Groningen: Wolters. Gregl, Z. 2009. Ranocarski stakleni pehari na nozi s dvije ručke iz Hrvatske. Histria antiqua 18(1): 433‒438. Gregl, Z. and I. Lazar 2008. Bakar: staklo iz rimske nekropole/The glass from the Roman cemetery (Katalozi i monografije Arheološkog muzeja u Zagrebu sv. 5). Zagreb: Arheološki muzej u Zagrebu. Lazar, I. 2003. Rimsko steklo Slovenije (Opera Instituti archaeologici Sloveniae 7). Ljubljana: ZRC SAZU. Lipovac Vrkljan, G. and R. Starac, R. 2014. Antička mikrotopografija Hrvatskog primorja. Prilog poznavanju arheološkog krajolika Vinodola prema nalazima crikveničke keramike, 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: Muzej Grada Crikvenice, Institut za arheologiju. Starac, R., 2013. Jadranovo – Lokvišće. Hrvatski arheološki godišnjak 9/2012: 503‒504. Starac, R. and T. Kružić 2014. Jadranovo – Lokvišće, Rtac. Hrvatski arheološki godišnjak 10/2013: 399‒401. Šiljeg, B. 2016. Nalazi rimskog stakla na području Crikvenice i Vinodola, in A. Konestra and T. Rosić (eds) 845°C Ad Turres (exhibition catalogue): 73‒76. Crikvenica: Muzej Grada Crikvenice. Transparentna 2013 = Gregl, Z. and M. Leljak 2013. Transparentna ljepota. Staklo iz hrvatskih muzeja od pretpovijesti do srednjeg vijeka (Exhibition catalogue). Zagreb: Arheološki muzej u Zagrebu. Uboldi, M. 1995. Diffusione delle lampade vitree in eta tardoantica e altomedievale e spunti per una tipologia. Archeologia Medievale 22: 93‒145.

Fragment of the cup wall with an oval appliqué. Greenish, free-blown glass. Dimensions: width of the appliqué 2 cm, wall thickness 0.1 cm Dating: 3rd - 4th ct. to the 5th ct. Analogies: Ising 107b; Lazar 3.10.2 20. Part of the handle (Plate 3: 11; Plate 4: 4) Fragment of the upper part of a band-shaped handle folded outward. Light-blue glass. Dimensions: height 1.5 cm, width 0.4 cm Dating: 1st - 3rd century Analogies: Ising 61; Lazar 8.3.3. 21. Window glass (Plate 3: 1) Achromatic glass. Mould casting technique. Dimensions: wall thickness 0.2 cm 22. Fragment of the handle with a part of the wall (Plate 3: 9; Plate 4: 7) Fragment of a bluish-green glass with an adhered handle (appliqué). Dimensions: height 2.8 cm, wall thickness 0.2 cm Dating: 1st - 2nd ct. Analogies: Lazar 2.4.7. 23. Fragment of the base (Plate 3: 13) The base of a vessel with an adhered ring-shaped foot. Achromatic glass. Dimensions: width of the foot 0.8 cm

Internet sources

24. Fragment of the base (Plate 3: 12)

BAZA = Baza antičkih arheoloških lokaliteta Republike Hrvatske Instituta za arheologiju (BAZA), Havišće, viewed 8.11.2019., .

Fragment of milky-white glass, foot of a vessel. Dimensions: width 2 cm 25. Glass fragment (Plate 3: 8) Fragment of achromatic glass with a visible rib on the wall. Dimensions: width 2 cm, wall thickness 0.1 cm 26. Fragment of the rim (Plate 3: 10) Fragment of achromatic glass. Rounded and thickened rim and a rib visible on the wall. Dimensions: height 2.6 cm 361

Experience in the conservation and restoration of archaeological glass at the Archaeological Museum of Istria in Pula Monika Petrović Abstract This paper describes a brief history of the founding of the Archaeological Museum of Istria and its conservation and restoration department. The focus is on the protection of archaeological glass and experience gained in the course of its restoration. Glass preservation procedures are described from the moment of excavation to the restoration treatment phase (cleaning, integration, reconstruction, consolidation). The desalination of finds is also discussed in light of a certain percentage of finds recovered from marine environments, as is the process of fabricating replicas and the proper storage and exhibition of glass artefacts.

Key words Preservation of archaeological glass, restoration of glass, consolidation and desalination, reversibility, reconstruction, replica.

Introduction During the first hundred years of its formation our museum has developed out of a municipal and then royal museum to its present form as the Archaeological Museum of Istria, with the Croatian culture ministry as the founding institution. There are few institutions with this kind of lengthy continuity in Pula, Istria County or Croatia as a whole. The founding of the first museum institution in Pula in the early years of the twentieth century was preceded by the collection of stone artefacts that lay unprotected in the city and its nearby environs and their relocation to the Temple of Augustus on the Forum and later to the Roman amphitheatre. There had been a number of initiatives to found a museum in Pula in the course of the nineteenth century. The investigation of Nesactium near Valtura by the Istrian Society for Archaeology and Regional History and the successful outcome of the excavation of this prehistoric and antiquity period site laid the groundwork for the founding of a museum. The Museum of the City of Pula (Museo Civico della Città di Pola) was founded on the 7th of January 1902 by the municipal council of Pula, acting on a proposal from the provincial council. After the First World War, in 1925, the institution was merged with the State Municipal Collection (stone monuments) and the Provincial Museum (Museo Provinciale) in Poreč to assume the stature of a provincial level institution as the Royal Museum of Istria (Regio Museo dell’Istria). By 1930, with the great number of monuments that had been collected kept at the existing museum building, the Temple of Augustus and the Roman amphitheatre, it was evident that a more appropriate and dignified setting was required

for the exhibition of these artefacts. The most suitable edifice in which to relocate the museum was the former building of the German gymnasium, located on the eastern slopes of the old town core, between the Roman-built Gate of Hercules and Double Gate (Porta Gemina) (Zenzerović: 2012). In 1947, following the departure of the Allied administration, the education ministry of the then People’s Republic of Croatia again renamed the institution as the Archaeological Museum of Istria. Numerous new investigations and, in particular, the museological processing of cultural property that had been returned, set the stage for the further development of the museological and scientific work of the institution. A preparator’s workshop is established in 1951, staffed by newly hired preparator Ljubica Horvat, to work on the preservation of stone monuments kept at the Temple of Augustus and the Roman amphitheatre. As the institution grew and ahead of the creation of a permanent exhibition in 1970 the museum hired Galliano Zanko (1961), Ljerka Krleža (1965), Ida Zanko and Jolanda Bilić (1969, pottery/ glass) – they worked tirelessly on the treatment and presentation of the museum holdings. Permanent collections were established, faithful to the specialist criteria, featuring stone monuments (lapidarium), prehistory, the antiquity, and the Middle Ages. As time passed and the scope of work grew, new staff members were hired to work in the preparator’s workshop. These were Hana Filiplić (1977, metal), Jadranka Homoky and Aldo Monfardin (1979, pottery/souvenirs), Slobodan Božičković (1993, stone restorer), and Zoran Grbin (1999, pottery/mosaic). There was also a growing need for other specialist staff, leading to the hiring of Đeni Gobić-Bravar (2002, restorer of pottery, glass, metal and stone), Andrea Sardoz (restorer of pottery, glass and

Roman Pottery and glass manufactures (Archaeopress 2022): 362–366

Experience in the Conservation and Restoration of Archaeological Glass

metal), and Monika Petrović (2006, restorer of pottery, glass and metal). In 2008 the workshop was formally recast as a fullyfledged conservation and restoration department. In its work it observes all the general norms pertaining to the preservation of archaeological material – documented in detail (descriptions in restoration record cards, photographic documentation prior to, during and after all the phases of work), and to consolidation and restoration. Conservation and restoration works are carried out conscientiously and the products used are reversible. The following text describes in detail the procedures used in the preservation and work on archaeological glass. Archaeological Glass Work Methodology Samples should be collected of the soil from which glass finds are recovered in the course of archaeological investigation. An analysis of these samples can assist us in understanding the degradation processes impacting the glass artefact and in identifying the cause of degradation. When working with very fragile fragments the ground (earth) surrounding it can protect the artefact, and these fragments are thus kept in suitable microclimatic conditions until they arrive at the conservation and restoration department for further treatment.

Figure 1. Consolidation of archaeological glass.

while stubborn deposits of impurities can be removed by rinsing in a solution of 50 % alcohol and 50 % demineralised water. The cleaning of the archaeological glass is followed by an examination of its condition and, if necessary, consolidation treatment.

Glass finds recovered from wet or waterlogged sites will very quickly lose moisture if suddenly exposed to light and air, leading to irreparable damage. Degradation also occurs to finds recovered from very dry sites, as they tend to absorb too much moisture from the atmosphere. In the course of archaeological excavation finds are impacted by a number of factors – the composition of the find (glass), the environment of the site, and temperature.

Consolidation or the reinforcement of ‘sick’ glass (glass disease) is a direct activity on and within an artefact with the objective of stabilising the object and retarding further degradation. Consolidation of archaeological glass is performed when there has been significant damage (fragile/brittle glass, flaking), or damage caused by the impact of external (humidity, temperature) and internal (the composition of the glass) factors, i.e. when the structural stability of the glass has been compromised. The consolidant is a 10 to 15 % solution of Paraloid B-72 in acetone. A layer of consolidant should be carefully applied to a glass find using a brush (Figure 1).

In general work on archaeological glass is very demanding and requires great precision, whether this pertains to curative conservation or to restoration interventions. The general norms for the preservation of archaeological glass are followed, valid for all restorations, and every intervention (cleaning, integration, reconstruction or consolidation) is documented in detail (described in restoration record cards, and photographed during all phases of work). Every artefact or fragment is treated separately in accordance with its condition, gradually until the artefact is in a stable condition.

Integration and bonding are phases of work that follow cleaning. The glass fragments are grouped in sets based on artefact typology. The selection of adhesive prior to the re-composition of preserved fragments is exceedingly important given the requirement of the reversibility of the procedure and of the materials used. Attention must also be given to the neutrality of the substance in relation to the original surface, that physical and visual characteristics are not altered over time and when exposed to various forms of radiation, and the low viscosity of adhesive required for capillary penetration at the joints (Perović 2008: 16–17) (Figure 2).

Primary processing, cleaning, is the first treatment in the preservation of glass finds. Finds are first cleaned of contaminants (soil and dust) with soft PVC brushes,

Restoring the form of the artefact begins with the preparation of adhesive tape cut into dimensions appropriate to the form of the object. Glass fragments 363

Monika Petrović

Figure 2. Integration and bonding of archaeological glass.

Figure 3. The reconstruction of missing sections.

are joined along the breakage lines and temporarily fixed in place with strips of adhesive tape. When the artefact has been reconstituted with the adhesive tape the join lines are reinforced with brass clips that are glued to the glass with quick bonding cyanoacrylate adhesive. With all the preserved fragments of the object held in place in this manner a two-component epoxy adhesive (Araldite 2020 A/B) is applied at the breakage lines. The adhesive is applied with a syringe to the joints only – due to its exceedingly low viscosity capillary action allows it to penetrate into the cracks and glue the glass together. The two-component epoxy adhesive is allowed to dry for 24 hours, whereupon the metal clips are removed with a gentle tapping action using cotton swabs soaked in acetone. Once the clips have been removed any residual adhesive is mechanically removed from the original using a scalpel, taking care not to damage the surface of the glass.

the original using a precise two-component vinylpolysiloxane silicone. The mould thus produced is then carefully positioned at the appropriate position right next to the original glass and attached with cyanoacrylate adhesive at the location of the missing section of the artefact. Perforations are made at suitable places on the outer mould through which the two-component epoxy adhesive will be poured and through which air will be vented out (one or more as required). PVC straws are then inserted at the perforations. The artefact is stabilised in a suitable position in a sand-filled vessel, whereupon the two-component epoxy adhesive is poured into the lowest straw. The reconstructed section is then allowed to dry for 24 hours before the silicone mould is taken off and any excess material is removed with swabs soaked in acetone and mechanically with the aid of a scalpel. The better the adhesion (integration) of the reconstructed section, the more faithful the reconstructed section will be to the original. Care must be given to the visual impression when undertaking a reconstruction. Differentiation of the original and the reconstructed section is achieved by differences in the tone or transparency of the reconstruction (Figure 3).

The next phase is reconstruction, a stage of work that requires the selection of the appropriate material(s) with which to reconstruct missing sections. A specific requirement of the adhesive substance is that it meets the condition of reversibility and that it is neutral in relation to the original artefact. The reconstruction of glass is a very specific technique that requires a capable restorer and a solid understanding of the materials. A test should be performed before an actual reconstruction is attempted: a silicone mould is made from the corner of the artefact (the mould is fabricated as a single piece for easier handling). The test serves to show how a reconstruction will turn out (tests are done when working with toned, engraved, decorated and similar glass artefacts). This is followed by the fabrication of the reconstructions of missing sections. A mould is made of the external and internal faces of

As has already been noted the procedures involved in the preservation of archaeological glass begin at the excavation stage, through the conservation and restoration treatment stage and, later, in the proper storage of glass finds. Desalination is a necessary first step when working with glass finds recovered in the course of underwater archaeological investigations. The procedure involves immersing the glass finds in a closed vessel filled with (distilled or de-ionised) water. The water is changed at regular intervals, with 364

Experience in the Conservation and Restoration of Archaeological Glass

Figure 4. Fabricating a replica of a glass artefact.

Figure 5. Packing archaeological glass.

samples taken for testing at each change of water. The desalination process is performed until the chloride concentration is reduced to a minimum level (Petrović 2014: 111–124). The final step in the preservation of archaeological glass is its adequate packing and storage in depots and its preparation for future exhibition (Figure 5).

from 45 to 55 per cent and a sustained temperature within the 15 to 25°C range. All sources of heat (including lighting) should be carefully shaded or regulated when exhibiting artefacts given that heat may induce degradation. All light, weak or strong, causes damage – the difference is only in the level of damage caused. The intensity of light (illuminance) is measured with a light meter in lux units. The maximum illuminance recommended for glass objects is in the 50 to 70 lux range (Osnove zaštite i izlaganja muzejskih zbirki, 1993).

The Storage and Exhibition of Glass Archaeological Artefacts Relative humidity in storage depots should range between 45 and 60 %, and the ambient temperature between 10 to 25°C, without any significant oscillation. Finds are stored in acid-free boxes to avoid any contact with dust and exposure to (sun)light. When there are good conditions for the storage of stable finds glass artefacts are wrapped in acid-free paper. The next step involves fabricating a lining of firm foam for additional protection of an artefact kept in a box. Glass fragments (in small groups) are kept in closed finelyperforated nylon bags, stacked upright and packed with foam lining, and stored in cardboard or plastic boxes. In unfavourable conditions glass finds must be kept in hermetically sealed plastic boxes. Silica gel (a hygroscopic amorphous silicon dioxide, SiO2) is placed in the boxes to provide additional protection. The quantity of silica gel that is added is equal to the weight of the stored glass find(s). In these conditions it is critical that the artefacts are subjected to regular monitoring.

Resolving issues related to reconstruction solutions is of particular importance when considering exhibition and presentation, as is the skill of the restorer in fabricating replicas of artefacts. When dealing with rare and highly fragmented archaeological glass, where there are insufficient elements for a reconstruction, a replica of the object is fabricated using an epoxy mixture. A lack of elements for a reconstruction make the fabrication of these replicas a demanding process. When undertaking this task one needs to carefully consider the methods and materials used in order to most faithfully replicate the basic characteristic of glass – its transparency. The object is fabricated using a modified lost-wax sculpting process. The entire object is shaped in clay and then coated with liquified wax, whereupon a silicone mould is made. Once the silicone mould has hardened the clay is carefully extracted. PVC straws are attached to the mould to serve as vents through which the epoxy mixture (Araldite 2020) will be introduced.

Degradation prevention measures must be in place when exhibiting glass artefacts. These include: a controlled and constant relative ambient humidity of 365

Monika Petrović The one-off silicone mould is removed after a 24hour drying period. The replica of the glass artefact is cleaned of excess epoxy and polished (Figure 5).

References Osnove zaštite i izlaganja muzejskih zbirki, 1993. Muzejski dokumentacijski centar, Zagreb. Perović, Š. 2008. Antičko staklo: restauracija. Zadar, Muzej antičkog stakla u Zadru. Petrović, M. 2014. Očuvanje arheološkog stakla. Histria archaeologica 44: 111–124. Pula. Zenzerović, K. 2012. Prošlost za budućnost: Izložbena djelatnost Arheološkog muzeja Istre 1902.–2012., Uz 110. obljetnicu Muzeja. Monografije i katalozi br. 20. Pula, Arheološki muzej Istre.

Advances are achieved in reconstruction solutions using systems applying 3D scanning and printing, and which have seen the successful resolution of the problem of scanning transparent and reflective materials such as glass. Conclusion Pula’s Archaeological Museum of Istria, which has operated under this name since 1947, continues on its tradition and mission to collect, preserve, treat and present archaeological material. A preparator’s workshop was founded at Pula’s Archaeological Museum in 1951 and transitioned in 2008 into the current conservation and restoration department. The department now has a staff of four restorers and one preparator working on the restoration and protection of our heritage monuments. The workshop is specialised in the restoration of stone, ceramics, glass and metals. The department has a thriving collaboration with the METRIS Istria County Materials Research Centre, which performs a range of chemical and physical analyses that are foundational to the planning of conservation and restoration procedures. The curative treatment and restoration department also serves other museums in Istria County and performs conservation, restoration, proper packaging and the monitoring of the condition of archaeological material. The Archaeological Museum of Istria and its conservation and restoration department have made the long trek from small municipal museum to major national culture institution. That, of course, is not the end of the road! Work is ongoing on the adaption of the museum building, and a new permanent exhibition is being created that calls for modern, interactive, multimedia and novel approachs. The museum curators and the staff of the conservation and restoration department will once again be put to the test in the treatment and presentation of archaeological material.

366

Experimental archaeology in the Museum of Ancient Glass in Zadar Marko Štefanac, Berislav Štefanac Abstract This paper presents the first results of the experimental research conducted in the glass workshop of the Museum of Ancient Glass in Zadar. The experiments are devised with an aim of providing answers to specific questions related to methods of manufacture of certain Roman artifacts. Various replicas of glass from antiquity were made in the experiments, using authentic techniques only. Tests based on objects of simple morphology (spindle-shaped bottles, bird-shaped dropper flasks, juglets with conical body) have offered a detailed insight into an interesting experiential aspect of the research problem. Authentic techniques were used to enable better understanding of scientific insights alongside precise perception of forms and contents.

Key words Museum of Ancient Glass, experimental archaeology, Roman glass.

Introduction Primary activity of the Museum of Ancient Glass in Zadar is based on exhibiting and presenting rich Roman heritage, alongside continuous preservation, protection and professional treatment of the material. In addition to rich museum activity, work in the field of experimental archaeology is another important segment. So far several projects have been realized based on experimental procedures. Experience acquired in the glass workshop is used to test a number of hypotheses with an aim of improving knowledge. All basic tests are performed on the basis of archaeological information and theoretical notions of ancient glassmaking. A number of methods, techniques and analyses have been used so far. All types of experimental procedures (replicating Roman-era artifacts, analysis of technological processes) are based on archaeological finds recovered from several sites in Croatia. Interdisciplinary approach was used to make a considerable number of ancient glass copies, by using original techniques. Important information about several different techniques used in antiquity has been obtained so far. Certain experimental studies conducted through the years were associated with replicating molds for the production of glass artifacts. Glass workshop led by glassblower Marko Štefanac offers a view into formation of many glass receptacles. Daily demonstrations mimic past phenomena in controlled experimental procedures. Owing to active research work and experimental procedures, the Museum of Ancient Glass became the first permanent institution in Croatia dealing with scientific experiments in the field of ancient glassmaking. First scientific-archaeological

experiments in the Museum of Ancient Glass were conducted in the year 2012 within the project Roman Glass in Croatia – Relief Workshop Stamps. All previous research results are regularly published in scientific papers and video clips. Glass workshop Specialized workshop of the Museum of Ancient Glass in Zadar is an ideal setting for implementation of longterm experimental studies (Figure 1). It is equipped with modern melting furnaces, chambers for glass cooling and tools. Various blowpipes and other small glassmaking implements (different types of scissors and tweezers) have been used in all educational and experimental works. Except for basic tools, molds made of various materials, primarily wood and clay, are used in the manufacture process. The main part of the production process relates to the electric glass melting furnace. Modern sodalime-silica glass is used in the workshop, as usual in contemporary glass studies. Melting point is 1150 °C, somewhat higher than the operating temperature for the Roman glass (1000 to 1050 °C) mentioned by Brill, used in maps brought by Stern and Schlick-Nolte (Stern 1995: 34–48). Regardless of that, its working properties are still comparable to the ones used by most Roman glassblowers. Metal pipe is used to shape the desired form by blowing, to manipulate molten glass and make the object. Modern pipe is almost identical to the ones used in the early days of glassblowing. In addition to melting furnace, the workshop is also equipped with annealing ovens. Constant control of

roman pottery and glass manufactures (Archaeopress 2022): 367–372

Marko Štefanac, Berislav Štefanac

Figure 1. Glass workshop in the Museum of Ancient Glass in Zadar (photo: B. Štefanac).

temperature variables is necessary for performing experimental procedures, in particular for making delicate thin-walled vessels. The object is first stabilized by short reheating in the annealing oven and then placed in the cooling chamber. If glass is cooled too quickly, it does not shrink evenly, and due to temperature changes in glass, structure becomes nonhomogenous and strain is created. In order to avoid permanent strain in the material, slow and continuous cooling is necessary. Experimental work Experiments conducted in the workshop of the Museum of Ancient Glass are devised with an aim of providing answers to specific questions related to methods of manufacture of certain Roman artifacts. Various replicas of glass from antiquity were made in the experiments, using authentic techniques only. Tests were performed on three different ancient items that feature early imperial glassmaking workshops. These artifacts are rather simple in terms of morphology, but each one is specific in some regards. Spindle flasks

Figure 2. Spindle flasks from ancient necropolis in Zadar – Archeological Museum Zadar, permanent exhibition of the Museum of Ancient Glass (photo: S. Govorčin).

One of the prominent products in antiquity were spindle flasks (Figure 2). They can be found in a number 368

Experimental archaeology in the Museum of Ancient Glass in Zadar

Figure 3. Process of making a spindle flask (photo: B. Štefanac).

of variants over a long period. Most finds of this kind in our region have been documented in Zadar and its environs (Štefanac 2011: 377–388). On the basis of morphological characteristics, vessels can be divided into two groups, small flasks with wide body and opening, and the bigger ones with narrow, elongated body. Their use was versatile. Smaller specimens were probably used for keeping and transporting cosmetic substances, essential oils and ointments, while bigger ones could have been used in households for storing oil and wine (Štefanac 2011: 377–388).

which will be inflated is formed at the tip of the pipe (Figure 3: 1). Blowpipe rotation and gravity are used to get a preform – a flask with short neck and ovoid body (Figure 3: 2). The second phase that refers to shaping of the basic form demands exceptional skill. At this stage additional heating of rounded walls of the flask is crucial until reaching certain stage of softness and viscosity. When the walls are soft enough, the object is taken out of the furnace, directed downwards and shaped by gravity (Figure 3: 3–4). This procedure should be repeated several times due to exceptionally thin walls of the flask that quickly lose softness and elasticity. Once the satisfying form is obtained, it is necessary to transfer the object from blowpipe to metal rod (pontil) to shape the rim.

During the experimental research, several different replicas of spindle flasks have been made. Theoretical insights related to their manufacture process have been tested in several repetitions. Process of making spindle flasks can be divided into three phases: 1. 2. 3.

Upper part is modeled in the third and final phase. The neck is reheated to certain stage of viscosity to enable shaping the rim in the following step. Tests have shown that simple pointed wooden stick is best for precise rim modeling (Figure 3: 5).

making preform shaping basic (spindle-shaped) form finishing touches (shaping rim)

Production of a preform is the first stage of manufacturing basic spindle-shaped form. The process starts with gathering molten glass from the melting furnace by a metal blowpipe. Incandescent cylinder

During the experimental procedure it is necessary to monitor closely the temperature of walls and pay attention to their timely reheating to avoid stress and possible fracture of the object. 369

Marko Štefanac, Berislav Štefanac antiquity. Owing to their practical form they could be used in medicine and pharmacy for precise dosage of healing substances, or in cosmetics for keeping and transferring various perfumed liquids. They could have been used in Roman daily life as a piece of common household inventory. Flasks could be used for refilling oil lamps, serving sauces and spices, preparing drinks, feeding children etc. The manner of their production can be defined in theoretical terms, but nevertheless experimental activities were performed. During the procedure, several different replicas were made suggesting certain conclusions.

Figure 4. Bird-shaped dropper flasks from ancient necropolises in Zadar and Nin – Archeological Museum Zadar, permanent exhibition of the Museum of Ancient Glass (photo: I. Jukić).

As with the spindle flasks, the production process can be divided into three phases:

Bird-shaped dropper flasks

1. making preform 2. shaping basic (bird-shaped) form 3. finishing touches (shaping rim)

A number of atypical forms have been made during the development and expansion of ancient glassmaking. Among them, examples of bird-shaped dropper flasks are particularly noteworthy (Figure 4). Different variants of such products have been found at several sites in Croatia. Archaeological research indicates that dropper flasks were used over a longer time span, from the mid-1st to late 4th century (Štefanac 2009: 115–148). Glass droppers were highly functional in

Making preform is the first stage in producing basic bird-shaped form. First molten glass is gathered to create incandescent cylinder which is then inflated (Figure 5: 1). Uniform spinning of the blowpipe and moderate air inflating are used to get the desired form that resembles a little bottle with spherical body (Figure 5: 2).

Figure 5 . Process of making bird-shaped dropper flasks (photo: B. Štefanac).

370

Experimental archaeology in the Museum of Ancient Glass in Zadar

In the third and final phase, neck of the dropper flask is shaped. This is preceded by a procedure of separating the object from the pipe and taking it with a metal rod. The neck is reheated to a certain stage of viscosity so that the vessel rim might be shaped in the following stage (Figure 5: 4). Rim is modeled by a simple pointed wooden stick that does not leave a trace of working (Figure 5: 5). After the procedure, slow and continuous cooling is required because the walls are very thin. Conical juglets with ribbed handle

Figure 6. Juglets with conical body and ribbed handle from ancient necropolis in Zadar – Archeological Museum Zadar, permanent exhibition of the Museum of Ancient Glass (photo: I. Jukić).

In the previous experimental activities, special attention was paid to manners of execution of minute details, by using authentic techniques. Glass juglets with ribbed handle offered good base for the study of details on the vessels. Such products were popular in the 1st century across the Mediterranean area (Kunina 1997: 302, cat. nos. 242–244; Stern 2001: 95, cat. no. 32; Lightfoot 2017: 134–135, cat. no. 141). The form was recorded on the eastern Adriatic coast, in particular in ancient southern Liburnia. So far over 20 complete examples dating to the second third of the 1st century have been recovered from the Roman necropolis in Zadar (Figure 6). Early imperial glass workshops produced two variants of products, juglets with flat base and juglets on raised cylindrical foot.

Shaping basic bird-shaped form follows after additional reheating of the vessel. In the reheating phase, special attention is paid to reduce it to the narrow parts of the bottle. When certain level of softness and viscosity is reached, the phase of body modeling begins. Soft viscuous walls are stretched by tweezers into desired form (Figure 5: 3). It is important to emphasize that creating stylized tail is performed in one movement because glass cools quickly. Since walls are exceptionally thin, the object is stabilized by short reheating in the melting furnace before the final phase.

Figure 7. Process of making a conical juglet with ribbed handle (photo: B. Štefanac).

371

Marko Štefanac, Berislav Štefanac Since reliable information on the manner of their production is lacking in professional literature, extensive testing has been conducted. Execution of this experiment is important to corroborate hypotheses based on the fact that complete form of the body was made by using the free blowing technique.

juglets with conical body) offered a detailed insight into an interesting experiential aspect of the research problem. Experimental works were conducted with an aim of testing hypotheses related to the manner of production and understanding of morphological characteristics, with a certain number of replicas made.

During the process of replication, several conical juglets were made to test certain conclusions. The results of the tests have provided new insights about freely blown products.

Variables and changeable values affecting the research were clearly identified in the experiment. In that context, we need to consider use of modern melting furnaces and glass raw material. Performing experimental activities led to more complex considerations and conclusions about the technological processes that happened in antiquity. The first results were visible after many repetitions of the experimental procedures. Alongside multiple repetitions of the experiments and monitoring control of temperature variables (analysis of cooling and reheating an object during the production process), clear information was collected about possible manners of executing minute details on freely blown products. Use of authentic techniques enabled better understanding of theoretical insights alongside precise perception of form and contents. All experimental tests have been presented in an illustrative and dynamic way which will definitely facilitate future research and learning.

Process of juglet production can be divided into three phases: 1. making preform 2. shaping basic (bird-shaped) form 3. finishing touches (shaping handle and rim) Preform production belongs to the first stage of shaping basic conical form. As in previous examples, the process starts with gathering molten glass from melting furnace with a blowpipe. Incandescent cylinder is formed at the tip of the blowpipe, which is then inflated (Figure 7: 1). Blowpipe rotation and gravity are used to shape preform similar to a tubular ampulla (Figure 7: 2). Tool (tweezers) is used only to shape the constriction between body and neck.

References

Skill and experience of the glassblower are crucial in the second phase. Shaping of conical (piriform) body follows after additional reheating of the vessel walls. Air is inflated to create the basic juglet form (Figure 7: 3), while form details (cone level, flat base) are flattened with a simple wooden tool or tweezers. Once the satisfying form is achieved, the object needs to be transferred from the blowpipe to the metal rod (pontil) to shape the upper part of the body.

Kunina, N. 1997. Ancient Glass in The Hermitage Collection. Sankt Peterburg. Lightfoot, C. S. 2017. The Cesnola Collection of Cypriot Art: Ancient Glass. The Metropolitan Museum of Art, New York. Stern, E. M. 1995. Roman Mold-blown Glass. The first through sixth centuries. The Toledo Museum of Art, Toledo (Rome: ‘ĽErma’ di Bretschneider). Stern, E. M. 2001. Roman, Byzantine, and Early Medieval glass, 10 BC–700 CE, Ernesto Wolf Collection. Ostfildern. Štefanac, B. 2009. Pticolike kapaljke s područja južne Liburnije. Asseria 7: 115–147. Zadar. Štefanac, B. 2011. Stakleni vretenasti recipijenti sa šireg zadarskog područja, in: G. Lipovac Vrkljan, I. Radić Rossi and B. Šiljeg (eds), Rimske keramičarske i staklarske radionice. Proizvodnja i trgovina na jadranskom prostoru, Zbornik I. međunarodnog arheološkog kolokvija, Crikvenica, 23.–24. listopada 2008: 377–388. Zagreb/ Crikvenica: Institut za arheologiju, Muzej Grada Crikvenice.

The third stage of production refers to shaping certain details of the vessel, primarily handle and rim. Rim is created first by using a pointed wooden stick (Figure 7: 4), then handle is made. Handle modeling is the most complex part of the process. The procedure starts with applying viscous mixture on the juglet body, and stretching it to the rim. Handle decoration is executed by tweezers used for pinching glass to achieve the desired ribbed ornament (Figure 7: 5). In this case as well, slow and continuous cooling follows after the procedure. Conclusion In the experiments in glass workshop of the Museum of Ancient Glass in Zadar we applied several methods aimed at sheding more light on technological processes taking place in ancient workshops (1st – 4th century AD). Tests based on objects of simple morphology (spindle-shaped bottles, bird-shaped dropper flasks, 372

List of Contributors Timka Alihodžić Archaeological Museum in Zadar Trg opatice Čike 1 HR-23000 Zadar [email protected]

Enrico Cirelli University of Bologna – Department of History Cultures Civilization Piazza S. Giovanni in Monte, 2 IT-40124 Bologna [email protected]

Iris Bekljanov Zidanšek Arhej d.o.o. Drožanjska 23 SLO-8290 Sevnica [email protected]

Carla Corti Landesamt für Denkmalpflege Baden-Württemberg, Berliner Straße 12 DE-73728 Esslingen am Neckar [email protected]

Gaetano Benčić Zavičajni muzej Poreštine Dekumanus 9 HR-52440 Poreč [email protected]

Ilija Danković Institute of Archaeology, Belgrade Knez Mihailova 35/IV RS-11000 Belgrade [email protected]

Martina Blečić Kavur University of Primorska, Faculty of humanities Titov trg 5 SI-Koper 6000 [email protected] [email protected]

Armand Desbat 59, rue Benoist-Mary FR-69005 LYON [email protected] Patrizia Donat Via dei Piccardi 1 IT-34141 Trieste [email protected]

Igor Borzić University of Zadar, Archaeology Department Obala kralja Petra Krešimira IV/2 HR – 23000 Zadar [email protected] [email protected]

Anamarija Eterović Borzić Museum of Ancient Glass in Zadar Poljana Zemaljskog odbora 1 HR-23 000 Zadar [email protected]

Emmanuel Botte Aix-Marseille Université, CNRS, CCJ 5 rue du Château de l’Horloge CS 90412 FR-13097Aix-en-Provence Cedex 2 [email protected]

Maja Grisonic University of Zadar/Department of Archaeology Obala kralja Petra Krešimira IV / 2 HR- 23000 Zadar [email protected]

Bridget Buxton University of Rhode Island USA-Kingston, RI 02881 [email protected]

Mato Ilkić University of Zadar, Archaeology Department Obala kralja Petra Krešimira IV/2 HR – 23000 Zadar [email protected]

Tania Chinni University of Bologna - Department of Cultural Heritage via degli Ariani 1 IT-48121 Ravenna [email protected]

Giuseppe Indino

Université de Pau et des Pays de l’Adour / Università Ca’ Foscari Venezia Laboratoire IRAA - CNRS (UAR 3155) France Viale Verona 102 IT - 36100 Vicenza [email protected]

Andrea Cipolato DSU, Università Ca’ Foscari Venezia Giudecca 410a IT- 30100 Venezia [email protected]

373

List of Contributors Ivana Jadrić-Kučan Department of Archaeology University of Zadar Obala Kralja Petra Krešimira IV / 2 HR-2300 Zadar [email protected], [email protected]

Valentina Mantovani Via Cadorna 21 IT-36051 Creazzo (VI) [email protected] Cristina Mondin Museo civico di Asolo Via Regina Cornaro, 74 IT-31011 Asolo (TV) [email protected]

Kristina Jelinčić Vučković Institute of Archaeology Jurjevska ulica 15 HR-10000 Zagreb [email protected]

Davor Munda Zavičajni muzej Poreštine Dekumanus 9 HR-52440 Poreč [email protected]

Ljubomir Jevtović Institute of Archaeology, Belgrade Knez Mihailova 35/IV RS-11000 Belgrade [email protected]

Ivana Ožanić Roguljić Institute of Archaeology Jurjevska ulica 15 HR-10000 Zagreb [email protected]

Ivanka Kamenjarin Museum of the Town of Kaštela Lušiško Brce 5 HR-21215 Kaštel Lukšić [email protected]

Mate Parica University of Zadar, Archaeology Department Obala kralja Petra Krešimira IV/2 HR – 23000 Zadar [email protected]

Ana Konestra Institute of Archaeology Jurjevska ulica 15 HR-10000 Zagreb [email protected]

Monika Petrović Archaeological Museum of Istria Carrarina 3 HR-52100 Pula [email protected]

Irena Lazar Faculty of Humanities University of Primorska Titov trg 5 SLO- 6000 Koper [email protected]

Irena Radić Rossi University of Zadar Obala kralja Petra Krešimira IV / 2 HR-23000 Zadar [email protected]

Goranka Lipovac Vrkljan Institute of Archaeology Jurjevska ulica 15 HR-10000 Zagreb [email protected]

Tea Rosić Crikvenica Town Museum Preradovićeva 1 HR – 51260 Crikvenica [email protected]

Paola Maggi Indipendent researcher Vicolo del Castagneto 24 IT-34137 Trieste [email protected]

Corinne Rousse Aix Marseille Université, CNRS, Centre Camille Jullian Maison Méditerranéenne des Sciences de l’Homme 5 Rue du Château de l’Horloge, CS 90412 FR-13097 Aix-en-Provence [email protected]

Émilie Mannocci Université Aix-Marseille / Università degli studi di Padova / École française de Rome 5 rue du Château de l’Horloge FR-13090 Aix-en-Provence [email protected]

Eleni Schindler Kaudelka Mariatrosterstraße 113 A-8043 GRAZ [email protected]

374

List of Contributors

Ranko Starac Maritime and History Museum of the Croatia Littoral Trg Riccarda Zanelle 1 HR-10000 Rijeka [email protected]

Christophe Vaschalde LabEx Archimède, Archéologie des Sociétés Méditerranéennes UMR 5140, Université Paul Valéry, route de Mende FR-34199 Montpellier Cedex Institut des Sciences de l’Évolution de Montpellier, Equipe Dynamique de la Biodiversité, Anthropo-Ecologie, Université de Montpellier Place Eugène Bataillon, CC 065 FR-34095 Montpellier Cedex 5 [email protected]

Nikolina Stepan University of Zadar/Department of Archaeology Obala kralja Petra Krešimira IV., 2 HR- 23000 Zadar [email protected] Roby Stuani Corso G. Garibali 74, IT-25017 Lonato del Garda (BS) [email protected]

Paola Ventura Soprintendenza Archeologia, belle arti e paesaggio del Friuli Venezia Giulia Piazza Libertà 7 IT-34135 Trieste [email protected]

Bartul Šiljeg, Institute of Archaeology Jurjevska ulica 15 HR-10000 Zagreb [email protected]

Petra Vojaković Arhej d.o.o. Drožanjska 23 SLO-8290 Sevnica [email protected]

Berislav Štefanac Museum of Ancient Glass in Zadar Poljana Zemaljskog odbora 1 HR-23000 Zadar [email protected]

Fabian Welc Institute of Archaeology, Cardinal Stefan Wyszynski University in Warsaw Wóycickiego 1/3 (23) PL-01-938 Warsaw [email protected]

Marko Štefanac Museum of Ancient Glass in Zadar Poljana Zemaljskog odbora 1 HR-23000 Zadar [email protected]

Tina Žerjal Arhej d.o.o. Drožanjska 23 SLO-8290 Sevnica [email protected]

Kristina Turkalj The Institute of Archaeology Jurjevska ulica 15 HR-10000 Zagreb [email protected] Marina Ugarković Institute of Archaeology Jurjevska ulica 15 HR-10000 Zagreb [email protected]

375

Archaeopress Roman Archaeology 94

Roman Ceramic and Glass Manufactures: Production and trade in the Adriatic region and beyond presents thirty-one papers read at the 4th International Archaeological Colloquium held in Crikvenica, Croatia, 8–9 November 2017. The papers deal with issues of pottery production in relation to landscape and communication features, ceramic building materials, as well as general studies on ceramic production, pottery and glass finds. Additionally, an invited contribution explores finds relating to clothing from the Roman pottery workshop at Crikvenica. Several papers are devoted to restoration and archaeological experimentation. Although the majority of papers tackle research conducted in the wider Adriatic area, several contributions deal with other provinces of the Roman world.

Goranka Lipovac Vrkljan graduated history and archaeology at the Faculty of Philosophy of the University of Zagreb. In 2001 Goranka obtained her PhD at the Department of Archeology of the same faculty, and has been working at the Institute of Archaeology in Zagreb since 1990. She led several scientific and research projects, including the Croatian Science Foundation project Roman Economy in Dalmatia: Production, distribution and demand in the light of pottery workshops - RED (IP-11-2013-3973), which resulted in the volume Pottery Production, Landscape and Economy of Roman Dalmatia. Interdisciplinary approaches (2018, Archaeopress Roman Archaeology 47). Together with her project collaborators, she organised four international archaeological colloquia, Roman pottery and glass workshops: production and trade in the Adriatic region, which were held in Crikvenica (2008, 2011, 2014, 2017). Her research interests are focused on topics related to pottery workshops, their products and distribution, and on landscape data. Ana Konestra graduated at the University of Udine (Italy), and obtained her PhD at the University of Zadar in 2016 with a dissertation on Italic fineware imports to the eastern Adriatic. Alongside collaborating on several scientific projects regarding pottery production, Hellenistic and Roman Dalmatia, as well as landscape archaeology, Ana co-directs survey and excavation projects on the Island of Rab, at Tar in Istria and at Danilo in central Dalmatia. Her research interests include aspects of material culture and Roman rural landscape organisation on the eastern Adriatic. She co-organised several international conferences, workshops and a summer school, co-edited several proceedings and edited Pottery Production, Landscape and Economy of Roman Dalmatia. Interdisciplinary approaches published by Archaeopress in 2018. Anamarija Eterović Borzić graduated archaeology and Latin language and Roman literature at the University of Zadar in 2009 and is currently a PhD candidate at the Department of Archaeology, Zadar University. Anamarija works as a senior curator and head of Department of Archaeology at the Museum of Ancient Glass in Zadar. She is author and co-author of a number of scholarly papers and books. She is also the head of research carried out on the archaeological site of Asseria near Benkovac.

Archaeopress Archaeology www.archaeopress.com