Wonders Lost and Found: A Celebration of the Archaeological Work of Professor Michael Vickers 9781789693812, 9781789693829

Citation preview

Wonders Lost and Found A celebration of the archaeological work of Professor Michael Vickers

edited by

Nicholas Sekunda

Wonders Lost and Found A celebration of the archaeological work of Professor Michael Vickers edited by

Nicholas Sekunda

Archaeopress Archaeology

Archaeopress Publishing Ltd Summertown Pavilion

18-24 Middle Way Summertown Oxford OX2 7LG

www.archaeopress.com

ISBN 978-1-78969-381-2 ISBN 978-1-78969-382-9 (e-Pdf)

© Authors and Archaeopress 2020 Cover illustrations:

AN1885.491 ‘Scaraboid seal depicting Achemenid winged horned lion’ © Ashmolean Museum, University of Oxford.

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. Printed in England by Severn, Gloucester This book is available direct from Archaeopress or from our website www.archaeopress.com

Contents Preface�������������������������������������������������������������������������������������������������������������������������������������������������������������iii Early Cycladic? Lead model boats in the Ashmolean Museum, Oxford������������������������������������������������������������ 1 Susan Sherratt Two Cushions, a Bes, a boar and a bead. New ‘discoveries’ in the Aegean collection at the Ashmolean�������� 8 Helen Hughes-Brock Ancient Colchis and the origins of iron: interim results from recent field survey work in Guria, Western Georgia�����������������������������������������������������������������������������������������������������������������������������������������������18 Brian Gilmour, Marc Cox, Nathaniel Erb-Satullo, Nana Khakhutaishvili and Mark Pollard The structure and function of ancient metrology�������������������������������������������������������������������������������������������33 John Neal The second stage of the Grakliani Culture������������������������������������������������������������������������������������������������������48 Vakhtang Licheli Owl skyphoi around the Adriatic���������������������������������������������������������������������������������������������������������������������58 Branko Kirigin Gyenus on stage: civic foundation and the comedy of Aristophanes’ Birds������������������������������������������������������������ 70 David Braund New archaeological finds at Pichvnari (November-December 2010)��������������������������������������������������������������82 Amiran Kakhidze A double-sided glass relief pinhead from ancient Colchis – the Pichvnari ‘Heracles’ �����������������������������������90 Sujatha Chandrasekaran Gold jewellery from Kavtiskhevi���������������������������������������������������������������������������������������������������������������������98 Darejan Kacharava Palynological analysis of organic materials from Pichvnari (including the earliest silk in Georgia)����������102 Eliso Kvavadze and Maia Chichinadze Mercurial metrics������������������������������������������������������������������������������������������������������������������������������������������108 Kenneth Lapatin The Erechtheion glass gems: classical innovation or Roman addition?�������������������������������������������������������117 Despina Ignatiadou Carp from the Danube delta? Notes on an unusual gold-glass in the Wilshere Collection���������������������������125 Susan Walker Mediterranean drinking habits in Roman Britain: celery-flavoured wine prepared in an Iron Age bronze strainer����������������������������������������������������������������������������������������������������������������������������������������������������������134 Eberhard W. Sauer, Mark Robinson and Graham Morgan

i

From an offshore island: classical art and the Britons in Late Antiquity����������������������������������������������������146 Martin Henig The siege-drill (trypanon): new archaeological evidence from Georgia �������������������������������������������������������155 Nicholas Sekunda An emphatic statement: the Undley-A gold bracteate and its message in fifth-century East Anglia����������160 Daphne Nash Briggs The Levant Company and British collecting��������������������������������������������������������������������������������������������������185 Arthur MacGregor Cryptography and vasology: J.D. Beazley and Winifred Lamb in Room 40���������������������������������������������������194 David W.J. Gill Dictators and Antiquity���������������������������������������������������������������������������������������������������������������������������������199 Clive Foss

ii

Preface This volume originated in a day of papers organised to celebrate the archaeological work of Professor Michael Vickers, Professor Emeritus of Archaeology at the University of Oxford, an Emeritus Fellow of Jesus College, and former Curator of Greek and Roman Antiquities at the Ashmolean Museum and Senior Research Fellow in Classical Studies at Jesus College. Michael Vickers now holds the post of Dean of Degrees of Jesus College, Oxford. He formerly taught at University College, Dublin, the University of Libya in Benghazi, the University of Texas at Austin, and the University of Colorado at Boulder. He has been a Visiting Member of the Institute for Advanced Study at Princeton and is a Corresponding Member of the German Archaeological Institute and of the Archaeological Institute of America. He was co-director of the joint British-Georgian Pichvnari Expedition 1998-2010. The commemorative day was held in the Habbakuk Conference Room, Jesus College, University of Oxford on Wednesday 18 May 2011. The title of the book, somewhat unimaginatively, repeats the title of that meeting. Sometime after the meeting I, somewhat unwisely, enquired whether there were moves afoot to publish the papers given there. On being told that there were not I, somewhat rashly, offered to take on the task. Since that time many things have changed. Over the last decade the amount of senseless administrative tasks required of academics by successive Polish governments has risen exponentially, making serious inroads into the time which I had available to work on the volume. Also the original intention was to publish this volume in the monograph series ‘Akanthina’, produced by the Department of Mediterranean Archaeology of Gdańsk University. Changes in policy at University and Faculty made this no longer possible. Thus the volume languished for several years without a publisher, until Michael Vickers directly intervened and spoke to David Davison of Archaeopress, to whom I am very grateful for taking on the task of publication. I hope this will be sufficient explanation of why this book has taken such a long time to appear. I would also like to take this opportunity to thank Rajka Makjanić, also of Archaeopress, who has done such a competent job of preparing the work for printing. I look upon this book as very much a joint effort with Michael, not only for his finding a publisher. At the original meeting, constraints of time meant that only nine speakers were asked to deliver papers, and, for various reasons, some of these were not able to offer their papers for publication. It was necessary to co-opt further authors to the task in hand. It was with Michael’s assistance that a list of further potential contributors was drawn up. All expressed enthusiasm for the concept, but not all were able to find the time to contribute. The resulting volume consequently reflects Michael’s wide range of archaeological interests. It is fair to say that Michael has taken an active part in this book in all stages of its appearance. One of Michael’s friends who expressed great enthusiasm for the project from the outset, is Elspeth Dusinberre. By way of compensation, we have reproduced as the cover of this book, the object which she would have discussed, if time and health had permitted. It is an Achaemenid gem (Oxford 1885.491), a Chalcedony scaraboid (height: 2.7 cm, width: 2 cm), dating to the end of fifth, or the beginning of the fourth century BC. It comes from a female burial (Grave 5) from the Crimean city of Nymphaeum, which was presented to Oxford University by Sir William Siemens in 1880. It seems to be quite appropriate to use in a book dedicated to a person of such eclectic tastes. Of course Michael’s interests are not confined to archaeology alone, but, given the history of the genesis of this volume, the papers contained in it are. Nicholas Sekunda Gdańsk 29 January 2020

iii

iv

Early Cycladic? Lead model boats in the Ashmolean Museum, Oxford Susan Sherratt Introduction

artefacts, not so much because they fitted neatly into their already formed visions of the past as because they added objects of a unique nature to private or museum collections of antiquities and seemed (at least at the time of initial acquisition) plausibly genuine.2 The essay that follows here concerns just such a set of objects whose authenticity could perhaps only be questioned with hindsight.3 It is offered to Michael, a good friend and inspiring colleague, who long ago encouraged me never to take too much for granted.

Michael Vickers has a long and distinguished history of effectively overturning the established received wisdom, the idées fixes, of earlier generations of Classical archaeologists, sometimes in the face of sustained opposition and in a very difficult environment. He has beamed probing searchlights into interpretations of archaeological and literary material which were once regarded as settled beyond question, and the effect has been that many of his colleagues and younger archaeologists will never be able to accept these with quite the same confidence again. Among his many inspired acts in the late 1990s as Senior Assistant Keeper in charge of Greek antiquities in the Ashmolean (as he was then) was to send the ivory statuette of the ‘Minoan Boy God’, acquired by Sir Arthur Evans from a dealer in the early 1930s and published by him in 1935 in volume 4 of The Palace of Minos at Knossos (Evans 1935: 468-83, figs. 394, 396, suppl. pl. LIII; more recently Galanakis 2013: figs. 139-40 in colour) as a ‘young male divinity’, the youthful consort of the Minoan Mother Goddess in Evans’s well developed view of Minoan religion, for carbon-14 dating at the Oxford Research Laboratory for Archaeology and the History of Art.1 The result - that the ivory of the statuette was modern (Bronk Ramsey et al. 1999: 203), and its implications, and those for a number of other supposedly Minoan ivory statuettes which surfaced on the art market between 1914 and the 1930s, are recounted in Kenneth Lapatin’s delightfully readable book, Mysteries of the Snake Goddess (Lapatin 2002).

The lead boat models and their history There are in the Ashmolean three lead boat models, which are purportedly of Cycladic provenance and of Early Bronze Age date. Together with a fourth, now in the Merseyside County Museum, Liverpool, they are the only examples of their kind so far known from the Early Bronze Age Cyclades. The first, and most complete, of the models (Ashmolean Museum 1929.26) (Figure 1) arrived in the Ashmolean in 1929 as the gift, through Sir Arthur Evans, of Professor R.M. Dawkins, who had been Director of the British School at Athens between 1906 and 1913. The other two (Ashmolean Museum 1938.725-6) (Figures 2-3) were given to the Museum by Evans himself in 1938 along with much of what then remained of his own personal collection of prehistoric Aegean antiquities. The fourth model (Merseyside County Museum 55.66.180) was deposited in the Merseyside County Museum in the early 1930s, on loan from the collection of Robert Carr Bosanquet, who immediately preceded Dawkins as Director of the British School from 1900 to 1906.4

Evans was taken in by his ‘Boy God’ largely because he was already conditioned to believe in his existence, not only because of the other, now equally suspect, ivory figurines of the Goddess or her youthful consort which had already surfaced, but largely because they all fitted in so well into his ideas of the nature of Minoan religion; and there is little doubt that this statuette, and probably also its companions, were designed and manufactured by someone (or more than one person) who knew exactly what Evans expected and hoped to find. However, there are reasons to think that he, and others, were also on occasion taken in by other

There is very little explicit information about the provenance and circumstances of finding of these models, which were barely mentioned by anyone 2  See, for example, Kevin Butcher’s and David Gill’s account of the history of the ‘Fitzwilliam Goddess’ (Butcher and Gill 1993). The Fitzwilliam statuette, acquired by the Fitzwilliam Museum in 1926, was in fact endorsed by Evans. 3  It was originally presented as a paper at the joint meetings of the American Institute of Archaeology and American Philological Association in New Orleans in January 2003, in a session on forgeries organised by Kenneth Lapatin, to whom Michael originally introduced me. 4  For the Ashmolean models, see also Renfrew 1991: 50, pl. 14, and more recently Sherratt 2000: nos. 5.2-5.4, figs. 55-57, pls. 69-71, col. pl. 3; Badisches Landesmuseum 2011: 306 no. 106; Galanakis 2013: fig. 146. For the model in Liverpool, see Mee and Doole 1993: 48 no. 490, pl. 26.

1  This was the second statuette of the ‘Boy God’ that Evans had acquired. The first, bought by him in the 1920s, ended up in the Seattle Art Museum (see Lapatin 2002: 98-9). The Ashmolean statuette was given to the Museum by Evans in 1938, along with most of the rest of his personal collection of Aegean antiquities.

1

Wonders Lost and Found

Figure 1. Ashmolean Museum 1929.26, gift of R.M. Dawkins.

Figure 2. Ashmolean Museum 1938.725, gift of A.J. Evans.

Figure 3. Ashmolean Museum 1938.726, gift of A.J. Evans.

Figure 4. Ashmolean Museum 1929.26. Drawing by Keith Bennett.

Figure 5. Ashmolean Museum 1938.725. Drawing by Keith Bennett.

All illustrations courtesy of the Department of Antiquities, Ashmolean Museum.

2

Susan Sherratt: Early Cycladic? Lead model boats in the Ashmolean Museum, Oxford

before their publication in 1967 by Colin Renfrew (1967: 5, 18, pls. 1:12, 3:12-14).5 In his 1909 publication, Scripta Minoa I, Evans referred in passing to lead boat models ‘from Amorgos’ (Evans 1909, 26), but thereafter never seems to have mentioned them again, despite the fact that two of them were in his possession probably from at least 1909 onwards. This is all the more surprising since in the second volume of The Palace of Minos Evans devoted a considerable amount of discussion to early Aegean boats, including those of the Cyclades (Evans 1928: 239-42).

which has been slit into three narrow strips for about one-third of its length in much the same way as a cobbler would cut the leather for a traditional Turkish slipper, two for the sides and one for the bottom of the vessel. The prows of the models are formed by clamping the two outer strips together with the middle one, producing a section which looks rather like an inverted clothes peg. The result of this is a narrow prow which rises fairly gently upwards with a keel-like ridge on its underside. The prow is topped by an upward projection of the ‘keel’ in the form of a rod which appears in all cases to have broken off either at or just above the tip of the prow. The stern on Dawkins’s model, which shows no sign of damage, takes the form of a flat rectangular structure, formed by the continuation of the bottom strip which is bent upwards. There are gaps between it and the ends of the sides, covering the entire depths of the sides, with the result that it looks rather like the tailgate on a ferry. The stern section of one of the Evans models is missing entirely, and on the other Evans model and the Bosanquet model the stern sections are damaged. Nevertheless, enough of the stern remains on each of the latter two to make clear that the design was essentially the same as on the undamaged model. On the Dawkins and Bosanquet models, on which the stern is either undamaged or less damaged, the hull is canted upwards at this end.

What little we do know about the lead boat models is contained in a letter written by Dawkins to Evans shortly before he presented his boat model to the Ashmolean in 1929. Dawkins says that he acquired his model in Athens around 1907, and that it was said to have been found on Naxos along with three other broken examples and at least three marble figurines, two of which were then in his possession. Dawkins also presented these two figurines to the Ashmolean at the same time as his boat model, in 1929. They are a couple of folded arm figurines of Kapsala variety, probably of Early Cycladic II date.6 A Dawkins label already attached to one of the figurines when it arrived in the Museum records that the two figurines and the lead boat model together originally cost Dawkins 550 drachmae. In 1946, Dawkins presented a number of other marble figurines from his own collection to the Ashmolean. One of these, a Plastiras figurine of pre-folded arm type,7 carried a label stating that it was bought in November 1907 and ‘was probably found with the lead boats’. The label also recorded that it came from Drakatis, Naxos. Try as I might, I have been quite unable to find any mention of a site or place of this name on Naxos in either the 19th or 20th century literature or maps, or to find anyone who has ever heard of it amongst some of those most familiar with the history of archaeology on the island. This is not to say that a location with this name does not or did not exist on Naxos: but it does not appear to be a location which is known to have produced any other antiquities.

Which end is which on Early Cycladic longboat representations, and how many different types of longboats were there? This decidedly odd stern construction poses some problems, since it is clear that the gaps left on either side of the hull on Dawkins’s model are quite deliberate, and not the result of damage, with the edges of both sides of the hull and the ‘tailgate’ remaining sheer and vertical. It thus presents us with a boat design which seems disconcertingly precarious for use on the open sea, not least since the gaps go right down to just above the waterline, even making allowances for the slight upward slope of the hull towards the stern. In the light of this, it seems particularly unfortunate that it is the stern end that is either missing completely or badly damaged on all but one of the models, especially since all of the prows are, by comparison, well preserved and extremely confidently executed.

Their design The two Evans models (Figures 2-3) appear to have been broken deliberately across the middle of the hull. However, as far as their state of preservation allows us to tell, all four boat models appear to be of identical design and manufactured by the same method (Figures 4-5). They seem to be made from a single strip of lead

The design of the models makes it clear beyond any shadow of doubt which end is intended as the bow and which the stern, which is more than can be said for the two-dimensional representations of Early Cycladic longboats found on the Syros frying pans, first brought to light by C. Tsountas in 1899 (Tsountas 1899), and on the Korphi t’Aroniou plaques excavated in the 1960s by C. Doumas (Doumas 1965), about which opinions have

Renfrew’s statement (1967: 18) that 1929.26 was bought by Dawkins in 1917 seems quite simply to be either a mistake or a misprint. 6  Ashmolean Museum 1929.27-8. See Sherratt 2000: nos. 7.18, 7.19, figs. 86-87, pls. 159-164. 7  Ashmolean Museum 1946.118: Sherratt 2000: no. 7.12, pls. 140-142, col. pl. 8. 5 

3

Wonders Lost and Found varied over the years.8 Tsountas himself believed that the high end on these representations was the prow, as did both Bosanquet and Dawkins (1923: 7) as well as Evans (1928: 240) and others. Others, however, starting with R. Dussaud (1914: 415) and including S. Marinatos (1933: 182-5), were more inclined to regard the high end as the stern, possibly acting as an aerodynamic feature to keep the boat steady in a following wind. In 1967, when he published the lead models, Renfrew believed that they settled this question once and for all. There was no doubt in his mind that they represented the same type of boat as those represented on the frying pans, and quite clearly the high end was the prow (Renfrew 1967: 5; see also more recently Wachsmann 1998: 69-70, fig. 5.1; Matthäus 2011: 117).

in profile bears a very strong resemblance to several of the Syros frying pan representations. From this model, it is quite clear that the high projection is solid, with (as on the frying pan boats) a sharply defined angle at the base. It is also quite clear that the horizontal projection at the other end is indeed a solid spur which sits nicely on the waterline. Yiannis Vichos has also carried out research (including some experimental replication) on the design of the frying pan longboats, and has shown conclusively (to my mind) that the interpretation outlined by Basch must be correct: the high end is indeed the stern (Vichos 1991).11 Moreover, Basch has also gone a long way to explaining the canting upwards towards the stern seen on some (but not all) of the frying pan representations by showing that, if these are viewed from a slightly different angle on the circular background of the frying pans, not only do the boats acquire a plausible keel, but the horizontal projection sits squarely on the waterline (Basch 1987: 87-8).

Renfrew was misled, not at all surprisingly in my view, by the superficial resemblance of the lead boats to the frying pan representations and by his very reasonable assumption that the lead models did indeed represent Cycladic longboats: but in fact he almost certainly did not look at them closely enough. At any rate, the arguments about which end was which on the frying pan boats continued, particularly amongst naval architects and others with direct experience of ships and seafaring.

Although not clear to Renfrew, it has seemed abundantly clear to Basch and others that the lead models differ in certain very significant features from the Syros and Korphi t’Aroniou representations and from the Palaikastro model. In particular, on the lead models the high end, which is hollow, slopes more gently upwards and lacks the sharp angle at the base, is quite clearly the prow. As a result, some scholars, such as Basch and Michael Wedde, have concluded that the lead models represent quite a different sort of boat from those represented on the frying pans, on the Korphi t’Aroniou plaques and by the Palaikastro terracotta model. Both Wedde (1991: 88) and Basch (1987: 79-80) have suggested that they represent a unique and otherwise unknown category of vessel, while Basch has argued that the odd double-canted outline of their hulls would make them potentially unstable, and that they could only therefore have been used for relatively unadventurous activities such as inshore fishing or coastal cabotage.

In 1987, Lucien Basch published a particularly convincing discussion of the frying pan boats in which he argued that the high projection, which has a sharply defined angle of 90 degrees or less between it and the bottom of the hull, would not only serve no useful purpose as the prow but would actually obscure the view of the crew and make it much harder to keep the boat steady in either a following or a head wind, while the sharp angle at the base would make beaching difficult (Basch 1987: 85-6). On the other hand, as the stern, it would actually contribute to stabilising the boat, particularly in a head wind. At the same time, Basch also argued that the curious lateral projection from the other end of the frying pan boats, whose interpretation had (by his own admission) more or less baffled Tsountas (1899: 91),9 was a horizontal spur which projected on the waterline and which made excellent sense at the prow of a longboat where it would facilitate cleavage through the waves, thus protecting the front of the hull from their full force and minimising the risk of frontal overwash. It would also have the effect of helping to hold the craft on course in a choppy sea. Basch was greatly aided in this interpretation by a terracotta boat model from Palaikastro, first published relatively unobtrusively in 1904 (Dawkins and Currelly 1903-4: 197, fig.1:k),10 which

At this point, we ought to consider not only the material of which the lead models are made but also the context from which they purport to come. Lead is intimately associated with the cupellation of silver, one of the salient characteristics of the Early Cycladic II period, published in 1904, it was done so very unobtrusively and in English. Moreover it does not seem to have affected the belief of Dawkins, Bosanquet, Evans etc. as to which end on either it or the Syros representations was the prow. In 1923 (in The Unpublished Objects from the Palaikastro Excavations, 1902-1906) Dawkins and Bosanquet argued (perhaps under the influence of the lead models, though these were not mentioned) that the high ends of both the Palaikastro model and the frying pan boats represented their prows (Dawkins and Bosanquet 1923: 7). The lateral projections at the other ends were seen as some sort of fixed rudder attached to the stern (cf. Evans 1928: 240). 11  See, too, the reconstruction by Thomas Guttandin in Badisches Landesmuseum 2011: 304-5 no. 105.

8  See now also the rock engravings of longboats from Strofilas on Andros, of slightly earlier, Final Neolithic, date (Televantou 2008: 4649, figs. 6.8, 6.10). 9  Tsountas suggested very tentatively that it might be some sort of steering contraption. 10  It should be noted that, although the Palaikastro model was first

4

Susan Sherratt: Early Cycladic? Lead model boats in the Ashmolean Museum, Oxford

and, like representations of longboats themselves, likely to symbolise one of the cornerstones of elite status and lifestyle. The fact that all four of the models are alleged to have been found in a single grave which also contained a number of marble figurines implies a relatively prestigious context, which in turn suggests a function whose prestige and symbolism was probably at least comparable to that of the longboats more usually portrayed on frying pans and marble plaques rather than something equivalent to a humble inshore dinghy. What the existence of these lead models, taken together with the material of which they are made and the context in which they are said to have been found, invites us to infer, in effect, is the co-existence of two quite different types of equally prestigious longboat, each designed to travel in the opposite direction to the other. Though this might not be completely impossible, when viewed in this way it does seem decidedly unlikely.

pan boats was a hollow prow, in which case the modified slope and angle in order to produce a more practical prow design becomes quite understandable; as does the provision at the same end of a strongly marked keel whose projection forms the broken-off stem on which we might envisage the attachment of a fish pennant like those seen on the frying pan longboats. It is the treatment of the stern, however, which perhaps more than anything else indicates that a rationalised interpretation of the two-dimensional frying pan representations is what we are actually witnessing. In place of the horizontal projection at the other end of these representations, which can have made little sense to anyone used to late 19th century craft (and which certainly baffled Tsountas), the model-maker has perhaps quite reasonably constructed the sort of raisable tail-gate with which some river ferries and shallow-draught livestock carriers were equipped by the late 19th century, which - when envisaged in a lowered position - provides an extremely good visual replica of this end of the frying pan longboats. However, that the model-maker himself may have felt understandably tentative about attaching this particular interpretation of the horizontal projections on the representations to sea-going craft is at least suggested by the curious coincidence that on three of the lead models it is this ‘tailgate’, rather than the prow, which is most consistently damaged or missing altogether.

A contextual explanation? It seems to me that there is a much more satisfactory explanation which can take account not only of the somewhat eccentric design of the unique lead models and in particular the curious stern construction which makes little sense in terms of vessels designed to cope with the open sea, but also of the timing of and decidedly murky circumstances surrounding their appearance in the first few years of the 20th century.12 The key to this lies in their superficial resemblance to the Syros frying pan representations which convinced Renfrew and others that they were indeed three-dimensional models of the longboats shown on the frying pans and the Korphi t’Aroniou plaques. This strange mixture of extremely plausible superficial similarity combined with what we can recognise (largely thanks to the Palaikastro terracotta model and to Basch’s expert analysis) to be radical differences in structural design can perhaps best be explained if we regard the lead models as the result of a rationalised interpretation of the frying pan representations (published in 1899), executed on the basis of Tsountas’ account of these by someone who knew enough about boats in general to attempt a plausible three-dimensional version.

The suspicions which arise from the very curious (not to say improbable) design of the four lead boat models, and their superficial similarity to the Syros frying pan representations, are compounded by the timing of their appearance just a few years after Tsountas’ publication of the latter, by the murk which surrounds their find circumstances (not least that all four appear to have come from the same dealer who seems to have led Dawkins at least to believe that they were found together), and by the fact that no other models or representations of boats of the same type have ever been found before or since. To this we can add the striking silence about their existence on the part of Dawkins, Bosanquet and Evans, despite the fact that the latter devoted several pages of one of his Palace of Minos volumes to early Aegean boats, including the Syros frying pan representations. This suggests that they, too, may well have had serious doubts about their authenticity. Evans’s single mention in 1909 of lead boats ‘from Amorgos’ (at a time when two of the boats may already have been in his possession) is curious, since almost certainly it refers to the same boats.13 Perhaps this is what he was told by the dealer, or perhaps he was merely giving them a plausible provenance. Amorgos, from the late 19th century, was an island from which

Whoever constructed these models seems to have assumed (like Tsountas) that the high end on the frying 12  The lead from which the models were made was originally thought, as a result both of lead isotope analysis and chemical composition, to be of Siphnian origin (Gale and Stos-Gale 1981: 213, fig. 13, table 10; cf. Sherratt 2000: 104 with nn.12-13). Subsequently, however, this was changed to become ‘isotopically consistent with Pb/Ag ore from Gümüşköy in [north-west] Anatolia’ (Gale and Stos-Gale 2008: 388, 402, fig. 37.7). All that this demonstrates to me is the inadvisability of using lead isotope analysis for positive identifications of provenance on the basis of databases that are inevitably less than comprehensive. In any case, lead continued to be extracted both on Siphnos and at Gümüşköy until well into the twentieth century (Sherratt 2000: 106, with further references; Kaptan 1981-1982).

A note attached to the register entries of 1938.725-6 refers to relevant correspondence in the archive - presumably Dawkins’s letter concerning the supposed association of all four models.

13 

5

Wonders Lost and Found collectors expected prehistoric Cycladic antiquities to come, and the centripetal pull of this expectation seems to have exerted an effect on several objects now in the Ashmolean register whose Amorgan provenance may be seriously doubted.

the long run entirely successful, one could even regard it as legitimate and worthwhile. Indeed, it was only at the point at which these models were sold as genuine antiquities (something which perhaps their maker might just conceivably never actually have intended) that any potentially lasting damage was done.

Epilogue

The third and final point that arises from a consideration of the lead boat models is that modern scholars should perhaps beware of seizing too enthusiastically on ‘discoveries’ of objects found languishing apparently ignored in museum collections, particularly when those of an earlier generation, who are known to have been fully aware of their existence, can be seen to have been unexpectedly reticent about them.14 This is especially so in cases where such objects are recruited to the cause of ‘solving’ major interpretational problems - and in such circumstances one can only recommend that their ‘discoverers’ examine them and what is known of their history, and the timing and context in which they first surfaced, with the greatest care.

What conclusions (or, if you like, what moral) can be drawn from all this? The first is the obvious general one: that objects bought through antiquities dealers, the provenances and contexts of which (if any) are vague and uncertain, can at best be of very limited use and at worst (as probably in this case) downright misleading. Although there is perhaps little point now in criticising the activities of late 19th and early 20th century collectors (who saw the collecting and classifying of archaeological no less than geological or zoological specimens as a primarily scientific activity), this is as undeniably true of artefacts collected a century ago as it is of those which surface in the hands of dealers today. As for the particular problem of forgery, there is little comfort to be found in the idea that an object safely deposited in a museum a hundred years or more ago can automatically be regarded as above suspicion. One need only glance at the contents of a lecture delivered by John Evans (Arthur Evans’ father) at the Royal Institution in 1865 (Evans 1893), in which he bewails the quantity and variety of forged antiquities of all types already washing around Europe, to realise that this is not the case; while, as far as the prehistoric Cyclades are concerned, objects such as obsidian tools, ceramic ‘kernoi’ and marble figurines (often illicitly rifled from graves) had been assiduously collected from the time of the Greek War of Independence and were particularly prized from the 1880s onwards. The price of 550 drachmae paid by Dawkins for the lead boat model and two figurines, probably something like 30 times the daily wage of an archaeological foreman, alone demonstrates the incentive to provide the kinds of antiquities that might be thought attractive to collectors.

Looking a little more closely at the models reveals that, despite their beguiling superficial resemblance to the frying pan representations, there are some significant differences - the noticeably gentler rise of the prows, for example, and the much more obtuse angle seen at the bases of these. One also has to wonder about the plausibility of the curious stern construction which apparently leaves deliberate gaps down the entire length of the hull (of which, incidentally, Renfrew made no mention in 1967). Either one has to conclude, with Basch and Wedde, that the models represent a quite different type of boat from those shown on the frying pans, the Korphi t’Aroniou plaques and by the Palaikastro terracotta, or that there is something decidedly odd about them. As it is, the timing of their collective (and unique) appearance in the hands of an Athens dealer in 1907, and their congruence with Tsountas’ interpretative description, published just a few years earlier, of the boat representations incised on the Syros frying pans seem to me to offer good reasons to doubt both their authenticity and their ability to supply an independent solution to the problem of which end was which on the frying pan boats.

In addition, it is worth pointing out that, whereas nowadays the best strategy for a forger is to produce objects which fit comfortably within known typological categories, sixty to a hundred years ago the premium put by collectors and museums on unusual or unique objects - ones which would stand out within any given class or general type of object - meant that the manufacture of antiquities provided scope for a greater element of original but informed creation. In this context, the creation of the lead boat models as an exercise in turning Tsountas’s frying pan representations into unique three-dimensional versions is particularly fascinating, since it fulfils both of these criteria at once. As an exercise in itself, though not in

Bibliography Badisches Landesmuseum 2011. Kykladen. Lebenswelten einer frühgriechischen Kultur. Karlsruhe: Badisches Landesmuseum. 14  Cf. Renfrew (1967: 5) who expresses surprise than nobody has mentioned them, and wonders if ‘the rather unexpected material of which they are made has led to doubts about their authenticity’. In his letter to Evans Dawkins includes the remark ‘I think that there is no doubt at all of the genuineness of these things’, which suggests that both he and Evans may indeed have harboured some doubts.

6

Susan Sherratt: Early Cycladic? Lead model boats in the Ashmolean Museum, Oxford

Basch, L. 1987. Le musée imaginaire de la marine antique. Athens: Institut Hellénique pour la Préservation de la Tradition nautique. Bosanquet, R.C. and R.M. Dawkins 1923. The Unpublished Objects from the Palaikastro Excavations, 1902-1906 (British School at Athens Supplementary Paper 1). London. Brodie, N., J. Doole, G. Gavalas and C. Renfrew (eds) 2008. Horizon. A Colloquium on the Prehistory of the Cyclades. Cambridge: McDonald Institute for Archaeological Research. Bronk Ramsey, C., P.B. Pettitt, R.E.M. Hedges and G.W.L. Hodgins 1999. Radiocarbon dates from the Oxford AMS system. Archaeometry datelist 27. Archaeometry 41,1: 197-206. Butcher, K. and D. Gill 1993. The director, the dealer, the goddess and her champions: the acquisition of the Fitzwilliam Goddess. American Journal of Archaeology 97: 383-401. Dawkins R.M. and C.T. Currelly 1903-4. Excavations at Palaikastro, III. Annual of the British School at Athens 10: 192-231. Doumas, C. 1965. Κορφὴ τ’᾽Αρωνιοῦ. Αρχαιολογικόν Δελτίον 20: 41-64. Dussaud, R. 1914. Les civilisations préhelléniques dans le bassin de la mer Egée, 2nd edn. Paris. Evans, A.J. 1909. Scripta Minoa I. Oxford. Evans, A.J. 1893. The Forgery of Antiquities [Reprint of a lecture delivered at the Royal Institution, 24 February 1865]. London. Evans, A.J. 1928. The Palace of Minos at Knossos 2. London. Evans, A.J. 1935 .The Palace of Minos at Knossos 4. London. Galanakis, Y. (ed.) 2013. The Aegean World. A Guide to the Cycladic, Minoan and Mycenaean Antiquities in the Ashmolean Museum. Oxford and Athens. Gale, N.H. and Z.A. Stos-Gale 1981. Cycladic lead and silver metallurgy. Annual of the British School at Athens 76: 169-224. Gale, N.H. and Z.A. Stos-Gale 2008. Changing patterns in prehistoric Cycladic metallurgy, in Brodie et al.: 387-408. Kaptan, E. 1981-1982. New discoveries in the mining history of Turkey in the neighborhood of Gümüşköy, Kütahya. Bulletin of the Mineral Research and Exploration Institute of Turkey 97-98: 60-67. Lapatin, K. 2002. Mysteries of the Snake Goddess. Art, Desire, and the Forging of History. Boston. Marinatos, S. 1933. La marine créto-mycénienne. Bulletin de Correspondance Hellénique 57: 170-235. Matthäus, H. 2011 Kostbares Gut: Geräte und Gefässe aus Metall, in Kykladen. Lebenswelten einer frühgriechischen Kultur: 108-117. Karlsruhe: Badisches Landesmuseum. Mee, C. and J. Doole 1993. Aegean Antiquities on Merseyside: The Collections of Liverpool Museum and Liverpool University. Liverpool.

Renfrew, A.C. 1967. Cycladic metallurgy and the Aegean Early Bronze Age American Journal of Archaeology 71: 1-20. Renfrew, C. 1991. The Cycladic Spirit. London. Sherratt, S. 2000. Catalogue of Cycladic Antiquities in the Ashmolean Museum: The Captive Spirit. Oxford. Televantou, C.A. 2008. Strofilas: a Neolithic settlement on Andros, in Brodie et al.: 43-53. Tsountas, C. 1899. Κυκλαδικά ΙΙ. Αρχαιολογική Εφημερίς 38: 73-134. Vichos, Y. 1991. L’extrémité haute des navires antiques: la poupe ou la proue? Une approche nautique au problème de l’identification des extrémités des navires antiques, in H. Tzalas (ed.) Tropis II. 2nd International Symposium on Ship Construction in Antiquity, Delphi 1987: 363-70. Athens: Hellenic Institute for the Preservation of Nautical Tradition. Wachsmann, S. 1998. Seagoing Ships and Seamanship in the Bronze Age Levant. College Station, Texas. Wedde, M. 1991. Aegean Bronze Age ship imagery; regionalisms, a Minoan bias, and a ‘thalassocracy’, in R. Laffineur and L. Basch (eds), Thalassa: l’Egée préhistorique et la mer (Aegaeum 7, Université de Liège, Histoire de l’art et archéologie de la Grèce antique, Liège): 73-95.

7

Two Cushions, a Bes, a boar and a bead New ‘discoveries’ in the Aegean collection at the Ashmolean Helen Hughes-Brock The subjects of this small offering to Michael are several small objects which lived in what was for 39 years Michael’s kingdom and which perhaps he handled a time or two early in that period. The word ‘discoveries’ in the title is not entirely truthful. In fact, what follows is the tardy correction of some old oversights - oversights by me and others, not by Michael!1

4 lot 13), registered as AN1952.107, diameter 1.65cm, height 1.3cm. Cook himself illustrated it in his monumental Zeus and actually recognised it as Minoan. He dated it to Middle Minoan II but saw the motif as a Gorgoneion, ‘the earliest Gorgon’s head known to me’ (Cook 1940: part 2, 845, fig. 659). Clark Hopkins (1961: 32) in a later discussion of the Gorgon was not entirely convinced, since, though ‘rather unpleasant in expression’, it was not the demon face with prominent teeth and protruding tongue which characterises the Greek Gorgon. V.E.G. Kenna, meanwhile, had excluded it from his catalogue of the Ashmolean seals as not Minoan on the grounds of ‘unusual’ size, style and motif (Kenna 1960: 154, pl. 20).

The Ashmolean volume of the Corpus der Minoischen und Mykenischen Siegel (CMS 6, published in 2009) is one of the largest in that series, containing entries for 516 engraved seals and rings as well as brief details of related items such as unengraved ‘seals manqués’, possible seals too worn to be sure about, genuine but foreign or post-Bronze Age seals and pieces condemned as fakes (CMS 6: pp. 26–29). The volume had a very long gestation. It could not have come into the world at all without Michael. For years he was pestered by unending requests to get this or that seal out of its display case, or, having put it back, to get it out yet again - always a fiddly job with seals, and made the more fiddly by the little individual boards with pin attachments which were used in the Arthur Evans Room. To say that Michael was always positively cheerful about it would be going too far, but good-natured he always was, and helpfulness incarnate throughout.

Unfortunately the preoccupation with Gorgons coupled with Kenna’s rejection deflected the attention of the CMS team from Cook’s correct Minoan dating. Once seen in the ‘rejects’ drawer, it was not looked at again. It was assumed that the piece was Archaic, for indeed it does have features in common with Archaic Gorgoneia (as e.g. Boardman 1968: 36 no. 68 = Boardman 1970/2001: 180, pl. 289). The shape, however, is good Minoan, the loop signet with perforation at the top of the handle which in Minoan seal literature often goes by its German name, Petschaft. Petschafte of the soft stones steatite and chlorite are dated to Middle Minoan I-II (Yule 1981: 85–6). Kenna was more familiar, however, with the more elaborate versions of the shape which are found commonly on the hard-stone examples. The size, pace Kenna, is hardly unusual, though on the big side for this shape (diameter 1.65cm, the general range being from 1cm to 1.5cm). The material is entirely unexceptional, a rather shiny black Cretan steatite in fashion at this period (its source not yet pinned down). What Kenna had in mind about the ‘unusual style’ we do not know, but whereas many loop signets were made of the new hard semi-precious stones and thus finely engraved with the new rotary tools (drills and cutting wheels), soft-stone seals like ours, which were engraved mainly with hand-held tools, do have a rather different appearance and Kenna in the 1950s was not used to seeing them in this shape.2 That leaves the motif - and so we come to our Bes.

At the earliest preliminary stage of work on CMS 6 a quick look was taken at a drawerful of anomalous and dubious-looking pieces. Some of these on later re-examination were ‘rehabilitated’ (Hughes-Brock 2000a). Several others, however, had been dismissed earlier, put aside and never looked at again. Alas! - for thus it was that after all the work on CMS 6 two fine Minoan seals were left out of it. The first is a loop or stalk signet said to be from Crete (Figure 1), which had been in the collection of A.B. Cook and after Cook’s death was bought by the Ashmolean from Sotheby’s in 1952 (catalogue 15 January 1952, p. Acknowledgements: Olga Krzyszkowska and Judith Weingarten made their contribution in Michael’s honour by taking the photographs (respectively Figures 1 and 2 and Figure 3) and Yannis Galanakis by scanning Figures 4 and 5, all reproduced by courtesy of the Ashmolean Museum, Oxford University. They and other friends deserve my warmest thanks for fruitful and entertaining conversations, for information and for sending forthcoming or published papers: Lucia Alberti, Robert Arnott, Lisa Bendall, John Boardman, Paul Collins (Ashmolean Museum), Oliver Dickinson, Sybille Haynes, Leonora Ives (aged nine) and Jennifer Moody.

1 

His first catalogue was Kenna 1960, in which most loop signets were high-quality hard-stone seals from Arthur Evans’s collection; the one steatite example is broken and he did not discern its shape (CMS 6 nos. 124–137 passim). His four CMS volumes, composed in the 1960s and 1970s, contain a greater number of soft-stone pieces. On the tools see Krzyszkowska 2005: 81–85.

2  

8

Helen Hughes-Brock: Two Cushions, a Bes, a boar and a bead

Figure 1 a and b Loop or stalk signet AN1952.107. Black steatite. Possibly Bes

It was Judith Weingarten who first suggested that the long-lived popular Egyptian demon Bes might have inspired motifs on some Minoan seals, in that instance some of the bizarre motifs on LM I seal impressions found at the palace of Zakro (Weingarten 1983: 101–3). Later, in a conversation with me, she wondered whether the crudely carved face on a Middle Minoan II steatite prism might not be an early Bes, a suggestion I thought worth mentioning re CMS 6 no. 71b. That face has fat bulging cheeks, bristling upright hair, round ears (like a bear’s) and round open mouth; the eyes are hard to discern. It might alternatively look like a boar’s head with the strokes at the sides seen as the tusks. Either way it is unusual. On three-sided prism steatite seals of this period boars are not unusual but they are depicted whole in profile.3 The face on our Petschaft has very fat bulging cheeks, a big nose, eyes apparently open, hair on the forehead and spiky hair or bristles above; its ears are not identical and neither may be complete; the mouth, also perhaps incomplete, looks open. It has some features in common with contemporary pieces, e.g. two four-sided prisms CMS 3 nos. 237 and 238, both carefully engraved hard stones (agate, quartz).

remarkable of all, with a pretty lattice pattern on one side, and the other side depicting a unique and detailed frontal figure with bulging cheeks, staring eyes, round ears and toothy mouth (Krzyszkowska 2012a: 153–5, fig.8). Krzyszkowska cautiously suggests Bes. This must be right, and Weingarten carries it further, arguing that more precisely this figure should be his lesser-known female variant Beset. The pose of the figure – frontal, standing with legs apart – distinguishes it from the familiar bow-legged squat Bes, as do the pendulous and un-Minoan-looking female breasts (Weingarten 2013). Our black Petschaft, then, fits very neatly into an emerging picture of Bes-like images from MM II. Three contemporary hard-stone seals show rather similar faces and Krzyszkowska conveniently illustrates these together on one page (CMS 3 nos. 237b and 238a allegedly from Mallia, CMS 6 no. 101a, allegedly from central Crete; Krzyszkowska 2012a: 154–5, fig. 9; see also Anastasiadou and Pomadère 2011). How did Bes and Beset come to appear in Crete? Bes was popular in both senses of the word, a demon without great temples and retinues of priests and one whose cult was widespread and had a longer life than almost any other (Weingarten 1983: 101–3). In this he matches another immigrant, the demon Taweret (to use but one of her names), whose curious transformation and assimilation in the Aegean has aroused interest and discussion for several generations of scholars now (see especially Weingarten 1991). Both demons play a protective role around childbirth and children. Put them together with the hints that there were Cretan women working as weavers (or ‘websters’, to use an older English term) in Middle Kingdom Egypt, and perhaps we can pin down who brought them both to MM Crete (Barber 1991: 76–7, 351–2; cf. Hughes-Brock 2000b: 122).4 This might explain why the only certain

Weingarten’s ‘possible early Bes’ prism in the Ashmolean has been discussed recently in a major study of Middle Minoan prisms by Maria Anastasiadou, who returns to the old Gorgoneion theory for the ‘Gorgo mask’ motif (as she calls it), arguing that it suggests that the Greek Gorgon had Minoan prototypes (Anastasiadou 2011: vol. I: 207–9; vol. II: 632 no. 494). Recent discoveries, however, at Petras in Eastern Crete have produced several remarkable additions to the body of MM II seals, both loop signets and prisms as well as a fine banded agate seal in the uncommon rectangular plate or tabloid shape. This last is the most Anastasiadou (2011 vol I: 177-8, pls. 17–18) knows of 34 examples. Add now a prism from Petrás with a handsome pair of boars handsome for this soft-stone prism-type which stand out among the other examples (Krzyszkowska 2012a: 149, fig. 4c).

3 

The weaver’s waste which Barber thought probably Aegean has since been radiocarbon-dated as mediaeval (Kemp and Vogelsang-

4  

9

Wonders Lost and Found

Figure 2 a, b and c. Cushion- shaped seal AN2014.1. Agate. Boar.

images of Bes in Minoan Crete are in glyptic, i.e. on small personal objects (Phillips 2008: vol. I, 153).

brown and light brown with narrow white bands, 1.78 x 1.33cm, thickness 0.42, perforation 0.21; its accession number AN2014.1 was only assigned at last in 2013 for the purposes of this article. The boar has a simple round dot eye, some degree of modelling of the body, curved bristly back with (a not uncommon feature) outline running above the bristles, short stumpy tail. He stands head downwards in the way of rooting swine. In the field above his head is a small plant or branch. Condition is good except for an unfortunate damaged place right in the centre, where there is a break through to the string-hole.

The Beset seal from Petras fits in neatly with this, for Petras was an important centre of textile production throughout the MM period and into LM and indeed had even being playing a part in the Eastern Cretan purple dye industry as early as MM I (Burke 2010: 36– 7, 60–1; Brogan, Betancourt and Apostolakou 2012; cf. Moody 2012: 257). The mainly MM II soft-stone prism seals from Eastern Crete may possibly illustrate textile production on one of their quite common motifs, a bar with several round objects attached to it by two little strokes. To Evans this suggested vessels slung on a pole, and most have followed him. Burke, however, has argued for loomweights on a warp-weighted loom (Burke 1997:417–9, pls. 160-1 and 2010: 44–8). Sometimes a human figure appears in conjunction. Burke does not pick up these cases but Anastasiadou (2011: vol. I, 303–4, vol. II, pls. 107–9) includes them in her much fuller list of ‘string vessels’, as she calls the motif. The Ashmolean alone has no fewer than six prism seals showing men holding or touching a bar with Burke’s ‘loomweights’ and three other prisms with both man and ‘loomweights’ but on different faces (CMS 6 nos. 36a, 51c, 59a, 60a, 70c, 71a, on different faces on nos. 50, 66, 68; cf. Wingerath 1995:26). As it happens, one of those with a man and ‘loomweights’ on one face has Weingarten’s ‘possible early Bes’ on another (CMS 6 no. 71a and b). Is this just coincidence? Note that the human figures with this motif are always male. Did women weave the cloth but men see to the marketing of it, rather as in some of the family cloth businesses known from their lively letters in Old Assyrian texts (Thomason 2013, esp. 94–8)?

It comes early in the corpus of agate seals, which were at their most popular rather later and on the mainland. Egypt, with abundant sources of agate in the Eastern Desert, probably provided most of the raw material, such as the chunks found in the Sanidakis plot workshop in Poros, Herakleion, destroyed in LM IA (Dimopoulou 1997: 436–7, pl. 172c). The Indian Subcontinent has good sources too; some agate from there reached the Aegean occasionally by way of Mesopotamia in the form of finished beads (Arnott 2019: 43-5).5The cushion shape in its rather short heyday, MM III–LM I, produced some of the most captivating of Minoan engravings, with vivid depictions of animals and interesting human scenes masterfully executed on beautifully coloured stones (e.g. CMS 6 nos. 177–184 with Krzyszkowska 2005: colour pls. 17, 21, 24, 26). Distinguished banded agate cushions include the famous bull and leaper at a much-discussed rectangular structure (CMS 6 no. 182), an agrimi with spectacular horns which was flattered by being copied on a well-known fake (CMS 6 no. 178), and - from the most dramatic archaeological setting! - the cushion from Archanes with a remarkable boat scene which the tall ‘priest’ engaged in a human sacrifice was wearing on his left wrist when the earthquake struck the

Our next seal (Figure 2a, b and c) unequivocally depicts a boar. It is a cushion-shaped seal (‘flattened cylinder’ in older writings) with the unusual feature of a shapely contoured back, made of a handsome agate, dark

5   The large collection of 29 stones from the Vaphio tholos tomb (LH IIA, contemporary with LM IB) includes a significant number of agates. Agate ex-bead seals are betrayed by their shapes: barrels, very long amygdaloids, natural ‘eye’ stones, and two multitubular beads with intriguing motifs on rectangular faces (Hughes-Brock 2000a: 114–5 and in CMS 6: 17; Krzyszkowska 2005: 123, 196, 239).

Eastwood 2001: 53).

10

Helen Hughes-Brock: Two Cushions, a Bes, a boar and a bead

shrine at Anemóspilia and killed him (Sakellarakis and Sapouna-Sakellaraki 1991: 148–51, fig. 128; Sakellarakis and Sapouna-Sakellaraki 1997: I, 294–308, II, 692–4 figs. 793–5). Our animal is not quite the masterpiece that the finest cushions show us but is well executed. The legs and body are treated in a slightly sketchy way, as on the frustrated barking dog of CMS 6 no.180, which can suggest an earlyish date, MM III–LM IA.

however, for tail behaviour is an indicator of state of mind in swine, as in cats and dogs. Generally speaking, a domestic pig’s tail curled means a happy, confident pig; straight, it can be negative or neutral. Our animal’s tail is held out straight behind, very short at the edge of the stone. On many other seals the tail is not clear. The pig on a rock crystal discoid seal a little earlier than ours (MM II) has a round lump eye and curved back like our pig’s, but its back is smooth and its tail curly (Boardman 1970/2001: pl. 19 = CMS 6 no. 153a). Smooth back and curly tail are also seen on two sturdy long-snouted pigs, the front one with head down as though rooting, on a sealing from Myrtos-Pyrgos (LM I) impressed by a hard-stone cushion (CMS 2,6 no. 232). The pig on a lentoid from the Royal Road at Knossos (LM IB) is so hefty that one might think of a fatted animal, like those mentioned on the tablets, but its back is bristly and it has a tusk; the tail is not clear but may be curled over the rump (Hood 1960: 23–4, fig. 27).6 Another fine lentoid of about the same date shows three animals with well-bristled backs, the front one’s tail curled over its rump (Evans 1935: part 2, fig. 549 = Boardman 1970/2001: pl. 87 = CMS 9 no. 136). Some pigs like these might give the impression of being domesticated (or boar captured and fattened?), but perhaps the engraver meant them otherwise. One might give thought here to Andrew Shapland’s consideration of peaceful ‘nature studies’, such as Evans and Shaw were inclined to see on the examples mentioned above, versus his own view of a more vigorous and subtle expression of an animal’s role in human society (Shapland 2010). Whether as sacrifice, formidable prey for the huntsman or provider of the ultimate in helmets (a Minoan invention, or elaboration of a Middle Helladic one?), the wild boar unquestionably interacts with human society in a more exciting way than a domestic pig does.

Minoan engravers delighted in animals both wild and domesticated, pigs included. Though coming a poor third after sheep and cattle, pigs were important for food and sacrifices (Moody 2012: 237–9, 242–3, 247; Halstead 2007: 28–31). Their pig-snouted sign is instantly identifiable in the Linear A and B tablets dealing with herd management and catering for banquets, and wild boar too occasionally figured in sacrifices and on the menu (Shapland 2010: 113–5; Palaima 2008: 101–3, fig. 12.39; Bendall 2007: 58, 118–21). Which kind is our animal? As Evans remarked, it is not always easy to tell the difference (cf. Krzyszkowska 2014: 344). In a passage on wild boar and domestic swine he takes the fine fat bristly-backed pig on a MM II chalcedony prism seal to be a domestic animal because of the gate in front of it (Evans 1935: 571–4 = CMS 6 no. 95b). The ‘gate’, to be sure, is a fairly common Hieroglyphic sign; earlier, Evans had suggested that the pairing of gate sign and pig ‘seems to indicate some such title as “Keeper of the Swine”’ (Evans 1909: 153 no. P22a, 199; Olivier and Godart 1996: 254 #256, sign no. 038). Olivier once mentioned this suggestion ‘just for fun’ because he found it appealing (his own suggestion was that the ‘gate’ could be read as part of a word or ‘sentence’ running over the three sides of the prism: Olivier 1990: 13; 1981: 113). Another appealing idea is Joseph Shaw’s, that the pig is portrayed at the gate to its sty (Shaw 1978: 247 n. 47). Rather sadly, both these ideas were best abandoned, and likewise Evans’s ‘whole litter of little pigs’ beneath a pair of all-over bristly pigs: the piglets, prosaically, are really nothing more than uneven or rocky ground (Evans 1935: 572 fig. 548 = CMS 2,6 no. 72, a sealing from Hagia Triada, LM IB).

This piece came to the Ashmolean in 1970 as a gift from Herbert Cahn, the director of Münzen und Medaillen, Basle. It was left unregistered because considered dubious or a fake. Who was responsible for that verdict, and why, would be interesting to know but is probably beyond us now. Why it came as a gift, rather than a purchase, is perhaps slightly puzzling, since Cahn was a dealer and collector. At Münzen und Medaillen A.G. he continued his family’s business, specialising in coins and medals, in Basle (originally in Frankfurt-am-Main, but removed from Nazi Germany), but he had a serious scholarly interest in Greek vases too, and also knew his way a little around Aegean seals, of which he included a few in a Münzen und Medaillen exhibition in 1965 (Cahn 1965).

The bristles on the back, as on our seal, do not exclusively denote a wild animal. All ancient swine were bristlier and leaner than modern farm pigs are, so that wild and domestic looked more alike (even in their bones and teeth, which can perplex zooarchaeologists: RowleyConwy, Albarella and Dobney 2012, esp. 2, 36–7). Their tails, however, can mark the difference, the wild animal letting its tail hang down limp and straight while the domestic pig’s tail is curled. Egyptian artists sometimes took care to depict either a straight hanging tail or a curly one (Osborn and Osbornová 1998: 142–3). So, I am told, does the original illustrator of ‘Asterix’, Albert Uderzo. The matter is not entirely straightforward,

The tusk, a line across the face, is clearer in a photo taken by Olga Krzyszkowska, who also corrects the identification of the material (Krzyszkowska 2012b: 742; cf. 2010: 253).

6 

11

Wonders Lost and Found The intermediary was the late John Betts. Betts was the scholar who knew the Aegean seal market and he was moreover engaged in serious study of forgers and forgeries. This was indeed the period at which the Corpus der Minoischen und Mykenischen Siegel, still in its early days, was much occupied with the problem of forgeries and how it should treat them in the CMS catalogues (see Hughes-Brock 2010: 231–2 and e.g. Betts 1981, Pini 1981a). Betts was also the compiler of CMS 10, the Swiss collections, published in 1980, which in fact included ten seals from the Münzen und Medaillen collection. Perhaps, then, Cahn showed our piece to Betts to ask his opinion.7 Did one of them, or both, have doubts about it? Was that perhaps why Cahn was content to donate it? Or perhaps he did think it genuine but unlikely to fetch a good price because of the conspicuous damaged place. Or perhaps, being very much a ‘networker,’ he saw it as a useful piece of ‘social currency’.8 Whatever the reason, he thought the Ashmolean a good place for it.

to Oxford on a work trip and a curious feeling about it nudged me to look at it with her. That we recognised it instantly as a good Minoan seal demonstrates very happily how well glyptic studies have advanced in the last 45 years (see Krzyszkowska 2005: 311–340; HughesBrock 2010: 231–2).9

In the event it was not John Betts but John Boardman who handed it over to the Ashmolean. The very unusual contoured back may well have aroused suspicion. (A well-known cushion from Knossos with flat-faceted back had been illustrated some seven times because of its significant motif of a ‘priestess’ carrying implements, but its profile was not published until 1984: CMS 2,3 no. 16, table p. 459). Boardman did not much like the engraving and was uneasy about the motif: the branch is so close to the animal’s head that it almost created the nonsensical impression of a boar with antlers. Nonetheless, in the normal way of things it would have been entered in the Accessions Register, as were several strange objects in 1968 and 1970, duly registered although some were clearly unorthodox (Hughes-Brock 1989: 79–86, fig. 1). Perhaps in the end the reason why it was side-lined was simply that the Aegean collection was in a state of change then as Hector Catling was preparing to leave the Ashmolean to take up the Directorship of the British School at Athens. By the time Michael Vickers took over in 1971 our seal was firmly settled in the fakes drawer. There it remained until 2012, when Olga Krzyszkowska came

Figure 3. Bead AE 312g. Transparent quartz.

Our second cushion (Figure 3) is unengraved, a bead which never progressed to its undoubtedly intended life as a seal. Its social life (to hijack Appadurai’s happy phrase)10 was clearly eventful but remains unclear. It is of translucent quartz 2cm x 1.25cm, thickness 0.4– 0.45cm; perforation 0.2cm, drilled from both ends, as often, with the join visible in the clear stone; AE 312g). The edges are fairly straight; one side is slightly more curved than the other. There are no fissures visible. The surface is slightly worn. A little patch of paper is stuck to the unmarked side, the remains of an old glued label or display device. It belongs to a group bought by Evans in Athens in 1893 from Athanasios Rhousopoulos. Rhousopoulos, a key figure in the early days of archaeology in Greece, was a collector and dealer like Herbert Cahn, but there the resemblance ends. Though he published little on archaeology, his work for the University of Athens, the National Museum and the Ephemeris Archaiologike was distinguished and lasting (Galanakis 2008: 297–8).11

Betts’s papers (and perhaps seal impressions) are now with the CMS in Heidelberg and await study. They will undoubtedly yield much of interest and must certainly include correspondence with Herbert Cahn. Had Cahn perhaps come to England in 1970, for the funeral or memorial service of Sir John Beazley? His obituary of Beazley is written in warm and personal terms. 8   A ‘social currency’ consideration perhaps lay behind Cahn’s gift to the British Museum in 1960 of a Mycenaean glass seal quite lacking in charm (CMS 7 no. 137 = Krzyszkowska 2005: 268 no. 538, colour pl. 48). Nobody could have foreseen its archaeological importance then. It is a mass-produced mould-formed seal of a significant class not systematically studied until 1981 (Pini 1981b). Moreover, it was formed in the same matrix as seals found later in 1980s excavations in the Elis region and in Thessaly. Such matches, of which there are now several known, indicate something about relations between the sites, though it is not yet certain what. See Hughes-Brock 2008: 139–141. 7  

The details of the Evans-Rhousopoulos deal which began this slightly murky chapter in the social life of our cushion bead are related by Yannis Galanakis 9   An important recent advance: the CMS database is on the internet now. 10   A. Appadurai (ed.), The Social Life of Things: Commodities in Cultural Perspective (Cambridge 1986). 11   For more on Rhousopoulos see Y. Galanakis’ recent articles in BSA 106 (2011), Anglo-Hellenic Review 45 and 46 (2012), Journal of the History of Collections 24 (2012), 25 (2013), AJA 117 (2013); also Hughes-Brock 2000a: 119 n. 37; CMS 6: 7.

12

Helen Hughes-Brock: Two Cushions, a Bes, a boar and a bead

Figure 4. Entries in the Ashmolean register for AE 306 and AE 308-311; two pots and three terracotta figurines found together with AE 312a-g in a chamber tomb at Kará on Mt Hymettus.

(2008: 299–301) together with details of the findplace, a chamber tomb at Kará (now Karéas) on the west slopes of Mt. Hymettus, the illicit excavation of which Rhousopoulos had some connexion with, though exactly how far it went is not clear. The two pots, three terracotta figurines (Figure 4) and seven small objects, including our cushion, were published together with similar material in the Allard Pierson Museum by Joost Crouwel (1973), who had spent time at Oxford working with Catling.

The seven small objects were registered together as AE 312a–g: three conuli, a mended glass bead, a small amber bead, an unperforated stone cylinder and our object (Figure 5; Crouwel 1973: 98–9). Our object apart, this is a rather nondescript handful, though the now perished amber bead plays its little part in the Mycenaean amber statistics and one wonders just what the cylinder was (limestone, not marble as published, 2.8 cm. long; oval in section, so not a vessel core; probably not an unfinished seal; a weight of about 10gm?) 13

Wonders Lost and Found

Figure 5. Entry in the register for AE 312a-g

cheated of its future as a fine Minoan seal, ended its active life a few generations later in a grave in Attica as a bead.

Our clear quartz bead AE 312g was described as ‘milky chalcedony’ in the Accessions Register but ‘chalcedony’ is most often used (and now limited to that use by the CMS) for the blue or bluish-grey quartz from which several dozen seals are made, including the ‘pig and gate’ mentioned above, and a few ‘star’ cushions (Pini 2010: 239). The sources of the blue stone are still uncertain, but the clear quartz has a source in Eastern Crete (Stamatatou 2004: 7). Just as there are several materials, blue chalcedony among them, which almost never occur for beads (the seal-makers evidently got first choice) so too the cushion shape, with rectangular face so obliging for the engraver, almost never occurs for beads (Hughes-Brock 1995: 111–3; 2008: 137–8). Our blank cushion must have come from a Cretan seal workshop and probably not later than LM I. Here begins another murky chapter in its social life. How did it leave the workshop? It was not someone’s private property there, for such workshop material very seldom ended up in graves and not often in sanctuaries.12 The same applies clearly to the moulds for vitreous and gold relief ornaments found near Knossos, one in the Kephala tholos tomb (Hughes-Brock 2008: 136–7), one at Mavrospelio (Evely 2000: 414 no. 21). Disorderly goingson, as e.g. when the Poros workshop mentioned above was destroyed in LM IA, no doubt offered opportunities for craftsmen to purloin, for mainlanders to ransack and so on. It can only have been in some unofficial and probably rather discreditable way that our object,

Bibliography Anastasiadou, M. 2011. The Middle Minoan Three-Sided Soft Stone Prism: A Study of Style and Iconography (Corpus der minoischen und mykenischen Siegel Beiheft 9). Mainz. Anastasiadou, M. and M. Pomadère 2011. Le sceau à “la figure féminine aux bras levés” du secteur Pi de Malia. Bulletin de correspondance hellénique 135/1: 63–71. Arnott, R. 2019. Crossing Continents: Between Meluḫḫa and the Aegean in Prehistory. Oxford. Barber, E.J.W. 1991. Prehistoric Textiles: The Development of Cloth in the Neolithic and Bronze Ages with Special Reference to the Aegean. Princeton. Bendall, L.M. 2007. Economics of Religion in the Mycenaean World: Resources Dedicated to Religion in the Mycenaean Palace Economy (Oxford University School of Archaeology Monograph 67). Oxford. Betancourt, P.P. 1983. Minoan Objects Excavated from Vasilike, Pseira, Sphoungaras, Priniatikos Pyrgos, and Other Sites (University Museum Monographs 47; The Cretan Collection in the University Museum, University of Pennsylvania 1). Betts, J.H. 1981. Some early forgeries: the Sangiorgi group, in Niemeier: 17–35. Boardman, J. 1968. Archaic Greek Gems; Schools and Artists in the Sixth and Early Fifth Centuries BC. London.

12   For other seal blanks see e.g. Evans 1905: 479, fig. 101 nos. 99a/8 (ivory cushion?), 99a/12 (prism?); Evely 1993: 164-5 with n. 98. From the Psychró (Dictaean) Cave: Betancourt 1983: 42 nos. 108–9; CMS 6: 27 no. AE 714. Cf. Hughes-Brock 2008: 136–7.

14

Helen Hughes-Brock: Two Cushions, a Bes, a boar and a bead

Boardman, J. 2001 Greek Gems and Finger Rings: Early Bronze Age to Late Classical 2nd edn. London. Bombardieri, L., A. D’Agostino, G. Guarducci, V. Orsi and S. Valentini (eds) 2013. SOMA 2012: Identity and Connectivity. Proceedings of the 16th Symposium on Mediterranean Archaeology, Florence, Italy, 1–3 March 2012, vol. 1 (British Archaeological Reports International Series 2581). Oxford: Archaeopress. Brogan, T. M., P. P. Betancourt and V. Apostolakou 2012. The purple dye industry of eastern Crete, in Nosch and Laffineur: 187–92, pl. 41. Burke, B. 1997. The organization of textile production on Bronze Age Crete, in Laffineur and Betancourt 2: 411–22, pls. 160-1. Burke, B. 2010. From Minos to Midas: Ancient Cloth Production in the Aegean and in Anatolia (Oxbow Ancient Textiles Series 7). Oxford. Cadogan, G., M. Iacovou, K. Kopaka and J. Whitley (eds) 2012. Parallel Lives: Ancient Island Societies in Crete and Cyprus (British School at Athens Studies 20). Athens.  Cahn, H. A. 1965. Early Art in Greece: The Cycladic, Minoan, Mycenaean and Geometric Periods, 3000–700 B.C. (André Emmerich Gallery, New York, May 7 – June 11, 1965). CMS 2,3 = N. Platon and I. Pini (eds) 1984 Corpus der minoischen und mykenischen Siegel 2,3: Iraklion Archäologisches Museum 3, Die Siegel der Neupalastzeit. Berlin. CMS 2,6 = W. Müller, I. Pini and N. Platon (eds) 1999. Corpus der minoischen und mykenischen Siegel 2,6: Iraklion Archäologisches Museum 6, Die Siegelabdrücke von Aj. Triada und anderen zentral- und ostkretischen Fundorten. Berlin. CMS 3 = W. Müller, I. Pini, N. Platon and A. Sakellariou (eds) 2007. Corpus der minoischen und mykenischen Siegel 3: Iraklion Archäologisches Museum, Sammlung Giamalakis. Mainz. CMS 6 = H. Hughes-Brock with J. Boardman 2009. Corpus der minoischen und mykenischen Siegel 6: Oxford, The Ashmolean Museum. Mainz. CMS 7 = V.E.G. Kenna (ed.) 1967. Corpus der minoischen und mykenischen Siegel 7: Die Englischen Museen II: London, British Museum – Cambridge, Fitzwilliam Museum – Manchester, University Museum – Liverpool, City Museum – Birmingham, City Museum. Berlin. CMS 9 = Henri and Micheline van Effenterre (eds) 1972. Corpus der minoischen und mykenischen Siegel 9: Paris, Cabinet des Médailles de la Bibliothèque Nationale. Berlin. CMS 10 = J.H. Betts (ed.) 1980. Corpus der minoischen und mykenischen Siegel 10: Die Schweitzer Sammlungen. Berlin. Cook, A.B. 1940. Zeus: A Study in Ancient Religion 3. Cambridge. Crouwel, J.H. 1973. Two Mycenaean tomb groups from Amsterdam and Oxford from Mt. Hymettos. Bulletin Antieke Beschaving 48: 91–100.

Dimopoulou, N. 1997. Workshops and craftsmen in the Harbour Town of Knossos at Poros-Katsambas, in Laffineur and Betancourt 2: 433–8, pls. 167-74. Duhoux, Y. and A. Morpurgo Davies (eds) 2008. A Companion to Linear B: Mycenaean Greek Texts and their World (Bibliothèque des Cahiers de l’Institut de Linguistique de Louvain 120) Louvain-la-Neuve. Evans, A.J. 1906. The prehistoric tombs of Knossos. Archaeologia 59 (1905): 391–562. (Also printed independently). Evans, A.J. 1909. Scripta Minoa: The Written Documents of Minoan Crete 1. Oxford. Evans, A. 1935. The Palace of Minos at Knossos 4. London. Evely, R.D.G. 1993. Minoan Crafts: Tools and Techniques. An Introduction Vol. 1 (Studies in Mediterranean Archaeology 92/1). Göteborg. Evely, R.D.G. 2000. Minoan Crafts: Tools and Techniques. An Introduction Vol. 2 (Studies in Mediterranean Archaeology 92/2). Jonsered. Galanakis, Y. 2008. Doing business: two unpublished letters from Athanasios Rhousopoulos to Arthur Evans in the Ashmolean Museum, Oxford, in D. Kurtz with others: 297–309. Halstead, P. 2007. Carcasses and commensality; investigating the social context of meat consumption in Neolithic and Early Bronze Age Greece, in Mee and Renard: 25–48. Hood, M.S.F. 1960. Archaeology in Greece, 1959. Archaeological Reports for 1959-60: 23–4. Hopkins, C. 1961. The sunny side of the Greek Gorgon. Berytus 14: 25–35. Hughes-Brock, H. 1989. The early Cretan white seals in the Ashmolean Museum, ancient and modern: some enigmatic materials, in Müller: 79–89. Hughes-Brock, H. 1995. Seals and beads: their shapes and materials compared, in Müller: 105–16. Hughes-Brock, H. 2000a. ‘Echt oder falsch? Trials, rehabilitations and banishments of some suspects in the Ashmolean Museum, in Müller: 107–21. Hughes-Brock, H. 2000b. Animal, vegetable, mineral: some evidence from small objects, in Karetsou: 120–7. Hughes-Brock, H. 2008. Close encounters of interesting kinds. Relief beads and glass seals: design and craftsmen, in Jackson and Wager: 126–50. Hughes-Brock, H. 2010. The many facets of seal research and the contribution of the CMS, in Müller: 225–37. Jackson, C. and E.C. Wager (eds) 2008 Vitreous Materials in the Late Bronze Age Aegean (Sheffield Studies in Aegean Archaeology 9). Karetsou, A. (ed.) 2000. Κρήτη − Αίγυπτος. Πολιτισμικοί Δεσμοί Τριών Χιλιετιών, Μελέτες [Crete − Egypt, Cultural Links over Three Millennia, Studies]. Athens. Kemp B.J. and G. Vogelsang-Eastwood assisted by others 2001. The Ancient Textile Industry at Amarna (Egypt Exploration Society Excavation Memoir 68). 15

Wonders Lost and Found Kenna, V.E.G. 1960. Cretan Seals, with a Catalogue of the Minoan Gems in the Ashmolean Museum. Oxford. Krzyszkowska, O. 2005. Aegean Seals: An Introduction (Bulletin of the Institute of Classical Studies, University of London, Supplement 85). Krzyszkowska, O. 2010. Material matters: some challenges past, present and future for Aegean glyptic, in Müller: 249–57. Krzyszkowska, O. (ed.) 2010. Cretan Offerings: Studies in Honour of Peter Warren (British School at Athens Studies 18). London. Krzyszkowska, O. 2012a. Seals from the Petras cemetery: a preliminary overview, in Tsipopoulou: 145–60. Krzyszkowska, O. 2012b. Worn to impress? Symbol and status in Aegean glyptic, in Nosch and Laffineur: 739–46, pls. 174-6. Krzyszkowska, O. 2014. Cutting to the chase: hunting in Minoan Crete, in G. Touchais, R. Laffineur, F. Rougemont (eds) PHYSIS: L’environnement naturel et la relation homme-milieu dans le monde égéen protohistorique (Aegaeum 37). Leuven. Kurtz, D. with others (eds) 2008. Essays in Classical Archaeology for Eleni Hatzivassiliou 1977–2007 (Studies in Classical Archaeology 4). Oxford. Laffineur, R. and P. P. Betancourt (eds) 1997. TEXNH: Craftsmen, Craftswomen and Craftsmanship in the Aegean Bronze Age. Proceedings of the 6th International Aegean Conference/ 6o Rencontre égéenne internationale, Philadelphia, Temple University, 18–21 April 1996 (Aegaeum 16). Liège and Austin. Mee, C. and J. Renard 2007 (eds) Cooking up the Past: Food and Culinary Practices in the Neolithic and Bronze Age Aegean. Oxford. Moody, J. 2012. Hinterlands and hinterseas: resources and production zones in Bronze Age and Iron Age Crete, in Cadogan et al.: 233–71.  Müller, W. (ed.) 1989. Fragen und Probleme der Bronzezeitlichen Ägäischen Glyptik. Beiträge zum 3. Internationalen Marburger Siegel-Symposium 5.–7. September 1985 (Corpus der minoischen und mykenischen Siegel Beiheft 3). Berlin. Müller, W. (ed.) 1995. Sceaux minoens et mycéniens. IVe symposium international, 10–12 septembre 1992, Clermont-Ferrand (Corpus der minoischen und mykenischen Siegel Beiheft 5) Berlin. Müller, W. (ed.) 2000. Minoisch-Mykenische Glyptik: Stil, Ikonographie, Funktion. Internationales SiegelSymposium Marburg, 23.–25. September 1999 (Corpus der minoischen und mykenischen Siegel Beiheft 6) Berlin. Müller, W. (ed.) 2010. Die Bedeutung der minoischen und mykenischen Glyptik: VI. Internationales SiegelSymposium aus Anlass des 50 jährigen Bestehens des CMS, Marburg, 9.–12. Oktober 2008 (Corpus der minoischen und mykenischen Siegel Beiheft 8). Mainz. Niemeier, W.-D. (ed.) 1981. Studien zur minoischen und helladischen Glyptik: Beiträge zum 2. Marburger

Siegel-Symposium 26.–30. September 1978 (Corpus der minoischen und mykenischen Siegel, Beiheft 1). Berlin. Nosch, M.-L. and R. Laffineur (eds) 2012. KOSMOS. Jewellery, Adornment and Textiles in the Aegean Late Bronze Age. Proceedings of the 13th International Aegean Conference/13o Rencontre égéenne internationale, University of Copenhagen, Danish National Research Foundation’s Centre for Textile Research, 21–26 April 2010 (Aegaeum 33). Leuven/Liège. Nosch, M.-L., H. Koefoed and E. Andersson Strand (eds) 2013. Textile Production and Consumption in the Ancient Near East: Archaeology, Epigraphy, Iconography (Oxbow Ancient Textiles Series 12). Oxford. Olivier, J.-P. 1981. Les sceaux avec des signes hiéroglyphiques. Que lire? Une question de définition, in Niemeier: 105–15. Olivier, J.-P. 1990. The relationship between inscriptions on Hieroglyphic seals and those written on archival documents, in Palaima: 11–24. Olivier, J.-P. and L. Godart 1996. Corpus Hieroglyphicarum Inscriptionum Cretae (Études Crétoises 31). Athens and Rome. Osborn, D.J. and J. Osbornová 1998. The Mammals of Ancient Egypt (The Natural History of Egypt 4). Warminster. Palaima, T. (ed.) 1990. Aegean Seals, Sealings and Administration: Proceedings of the NEH-Dickson Conference of the Program in Aegean Scripts and Prehistory of the Department of Classics, University of Texas at Austin January 11–13, 1989 (Aegaeum 5). Liège. Palaima, T. 2008. Scribes, scribal hands and palaeography, in Duhoux and Morpurgo Davies vol. 2: 33–136. Phillips, J. 2008. Aegyptiaca on the Island of Crete in their Chronological Context: A Critical Review (Österreichische Akademie der Wissenschaften, Denkschriften der Gesamtakademie 49, Contributions to the Chronology in the Eastern Mediterranean 18). Vienna. Pini, I. 1981a. Echt oder falsch? Einige Fälle, in Niemeier: 135–57. Pini, I. 1981b Spätbronzezeitliche ägäische Glassiegel. Jahrbuch des Römisch-Germanischen Zentralmuseums 28: 48-81. Pini, I. 2010. An unusual LM I four-sided prism, in Krzyszkowska: 239–42. Rowley-Conwy, P., U. Albarella and K. Dobney 2012. Distinguishing wild boar from domestic pigs in prehistory: a review of approaches and recent results. Journal of World Prehistory 25: 1-44. Sakellarakis J.A. and E. Sapouna-Sakellaraki 1991. Αρχάνες [Archanes]. Athens. Sakellarakis, Y. and E. Sapouna-Sakellaraki 1997. Αρχάνες, Μια Νέα Ματιά στη Μινωική Κρήτη [Archanes: Minoan in a New Light]. Athens.

16

Helen Hughes-Brock: Two Cushions, a Bes, a boar and a bead

Shapland, A. 2010. Wild nature? Human-animal relations on Neopalatial Crete. Cambridge Archaeological Journal 20: 109–27. Shaw, J. 1978. Sliding panels at Knossos. Annual of the British School at Athens 73: 235–48. Stamatatou, Ε. 2004. Gemstones in Mycenaean Greece: Their Use and Significance (British Archaeological Reports International Series 1230). Oxford. Thomason, A. K. 2013. Her share of the profits: women, agency, and textile production at Kültepe/Kanesh in the early second millennium BC, in Nosch, Koefoed and Andersson Strand: 93–112. Tsipopoulou M. (ed.) 2012. Petras, Siteia — 25 Years of Excavations and Studies. Acts of a Two-day Conference Held at the Danish Institute at Athens, 9–10 October 2010 (Monographs of the Danish Institute at Athens 16). Weingarten, J. 1983. The Zakro Master and his Place in Prehistory (Studies in Mediterranean Archaeology 26). Göteborg. Weingarten, J. 1991. The Transformation of Egyptian Taweret into the Minoan Genius: A Study of Cultural Transmission in the Middle Bronze Age (Studies in Mediterranean Archaeology 88). Partille. Weingarten, J. 2013. The arrival of Egyptian Taweret and Bes[et] on Minoan Crete: contact and choice, in Bombardieri et al.: 371–8. Wingerath, H. 1995. Studien zur Darstellung des Menschen in der minoischen Kunst der älteren und jüngeren Palastzeit. Marburg. Yule, P. 1981 Early Cretan Seals: A Study of Chronology (Marburger Studien zur Vor- und Frühgeschichte 4). Mainz.

17

Ancient Colchis and the origins of iron: interim results from recent field survey work in Guria, Western Georgia Brian Gilmour, Marc Cox, Nathaniel Erb-Satullo, Nana Khakhutaishvili and Mark Pollard Introduction

published in Russian as Proizvodstvo Zheleza v Drevney Kolkhide, ‘The Manufacture of Iron in Ancient Colchis’ (Khakhutaishvili 1987), and later translated into English (Khakhutaishvili 2009). Radiocarbon and archaeomagnetic dates indicated that most of these sites were operating in the late 2nd and early 1st millennia BC, at least as old as the oldest known iron smelting sites in the Mediterranean. Importantly, however, smelting sites from one region: called the Supsa-Gubazeuli production area, yielded several dates in the early 2nd millennium BC.

Historical sources have long named the southeastern coast of the Black Sea as an early focus of iron production. Greek and Roman sources hint that this area had, by the 5th century BC, gained the reputation as being the region where ironworking originated. Some argue that biblical references to Tubal (Tuval) discovering iron metallurgy refers to the invention of iron smelting by people living in the South Caucasus or north-eastern Anatolia (Kuparadze 2008). However it is also clear that this region was also a known source of some of the other main metals of antiquity, particularly copper and gold. The antiquity of the tradition for early gold working here can be gauged by the ancient myth (mentioned in Homer’s Odyssey, 12.69-72; Lattimore 1967) in which Jason and his fellow Argonauts who sailed to ancient Colchis and found, stole and fled with the ‘golden fleece’ pursued by the Colchian King Aeëtes (Apollonius Rhodius, Argonautica, 4.160-235; Race 2008).

The early fieldwork focused on four areas where prehistoric smelting sites were found to be common, the areas being centred on four particular river systems in this coastal part of western Georgia. Listed from south to north these are the Chorokhi, the Choloki-Ochkhamuri, the Supsa-Gubazeuli, and the Khobi-Ochkhomuri river systems (Figure 2). In Khakhutaishvili’s report (2009 [1987]: 20) it was also noted that a further fifth area of sites was thought to exist in the Black Sea coastal region to the west of Trebizond (modern Trabzon) in the vicinity of which the ancient group of people known as the Chalybes/Khalybs lived (Figure 2). He further suggests, that the names of the Chalybes may derive from their skills in iron metallurgy, meaning literally ‘iron-makers’ (Khakhutaishvili 2009 [1987]: 14-15). Little to nothing is known of sites in this area, however.

Discovery and exploration of prehistoric smelting sites in Western Georgia Sometime after the Second World War prehistoric smelting sites - initially all identified as belonging to ironworking - began to be noticed in the hilly coastal region of western Georgia, the region known to the ancient Greeks as Colchis (Figure 1). The earliest field project investigating these sites was initiated in 1960 by I.A. Gzelishvili (1964), but the most intensive period of investigation was undertaken between 1970 and 1984 by Professor David Khakhutaishvili. During this survey approximately 400 prehistoric smelting sites were noted, and a number of sites from each area were excavated. The great majority of these sites were found as a result of agricultural activities, mostly involving the cultivation of tea, which formed part of the Soviet collective farm system. At the time of considered some prehistoric iron A report on the

In addition to the smelting sites noted earlier, a new group of similar prehistoric copper smelting sites was much more recently identified during fieldwork (by one of the present researchers) in the Chakvistskali river valley, north-east of Batumi (Khakhutaishvili and Tavamaishvili 2002), and a few further sites investigated archaeologically, at Avgia and Khopcho, in the Choroki river area south-east of Batumi (Khakhutaishvili 2006). It has become clear from more recent analytical and dating results that both represent copper smelting, which in the case of Avgia (where two furnaces have been radiocarbon dated) was operating for some period between about the mid-10th century to mid-8th century BC. According the radiocarbon dates, Khopcho belongs to a period several centuries earlier, around the 15th-13th centuries BC (Khakhutaishvili 2006) (see Table 1 below).

their discovery, these sites were of the earliest and most widespread smelting remains yet discovered. excavation of 26 of these sites was 18

Brian Gilmour et al.: Ancient Colchis and the origins of iron

Figure 1. Location map showing the approximate extent of ancient Colchis, with the location of the survey.

time, their use was highly restricted. In addition, GPS technology was unavailable, so the positions of the sites were simply noted in reference to existing local topographical detail, not all of which still survives. The only other possible source of information as to the location of many of the sites noted was the (fading) memory of those people involved with the investigations of particular localities at the time.

Having discovered the existence of this early smelting industry the principal aim of the earlier field project was to undertake a large scale preliminary exploratory survey to investigate the scale, extent, identity, survival and date of the industry. The damage to these early industrial remains caused by the cultivation of the land, largely for tea plantations, was recognised in the interim report (Khakhutaishvili 2009 [1987]). Despite the substantial amount of work carried out over the last 50 years by Georgian archaeologists, a number of key questions remain, especially those relating to the technology, chronology, and spatial distribution of the industry.

Background to the new landscape field project Unfortunately, early Georgian archaeological field research remained almost completely unknown to the wider world of archaeological scholarly research, partly because of restrictions placed on the flow of information from this region, and partly because much of the work was published in Georgian and Russian (Gzelishvili

Mapping the widespread scattering of sites that were discovered was still a big problem, because although accurate maps at different scales did exist at the 19

Wonders Lost and Found

Figure 2. Location map of known/suspected prehistoric smelting sites in western Georgia and north eastern Turkey (reproduced from Khakhutaishvili 2009 [1987]: 20).

1964; Khakhutaishvili 1987). Khakhutaishvili’s 1987 book detailed the excavations of 26 sites subjected to more intensive examination. Information on the whereabouts of the rest of the circa 400 sites mentioned by Khakhutaishvili (2009 [1987], 17) survived only as topographical descriptions and sketches in the original field note books kept at the time of the survey.

area contains substantial Bronze Age and Early Iron Age occupation as well (Mikeladze and Khakhutaishvili 1985). This collaborative research laid the groundwork for our renewed exploration of smelting landscapes in the broader region around Pichvnari. The 2009 republication in English of David Khakhutaishvili’s interim report also included an appraisal both of what had been achieved and what remained to be done. This assessment revealed that, although very extensive work had been undertaken, many of the sites would need relocating and mapping, then examining (in a few cases re-examining) using the most up-to-date archaeological research techniques, before all knowledge of even their whereabouts was lost.

After the collapse of the Soviet system, the development of new international collaborations in western Georgia opened up the possibility of following up this earlier work before the knowledge of the site locations, or the sites themselves, became lost. Foremost among these was the British-Georgian collaboration on the excavations at Pichvnari by Professors Michael Vickers (Oxford University, UK) and Amiran Kakhidze (Batumi Archaeological Museum). While Pichvnari is best known for its Classical and Hellenistic remains (Vickers and Kakhidze 2001; Vickers and Kakhidze 2004), the

Although a programme of dating was carried out earlier on some of the sites, many questions remained 20

Brian Gilmour et al.: Ancient Colchis and the origins of iron

about the dating and chronological development of individual sites, some of which had radiocarbon dates spanning several centuries. Modern dating techniques such as accelerator radiocarbon determination and optically stimulated luminescence (OSL) should allow a much clearer dating picture to emerge. It was also clear that modern archaeological survey techniques - for land survey as well as geophysical prospection - would greatly aid the pin-pointing of individual sites and the planning of systematic archaeological investigation.

Gubazeuli river zone was selected for the initial pilot season of exploratory fieldwork in the autumn of 2010, to be carried out by a joint UK (University of Oxford, Research Laboratory for Archaeology and the History of Art) and Georgian (Shota Rustaveli State University, Batumi and the Batumi Archaeological Museum) team, and was made possible by a British Academy grant. After an initial field visit in 2009, the joint GeorgianBritish field project was begun in September 2010. Its main aim is to build on the earlier pioneering archaeological survey work carried out across western Georgia. This potentially large database of sites was to be investigated using a combination of field survey techniques, limited ‘pin-point’ excavation and scientific identification and dating. One main objective was to examine the evidence for a transitional phase in the Late Bronze Age and Early Iron Age, when the manufacture and exploitation of iron first emerged in the region, a technological transition that is still poorly understood.

With regard to the nature of metal production activities occurring at these sites, some chemical and micro-structural analyses of slags were conducted by Georgian specialists as part of this earlier program of investigation. These reports concluded that the smelting remains are that of iron production (Inanishvili 2007; Tavadze et al. 1984). Unfortunately, most of the chemical analyses do not report copper and zinc contents, and the few published photomicrographs are difficult to interpret. A recent large-scale analysis of slags from many sites, however, has clearly shown that many of the smelting sites across the region relate to copper smelting, although the slags (and the ores) were also rich in iron (Erb-Satullo et al. 2014). The significant iron content of the slags may have been the source of some of the confusion.

The first field season was based in the modern Georgian administrative region of Guria, with the overall aim of locating, mapping, examining recovering identifiable field evidence from the industry in the Supsa-Gubazeuli area where the earliest dates were previously obtained (Figure 3). As elsewhere in Western Georgia, only a small proportion of the sites were previously examined in any kind of detail and an analytical appraisal of the more detailed nature and development of the industry had also been left for future work. The intention was also to look at how we might identify and explore the exploitation of the landscape and its development for this prehistoric industry. During this initial exploratory field season nearly 30 sites were found and mapped, and although most of these were sites observed in the original survey, some new sites were also noted.

Research from elsewhere in the eastern Mediterranean region has suggested that, although iron occurs in archaeological contexts before or during the mid-2nd millennium BC, it is not until the late 2nd millennium BC, or even the early 1st millennium BC that iron becomes increasingly common in the archaeological record (Waldbaum 1999; Yahalom-Mack and EliyahuBehar 2015). Thus, while there was no particular reason to doubt the overall dating results from the early survey work in western Georgia, the very early dates (see above) for a series of sites in the Supsa-Gubazeuli production area were surprising, at least for a series of iron smelting furnaces. Moreover, in the 1987 publication Khakhutaishvili suggested the likelihood that iron smelting in this region developed from copper smelting, but no definitive analytical research was carried out to substantiate this idea (Khakhutaishvili 2009 [1987], 112).

In the summer of 2012 a second season of exploratory field survey work was carried out in this same Guria area with the aim of continuing the work started in 2010. Roughly 20 additional sites were located in 2012, and with a short return visit to re-evaluate some sites in 2014, a total of 49 sites in the Supsa-Gubazeuli River region were mapped – more or less the same overall number as in the original survey (Figure 4). Most of the sites lay inside a 5 km square area within the districts of the modern villages of Mziani, Askana, with some situated in the adjacent villages of Mshvidobauri and Nagomari. However, not all the original sites were located, but this was balanced by the discovery of several new ones. This second season of investigative work in Guria was made possible by an exploratory grant from National Geographic.

The results of the early survey work not only demonstrated that metal production was very widespread along the hilly inland part of the Black Sea coastal region of western Georgia, but also that it appeared to have been active, in the Supsa-Gubazeuli region at least, between about the first half of the 2nd millennium and the mid-1st millennium BC. This is just the period when the emergence of iron smelting technology, possibly from a copper smelting tradition, might be expected. For this reason the main cluster of smelting sites reported in the area of the Supsa21

Wonders Lost and Found

Figure 3. Google earth view southwestern Georgia with the field survey area outlined in red, lying just south of the Supsa River. Note the position of the survey area in an intermediate zone between the Colchian lowlands and the large Rioni river basin to the north, and the more mountainous regions of the Lesser Caucasus to the south.

2010 and 2012 Guria field survey results

The base for the project was established at Mziani, near the centre of the five km square area (see Figure 4) where nearly all the sites seen earlier were said to occur. The first survey priority then was to list, identify and plot the approximate position of as many as possible of all smelting sites previously noted in the study area. As far as possible this was done by close examination of the original field notes together with what local knowledge was still available. The approximate positions of 50 smelting sites which had been observed previously were thus plotted. The next task was to hunt for them on the ground so that they could be plotted more accurately.

Those areas where the smelting sites had been located lay inland beyond the low-lying coastal area in the hilly zone between the various rivers which flowed westwards into the sea. The region is characterized by thick layers of diluvial clay, which sit on top of erosionderived laterites (Khakhutaishvili 2009 [1987]: 53). In several areas along the coast, magnetite-rich sands, originating from inland outcrops, are found. The main concentration of sites had been noted in hills above the middle region of the Supsa River, near the valley of its tributary, the Gubazeuli River (Khakhutaishvili 2009 [1987]: 53).

It soon became clear that the vast majority if not all the sites previously noted lay in the large areas exploited for tea plantations or smaller scale-agricultural operations (such as hazelnut growing). Additional exploration was also carried out in several woodland areas to begin to assess the potential for the discovery of previously unknown sites. Some sites which were identified in earlier fieldwork were difficult to locate, but in other cases sites were found by using a magnetic susceptibility probe in a ‘free-form’ exploratory mode. Sites were sometimes nearly invisible at the surface, hidden beneath dense vegetation or disturbed by modern activities. The potential for locating industrial

Given the extremely dense subtropical vegetation, sites, even previously excavated ones, were not easy to find. However, it proved possible to locate many of them using a combination of the original notes as to their whereabouts together with local knowledge where this still existed. Thus our first task was to use the original notes as to the location of the sites and plot the approximate positions of the sites on the relevant part of a copy of one of the old (1:50,000) military maps before we set out to look for them. 22

Brian Gilmour et al.: Ancient Colchis and the origins of iron

  Figure 4. Map of the Supsa-Gubazeuli copper smelting sites area overlaid onto the 1:50,000 Soviet military map of this area. One site (Site 34/Vakijvari I) was located about 10 km to the south of the main cluster and is not shown on this map.

and habitation sites by measuring the magnetic component of the topsoil tends to be overlooked by archaeologists, but is particularly effective in the case of smelting sites.

well suited to finding, mapping and investigating this type of site. This approach yielded good results and enabled the extents of (largely invisible) slag scatters to be mapped, and the central positions of furnaces to be pin-pointed to within about 25cm. Limited ‘keyhole’ testing was then carried out to test the results of the geophysical survey, confirm the identifications of features, examine site survival, and recover evidence as to their nature, use, layout and development. Stratified remains were recorded and samples taken for scientific analytical identification and dating.

In these areas south of the Supsa River, the very overgrown nature of the old tea plantations occupying much of this landscape meant that this approach had to be applied in a more targeted way to less vegetated areas where systematic geophysical survey was more feasible and where traces of slag scatters indicated the presence of smelting in the vicinity. One area of former tea plantation, where the presence of one or more now disappeared sites could only roughly be estimated from previous notes, was selected for more detailed field survey (Figure 5). We used topsoil magnetic susceptibility to locate the sites, followed by gradiometry (magnetometry) to look at their layout and select specific targets suitable for examination by excavation.

The target area for magnetic susceptibility survey was laid out as 100m grids with susceptibility measurements being taken at 10m intervals. Magnetic susceptibility is a measure of the content of magnetic particles: mostly magnetite, the magnetic form of iron oxide, in the topsoil. The results from this magnetic susceptibility survey (using a Bartington MS2 susceptibility meter) were then plotted so as to produce a magnetic susceptibility concentration or ‘contour map’ of each 100m grid

This sequential method of archaeological geophysical survey, moving from larger to small scale, is particularly 23

Wonders Lost and Found

Figure 5. View looking north-west across the abandoned tea plantations south-west of Mziani village. Site 5/Askana V occupies the centre of the low hill in the middle distance, towards the left. The partially exposed site of Site 1/Askana I is just out of the photograph to the lower left.

(Figure 6). This map shows the position and approximate extent of two smelting sites. One site (Site5/Askana V), was possibly seen before but subsequently lost, while the other (Site 1/Askana I), was previously excavated and was partially still visible. In addition to these more obvious foci of activity, a less strong signal was noted about 150m towards the south of the area mapped for magnetic susceptibility.

(50m or so) west of the area where a previously noted site (Site 6/Askana VIII) was thought to be. The overgrown nature of the undergrowth meant that only limited magnetometry was possible, and that precision in targeting this method of survey was important. However, the results of the magnetometry survey over the two target areas were both very informative and interesting, and at first slightly puzzling in the case of thThe magnetometry results were all instructive (Figure 7). It would appear that the flat and relatively low lying field on one side of this first study area (the paler area in the aerial view in fig. 5) has long been used for agriculture, it being relatively featureless except for a relatively modern land-drain running across it near the northern end (Figure 7, Area 1). This was suspected to be the case from the relatively even, low magnetic susceptibility results for this area. The north-eastern magnetometry grids were centred on the area of high susceptibility here (see Figure 6) which was found to consist of two distinct slag scatters on either side of a furnace (Figure 7, Area 2).

Next, we carried out a gradiometry (magnetometry) survey over the (hot-spot) areas of highest topsoil magnetic susceptibility to plot the layout of these areas, mainly to reveal the extent of undisturbed slag dumps, positions of any surviving furnace(s) which lay beneath the disturbed topsoil. A single 30m square magnetometry grid was laid out over the main areas of slag concentration of Site5/AskanaV, the previously noted site which was no longer visible (Figure 7). A third grid was also laid out over the part of the area to the south showing a lower, more diffuse magnetic susceptibility concentration. This area was slightly 24

Brian Gilmour et al.: Ancient Colchis and the origins of iron

Figure 6. Magnetic susceptibility contour map of the area to the north-west of the nearby village Mziani (but still mainly within the district of Askana) showing the areas of two smelting sites as concentrations of magnetically active particles in the topsoil (image courtesy AE Johnson).

The sandstone slabs and the few grey blocks immediately above them would appear to have been packed in at the same time whereas the few stone blocks above were set at a different angle and were packed into burnt clayey material which was itself placed on top of the lowermost grey lining blocks. Thus the lower stones forming the remnants of the furnace shaft lining belonging to the same earlier phase of use after which the contents of much of the furnace were dug out. Then, the surviving lowermost blocks were left in place and the upper part of the furnace shaft rebuild ready for the next and final furnace firing. After this process, much of the contents of the furnace were again dug out followed by the backfilling and abandonment of this particular site. The volume of slag and related waste on the site – perhaps in the order of about 5 -10 tonnes based on volume estimates for sites like these (e.g. Khakhutaishvili 2009 [1987]: 55, 57) would indicate that this pit furnace was reused in this way many times.

Magnetometry thus pin-pointed the site of a single furnace which was uncovered and investigated with a limited 2m x 2m test trench (Figure 8). The goal of the test trench was to investigate the form of the furnace and to recovering stratified samples for both optically stimulated luminescence (OSL) and AMS radiocarbon dating. The furnace was of a very distinctive form and consisted essentially of a sunken shaft. This was built by first digging a circular, slightly tapering pit cut 1.2m down though the buff yellow clay which forms the subsoil in this region. At the top this pit measured approximately 1m in diameter narrowing down to approximately 0.5m at the bottom. Then pit was lined with stone of two main types so as to form the working shaft of the furnace. The lowermost 35cm was then lined with 5 slabs of dark rusty brown, relatively soft sandstone and these were still in place at the base of the pit (Figure 8). Above this the lining consisted of chunks, or some complete water-worn lumps, which may have come from the nearby Natanebi River, of a harder grey rock probably of igneous origin. It was clear that the stone packing that survived above the sandstone slabs showed evidence of two phases of use of the furnace.

After the furnace went out of use the shaft was backfilled, presumably shortly after it was last used, with a mixture of burnt ashy material, cinders, slag, fragments of furnace shaft lining, clay fragments and the like. Much of this material must simply have been what was dug out of the furnace after the last firing so as to extract the final product of the smelting, either a copper ingot or block of copper sulphide matte, from 25

Wonders Lost and Found

a b

Figure 7. Overall results of the magnetometry survey over the first three areas covered, showing (a) a relatively featureless (except for a modern land-drain) meadow to the west and two former iron smelting areas to the north-east (a and b) and south-east (b), the latter having been previously disturbed (images courtesy AE Johnson).

26

Brian Gilmour et al.: Ancient Colchis and the origins of iron

a

b

c

Figure 8. Site 5/Askana V (a) Remains of the furnace after the removal of the brown silty accumulation (visible in the top section here looking east) over the abandoned furnace. (b) Detail of the same view showing the remains of two periods of furnace use, the upper stones relating to secondary furnace re-use, while the lower stones belong to an earlier phase of use. (c) View looking west down into the furnace after removal of the material backfilled before the site was abandoned.

27

Wonders Lost and Found the lower part of the shaft. There was no evidence of any superstructure to the furnace, although one may have existed, perhaps in the form of a low circular chimney continuing the shaft above the ground.

magnetic susceptibility and magnetometry, plus some key-hole testing. It was clear that the most eastern of these two sites (Site 45) had been quite heavily eroded, probably well before the recent ploughing, and that this ploughing had caused further damage.

The third set of magnetometry grids was centred over a more diffuse and less intense area to the southeast of the area measured for magnetic susceptibility. Interestingly the magnetometry results here gave a distinctively stripy pattern (Figure 7a). This area was also found to be just downhill (to the west of) a deep erosion gully, and careful inspection of the surface showed a wide scatter of slag but no central focus. What seems to have happened here is that an original site, situated 50m or so to the east, has been eroded and/ or washed away (probably by the heavy rains typical of the region) the slag having been re-deposited over the lower lying land downhill to the west. Subsequently this land has been ploughed, perhaps before planting of the tea bushes here, this operation having been responsible for the stripy effect on the magnetically rich topsoil here.

Topsoil magnetic susceptibility on Site 45 suggested that only half the site lay in the hazelnut plantation, and that the rest lay further to the east in an area of dense undergrowth and small trees, which was not possible to investigate. Magnetometry over the western half of the site indicated this to be a much disturbed area of prehistoric slag dumping, with any surviving furnaces lying slightly uphill to the east, beyond the area where investigation was possible. By contrast Site 46 lay well within the study area and magnetic susceptibility survey gave a good target area for magnetometry survey which confirmed the visible impression that the slag dump here had been heavily disturbed and further spread out by recent ploughing. However the furnace was pin-pointed by magnetometry enabling it to be investigated with a keyhole (2m x 2m) test trench (Figure 9).

A similar geophysical survey was carried out in an area of land, now used as a hazelnut plantation, approximately 1km east of the first study area. The hazelnut plantation was only established within the past 5-10 years, in an area which had formerly been part of one the Soviet collective farm tea plantations. Before planting, the land was heavily ploughed, scattering the slag dumps associated with two new prehistoric smelting sites: Site 45/Askana XXVI and Site 46/Askana XXVII, neither of which was previously noted by David Khakhutaishvili during his earlier investigations.

As with Site 5/Askana V, the magnetometry investigation of the magnetic susceptibility ‘hot-spot’ here, covering the approximate extent of Site 46/Askana XXVII, showed the more-or-less exact position of the furnace, enabling a detailed archaeological investigation to be carried out over a precise 2m square area. Again, the furnace was found to have been backfilled, most likely with the mixed burnt slag and charcoal-rich debris left over when the furnace was dug out to remove the lower contents, shortly after the last firing of the furnace. The structure of the furnace was quite similar to that found on Site 5/Askana V. However, in this case the furnace pit was slightly wider and shallower, measuring

Both sites were investigated in 2012 using the same combination of geophysical techniques; topsoil

Figure 9. Site 46/Askana XXVII, two views of the furnace after removal of post-abandonment backfill debris showing the largely intact (possibly primary) sandstone packing of the lower shaft, plus some surviving clay packing, uppermost here, of a secondary phase of furnace use. A few rounded, water-worn stones, the remainder of the upper packing of the shaft, are also visible.

28

Brian Gilmour et al.: Ancient Colchis and the origins of iron

approximately 1m deep by 1m across at the top (Figure 9). Again, the lower part of the furnace pit lining forming the actual furnace shaft survived in place. As before the lowermost part of this lining was found to be made with blocks (rather than slabs as at Site 5/Askana V) of rusty brown, iron rich sandstone, above which what little remained of the rest of the lining consisted lumps of grey igneous rock. It would appear that this stone lining belonged to the penultimate reconstruction (possibly of many reconstructions) of the furnace. Some of the thick clay packing above this stone lining appeared to have been put in place later and to be last remnant of the final rebuilding of this furnace before it was backfilled and abandoned.

spatial patterns of ore acquisition within this landscape (Erb-Satullo et al. 2017). Results from the earlier survey and excavation work, together with the more detailed recent work has shown the multiple use of furnaces which essentially consist of pits sunk up to about 1.2m into the clayey subsoil of the lower hilly part of this region, these sunken shaft furnaces being supplied from above by an air blast via multi-part tuyère tubes. It was clear that the slag was of two distinct types, the first of which was more heterogeneous, relatively dense, but still quite porous with many small gas bubble holes. This contrasted with the second form of slag which was more homogeneous, very dense and heavy and fewer, larger gas bubbles. On most of the sites, both forms of slag occur together, though some sites appear to have more of one type than the other. This may be due to the nature of surface exposure, however, especially if the two types of slag were dumped in slightly different areas.

In addition to the excavation of the two furnaces a series of small exploratory test pits were dug on the areas of waste debris (slag and related material) on these two sites as well as various other sites where the slag dumps survived relatively well. The aim here was to get a more controlled and representative view of the proportions of different types of slag remains and other waste debris present. On many other sites only a surface collection of waste debris was feasible although as far as possible a representative sample of the waste debris encountered on each site was collected from all of the sites. In some cases, site preservation was very poor either due to later disturbance or erosion or both, making the slag assemblages more partial for some sites.

Strong evidence for this comes from our more detailed study of Site 5/Askana V, where two slag dumps showed up on the magnetometry, one on either side of the central furnace (Figure 7b). The more complete dump to the south-east of the furnace consisted almost entirely of the very inhomogeneous, more porous slag which matched the type of slag most common in the furnace backfill. By contrast, the slag scatter to the north-west of the central furnace consisted almost entirely of the homogeneous, very dense form of slag. Unfortunately this slag tip was much less well preserved and much of it seems to have been lost to erosion down the steep gully to the north.

Identification, analysis and dating of field remains and production debris Most of the waste debris collected during the two seasons of work fell into two main categories; firstly slag, the wholly or partially fused stone-like by-product of the metal production at these smelting sites, and secondly sherd-like pieces of very coarse ceramic material that served as either crucibles or furnace lining. Occasional pieces of tuyère and non-metallurgical ceramic were found, but these were generally rare.

When both forms of slag are freshly broken open the occasional greenish patch can be seen, strongly suggesting that these slags relate to a copper smelting process. Large numbers of the slag samples from various sites has been subjected to qualitative X-ray fluorescence (XRF) analysis shows, as expected for copper smelting, the presence of a relatively small proportion of copper and a much larger proportion of iron which is also evident from the rusty appearance of some of the slag.

Overall it is clear from a substantial program of analytical work (X-ray fluorescence spectrometry, scanning electron microscopy, and optical microscopy) that all the sites encountered in the Supsa-Gubazeuli area were used for copper smelting, although, as is typical of copper smelting slags, iron formed as a significant component of the slags (Erb-Satullo et al. 2014). This may to some extent have misled the earlier researchers looking at this material. More unexpected was the large proportion of zinc in the slags, although this simply reflects the polymetallic nature of the ores. The presence of zinc in the slag from some sites but not others, suggested that different ores, with a greater or lesser presence of zinc, were exploited (Erb-Satullo et al. 2014: 153-156). This fact has allowed us to map

Quantitative chemical and microscopic analysis demonstrated unequivocally that these slags relate to copper production. Moreover, the results of these analyses strongly suggest that the main ore source used was chalcopyrite, possibly with some weathering (ErbSatullo et al. 2014: 155-156). The smelting process would likely have been carried out in two or more stages, possibly with roasting in an oxidizing atmosphere as a component of the process. The roasting makes the copper compounds more friable, creating a high surface area for reducing reactions and removes some of the sulphur. Copper smelting technologies at these 29

Wonders Lost and Found sites are discussed in more detail in separate papers (Erb-Satullo et al. 2014; Erb-Satullo et al. 2015).

Guria industry (2009 [1987]: 105-106). None of the dates obtained so far match the early second millennium BC radiocarbon dates obtained from a couple of sites in earlier fieldwork.

Discussion, conclusions, and future directions Four radiocarbon dates from the Supsa-Gubazeuli area sites conform with earlier results, showing that the sites mostly date to the Late Bronze and Early Iron Age. One of the charcoal samples came from lumps of slag recovered from surviving waste dumps whereas the remaining three came from the two furnaces identified (from Site 5/Askana V and Site 46/ Askana XXVII respectively) and examined during this campaign of work. The calibrated dates span the period from the early 12th century to the late 9th century BC (see Table 1) which is in rough agreement with the dating obtained by David Khakhutaishvili for the

A comprehensive analysis of all new radiocarbon dates (Erb-Satullo et al. in press) has clarified and adjusted our understanding of the exact timing and intensity of copper production. Nevertheless, David Khakhutaishvili’s radiocarbon and archaeomagnetic dating of the sites appears roughly accurate, though the earlier early-2nd millennium dates have not be confirmed. It is still a puzzle, however, as to why the earlier unfinished work on this region concluded that all the sites related to ironworking when, during our study of these sites, it soon became clear that they were associated with copper smelting. It is possible that the high iron content of some of the Guria (and other) copper smelting slags may have misled the earlier researchers. Our study of the Guria slags suggests that the very high iron content of these slags is the result of the smelting of chalcopyrite (CuFeS2) ores with other iron-bearing compounds (e.g. hematite) as accessory minerals.

Table 1

It seems most likely that copper smelting, and iron smelting when it began, would have been carried out in areas where the combination of ores, fuel, and other resources were readily available. Detailed analysis of these copper smelting landscapes (Erb-Satullo et al. 2017), suggests that these sites formed part of an expansive system of numerous small independent metalworkers, carrying out largescale metal production without the oversight of a centralized authority. The detailed mapping, recording, topographical contexts, and descriptions of the sites are now mostly complete for our study area in the Supsa-Gubazeuli River region of Guria. However, further work in the form of the detailed geophysical study and more substantial excavation of one or 30

Brian Gilmour et al.: Ancient Colchis and the origins of iron

more specific sites (or groups of sites) remains. This future work will allow us to assemble enough evidence to reconstruct exactly how this particular prehistoric smelting industry worked, from the mining of the ores through to the production of the metal objects.

which have not yet been identified but are waiting to be found. Our exploratory surveys in other regions of Western Georgia (mountainous Adjara and Samegrelo) have identified iron smelting sites, but as yet, none are contemporary with the copper smelting landscapes of Guria (Erb-Satullo et al. 2014: 154-155; Erb-Satullo et al. in press).

Expansion of the initial dating program has provided a more reliable idea of the chronology and longevity of this industry, and provides clues as to how long each site may have remained in operation (Erb-Satullo et al. in press). A key issue, discussed in more detail elsewhere (Erb-Satullo et al. 2017; Erb-Satullo et al. in press) is the extent to which smelting operations moved about or stayed in one place, and the extent to which copper smelting was a seasonal activity. It is likely that the landscape at the time was more wooded than it is now, and it may be that the woodland landscape would have been managed, or at least exploited, quite carefully to allow for the regeneration of fuel resources. Similar practices have been documented for early smelting industries elsewhere (as in the UK for the Roman and later iron industry of the Weald; see Rackham 1986). Comprehensive paleobotanical analysis at these Colchian smelting sites, or sites in their immediate vicinity, would also reveal the impact that this longstanding smelting industry had on local vegetation.

There are many issues relating to the prehistoric Guria copper smelting industry that still need addressing, and these form part of our ongoing research in this region. We are also beginning to study both the other areas looked at by David Khakhutaishvili, as well as new areas with the aim of putting together a more cohesive picture of this giant jigsaw puzzle. Acknowledgements This exploratory research project would not have been possible without the financial support of the British Academy (SG100285) and National Geographic (GEFNE 30-11). The research is also supported by a NSF Graduate Research Fellowship (DGE0644491 and DGE1144152) and a NSF Doctoral Dissertation Improvement Grant (BCS-1338893) awarded to Nathaniel Erb-Satullo. We have also received financial support from the John Fell fund of the University of Oxford. In addition to this we are also grateful for the support and encouragement we have received from various people, in particular the Batumi Archaeological Museum and its director Professor Amiran Kakhidze, and to Dr Emzar Kakhidze for helping to set up and get the project going. We would also like to acknowledge the encouragement and advice we have received from Professor Michael Vickers and Dr Manana Odisheli, for their invaluable help and encouragement in getting the project started in the first place, and to Dr Tony Johnson for his work in setting up the geophysics element to the project, running this in 2010, and for producing the resulting images shown in this paper.

A key question which we are working towards is how, when and where the smelting of copper in western Georgia may have led, or given way to the smelting of iron. The late 2nd and early 1stt millennium BC (that is our period of interest) is the period where we might expect to see evidence for the emergence of iron as a material used alongside bronze for tools and weapons. Indeed, the assemblages of metalwork in mortuary contexts show that iron and bronze were used to make very similar artefacts (Papuashvili 2011), suggesting that the eastern Black Sea region is an ideal place to examine this technological transformation. We need to link the industries we are studying to products of these industries. In the case of bronze objects we also need to at least be able to suggest where the copper might have been alloyed with tin, as well as where the tin might have come from, and where the objects might have been made, and if processes such as recycling may have been an issue. Some analyses have shown that alloying with tin did occur at at least one copper smelting site, but this does not seem to have been the normal practice (Erb-Satullo et al 2015).

Bibliography Erb-Satullo, N., B.J.J. Gilmour and N. Khakhutaishvili, 2014. Late Bronze Age and Early Iron Age copper smelting technologies in the South Caucasus: The view from ancient Colchis c. 1500-600 BC. Journal of Archaeological Science 49: 147-159. Erb-Satullo, N., B.J.J. Gilmour and N. Khakhutaishvili, 2015. Crucible technologies in the Late Bronze-Early Iron Age South Caucasus: Copper processing, tin bronze production, and the possibility of local tin ores. Journal of Archaeological Science 61: 260-276. Erb-Satullo, N., B.J.J. Gilmour and N. Khakhutaishvili, 2017. Copper production landscapes of the Late Bronze-Early Iron Age South Caucasus. Journal of Anthropological Archaeology 47: 109-126.

In the case of iron production, we need to work out where the metal was being smelted, and whether other production activities (e.g. refining and smithing) were performed in different places. In the case of iron, these considerations do not appear to relate to any of the Guria sites. There may be areas of prehistoric iron smelting in Guria, and elsewhere in western Georgia, 31

Wonders Lost and Found Erb-Satullo, N., B.J.J. Gilmour and N. Khakhutaishvili, The Ebb and Flow of Copper and Iron Smelting in the South Caucasus. Radiocarbon (in press). Ghambashidze, D. 1919. Mineral resources of Georgia and Caucasia: Manganese industry of Georgia. London: George Allen. Gzelishvili, I.A. 1964. Zhelezoplavil’noe Proizvodstvo v Drevney Gruzii [Iron smelting production in ancient Georgia]. Tbilisi: Metsniereba. Inanishvili, G. 2007. About the history of iron production in Georgia. Metalla 14: 1-62. Khakhutaishvili, D.A. 1987. Proizvodstvo Zheleza v Drevney Kolkhide [The production of iron in ancient Colchis]. Tbilisi: Metsniereba. Khakhutaishvili, D.A. 2009. The manufacture of iron in ancient Colchis. Oxford: Archaeopress. Khakhutaishvili, N. 2006. Ancient iron production related to the recent findings on Gonio Castle surroundings (2001-2003). Eirene 42: 222-234. Khakhutaishvili, N. and G. Tavamaishvili 2002. Udzvelesi rkinis metalurgiis akhali kera chakvistsqalis kheobashi [One more centre of ancient iron metallurgy in Chakvistsqali Basin]. Dziebani: The Journal of the Centre for Archaeological Studies, Georgian Academy of Sciences: 34-40. Kuparadze, D.M. 2008. Ancient Georgian iron metallurgy and its ore base, in . R. I. Kostov, G. B. and M. Gurova (eds) Geoarchaeology and Archaeomineralogy. Sofia: St. Ivan Rilski. Lattimore, R. 1967. The Odyssey of Homer. Translated with an introduction by Richmond Lattimore. New York: Harper Collins. Mikeladze, T.K. and D.A. Khakhutaishvili 1985. Drevnekolkhidckoye poseleniye Namcheduri I [Ancient Colchian Settlement Namcheduri I]. Tbilisi: Metsniereba. Papuashvili, R. 2011. K voprosu ob absolyutnoy khronologii mogil’nikov kolkhidy epokhi pozdney bronzy-rannego zheleza [On the question of the absolute chronology of the cemeteries of Colchis in the Late Bronze - Early Iron Age]’, in Z. K. Albegova, M. K. Bagaev and S. N. Korenevskiy (eds) Voprosy Drevney i Srednevekovoy Arkheologii Kavkaza [Questions of Ancient and Medieval Archaeology of the Caucasus]: 82-94. Grozny: Uchrezhdeniye Rossiyskoy Akademii Nauk Institut Arkheologii. Race, W.H. 2008. Apollonius Rhodius, Argonautica; edited and translated by William H. Race (Loeb Classical Library). Cambridge MA and London. Rackham, D. 1986. The History of the Countryside. London: Dent. Tavadze, T.N. G.V. Inanishvili, T.N. Sakvarelidze and T.H. Zague 1984. Issledovaniye drevnikh shlakov zhelezhogo proizvodstva na territorii Gruzii [Investigation of Ancient Slags of Iron Production on Georgian Territory]. History of Science: 21-28.

Vickers M. and A. Kakhidze 2001. The British-Georgian excavations at Pichvnari 1998: the ‘Greek’ and ‘Colchian’ cemeteries. Anatolian Studies 51: 65-90. Vickers M. and A. Kakhidze 2004. Pichvnari 1: Results of Excavations Conducted by the Joint British-Georgian Pichvnari Expedition. Oxford: Batumi. Waldbaum, J.C. 1999. The coming of iron in the eastern Mediterranean: Thirty years of archaeological and technological work, in V. C. Pigott (ed.) The Archaeometallurgy of the Asian Old World. Philadelphia: University of Pennsylvania Museum. Yahalom-Mack, N. and A. Eliyahu-Behar 2015. The transition from bronze to iron in Canaan: Chronology, technology and context. Radiocarbon 57: 285-205.

32

The structure and function of ancient metrology John Neal Gaetano De Sanctis, considered by many to have been the most eminent Roman historian of the twentieth century, was reported by his student Livio Catullo Stecchini to have stated that ‘[ancient] Metrology is not a science, it is a nightmare.’ By contrast, it will be argued here that ancient metrology is elegant and all encompassing. The evidence for the structure of ancient metrology strongly suggests that all the measurement systems in use throughout the ancient world formed part of a unified system. This may at first sight appear to be a bold claim but it is justified since similarities far outweigh any differences.

Both fractions are most easily demonstrable in the structure of metrology with commonly known measures such as ‘Greek’ and ‘Roman’ that are universally acknowledged to be linked by the unit fraction 25 to 24. If we consider the variant of the ‘Roman’ foot given by Petrie at Stonehenge (Petrie 1880) it perfectly illustrates both the precision and permanence with which standards could be maintained in a structure, and the international nature of the modules. We now know that the absolute value of the ‘Roman’ foot that he deduced from the inner diameter of the sarsen circle was .9732096ft and Petrie measured the whole 100 Roman feet to within one twentieth of an inch of this value. He termed it a Pelasgo Roman foot and later termed the same value as Etrurian, the implication being that he had found evidence of the same foot length being used at these locations. Stecchini measured this value to within a millimetre overall as the length of the naos or cella of the Parthenon. The same foot value is given with extreme accuracy as the measured length of the running track of the Nemea stadium, given as 178 metres, equivalent to 600 feet (a stadion) of .9732096ft (296.63mm) to within three quarters of an inch overall (Miller 2001).

These similarities are clear when considered in the light of variations in individual modules. It is generally acknowledged that variations in the same module exist, but this is largely attributed to errors of copyists when making new rulers. This is, however, inherently unlikely since canonical rulers were presumably issued by the custodians of the temples where prototypes were maintained. If these were lost then the known dimensions of the temple itself would guarantee a faithful copy. Furthermore, the variations in the modules were universal and occurred in every ancient standard of mensuration.

At the recognized 25 to 24 ratio of this ‘Roman’ foot the ‘Greek’, or ‘Olympic,’ foot is exactly 1.01376ft (309mm) and is known at this precise length from a number of sources, not the least of which is Francis Penrose’s estimate of the width of the Parthenon stylobate, again with the tiniest margin of error, in this case within a quarter inch over the hundred feet. Although Penrose did not pursue the point, he remarked at that time that it had a geodetic appearance, noting that the 100ft stylobate equalled one second of a geographic degree. The meridian degree at the latitude of southern England is measured as 69.12 statute miles and this translates to the geographic foot at this latitude being 360,000 feet of 1.01376ft (308.99mm).

The eminent Egyptologist W. Flinders Petrie was aware of such variations and observed two that occurred repeatedly that were variations of the 450th and 170th part. (Petrie 1910-11, 482). Subsequent research has enabled these variations to be more exactly expressed as the 440th and the 175th part. The reasons for their existence lie in the need to maintain integers in both diameters and perimeters of circular designs, wheels, rotundas, storage vessels and surveying at geographic scales etc. Ancient sources, from cuneiform texts to Vitruvius state that if the diameter of a circle is a multiple of four units then 3.125 (25/8) is the ratio used for pi and the difference between this and the universally used 22/7 is the unit fraction 175 to 176. This means that an integer is maintained by using a module the 175th part longer on the perimeter. In all other cases seven is the number that defines the radius of a circle and the design module is found by dividing the radius by seven.1

These values, ‘Roman’ .9732096ft and ‘Greek’ 1.01376ft, among others, were both first identified with exactitude by John Michell in 1981 (Michell 1981). He also demonstrated that each of these values related as the unit fraction 176 to 175 with equally well-identified variants. Reduced by its 176th part the ‘Greek’ foot is

When archaeologists today analyze ancient measures, they tend to create averages of the variations as the definition of the units, but this destroys the integrity of the system to which they belong and obscures the fact that the variations in question were both practical and deliberate. It is important at the outset to realise that the national terminology, such as ‘Greek’, ‘Roman’, ‘Egyptian’, ‘Persian’,

1 

‘Belgic’ etc. that have been assigned to the various modules have little or nothing to do with the nations that adopted them; rather, they are but branches of a single organisation that was used concurrently in all nations.

33

Wonders Lost and Found 1.008ft and the ‘Roman’ .96768ft and these were close to the values reported by John Greaves for these feet (Greaves 1647).

Interestingly, the Root Reciprocal value of the Greek foot at .994318ft is the length of the Japanese shaku, the Korean cha and a Chinese chi measure of the Tang dynasty. When classifying the various ‘national’ feet the ‘Root’ value must be determined and this too, generally has a unit fractional link with the ‘English’ foot. Take for example the ‘Assyrian’ cubit as defined by Julius Oppert in a study of the foundation tablet of Assyrian Khorsabad (Oppert 1857-59). The cubit translates to 1.8ft whose constituent foot would be .9ft; the unit fraction 9:10 English. (Two-feet cubits were termed long cubits and one-and-a-half feet cubits were termed short cubits). This length is generally known as the ‘lesser foot’ and is found throughout the ancient world in all of the same variants as the above table if .9 is placed in the ‘Root’ position and decreased or increased by the fractions as shown. Because we are expressing the lengths in English feet then the above variations are also ratios when applied to other feet.

Although often encountered the former values are slightly longer than the definitive lengths most widely accepted in archaeological circles for these ‘Greek’ and ‘Roman’ feet. Stecchini identified the lesser values most closely; the ‘Greek’ foot as 308.276mm and the ‘Roman’ as 295.95mm (Stecchini. Web.), which are respectively 1.0114ft and .97095ft, values that are almost the unit fraction 440 to 441 of the longer measures deduced by Michell. Michell’s values prove the more soundly based, and we might correct Stecchini’s measurements accordingly. Both of these ‘Greek’ and ‘Roman’ feet as identified by Stecchini are also related by the same 175th fraction to lesser values of his examples than the values reported by Michell. Both of the lesser values were recorded on a white stone in the Forum, now in the Capitoline Museum. They were measured by Martin Folkes in 1736 (Folkes 1736-1737) and (converted from his inches) were for the Roman foot .965833ft and for the Greek foot 1.005833ft. These values may be very minimally adjusted to be absolute values; the Roman foot is .965487ft and the Greek 1.005714ft. These are very obviously the Roman .96 increased by the 175th part and when the Greek foot is similarly viewed it is seen that 1.005714 is simply the decimalised fraction of one plus the one hundred and seventy fifth part.

The explanation as to why these simple arithmetical interconnections have never previously been fully deciphered is because comparisons have been made at the wrong classifications and the straightforward unit fraction linkage thereby fragmented. These unit fraction connections are easily demonstrated by taking recognized values of modules to compare at their correct classifications. The link nine ‘Assyrian’ to ten ‘English’ is established. Also ascertained is the fact the ‘English’ foot is ‘Greek’ and the commonly accepted ‘Roman’ to ‘Greek’ ratio is twenty-four to twenty-five gives a ‘Roman’ foot of .96ft, a length that is very often found (e.g. at Woodhenge: Cunnington 1926). It is also generally recognised that the ‘Belgic’ foot stands at a ratio of nine to eight of the ‘Roman’ (according to the exchange rate given by Nero Claudius Drusus, governor of Gaul), and the resulting comparison looks like this:

By far the best medium for expressing ancient metrology is the English foot for the simple reason that it is demonstrably a variant of these ‘Greek’ ‘Olympic’ feet. Therefore the 441st part of the Greek foot less than 1.01376ft, would be 1.011461224ft and this number is (176/175)2 meaning that it is the English foot twice increased by its 175th part; this is how metrology was first seen to possess these absolute values.

Assyrian Roman Greek/English Belgic

When the range of values of a single module was first compiled (Neal, 2000) a certain distinguishing terminology had to be devised in order to classify the variants as shown in the table below. All of the values relate horizontally as 175 to 176 and the numbers vertically as 440 to 441, the English foot is in the ‘Root’ position. These exact variations apply to the feet of all nations.

.9ft 15 9 5 .96ft 16 24 1ft 10 25 1.08ft 6

(30.549cm)

9

As well as the ‘Belgic’ being nine to eight of the ‘Roman’ it is seen to be six to five of the ‘Assyrian’. This unit fraction linkage continues throughout the bureaucratic measurement standards adopted by all nations.

Root Reciprocal Root Root Canonical Root Geographic .994318 1 1.00571426 1.0114612 (30.307cm) (30.479cm) (30.654cm) (30.829cm) Standard Reciprocal Standard Standard Canonical Standard Geographic .996578 1.002272 1.008 1.01376 (30.376cm)

8

(30.724cm)

34

(30.899cm)

John Neal: The structure and function of ancient metrology

English feet Assyrian

0.9

Comm’n Assyria Iberian Roman Egypt 1 63/64 15/16 1 1/63 1 20/21 14/15 1 1/15 1 1/20 1 49/50

Iberian

0.91429

Roman

0.96

Comn Egypt

0.97959

Greek/English

1

1 1/9

Comn Greek

1.02857

1 1/7

Persian

1.05

1 1/6

Belgic (lesser)

1.07143

Sumerian

1.09714

Grk/ Eng

1 1/14 1 1/49 1 1 1/14 1 1/20

1.14286

8/9

5/6

24/25

14/15

7/8

48/49 1

20/21

1 1/14 1 1/24 1 1/7 1 1/9 1 1/4

1 1/6

1 1/7

4/5 6/7

20/21

14/15

48/49 1

24/25

9/10 15/16

49/50 1 1/49 1

7/8

6/7

9/10 9/10 27/28

1 1 1/27

1 1/9

Russian

6/7

1 1/15

English archaic 1.11111 Royal Egyptian

7/8

1 1/20 1 1/48 1 1/5

English Royal Sumer’n Archaic Egypt

9/10

35/36 1 1/35 1

1 1/24 1 1/48 1 1/8

Comm’n Greek Persian Belgic

1 1/15 1 1/24

15/16 24/25

1

35/36 1 1/35 1

20/21 48/49

Table 1

Nevertheless, a number of both ancient and more recent sources (Hero of Byzantium for example) state that the relationship of the royal cubit to the ‘Roman’ foot has a ratio of nine to five; thus the ‘Roman’ foot by this reckoning is related to the ‘royal Egyptian foot’ as five to six, and this can only can only happen if the ‘Roman’ foot is reduced by a factor of 1.008. This reduction happens with several of the ancient modules, they have a ‘lesser’ and ‘greater’ Root values. In particular, the ‘Persian’, ‘Belgic’ and common ‘Greek’ can change their ‘Root’ denomination.

Although there are nineteen separate mathematical feet, taking just twelve of them (or the table becomes too elongated to show on a single page) the unit fraction link at one of the correct classification values with the English foot as 1, would lookas shown in Table 1. The complexity is compounded by the fact that certain of these ‘Root’ values are not firmly fixed; they may be altered by a magnitude of 1.008 (the fraction 125/124), and the unit fraction link then changes to another unit fraction. For example, if the ‘Roman’ foot of .96ft is reduced from 24 to 25 of the English by a factor of 1.008 it then becomes 20 to 21 of the ‘English’; all of the unit fractions then budge along, similar to key or octave changes in music.

There are nineteen distinct ‘mathematical’ feet, as opposed to the anatomical (pes naturalis) between the minimum of the ‘Assyrian’ .9ft and the maximum of the ‘Russian’ 1.166ft. The English/Greek feet fall neatly toward the centre of the spread and are for this reason the most suitable medium with which to express the structure of metrology. Anything less than .9ft would be a half cubit of a longer measure, and these are the lengths of the ‘natural’ feet which are one seventh of the overall height of the six feet canonical man (cf. Leonardo da Vinci’s ‘Vitruvian Man’). One seventh of six English feet is the half royal Egyptian cubit of .857142ft. Thus the minimum .9ft parameter for a ‘mathematical’ foot length is established.

Another example is the ‘Persian’ foot of 1.05ft that is 21 to 20 of the English, (and which was the half royal cubit of Darius the Great [Skinner 1967]), and yet the value used in Persia as the exchange rate with the British measures during the occupation of 1941-45 was 25 inches for this two-feet cubit, which then translates to a foot 1.008 (126/125) less than the original, making it 25 to 24 of the British. This particular point is worth elaborating upon. It is established that the ‘Roman’ foot is twenty-four to twenty-five of the ‘Greek’; it is also the case that the ‘royal Egyptian’ foot is eight to seven of the ‘Greek’. Therefore, any of the Greek cubits increased by its seventh part is a royal Egyptian cubit. The most commonly encountered length of the royal cubit is 522.5mm, being the length of the cubit rod of Maya, the treasurer of Tutankhamun, in the Louvre, and the rod of Hormheb in the Museo delle Antichità Egizie in Turin; dozens more examples of this length may be cited. It translates to 12/7ft or 1.714285ft, being the English (and therefore Greek) cubit of 1½ft increased by its seventh part, and thus exactly 440 to 441 of the Great Pyramid cubit (523.7mm) precisely identified by Petrie.

At the other end of the scale anything greater than maximum foot of the ‘Russian’ 1.166666ft moves into the realm of longer modules. 1.2ft for example exceeds the length of a foot because it is the Roman 20 digit palimpes. Also if 1.2 is mistaken for a foot and multiplied by 1½ to determine its cubit, it is 1.8ft which is the two feet cubit of the ‘Assyrian’ .9ft foot, and reason would have it that a cubit cannot be a long cubit and a short cubit simultaneously. These factors define the rigid bounds or parameters of the mathematical foot values.2 With the exception of the common Greek foot, all the feet relate as six to seven of another, as the anatomical foot is one seventh of the overall height and of the arm span, as established by Leonardo’s corrected rendition of Vitruvian Man. This being so, then every man

2 

35

Wonders Lost and Found

Figure 1. A probable market trading standard from Tlemcen dated AD 1328. Said to be 47cm, its calculated value would be 47.026cm

It is important to make these distinctions of the type of foot, both for module recognition and to clarify magnitudes from ancient sources. A good example is seen in the dimensions of Old St Paul’s in London; it was recorded in 1313 as being 690ft long yet was measured in 1657 before the Great Fire as being 560 English feet long (Storer 1819). The earlier measurement was done in terms of a pes naturalis of the Belgic foot, (its half cubit). The use of a half cubit termed a foot persisted until relatively modern times, for example the ‘Welsh’ foot of .825ft which is the half Saxon cubit of 1.65ft; or the official foot length of Malta was the half royal Egyptian cubit at .857142ft. These anatomical feet are also termed twelve-digit hand spans.

ft cubit of this common Greek measure, given as 47cm on a marble trading standard from the Algerian city of Tlemcen. The metric system When the metric system is considered from the point of view of ancient metrology, the introduction of the Système international d’unités (SI) units has been both boon and bane. It was a boon because everyone was finally able to state a firm value for the length of any module. Previously, European city-states or regions used differing lengths for the ‘bureaucratic’ foot and they had to clarify the value of the foot that they were using in a transaction or comparison in terms of certain fixed values that they all agreed upon. By and large the measures considered stable were the ‘London foot’ the ‘Paris foot’ and the ‘Rhineland foot.’

There is nothing nationalistic in regarding the English foot as the ideal tool for metrological analysis. Since the English foot is just as old as the others and is quite central to the spread of foot values, whatever the module employed, from feet to miles etc., it is easier to recognize because the prefixing numbers are maintained. 185 metres is commonly given as the length of a Greek stadion for example; converted to feet this is 606.956ft or 600ft of 1.011593ft and may therefore be corrected to be ‘Root Geographic’ which makes it 184.976 metres.

Problems associated with this custom became apparent with the sudden instability of the French foot. This was brought about initially by the physical loss of their standard fathom measure known as the toise du Châtelet which had become damaged then disappeared altogether. In 1688 Jean-Baptiste Colbert in his capacity as Superintendent of Buildings introduced a reformed toise (fathom) shorter by almost five lignes (11mm) than the one that scholars and artisans had regarded as their measure, the one known as the toise de l’Écritoire, and which they believed was the original standard of Charlemagne.

It is immediately recognizable as a six hundred foot stadion because of the prefixing number ‘six’ – which is lost by being expressed in Système internationale (SI) units. Similarly, Eratosthenes’ stadion is given as seven hundred to the degree. If we therefore divide the 360,000 feet degree by seven hundred, we get 514.2857, a figure recognizable as a 500 feet stadion because of the prefixing ‘five’ but expressed as 156.754 metres it lacks this clarity. Each foot is of 1.028571ft, which is the Root value of the Common Greek foot at 36 to 35 of the English/Greek. This length was known throughout Europe as the ‘Rhineland’ or ‘Prussian’ foot and its yard is the Scottish ‘elwand.’ The illustration below is the 1½

This seemingly deliberate but unnecessary destabilization of the existing system, brought about by the reform, resulted in scientific confusion and later became the principal excuse for the Revolutionary Cause’s introduction of a totally different system of measurement, namely the SI system. The metre then became the touchstone by which all other values could be accurately expressed and this is its one and only advantage.

between five feet four and seven feet i.e. just about all of us, may be regarded as canonically six feet tall in one or another of the feet.

36

John Neal: The structure and function of ancient metrology

The disadvantages of the metric system as a tool to express ancient measurements are primarily that the module has no provision for a sub-multiple of three, although this is what the metre actually is – a yard from a previous system. It is based, in fact, on the ‘Belgic’ foot. Petrie unwittingly said as much in the course of an account of ancient measures (Petrie 1910-11). He stated that had the British adopted the ‘Belgic’ foot instead of the ‘English’ at the time of the statutes instituted by Edward I in 1305 (it was widely used in Britain at that time), ‘Then but little adjustment would have been required of us to conform to the metre’.

scientists had finally dug in against such ill-considered reforms for the simple reason that time and spheres must have a sexagesimal division. Since they could not divide the metre by three, the newly introduced metric system was utterly divorced from the systems that it superseded and this is why ancient measurement systems cannot be properly understood in the SI units of metres and centimetres. The metric system is a typical committee contrivance of compromise that took the best part of a century to devise and get utterly wrong. Systemic comparisons must be conducted at the foot level, and because the metric foot has been lost then the basic numbers have been lost. Older British metrologists similarly occluded the study of the field by their fondness for expressing even considerable distances in terms of the inch. This produces the same result, namely the destruction of the founding numbers, which should be expressed as feet. This term has little to do with the human appendage; in this context foot means basis. For the purposes of comparative metrology even large units such as miles and leagues etc. should be reduced to their basic feet and the comparisons made at that level.

This statement is the literal truth; consider that there are seventy-five Roman miles, therefore 375,000 Roman feet to the geographic degree. The ‘Belgic’ foot is the ratio of nine to eight of the Roman so there are 333,333.333 Belgic ‘Belgic’ feet and consequently 111,111.111 ‘Belgic’ yards, and this is the number of metres to the degree. The ‘Belgic’ foot was used as a geographic measurement in antiquity and is the basis of Ptolemy’s reckoning of 66 ⅔ miles to the degree. 360,000ft divided by this number gives the number of feet to the mile as 5,400 each foot of the Belgic 1.08ft to the 5,000ft mile. These are of course Root measures, giving the Root ‘metre’ as 3.24ft. It is more recognisable as a metre if the degree length is taken from the average degree at approx 45 degrees; the metre length is then close to the accepted 3.2808398ft..

How modules are deduced from ratios A few examples will show how the system was applied; this will illustrate both the universality and longevity of the system. First, from a wide choice of modules the architect selected modules to fit certain proportions to rational integers. Andrea Palladio stated there were ‘….. seven sets of the most beautiful and harmonious proportions to be used in the construction of rooms’ one of which was

It had this awkward number, 111111.111 to the degree because the French reformers failed to decimalise the degree or the number of hours of the day. They had attempted to do so by devising the grad, but the

Figure 2. Main bar of the George and Pilgrim Hotel, Glastonbury

37

Wonders Lost and Found the proportion three by four, a proportion that had been used from remote antiquity. It is evident in the pyramid of Kephren at the identical module value used by Palladio (the Vicentine foot, which is a variant of the ‘Russian’ foot, ‘Root Geographic Russian’ at 1.18ft or 359.67mm).

common Greek cubit has a six to five ratio with the royal Egyptian short cubit and this ratio is universal, generally used as the fixed proportion for the builder’s cubit to the mercantile cubit in all nations, and the same ratio obtains between the lesser Roman foot and the royal Egyptian.

The module of construction invariably stems from the proportion, in this case the vast pyramid is two backto-back Pythagorean 3,4,5 triangles, then in terms of the Russian foot it is 300 to half base, 400 height and 500 apothem, and in terms of the 100ft ‘plethron’ it is 3,4,5. The ‘Russian’ foot is also the basis of the Arabic ‘black cubit’ that had been used in Egypt by Al Fargani to rebuild the Nilometer on Rawda Island in Cairo and Schwaller gave many examples of its use in ancient Egypt; but archaeologists are largely ignorant of the structure of metrology and constantly try to make everything in Egypt conform to dimensions expressed in the royal cubit alone.

Secondly, the longevity of the units is apparent most obviously in the prehistoric ruins of Stonehenge where each of the above units is fundamental to the design employing the identical variants of the mediaeval modules. The outer diameter of the sarsen ring employs the same common Greek foot module as used by the medieval designer of the George and Pilgrim. Without having to resort to prehistoric-medieval comparisons, Stonehenge as a single monument displays the rigidly maintained continuity from the origins of this highly refined system of mensuration. As can be seen from three rather obvious features of the monument’s geometry, the inner lintels of the sarsen ring are one hundred ‘Roman’ feet, the outer is one hundred common ‘Greek’ feet and the width of the station stone rectangle is one hundred lesser ‘Belgic’ feet, dimensions obviously quite deliberately contrived. Each of the values would give the diameters as a unit, in these cases one plethron. But the station stone rectangle is just within the diameter of the Aubrey hole circle, dating from the late Fourth Millennium BC, while the lintel circle dates from the mid-Second Millennium BC. This most clearly indicates a developed culture that had the scientific ability to maintain standards for millennia in order to make continuous refinements on the older foundations while adhering to the same design principles.

Closer to home, this same ‘Palladian’ proportion was used in the construction of the main bar of the George and Pilgrim Hotel, Glastonbury (Figure 2), built in the fifteenth century. It measures twenty common Greek feet by fifteen; this is the enclosed area, centre to centre of the framing beams. The module is therefore a five feet pace of these feet; the integers five by three paces are the same number as the design proportion. The common Greek foot was widely applied in mediaeval English architecture alongside many others and the particular module would be selected to integrally fit the available space. Then, what happens when the length of the room is divided by three, which is the distance between the cross beams, is that there is a module change to maintain an integer; the distance between the beams is one royal ‘Egyptian’ six feet fathom. These modules, the ‘royal Egyptian’ and the ‘common Greek’, are often found together because they are a harmonic. The long

Once again, the illustration below exemplifies how the module invariably fits the proportion as shown by the rectangle. Stonehenge is patent in this respect inasmuch there are thirty lintels on thirty upright sarsen stones and the lintel width is one thirtieth of

Figure 3. Station stone rectangle based upon Petrie’s inner sarsen measurement. Note that there are three plethra lengths in one design

38

John Neal: The structure and function of ancient metrology

the diameter. This length at 3.4757485ft is two royal Egyptian cubits.3 This cubit is known to have a five to six relationship to the so-called ‘sacred cubit’ of the Jews (Newton, in Greaves 1737), and Newton defined this sacred cubit by a series of eliminations to within very good parameters. Because this Jewish cubit is also two common Greek feet there are fifty of these sacred cubits in the outer lintel diameter.4

who had first inspired him with an interest in the pyramids of Egypt. Petrie’s own measurements of the monuments enabled him demolish the fanciful theories that were being circulated, and ever after he was hostile to the possibility of any high science in antiquity, thereby taking the romance and life out of metrology. However, it is worth remarking that in the introduction to his Pyramids and Temples of Gizeh Petrie reports the observation of a visiting American to the site who expressed his disappointment that his fantasies had been shattered by accurate measurements, observing ‘Well, sir! I feel as if I had been to a funeral!’ Petrie noted ‘By all means let the old theories have a decent burial; though we should take care that in our haste none of the wounded ones are buried alive.’ This suggests that Petrie did not altogether abandon his deep respect for the ancient people or their yet-to-be-discovered accomplishments.

The distinguishing feature of the above cubit is that it is geodetic; it is the ten millionth of the polar radius. This is the cubit identified by Berriman and which he termed ‘Cubit A’ and thereby linked it to the metre by the formula (2/π) metre (there being ten million metres in the quadrant arc) (Berriman 1953, 19). It is worth remarking that virtually all researchers into metrology reach the conclusion that the ancient measurements seem to have a geodetic basis.

Grounded in Darwinism, an increasingly secularized archaeological community supported him in his newly adopted scepticism; earlier beliefs in the importance of metrology as a cultural relic of older civilizations were sufficient to bring the study of the subject virtually to a halt in the twentieth century. Before this, scientists of every persuasion had taken it for granted that our own civilization stemmed from a greater, now forgotten, culture and regarded their own knowledge as the vestiges of the higher wisdom of their ancestors. Newton was obsessively of this opinion and dissected every stanza of the Gematria in the New Testament and the Kabbalah of the Old in order to reconstruct the dimensions of the Temple of Jerusalem that he openly stated was a model (fractal fashion) of the earth. He did this in an attempt to identify the ancient modules of measurement which he knew were geodetic in order to complete the Principia, which could not be done without an accurate knowledge of the earth’s dimensions.

The necessity of an ancient survey This assumption will be dealt with at an early stage. In order for this to be the case, a scientifically accurate survey of the earth must have been conducted in remote antiquity. This is not in fact as fanciful as it may at first appear. The French surveyors who first accomplished the modern survey upon which the metre was founded were subject to transport conditions little different from those that applied in the Neolithic, and the equivalents of the instruments they used were quite capable of being created at that remote time. Lenses with the ability to produce any kind of magnification are found, literally by the thousand, in ancient barrows and archaeological sites throughout human history (Temple 2001). Put a telescope upon a simple dioptra and it is a theodolite, and Pythagorean triangulation was practised throughout the ancient world. Fully accurate surveys are known to have been conducted in our own era without modern technology such as satellites or lasers.

He was on the right track because the structure of metrology is sufficient to establish the statistics of the ancient reckoning of the earth’s dimensions; which proves to be as accurate as, and a far more rational model than, the one designed by French savants in the eighteenth century. The evidence to support these statements is voluminous and unambiguous. The fact that this impeccable evidence as to the dimensions of the world was passed on to us by people who had lost the ability actually to make the measurements for themselves implies that we are dealing with inherited traditions.

It is conventional today to regard people in the distant past as unscientific, and this has been largely brought about in the attitude personified by Flinders Petrie. There is a certain irony in this because Petrie was one of the most prolific authors on the subject of ancient metrology, and the whole understanding of the subject relied to great extent on his input into the database. The fact of the matter is that Petrie had been totally sucked into the pyramidiocy fads of the late nineteenth century through his involvement with Piazzi Smyth,

Metrology and the Earth’s dimensions

Stone 1924 gives the lintel width as 3.5ft; to the author’s knowledge no other surveyor mentions this particular and vital statistic. 4  Although many different modules may befound in any monument, they will all be of the same classification value. For example the modules in Stonehenge are all of the Standard Geographic classification (Root x 1.01376); the modules in the Great Pyramid are all of the Standard classification (Root x 441/440). 3 

It rather goes against the grain to hold up one compound measurement as more important than another, but we may start with one that has already been dealt with here as highly significant. It is the sacred cubit or Cubit 39

Wonders Lost and Found

Figure 4. Geoid radius lengths and their ratios from the structure of metrology Polar 63,356.45kmmean 6,370.9kmequatorial 6,378.12km

A inaccurately defined by Berriman, or better still the Standard Geographic common ‘Greek’ long cubit. The length is calculated thus: one common ‘Greek’ foot at 36/35 = 1.028571 multiplied by 2,5 then multiplied by the Standard Geographic formula of 1.01376, and this equals 2.0854491ft, or one ten millionth of the polar radius. The Polar radius of 20,854,491ft multiplied by 441/440 equals the Mean radius of 20,901,888ft and the Equatorial radius is then half the fraction between the Polar and Mean – longer than the Mean. These three earth radii are then the relative series 880 to 882 to 883. Thus the complex geoid is rendered into the simplest of terms.

Take the Polar Radius of 20,854,491ft, and by using the 22/7-pi ratio, the degree length of its meridian may be computed as 360,000 geographic feet of 1.011461224ft (308.27mm), which is the most commonly quoted value of the Greek foot. This degree is 3/7th of the distance equator to pole. Near its centre (within five miles) is Delphi which is the home of Greek geographic lore; this in as much as the omphalos of Delphi marks the point where the eagles released by Zeus to compute the size of the earth flew around the world in opposite directions and met back up. This foot value as a formula is common to all measurement systems, the Assyrian .9ft for example when multiplied by this number equals .910315ft and would fit this degree 400,000 times. It is therefore the ‘geometric foot’ proposed by Alexis-Jean Pierre Paucton to be a more rational measurement than the metre long before the latter’s adoption (Paucton 1780; a definitive work of its time).

It is sufficiently scientifically accurate, giving the Polar Radius as 6,356.45km, the Mean as 6,370.89km, and the Equatorial as 6,378.12km and these are the absolutes. Immediately we have an advantage over the modern system because whatever modern reference ellipsoid one consults gives a slightly different result for these simple dimensions; this means that there is no common agreement on the exact size of our spheroid; although seemingly good reasons are given for these differences, it remains an unsatisfactory fudge. The modern system of geographic notation, the metre, attempts to average all of the different latitudinal degrees of the meridian. Averages often give nonsensical results such as the average human having less than two legs, but the ancient system of mensuration takes account of these considerable latitudinal differences in its basic structure through the variations of certain of the foot lengths that have been tabulated.

Then if the same multiplication is used on the Earth’s Mean Radius of 20,901,888ft, the 440th part longer than the Polar, then the degree length is 364,953.6ft = 360,000ft of 1.01376ft. It is known at this precise length as the ‘Olympic foot’, and if this degree is divided by any of the known modules that are associated with degree lengths, fully engaging facts emerge. We know from Eratosthenes, for example, that there were reckoned to be 210,000 royal Egyptian cubits to the degree. If we therefore divide 364,953.6 by 210,000, we get 1.73787ft. This number is 12/7 or 1.714285ft, and when multiplied by 1.01376 it is the Standard Geographic royal Egyptian cubit. This is 52.97cm and Schwaller records two cubit rods from the tomb of Tutankhamen of 53cm.

Skinner noticed that the sacred Jewish cubit was two common Greek feet.

The beauty of this is that all of the modules that fit this degree integrally are then integral to the Polar Radius.

5 

40

John Neal: The structure and function of ancient metrology

Ancient and more recent estimates of the earth’s dimensions

The cubit of 1.73787ft fits the Polar Radius 12,000,000 times. And conversely the sacred cubit of the Polar Radius then fits this Mean degree 175,000 times. This strange coincidence of correspondence happens because these measures are using two radii separated by the fraction 440 to 441and number itself distorts by this identical ratio. Simplified thus:

Ancient estimates of the length of the terrestrial degree go back as far as the Old Kingdom in Egypt but it is best to approach the subject from the surge of interest in such matters that was provoked by the cultured Abbasids who had governed the Caliphates from Baghdad. The fifth caliph, Harun al Rashid had begun the quest for scholarship that was continued by his son Al Mamun who was born in AD 786 and became caliph in 813. It was under his patronage that embassies were set up in India where Arabic scholars learned to use the numeral system now referred to as Arabic. Muhammed Al Fazari through his translation of the work by Brahmagupta titled Correctly Established Law of Brahma that contained the axioms of decimal notation, is how our numerical system came into Arabia – the astrolabe was introduced from the same source. According to Al Biruni, Al Fazari had noted the Indian correspondence of the earth’s circumference expressed by the Indians as 3,300 ‘Yojanas’ with that of the traditional Arabic reckoning of 6,600 farsakh. There are as many differing yojanas as there are farsakhs so all one can do with such information is to divide these numbers into 129,600,000 or the number of geographic feet in the perimeter. In the case of the 6,600 farsakh it leaves one with a number that is 19636.3636; by its prefixing number this can be divided by either the sexagesimal 1,800 or decimal 2,000 in order to see what module is revealed.

As can be seen in the above diagram a canonical perimeter number of 360, if regarded as English /Greek feet then the diameter is exactly one hundred royal Egyptian feet at the exact length identified by Petrie, and for this reason it is the most widely accepted value of the Egyptian foot at 39.4cm. However, the royal Egyptian foot as related to the English is an eight to seven ratio, which would make the above diameter 114.2857, but it is given that extra 440th fraction, and then the proportions of the earth knock it off again leaving the Mean degree and Polar Radius numerically compatible. Millions of people must have seen the above equation because circles are always 360, all that is lacking is number recognition. The present writer has long since ceased to wonder if it is coincidence or design; it is simply a very beautiful fact that the earth distorts by the identical ratio inherent in number and geometry.

By the latter it would yield the common Egyptian foot of .981818ft and the former the Sumerian foot of 1.090909ft. The other approach is to divide the 6600 perimeter by 360 to find the number of farsakh to the degree and it equals 18⅓, it is accepted that one of the

Figure 5. Diagram showing the increases in the geographic feet towards the north and the fractions by which they increase

41

Wonders Lost and Found solutions to the length of a farsakh is that it is three miles. When therefore the 360,000 feet of the degree is divided by 55 (as the three miles) it equals 6,545.4545, and this is clearly 6,000 feet or a mile of the Sumerian foot of 1.090909ft. This is the obvious solution because the various Arabic miles are generally given as 4,000 cubits, therefore 6,000 feet. This is an appropriate unit for this Baghdad area since the ziggurat of Etemenanki is known from cuneiform texts to have dimensions of 180 Sumerian feet sides for the core pyramid and 180 Sumerian cubits for the outer shell. Measurements of the monument revealed the use of a variant of this Sumerian foot at the Standard Geographic value, 1.112238ft, (1.090909ft x 1.01376) or 33.9cm (Koldewey 1914).

known unit that would vary by the degree classification. If 360,000 or the Root value is used then it will give a Root solution. Take the 56 miles of Habash’s degree; it will be 4000 cubits whose constituent foot will be 1.07142ft or the lesser Belgic Root, therefore a feasible solution. Similarly the degree of 360,000 divided by 56¼ miles is 6400, which has a foot value of 1.066666ft, which is the Persepolitan foot (Petrie identified its cubit at exactly 19.2 inches). These feet at these classifications relate as 224 to 225 as does 56 to 56¼. Modern commentators who shed doubt upon the validity of any kind of survey conducted by the Arabs, concede that whenever correct solutions are offered for lengths of the degree, it is invariably referenced to pre-Islamic customary usage. Throughout Europe and Arabia there had been a continuous reckoning of seventy-five Roman miles to the degree and this is exactly right, in spite of the fact that the Romans too were incapable of measuring a degree. It was an inherited tradition. The same could be said of the earlier Greeks who wrote at length on the dimensions of the earth. Ptolemy in his Geographia gives a different length for the number of stadia to the degree than he does in the Almagest. We most often acknowledge the value of 700 stadia for this length, largely because this was also the value used by Eratosthenes. But it is when, as in the Almagest, he claims as few as 500 stadia for this length that it arouses curiosity.

Al Mamun harboured a great desire to understand the works of Ptolemy and other Greek writers concerning the lengths of their miles, stadia and the geographic degree. Often cited in Arab literature, he is reputed to have financed and instructed a group of scientists to measure the length of the terrestrial degree in a flat area of desert; some say Palmyra in Syria, some the area of Baghdad and Mosul in Iraq. There are varying stories as the methods employed and to the results; the most common of these accounts states that the length of the degree was fifty-six and two third miles, each mile of 4,000 ‘black cubits’ and there is compelling evidence as to the value of the black cubit. The length is exactly 1.77408ft (54.074cm) and this is the Standard Geographic value of the ‘Russian’ short cubit: 1.75ft multiplied by 1.01376, most exactly preserved on the Rawda nilometer. This nilometer had been rebuilt by Al Farghani for Al Mamun’s successor in AD 861 using the cubit that had been adopted by Al Mamun as the national standard. Al Farghani was one of the scientists who gave the length of the degree as 56⅔ miles and this would make the degree over 400,000 feet in length, an error approaching twelve per cent. The modern scholars who were consulted for the following data have expressed grave doubts that this degree measurement had ever been undertaken (Mercier 1992; King 2000). 56⅔ miles is an extremely improbable degree division whatever the basic cubit.

360,000 divided by 500 equals 720 feet which is rather long to term a ‘stadion’; divide this sexagesimal number by its closest decimal and we get 750, which equals .96 or the greater Root value of the Roman foot. The length of Ptolemy’s stadion is therefore the tenth of the Roman league of 1½ miles in terms of which the Roman Empire was mapped in the Tabula Peutingeriana. The point here is that we should not be sidelined by terminology; many module terms are used to apply to significantly different lengths. Mid-range values in particular such as farsakh, league, parasang, schoinus and iteru have no single identity. One simply works with the given numbers and checks if they are rational. The bullseye is hit with such frequency that one begins to realize that this is exactly how the system was designed to be used.

One may accept or reject solutions by a simple process, other lengths quoted for the degree examination as ordered by Al Mamum are 56 miles quoted by both Habash and Al Biruni; ibn Yunus gave 57 miles but quoted Habash as also recording the degree at 56¼ miles.6 57 miles cannot be a sensible reckoning either because 360,000 divided by this number must give a

From Hellenistic sources there were five principle solutions given for the earth’s circumference. Posidonius gave the same solution as Ptolemy’s longer stadion at 180,000 for the entire meridian. Posidonius also proposed 240,000 stadia for this dimension, Eratosthenes gave both 250,000 and 252,000, and Aristotle 400,000. Perfect sense can be made of all of them apart from the value proposed by Archimedes of 300,000 stadia. If the circumference of 129,600,000 geographic feet is divided by Archimedes’ 300,000 one ends up with a length of 432ft. The only connection

Ahmad ibn ‚Abdallah Habash Hasib Marwazi, died a centenarian in 869. The Book of Bodies and Distances Abu al Rayhan Muhammed ibn Ahmad al-Biruni, 973-1048, proposed a way of measuring the earth’s radius. Abu’l-Hasan ‚Ali Ibn ‚Abd al-Rahman Ibn Ahmad Ibn Yunus alSadafi, Astronomer, ca 950 to 1009.

6 

42

John Neal: The structure and function of ancient metrology

with this length as a stadion is that it may be viewed as a 500 feet length of the natural foot which at .864ft is the half royal Egyptian cubit of 1.728ft.

and they would obviously get rather good at it and know the extent of their land intimately. Inscriptions from the Old Kingdom clearly state the dimensions of Egypt in terms of the ituru (a word meaning ‘river’). The stadia measures used by Eratosthenes are compatible with the ituru and by simple comparisons assisted by information gathered by Gyula Priskin, almost entirely consulting Strabo’s Geographia, who in turn was quoting Eratosthenes as to the distances involved, we may exactly define the units of measurement and the dimensions of Egypt (Priskin 2006).

This precise length for the royal cubit is that of the wooden cubit rod of Kha in the Turin Museum; at double this length it is also the width of no fewer than nineteen passageways in various Pyramids (Lepsius 1865, 7). Given by Lepsius as three English feet, five and a half inches, this an accuracy of better than 99.9 per cent about a third of a millimetre. It is exactly 175 to 176 of the Stonehenge double cubit of the lintel stones, and 176 to 175 of Petrie’s cubit 1.718ft this however is perhaps a little unwieldy as a solution to Archimedes’ statement.

Eratosthenes first stated that Alexandria and Syene were separated by 7.2 degrees and this amounted to 5,000 stadia, he was speaking in latitudinal difference, not the literal straight line between these points. This is obviously one fiftieth of the circumference thereby implying a total circumference of 250,000 stadia. But all Eratosthenes had actually done was to use the rounded number of 5,000 by using a module 1.008 times longer than the Root value that the other modules are expressed in; it would have been 5,040 stadia using the Root classification of the module. The lack of understanding by – and about – the Greeks concerning Eratosthenes’ reckoning, is displayed by the general belief that he had changed the total perimeter number to be 252,000 stadia (e.g. Evans, 1998) in order for it to be compatible with a sexagesimal count of the sixtieth and three hundred and sixtieth parts, but it is not quite like that.

Aristotle’s 400,000 stadia is far more plausible as a canonical solution because it has an obvious numerical affinity with the metre-‘Belgic’ yard. The meridian perimeter divided by 400,000 yields a half stadion of 324ft which is 300 ‘Belgic’ 1.08ft. The degree contains 1,111.111 of them, which is the equivalent of 100 metres to the unit. This is a solution that is quite often encountered. Ptolemy used 66⅔ miles for the length of the degree. This is a 5,000 feet mile, with each foot of 1.08ft. Posidonius’ 240,000 stadion circumference solution is also a ‘Belgic’ measure. In this case they would be 540 feet stadia, or 500 feet of 1.08ft. The metre has indeed been used since antiquity but it was not a fixed singular length and was divisible by three.

There is unanimous agreement that Eratosthenes’ stadia were 700 to the degree, and there is also wide agreement that when writers such as Strabo quoted the overall dimensions of Egypt south to north that it was a literal straight line from the boundary of Egypt at 24° to the northernmost part on the ‘Egyptian Sea’ at 31.5°, then this seven and a half degrees times 700 must be 5,250 stadia. Yet Strabo, more than once, states that this distance is 5,300 stadia (Strab. 17.1-4). This single statement is the clue that provides the solution to the problem of the metrological division of Egypt by geographic degrees.

One often encounters the half 600ft stadion also called a stadion. It was used by Herodotus to give distances of Egypt; a close examination of the numbers of ‘stadia’ quoted by him for the distances between places in Egypt reveal that the module he referred to was a length of 300 royal Egyptian feet of 1.142857ft or 342.85714ft (104.5m). It has the exact ratio of two to three of Eratosthenes’ stadion; they are also respectively 200 and 300 royal Egyptian cubits of 1.71428ft. The scholarly consensus regarding the Arab’s and the Greek’s attempts to rationalize the dimensions of the world is that both were incapable of conducting a sufficiently accurate survey to reach their often accurate conclusions. In all cases, however, the correct solutions seem to have been part of their inherited traditions.

Taking the length of the degree as being 360,000 feet, then Eratosthenes’ stadion is 514.2857ft, or the ‘English’ 500ft stadion plus its 35th part, resulting in 500 common ‘Greek; feet. In simple number terms this gives 5,250 as the stadia length of Egypt. Yet the literal number of stadia in the overall length is stated to be 5,300 and it is the difference between these lengths that reveal the geographic foot of Egypt. The difference is the unit fraction 105 to 106 and if this is expressed as a decimal we get 1.00952381ft, which is the length of the geographic foot of Egypt.

By far the most credible and often quoted of the ancient geographers is Eratosthenes. In his capacity as librarian of the Library of Alexandria it must assumed that he had an enormous amount of geodetic information to hand, and he was in a perfect environment to witness techniques of surveying. Due to the annual inundation of the land during the Nile flood, vast tracts of Egypt had to be accurately surveyed on an annual basis in order to re-establish land boundaries. This had been done since time immemorial,

The degree calculated from this geographic foot is 363,428.5714ft (110,773m) and would be centred at 2/7ths of the distance equator to pole, which is the site 43

Wonders Lost and Found of the Temple of Amun, at the centre of the Thebaid, or Upper Egypt. But this is not the only way to arrive at this position.

Now we must make sense of the specific statements, from papyri at Tanis (Petrie), cubit rod inscriptions (Borchardt) temple inscriptions at Aswan (Priskin) and from the White Chapel of Sesostris – that the overall length of Egypt is 106 iteru of 20,000 royal cubits. The division of 360,000 feet of the degree by 20,000 cubits is 10½ (already it is pointing to 105). By the same reckoning as before, 106 of the longer iteru are exactly ten degrees and this would take the overall length of Egypt a further two and a half degrees south of the twenty-four degree southern boundary of the two kingdoms of Upper and Lower Egypt.

To quote Priskin: ‘Eratosthenes took over an Egyptian tradition of converting different units of measurement according to which one iteru equalled fifty stadia.’ The iteru has more than one length; it is variously given as 15,000 royal cubits and 20,000 royal cubits (a little under five English miles and nearly six and a half miles respectively). Many metrologists accept the 15,000 cubit iteru on logical grounds; others insist on the 20,000 cubit version because some inscriptions state it to be so. The earliest of these date to the early second millennium BC on a building block from a temple of Amenemhet I; while others from the White Chapel of Sesostris I state the 106 iteru length of Egypt and define the iteru as 20,000 cubits (Schlott-Schwab 1981).

At 21.5°, Abu Simbel at 22 ¼° N is within this southern degree and attests to an enormous Egyptian presence in this area, particularly as a buffer zone against the Meriotic people. The Ptolemaic Greeks knew these southern lands beyond the borders of Egypt proper as the Dodekaschoinos and Triakontaschoinos (Bevan 1927). The Ptolemaic record refers to the twelve and thirty schoinoi of these southern lands as twelve and thirty Ar which is a form of atur or iteru. This interpretation of the ancient records is confirmed persuasively by the placement of the boundary stele of Sesostris (Figure 6) exactly at 21½ degrees.

Because the word iteru means river, and because the length of 106 iteru of 20,000 cubits far exceeds the dimension of Upper and Lower Egypt the iteru was believed to be derived from the length of the meandering course of the Nile. Absurd. As laconic as it is, we have enough information to exactly resolve this length and the exact dimensions of Egypt with it. Stecchini took for granted that the iteru, which he termed atur, was 15,000 cubits and this agrees with Priskin’s observation that: ‘Eratosthenes continued the Egyptian tradition of the reckoning of the iteru as fifty stadia’. This would be his stadion of 514.285ft, fifty of which is exactly 15,000 cubits of 1.714285ft Stecchini also mentioned inscribed cubit rods that had been described by Ludwig Borchardt that stated the dimensions as 106 iteru and broke the distance down from Bedhet at the northernmost point of Egypt to a Nile island called PiHapy as twenty iteru and from Pi-Hapy to Syene on the southern border as eighty six iteru. We can define this length exactly as we did with Eratosthenes’ stadia. Take the geographic degree number of 360,000, divide it by 12/7ft (Root cubit), and we get 210,000. Divide this by the iteru of 15,000 cubits and it is exactly fourteen iteru to the degree. Multiply this by the number of degrees, which is seven and a half, and we get exactly 105. Therefore, the length of Upper and Lower Egypt is the 106 iteru of the Egyptian reports in terms of the Root cubit, but is geographically 105, only in terms of the Root measure it is one hundred and six. 106 to 105 once again gives us the same geographic foot that we deduced from the record of Strabo. This means that in matters of module identification ancient people, in order to simplify the equations, thought in the Root numbers in exactly the same way as the present writer has learned do in order to make things clear.

Figure 6. Boundary stele of Sesostris from Semna located at 21.5º N. (Altes Museum, Berlin)

44

John Neal: The structure and function of ancient metrology

From the inscription on the boundary stele it would indicate that it was he who established this southern boundary c. 1850 BC; it begins:

iteru measurements for Egypt proper and the extended Egypt. Thus, Upper Egypt, the Two Kingdoms of Egypt and extended Egypt were respectively measured as, twelve, fifteen and twenty differing iteru composed of the same submultiple; the royal cubit, that they regarded as sacred.

Year 16, third month of winter: the king made his southern boundary at Heh. I have made my boundary further south than my fathers. I have added to what was bequeathed me.

This identical schoinos measurement was also used in Europe; as late as 1568 King Philip I of Spain issued a proclamation adopting the vara of Castile as the national Standard, and in the same proclamation he defined the degree length as seventeen and a half maritime leagues, each league to be 7,500 varas of Castile. Since the vara was 2.74285ft – 83.6cm, when multiplied by 7,500 it comes to 20,571.428ft, which is the same as the 12,000 royal cubit schoinos. This illustrates that the Spanish scientists also used the Root numbers to simplify calculation because 17.5 of these leagues equals 360,000 feet. They could maintain the correct solution for degree lengths at different latitudes by using the longer, well recorded variants of the vara, that accurately differ by the same fractions/ratios given in Table 1.

What is truly remarkable concerning these Egyptian solutions is their level of accuracy. In 1907 F.R. Helmert completed his geodetic survey, and if his accurate degree definitions are applied to all of these degrees along the course of the Nile; added together for the seven and a half degrees of Upper and Lower Egypt then the 106 iteru length by using the 12/7ft royal cubit, there is only about 600ft of shortfall. But if one adds the shorter degrees to the south by using the longer 20,000 cubit 106 iteru then the solution is virtually exact for the overall ten degrees, each being of the 363,428.714ft degree of the Thebaid – 110,773m. Could the Egyptians have measured to this degree of accuracy? On the above evidence it seems probable, but this is far from all of the evidence and taken all together it is a certainty that they could.

It must be pointed out that the league length of 20,571.42ft is also 20,000 common Greek feet. It was recognised as such by English navigators of the sixteenth century. Edward Wright stated in 1599: the meridian circumference is reckoned to be 6,300 Spanish leagues each league composed of 20,000 feet (Wright 1599) 6-7. The common Greek foot was also the half sacred Jewish cubit, the Leyden foot, the Rhineland foot, the third of the Scottish elwand and is the foot of Eratosthenes’ stadion – exactly.

Consider that before the arrival of the Greeks the Egyptians had three millennia of continuous traditions and their statements concerning their metrology are uttered with more aplomb than the tentative Greeks. The Greeks may have continued certain customary usages practiced by the Egyptians but the Egyptian geomancers, probably because of the esoteric nature of the scribal guild, seem not to have communicated all that they knew to them.

What becomes apparent in the course of the study of ancient metrology is that its purpose was to express everything as an integer. This is most apparent where the recorders have a choice; that is to select a module to rationally and integrally fit a fixed length. The Pharos of Alexandria is one example; the islet upon which it was built was connected to the mainland by a causeway known as the Heptastadion. One figure we have for its length is 1,260 metres (McKenzie 2007) converted to feet it becomes 4,133.86ft. Divide by seven for the stadion length and we get 590.55ft. The closest number being 600, divided by that number produces .98425ft. The closest identification of this foot would be a Root Canonical value of the common Egyptian foot of .98518ft.

Thus, it has been demonstrated that the lengths, terminology, textual and physical evidence, constitute proof of extremely ancient knowledge of the accurate dimensions of the world and the sophistication of the directly related measurements with which to express these data on a practical level. We know that throughout the ancient world, for bureaucracies of any nation or state it was of primary importance to establish, exactly, the boundaries of their governance. It was this criterion that ensured that techniques of surveying and accurate measurements were rigorously observed and maintained. The identical system, techniques and lengths used elsewhere

Another example is the length of the Siloam Tunnel in Jerusalem (also known as Hezekiah’s Tunnel, ca. 700 BC) which was a fixed length connecting the Gihon spring to the Siloam Pool. A contemporary inscription states that the length of the tunnel is 1,200 cubits. C.R. Conder reported this length as 537.6 metres (Conder 1882); converted to feet and divided by 1,800 to identify the foot length, we get the Root common ‘Egyptian’

The schoinos was an Egyptian measure adopted by the Greeks, this was sometimes called iteru and was twelve thousand royal cubits, or forty stadia. And if these schoinoi are used to measure the six-degree length of the Thebaid, we get the same solution of 105 geographic schoinoi and 106 Root schoinoi, exactly as do the two 45

Wonders Lost and Found Bibliography

foot of .979591ft that is forty-eight to forty-nine of the ‘English’. This particular foot is 175 to 176 of the common Egyptian foot of the Heptastadion. These lengths were selected from a wide range of modules after the measurement had been taken, that is they were not necessarily measured in these modules but the module was selected afterwards to be a rational number. Much as the French had measured the meridian with the toise and it was afterwards expressed in metres.

Berriman, A.E. 1953. Historical metrology; a new analysis of the archaeological and the historical evidence relating to weights and measure. London. Bevan, E.R. 1927. A History of Egypt under the Ptolemaic dynasty. London. Conder, C.R. 1882. The Siloam Tunnel. Palestine Exploration Fund Quarterly Statement 14: 122-31. Cunnington, M.E.P. 1929. Woodhenge: a description of the site as revealed by excavations carried out there by Mr. and Mrs. B.H. Cunnington, 1926-7-8. Devizes. Evans, J. 1998. The History and Practice of Ancient Astronomy. New York. Folkes, M. 1735-1736. An Account of the Standard Measures Preserved in the Capitol at Rome. Philosophical Transactions 39: 262-6. Greaves, J. 1647. A Discourse of the Roman Foot and Denarius from whence, as from two principles, the measures and weights used by the ancients may be deduced. London. Greaves, J. 1737. Miscellaneous Works many of which are now first published ... : to the whole is prefix’d, An historical and critical account of the life and writings of the author. London. King, D. 2000. Too Many Cooks … A New Account of the Earliest Muslim Geodetic Measurements. Suhayl: International Journal for the History of the Exact and Natural Sciences in Islamic Civilisation 1: 207–241. Koldewey, R. 1914. The Excavations at Babylon. London. Lepsius, R. 2000. The Ancient Egyptian Cubit and its Subdivisions. London [1865]. McKenzie, J. 2007. The Architecture of Alexandria and Egypt 300 B.C.-A.D. 700. New Haven and London. Mercier, R.P. 1992. Geodesy, in B. Harley and D. Woodward (eds) The History of Cartography 2:1, Cartography in the Traditional Islamic and South Asian Societies: 175–188. Chicago. Michell, J. 1981. Ancient Metrology, Stonehenge and the Whole World as Therein Symbolised. Bristol. Miller, S.G. 2001. Nemea II: The Early Hellenistic Stadium. Berkeley and Los Angeles. Neal, J. 2000. All done with Mirrors: an exploration of measure, proportion, ratio and number: opus 2. London. Noback, C. and F.E. Noback 1851. Vollständiges Taschenbuch der Münz , Maß- und Gewichtsverhältnisse. Leipzig. Oppert, J. 1857-59. Expedition Scientifique en Mesopotamie. Paris. Palladio, A. 1570. I Quattro Libri dell’Architettura. Venice. Paucton, A.-J. P. 1780. Métrologie, ou, Traité des mesures, poids et monnoies des anciens peuples & des modernes. Paris. Penrose, F.C. 1888. An investigation of the principles of Athenian architecture:  or, The results of a survey conducted chiefly with reference to the optical refinements exhibited in the construction of the ancient buildings at Athens. London.

How the system became forgotten This esotericism of the guilds is partially responsible for the disappearance of the system. By the beginning of the Christian era these ancient skills of numeracy, geography and metrology had all but disappeared, practiced only by a shrinking number of initiates. All that carried it forward through the centuries was its universality and momentum; it was altogether forgotten by the time of the introduction of the SI units. It would seem that the principal cause of the system’s demise was the repeated book burnings of aggressive usurpers. Whenever one tries to trace its existence back in time there are definite cut off points invariably instigated by the victors in a situation. Every dominant race, Mongols, Nazis etc. burnt the books, the destruction of Gnosticism after the council of Nicea that prompted the burning of the Library of Alexandria, the persecution of Confucianism and killing of scholars after the unification of China by the Qin are typical examples of attempts to expunge all cultural traces from vanquished people. Multiply this by thousands and such events are certain to erode even the best grounded of traditions. This is a small example of how metrology was applied and it partially reveals the philosophy of its practitioners. Everything had to be expressed as a rational integer or unity as a measurement. This is the diametric opposite of the single unit SI measurement, wherein everything is expressed as a fractured number. It is an intellectual crime that traditional metrology, which is perhaps the most powerful evidence of an advanced ancient technology, should be discarded as worthless by a group who would seek to inflict a system of evaluation fit primarily for bankers upon the population at large. There is not a reason in the world that dual systems should not continue with the artisans keeping their customary measures peculiar to their trade and all legal and scientific matters expressed in the SI terminology as a lingua franca. In all nations the decimal system has had to be, sometimes brutally, inflicted upon a populace unwilling to relinquish their traditional and rational systems of measurement. This process is still going on.

46

John Neal: The structure and function of ancient metrology

Petrie, W.M.F. 1880. Stonehenge: plans, description, and theories. London. Petrie, W.M.F. 1883. The  pyramids  and  temples  of  Gizeh. London. Petrie, W.M.F. 1910-11. Weights and Measures, Ancient Historical. Encyclopedia Britannica 11th edn. 28: 48088. Powell, M.A. 1987-90. Masse und Gewichte. Reallexicon der Assyriologie 7: 457–517. Priskin, G. 2006. The Egyptian Heritage in the Ancient Measures of the Earth. Göttinger Miszellen 208: 75-88. Schlott-Schwab, A. 198.1 Die Ausmasse Ägyptens nach altägyptischen Texten. Wiesbaden. Schwaller de Lubicz, R.A. 1998. The Temple of Man: Apet of the South at Luxor. Rochester. [translation of Le temple dans l’homme. Cairo 1949.] Skinner, F.G. 1967. Weights and measures:  their ancient origins and their development in Great Britain up to A.D. 1855. London. Stecchini, L.C. 1978. A History of Measures, http://www. metrum.org/measures/index.htm; ‘Notes of the Relation of Ancient Measures to the Great Pyramid’ Appendix to Tompkins. Stone, E.H. 1924. The stones of Stonehenge: a full description of the structure and of its outwork. London. Storer, J.S. 1819. History and antiquities of the Cathedral Churches of Great Britain 3. London. Temple, R. 2001. The Crystal Sun: rediscovering a lost technology of the ancient world. London. Thom, A. 1960. Megaliths and Mathematics. Antiquity 40: 121–128. Tompkins, P. 1978. Secrets of the Great Pyramid. London. Wright, E. 1599. Certaine Errors in Navigation Detected and Corrected. London.

47

The second stage of the Grakliani Culture Vakhtang Licheli 3. Early Bronze Age

The Grakliani Gora/Grakliani Hill settlement and necropolis are situated in the central part of the South Caucasus, in Georgia, within the territory of Samtavisi village in the region of Kaspi, near the village of Igoeti, on a hill rising from the right bank of Lekhura river, alongside the Tbilisi-Senaki-Leselidze highway (Figure 1). The hill overlooks two small rivers, the Lekhura and the Tortla. A multi-level settlement and burials of different periods were unearthed in this area.

4. Late Bronze Age/Early Iron Age. This period, starting from the 13-12th centuries BC, and lasting down until 900/800 BC, is far the best represented at the site, both at the settlement and the necropolis. Materials dating from this period were frequently found even at the surface level, in the eroded earth. Among the Late Bronze Age structures is a sanctuary with its interior partially damaged. The surviving parts of the sanctuary are very informative. A structure was unearthed in the eastern part of the 2nd terrace, the eastern and western parts of which are preserved, but the central part is heavily damaged. It is 5.2m wide, while the construction has survived up to a height of 0.3m (Figure 2). It shares its western wall with a unique sanctuary consisting of five partitioned clay sections along its NS axis. They all are all quadrangular in shape, while the surviving width of each section along the NS axis is successively 0.5m, 0.2m, 0.5m, 0.2 and 0.5m. Clay vessels were discovered on a leveled floor 3.8 metres away from this structure to the east. Among them is a vessel with a single zoomorphic handle,

The materials excavated in the settlement and necropolis date from different periods and comprise the stratigraphical sequence which is presented below.1 Naturally, future work at the site may well add new insights to this picture. 1. The Palaeolithic period is represented by a large number of primitive tools. Professor G. Grigolia identifies these finds as remains from the OlduvaiDmanisi period and consequently dates them to 2-1.5 million years before the present. 2. The Chalcolithic period

Figure 1. Location of Grakliani Gora: aerial photo. 1  The stratigraphy of Grakliani Hill has already been published, see Licheli 2011.

48

Vakhtang Licheli: The second stage of the Grakliani Culture

Figure 2. Grakliani Gora. Damaged Temple, plan and section.

10. 2nd-1st centuries BC

similar to those from the Treli necropolis dating to the Late Bronze Age. The altar in the southern part of the structure is especially interesting. It is a bowl-shaped structure with a circular upper part and a shaft-shaped edge, accurately leveled above. In the western part, the shaft has three horn-shaped prominences that culminate on the eastern side in a double-protome ram representation. In the light of this remarkable evidence, the genesis of double-protome figures, especially of those discovered on the territory of Colchis and dated to a much a later period, the 8th-7th centuries BC, remains open. This is the second Grakliani Gora sanctuary assigned to the final period of the Late Bronze Age. This is the period of the first stage of the Grakliani culture, flourished in the 11th – 9th centuries BC and includes a unique inscription of the 10th century BC (Figure 13).

11. 3rd-4th centuries AD. This layer was discovered in 2014 on top of an isolated hill, where Roman period tiles (of the 2nd-4th centuries AD) were unearthed in large quantities. They are identical to tiles found in Bagineti. A limestone capital was also unearthed (Figure 3). The Grakliani Hill settlement in the 5th-4th centuries BC The deposits dating from this, the 7th, period survive in a far better condition than those from the other periods. Sometimes there is the possibility for more detailed stratigraphic definition and there are chronological horizons within this layer. Specifically, the following three horizons have been distinguished in the 7th layer (the 5th-4th centuries BC) so far: 7A – mid and late 5th century BC; 7B – the late 5th century and early 4th century BC; 7C – the late 4th century. This 7th layer deserves our particular attention, as it contains evidence for relations with the Achaemenid Empire.2

5. 8th-7th centuries BC 6. 6th-5th centuries BC 7. 5th-4th centuries BC 8. 4th-3rd centuries BC 9. 3rd-2nd centuries BC

2 

49

Licheli and Samsonia 2010; Licheli 2011a: Licheli 2011b.

Wonders Lost and Found room structures built to the same design. A study of structures from the second terrace has already been published (Licheli 2011). The structure from the third terrace has been built to a rectangular design elongated EW. According to the evidence so far unearthed, the structure is 23 m wide EW and 3.98 m long NS. The southern wall of the building is lost. Evidently, it was gradually destroyed due to the terraced design, landslides and anthropogenic factors. The structure is turned to the south-east by 22 degrees from absolute north, which evidently is related to the seasonal changes in the solar path. Taking into account the structural and architectural data, the building shows close affinity with the basic structural principles underlying earlier buildings discovered in this area (Licheli 2011). The building is constructed from beaten clay walls (as well as altars and ovens) and beams of different thickness (mainly used as supports). An important element of the temples of 2nd and 3rd terraces are sizeable rectangular ritual ovens. One of them (3rd terrace, in Room III A; size: 1.48 x 1.35 x 0.8m) situated in its north-eastern corner (Figure 4).

Figure 3. Limestone capital.

The beaten clay flooring of the room descends to the South. The central and part of the eastern rooms have also been examined. In terms of planning, a similar structure can be found near the ‘western temple’ of Grakliani Gora. However, such structures are not common for the south Caucasus.

This paper sums up the results of the latest years of work that focused on the 5th-4th centuries BC, i.e. on the traces of the Achaemenid period (3rd and 4th terraces; Figure 5). The Grakliani Gora evidence is crucial for the study of Achaemenid economic and cultural influence in the South Caucasus, and for indentifying the economic and cultural background leading to the establishment of the Kingdom of Kartli (Iberia).

The Grakliani archaeological layers are also important because they offer a totally different picture of the political events taking place in the central part of South Caucasus in the times of the Achaemenid ‘palaces’. While the ‘palaces’ reveal no traces of struggle and suggest a ‘peaceful’ withdrawal of the Achaemenids, the Grakliani Gora structures are burned and destroyed, which suggests a different picture. The above-mentioned Room III C, where arrowheads were discovered, also bears the traces of a struggle. Thus, the population residing in East Georgia must have suffered from a devastating military attack. However, the finds from the dwellings suggest that the population must have been informed previously since most of the household goods had evidently been removed (as from the ‘palaces’).

In this paper, I will refrain from discussing the problematic question of the date of the Achaemenid invasion of the Caucasus (which most scholars associate with Darius I), since ‘despite some chronologically fixable points, the course of events is often unclear’ (Kuhrt 2010: 181). But whatever the precise date, it is clear that the Achaemenids had gained a firm foothold in the South Caucasus by the 5th-4th centuries BC. Part of the 6th-4th century finds from the archaeological sites of Iberia are related to the Achaemenid world (Licheli 2001; Licheli 2014; Makharadze 1999; Narimanishvili 2009). The most obvious evidence consists of Persian ‘palaces’ discovered in Georgia (Gumbati, Sabatlo) as well as in Azerbaijan (Sari-Tepe, Qarajamirli) and Armenia (Beniamin, Erebun. See: Knauss 2001, 125145). Later discoveries dated to the same period (the 5th-4th centuries BC) from Grakliani Gora specify some problems related to Iberian-Achaemenid relations.

The Grakliani Gora finds abound in ceramics of the so-called ‘Achaemenid type’, which are known from earlier discoveries as well. However, the Grakliani Gora settlement and necropolis enables us to make some inferences. First, there are analogies among the ‘palace’ ceramics; bowls for example. Ceramics found at Qarajamirli partly resemble the Grakliani Gora vessels, but a full-scale comparison is not possible as they are

Shrines found on the second and third terraces of Grakliani Gora are especially important for the study of the Achaemenid influence. They are three50

Vakhtang Licheli: The second stage of the Grakliani Culture

Figure 4. Oven on the 3rd terrace, in Room III A.

badly preserved. In any case, Qarajamirli bowls Nos 9 and 28 (Babaev, Gagoshidze and Knauss 2007: fig. 12) are similar to Grakliani Gora type 2 bowls. The comparison of ceramic ware from these sites reveal that after the Achaemenid withdrawal from Qarajamirli, that is after 330 BC, and at the time of sporadic occupation of the territory, ceramics were painted in red. They were found together with vessels bearing circumscribing strips done in white paint that appear in East Georgia after the 4th century BC (the so-called ‘Herian’ pottery) (see Pitskhelauri 1973). The Grakliani Gora finds may serve

as a clue to dating the appearance of painted vessels in the central part of the South Caucasus. In any case, the absence of the so-called ‘Herian’ ceramics among the Grakliani Gora painted vessels of the Achaemenid type indicates that the latter date earlier than those unearthed at Qarajamirli (generally, the Grakliani Gora finds contain ‘Herian’ clay vessel fragments as well). Interior altars built agaisnt the western wall of the sanctuaries are vivid examples of Achaemenid cult tradition. These are typical fire altars having exact

Figure 5. Masterplan of the 3rd and 4th terraces

51

Wonders Lost and Found

Figure 6. A Zoroastrian open fire altar discovered on the 4th terrace.

analogues in Western Iran. Specifically, an altar is built against the western wall of the fire temple No.1 at  Nush-i-Jan. The altar has a fire grate with obvious traces of a fire (as at Grakliani Gora). According to the archaeologists working at  Nush-i-Jan (Roaf and Stronach 1973), this must be a Persian fire altar transformed under the reign of Cyrus (590-530 BC). However, some researchers identify it as a Median altar later adapted by the Persians. I suppose that this construction is of special importance: it could be rare example of a Median tradition, one that continued in the same form into the Achaemenid period, but in this case outside Iran. Such altars can be found in almost all rooms of the 2nd and 3rd terraces. The following deserves a special mention: In Room III A, which obviously had a dominant function, a cult structure was found built against the eastern wall immediately next to the sacred oven, opposite the fire altar. This is a decorative composition, 68 x 16cm, arranged on a protruding platform south of the oven. Three shallow quadrangular pits (full of pure ash) of almost the same size are arranged in a horizontal sequence. Along the vertical axis of the northern pit, there is a truncated cone-shaped bulge linked to the pit at right angles. In my opinion, this unique combination could be a later reminiscence of the religious tradition observed at the Late Bronze Age sites. Thus, this one room alone provides enough grounds to suppose that despite the Persian/Eastern influence, the local ethnic picture had not been altered.

cm long and 4 cm wide, and an iron knife blade were found on the surface of the altar (cf. an altar from Tsikhia-Gora). There were also lamp fragments on the altar. The eastern part of the altar surface is leveled, while the western part is deformed. There is a step of cobblestones built to the altar to the south-east (44 x 53 x 25cm), which was plastered with a thin layer of lime solution (Figure 6). In the west, the altar has a 65 cm wide and 21 cm long platform. There is a faint trace of a fire-place above the platform. The bow-shaped fire-place is 3.1 cm high. Obviously, this altar was at first shaped as an oven and then was converted to an open fire altar. The baking platform was preserved, the horizontal length of which is 48 cm. A thin wall plastered with a lime solution starts 19 cm east of the altar and continues to the north. The distance between the eastern wall and the altar varies between 19 and 39 cm, and between the northern wall and the altar between 10 and 20 cm. The wall is 89 cm long in the east and 119 cm long in the north. The north-eastern angle of the wall is rounded. To the east of the wall, there is a fragment (49 x 43cm) of the same structure that had fallen down. 77 cm south of the eastern wall there is a 12 cm deep hole, 1.5 cm in diameter, cut in the floor to support a beam. Part of a collapsed wall west of the altar was observed on the ancient surface of the court. Obviously, the fire altar was surrounded by a thin wall on three sides (from the east, north and west) to protect the fire from the wind. This reflects the trend of transferring Zoroastrian practices outside before the reign of Artaxerxes II (404-359 BC), when the fire cult was established, and temples were built to worship the goddess Anahit (Choksy 2007: 230). As is well-known, such Persian temples were modeled after Elamite, Babylonian, Assyrian and Greek patterns.

Together with this type of altar we can also discuss a Zoroastrian open fire altar discovered in the 4th terrace (Figure 6). The altar appeared at a depth of 136 cm below ground level. This is a rectangular altar with rounded corners. The western face is 128 cm, the northern 122 cm and the eastern and southern faces 122 and 119 cm respectively. The height is 58-59 cm. The altar was erected outdoors. A bovine shin bone 11

An analysis of temple planning on Grakliani Gora may reveal Achaemenid influences: temples on the 2nd and 3rd terraces have auxiliary rooms, which was quite 52

Vakhtang Licheli: The second stage of the Grakliani Culture

usual. This can even be seen as an adaptation of a Persian practice. There are special rooms on Grakliani Gora where animal skeletons were found. These rooms are directly connected with the main ritual space that accommodated shelves for ceremonial items and a fire altar (Persian adurgah – ‘fire place’, a bomos according to Strabo, who describes an open air precinct at Zela for Ahura Mazda, Anahita and other divinities (Geography 11.8.4) (see Choksy 2007: 244). According to the texts from Persepolis fortification tablets (discovered by Herzfeld during 1933-34 excavations) such altars/rooms were used for different religious rituals honouring various deities (Razmjou 2004). This room on Grakliani Gora might have accommodated one of the main ceremonial ovens. There are two details that lead to this conclusion: a. the unique decoration of the oven having no analogue in Georgia or the Caucasus, and what is even more important, b. the altar arranged directly next to the oven, which, without any exaggeration, is the direct analogue of the Late Bronze Age altar from the Kartli area (as mentioned above), showing continuity of local traditions. This detail has yet another significance for the identification of ethnical identity: the practice of the same burial ritual as attested already in the Early Bronze Age and especially, the total uniformity between the 8th-7th century BC and the 5th-4th century BC burial rites suggest that, as mentioned, despite the Achaemenid influence, the ethnic affiliation of the population was not affected.

pots. They have a distinctive ornament – a jagged curve incised between two circumscribing grooves over their shoulders. Such pots were also discovered on the 4th terrace in 2014. Fragments of these pots prevail over others, which means that the inhabitants of Grakliani Gora used locally manufactured pottery. The discovery of metal workshop remains on the 2nd terrace is no less important. The chemical treatment of a crucible found at this place attests to a small bronze casting workshop. I should also mention an unusually large number of boat-shaped grinding stones (Figure 8). They were discovered almost in every room, while in room III B a small mill was unearthed in situ – three grinding devices arranged against the western wall. A space was left in between for millers to adopt a comfortable position during this laborious and time-consuming process. All the three hand grinders had finely plastered knee supports on each side. The flour would pour into a clay-plastered trough-shaped basin. There were also discovered in situ two ‘family mills’ in building N IV-6 (Figure 10), where two vessels for fire on the oven were also unearthed. Pottery of the 4th century BC was also discovered also in the building N IV-1. The grind-stones were made of basalt, although examples of sandstone and gabro-diorite were found as well (analyses by A. Tetruashvili). Paleobotanical analyses confirmed two types of wheat (Triticum Dicoccum, Triticum Aestivum) and palynological analysis confirmed agricultural practices favoured by the relatively mild climate (mixed forest and pteridophytes). The palynological analysis (made by Prof. E. Kvavadze) of bowls revealed the prevalence of walnut (Juglans Regia) pollen followed by hornbeam (Carpinus Caucasica) pollen. Small quantities of oak, pine, beech, lime, oriental hornbeam and wild grape pollen were also discovered. Wheat, thistle, buckwheat and goosefoot pollen were also discovered.

This is a concise account of the material culture illustrating the events unfolding on Grakliani Gora in the 5th-4th centuries BC. We should only add a few words about the industrial profile of Grakliani Gora, namely the pottery. A kiln (Figure 7) was unearthed on the 2nd terrace together with single-handled round-bellied

Figure 7. A kiln on the 2nd terrace.

53

Wonders Lost and Found

Figure 8. Boat-shaped grinding stones.

Figure 9. Zoroastrian altar, 3rd terrace, BIII-C.

Figure 10. The ‘family mill’ in the building N IV-6 (5th-4th centuries BC)

54

Vakhtang Licheli: The second stage of the Grakliani Culture

Figure 11. Oven in the building N IV -1 with pottery in situ.

Figure 12. Shrine of the 10th century BC.

55

Wonders Lost and Found

Figure 13. Inscription in the 10th century BC shrine.

Stockbreeding is suggested not only by the cattle bones in the graves, but also by remains of animal-based food in cultural layers (cattle, pig, goat, sheep). Boar and deer bones prevail among the remains of wild animals. Horses were used in husbandry and probably, in warfare. Moreover, horse skeletons were found in burials as well, which suggests the existence of affluent upper classes. Social strata are evident not only from qualitative and quantitative differences in burial goods but also from fibulae that obviously served as insignia. The burial inventory may include gold, silver and bronze diskfibulae, that have direct analogues in the Achaemenid world (Susa) and can be occasionally found in Colchis as well (one example). Roundels of the 14th century BC, analogues of Grakliani Gora finds, were interpreted by O.W. Muscarella as of Elamite provenance (1988, 226). Chronologically closer are Elamite roundels, one dated to the 16th century BC, discovered in Atskuri grave no. 1 with a stone embankment (Licheli 2004. Licheli and Rusishvili 2008).

Intensive trade relations (Colchian, Phoenician, Persian and Greek produce), Sanitary and hygienic structures (a 5th century BC sanitary system was discovered), Standard planning for residential terraces, Architectural solutions attested only on Grakliani Gora, Specific planning of sanctuaries and a novel interior design.

For the assessment of the Grakliani Gora site, we should take into account a set of urban characteristics attested in this area:

Bibliography

No other archeological site of this period in Eastern Georgia possesses all these urban characteristics. Thanks to these Grakliani Gora stands out among other 5th-4th century BC sites not only in Georgia but also in the entire South Caucasus. The combination of these features together is a new phenomenon first attested at Grakliani Gora. This raises the likelihood that 5th-4th century BC social, economic and cultural developments in this area can be identified as a separate culture. namely the Grakliani Culture.

Babaev, I., I. Gagoshidze and F. Knauss 2007. An Achaemenid ‘Palace’ at Qarajamirli (Azerbaijan) Preliminary report on the excavations in 2006, in A. Ivantchik and V. Licheli (eds) Ancient Civilizations from Scythia to Siberia 13: 31-45. (Achaemenid Culture

Population density, Local industry, The availability of a shrine in public spaces, 56

Vakhtang Licheli: The second stage of the Grakliani Culture

and Local Traditions in Anatolia, South Caucasus and Iran. New Discoveries): 31-45. Leiden-New York. Choksy, J. 2007. Reassessing the material context of ritual fires in ancient Iran. Iranica Antiqua 42: 229-69. Khakhutaishvili, D. 1964. Uphlistsikhe 1. Tbilisi. Kuhrt, A. 2010. Achaemenid images of royalty and Empire, in G. B. Lanfranchi and R. Rollinger (eds) Concepts of Kingship in Antiquity (Proceedings of the European Science Foundation Exploratory Workshop): 87-106. Padua. Knauss, F. 2001. Persian rule in the North. Achaemenid palaces on the periphery of the Empire, in I. Nielsen (ed.) The Royal Palace Institution in the First Millennium BC. Regional development and Cultural Interchange between East and West, (Monographs of the Danish Institute at Athens 4): 125-143. Athens. Licheli, V. 2004. A burial with a stone embankment at Atskuri. Journal of Georgian Archaeology 1: 218-225. Licheli, V. 2011a. Cultural Relations in Central Transcaucasus,, in The Caucasus: Georgia on the Crossroads. Cultural exchange across the Europe and Beyond (Proceedings. 2nd International Symposium of Georgian Culture, November 2-9, 2009, Florence, Italy): 72-75. Tbilisi. Licheli, V. 2011. Urban Development in Central Transcaucasia in Anatolian Context: New Data. Ancient Civilizations from Scythia to Siberia 17: 135-156. Licheli, V. 2014. Grakliani Culture. Tbilisi. Licheli, V. 2017. Kommentaren zum archäologischen Kontext des Grakliani Gora von 5.-4. Jahrhunderten v. Chr. Georgica. Zeitschrift für Kultur, Sprache und Geschichte Georgiens und Kaukasiens 38: 41-58. Licheli, V. and R. Rusishvili 2008. A Middle Bronze Age Burial at Atskuri. Archaeology in Southern Caucasus: Perspectives from Georgia (Ancient Near Eastern Studies Supplement Series 19): 205-228. Leuven/ Dudley ME. Licheli, V. and N. Samsonia 2010. Mesopotamian Seals from Igoeti -Grakliani settlement, https://www.tsu. ge/data/image_db_innova/samsonia.grakliani.pdf Makharadze, Z. 1999. New Discoveries at Tsikhiagora. Dziebani 4: 57. Muscarella, O.W. 1988. Bronze and Iron. Ancient Near Eastern Artifacts in the Metropolitan Museum of Art. New-York. Narimanishvili, G. 2009. daraqois namosaxlari da samxreT kavkasiis aqemeniduri xanis istoriis zogierti sakitxi. Iberia – Kolxeti 5: 125-145. Pitskhelauri, K. 1973. aRmosavleT saqarTvelos tomTa istoriis ZiriTadi problemebi (Zv.w. XV-VII ss). Tbilisi. Razmjou, S. 2004. The Lan ceremony and other ritual ceremonies in the Achaemenid period: the Persepolis Fortification Tablets. Iran 42: 103-117. Roaf M. and D. Stronach 1973. Nush-i-Jan: Second Interim Report, 1970. Iran 11: 129-40.

 

57

Owl skyphoi around the Adriatic Branko Kirigin In earlier scholarly works the presence of Greek pottery on Adriatic sites was interpreted as a result of cabotage sailing (along the coasts) of merchant ships. Arriving at different ports the pottery was transported inland. Recent archaeological excavations at Palagruža, the Isle(s) of Diomedes, has shown that the Greeks used the open sea, long-distance trade route as early as the late 6th century BC. This open sea route via Palagruža was much shorter then sailing along the coasts and was

used by trading boats that connected Athens with Adria and Spina via Corfu, ancient Kerkyra, the main Greek port just before the Adriatic Sea, carrying cargo that was intended for these ports only (Kirigin at al. 2009). English colleagues, especially Michael Vickers, take most of the credit for spreading news about our work on Palagruža. In 1999 he invited me to Oxford to present the results of our excavations on this small, isolated,

Figure 1. Map of the Adriatic from Google Earth with sites where owls have been found (numbers correspond with site numbers in the list of sites)

58

Branko Kirigin: Owl skyphoi around the Adriatic

and waterless island in the centre of the Adriatic Sea (no. 6 on Figure 1). Michael is also the person who was most responsible for bringing our exhibition ‘Palagruža: the Isle of Diomedes’ to the Ashmolean Museum in 2002; the first exhibition from Croatia to be mounted at this famous Museum. Due to his efforts and wisdom I am delighted to make this small contribution to his Festschrift with the ancient Greek symbol of Athens – an owl. Owls are a popular motif on Greek vases, predominantly on skyphoi. Many have been found at sites on the north Mediterranean, at Greek colonies, emporia and local (native) settlements, including sanctuaries and burials, male and female. Also, and unfortunately, a good deal of the published owl skyphoi have no context, and the data thus can only be of interest to art historians. As an archaeologist I do not feel competent to attribute these owl skyphoi from the Adriatic, especially to put them into groups (Johnson 1953; 1955), less still to attribute new ones. Still, a catalogue of all the known sites that have produced owls has not been published so far, and I hope that this paper on owls from the Adriatic will be of some use to whoever will do it in the future.

Figure 2. Palagruža. Sherd with legs of a bird. (Photo B. Kirigin).

As far as I could find there are no vases with mythological scenes where owls are represented on the Adriatic. Owls that form part of mythological scenes on early Greek vases (6th-5th century BC) have been recently examined in detail by Kreuzer (2010). There is a slight possible exception. A small sherd from Palagruža (Inv. no.72342) is decorated with the lower parts of two legs of a bird on a black line, and on the right with a curve (Figure 2). The sherd belongs to a small vessel (skyphos?) and is of a dull, brownish clay. One wonders whether this could be a representation of a BF owl on a shield.1 The claws on our sherd have some similarities with the owl shown on a BF neck-amphora from Bochum (Kreuzer 2010: 174, fig. 35).

Surveying the appropriate publications, and by consulting with dear colleagues from abroad, I have come across 32 sites where owls have been found on or near the Adriatic Sea (Figure 1). Of these 8 are on the east coast and 24 on the west, i.e. Italian, coast. The sites are dealt with in more detail in the Appendix to this paper. Most of the owls are on Attic and South Italian Red–Figure skyphoi. Other shapes with owls are rare: a bolsal and two kantharoi from Adria (no. 12 in the Appendix), a kantharos from Numana (no. 21 in the Appendix. As well as those decorated in Red-Figure technique, several examples have been decorated by applying colour to Black-Glaze skyphoi, (Vilina špilja, Pharos, Numana and Salapia: nos. 2, 4, 20 and 23 in the Appendix), and one owl appears on a late Gnathia oinochoe from Issa (no. 5 in the Appendix).

Watson has noted the various problems involved in the dating of owl-skyphoi, emphasizing that it is ‘difficult to relate them to other painted vases…’ (Watson 1998: 35-37). According to the dates and attributions as presented in the publications, the earliest owl skyphoi in the Adriatic date to the first half of the 5th century (Dyrrachion, Adria, Certosa, Aureli, Pianello, Camerano), mostly from the mid-5th (Nesactium, Adria, Forcello, Spina, Numana). A high concentration occurs in the late 5th (Čitluci, Tolmin, Adria, Forcello, Spina, Pitino, Camerano, Canosa, Banzi, Oria, Lavello, Rutigliano, Botromagno, Brindisi, Leuca), and a few are from the first half of the 4th century BC (Attic RF from Canosa, South Italian RF from Vilina špilja, Palagruža, overpainted from Numana, Salapia, Vilina špilja and Pharos, and Late Gnathia from Issa).

On 17 sites owl skyphoi have been found in graves (nos. 1, 3, 5, 7, 8, 15-19, 21-27). Among these one is found in a female grave at Certosa (no.16). Many were found in settlements: Faros (no. 4), Padova (no. 9), Castelrotto (no. 10), Oppeano (no. 11), Adria (no. 12) Este (no. 13), Forcello (no. 14), Spina (no. 15), Pitino di S. Severino (no. 19) and at sanctuaries: Vilina špilja (no. 2), Palagruža (no. 6), Oria (no. 29) and Leuca (no. 32). The exact context is not known for owls from Egnazia (no. 28), Brindisi (no. 30) and Rudiae (no. 31), but as they were completely preserved it is possible that they are from graves. In general it can be stated that among the different native communities around the Adriatic the owl skyphoi were found in the same contexts (graves, settlements, sanctuaries).

Owls on shields Kreuzer 2010: fig. 5 on p. 129, fig. 26 on p. 162. We have representations of cocks on shields too, but owls on shields are all in BF and the persons that hold shields are in RF: See for example: http://www.beazley.ox.ac.uk : nos: 200022, 200080, 200524, 201956, 211334.

1 

59

Wonders Lost and Found to be of a type known from the Agora excavations and dated to c. 480-460 BC (Sparkes and Talcott 1970: 257 no. 313, 259 no. 342).

The sites at Glasinac (no. 3) and at Most na Soči (no. 8) were associated with iron ore and the production of iron artefacts. So it is possible that there some trading activity existed between Greece and the communities that controlled these mines.

2. Vilina špilja (Cave of the Fairies) This cave is situated above the source of the river Ombla (Rijeka Dubrovačka) near Dubrovnik. There were no excavations, only surface collecting. Apart from the Late Neolithic, Copper and Bronze Age pottery, some 20 finely-painted sherds of pottery have been found recently. Among these six skyphoi sherds with different owls (five Red-Figured and one with applied colour) are mentioned: three are visible on a photo with other sherds (Figure 3). They are attributed to South Italian and dated to the first half of the 4th century BC (Perkić 2010: 159-161, 206 no. 404).

No owls have been found at the Diomedes sanctuary at Cape Ploča on the mainland between Split and Šibenik (Čače, Šešelj 2005), or at the cave sanctuary at Spila near Nakovana on the Pelješac peninsula (Forenbaher and Kaiser 2003). Neither site can be dated before the second half of the 4th century BC. Nevertheless, the puzzle surrounding owls is still unsolved. One wonders why owls would be made in South Italian workshops. Was it because owls were popular because Athens was popular, or because owls represent an attractive decorative design, or was it because they were a representation of some local cult(s)?

3. Glasinac A very worn BG miniature skyphos with owls and olive branches (Figs. 4a and 4b, photo by Ivo Dragičević) was found, together with bronze vessels, bronze, silver and glass jewellery and iron spearheads at Glasinac (Čitluci, Bosnia and Hercegovina) in Barrow V excavated in 1892 (Fiala 1892: 405-6, pl. II, no. 2; 1893: 139, fig. 36). The skyphos is of reddish clay, 5.8 cm in height, and dated to the end of the 5th century BC. It has been identified as an Etruscan imitation of Attic pottery (Parović-Pešikan

Sites with owl skyhoi found on the Adriatic 1. Dyrrachion Hidri has published drawings of pottery finds from grave 9 in Dyrrachion. Among these is an Attic type A skyphos with an owl (in sketch). No other details are mentioned (Hidri 2006: 40, 93, T. II). The skyphos looks

Figure 3. Vilina špilja. Painted pottery sherds (after Perkić 2010)

60

Branko Kirigin: Owl skyphoi around the Adriatic

Figure 4. Čitluci. Front side (left) and handle side (right) of the Attic owl skyphos (photos by Ivo Dragičević)

1985/1986: 41, Pl. I, no. 2).2 It is of Attic type A shape (with horizontal handles). This type can be dated from the end of the 6th to the second half of the 4th century BC (De Juliis 2002: 135-136).3

to two miniature owl skyphoi with added colour from Apulia which are in the Sèvres Museum. These are 6.5 cm in height, made of buff-coloured clay, and come from an unknown site.4 One wonders whether the Čitluci skyphos was in fact originally decorated in RF style, but the decoration has had the same fate as the owl skyphos from Oria (no. 29 here), with its decoration only preserved on one side in negative (Maninno 2006: 110 nos. 107 and 112, fig. 121). A similar effect can be observed on the skyphos from Čitluci. If this is correct, then the Čitluci skyphos is Attic.

Katić (2002: 431) notes that this skyphos ‘originated in Pharos where there are numerous examples of the same type of skyphos’ (see no. 4 below). In one of his earlier works Katić suggests that the skyphos from Čitluci could not be from the late 5th century BC, but rather from the second half of the 4th, and that this type of skyphos with owls in added colour were most probably produced in Pharos (Katić 1995: 123 and note 5; 1999-2000: 53). The problem with the skyphos from Čitluci is that the shape is of Attic A type (the ratio of base to rim diameter is 1: 1.3) and that the Pharos ones have a more narrower base (ratio of base to rim 1: 2.1). On the other hand, Etruscan owl skyphoi are, as far as I know, all of type B, and the representation of the owls on Etruscan owl skyphoi are not as near to the base ring, as the Čitluci owls are. Furthermore, the eyes of the owls on the Čitluci skyphos are made in yellowish colour with the pupils marked by a dark dot in the centre, in contrast to the Pharos owls that have the outline in white i.e. the other way around (Figure 5).

To get a closer date for the owl skyphos from Čitluci one has to look at the other offerings in the same cremation grave. Among some 20 other objects from this grave, a bronze strainer (Fiala 1892: pl. II no. 5; 1893: 140 fig. 35) has close parallels with examples from Apulia that are dated to the mid-5th century BC (Tarditi 1996: 54, nos. 100-102) and from Olbia on the Black Sea (ParovićPešikan 1960: 29). Tarditi stresses that those from Apulia could be local imitations of strainers produced in Etruria or Campania, and that they do not differ from examples in different regions (Tarditi 1996, 143). A silver bracelet with a hammered end and incised decoration (Fiala 1892: 406 fig. 26; 1893: 139 fig. 32) has close parallels with a gold specimen from a tumulus in Novi Pazar, which is dated to the 5th century BC (Popović 1994: 142 no. 133).5 And finally, a ribbed bronze jug with silver wire connecting the handle to the rim (Fiala 1892: 406 pl. I, no. 1; 1893: 140 fig. 34) which is dated to the end of the 5th century BC (Babić 2004: 85, 111-112, 156 no. 20, 168 no. XXI).6 This evidence clearly shows that

It could be that the decoration on the Čitluci skyphos has been affected by burning during cremation. Also, the skyphos from Čitluci is half the size of those from Pharos. Another factor is that we do not know the fabric(s) of the Pharos owl skyphoi. The fabric of the Čitluci skyphos is reddish/yellow in colour, and it is not decorated with any circles or a dot on the base. It seems as though the skyphos from Čitluci is nearer

CVA Sèvres, p. 91 pl. 46, nos. 24 and 30. See also: De Juliis 2002: 108, no. 6/M; http://www.beazley.ox.ac.uk/databases/cva.htm. Biba Teržan (personal communication) suggests that this bracelet is, in fact, an earring. 6  For clay pots of similar shape and date see: Baldoni 1993: 105 nos. 98-102; 126 no. 99; 128 nos. 100-102; 129. 4 

In an earlier paper Parović Pešikan (1960: 32) identified this skyphos as of Greek origin: see also: Babić 2004: 159, no. 11. 3  For type A owl skyphoi in added colour from Apulia see no. 29 (Oria) and De Juliis 2002: 107 no. 5 and 108 no 6 (CVA, Sèvres, pl. 46, 24 and 30 with laurel branches) from unknown sites, 7.5 and 6.5 cm in height. 2 

5 

61

Wonders Lost and Found

Figure 5. Pharos. Owl skyphoi (after Katić 1999/2000)

62

Branko Kirigin: Owl skyphoi around the Adriatic

weights (Jeličić Radonić, Rautar Plančić 1995: 89 no 1). Altogether some 10 skyphoi with owls in applied colour are known from Pharos. According to the drawings they all seem to be of similar shape (Agora 349) and have a similar painted design. Thus the mid-4th century BC date proposed by Katić is acceptable, but, as Katić himself mentions, the production of them in Pharos still needs to be confirmed (Katić 1999-2000: 53, 55). As Katić has also observed, the Pharos fragments seem to be very close to the two owl skyphoi from Numana (see below no. 21).

the Čitluci owl skyphos dates to the second half of the 5th century BC, and is most probably an Attic product. 4. Pharos Some 20 sherds of different BG skyphoi with applied white colour decoration containing parts of owls and/ or olive branches, have been found during rescue excavations in the 1990s within the Greek colony of Pharos on the island of Hvar. No precise location or stratigraphical data are given in the original publication, except a note that they were found in the same context alongside locally produced table amphorae and small bell-shaped clay kiln supporters (Katić 1990-2000: 53). The skyphoi are dated around the mid-4th century BC (Katić 1999-2000, 53, 58). Katić (1999-2000, 53) thinks they may be locally produced and exported to Bosnia (see no. 3 here). They are published only in drawings (Katić 1999-2000, 58 = T. 3) but without a scale. Not a single one is complete. An exception is a reconstructed drawing of a fragmentary skyphos, decorated in its lower portion with part of an owl (Fig. 5) with a total height of c. 13 cm (Jeličić Radonić, Rautar Plančić 1995: 62 no 1; this one is reproduced in Katić’s pl. 3 upper left: Katić 1999-2000: 58 no. 1). There is no description of the fabric, nor is it mentioned where they are held. Another fragment of a BG skyphos with an olive branch was found in 1985 in a trench excavated in Vagonj Street no. 5, together with other Greek and Hellenistic fine wares, terracottas, tegulae, amphorae taps, and loom

5. Issa A rare, if not the only known example, of an oinochoe of late Gnathia style with an owl on the front (Figs. 6a and 6b) comes from Issa, a mid-4th century Syracusan colony on the island of Vis. Most probably it comes from a grave, as all the rest of the completely preserved fine pottery from this site. It is a local mid-3rd century product (Kirigin et al. 1986: 24 no. 59; Kirigin 1990: 63, 64 no. 9, pl. 30 b). 6. Palagruža Excavations on Palagruža (the island of Diomedes), have revealed fragments of 3-4 skyphoi with owls. These have been found in a dark layer (4050) on the south slope of Salamandria, the main archaeological site on the main island on this small archipelago: Velo Palagruža (Kirigin, Miše and Barbarić 2010). This layer

Figure 6. Issa. Detail from the front side (left) and handle side (right) of Gnathia style oinochoe (photos T. Sesar)

63

Wonders Lost and Found

Figure 7. Palagruža. Sherd of an owl skyphos. Inv. No. 40216. (Photo B. Kirigin)

Figure 9. Palagruža. Sherd of an owl skyphos. Inv. No. 64509. (Photo B. Kirigin)

Figure 8. Palagruža. Two sherds of an owl skyphos. Inv. Nos. 72344 and 72343. (Photo B. Kirigin)

Figure 10. Palagruža. Sherds of owl skyphoi. Inv. No. 72352, 72340 and 72341. (Photo B. Kirigin).

Figure 11. Palagruža. Seven sherds of owl skyphoi. From left to right. Upper row: Inv. Nos. 72350, 72345, 72346, 72351; lower row: 72347 (seemingly a palmette), 72348, 72349 (Photo B. Kirigin).

64

Branko Kirigin: Owl skyphoi around the Adriatic

Figure 12. Nesactium. Reconstruction drawing of the owl skyphos (after Mihovilić 2004) and the sherd showing a part of an owl shown on the left (Photo courtesy of Kristina Mihovilić).

has been interpreted as a result of clearance of the plateaux when the Late Roman fort was built and all the previous buildings were removed from it. The layer includes prehistoric (Early Neolithic, Late Copper/Early Bronze Age), Greek Late Archaic, Classical, Hellenistic, Roman Republican and Early Roman artefacts. The abnormal circumstances in which these artefacts were deposited is demonstrated by several examples of connecting sherds which have been found in this layer were recovered lying more than two metres apart. In the case of owl no. 1 (Inv. No. 40216: Figure 7) the sherds were found in squares V6 and V8. These are of pale buff clay. Some other sherds shown in Figure 8 (Inv. No. 72344 and 72343) could belong to this owl skyphos. They are clearly non-Attic and could belong to a South Italian 4th century BC product. Another more elaborate sherd with an owl (Inv. No. 6450: Figure 9) is made of pale brown clay, and as Ian McPhee has suggested (via email) it ‘may come from northern Apulia, i.e. area of Canosa’ and is ‘probably 350-300’ BC in date (Kirigin, Katunarić, Miše 2005: 385). Another two sherds from squares V 6 and 8 (Inv. Nos. 72340 and 72341) with dots in rows seem to belong to owl skyphoi (Figure 10). The sherd shown uppermost (Inv. No. 72352) in Figure 10 could also belong to an owl skyphos. Both are of pale buff clay and have traces of BG on inside. Some other sherds from Palagruža could also belong to owl skyphoi (Figure 11).

cm in height and is made of buff brown clay (Mihovilić 2002: 503-504, pl. 2,4; 2004: 107d). 8. Most na Soči Recent excavations at the well-known cemetery of the Iron Age St. Lucia culture (some 6000 graves) have produced an Attic type B skyphos decorated with an owl (Figure 13), found in a cremation burial (grave no. 1). The grave contained a Certosa fibula, two bronze serpentine fibulae, and a decorated bone handle of an iron knife. The grave is dated to the third quarter of 5th century BC (Mlinar 2002: 28-30). 9. Padova Four owl skyphoi are known. These come from the preRoman settlement. Inv. Nos: 357902 (De Min, Gamba, Gambacurta, Ruta Serafini 2005: 76-77, fig. 79; WielMarin, 2015), and 357903-357904 and 360189 (all three of them in Wiel-Marin, 2015). 10. Castelrotto (Verona) One owl skyphos is known. It was found in the ancient settlement. Inv. No. 87265vr (Wiel-Marin, in press) 11. Oppeano (Verona) One owl skyphos is known (Inv. no. 87280vr). It was found in the old settlement (Gamba 1986: no. 26; WielMarin in press).

7. Nesactium (Vizače near Pula) During excavations carried out in 1981 within Temple B (Roman) 11 fragments of an owl skyphos were found in the area of a tomb containing multiple burials of the local Histrian population dating from the end of the 8th to 3rd/2nd centuries BC (Mihovilić 1996). The skyphos is dated to the mid-5th century BC (Figure 12). It is 6.6

12. Adria Recently Wiel-Marin has published the Attic RedFigured vases from Adria, among which 36 have representations of owls: two on kantharoi, one on a bolsal and the rest on skyphoi. 11 skyphoi are of type A, 8 of type B and the rest are fragments that could not 65

Wonders Lost and Found

Figure 13. Most na Soči. Attic owl skyphos. (After Mlinar 2002)

16. Certosa (Bologna) In a female inhumation, in a wooden coffin, an RF owl skyphos type B was found (no dimensions are given), lying next to two Attic RF kraters (by the Painter of Louvre C11266 and the Pan painter), a Saint Valentin kantharos, metal and ivory objects. The grave is dated to the second quarter of the 5th century BC. The skyphos belongs to Johnson’s Group VI (Govi 1999: 52, note 94, fig. 22).

be indentified by shape. There are also many fragments with representations of owl branches. The above are dated to the second to the third quarter of the 5th century BC. They are all placed in ‘Gruppo degli Skyphoi con Civetta’ and to several groups proposed by Johnson in 1953 (eight belong to his group I, three to group II, one to group III and IV, and two to group VIII). All of them belong to the Bocchi Collection, which was mostly formed in 19th century, and come from the settlement (Wiel-Marin 2005: 192-193, 235-240, 260-263).

17. Aureli (Bologna) An Attic RF owl skyphos type B was fund in a simple inhumation grave, along with a krater by the Painter of Bologna 228 and a kylix decorated in the manner of Douris, together with some seven other objects that are now lost. The grave is dated to c. 465 BC. No other details are given except that it has similarities to the one from Camerano, see below (Govi 1999: 90-91, fig. 40).7

13. Este (Padova) One owl skyphos is known. F. Wiel-Marin, personal communication, inv. no. 39866. 14. Forcello (Bagnolo di San Vito, Mantova) Some 60 owls are known: 40714 (de Marinis, Rapi 2007: fig. 72,4), 99331 (Marinis and Rapi 2007: fig. 72,5), 99332 (de Marinis, Rapi 2007: fig. 72,1), 99333 (de Marinis 1986: fig. 76), 99334 (Marinis and Rapi 2007: fig. 72,3), 99335 (de Marinis, Rapi 2007: fig. 72,2), 99336, 99337 (Wiel-Marin 2012: fig. 10), 99338 (Wiel-Marin 2012: fig. 11), 99339-99342, 99367-99368, 145865+145866 (Marinis and Rapi 2007: fig. 97; Wiel-Marin 2008: fig. 17), 163290163291, 163295-163296, 163297 (Wiel-Marin 2012: fig. 8), 163298, 163299 (Wiel-Marin 2012: fig. 9), 163300-163306, 164225-164231, 164387, 168603-168614, 168797, 168851, 168859-168864, 169099.

18. Camerano (some 11 km southwest of Ancona) Two owl skyphoi are known to come from the Picenian Iron Age cemetery. One type B Attic RF example, 10.8 cm in height, was found in grave 10. Canuti dated it to 475-450 BC (Baldelli, Landolfi, Lollini 1991: 90, no. 11). The other is type C, 7.1 cm in height, comes from the area of grave 90, and is dated to 425-400 BC by Scandalli (Baldelli, Landolfi, Lollini 1991: 993-94 no 15). 19. Pianello di Castelbellino (some 31 km southwest of Ancona) A rather worn Attic owl skyphos type B was found in grave 5 in the Picenian Iron Age cemetery. It is 7.9 cm

15. Spina Owl skyphoi have been found in graves 203, 374, 553 (Aurigemma 1960: pls. 151, 195, 209, 212) 740, 981, 1095, 1168 (Aurigemma 1965: pls. 5, 42, 116, 121). Some of them come from the settlement (F. Wiel-Marin, personal communication).

7  For other owls from Bologna see: Johnson 1953: 97 note 9, 101 (Group VI); 1995: pl. 35 fig. 1-18 (skyphoi), and fig 36 fig. 19 (pelike). Altogether some 33 vases decorated with owls.

66

Branko Kirigin: Owl skyphoi around the Adriatic

in height and dated to 475-450 BC (Baldelli, Landolfi, Lollini 1991: 115 no. 13).

is dated to the end of the 5th century BC (Baldoni 1993: 57-69, 105 no. 92, 123).

20. Pitino di S. Severino (some 42 km south of Ancona) A rim-sherd coming from a skyphos decorated with an olive branch comes from the native settlement site. It is dated to 430 BC (Baldelli, Landolfi, Lollini 1991: 140 no 12).

25. Banzi In a rich grave (no. 534) in a Daunian necropolis situated some 56 km inland from the Adriatic Sea (Molfetta), an Attic owl skyphos of type A, 7.7 cm in height, was found together with local painted and BG, Attic painted and BG pottery and a lamp, iron, bronze and gold jewellery, vessels and arms: 47 objects all together. It is dated to the end of the 5th and the 4th centuries BC (Settis and Parra 2005: 433-438 no. III, 334 on p. 436).

21. Numana The Rilli Collection has an Attic kantharos from Numana decorated with an owl between olive branches, dating to 450 BC and attributed to ‘Classe del kantharos Czartorsky’ (Landolfi 1982). In a grave of a ‘guerriero’, no. 123 in the Sirolo necropolis (area Quagliotti) two BG skyphoi with owls between olive braches in white applied colour were found, together with 6 plates and skyphoi of Alto-Adriatico style, three BG skyphoi and a small chous, local olpai and metal finds: a bronze cauldron, three fibulae and three rings, and an iron spearhead. The grave is dated to the second half of the 4th century BC. The two owl skyphoi are 8.2 and 7.9 cm in height, are attributed thus: ‘Più che agli skyphoi etruschi del Sokra Group il nostro esemplare sembra vicino ai prodotti dello Xenon Group’. The smaller skyphos has only three olive branches (Berti, Bonomi, Landolfi 1997: 86-90 owl skyphoi nos. 02.12 and 02.13). The shape and decoration of these two skyphoi are similar to those from Pharos (no. 4 above).

26. Rutigliano In grave 49, described as a ‘sarcofago’ from a Peucetian necropolis situated some 10 km from the Adriatic sea to the south-east of Bari, an Attic owl skyphos type B, 7.3 cm high, was found, along with 27 BG imported and local vases, an amber pendant and a bronze strainer, The grave is dated to the end of the 5th century BC (De Juliis 2006: 146- 49 no 13). 27. Botromagno Two Apulian RF owl skyphoi of Corinthian type with one vertical and one horizontal handle, come from grave no. 2 from a Peucetian necropolis situated near Gravina di Puglia some 50 km from the Adriatic to the south-east of Bari. One is 7.6 and the other 7 cm in height. They are dated to the end of the 5th and the 4th centuries BC (Ciancio 1997: 200 nos. 199 and 200).

22. Canosa Two Red-Figured owl skyphoi of type A come from the ‘Ipogeo dei Vimini’, cella A and B. They were found together with local fine wares with added colour. The one found in cella A dates to the first quarter of the 5th century BC, while the skyphos found in cella B dates to the second quarter (De Juliis 2002: 9 fig. 1, 129 fig. 7).8

28. Egnazia Johnson (1953: 98 note 17) mentions an owl skyphos from Egnazia (CVA IV D r pl. 56 no. 7; non vidi). 29. Oria Oria is a Messapian site situated some 27 km southwest from Brindisi. The published fragments of three or more Attic owl skyphoi dating to the 5th century BC have been found in a sanctuary dedicated to Demeter there. Mannino (2004: 340-341 fig. 2; 2006: 110 nos. 107 and 108, fig. 121-123) mentions ‘Numerosi skyphoi con civetta’ as coming from this site.

23. Salapia A type A owl skyphos, decorated in added colour, 7.5 cm in height, in grey clay, was found in tomb 114 which was excavated in 1968. It dates to the third quarter of the 4th century BC. Unpublished (De Juliis 2002: 80 no. 1). 24. Lavello (Forentum) In a very rich grave (no. 955) of a female in a Daunian necropolis (end of the 6th – beginning of the 4th century BC) situated some 47 km from the Adriatic sea (between Manfredonia and Molfetta) an owl skyphos of local Daunian production, 8.7 cm in height, was found together with objects for libations, ceremonies, cooking and tableware in iron and bronze, a bronze lamp and a candelabrum, 30 fibulae, amber pendants, an ivory disc and 22 fine pottery vessels (Attic and local). The grave

30. Brindisi An Attic owl skyphos of type B, 7.5 cm high, dated to the third quarter of the 5th century BC. Close location and context unknown (Mannino 2006, 43-45, no. 10, fig. 24) 31. Rudiae Johnson (1953: 98 note 17) mentions an owl skyphos from Rudiae (CVA IV D r pl. 56, no. 6 non vidi). 32. Leuca Fragments of an Attic owl skyphos come from the sanctuary site by the seaside at Grotta Porcinara.

8  The original publication of De Juliis, L’Ipogeo dei Vimini di Canosa, Bari 1990 was not available to me.

67

Wonders Lost and Found It is dated to the third quarter of the 5th century BC (Mannino 2006: 77-79 no. 55).

Fiala, F. 1892. Rezultati prethistoričkog ispitivanja na Glasincu u ljeto 1892. Glasnik Zemaljskog muzeja u Sarajevu 4: 389-444. Fiala, F. 1893. Die Ergebnisse der Untersuchung prähistorischer Grabhügel auf dem Glasinac im Jahre 1892: 127-168. Vienna. Forenbaher S. and T. Kaiser 2003. Spila kod Nakovane, Ilirsko svetište na Pelješcu. Zagreb. Gamba, M. 1986. Nuovi ritrovamenti di ceramica attica nel Veneto. Aquileia Nostra 57: 642-664. Govi, E. 1986. Le ceramiche Attiche a vernice nera a Bologna. Bologna. Hidri, H. 2006. Popullsia e lashtë e Durrësit. The ancient population of Durres. Tiranë. Jeličić Radonić J. and B. Rautar Plančić (eds) 1995. Pharos antički Stari Grad (exhibition catalogue). Zagreb. Johnson, F.P. 1953. An owl skyphos in G.E. Mylonas (ed.) Studies Presented to David Moore Robinson on his Seventieth Birthday 2: 96-105, pl. 32b. Saint Louis. Johnson, F.P. 1955. A note on owl skyphoi. AJA 59: 119124, pl. 35-38. Katić, M. 1995. Antička keramika, in Jeličić Radonić and Rautar Plančić: 123-124. Katić, M. 2002. Greeks and the hinterland of Western Balkans, in N. Cambi, S. Čače and B. Kirigin (eds) Grčki utjecaj na istočnoj obali Jadrana: 423-433. Split. Katić, M. 1999-2000 Uvod u proučavanje keramičkih radionica Farosa Opuscula Archaeologica 23-24: 49-58. Kreuzer B. 2010. “…ἐν Ἀθήναις δέ γλαῦκας…” Eulen in der Bilderwelt Athens. Jahreshefte des Österreichischen Archäologischen Institutes in Wien 79: 199-178. Kirigin, B., B. Djurić, P. Kos and B. Teržan 1986. Issa (exhibition catalogue). Ljubljana. Kirigin, B. 1990. Late Gnathia: a glimpse at the Issa case. Πρακτικά της Β’ Συνάντησης για την Ελληνιστική Κεραμεική: 58-65, pls. 27-30. Kirigin, B., T. Katunarić and M. Miše 2005. Salamandrija (otok Palagruža) Hrvatski arheološki godišnjak 2: 283286. Kirigin, B., A. Johnston, M. Vučetić and Z. Lušić 2009. Palagruža - The Island of Diomedes and Notes on Ancient Greek Navigation in the Adriatic, in S. Forenbaher (ed.) The Connecting Sea: Maritime interactions in Adriatic Prehistory (British Archaeological Reports International Series 2037): 137-155. Oxford: Archaeopress. Kirigin, B., M. Miše and V. Barbarić 2010. Palagruža The island of Diomedes. Summary excavation report 2002-2008. Hesperìa 25: 65-91. Landolfi, M. (ed.) 1982. La ceramica attica figurata nelle Marche. Castelferretti. Mannino, K. 2004. Vasi di età classica nella Puglia anellenica, in L. Braccesi and M. Luni (eds), I Greci in Adriatico 2 (Hesperìa 18): 333-355. Rome. Mannino, K. 2006. Vasi attici nei contesti della Messapia (480-350 a. C.). Bari.

Acknowledgements During work on this paper I have received much help from Federica Wiel-Marin from the University of Padova, Katia Mannino from the University of Lecce, Biba Teržan from the University of Ljubljana, Ana Marić from the Land Museum in Sarajevo (closed since Oct. 4th 2012!), Kristina Mihovilić from the Archaeological Museum of Istria in Pula, Alan Johnston from the Archaeological Institute in London, Ian McPhee from the Trendall Research Centre in Victoria, Domagoj Perkić from the Conservation department in Dubrovnik, Miha Mlinar from the Museum at Tolmin and Toni Ivančić from the Archaeological Museum at Split. To all of them I owe a debt of gratitude. Work on this paper was supported by the Ministry of Science, Education and Sports of the Republic of Croatia through project no. 244-2440820-0810. This paper was written in 2012. Bibliography Aurigemma, S. 1960. La necropoli di Spina in Valle Trebba 1. Rome. Aurigemma, S. 1965. La necropoli di Spina in Valle Trebba 2. Rome. Babić, S. 2004. Poglavarstvo i polis. Beograd. Baldelli, G., M. Landolfi, D.G. Lollini (eds) 1991. La ceramica attica figurate nelle Marche (exhibition catalogue). Ancona. Baldoni, D. (ed.) 1993. Due donne dell’Italia antica. Corredi da Spina e Forentum (exhibition catalogue). Padova. Berti, F., S. Bonomi and M. Landolfi (eds) 1997. Classico Anticlassico. Vasi alto-adriatici tra Piceno, Spina e Adria (exhibition catalogue). Bologna. Ciancio A. 1997. Silbìon. Una città tra greci e indigeni. Bari. Čače S. and L. Šešelj 2005. Finds from the sanctuary of Diomedes on Cape Ploča: new contributions to the discussion of Hellenistic period on the east Adriatic. Illirica antiqua: 163-186. De Juliis, E. 2002. La ceramica sovraddipinta apula. Bari. De Juliis, E. (ed.) 2006. Catalogo del Museo Nazionale di Taranto 2.2. Rutigliano 1. La necropoli di contrada Purgatorio. Scavo 1978. Taranto. de Marinis R.C. (ed.) 1986. Gli Etruschi a nord del Po (exhibition catalogue Mantova 1986-1987). Mantova. de Marinis R.C. and M. Rapi (eds) 2007. L’abitato etrusco del Forcello di Bagnolo S. Vito (Mantova): le fasi arcaiche, (exhibition catalogue Bagnolo San Vito 2005, 2nd edn). Firenze. De Min, M., M. Gamba, G. Gambacurta and A. Ruta Serafini A. 2005. (eds) La città invisibile. Padova preromana. Trent’anni di scavi e ricerche, Ozzano Emilia. Bologna. 68

Branko Kirigin: Owl skyphoi around the Adriatic

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: 499-519. Split. Mihovilić, K. 2004. Ceramica greca in Istra, in L. Braccesi and M. Luni (eds) I Greci in Adriatico 2 (Hesperìa 18): 101-121. Rome. Mlinar, M. 2002/ Nove zanke svetolucijske uganke. Arheološke raziskave na Mostu na Soči: 2000 do 2001 / New stigma to the Enigma. Archaeological excavations at Most na Soči: 2000 - 2001. Tolmin. Parović-Pešikan, M. 1960. O karakteru grčkog materijala na Glasincu i putevima njegovog prodiranja. Starinar 11: 21-45. Parović-Pešikan, M. 1985/1986. Grčko-italske i helenističke vaze iz Zemaljskog muzeja u Sarajevu. Glasnik Zemaljskog muzeja u Sarajevu 41-42: 39-60. Perkić, D. 2010. Svetište u Vilinoj špilji iznad izvora rijeke Omble, in Antički Grci na tlu Hrvatske (exhibition catalogue): 158-161. Zagreb. Popović, L.B. 1994. Collection of Greek Antiquities. National Museum Belgrade. Belgrade. Schauenburg K. 2001. Eulen und der Vögel, in Studien zur unteritalischen Vasenmalerei 3. Kiel. Settis, S. and M.C. Parra (eds) 2005. Magna Grecia. Archeologia di un sapere (exhibition catalogue). Milan. Sparks B.A. and L. Talcott 1970. The Athenian Agora 12. Black and Plain Pottery of the 6th, 5th and the 4th Centuries B.C. Princeton. Šešelj, L. 2009. Promunturium Diomedis: svetište na rtu Ploča I jadranska pomorska trigovina u helenističko doba, PhD dissertation, University of Zadar. Tarditi, C. 1996. Vasi di bronzo in area Apula. Produzioni greche ed italiche di età arcaica e classica. Lecce. Watson M. 1998 The owls of Athena: some comments on owl-skyphoi and their iconography. Art Bulletin of Victoria 39: 35-44. Wiel-Marin, F. 2005. La ceramica attica a figure rosse di Adria. La famiglia Bocchi e l’archeologia. Padova. Wiel-Marin, F. 2008. La ceramica d’importazione, in M. Baioni and C. Predella (eds) Archeotrade. Antichi commerci in Lombardia orientale (exhibition catalogue 2008-2009, Rete Museale “ma_net”): 236-248. Milan. Wiel-Marin, F. 2012. Adria e Forcello: Alcune considerazioni sulla ceramica attica figurata proveniente dagli abitati, in D. Paleothodoros (ed.) The Contexts of Painted Pottery in the ancient Mediterranean world (seventh-fourth centuries BCE): 93100. Oxford. Wiel-Marin, F. (in press) Ceramica attica dagli abitati d’altura del Veronese. Wiel-Marin, F. 2015. La ceramica attica degli abitati a nord-est del fiume Po, in Bonomi S. and Guggisberg M. (eds) Griechische Keramik nördlich von Etrurien: mediterrane Importe im archäologischen Kontext (Atti del Convegno Internazionale Basilea 2011): 45-65. Wiesbaden.



69

Gyenus on stage: civic foundation and the comedy of Aristophanes’ Birds David Braund This paper addresses two large themes which ultimately will come together around a single phenomenon, namely the foundation of cities. For in what follows I consider first the ubiquitous theme of city-founding across Greek culture, with particular reference to Aristophanes’ Birds and the complex of comic themes in that play. Secondly, I shall then offer some reflections on the foundation story of a particular settlement, namely Gyenus in ancient Colchis on the east coast of the Black Sea, otherwise known as Cycnus. Readers will immediately perceive that, with regard to both these themes, we shall be dabbling in the pool of research to which Michael Vickers has made a series of notable contributions.

might have done and how they might have fared with a different society and, in particular, a different political system. Meanwhile, Thucydides in his critical analysis of Athenian democratic imperialism offers little encouragement for any idea that a Persian democracy would have been less imperialistic or better governed or more successful in other terms than had been Persian monarchy. We should not be surprised or persuaded by the fact that Athenian tragedy tended – implicitly and explicitly – to assert the superiority of democracy in general and of the democracy of Athens in particular, whether by comparison with Persians (as in Aeschylus’ Persians) or by comparison with other Greek poleis, such as Thebes, for example in Euripides’ Suppliants. As we shall see, Athenian comedy is more interesting, in the fifth century at least, because it offers mockery, criticism and carnivalistic rejection of that largely conformist tragic vision.

Greek culture and civic foundation In antiquity the foundation of cities was both commonplace and extraordinary. For Greek culture was very much a culture of settled communities, whether named as poleis, emporia, villages or some other term. Most famous (and most often misused) is Aristotle’s assertion (Politics, 1253a) at the beginning of his analysis of the state, that man is a ‘political animal’, words more accurately to be translated as a ‘polis creature’. Indeed, the whole work might better be translated (however inelegantly) not as Politics, but as ‘Polis matters’. In fact, Aristotle’s work stands in a long tradition of Greek thought about how communities should be created and operated, with the question of the best political system as a central theme. Indeed, such issues of ‘politics’ in this broad sense were at once so widespread and so fundamental to the thought and reality of the Greek world that they regularly appear also as dominant themes elsewhere in Greek culture too, not only in dedicated political philosophy. Herodotus’ Histories, for example, not only includes a set-piece debate about the best political system (3.80-82), but also embraces the whole range of issues about human social organization and behaviour, with the Persian Wars as a clash of political cultures. As one important feature of that expansive concern, the set-piece debate on political systems emerges as foundational in the sense that it sets the future direction of Persian rule as monarchical, while also airing the problems inherent in that course, which appear subsequently and starkly enough in Persian kings’ contributions to Persian failures in wars of imperialism, notably in Ethiopia, in Scythia and, ultimately, in Greece. In that way the Histories leave open the question as to what Persians

Already we begin to see the interplay of ideas (sometimes extremely theoretical) and everyday reality. Much of that entailed the close consideration of actual communities, whether Greek or otherwise. The Aristotelian programme of collecting such ‘constitutions’ seems to have been part of the philosophical endeavour that underpinned the writing of the Politics. Among the 158 case-studies (as we might now call them) gathered for Aristotle’s project, were certainly non-Greek examples. We should pause to note that, in this as in much else, Greeks were very ready to learn from non-Greeks, whatever other attitudes they might harbour towards ‘the barbarian’. While the Aristotelian programme was unusual in scope, the tendency to such comparison is already clear in earlier philosophy, as in Plato’s interest in Sparta and Carthage, notably in his own Politeia (unhelpfully known as his Republic, but also a work on the polis). And Herodotus, again, is replete with comparisons that underpin his overarching analysis of human society (notably clear at Hdt. 3.38). His Histories are especially helpful in showing how the interactions of Greek and non-Greeks, with their various forms of community, might serve to answer the largest questions not only about matters of community, or ‘politics’, but also about the human condition as a whole phenomenon. Again, we should not be surprised that other Greeks might recoil from his inclusive approach, as famously did Plutarch at the time of renewed Greek chauvinism under the Roman empire c. AD 100. 70

David Braund: Gyenus on stage: civic foundation and the comedy of Aristophanes’ Birds

In modern studies of these matters, understandably enough, there is much focus on political theory, historical writing and such documents as may be brought to bear. However, in view of the ubiquity of the theme of civic foundation, it should not surprise anyone that much may also be gleaned from closer consideration of quite different creations of Greek culture. A case in point is the Birds of Aristophanes, first produced at Athens in spring 414 BC. Here of course we have Old Comedy, with all its scope for fantasy, ribaldry and the rest. And yet that ‘rest’ includes also political comment of various kinds and, more broadly, a strong and critical engagement with the real world of democratic Athens of the later fifth century in which such plays were fashioned and first performed. That range and admixture of topics and approaches in a single play or group of plays, all directed to the creation and celebration of comedy, can leave scholars in a daze of uncertainty and even confusion about how to interpret (one might say, ‘use’) such plays to advance understanding of history and political thought. Accordingly, interpretation is fraught with controversy on the most fundamental of points, while there seems to be no slowing of the torrent of modern publications about Old Comedy, nevertheless.

Birds, despite the fact that it was first performed almost twenty years earlier. The two Athenians stress their patriotism to Athens, which mitigates their rejection of life in Athens. They are not anti-Athenian, but they are unable to bear the various burdens of their Athenian life, which are those regularly complained about in Old Comedy, with the lawcourts very much to the fore. The audience no doubt felt the sympathy needed to sustain the comedy, for the play was no failure: the hypothesis informs us that Birds achieved a second place in the competition, behind Ameipsias’ Komasts but ahead of Phrynichus’ Monotropos. In that spirit they want Tereus to advise them of an alternative to Athens, where they may live without endless litigation and the like. Why Tereus? In plot terms, because he has been a man and so will understand human desires, while he is also a bird. Aristophanes makes clear the underlying notion that birds fly far and wide and so will know about the world at large much more than a human might. The idea is not peculiar to Aristophanes. It suffices to consider Aristeas of Proconnesus and his archaic epic on the Arimaspians, a one-eyed people of the distant north, who were imagined as locked in conflict with neighbouring griffons over their gold. Although we have only a very few fragments of this work, it is clear that its author, Aristeas, claimed to have flown far and wide across the world, surveying an array of obscure cities, peoples and places (Bolton 1962). Herodotus reports that he was honoured at Metapontum in the form of a bird, specifically a crow. Aristeas claimed too that he had been taken on this extraordinary journey by Apollo himself – the god at once of poetry, the obscure north (including the Hyperboreans) and distant settlement or colonization in general (Hdt. 4.13). No doubt we glimpse in the Proconnesian something of the broadly Orphic tradition that appears for us elsewhere around the western Black Sea and its environs, notably in the riddling bone plaques of Olbia (and Berezan, apparently), the Getan Zalmoxis and the Derveni papyrus (further, Barr-Sharrar 2008). We need not suppose that Aristeas’ corvine has any direct link to the jackdaw and crow with which the two Athenians begin the play, their unsatisfactory avian guides to Tereus. The members of the crow family are among the cleverest of birds, and can also ‘talk’: to that extent they were plausible guides, while there was eminent scope for humour about individuals or the general class of flatterers (kolakes/korakes: Dunbar 1995: 143). A particular relevance for Apollo in Birds may be claimed, however, insofar as the god was not only associated with Thrace (important to the play: see below), but seems also to have been key to Sophocles Tereus, whose significance for Aristophanes’ play is made explicit (Birds, 100). All the more so if we accept the plausible argument that Apollo appeared ex machina at the end of Tereus (Fitzpatrick 2001, esp. 100).

Reflections on Aristophanes’ Birds My purpose here is not to engage in detail with the mass of such scholarship, which Michael Vickers memorably described as ‘scholarly rubble’ when taking into account the whole interpretative tradition on Old Comedy (Vickers 1995: 339). Rather I shall draw attention to issues within Birds which seem to contribute to our understanding of the play and the role of civic foundation within it. The plot develops with a simple logic that has caused controversy of its own, quite unnecessarily in my view. The play begins with two elderly Athenians, who are making their way across hard country, which is not given any specific location geographically. They are so dissatisfied with life in contemporary Athens that they seek a new community. For information on the best place to go they are journeying in search of Tereus, the hoopoe. In myth, and indeed in Athenian tragedy, Tereus had been a king of Thrace. He had been turned into a bird – specifically the hoopoe – in consequence of a gruesome tradition, whereby he had cut out the tongue of his sister-in-law Philomela to stop her telling his wife, her sister Prokne, about the fact that he had raped her. Philomela communicated his crime to her sister, even so, by means of her weaving. And so Prokne punished her husband Tereus by serving up the flesh of their son, Itys, as a meal for him. All three were turned into birds as Tereus set about killing the sisters: Tereus – the hoopoe, Prokne – the swallow, and Philomela – the nightingale. At Athens, Sophocles had written a tragic Tereus in 432, which is key to understanding 71

Wonders Lost and Found Once the two old Athenians have reached the marvellously-absurd Tereus, their purpose soon develops from the search for a new home somewhere to the foundation of a new city. Again, there is an affirmation of Athenian patriotism here, on two counts. First, there is nowhere better to live than Athens, unless one starts to consider the fabulous world beyond Egypt towards Arabia and India (Birds 145; so fantastical as to suit the claims of the Sausage-seller at Knights 10869). The Athenians rule out such distant adventures to the mysterious east. We may well observe the absence of any wonderful place to go in the west, whether in Sicily or elsewhere. Secondly, there is much of Athens in the walls and institutions of the fantastical city of the birds which ultimately will be their settlement among the clouds (reasonably translated Cloudcuckooland). And the new foundation will resemble, at the same time, a community which resides within the Athenian empire, albeit with an assertive independence from the officialdom that seems to have been unpopular in Athens itself as well as among those states of the empire. All that was vital to the play’s chances of success before an Athenian audience. The new foundation is a reconfigured Athens, comically built in the air, where it still lay within the long reach of burdensome Athenian officialdom.

exploits the easy opportunity to have fun with the politics of boasting exotic birds in democratic Athens (cf. Cartledge 1990). The Athenian audience might well see Athens at every turn of the comedy, especially when prompted by Aristophanes’ recurrent links to their city and its practices. Tereus himself was current enough in Athens of 414, not only because of the Sophoclean tragedy some twenty years earlier, but also because of his role in connecting Athens with Thrace. For we have seen that Tereus the bird had been not only a human, but also and more specifically a Thracian king. His former royalty may help to explain his apparently elevated position among the birds, having even his own slave (it may matter that a slave was evidently prominent in opening Sophocles’ play: Fitzpatrick 2001). Dunbar likens him to a tragic warrior-king at Birds 270. Aristophanes does not bring out that image, but he hardly needed to do so for his Athenian audience. The myth of Tereus entailed his marriage to an Athenian woman, Prokne and savage abuse of her sister, Philomela, also Athenian, for these were the daughters of King Pandion of Athens. Moreover, around the time of Birds (close dating is impossible) the sculptor Alkamenes dedicated a statue group on the Athenian acropolis, depicting Prokne about to kill her son Itys to punish Tereus (Paus. 1.24.3). The substantial remains of this piece, presumably Alkamenes’ own work, has survived and is now displayed in the Acropolis Museum at Athens. There must have been some public discussion of the dedication, even if the sculptor himself met all the costs. Indeed, there is a faint possibility that we have even an allusion to the sculptor in the play itself at lines 272-3. For there we have the arrival of a red-winged bird (phoinikopteros, usually taken to be a flamingo), evidently a Spartophile, if not an avian Spartan. However, he is specifically a bird of the marsh, limnaios, which is a rather curious way to refer to a Spartan, notwithstanding the limne there. The problem is soon echoed by the appearance of a Mede-bird, who is a creature of the mountains, for reasons that remain similarly unclear, despite Dunbar’s best efforts. What is clear is our inability to grasp the humour in full. Much would probably be clarified if only we had context for Cratinus’ allusion to a phoinikopteros in his Nemesis, probably of 431 and appropriately concerned with Sparta (Bakola 2010). The case for Alkamenes rests on his evident concern with Prokne and consequent special relevance to Birds, together with the complex question of his origins, which might well make him limnaios and so give some point to the adjective (cf. Stewart 1990: 267-9). If he is our man/bird, it would follow that he is in some sense a Spartophile, for the coincidence of his name with an early Spartan ruler hardly suffices to explain his red wings. Questions overwhelm answers.

(a) Thrace The world of the birds is poised in the air between earth and the Olympian heavens. Of course, in comic logic that is no obstacle to the construction of great walls around the new city, very much castles in the air and again reminiscent of Athens and its famous walls. The whole undertaking was first agreed by an assembly of the birds themselves, summoned and presided over by Tereus, whose role is intriguingly ill-defined. As a former human, he has special knowledge, so that he is able to teach the birds Greek (Birds, 199-200 with Dunbar 1195: ad loc., who might have said more about the joke of this Thracian as Greek-teacher and probably also the theme of communication so important to the Tereus myth). Moreover, Tereus sports a prodigious costume with absurd crests (though shabby feathers, since he is in moult) that seem to show some remnant of his human kingship. At any rate he has sufficient authority and skill to keep a kind of comic order in the bird-assembly which he has himself called. In general, as bird and former human Tereus is a gobetween for the Athenians with a collection of birds that are notably hostile to humankind in general and to the Athenians before them in particular. Again we are among birds, but also in a kind of Athens, with speakers before a rowdy and aggressive bird-public, such as we see, for example, in Acharnians. Here it is the persuasive Athenian Peisetaerus who is the envoy from another land (Birds, 255), while the assembled birds are at once foreign and Athenian: Aristophanes

We are on much firmer ground, however, in recognizing also the enormous importance of Thrace and Thracians 72

David Braund: Gyenus on stage: civic foundation and the comedy of Aristophanes’ Birds

for Athens during the Peloponnesian War. In the later 420s, of course Thrace had been the principal theatre of war, and Aristophanes unsurprisingly exploited that fact to comic effect both in general and in detail (further, Braund 1999). He had had much fun too with the Athenian use of Thracian mercenaries, notably the Odomantians of Acharnians (see esp. Vickers 1997, ch. 4, with Fragoulakis 2013: 263). Their name somewhat resonates with Birds, where we find the Mede-bird as mousomantis, which (as Dunbar notes) was taken from a Thracian context: it seems to have indicated Orpheus in Aeschylus Edonians (fr. 60). Aristophanes had a taste for such poetic allusions on Thracian matters (cf. Ar. Thesmo. 134ff.).

To that extent we may compare Tereus of the early part of the play with another Thracian archetype who appears with some prominence in its closing section, namely the Triballian deity. For the Triballians were a real people, who belonged to the Thraceward region, like mythical Tereus. Given the history of Athenian involvement in Thrace, we should be optimistic about Athenian knowledge of this important people of the hinterland (broadly western Bulgaria and part of Serbia, as it seems). After all, the Triballians were responsible for the death of Sitalkes himself in 424 BC, as Thucydides reports (Thuc. 4.101: cf. Vickers 1989, 296, inferring that the Triballians were Spartan allies). That will have attracted Athenian attention. They are well characterised in Birds as inhabitants of a barbarian interior beyond the Greek-settled coast: the Athenian empire could not ignore them, as their exploits of the fourth century demonstrated well enough (see Graininger and Valeva 2015).

All this contributes in different ways to the powerful resonance of Tereus himself in Athenian relations with Thrace. Of course, Thucydides famously insists that the Tereus who married Prokne was not a king of Thrace, but ruler of Daulis (alias Phokis). However, it is well understood that his protestation indicates that others in Athens claimed quite the opposite, and that the TereusProkne link was part of the diplomatic underpinning to the various dealings of the 420s, in particular, which included the bestowal of citizenship upon Sitalkes’ son, Sadokos (Thuc. 2.29; cf. 67 for an example of the benefit to Athens). In all likelihood herein lies the rationale for Alkamenes’ Prokne and Itys on the acropolis, too. Once again there was scope for comedy, which Aristophanes memorably exploited with his extraordinary image of Sitalkes scrawling ‘the Athenians are beautiful’ on the walls of his palace in Wasps. Here too there may be more than comic nonsense (pace MacDowell 1995: 51). For Xenophon (Anab. 7.3.39) shows enough of royal Thracian enthusiasm for Athens (however superficial) to indicate that there may be some substance in the image of an enthusing Sitalkes, when he writes of another Thracian king giving the watchword as Athena on account of his kinship with the Athenians. For the Tereus myth gave a basis for kinship with Athens and so facilitated the award of Athenian citizenship. Xenophon’s king had earlier declared his faith in the loyalty of the Athenians as being his kin (7.2.31), because he claimed Tereus as his ancestor (7.2.22). The egregious behaviour of the comic Sitalkes may catch something of the flavour of Thracian rhetoric in enthusing on the convenient mythical bond with Athens, which would help to explain the sour remarks of Thucydides, no stranger to Thrace and Thracians. While in Birds there is no insistence on Tereus’ Thracianness, we may presume that the poet shared in the usual view, not the revisionism of Thucydides. His explicit description of Tereus as Thracian in returning to him in Lysistrata (563) a few years later surely confirms as much, while showing too that the theme of Thrace was still of key importance in Athens. Indeed, in that play Tereus is mentioned not simply as a Thracian but as a Thracian par excellence.

There can be no doubt that Triballians figured in Athenian discussions of Thrace, and it may well be that they were especially topical in 414 (so MacDowell 1995: 217, who nevertheless thinks that Athenians knew little about them). Indeed, they evidently had a reputation for abnormal (one might say, barbarian) eating in sacrificial contexts, for we hear (Dem. 54.39) that a group of young Athenians who did as much in early fourth century Athens were therefore known as ‘Triballians’. In similar vein a fragment of Alexis (fr. 243) cites the Triballians as the nadir of bad behaviour in sacrificial dining. A little later, Aristotle (a man of Chalcidice) says that among the Triballians it was good to sacrifice one’s father (Arist. Top. 115b; cf. Hdt. 3. 38 and Dunbar 1995: 702). Evidently Triballians were the stuff of comedy in regard to their sacrificial conduct. Moreover, there is every reason to suppose that the prominent Triballian of Birds was no shock to the audience in that context, not least because Aristophanes did not feel the need to explain the Triballian god’s particular relevance to sacrificial consumption and so his appropriateness in this scene of divi hunger now that Cloudcuckooland had blocked the way (further, Kanavou 2011: 125 n. 558, where general talk of savagery is unhelpful; cf. Sidwell 2009: 266-7 on authorial games at stake here). The Thraceward region had many resonances at Athens. Moreover, Athens continued to exploit the manpower and other resources of Thracians. Athenian claims to control the Greek cities of Thrace demanded as much, while Thracians were also vital to the whole issue of the emerging Macedon, whose resources were also of great importance. Thrace had long been important to Athens, but the war had given it a new importance, which had in turn brought a fresh focus on Tereus. Accordingly, Sophocles’ production of a play on the theme in 432 might well have been affected by contemporary issues, not least concerning Amphipolis. In much the same way, Euripides had chosen to produce a Medea in 431, 73

Wonders Lost and Found where Corinth was a major focus for Athenians on stage as in life.

notion of world-travelling birds and the two Athenians in search of the hoopoe. Thrace apart, we find occasional allusions to other parts of the world, such as the Red Sea (above) and Libya, together with occasional foreigners, like the Mede (above) and the Persian. The noisiest absence, of course, is Sicily. There has been ample consideration of the possible reasons for the play’s failure to exploit the wealth of humour to be had there. However, the nearest we come to a reference to Athens’ principal opponent there, Syracuse, is not a reference at all: at 1297 we have Syrakosios. His name echoed the city, while a probable explanation for it is his family’s links there. He is the possible author of a much-discussed measure limiting comedy, while he was himself known as the raucous jay, presumably in much the same way that Kleon was a barking dog, both adjudged by their critics to be improper speakers in the Assembly (cf. Dunbar 1995: ad loc.). It is at least an interesting coincidence that he appears with another criticised political figure who might also evoke Syracuse. This was a certain Meidias (ibid.), who is known as Quail, apparently in view of his fondness for quail-fighting and perhaps his general pugnacity. For the common ground that might have been exploited in the play with regard to Sicily and birds, is specifically the important part of the city of Syracuse that was Quailisland, or Ortygia. All this hovers over these lines of the Birds, but what is far more striking is that none of these opportunities for humour seems even to be attempted, let alone cashed to good effect. Even those who perceive allegory in the play that bears on the Sicilian expedition (such as one might reasonably anticipate in view of the scale, ambition and controversiality of that expedition) must allow that there is little strongly to connect the event with the comedy. As MacDowell points out, we have only two passing jokes about Nicias (lines 363, 639), which are relevant only insofar as he was with the expedition at the time of the play’s first performance, plus a mention of the Salaminia (lines 145-7). The latter is so evocative of the recall of Alcibiades that it is the strongest and surest allusion to current events bearing on the expedition (MacDowell 1995: 223). For whatever reason, Aristophanes chose not to engage directly with the expedition, while the case that he has done so by profound allegory is very hard to sustain without stronger clues in the play itself. One wonders whether the expedition was a matter of such anxiety among many Athenians, who had family members far away in Sicily, to make critical humour too uncomfortable. In his Komasts, Ameipsias may have directed his humour against the wealthy young men who had parodied and profaned the Mysteries, perhaps together with the mutilation of the herms. But the great expedition was still in train and quite probably no laughing matter for most Athenians. Instead the poet directed his humour at the aspects of contemporary reality which had been his targets before, and which seem to have appealed to his audience, primarily the failings of the democratic

Certainly, Thrace figures rather prominently in Birds, quite apart from Tereus (cf. also below on the swans of the River Hebrus). At lines 1040-1 the Decree-Seller announces that Cloudcuckooland may have the same weights, measures and decrees as the Olophyxians. The mention of Olophyxians is a surprise: the audience expected ‘Athenians’, for theirs were the standards to be used (Mattingly 1999: 712). ‘Olophyxians’ injects an extra note of absurdity into the scene. And a pun soon follows (1042), but that was not the only reason to adduce the small community of the Olophyxians, located on the Athos promontory of Chalcidice. Evidently, there had been some particular discussion of the place that was still in Athenian memories in 414. By now Athenian weights and measures were common across the empire after the Standards Decree (now understood to be no later than the 420s: further, Vickers 1996; cf. Lane Fox 2010), but the mention of decrees, which has puzzled commentators (notably Dunbar 1995: ad loc), surely suggests particular decisions taken by the Athenians in regard to the Olophyxians. Once again we seem to have a Thraceward reference that we cannot fully grasp on the present evidence, except that we may suspect the relevance of the town’s return to Athens after flirtation with Brasidas in the late 420s (Thuc. 4.109). Finally, we can do no more than observe the curious coincidence (if it is no more) that it was in the vicinity of the Olophyxians on the Athos peninsula that another city of the heavens, called Ouranopolis, was established about a century after Birds, also apparently an experiment in ideal civic foundation (e.g. DuBois 2006). Thrace as a whole is only mentioned once in Birds, and late in the play. However, there we find its association with war very clearly expressed. The Thraceward region is the place to go if you want a fight. For Peisetaerus advises the Father-beater not to take his aggression out on his father, but instead to go to fight in Thrace (Birds, 1369). The advice only makes sense if the audience understood Thrace to be full of such opportunities for fighting in 414. Meanwhile, we may readily understand that the turmoil caused in Thrace by Brasidas took time to undo in the region, whatever the Peace of Nicias might have ordained from afar. The case of the Olophyxians, with their rather comic name, was doubtless not an isolated one there. And, finally, if the Father-beater is indeed in some sense Alcibiades (e.g. Vickers 1995: 346), the suggestion that he head for Thrace has a special relevance, given Alcibiades’ famous Thracian connections. (b) Sicily In fact there is a noteworthy lack of geography in Birds, given the potential for exotic places that resided in the 74

David Braund: Gyenus on stage: civic foundation and the comedy of Aristophanes’ Birds

superstructure and those who seemed to prosper by manipulating its onerous ways to their own advantage.

of Birds was precisely not like those other foundations, themselves different enough from one to another. For this comic Ouranopolis was very much the kind of private initiative that Aristophanic comedy enjoyed so much: in that sense the two Athenians resemble Trygaeus on his dung-beetle mission, another aerial adventure in which the ordinary man of (comic) good sense takes matters in hand.

(c) Founding Cloudcuckooland In other Aristophanic plays the comic alternative foregrounded the initiative of individuals who cut through the ‘red tape’ with an appropriately fantastic scheme – making a private peace in Acharnians, bringing the divine Peace in Peace and so on. Peace had been the key issue in the 420s when war was raging, and it would be again soon enough with Lysistrata. In 414 war persisted, but in a distant and untested zone of conflict in far-off Sicily, where Athens had just committed a large force: the achievement of peace was an improbable theme and so private initiative in Birds centres on systemic failings outside the realm of peace and war. In that sense, Birds may recall Wasps or look forward to, for example, Thesmophoriazusae, or further forward still to Ecclesiazusae. For the cunning plan of the two Athenians in Birds is to found a new city which will be free of the oppressive burdens of the contemporary system in Athens itself. At the same time, as we have seen, the poet is careful to insist on the powerful patriotism within which this initiative is to be carried through. This is not to be a foundation directed against Athens (though ultimately it must have that result), but rather to create a city that would reproduce Athens without much of its undesirable superstructure, practices and individuals. We have observed how in that sense it was a reproduction of Athens, while the (ludicrous) imperialistic ambitions of the new foundation may also seem to express something of imperialism Athensstyle, with even the Olympian gods to pay tribute. Some in the audience may have reflected on the imperialist ambition inherent in the Sicilian expedition, but there was little encouragement to do so in the language and action of the play.

The private foundations of cities that we can hope to discern in the later fifth century are foundations of the mind. We must seek them in the ideological debates that lie behind Plato’s Republic. Perhaps we should consider in passing also the strong tradition that it was precisely in Sicily that Plato later sought to bring his theoretical vision of the best state into reality at the court of Dionysius I of Syracuse. Meanwhile, it would be good to know much more in detail about the political aspects of the vibrant philosophical debates that characterized Athens of the later fifth century. We should take seriously Plato’s location of the debates of the Republic in that time, in the broad context of the Peloponnesian War and with Socrates to the fore, even if we must also refrain from any suggestion that these were debates which took place then in quite this way. For, however we nuance the matter, there is no real doubt that Plato’s work is deeply rooted in the intellectual exchanges of the last decades of the fifth century. Accordingly, although this line of interpretation has been rather neglected, it is important too that we locate the comic utopia of Cloudcuckooland in the larger history of utopian thought of which the Republic is a part (Dawson 1992; Morrison 2007). In that way Birds comes to resemble another Aristophanic play, Clouds, and thereby shares in a broader comic trend of mocking the innovative ideas of contemporary thinkers, with or without their associates among the elite youth of the city. Accordingly, while the foundation of Cloudcuckooland might offer a comic alternative to Athens, it also offered ample opportunity in its very absurdity for fun too at the expense of these well-heeled intellectual circles. And of course these were also deeply implicated in the shenanigans surrounding the despatch of the great armada, possibly (as suggested above) the focus of Ameipsias’ Komasts. For all those reasons in Birds we are faced with a compressed and of course highly comic account of a foundational process, as imagined in the late fifth century. To that extent, the play develops its fantastical comedy around a process of civic foundation that was fundamentally a reality. And as an actual foundation would have demanded, the new community is inaugurated with a religious ceremony, a sacrifice and dining. In this case, however, the proceedings are comic. In particular, the deities invoked are comically bird-related forms of familiar gods appropriate to such a context: in turn, Hestia, Poseidon, Apollo (the Pythian and Delian swan), Leto and Artemis, Dionysus and Cybele, and a host of birds. Peisetaerus presides, with

Instead the play maintains its focus on the foundation of a new city in the sky. It would be convenient to know of an Athenian plan to plant a colony c. 414, but none is known to us. Conceivably there were those at Athens who thought that a new foundation might be planted in Sicily in consequence of a successful expedition there. Certainly, new foundations were not to be excluded from Athenian imperialistic thinking. We happen to know at least of the foundation of Brea (once more in Thrace), whose date remains unclear, but may well be in the 430s (see above). Foundation had been central to Athenian imperialism in Thrace, where the foundation of Amphipolis had been a major event of recent decades too. Elsewhere, we may add also Athenian colonial adventures on the south coast of the Black Sea at Sinope and perhaps to its east at Amisus (Braund 2005). Accordingly, in 414 Athenians were no strangers to the realities of civic foundations, even if we hear of none in very recent years. Crucially, however, the foundation 75

Wonders Lost and Found a comic measure of grumpiness, while a corvine priest performs the meagre sacrifice and a series of greedy individuals comes and goes (Birds 895ff.).

shortly to be completed as Birds appeared in 414. He had also featured in a scandal which entailed his (for himself or for his son) evasion of military service in the expedition to Sicily (e.g. Plut. Nic. 13; Alc. 17). Indeed, his topicality is confirmed by his appearance in a fragment of Phrynichus’ Monotropos, which was produced in the very same festival as Birds, to which it came third. There (fr. 22) he is said to be ‘the one who brings fountains’, which has been taken to mean his involvement in the water supply of Athens. However, we may have a joke here: his famous evasion of military service was based upon his burning of his own house as an act to show his (feigned) madness. In that context critical comedy might well jibe at Meton as a water-bringer. In 414 he was famous for not bringing water to douse the fire he had set in his own house, whatever he may otherwise have done or not done with fountains. In fact, it remains unclear how much Meton had to do with the actual design of cities and matters of construction: Dunbar (1995: 551) may very well be right that he is introduced here because this is a foundation in the air, where his interests certainly lay. That would be concordant with his scheme to plan the new city in the shape of a star (Birds 1007). The creation of Cloudcuckooland required not an urban planner but an astronomer.

These several strands of humour are duly played out as the unwelcome newcomers are rebuffed in turn from the sacrifice and new foundation. While modern sensibilities might find some of these much more clearly undesirable than others, the comic outlook here sees them all as part of the problem that the new foundation has been built to avoid. Their rejection from the sacrifice (and the opportunity to feed there) further expresses that this is indeed an alternative to Athens. Strikingly, the first to arrive at the new foundation is an unnamed poet (Birds 905ff.). His prominence here may be explained in many different ways: for example, it may be important that the play itself is by its very nature a poet at work in a civic context. The unnamed poet, however, has come in search of gain. He has a range of verse to offer, evidently to aggrandize the foundation. It is especially interesting that he is depicted as immediately suggesting the new city to be an old one. That was part of aggrandizement, while expressing also his deceitfulness in pursuit of profit. We shall return to those verses shortly, for their significance is considerable.

Bogus wordsmiths – poetic and oracular – had been sent on their way unceremoniously, and the dubious stargazing intellectual had also been driven off. Next was the rejection of officialdom itself in the person of an episkopos (Birds 1021-34). Peisetaerus wants no part of him and drives him off too. Here we need not suppose the audience’s sympathy with its imperial subjects. While we cannot discount a measure of such sympathy, nor the more calculating analysis whereby such officials caused problems for the empire, the larger point is that the episkopos embodies an official and officious burden inherent in the democratic superstructure that is not wanted by anyone, Athenians and allies alike. Presumably in the event of success in Sicily, there would be a busy band of such men there. Again, however, the alternative Athens of Cloudcuckooland has no truck with them: they do not represent what is good about Athens. Arguably worse is the decree-seller who soon follows (Birds 1035-57), exemplifying the faults of the system by which both Athenians and allies live. He too receives short shrift. A choral parabasis rejoices in the new foundation and marks the end of this first stage of the foundation process.

As the poet departs another kind of charlatan appears, an oracle-monger (Birds 959ff.). If the poet sought to invent a civic past, the oracle-monger seeks to prosper by his absurd representations of divine will and so the city’s future. The audience presumably enjoys seeing him beaten and driven off by Peisetaerus, claiming his own kind of oracle from Apollo, written in his own book. Of course, adherence to such a book of sacred learning might well seem humorous in the context of classical Greek religion (cf. also Dunbar 1995: 547). Again, we may wonder how much such men had had to say about the Sicilian expedition: Thucydides tells us of the Athenians fury at them and their kind once they had grasped the shocking fact that the expedition had in fact failed totally (Thuc. 8.1). Such individuals were an easy target at all times, but they were particularly topical in 414. Moreover, they could be mocked without any need to mock or otherwise undermine the expedition in progress. Next to arrive is a certain Meton (Birds 992ff.), who proposes to measure the air. The echo of Clouds is again loud enough: this was the sort of nonsense allegedly taught by Socrates (Dunbar 1995: ad loc. appropriately cites Clouds 201-3). We may well understand why Plutarch brackets Meton with Socrates: neither was a supporter of the Sicilian expedition (e.g. Plut. Alc. 17). And that is but one element of a strong sense of topicality here, for Meton was a real and remarkable man. He became well-known for his 19-year astronomical cycle: that seems to have begun in 432/1 (Diod. 12.36) and was

(d) Topicality and Sicilian allusions We have seen enough topicality in these scenes to force at least a recalibration of the scholarly tendency to insist on the ahistorical emphasis of Birds, as intelligently formulated by MacDowell, for example (above). In particular, Meton’s central inclusion, framed by two wordsmiths and two instances of the state at work, not 76

David Braund: Gyenus on stage: civic foundation and the comedy of Aristophanes’ Birds

only played to the anti-intellectualism of Old Comedy, but also suggested at least impatience with those (like Meton) who did not participate fully in the communal endeavour in Sicily. Of course, that is not at all to claim that Birds or its author favoured the expedition as such: one could mock Meton for his ideas and conduct without going that far. But it was probably hard to laugh at Meton in 414 without thought of his scandalous evasion of the expedition. And that inference is all the stronger when we examine the verses of the poet who had been first to attend. For here too is Sicily.

least take the Sicilian allusion to be more than a matter of chance. While civic foundation was very relevant to the drama, Sicily was very much its great historical context in 414. Meanwhile, we may also wonder whether the armada and the debates surrounding it had made these Pindaric verses somehow current in Athens of the day. Moreover, Aetna and the relevance of Magna Graecia help to explain another feature of Birds, namely the play’s concern with volcanoes. For among the extant plays of Aristophanes it is only at Birds 824 that we find mention of the Phlegraean fields, usually identified in antiquity near Cumae in southwestern Italy. There seems to have been a rival claim in Chalcidice (Dunbar 1995: 494), but we may draw conclusions about its westerly location and association with Aetna itself from Pindar’s verses on the Gigantomachy there in an epinician for a certain Chromius of Aetna (Nem. 1.67). The burial of conquered Giants was taken to account for the unrest below the earth that manifested itself in volcanic activity of all kinds, not least in southern Italy and Sicily (cf. Strabo, 5.4.6). Moreover, in addition to their geographical orientation westwards, Giants were relevant to Birds insofar as the new foundation was also an assault on the Olympian gods, rather as the Giants had attempted in their more brutal fashion. Aristophanes had already associated the Giants with the various pseudo-intellectual notions that were said to be taught by Socrates, ousting the Olympian gods and replacing them with new-fangled deities (esp. Clouds 853). In Birds therefore Giants suited both the theme of anti-intellectualism and the broader contemporary context, much as mockery of Meton had done.

For the unnamed poet begins adapting Pindaric verses (Birds 926-30, with Dunbar 1995: ad loc), and then mentions explicitly Pindaric poetry (Birds 930). In that way Aristophanes ensured that even the least learned of the audience should have a sense of his literary game. From this point the unnamed poet proceeds to offer parody of further Pindaric verses (frr. 105a-b, with Martin 2009: 94-5), rendered suitably comic. Pindar’s original had developed a poetic strategy around the already-familiar notion of the Scythians as nomads who had their houses on wagons, not fixed in cities (cf. the Hesiodic Catalogue of Women, fr. 151, with Dunbar 1995: 537, who lists also Hdt. 4.46; PV 709-10 and Airs 18). Pindar’s trope provides all the confirmation that we might need for the familiarity of the notion well before Herodotus’ histories, since he clearly expected his patron and audience to grasp the ethnographic detail without difficulty. The patron was Hiero, tyrant of Syracuse, while the audience was presumably, at least in primis, Sicilian. For Pindar had praised Hiero’s foundation of the city of Aetna in part by adducing Scythian nomads. The Pindaric trope is not wholly clear, but seems to centre upon the idea that even among nomads a house is basic to social standing, for a wagonhouse is required if the Scythian is to enjoy any fame. Given this human respect for houses and the like, Hiero has achieved outstanding fame with his establishment of a whole city, namely the refoundation of Catana as Aetna, a theme on which he supported substantial poetry (Dougherty 1993). While the Pindaric thought seems to follow these broad lines, much about Pindar’s verses and their Aristophanic adaptation demands discussion (further, Dunbar 1995: 536-8). We may wonder, for example, whether the focus on cold weather in these exchanges also owes something to what Pindar may have said about the world of Scythia, whose cold was clear enough in Herodotus (4.20; cf. 8-10 etc.). And that in turn would make the unnamed poet’s angling for clothing much more coherent as an issue here. The key point, however, is that this literary game takes the audience to Syracuse and Sicily. Lurking here, of course, is a much more general question about the extent to which the audience of Birds was likely to grasp such allusions, and whether Aristophanes would even have expected as much from the most learned in the theatre. While that issue cannot be explored here, we may at

At the same time, Porphyrion the Giant suited a play wherein birds behaved towards the Olympians after the manner of Giants, for the simple reason that there was also a species of bird named Porphyrion. Aristophanes takes full advantage of the double significance of the name at several points in the play (Birds 553 with Dunbar 1995: ad loc). This was the clearest instance of this double usage, but there seem to have been others too. A scholiast (on Birds 553) tells us that Cebriones was the name of a Giant as well as the bird which appears there in the play, bracketed with Porphyrion. This isolated testimony, if right, may lead us to wonder how many of the birds in the play evoked the names of Giants. In view of these two bird-Giants, the alert might find the famous Giant Alkyoneus even in the calm vision of halcyon-birds, especially as these alkyones were not aspirated in Greek. Suidas (s.v. alkyonides) and others report a myth that had the daughters of the Giant Alkyoneus metamorphosed into such kingfishers (note too the birdsong evoked by the name Enkelados). However, Aristophanes does nothing to develop the idea in his three mentions of the bird in this play (Birds 251, 298, 1594), presumably because the image of the 77

Wonders Lost and Found species did not readily suggest the behaviour of Giants. However, the halcyon was present in the gathering of birds that gave such an aggressive reception to the two Athenians (Birds 298): the birds’ capacity for violence is not to be underestimated.

attended Apollo’s birth on Delos (Hymn 4. 249-55; cf. 2.4-5). The swan was musical, prophetic and inspiring (cf. Papaioannou 2004). It is in this context that we must understand Orpheus’ rebirth as a swan in Plato’s Myth of Er, though it remains unclear how Orphics felt about the idea (Republic 620a). It is probably relevant too that Aristophanes locates swans at the River Hebrus, a further indication of Thrace in the play (Birds 769). Throughout we have here the Whooper Swan, which migrates annually each winter from the north down into the Black Sea and the northern Aegean (Dunbar 1995: 476-7). Immediately we see why it was a corvine, not a swan, that showed the way to Cyrene: the migrant whooper swan did not venture as far south as Africa.

The purpose of the foregoing remarks has been to open the large theme of birds in the context of foundation. While exploring Aristophanes’ Birds to that end, we have started to elucidate the several large comic strategies in the play. We have seen that power and agency might be claimed for birds, comically of course, but not without a strong underpinning, whereby birds might be thought to have extensive knowledge of the world’s geography and settlements, and also to enjoy a range of relationships with the supernatural, including the Olympian gods and the medium of sacrifice (note esp. Romer 1994 on Diagoras of Melos). For while birds like Porphyrion might also be Giants, we have seen too, particularly in the inauguration of the new foundation of Cloudcuckooland, that birds might also be gods. In that regard, Aristophanes’ comic strategy in Birds embraced mockery of contemporary intellectual trends about the nature of the heavens and the supernatural. For we have seen how this comedy of Cloudcuckooland brings birds (not only cuckoos) together with those themes as they had been explored in Clouds: the echo of names is no coincidence. Moreover, we have also seen how these contemporary concerns included not only the long-standing comic fascination with Thrace and current issues there, but also evocations of Sicily and the west that seem to have been underplayed or overlooked in the many treatments of the play.

In Birds we saw how the unnamed poet was among the first to arrive at the new foundation, eager to benefit from his (borrowed) verses on the city. As Callimachus illustrates on Cyrene, there was ample scope for literary creativity in the traditions of each Greek community, whether in the form of poetry or in prose history and mythography, such as Herodorus wrote of Heracles at Heraclea Pontica c. 400 BC or Syriscus at Chersonesus some two centuries later. The public honours bestowed upon the latter for his writing and presentation of his work sufficiently illustrate how far the community was invested in its myths, cults and origins (IOSPE i2 344). Colonial foundations were hungry for a past, even where there was no great patron to ensure that a past was created for it, as Hiero had done for Aetna (Dougherty 1993). Elsewhere I have tried to draw attention to the nest of such traditions that developed in the eastern Black Sea region, where our information is fragmented and scant, and does not have the benefit of significant inscriptions or the like (Braund 1994). It is worth, perhaps, revisiting one of those traditions in the light of the foregoing discussion on foundation and birds.

II The concept of the animal-guide which leads humans to found a city, as to take some other course of action, is common enough in Greek culture as elsewhere (Krappe 1942 collects examples). Birds figure prominently. For example, Callimachus of Cyrene relates a tradition of his city whereby it was founded from Thera under the guidance of a corvine which was Apollo in bird form (Hymn 2.66). We saw in Birds how Comedy could deal with this notion of the bird-deity. There, at the foundation of Cloudcuckooland, the god Apollo (as often, with Artemis and Leto) has a prominence concordant with his usual role in colonial foundations. Here he is the Pythian and Delian swan (Birds 868).

For Pomponius Mela, who completed his work in the mid-first century AD, states that a swan played a key role in the foundation of what is now Ochamchire in ancient Colchis (cf. Kacharava and Kvirkvelia 1991: s.v.). His information is all the more worthy of consideration in that this is one of several scraps of information that we have on the subject of foundation in the ancient Black Sea only from him. Another is the name of nearby Phasis’ oikist, Themistagoras (Mela, 1.108). A cautionary note is needed: Stephanus of Byzantium does not mention this oikist, as sometimes implied or claimed. Now, after the foregoing discussion, we are rather better placed to understand what is at stake in Mela’s story, which is very compressed, as usual in his work:

In a brilliant study, Frederick Ahl has explored Apollo’s association with the swan within a wider analysis of the northern aspects of the deity (Ahl 1982). Plato (Phaedo 84dff.) develops the Apollo-swan connection to claim Apolline powers for the swan, in this case the power of prophecy. Meanwhile, the influence might also be imagined as travelling in the opposite direction, from bird to god, as with Callimachus on the swans who

At in primo flexu iam curvi litoris oppidum est quod Graeci mercatores constituisse, et quia cum caeca tempestate agerentur, ignaris qua terra esset cycni vox notam dederat, Cycnum adpellasse dicuntur.  78

David Braund: Gyenus on stage: civic foundation and the comedy of Aristophanes’ Birds

But in the first turn of a now curving shoreline is a town which Greek merchants are said to have founded. And, as they were driven by a blinding storm, it is said that the call of a swan gave a sign to them in their ignorance as to where lay land, and so they called the town Cycnus (Swan). (Mela, 1.99)

was not involved in the foundation of Swan-town? No doubt, in ancient Ochamchire the local elite had firm enough ideas about which deity was involved, and we may reasonably assume that there was a cult there – as there was regularly elsewhere – which entailed the supernatural aspects of the community’s foundation. Given his energetic colonization, Apollo is most likely. The fact that it is the swan’s voice that is crucial seems further to confirm the point. However, we may also reasonably suppose, given the tendency for contestation on mythical matters in antiquity, that there may have been competing versions told at Dioscurias and in the Bosporan kingdom, for example. We have no knowledge of these debates and disagreements, but we should at least observe the opportunities here also for constructive diplomacy whether on a private or public level.

The swan has not led a colonial expedition, as for example the corvine of Cyrene, but it has made a crucial intervention which has turned disaster into opportunity. The sailors have been saved from a watery grave, while the town has been founded. A voyage for trade has been transformed by the storm and the swan into an act of foundation, the creation of a new settlement in Colchis. But how are we to understand this swan? While Mela’s compressed account says nothing of the supernatural, we have seen enough already in this paper to realise that supernatural forces are at work in the story, as befitted civic foundation, and as Syriscus and others demanded. The storm may have been the work of a deity, but it is the swan who points us towards a sharper image. For, while the swan had a range of divine connections and mythical appearances (witness the many mythical heroes named Swan), we have seen its strong association with Apollo in particular. And we have also observed Apollo’s particular role in the foundation of cities, as at Cyrene. Closer to hand, of course, we seem to have him in such a role at Phasis, where he received cult as Apollo Hegemon by the end of the fifth century BC and perhaps much earlier (Braund 2010; cf. more generally Lordkipanidze 2000). For the community of Swan-town, however, there was little to be gained by exclusivity, especially in view of the fact that its neighbour on the other side from Phasis along the coast westwards was Dioscurias, which traced its foundation to the Dioscuri, Castor and Pollux (Braund 1994; 27-32; see further Dana and Dana 2001-3, esp. 106 n.77 on Dioscuri and foundation on the west coast of the Black Sea). Of course, these were the sons of Leda the swan, born from an egg. Moreover, they were much honoured as deities who saved sailors from storms and other perils of the sea (as at Istria, ISM 1. 112). Who was to say that the swan which had saved Mela’s stormtossed merchants were not connected with them? Substantially further to the west was the important cult of Aphrodite Ourania at Apatouron on the Taman peninsula, where migrating whooper swans still flock in numbers. A Bosporan dedication of the later second century BC from Panticapaeum made to the goddess at Apatouron (CIRB 75) shows the goddess riding a flying swan: this was the work of her thiasos and entailed the Bosporan rulers of the day, so that the iconography must be more than a random quirk (for much more, see Braund 2018: ch. 5). Moreover, Aphrodite too was a marine deity, very capable of saving those in trouble at sea. A Naucratite tradition shows her saving and founding in much that way, albeit without mention of a swan (Gutzwiller 2010). Who was to say that Aphrodite

In many ways, these processes are commonplace in the context of Greek settlement and mythmaking. However, there is an unusual dimension to this case which makes it especially interesting, for the whole tale of the saviourfounder swan can be shown with some confidence to have been created well after the foundation of this particular community. Crucially, Ps.-Scylax not only omits the story, but also gives a different name to the community itself. In his account (Periplus, 81) this is nowhere said to be Cycnus, but is instead Gyenus. There is no scope to suppose two different towns, for Colchis is not so rich in Greek settlements, the names are very similar and (most important) the location seems to be one and the same. We may conclude, therefore, that Gyenus became Cycnus after the completion of PsScylax’s work in the fourth century BC. In principle we might prefer to condemn one (or even both) of the two authors for error or invention, but there seems insufficient basis for such a large step. We seem to have, rather, the evolution of an awkward name, evidently of non-Greek origin, Gyenus, into a name that made immediate sense in Greek and Latin, Cycnus, and which brought with it the opportunities for mythmaking that we have discussed. We might also be tempted to consider emendation of Ps. Scylax’s text on Gyenus (section 81) in order to change the name, even to Cycnus. However, a closer look at his text tells against such a response: ΚΟΛΧΟΙ. Μετὰ δὲ τούτους Κόλχοι ἔθνος καὶ Διοσκουρὶς πόλις καὶ Γυηνὸς πόλις Ἑλληνὶς καὶ Γυηνὸς ποταμὸς καὶ Χερόβιος ποταμὸς, Χόρσος ποταμὸς, Ἄριος ποταμὸς, Φᾶσις ποταμὸς καὶ Φᾶσις Ἑλληνὶς πόλις, καὶ ἀνάπλους ἀνὰ τὸν ποταμὸν σταδίων ρπʹ, εἰς  πόλιν  μεγάλην βάρβαρον, ὅθεν ἡ Μήδεια ἦν· ἐνταῦθά ἐστι Ῥὶς ποταμός· Ἴσις ποταμὸς, Λῃστῶν ποταμὸς, Ἄψαρος ποταμός.

79

Wonders Lost and Found COLCHIANS. After the aforementioned are Colchians, a people, and Dioskuris [= Dioscurias] a polis and Gyenus a Greek polis and Gyenus River and Kherobios River, Khorsos River, Arios River, Phasis River and Phasis, a Greek polis, and an upriver voyage of 180 stades to a great barbarian polis, where Medea was. Then the Ris River, Isis River, Bandits’ River, Apsaros River.

Borysthenes, as in Herodotus, notably at 4.18, and Dio Chrysostom, Oration 36). However, the Olbian case was encouraged by the geographical prominence of the River Borysthenes (Dnieper) in Greek culture generally, together with its considerable local separation from Olbia proper. This example, at least, does not map well onto Gyenus, where too we have a myth. Indeed, we may even suppose there that the swan was on or beside the homonymous river (for its name must have changed with the town). It would be good to have the fuller version of the story of the saviour swan upon which Mela was drawing, but we do not have that luxury and so run out of knowledge about detail, whereas we can construct a plausible general model of mythmaking there and change between the fourth century BC and first century AD.

Immediately we see the author’s concern with rivers as well as settlements, both Greek places (as he judges) of the coast and non-Greek (like Aristophanes’ Triballians) of the interior. Such was appropriate to the genre: this is a Periplus after all. These twin concerns helpfully reveal that Gyenus is not only the name of a Greek polis, but also that of a river. Moreover, the river is listed next after the town of the same name. There can be no real doubt that this is the River Galidzga on which sits the modern town of Ochamchire, though one must always be alert to the various complexities of river-names in ancient texts on the region. Evidently the ancient town of Gyenus took its name from the river on which it was founded. Indeed, we seem to have the same process at work in nearby Phasis, as in many a Greek foundation elsewhere. The naming-practice testifies to the vital importance of the local river in sustaining the new foundation, both supplying water and facilitating movement, while at the same time also marking territory and contributing to defensibility. The name of the river came first because the river preceded the foundation and already had a local name. We now understand how large a part local peoples had in the foundation and development of Greek colonies, but we need not suppose that the Greek form of the name Gyenus was more than an approximation to the name current among locals in a language alien and awkward to the settlers and, all the more so, to distant writers such as Ps. Scylax.

III This paper has brought together two major strands of the work of Michael Vickers, to both of which his contribution has been very considerable: that is, Old Comedy and Georgia, ancient and modern. I hope that it is by now obvious how these two distinct fields may not only be studied together, but also benefit from being brought together. When we see the unnamed poet turning up to produce myth-laden verses for the new foundation, we have a comic representation of a process which was real enough, even in a modest town like Gyenus. For we have seen how each community valued and developed its own stories, not least of origin, descent and ancient achievements. Regularly all this was couched in a discourse of the supernatural which showed that the community was not simply a human foundation but was a matter of divine will. That was as true of great cities like Athens and Ephesus as it was of little Gyenus-Cycnus. Bibliography

If these rather conservative inferences are correct, based on the texts and on usual practice elsewhere around the Greek world, we must take Gyenus to have adapted its name to become Cycnus at some stage before the first century AD. Certainly towns could and did change their names: in the Crimea, for example, Megarike had gone through a much greater change when it became Chersonesus, being also it seems Heraclea (Pliny, NH 4.85). In the Hellenistic world and after such changes became commonplace insofar as communities were renamed to bear a name that evoked or commemorated a member of the ruling dynasty. We are left to wonder how and when the change from Gyenus to Cycnus may have occurred. Finally, however, we should consider another possibility, whereby the change would be more of perspective than of substance. For we know in the case of Olbia that a city might be known to itself routinely and officially by one name (here, Olbia) and quite another to outsiders (here,

Ahl, F. M. 1982. Amber, Avallon, and Apollo’s Singing Swan. American Journal of Philology 103: 373-411. Bakola, E. 2010. Cratinus and the art of comedy. Oxford. Barr-Sharrar, B. 2008. The Derveni krater. Princeton. Bolton, J.P.D. 1962. Aristeas of Proconnesus. Oxford. Braund, D. 1994. Georgia in antiquity. Oxford. Braund, D. 1999. Laches at Acanthus Classical Quarterly 49: 321-5. Braund, D. 2010. The religious landscape of Phasis, in E. Petropoulos and A.A. Maslennikov (eds) Ancient sacral mounments in the Black Sea: 431-40. Thessaloniki. Braund, D. 2018. Greek religion and cults in the Black Sea region. Cambridge. Cartledge, P.A. 1990. Aristophanes and his Theatre of the Absurd. Bristol. Dana, M. and D. Dana 2001-2003. Histoires locales dans le Pont-Euxin ouest et nord. Identité grecque et le construction du passé. Mar nero 5: 91-111. 80

David Braund: Gyenus on stage: civic foundation and the comedy of Aristophanes’ Birds

Dawson, D. 1992. Cities of the Gods: Communist Utopias in Greek Thought. Oxford. Dougherty, C. 1993. The poetics of colonization. Oxford. DuBois, P. 2006. The History of the Impossible: Ancient Utopia. Classical Philology 101: 1-14. Dunbar, N. 1995. Aristophanes, Birds. Oxford. Fragoulakis, M. 2013. Kinship in Thucydides. Oxford. Graininger D. and J. Valeva 2015. A Companion to Ancient Thrace. Chichester. Gutzwiller, K. 2010. The Demon Mosquito. Zeitschrift für Papyrologie und Epigraphik 174: 133-8. Kanavou, N. 2011. Aristophanes’ comedy of names. BerlinNew York. Krappe, A.H. 1942. Guiding Animals’ Journal of American Folklore 55: 228-46. Lane Fox, R. 2010. Thucydides and documentary history. Classical Quarterly 60: 11-29. Lordkipanidze, O. 2000. Phasis. Stuttgart. MacDowell, D. 1995. Aristophanes and Athens. Oxford. Martin, R.P. 2009. Read on arrival, in R. Hunter and I. Rutherford (eds) Wandering poets in ancient Greek culture: 80-104. Cambridge. Mattingly, H. 1999. Review of T. Figueira, The Power of Money: Coinage and Politics in the Athenian Empire. American Journal of Archaeology 103: 712-13. Morrison, D.R. 2007. The utopian character of Plato’s ideal city, in G. R. F. Ferrari (ed.) Cambridge companion to Plato’s Republic: 232-55. Cambridge. Papaioannou, S. 2004. Ut non [forma] cygnorum, sic albis proxima cygnis’: Poetology, Epic Definition, and Swan Imagery in Ovid’s Metamorphoses. Phoenix 58: 49-61. Romer, F.E. 1994. Atheism, Impiety and the Limos Melios in Aristophanes’ Birds. American Journal of Philology 115: 351-65. Sidwell, K.C. 2009. Aristophanes the democrat. Cambridge. Stewart, A. 1990. Greek Sculpture: An Exploration. New Haven. Vickers, M. 1989. Alcibiades on Stage: Aristophanes’ “Birds”. Historia 38: 267-99. Vickers, M. 1996. Fifth Century Chronology and the Coinage Decree. Journal of Hellenic Studies 116: 171-4. Vickers, M. 1997. Pericles on stage: political comedy in Aristophanes’ early plays. Austin TX.

81

New archaeological finds at Pichvnari (November-December 2010) Amiran Kakhidze 4 metres in height over an area measuring 160m x 120m. Needless to say, we immediately approached the relevant department to prevent further destruction of the site. Thanks are due to Niko Vacheishvili, the head of the Agency for the Preservation of the Cultural Heritage of Georgia who acted decisively, signed a contract with the Batumi Archaeological Museum, and allocated funds for the proper archaeological investigation of the site. This took place over two months and secured much valuable material. We give a brief account here, starting with the sand-dunes (Figure 1).

The British-Georgian Archaeological Expedition brought its activities to a close, after thirteen seasons of work, at the end of August 2010, little suspecting that further work on the site was imminent. The Batumi Archaeological Museum found itself, however, conducting wide-scale field-work in November and December 2010. The Georgian government had decided to give what was termed a ‘free tourist zone’ to developers to build hotels, but without any consultation with archaeologists. Preparations for the infra-structure for the hotel development began in November with clearing the ground, digging foundations for 32 hotels and making space for a carpark. Unfortunately, this area included two of the most important archaeological sites at Pichvnari, namely the sand-dunes containing the remains of Colchian culture of the 8th-7th centuries BC, as well as part of the Hellenistic burial ground of the 4th-1st centuries BC.

The sand-dunes of the eastern Black Sea are associated with a process of terrace formation that occurred over the past 10-12 millennia. According to studies of vegetation of the Kobuleti peat-bog, a new terrace had formed here some 6000 years ago, although others believe that it occurred 4-5000 years ago. These sites are spread over a large area along the Black Sea coast of Georgia (Tavamaishvili 1999: 29-39). The Pichvnari sand-dunes were studied during the

Work progressed at great speed and within two days heavy machinery had flattened dunes between 3 and

Figure 1. Sand dunes, an overview

82

Amiran Kakhidze: New archaeological finds at Pichvnari (November-December 2010)

1960s, but an even more intensive survey took place in November and December of 2010.

The trench was extended towards the east over a large area. Part proved to be sterile sand, but elsewhere there were traces of small sacrificial areas. A larger sacrificial area was found in the SW I 78 sector. The sacrificial area measured 16 x 8.5 m. The eastern part had already been disturbed by the developers. The finds were dominated by fragments of bath-shaped vessels, stone net weights.

The southern part of the dunes had been cut during development work, and the first thing we did was to control a trench which revealed the stratigraphy of the site. Observations were made regarding the creation of the dunes, the frequency and intensity of tides, and the stages of dune formation. We identified the outlines of the sand-dunes and collected relevant material.

It was planned to build one of the hotels 28 m north of the south trench on the back of the sand-dune. Here we excavated a trench measuring 16 x 8 m. The first terrace of the western area produced an altar measuring 3.3 x 2.2 m. There were two well preserved ceramic vessels nearby, one of them deposited, somewhat unusually, upside-down. There was a terracotta head of a dog on top of the altar, which is of a special interest for identifying the meaning of the site (Figure 2). The altar had been built with square shaped horn-like stands related to a bull cult (Figure 3).

Besides cleaning and recording the trench in the southern dune, we excavated the south-west part. Large areas revealed dark spots and traces of sacrificial areas at a depth of about 1.5 m. The surface was levelled and two monumental altars situated to the north emerged that incorporated several clay hornlike stands. They measure 2.6 x 1.6 m and 2.4 x 1.3 m. We noted that on the north side of the trench a layer of small pebbles indicated a line of waves. Here too we made interesting finds, including a large conical clay vessel probably used during religious rituals.

The second terrace yielded fragments of ceramic vessels and a sacrificial area, part of which lay beneath

Figure 2. Terracotta dog’s head

Figure 3. Altar fragments with hornshaped finials

83

Wonders Lost and Found an unexcavated area. Here too there was a sandy sterile layer left by the waves of the sea.

of clay and connected with a cult of a bull. The use of stands, round in cross-section and used in connection with hearths, was widespread throughout the Caucasus and Asia Minor from the early Bronze Age and are thought to be somehow connected to the Kura Araxes culture (Pkhakadze 2000: 16-23). This kind of material was found in Pichvnari in the cultural layers of the last centuries of the 2nd millennium BC and are believed to be associated with bull- and fertility cults. The altars from the sand-dunes had been built with larger square stands; one has a phallic shape with a cross on its tip, perhaps symbolising the sun in motion. There were some clay tiles with cult symbols (Figure 4).

The third zone also had sandy areas, altars small and large, black spots indicating ritual burning and a ball of clay. We might speculate that the bath-shaped vessels were made here in wooden moulds. The same kind of clay balls had been attested elsewhere in the sanddunes. We found a fifth altar, divided into two with horn-like stands that measured 2.35 x 1.1 m. There were fragments of pitchers and other kinds of clay vessels and a large spherical stone polisher. The fourth terrace was situated at the top of the sanddune and illustrates the last stage of dune-formation. The large area was used as a sacrificial area and there were two further altars Nos 5 and 6. The second altar was also divided with large square horn-shaped stands found in situ.

The sculptural head of a dog could give us a clue towards the functional meaning of the sand-dunes. A preliminary study of Georgian folklore and ethnography suggests that it may have been connected to the sea deity Mesepe. A large conical clay vessel with thick walls inserted into the eastern wall of the third altar was probably also connected with the unique ancient rituals of the Colchians. We shall not dwell on other kinds of vessels, apart from noting that Colchian pithoi, drinking vessels and tubular-handled jugs began to have been made in the 8th-7th centuries BC. There were substantial amount of stones used as net weights, and large spherical stones used as polishers.

The peripheral cult places and other remains were located on the general map and two large control trenches were dug in the eastern part of the sand-dunes which happened to be in the area of the ‘free tourist zone’. One of them produced fragments of vessels dating to the end of the 2nd millennium BC. One of the decorative vessels has been fully restored. A second vessel had fallen down a hole to the bottom. It would appear that the tradition of sacrificing liquids to the gods of the underworld was practised at Pichvnari from the Late Bronze Age.

The excavations of the sand-dunes have thus enabled us to rescue many artefacts related to Colchian material and religious culture and enriches our understanding of what happened at the sand-dunes in antiquity. The new material supported the hypothesis put forward in the 1960s regarding the existence of highly developed Colchian religious sanctuaries in the dunes of the

We will present here but a small part of the rich finds, 90% of which were bath-shaped vessels. There were many stands both round and square, horn shaped, made

Figure 4. Tile fragments

84

Amiran Kakhidze: New archaeological finds at Pichvnari (November-December 2010)

Hellenistic site

Georgian Black Sea coast, where rites associated with sea deities and a bull-cult were conducted at certain times of the year. The horn-shape stands have a special role and significance among the new finds, and require further explanation.

Excavations were also conducted at the Hellenistic cemetery. Some 420 burials of the local Colchian population had been found over the years dating to the 5th century BC. The so-called Greek necropolis of the 5th-4th centuries is of a special significance in that it is the only ancient Greek cemetery in the Caucasus. We excavated 450 burials and dozens of sacrificial areas. More recently, some early medieval graves were found here as well.

Water was considered to be one of the primary sources of the universe according to myths of many ancient cultures. Water thus became a cult object. In Georgian mythology many water deities had many functions; there were the ‘Mother of water’, the ‘Water-Witch’, the ‘Water Man’, the ‘Water Angel’, the ‘Mesepi’, the ‘Larsa’, the ‘Lazare’, the ‘Fish whale’ and so on. Some of them are animistic in origin, but underwent some development, or transformation. We should pay a special attention to the religious traditions of the native Mingrelian population of western Georgia. There are legends about the ‘Mesepi’, a sea woman and a sea man, who were described as being immortal and anthropomorphic; female Mesepi were pretty with blonde hair and white faces, men were ugly and frightening, and wore black. According to other legends these sea creatures were zoo-anthropomorphic, they lived in the depths of sea and barked like dogs; they came out on land at a certain time of the year (between October 28 to November 3) and carried a magic sceptre and were accompanied by a hound. Their aim on land was check their domain and collect taxes for winter. They protected animals. The Mingrelians believed that female Mesepi would bring good weather and good harvests, but the male Mesepi was a destroyer who brought bad luck. They would hunt in the deep forests and valleys, feast on game, then collect and wrap the bones in skins and bring the animals back to life with a wave of their sceptres.

The Hellenistic cemetery had been excavated since the 1960s. Burials are situated in two areas. One lies within the south-eastern part of the Pichvnari necropolis, but the other, studied this year, is actually at the sand-dunes. In 2003-2004 the British-Georgian joint archaeological expedition had found 46 burials here (Kakhidze and Vashakidze: 19-20; 231-237, figs 71-108). The rescue excavations of 2010 revealed a further 44 graves. We give a brief account here. We determined the western, and part of the northern border of the necropolis. Most of the burials were inhumations apart from a couple of amphora-burials. Iron and bronze nails found in the graves show that some of the deceased had been interred with coffins and others beneath wooden roofs. Greek influence can be seen in the tradition of an eastern orientation for many graves (26 graves). Ten pointed to the north and the rest were oriented to the south-east. The moist sandy soil does not preserve skeletons but remains of teeth were found in 16 graves. The Greek tradition of sacrificial areas (for funerary meals) seems to have been established among the local population as well. Twelve ritual platforms were excavated recently; the grave sizes indicate that the deceased were mostly buried in a crouched position on their left or right side, while others were buried in a supine position.

The extraordinary dog that accompanied them (the ‘Mesepish Jogori’) was believed to have been tall and hairy, clever and swift. They were believed to be sacred creatures and anyone who tried to harm them would be put to death. After 6 days the Mesepi would go back to sea taking with them food for winter. The Mingrelians believed that iron would repel the male Mesepi and they would place a piece of iron in their barns and store rooms to prevent the Mesepi from stealing their food. Their return to the sea would have been followed by torrential rains that washed away their footprints. In case of female Mesepi all gates and doors were left open.

The grave goods were situated mostly in the areas of the head and hands of the deceased, with a few in the leg area. The majority of the grave goods consist of locally produced ceramic vessels. There were traditional forms typical for Colchian pottery, various sizes of jars, jugs, bowls and large pots, some with tubular handles (Figure 5) and specifically Pichvnari type of jugs. Many pots were made of brownish clay with plain but elegant proportions. There were rare and in some cases even unique material; one of the small jugs had on its upper body three reliefs of a sun-disc and a half-moon (Figure 6). There were more vessels with fluted decoration; some pots had inscribed circles as potters’ signs. There were also Greek shapes like Colchian amphorae, spouted vessels, oinochoai, etc.

Together with water deities Mesepi were also deities of weather and fertility. The animal protective nature of the god is understood to be a later development (Tandilava 1978); 1986; 1996: 132-148). It would appear that the ancient Colchians conducted their sacrificial rites in order to appease sea deities in the coastal areas of the Black Sea. Excavations at Kalota in the mountainous region of Khulo attested the existence of the cult of ‘Lazare’, another water deity (Kakhidze and Mamuladze 1993).

Among the imported pottery there were some fine kantharoi, one of them inscribed with the letter A, 85

Wonders Lost and Found

Figure 5. Spout-handled jug

perhaps the first letter of a proper name. Another small drinking bowl had a graffito underneath: dekas. A remarkable Attic ‘West Slope’ ware deep phiale was inscribed diossatyros (Figures 7-8). There was a considerable amount of plain or black-gloss perfume vessels, aryballoi, but only one black-gloss salt-cellar. The grave goods included local silver coins, the ordinary triobols. From the 5th century BC there had been a continuous tradition of burying the dead with a so-called ‘Charon’s obol’ in the mouth. The local silver coins were used in Colchis throughout the early Hellenistic period.

Figure 6. Jug decorated with sun disc and moon

Figures 7-8. Attic ‘West Slope’ phiale

86

Amiran Kakhidze: New archaeological finds at Pichvnari (November-December 2010)

Figure 9. Gold and glass earrings

Figure 10. Silver earring

Jewellery was present in abundance, including gold earrings. Some spherical earrings were made by coating a glass bead with gold leaf (Figure 9). There were several gold earrings with suspended rings and four graves had a pair each. Many more were of silver, bronze and iron. Silver lion-head earrings (Figure 10), of a kind found at Pichvnari in gold in 2004, were found for the first time. It should be said that this kind of earring is known from over a wide geographical area, but has not been found elsewhere in Georgia. There was also a choker of socalled Kirkali type. Our collection of bracelets was enriched: bronze for the most part, but there was a pair of silver bracelets with animal head finials. Most were round in section, but some were flat or square. Usually there was more than one pair in a single burial. They were usually decorated with knobs or close-set grooves.

interesting group is represented by mosaic beads, but the most exciting were those with human faces (Figure 11), of a kind found earlier at Pichvnari. E. Alekseeva dates similar beads from the northern Black Sea to the 3rd century BC and they have been found in the Taurian Chersonese and Maikop. This kind of bead is well known from central Europe and throughout Mediterranean during the 5th-4th centuries BC, but apparently appear in Colchis only from the Hellenistic period. A miniature amphoriskos-shaped pendant-bead resembles pendants from the rich burials at Dablagomi, Melitopolis and other sites. There were also gold-leaf beads, a technique that is believed to have originated in Egypt in the 4th century BC spreading to Europe from the 3rd-2nd centuries BC. Large metal artefacts were found for first time in Hellenistic burials at Pichvnari. A bronze mirror and couple of iron strigils were unearthed during the rescue excavations. The excavations also enriched our knowledge of the fishing activities of the Pichvnari population. There was an entire net deposited in one of the burials, complete with 57 lead weights. Another burial contained a net with 86 lead weights and tied together with ten hooks (Figure 12) to catch a larger fish, probably plaice. A third burial was especially interesting for it contained three smaller fish-nets together with four-star bait and some floats consisting of two hollow halves soldered together. This was unparalleled for Pichvnari so far. Other interesting items included a bronze chain, bells, an iron knife, and a coffin lock.

There were especially interesting examples of glyptic art. It is difficult to identify the images until they are properly cleaned; provisionally, one shows a Pegasus, another a seven petalled rosette. There were also images of a cornucopia and a tree of life. One of the rings had a horse on its bezel. Several rings were made of bronze with a flat ring and a round bezel, and were probably made locally. There were also rings with gem-stones, but one of the deep set stones is unfortunately missing. Glass beads and pendants are found in great quantity at the Hellenistic cemetery. The majority of them are plain black, blue, yellow and white. Similar glass beads have been found at many sites in Georgia and the Crimea. There were colourful pear-shaped beads, monochrome glass pendants in the shape of a heart, an amphoriskos and a dolphin, also attested elsewhere.

As we can see, the new finds speak volumes about the lifestyle, and the industrial, cultural and religious activities of ordinary people in Hellenistic Pichvnari. We can observe a continuation of Greek influence, such as the

There were a few polychrome beads, two of them gem-stones, square or pointed ovals in shape. An 87

Wonders Lost and Found

Figure 11. Polychrome glass head beads

Figure 12. Bronze fish hooks

individual inhumations oriented to the east, sacrificial platforms, and the tradition of Charon’s obol, at the same time preserving local Colchian traditions. Grave goods included pots containing food and drink for the afterlife, and the deceased were adorned with jewellery, beads and coins. The inventory implies that the principal activities of the local population were agriculture, fishing, viticulture, animal husbandry, pottery making, metallurgy and textile production. Fishing was naturally important due to the proximity to the sea and the Choloki-Ochkhamuri river basin. We can presume that salted fish and dolphin fat were trading commodities here as was the case for the rest of the coastal population.

The jewellery was enriched with new types and shapes, such as the coated suspended earrings and the silver bracelets with lion head finials. There were fine examples of other varieties of bracelet, earrings, finger rings etc. The astral deities, the Sun and Moon, had a special role in the religious pantheon of Hellenistic Colchians. Greek inscriptions together with the strigils suggest that sporting contests and religious feasts were held in Pichvnari. Colchis was never directly affected by the expansion of Alexander of Macedon but a close relationship with the Hellenistic kingdoms meant that the eastern Black Sea region became involved in the process of Hellenization. Trans-Caucasian trade routes continued to traverse the Pontic shores.

The new material is significant for the study of Colchian pottery. Although there were many traditional forms, most of them were distinctive new shapes. It would appear that Pichvnari was a significant centre of pottery manufacture. Imported pottery is scarce, mostly consisting of Attic black-gloss kantharoi, some of them inscribed. 88

Amiran Kakhidze: New archaeological finds at Pichvnari (November-December 2010)

Bibliography Kakhidze A. and N. Vashakidze 2010. Pichvnari 3, Elinizmi da Kolkheti (The Hellenistic Period and Colchis). Batumi-Oxford. Kakhidze, A. and S. Mamuladze 1993. Acharistsklis kheobis udzvelesi arkeologiuri dzeglebi (Ancient archaeological sites in the gorge of the Acharistskali river). Batumi. Pkhakadze, G. 2000. Kerebisa da sadgrebis gavrcelebis sakitkhisatvis (On the distribution of hearths and stands). Dziebani 5: 16-23. Tandilava, Z. 1978. Masalebi mesepeebze (Materials concerning Mesepe) Samkhret-dasavlet sakartvelos dzeglebi (Monuments of Southwest Georgia) 4. Tandilava, Z. 1986. Gvtaeba mesepis mitologiuri sakhis gagebistvis (Understanding the mythological meaning of the deity Mesepe) SDSZ (Monuments of Southwest Georgia) 8. Tandilava, Z. 1996. Tsklis kulti da kartuli polklori (The Cult of Water and Georgian Folklore. Batumi. Tavamaishvili, G. 1999. Udzvelesi namosakhlarebi samkhret-dasavlet sakartvelos akhal shavzgvur terasaze (Ancient settlements on the south-west coastal terraces of Georgia): 29-39. Tbilisi.

89

A double-sided glass relief pinhead from ancient Colchis – the Pichvnari ‘Heracles’ Sujatha Chandrasekaran The archaeological site of Pichvnari in ancient Colchis is characterized by its vast necropolis of the Classical and Hellenistic periods (Figure 1), with numerous finds attesting to the direct contacts between Greeks and Colchians. This paper concentrates on one of the singular finds from the site - a small decorative pinhead of yellow glass with a double-sided relief depiction of a bearded man commonly designated as ‘Heracles’ (Figure 2).1 The ornamental relief pinhead measures approximately 1.7 x 1.7cm and was retrieved from burial 209 of the Hellenistic necropolis in Pichvnari and

is dating accordingly.2 Discovered in 2004, the ‘Heracles’ pinhead is a unique Colchian example of a small yet widespread category of glass items from the ancient world, and its discovery attests to the role of ancient Colchis within the global Hellenistic network. Ornament category Double-sided relief ornaments of glass belong to a class of small glass items of the Classical and Hellenistic periods – pendants or amulets – formed around a

Figure 1. The Hellenistic necropolis of Pichvnari. (Photograph S. Chandrasekaran) 1  See Kakhidze and Vashakhidze 2010: 151, 258; Pl. 155, 3, 158, 5 (a ‘pendant’); also Vickers 2011: 140, 142, fig. 7 (‘glass bead’). The ornament is also mentioned by M. Turmanidze (2007). The first mention of this item (as a bead) is found in Vickers and Kakhidze (2004: 22). See a preliminary study of this find by the author (Chandrasekaran 2013).

The burial is dated to the Hellenistic period based on its finds, in particular jewellery, a clay jug and a coin. See Kakhidze and Vashakhidze 2010: 258.

2 

90

Sujatha Chandrasekaran: A double-sided glass relief pinhead from ancient Colchis

Figure 2, a-c. A decorative pinhead from Pichvnari with double-sided relief depiction of ‘Heracles’. (After photographs by M. Vickers and T. Vashakidze)

details.4 As such, double-sided relief ornaments stand in distinct contrast to the bulk of core-formed glass face ornaments from the Hellenistic period or earlier, which were shaped and detailed by hand and hand tools to form individual heads or face masks with highly stylized features.5

core and shaped into a rounded ornament with facial depiction. Such items are found in varying quantities throughout the Mediterranean and in the Black Sea region, with highest concentrations in North Africa and Northern Black Sea area (Figure 3).3 The core-formed glass technique was popular in the Mediterranean from the 6th-5th centuries BC, with a brief resurgence in the late 4th century BC (Lightfoot 2001: 59). Based on examples with known find contexts, T. Haevernick has convincingly dated double-sided relief ornaments to the 4th and 3rd centuries BC (see Haevernick 1968: 6489), placing them in this period of ‘resurgence’.

Production technique The ‘Heracles’ pinhead was produced of milky, yellow glass paste as follows: a metallic stick or rod was swirled into the glass paste, scooping up the necessary portion, and then placed in a double-sided mould which pressed the mass into a relief.6 At some point the rod was removed, leaving a hollow cylindrical ‘core’ in the object. This so-called basal perforation (as defined by T. Rossell [2003]) is preserved on the Pichvnari pinhead,

The production of these items involved a two-step process that was fairly new for the period (Haevernick 1968: 647) – a mass of glass paste was swirled onto a core and then pressed into a double-mould, resulting in a flat, disc-shaped ornament with detailed facial relief on both sides – this process will be described in detail below. The resulting relief is offset from the surrounding ‘frame’ of the glass disc and characterized by Hellenistic portrait-style iconography with realistic

4  Note that this surrounding ‘frame’ is often damaged, as can be observed in Haevernick 1968: pl. 6. 5  See Seefried (1982) and Haevernick (1977) for detailed studies of various types of glass face ornaments (pendants and beads), including many illustrations. Tait (1991: 216) notes the use of rod tools and pincers to render details on face ornaments produced without a mould. 6  The manufacture technique of double-sided relief beads-pendants is described by Alekseeva (1978: 59, 62) and also Haevernick (1968: 647).

See the maps of general distribution of such items in Seefried 1982: figs. 46A, 46B.

3 

91

Wonders Lost and Found

Figure 3. Map with find sites of double-sided relief pendants and ornamental pinheads. (S. Chandrasekaran using AMWC Map Tiles – www.amwc.unc.edu [© CC BY-NC 3.0])

with traces of metal near its opening (Figure 2b)7 attesting to this production step.

The final production step involved the insertion of a metal pin – now lost – into the centre of the moulded pinhead from one side (the ‘reverse). This action was clearly carried out while the glass paste was still pliable and with no care for the relief depiction on that side. The slit-like insertion has damaged the form of the relief and smudged the facial details. Metal remains in the insertion hole attest to this final step (see Figure 2c). The fact that no traces of the second production step – the definition of features – are visible on the damaged reverse side (compare Figures 2a and 2b), suggests that the ‘Heracles’ ornament was planned for use as a onesided pinhead from early on in production. As such, this pinhead differs from other known double-sided relief pinheads, which were mounted on a pin using the basal perforation only, leaving both ornamental sides untouched.10

While the exact construction of the mould itself cannot be determined,8 it is clear that the pinhead was fashioned by the clamping or pressing together of two moulds – each with a similar relief.9 In the process, the overflow of excess glass paste was flattened by the edges of the mould, resulting in a wide, flattened border around the facial relief (Figures 2a-b). The second step involved final touches to the glass relief by hand – using fine tools – to add further definition to facial features. Thus we see on the front side of the ‘Heracles’ ornament fine lines pressed into the glass around the eyes to form an ornamented frame, or into the hair and beard to clearly define individual locks or curls. It is unclear whether these touches were made before or after the core-forming rod was removed, but clearly they were made while the glass was still pliable. Interestingly, this step can only be recognized on the undamaged ‘front’ side of the ‘Heracles’ ornament and not on the damaged reverse side.

The flat border above the relief face is damaged, leaving it unclear whether the ornament originally had a loop for hanging – even though this appears unlikely. Note that the yellow colour of the ‘Heracles’ pinhead is quite rare for this type of ornament, which was usually produced in blue and green hues.11

See also Turmanidze (2007), who records traces of metal here. Haevernick (1968: 647) notes the possibility that one-off clay moulds were used for production of the double-relief ornaments; It is impossible to determine whether both mould halves were attached (e.g. using tongs or hinges). 9  In some cases we can observe some variation in the rendering of details on both reliefs, for example on one of the pendants from Elizavetovskoye Gorodishche (Kopylov 2006: fig. 1 no. 5). Note the differing eye form, and especially the necklace depicted only on one side. 7  8 

10  Arvellier-Dulong and Nenna (2011: nos. 481, 487) show five examples where the pins are still preserved. 11  Haevernick 1968: 648, see also the catalogue listing in 649-52, as well as most of the pieces mentioned in this paper.

92

Sujatha Chandrasekaran: A double-sided glass relief pinhead from ancient Colchis

Description of the relief

items depicting females,13 mainly in a form recalling Astarte, with her characteristic thick hair falling to the shoulders and framing her rounded face.14 Male depictions – distinctly fewer in number – show an adult bearded man in most cases and occasionally a beardless youth, sometimes with clearly negroid features.15 In one case, on a specimen from Elizavetovskoye Gorodishche, we even find both a bearded adult male on one side and a beardless youth on the other (see Kopylov 2006: fig. 1 no. 10).16

Both sides of the pinhead show a relief depiction of a bearded man commonly referred to as ‘Heracles’ due to the similarity with well-known depictions of the Greek hero as embodied by for example, the Hercules Farnese. As already mentioned, the relief details are best visible on the undamaged side (Figure 2a): here we see a round face framed from the top by vertical locks of hair, from the bottom by a full beard of similarly arranged locks finishing at the bottom with horizontal rows of curls. The face is defined by almond-shaped eyes, a small, straight nose with rounded nostrils, and a small mouth with full lips draped with a thick moustache that reaches down to the beard.

No item is known to combine male and female faces.17 The ‘Heracles’ from Pichvnari belongs to the category of items showing a bearded man on both sides. Of the known examples published with images,18 only the Pichvnari ‘Heracles’ was fashioned as a one-sided pinhead.19

The relief on the reverse, while less clear (Figure 2b), shows the same bearded facial form as on the front, but with less detail of hair and eyes and no idea of the nose and mouth.

Pendants vs. pinheads

History of research of similar objects

Most double-sided relief ornaments appear to have been fashioned as pendants with a small loop or attachment perforation at the top of the flat border above the relief face. Often, the upper section where a loop might have been is broken, and nothing remains of a pin by which an ornament might clearly have been determined as a

Double-sided relief ornaments of glass represent a small category of ancient glass ornaments. While a decent number of examples are known from various museum collections throughout the world, only a small portion have actually been published with varying degrees of detail. The 1968 article of T. Haevernick is the earliest comparative and overall study of these items (see Haevernick 1968) and a main point of reference for those studying comparable pieces. While Haevernick does her best to examine pieces from all find regions, she (and other authors) concedes the problem of limited access to materials, especially those from the Black Sea region, and makes no claim towards a comprehensive study. Useful material on double-sided relief ornaments is also found in E. M. Alekseeva’s 1978 work on ancient beads of the Northern Black Sea region, which presents a sample of regional finds together with other core-formed face beads.12 T. Rossell completes the regional overview with a brief compilation of the finds from present-day Catalonia and Ibiza (see Rossell 2003). For the remaining geographic distribution, scholars must look at a variety of individual publications and museum catalogues to find more than a fleeting mention of such items and perhaps a poor image. An up-to-date and detailed comparative analysis of all items known from the main centres of find concentrations – Carthage, Eastern Spain and the Northern Black Sea – is still pending (see Dan (2011: 220).

13  This contrasts the predominance of male depictions in nonmoulded core-formed glass face ornaments. Compare for example the pendants in Seefried 1982. 14  Female depictions: Haevernick (1968: pl. 6, nos. 1–7, see also exhaustive catalogue listing) Young (1949: pl. 64, nos. 1–2), Alekseeva (1978: pl. 34, nos. 11–14), and Rossell (2003: figs. 3, 5). Also see Zwierlein-Diehl (1991: no. 2482), Sternini (1998: nos. 101-102), Spaer (2001: nos. 321-322, cf. fig. 72), Adam-Veleni (2010: no. 495), ArvellierDulong and Nenna (2011: nos. 481-486), Platz-Horster (2012: no. 138), Schwarzer and Rehren (2015: figs. 7,74-7,75). 15  Note that with the exception of illustrated pieces, Haevernick’s catalogue does not indicate whether the male faces are bearded or not. Therefore, I present this generalization with some caution, but judging by accessible published images, bearded male faces predominate. Note, however, that it is important to study all images and finds to understand the real percentage of beardless or negroid faces; Male depictions: Haevernick (1968: pl. 6, nos. 8–10 – all bearded), Cramer (1908: 83-84, no. 1101, fig. 53 b, d – both bearded, fig. 53e – beardless – negroid), Alekseeva (1978. pl. 34, nos. 15 – beardless, 16 – negroid, 17 – bearded), Belov (1962: 175, fig. 38 no. 8 – bearded), Kopylov (2006: fig. 1, nos. 8–9 – both bearded (the latter = Cat. Paris 2001: 105 no. 65), 10 – combination of bearded + beardless; Note that the figures are incorrectly labelled within the text) and Erlikh (2011: 69 fig. 22, nos. 26–27 - bearded), OIAK (1893 for 1891: 5, no.3), Rossell (2003: 24 figs. 2, 4 – both bearded). Also see Spaer (2001: nos. 323324 – bearded), Adam-Veleni (2010: no. 496 - bearded), Wünsche and Steinhart (2010: no. 21), Arvellier-Dulong and Nenna (2011: nos. 487488 – both bearded), Platz-Horster (2012: nos. 140-142 – all bearded). 16  A combination of two contrasting facial types – bearded and beardless – is quite rare. Note a similar contrast in a double-sided relief pendant showing a squatting Bes figure, bearded on one side and beardless on the other (Platz-Horster 2012: no. 139). 17  However, a unique double-sided relief pinhead from Pergamon shows a male-female combination with a ‘Medusa’ head on one side and a standing ‘Apollo’ figure on the other. See Schwarzer and Rehren 2015: 112, Taf. 7,76, also Schwarzer 2008: G44. 18  See the male examples listed in Haevernick 1968 as well as the bearded examples given in n. 15. 19  For other known pinheads see n. 20.

Overall analysis – male and female depictions All double-sided relief ornaments of glass depict portrait faces in frontal view, with most known 12  See Alekseeva 1978. It is worth noting that Haevernick’s catalogue is more comprehensive than Alekseeva’s work.

93

Wonders Lost and Found pinhead. Pinhead ornaments were fashioned from the same forms as pendants. Instead of attaching a loop at the top, however, a pin would have been inserted into the basal perforation. Only a small handful of ornamental pinheads have clearly been identified as such and published.20 Nonetheless, their distribution – like that of all other double-sided relief ornaments – touches upon various parts of the Mediterranean as well as Colchis (Figure 3), demonstrating a widespread yet limited popularity throughout the Mediterranean network as a whole.

workshops,23 most likely Carthaginian, based on the extremely high concentration of such finds in Carthage, where numerous glass workshops have been identified.24 M. Seefried has pointed out that 90% of all glass mask ornaments – very similar in their general form to the double-sided relief ornaments – come from find sites of Carthage (Seefried 1982: 40). Certainly, it is worth considering the possibility of production – perhaps using imported moulds – in other significant areas, for example in the Northern Black Sea region, where glass production has also been identified for the Hellenistic period.25 On the other hand, the relatively brief existence of decorative pendants and pinheads during the 4th-3rd centuries BC may not have sufficed to popularize such items and facilitate their manufacture outside the area of their main find concentration (Carthage/North Africa).

The smaller number of pinheads as compared to pendants may suggest a smaller demand. However, if we consider that glass objects, in particular jewellery, were often manufactured with the idea of imitating – to some extent – precious stones with relief decorations, then these decorative pinheads would have been considered as valuable as their pendant counterparts.21 Certainly, the fact that the known find contexts are funerary demonstrate that these items were considered of some value.

The overall features of the Pichvnari ‘Heracles’ demonstrate the realistic proportions and features common to Hellenistic portrait images and contrasting the bulk of core-formed glass pendants and amulets. However, certain features are rendered and emphasized in a fashion reminiscent of the more highly stylized, hand-fashioned bearded male faces or ‘masks’ of contemporary or earlier periods (lists may be found in Seefried 1982 or Alekseeva 1978).

Place of Manufacture – find distribution and stylistic criteria A serious discussion of possible production centres for this object category is rendered quite difficult due to a small number of known finds and glass production sites, combined with a lack of documentation. However, the overall distribution of known finds shows the largest concentration of glass relief pendants and pinheads – approximately one third of all known finds – to be in Carthage (Haevernick [1968: 649–50] lists 37 examples). A significant concentration is also found in the Northern Black Sea region, with many items found in and around the ancient Bosporan Kingdom and Colchis (see Figure 3).22 We also find a number of ornaments coming from present-day Catalonia and Ibiza (Rossell 2003).

Particular resemblance can be observed between ‘Heracles’-type double-relief ornaments and Hellenistic bearded face pendants decorated with twists of glass to mark facial hair (Figure 4) – pendants of this type have been described in detail by M. Seefried under the category of CIII.26 Full facial hair features in both CIII pendants and the Pichvnari ‘Heracles’ and frames the entire face: while abundant hair and beard are denoted by glass curls on the former, ‘Heracles’’ thick hair and lush beard are moulded and defined into the relief. In both cases, hair and beard sections meet at cheekboneeye level. Note the outward radiation of the individual beard curls in the CIII pendants, which is clearly

To date, scholars concur that these pinheads and pendants were produced in Punic-Phoenician

For example, Haevernick (1968: 649), Young (1949: 432). Haevernick (1968: 649) also notes the frequent use of ‘Phoenician’ instead of ‘Punic’ in such discussions. 24  Haevernick (1968: 649). See also Seefried 1982: 38-40. While Seefried lists specific finds of glass ovens, Docter and Sonneveld (2009: 132) note the probability of other glass production sites in the Carthage settlement area indicated by finds of glass wasters. Glass-working ovens of the Hellenistic period have also been found elsewhere, e.g. on Rhodes – see Seefried 1982: 38. 25  Evidence for Hellenistic glass workshops is limited for the Northern Black Sea region – in contrast to the evidence for the Roman period. See Ostroverkhov (1981) for a glass workshop with bead production of the 4th century BC discovered in the Yagorlyk settlement, Northern Black Sea region (Ukraine, between the Bug and Dnepr Rivers). N. Kunina (1997: 24) notes this as the oldest known example of glass production in the Black Sea area. She also notes the presence of Hellenistic glass manufacture in the nearby Caucasus (Armenia). 26  Seefried 1982: 105-16, 159; 1979: 20, Figs. 9-10. Haevernick (1977: 305-8) characterises this pendant type as Group 1 ‘klassisch karthagisch’. 23 

20  Haevernick 1968: 649–50, 652, nos. 25 – bearded? male, 43–44, 50, 56–7 – all female, 104–5. The latter two = Filarska 1962: 84, nos. 86 – female, 87 – bearded male, figs. 25, 1-2. See also Schwarzer and Rehren 2015: figs. 74–5 – both female. 21  See Pliny’s description (35.30; 37.75) of the growing industry in glass imitations of clear and milky semi-precious stones. See also Haevernick (1968: 649). Seefried (1982: 53) points out that glass pendants are frequently found in conjunction with beads or necklace details of gold, emphasizing the value of glass as a material. For a discussion of milk-glass imitations of murrhine ware see Tressaud and Vickers 2007. 22  See Haevernick 1968: 651, nos. 63–71 – with find sites indicated – as well as nos. 72–83 – which are non-inventorized items from Russian collections, including bibliographies. For the pendants from Elizavetovskoye gorodishche see Kopylov (2006: 71, 73, fig. 1, nos. 5–10). The pendant from Tenginskaya is published in Erlikh (2011: 69, fig. 22, nos. 26–27). Arvellier-Dulong and Nenna (2011: nos. 481, 487) have also published an additional five ornamental pinheads from Kerch.

94

Sujatha Chandrasekaran: A double-sided glass relief pinhead from ancient Colchis

CIII finds and double-sided relief ornaments, come from (see Haevernick 1968: 649-50, nos. 6-37). Taken together with the predominance of finds from this area, it seems fairly certain that double-sided glass relief ornaments were produced in North Africa or Carthage in particular and distributed to sites of the Northern and Eastern Black Sea through the Mediterranean trade network.29 At this point, it is worth briefly addressing the problem of identifying workshops and/or individual craftsmen for these items. Haevernick believes the level of variation amongst the double-sided relief pendants too insignificant for production in multiple centres and suggests a single workshop or even a single craftsman (Haevernick 1968: 648). At the same time, however, enough variation can be observed in the individual male (and female) reliefs – particularly in hair form and execution of detail – to demonstrate the existence of multiple forms and the possibility of multiple craftsmen or workshops.30 Function While the original function of pinheads and pendants in the Mediterranean is generally considered to be of apotropaic nature, it remains unclear whether they embodied the same protective function in the Black Sea region or were simply of a more decorative nature (see Haevernick 1968, 649). It is noteworthy that the bulk of the double-sided relief ornaments has been recovered from tombs, both in North Africa and the Black Sea region, perhaps indicating a similar functionality in both regions.

Figure 4. A glass face mask from Carthage, 350-200 BC. Warsaw National Museum, Inv. no. 142640. (After Filarska 1962, pl. 1.3)

mimicked by the radiation pattern of ‘Heracles’’ thick beard strands.

Conclusion

Additional similarities between the two ornament categories include special emphasis of eyes and lips. While defined through oversizing (eyes) or pronounced rounding (lips) in CIII pendants, the Pichvnari ‘Heracles’ uses more subtle framing of said features – the marked borders contouring the eyes, and the lush moustache draped around the lips to emphasize their roundness (compare Figures 2a and 4).27

The ‘Heracles’ from Pichvnari is a unique piece on two levels: it is a rare example of a glass ornamental pinhead with bearded male relief as well as a singular find of this category from ancient Colchis. It was clearly of some value as a rare glass import in Colchian territory, and was placed in the grave of the deceased either as an apotropaic feature or simply as a valued decorative item. Its excellent preservation, rare milky-yellow colour and defined features render it an excellent piece by which to study items of this kind, especially in order to analyse the iconographic style. ‘Heracles’ came to Colchis as an import, perhaps brought directly to the region from the Mediterranean or indirectly through

The resemblances discussed here illustrate a shared iconographic tradition between double-sided relief ornaments with bearded male heads and bearded pendants of the CIII category, not to mention with many other categories of core-formed face pendants and amulets.28 This also speaks for a shared provenance in North Africa, most likely in Carthage, where the bulk of known core-formed face ornaments, in particular

29  Yu. L. Shchapova (1978) notes the wide scope of sites from which these items may have entered the Black Sea region, naming Egypt, Mesopotamia, Phoenicia and even Syria, Rome (Italy) and its European provinces. 30  A detailed comparative analysis goes beyond the scope of this paper. However, it is enough to observe variations amongst female reliefs (see those listed in n. 18) and the male reliefs discussed here (see n. 19).

Compare ‘Heracles’ with other CIII pendants indicated in the previous footnote. 28  Compare, for example, the male faces of groups BA, BII and F1 in Seefried (1982: Pl. 1, 4), also the male beads of Group 2 according to Haevernick (Haevernick 1977: 310, fig. 3). 27 

95

Wonders Lost and Found the Cimmerian Bosporus, where many similar items have been found.

Материалы международной археологической конференции (г. Краснодар, 27-29 мая 2013г.): 146-151. Cramer, M. 1908. Griechische Altertümer südrussischen Fundorts aus dem Besitze des Herrn A. Vogell, Karlsruhe. Kassel. Dan, A. 2011. La Mer Noire et le Levant ancien: quelques domaines d’enquȇte’ Rivista di Studi Fenici 39/2: 211-58. Docter, R. and J. Sonneveld 2009. Punic Glass from Carthaginian Settlement Excavations (Carthage Studies 3). [Erlikh] Эрлих, B.P. 2011. Святилища некрополя Тенгинского городища II, IV в. до н.э. Москва. Filarska, B. 1962. Sykla starozytne. Warsaw. Haevernick, T.E. 1968. Doppelköpfchen. Wissenschaftliche Zeitschrift der Universität Rostock 17: 647-53. Haevernick, T.E. 1977. Gesichtsperlen. Madrider Mitteilungen 18: 152-231. Kakhidze, A. and N. Vashakhidze 2010. Pichvnari 3. The Hellenistic World and Colchis. Types of Burials and Burial Customs in South-Western Georgia in the Hellenistic Period. Batumi/Oxford. [Kopylov] В.П. Копылов, 2006. О проникновении пунийских товаров на нижний Дон в IV в. до н. э., in Международные отношения в бассейне Черного моря в скифо-античное время. Ростов-на-Дону. Kunina, N.Z. 1997. Ancient Glass in the Hermitage Collection. St. Petersburg. Lightfoot, C.S. 2001. The Pendant Possibilities of CoreFormed Glass Bottles’ Metropolitan Museum Journal 36: 59-66. [Ostroverkhov] Островерхов, A.C. 1981. Древнейшнее античное производство стеклянных бус в Северном Причерноморье. Советская археология 4. Platz-Horster, G. 2012. Erhabene Bilder. Die Kameen in der Antikensammlung Berlin. Wiesbaden. Rossell, T. 2003. Punic Two-Faced Masks. Annales du 15e Congres de l’Association Internationale pour l’Histoire du Verre, New York – Corning 2001: 23-5. Nottingham. Schwarzer, H. 2008. Die Stadtgrabung. Teil 4. Das Gebäude mit dem Podiensaal in der Stadtgrabung von Pergamon. Studien zu sakralen Banketträumen mit Liegepodien in der Antike (Altertümer von Pergamon 15,4). Berlin/ New York. Schwarzer, H. and T. Rehren 2015. Antikes Glas aus Pergamon. Erste Ergebnisse archäologischer und naturwissenschaftlicher Untersuchungen, in U. Kästner, A. Scholl (eds) Pergamon als Zentrum der hellenistischen Kunst. Bedeutung, Eigenheiten und Ausstrahlung: 106-134. Petersberg. Seefried, M. 1979. Glass core pendants found in the Mediterranean area. Journal of Glass Studies 21: 17-26. Seefried, M. 1982. Les pendentifs en verre sur noyau des pays de la Méditerranée antique (Collection de l’École Française de Rome 57). Rome. [Shchapova] Щапова, Ю.Л. 1978. Приложение: О происхождении некоторых типов позднеантичных стеклянных бус, Свод археологических источников (САИ) Вып. Г 1–12. Т.2. Mосква.

Acknowledgements I am honoured by N. Sekunda’s invitation to contribute to this volume dedicated to my former PhD supervisor Michael Vickers, who carefully guided me through the adventurous maze of doctoral research at Oxford University and has been a solid source of encouragement at all times. It was he who introduced me to the archaeological sites of Colchis, in particular to Pichvnari, where I had the privilege to work with him in 2008 and also became acquainted with the Pichvnari ‘Heracles’. Michael Vickers was also so kind as to provide me detailed images and the measurements of the ornamental pinhead when I expressed the wish to work with it. In this respect, I am also grateful to Tamar Vashakidze from the Batumi Museum for the opportunity to study the ‘Heracles’ directly as well as providing me detailed information on it, and also for permission to reproduce photographs of the object. I would further like to thank Holger Schwarzer from the University of Mainz for sharing information about two similar objects from Pergamon with me and kindly providing me photographs as well as alerting me to the relevant publication (see Schwarzer 2008, Schwarzer and Rehren 2015). I am also grateful to Nicholas Sekunda for directing me towards T. Rossell’s publication and the finds from Eastern Spain. My thanks also extend to Anca Dan from the École Normale Supérieure in Paris for engaging in discussions on glass wares of ‘North African’ provenance from Northern Black Sea sites, and also for providing me a copy of her 2011 article on the topic.

Bibliography Adam-Veleni, P. 2010. Glass Cosmos. Thessaloniki. [Alekseeva] Алексеева, E.M. 1978. Античные бусы Северного Причерноморья Свод археологических источников (САИ), вып. Г: 1–12. Т. 2 . Москва. Arvellier-Dulong, V. and M.-D. Nenna 2011 Les verres antiques du Musée du Louvre 3. Parures, instruments et elements d’incrustation. Paris. [Belov] Г.Д. Белов 1962. ‘Эллинистический дом в Херсонесе’ Труды Гос. Эрмитажа 1962/3. Leningrad. Cat. Paris 2001. L’Or des Amazones. Peuples nomades entre Asie et Europe VIe siècle av. J.-C. - IVe siècle apr. J.-C. Musée Cernuschi 16 mars – 15 juillet 2001. Paris. [Chandrasekaran] Чандрасекаран, С. 2013. Об одном стеклянном двухстороннем рельефном лицевом орнаменте-маске из Пичвнари с изображением «Геракля», в Т.А. Павленко, Р.Б. Схатум, В.В. Улитин (ред-ы), III «Анфимовские чтения». 96

Sujatha Chandrasekaran: A double-sided glass relief pinhead from ancient Colchis

Spaer, M. 2001. Ancient Glass in the Israel Museum. Beads and Other Small Objects. Jerusalem. Sternini, M. 1998. La collezione di antichità di Alessandro Palma di Cesnola. Bari. Tait , H. 1991. Epilogue, in H. Tait (ed.) Five Thousand Years of Glass. London. Tressaud A. and M. Vickers 2007. Ancient murrhine ware and its glass evocations. Journal of Glass Studies 49: 143-52. Turmanidze, M. 2007. Beads and Pendants from the Pichvnari Necropolis of the Hellenistic Period. (IberiaColchis 3). (Georgian with English summary). Vickers, M. 2011. Coins and pebbles from the AngloGeorgian excavations at Pichvnari, in Homines, Funera, Astra, Proceedings of the International Symposium on Funerary Anthropology: 129-33. Alba Iulia. Vickers, M. and A. Kakhidze 2004. Pichvnari, Ajarian AR, Georgia. Anatolian Archaeology 10: 21-2. Wünsche, R. and M. Steinhart 2010. Schmuck der Antike. Ausgewählte Werke der Staatlichen Antikensammlung München. Munich. Young, R. 1949. An Early Amulet Found in Athens, in Commemorative Studies in Honor of Theodore Leslie Shear (Hesperia Supplement 8): 427-33, 500. Zwierlein-Diehl, E. 1991. Die antiken Gemmen des Kunsthistorischen Museums in Wien 3. Munich.

97

Gold jewellery from Kavtiskhevi Darejan Kacharava In 1983 excavations took place in Kaspi region some 45 km from Tbilisi, in the village of Kavtiskhevi at a place called Sakaraulo Seri. A rich burial of the early 3rd century BC had been unearthed in the Tsikhiagora necropolis (Tskitishvili 2001: 41-46). Outstanding among the gold objects worn by the deceased were a pendant and a pair of earrings, remarkable on account of their distinctive openwork filigree technique. The objects in question are now in the Simon Janashia Georgian National Museum. The pendant (inv. no. 103-984:6; Figure 1) has a flat and circular hook and is decorated with three spiral wires in the centre and at the edges. The hook is attached by means of a ring that has a plain wire on the lower edge. The pendant was created by means of an openwork net. The net itself is quite sophisticated, and forms a rosette in its upper part. In the centre there are two twisted wires soldered at the crossing points and decorated with granules. This net is attached to the rosette and the soldering point is concealed with a granule. The lower part of the net consisted of four wires attached to the finial in the form of a cluster of four granules. The lower wires are connected to the central part of the pendant by means of hooks. The net had a brownish gem-stone (?) inserted in the net. The pendant weighs 0.89 grams and is 1.9 cm high.

Figure 1. Pendant from Kavtiskhevi, Georgian National Museum, inv. no. 103984:6.

The earrings (inv. no. 103-984:5; Figure 2) are identical; they both have circular hooks slightly parted made with

Figure 2. Earrings from Kavtiskhevi, Georgian National Museum, inv. no. 103-984:5.

98

Darejan Kacharava: Gold jewellery from Kavtiskhevi

round wires. Pendant parts are loosely attached, rather as with the larger pendant described above. There are minor differences between the two earrings; one has a hook decorated with beaded instead of twisted wire, attached to the suspension with a plain ring. The net is broader and of a different colour and the inserted bead has a bluish tint. The suspension is 1. 9 cm long and the rings are 2.8 and 2.9 cm in diameter.

hair decoration and is now in the Pergamum Museum, Berlin (Platz-Horster 2005: 92). In a later period the open-work filigree technique was modified and became simpler: the net was attached to a gold background and ceased to be open-work, although the granulation was still applied to the soldering points. This development can be observed on a bracelet with lynx finials in the Benaki Museum at Athens (Segall 1938: No. 35; Pfrommer 1990: 340, No. TA109, fig. 16, 53). Details of a two-part bracelet in de Clerq-Boisgelin collection were also made with the same technique; the bracelet was finished with a lynx’s head at one end and the head of an antelope at the other (Coche de la Ferté 1956: 65, pl. 25, 1; Deppert-Lippitz 1985: 266-267, fig. 200; Pfrommer 1990: 332, no. TA6, fig. 16, 47). Bracelets with bulls’ heads in the Stathatos collection also display a similar technique (Amandry 1953: 113, nos. 253-254, pl. 44; Pfrommer 1990: 332, no. TA12-13, fig. 16, 33).

The same type of jewellery is known from elsewhere Georgia: as chance finds at the Samadlo settlement (Gagoshidze 1981: 45, no. 11, pl. X, 11), and at Takhtisdziri Necropolis (Burials nos. 3 and 8). The complex at Samadlo is dated to the late 3rd century BC (Gagoshidze 1979:76); at Takhtisdziri, Burial no. 3 has been dated to the 4th century BC and Burial no. 8 to the late 4th century BC. The latter has not yet been published. Openwork filigree jewellery is unusual among the gold jewellery of Georgia. The technique was not very widespread in Classical times. Decoration combining openwork and filigree are more characteristic of Etruscan gold jewellery of the 7th century BC in the second phase of the Orientalising style (Coche de la Ferté 1956: 37, 76, pl. 29, 1-2). Openwork cylindrical beads entirely covered with filigree have been found in one of the burials at Sindos dating to 560 BC. It is noteworthy that the same technique has been attested on jewellery of the 8th-7th centuries found in Georgia (Gagoshidze 1976: 5-38).

These bracelets are dated to the 2nd century BC. A diadem from Kerch now in Munich is also decorated with a combination of reticulation and granulation, but the granules are not always placed on the points where wires were soldered. The Kerch diadem is also dated to the 2nd century BC (Deppert-Lippitz 1985: pls. 28-29). A small cylindrical bead in the British Museum is also decorated with a net and filigree, but its poor state of preservation does not allow us to determine whether the net had a metal background or whether it had an open-work effect (Marshall 1911: 233, no. 2071).

The technique of making openwork nets with hundreds of small criss-crossing wires with granules over the soldered points came into use in the 4th century and lasted until the 2nd century BC. A sceptre from the Tarentum Treasure in the British Museum was decorated with this technique and dates to c. 350-320 BC (Marshall 1911: 232-233, no. 2070; Williams, Ogden 1994: 203, no. 134). The only difference here is that the soldered points were decorated not with filigree, but with small gold rings filled with enamel inlays. The same technique with filigree can be seen on a gold cylindrical object now in the Museum of Arts in Brussels. This piece of jewellery is crowned with acanthus leaves and has three chains suspended from the lower part. It is thought to be a pin of the Hellenistic period (Oliver 1999: 85-90, fig. 11).

An open-work technique with filigree has been attested in jewellery in the Orient; for example, three pairs of earrings in the Pasargadae Treasure were executed in the same technique. David Stronach considers that the Pasargadae Treasure to have been buried around 331330 BC (Stronach 1978: 168-169, nos 2-4, pls. 148-150, fig. 85, 1-3; Rehm 1992: 152-153, fig. 127-129, F100-F102; Musche 1992: 273-274, no. 6.3.1; 6.3.2; 6.3.3). All three belong to an Achaemenid type of circular earrings. On one occasion the same technique was used to decorate only the outer edge of an earring: the outer surface was decorated with a row of 16 rosettes made of gold wire and the centre with double discs. The petals of each rosette were decorated with filigree, so that here we see the combination of both techniques of granulation and of filigree. Another pair of earrings had spirals made of wire in the central part which served as a background for oval sheets with blue paste inlays. The earring had three suspensions also incrusted with blue paste. The third pair of earrings had open-work nets decorated with filigree and a suspension in the shape of an openwork net with a pyramidal filigree cluster; a lapis lazuli was inserted into the mount.

Cylindrical beads on a necklace from Panticapaeum (Hermitage, St Petersburg) are decorated with an open-work filigree technique and are dated to the 3rd century BC (Deppert-Lippitz 1985: 218, fig. 154-155); similar cylindrical beads on a necklace from Pydna necropolis near Thessaloniki (now in the Virginia Museum of Fine Arts, Richmond) are also made in the same open-work filigree technique and are dated to c. 200 BC (Gold of Macedon 2000: 40, fig. 32). One of the finest examples of the technique from Tarentum served as a 99

Wonders Lost and Found A suspension and a ring of the earring were decorated with filigree concealing the soldered points. The earrings from Pasargadae have no parallels elsewhere although the suspension with a lapis lazuli gem is similar to the earrings and the pendant from Kavtiskhevi. It is probable that the jewellery from Kavtiskhevi as well as the jewellery pieces from Samadlo and Takhtisdziri had been imported from the Achaemenid Empire. This is scarcely surprising or unexpected since complexes of the 4th and early 3rd centuries in both eastern and western Georgia (for example: at Akhalgori, Tsintskaro: Algeti, Kanchaeti, Vani, Itkhvisi, Sairkhe) contain Achaemenid type jewellery (Chkonia 1981: 48; Lordkipanidze 2003: 32-34, 40-41, pls 2.3-4, 3.1, 7.3, 16.3; Margishvili 1992: 18, 56; Nadiradze 1990: 72). Imported gemstones, including carved examples, also made their way to Georgia from the ancient Persian Empire (Javakhishvili 2009: 8593; Lortkipanidze 1981; Javakhishvili, 2007: 117-128). Georgian archaeologists attested Achaemenid toreutic imports in the above-mentioned complexes dated to the 4th and 3rd centuries BC (Gigolashvili 2001: 3336; Guigolachvili 1990: 279-281; Kacharava 2006: 351364; Lordkipanidze 2003: 38-71) as well as Achaemenid glass imports (Pirtskhalava 1983: 79-86; Gagošidze and Saginašvili 2000: 67-73).

Gagoshidze, I. 1979. Samadlo, arkeologicheskie raskopki (Samadlo, Archaeological excavations). Tbilisi. Gagoshidze, I.1981. Samadlo, arkeologicheskie raskopki (Samadlo, Archaeological excavations). Tbilisi. Gagoshidze, I. and M. Saginašvili, 2000. Die achaimenidischen Glasgefäße in Georgien. Archäologische Mitteilungen aus Iran und Turan 32: 6773. Guigolachvili, E. 1990. Les coupes en argent de Vani, in O. Lordkipanidzé and P. Lévêque (eds) Le Pont-Euxin vu par les grecs. Sources écrites et archéologie: 279-281. Paris. Gigolashvili. E. 2001. Omphalosiani pialebi vanis dz. ts. 4 saukunis mdidruli samarxidan’ (Phialae with omphaloi from rich burials of the 4th century BC in Vani), in Antikuri khanis kolkhetis arqeologiis sakikhxebi (The archaeological problems of the Classical Colchis): 33-36. Tbilisi. Javakhishvili, K. 2009. Saqartveloshi aghmochenili Aqemeniduri sabechdavebi (Achaemenid seals from Georgia). Iberia-Kolkheti 5: 85-93. Kacharava, D. 2006. Vertskhlis chamcha Vanidan’ (Silver ladles from Vani), in Shota Meskhia 90, saiubileo krebuli, midzgvnili Shota Meskhias dabadebis 90 tslistavisadmi (Essays in honour of the 90th anniversary of the birth of Shota Meskhia): 351-364. Tbilisi. Kypraiou, E. (ed.) 2000. The gold of Macedon. Archaeological Museum of Thessaloniki. Athens. Lordkipanidze, O. 2003. Akhalgoris gandzi (datarigebisa da istoriuli interpretatsiis tsda) (The Akhalgori Treasure [Studies for dating and historical interpretation]). Ziebani 11: 28-71. Lortkipanidze, M. 1981. Udzvelesi sabechdavebi Iberiidan da Kolkhetidan (The Earliest Seals from Ancient Iberia and Colchis). Tbilisi. Margishvili, S. 1992. Antikuri khanis mdidruli samarkhebi Algetis kheobidan (Rich Burials of the Classical Period from the Algeti Gorge) (Tbilisi). Marshall, F.H. 1911. Catalogue of jewellery, Greek, Etruscan and Roman in the Department of Antiquities, British Museum. London. Musche, B. 19912. Vorderasiatischer Schmuck von den Anfängen bis zur Zeit der Achaemeniden (ca. 1000-330 v. Chr.). Leiden-New York-Copenhagen-Cologne. Nadiradze, J. 1990. Sairkhe, Sakartvelos udzvelesi qalaqi (Sairkhe, an Ancient Town of Georgia). Tbilisi. Oliver, A. Jr. 1999. Aspects of Hellenistic jewellery from Italy, in D. Williams (ed.) The Art of the Goldsmith: 8590. London. Pfrommer, M. 1990. Untersuchungen zur Chronologie frühund hochhellenistischen Goldschmucks. Tübingen. Pirtskhalava, M. 1983. Minis churcheli (Glass vessels). Vani 7: 79-86. Platz-Horster, G. 2005. Gold jewellery from Tarentum. Berlin. Rehm, E. 1992. Der Schmuck der Achämeniden. Münster. Segall, B. 1938. Katalog der Goldschmiede-Arbeiten. Athens.

Bibliography Amandry, P. 1953. Collection Hélène Stathatos, Les bijoux antiques. Strasbourg. Apakidze, A., A. Gobejishvili, A. Kalandadze and G. Lomtatidze 1955. Mtskheta, arqeologiuri kvleva-dziebis shedegebi (Mtskheta, Results of Archaeological Studies). Tbilisi. Bromberg, A. 1991. Gold of Greece, Jewelry and ornament from the Benaki Museum. Dallas, TX. Chkonia, A. 1981. Oqros samkaulebi Vanis naqalaqaridan (Gold jewellery from the Vani settlement). Vani 6 Tbilisi. Coche de la Ferté, E. 1956. Les bijoux antiques. Paris. Deppert-Lippitz, B. 1985. Griechischer Goldschmuck. Mainz. Dzhavakhishvili, K. 2007. Achaemenid seals found in Georgia, in A. Ivantchik and V. Licheli (eds) Achaemenid Culture and Local Traditions in Antolia, Southern Caucasus and Iran. New Discoveries: 117-128. Leiden-Boston. Gagoshidze, I. 1964. Adreantikuri xanis zeglebi qsnis xeobidan (Early Classical Sites from the Ksani Gorge). Tbilisi. Gagoshidze, I. 1976. Masalebi qartuli oqromchedlobis istoriisatvis (chvens tseltagritskhvamde I atastsleulis pirveli nakhevari) (Materials for the history of Georgian goldsmithing in the first half of the first Millennium BC), in S. Janashias saqartvelos sakhelmtsipo muzeumis moambe (Reports of the Simon Janashia Georgian State Museum): 5-38. 32-B. Tbilisi. 100

Darejan Kacharava: Gold jewellery from Kavtiskhevi

Stronach, D. 1978. Pasargadae, A report on the excavations conducted by the British Institute of Persian Studies from 1961 to 1963. Oxford. Tskitishvili, G. 2001. Mdidruli samarkhi Tsikhiagoris samarovnidan (A rich burial from the Tsikhiagora necropolis) Dziebani 7: 41-46. Williams D. and J. 1994. Ogden, Greek gold jewellery of the Classical World. London 1994.

101

Palynological analysis of organic materials from Pichvnari (including the earliest silk in Georgia) Eliso Kvavadze and Maia Chichinadze Ten samples from the Pichvnari necropolis were submitted for analysis in the laboratory of the Institute of Palaeobotanics of the Georgian National Museum by Maka Bokeria, the palaeoethnobotanicist who had participated in the 2008 season of excavations in the Pichvnari necropolis conducted by the joint BritishGeorgian team from Oxford and Batumi. The material mostly consisted of organic charcoal, but there was also a large iron nail from a 4th century BC burial that had wooden remains attached to it. The wood showed traces of some dark-coloured organic matter and textile threads (Figure 1). Textile remains were also visible on a small bronze bell (Figure 2), from which three samples were taken for study. Sand and some organic matter were extracted from inside the body of the bell. The materials were analysed in keeping with current practice (Moore et al. 1991). The most striking result was the discovery of silk fibres, the earliest so far known in Georgia, antedating by some centuries evidence for silk found at Dedoplis Gora (Kvavadze and Gagoshidze 2008). The results Figure 2. Textile remains were found on a small bronze bell.

Wood attached to the nail. Our palynological survey found pollen grains of a variety of plants, as well as a number of fibres of different textiles. A total of 410

Figure 1. Pichvnari, Tomb. Large iron nail with wooden remains and textile attached.

102

Eliso Kvavadze and Maia Chichinadze: Palynological analysis of organic materials from Pichvnari

Figure 3. Pichvnari, Tomb. Arboreal pollen found in the organic remains of wood attached to the nail: 1, 2: Pinus; 3: Pterocarya pterocarpa; 4: Juglans regia; 5: Carpinus caucasica; 6: Alnus; 7: Polypodiaceae.

palynomorphs were counted of which 119 were pollen and 291 samples were of non-pollen palynomorphs. There was much arborial pollen, notably of alder (Alnus), pine (Pinus) and walnut (Juglans regia) (Figure 3).

There were only a few cotton fibres (Figure 7) and even twisted flax (Figure 7). Worthy of note is the fact that we also found fibres that were dyed in different colours such as pink, brown, violet and dark grey. There were spores of Chaetomium, a fungal contaminant that settles on textiles made from plants and destroys their fibre. Other fungal spores included the manure fungus Sordaria and some phytoliths of various cereals.

Pollen of oak (Quercus) and Pterocarya pterocarpa were found in smaller quantities. Other trees included elm (Ulmus), zelkova (Zelkova), hornbeam (Carpinus caucasica), oriental hornbeam (Carpinus orientalis), maple (Acer), and hazelnut (Corylus). Pollen grains of the cultivated grape were also (Vitis vinifera) well represented (Figure 4).

Zoological material was included in our palynological analysis. The remains of ticks and insect epidermises were present in great quantity.

Numerous samples of pollen of cereals, including wheat (Triticum), were identified within the herbaceous group. There was also pollen of different kinds of ruderal vegetation (weeds that occur opportunistically on land that has been disturbed by human activity. These included: great plantain (Plantago), wormwood (Artemisia), and cocklebur (Xanthium).

Funerary platform No. 3. The material from this area included a few pollen grains, but non-palynological fossils were better represented. The pollen of pine, oak, alder, spruce, beech and the cultivated grape (Vitis vinifera) were identified. The pollen of wheat, wormwood and of some wild cereals were found as well as the spores of ferns such as Crimean bracken (Pteridium tauricum).

The remains of the non-pollen palynomorphs complex are of great significance. The tracheal cells of wood bark form the majority in this group. The parenchymal cells of pine were identified. There were many textile fibres made from flax (Figure 5) and silk (Figure 6).

The cells of wood bark predominate among the nonpollen palynomorphs, where the tracheal 103

Wonders Lost and Found

Figure 4. Pichvnari, Tomb. Pollen of cultivated landscapes found in the organic remains of wood attached to the nail, 1-6: Vitis vinifera; 7: Triticum; 8: undiff. Cerealia; 9: Artemisia; 10: Xanthium

Figure 5. Pichvnari, Tomb. Fibres of flax textile (Linum) taken from organic remains of the nail.

Figure 6. Pichvnari, Tomb. 1: fibre of wool textile; 2-6: fibre of silk from the organic remains of the nail.

104

Eliso Kvavadze and Maia Chichinadze: Palynological analysis of organic materials from Pichvnari

Figure 7. Pichvnari, Tomb.1: twisted flax fibre; 2, 3: fibre of cotton (Gossipium).

Figure 8. Pichvnari. Funerary platform. Tracheal cells of pine wood (Pinus).

The remains of flax fibre (Figure 11) and the tracheal cells of pine wood were found in quantity.

cells of pine wood are found in considerable quantity (Figure 8). Many crystals of animal bones were identified (Figure 9).

The spores of fungi, including Chaetomium, were also well represented, as were the phytoliths of cereals. There was a small quantity of cotton fibre, as well as some dyed flax. Sample No. 1 contained one human and three animal hairs, and the remains of a single bird feather. Zoological fossils included the micro-remains of ticks and the bristles and epidermises of insects.

There was a sizeable amount of flax fibre but less cotton. The spores of fungi and the remains of some water plants were found. The zoological material lacked diversity, being mainly represented by tick bristles and the epidermises of insects. Funerary platform No. 8 revealed the same palynological spectrum, with the addition of the indumentum of a moth caterpillar. The small bronze bell. Sample No. 1 taken from inside the bell was rich in palynological material, producing a total of 345 palynomorphs, of which 131 were of plant pollen belonging to 19 taxa. The arboreal group consisted mostly of pine (Pinus) and alder (Alnus); there were a few samples of fir pollen (Abies), spruce (Picea) and oak (Quercus) (Figure 10).

Sample No. 2 was taken from a textile that had been attached to the bottom of the stem of the bell. Here only four pollen grains were recovered (of pine, spruce and cereals) and one spore of Crimean bracken. There was, however, flax fibre in considerable quantities, some dyed. A total of 104 flax fibres were counted, together with the micro-remains of dye accompanied by the remains of the fungus Chaetomium that attacks the fibres of plant textiles. There were some spores of other fungi and the parenchymal cells of pine wood.

The grains of elm pollen (Ulmus), walnut (Juglans), beech (Fagus), hornbeam (Carpinus) and hazelnut (Corylus) were also identified. The herbaceous group contained much fern pollen such as that of Crimean bracken (Pteridium tauricum), as well as a cereal pollen. Wheat (Triticum), millet (Panicum) and other cereals were found, but their state of preservation made identification difficult. Weed pollen was also present.

Sample No. 3 came from the top of the stem of the bell. Examination beneath a binocular microscope (x 20) revealed textile remains. Palynological analysis of the sample revealed 36 flax and two cotton fibres. There were a sizeable number of brown dye particles. An unknown textile fibre was also found as well as spores of the Chaetomium fungus and the tufts of hair of carpet-beetle larvae. 105

Wonders Lost and Found

Figure 9. Pichvnari. Funerary platform. 1-4: bone salt crystals: 5-feather of bird.

shroud of the deceased. The wood itself probably came from a funerary couch.

Analysis of the results The results of the survey carried out on the nail sample and from the sacrificial area of the Pichvnari necropolis enable us to draw certain conclusions. The arboreal pollen samples indicate that at the period when the burials took place, the area was well assimilated to human occupation, and both agriculture and viticulture were well developed. There must have been walnut trees and hazels in gardens and arable farming thrived, as the wheat pollen bears witness. The scarcity of manure fungi might be indicative the fact that while animal husbandry may have been developed, agriculture predominated.

The great quantity of silk fibres and other textile fibres of flax, cotton and wool, together with the particles of different colours of dye, give the strong impression that the deceased were buried wearing beautifully woven clothing. We might even speculate that women especially wore woven silk dresses. The materials that came from the funerary platforms revealed a great quantity of bone crystals that emanated from food that contained meat. The cells of pine wood could be interpreted as an indication of wooden vessels, in which the food had been stored. The flax fibre from the sacrificial area could mean that a linen cloth had been used to wrap the food or cover the vessels.

As for the wild plants, it is probable that trees were also occurring naturally on the landscape including broadleafed trees such as oak, hornbeam, zelkova and elm. The Caucasian wing-nut (Pterocarya fraxinifolia) and alder were probably located along the river. The inventory of plants allows us to speculate that the climate of the time in question (4th century BC) was warm and mild, otherwise it would have been impossible for the zelkova and wing-nut trees to grow here.

It should be also noted that the palynological samples of organic materials that came from the bronze bell is quite different from the spectrum of the finds from the nail sample. Among the arboreal pollens there were examples of fir (Abies) and greater quantities of spruce (Picea) and beech (Fagus). These usually grow in mountainous areas at high altitude, and the fact that these palynological species appear here should indicate that the bell was not local, and for a long time before

The remains of the various textiles attached to the nails could have been the remains of clothing or of the 106

Eliso Kvavadze and Maia Chichinadze: Palynological analysis of organic materials from Pichvnari

Figure 10. Pichvnari. Bronze bell. Pollen taken from organic remains: 1: Abies nordmanniana; 2, 3: Pinus; 4: Picea; 5: Juglans regia; 6: Alnus; 7: Ulmus; 8: Carpinus caucasica; 9: Pteridium aquilinum; 10: Poaceae.

burial it was located in the high mountainous part of Colchis. The heat-loving plant pollen of zelkova, lime and the cultivated grape are also absent from the samples taken from the bell. The bell seems to have been deposited in the grave after it was wrapped in a linen cloth and was fastened with a leather strap, indicated by the discovery of some tufts of hair of leather beetle larvae. The character of the pollen grains, spores and tick remains in the samples from the nail and the bell allow the conclusion that the burial containing the nail took place in the springtime, while the bell formed part of a burial that took place in the autumn. Bibliography

Figure 11. Pichvnari. Bronze bell. Flax fibres taken from organic remains of the bell.

107

Kvavadze, E. and I. Gagoshidze 2008. Fibres of silk, cotton and flax in a weaving workshop from the first century AD palace of Dedoplis Gora. Vegetation History and Archaeobotany 17: 211-215. Moore, P.D., J.A. Webb and M.E. Collinson 1991. Pollen analysis. 2nd edn. Oxford.

Mercurial metrics Kenneth Lapatin Silversmiths regularly tried to make silver vessels in ‘round’ figures in terms of local, or traditional, weight standards* Michael Vickers’ broad research has cast new light on subjects as diverse as Greek history and theater, ancient chronology and numismatic standards, and Byzantine and Renaissance architecture and painting. His studies of materials ranging from hard stones to glazed ceramics and their changing relative values have opened up important new avenues of study. In the field of ancient plate, he has long argued that finely figured gold and silver cups and unadorned vessels alike not only expressed their owners’ wealth, status, and prestige, but also functioned as ‘large denomination bank notes’ in the monetized Mediterranean world, and that despite their fine craftsmanship they served as a reserve of bullion that could be melted down in time of need. Greek temple inventories inscribed on stone frequently record the fixed weights of such objects, and surviving artifacts seem to tell the same story.1

the inscriptions recording weights of objects are remarkably meticulous, utilizing notations as accurate as 1/24 of an ounce (a scrupulum, hence the English scruple). Such exactitude was evidently thought to be possible and, evidently, desirable. Obviously, objects with such precise inscriptions did not have round, even weights – or no longer did when they were inscribed. Perhaps (some of?) these inscriptions, often incised lightly into the bottom of a vessel, are the result of some kind of inventorying and/or reweighing, either periodically, prior to exchange, or following some wear, damage, or repair. The Berthouville Treasure (Figure 1), unearthed on March 21 – the first day of Spring – 1830, by Prosper Taurin, a farmer plowing a newly acquired field near the village of Berthouville in northern France, recently conserved by the J. Paul Getty Museum and exhibited across the United States, in Southern France, and in Denmark before returning to its permanent home in the Département des Monnaies, médailles, et antiques of the Bibliothèque nationale de France (aka the Cabinet des médailles), provides more fodder for such discussions.2 The hoard of silver vessels, weighing approximately 25 kilograms in toto, consists primarily of vessels, but also includes two statuettes and various components and fragments, many of which retain traces of gilding (the contribution of which to the objects’ ancient value is another open question). Dedicatory inscriptions on about 30% of the vessels indicate that the hoard was dedicated to Mercury Augustus and Mercury Cannatonensis, and later 19th-century excavations at the find site revealed the foundations of a typical Gallo-Roman fanum and a nearby theater-like building, which have also been visualized more recently through electrical resistivity imaging.3

Gold and silver vessels sometimes bear ancient inscriptions that record their weights. These inscriptions, however, are often less straightforward than might be imagined. First, whether scratched lightly as graffiti or carefully punched, they are not always easy to read or interpret; second, the objects themselves are often worn, damaged, or incomplete, missing lids, handles, or other parts, making it difficult to correlate the figures inscribed with the objects’ present weights; third, some inscriptions evidently record the weights of multiple objects – pairs or groups – together, rather than just those of the individual pieces; fourth, how are we to calculate the ancient units of measurement? And how precise can we expect the ancient craftsmen or caretakers to have been? Modern studies have customarily begun with coins, averages, and division to determine today’s equivalent of an ancient obol or drachma, shekel or siglos, denarius or sestertius, and then multiplied up to a mina or talent, uncia, pound, or some other measure. Yet the coins we have at our disposal have suffered wear and loss over the centuries as have silver vessels, which have often also been restored or, more recently, conserved. How closely do these object’s current weights approximate their original ones? Some of * 1

While at the Getty, the Treasure was carefully examined, both visually and analytically, using both invasive and non-invasive technologies. Most of the objects were sampled and analyzed with inductively-coupled plasma The standard account of the Treasure remains Babelon 1916. See also Lapatin 2014 and more recent publications mentioned in the acknowledgments. 3  For an account of the discovery and acquisition of the Berthouville Treasure, as well as the archaeology of the site, see Babelon 1916, and, more recently, Avisseau-Broustet 2014, with further references. 2 

Vickers 1992. Vickers 1990; 1992; Vickers and Gill 1994. See also Lapatin 2015a.

108

Kenneth Lapatin: Mercurial metrics

Figure 1. The entire Berthouville Treasure

Figure 2. Dedications of Quintus Domitius Tutus

mass spectrometry (ICP-MS) to determine the purity of their silver and its composition. Of 80 samples, 52 are over 90% pure.  Twenty-one of the remaining 28 are above 80% pure. Analyses identified some modern components added to vessels as replacements in the 1830s, and other outliers may well be the result of sample contamination or other error.4

Some of the vessels from Berthouville bear weight as well as dedicatory inscriptions. While such inscriptions are usually located on the bottom of vessels, or the rims, the most extraordinary were revealed by X-rays: the smooth inner liners of two silver scyphi depicting Centaurs dedicated by Quintus Domitius Tutus (BnF 56.6-7, Figure 2, bottom left and right) each have punched inscriptions (Figure 3). These each record the weights of those components not of the entire cup – and thus testify to the careful tracking of the precious metal in the workshop, for the inscriptions were never intended to be visible to the

The analyses were conducted by Lynn Lee of the Getty Conservation Institute and Patrick Degryse of the Katholieke Universiteit Leuven in collaboration with Mark Walton and Karen Trentelman. See also Sánchez and Lansing-Maish 2014.

4 

109

Wonders Lost and Found cups’ users. Written with standard Roman notation, the inscriptions also indicate that the cups were manufactured in the Latin west, rather than the Greek east (see Sánchez and Lansing-Maish 2014: 116-117). Rather different is the inscription on the back of a large plate dedicated by Gaius Propertius Secundus (BnF 56.15, Figure 4) with a lion hunt incised in the tondo and a relief border depicting wild animals, theatrical masks and other Bacchic implements, musical instruments, and sacred buildings. The lion hunt in the tondo, with a sword-bearing horseman seen from three-quarter rear view, recalls the mounted deer hunter from the hunting fresco on the façade of the late fourth-century B.C. Tomb II at Vergina.5 Unlike many of objects from Berthouville, this plate is unlikely to have been made originally for dedication to Mercury. For nowhere in its rich iconography does the god appear. However, around the tondo an inscription has been neatly cut between two circular guidelines, evidently added prior to dedication:

Figure 3. X-ray of skyphos BnF 56.6 revealing inscription

Figure 4. Hunting plate BnF 56.15 See, e.g. Andronicos 1984: 107; and, more recently, H.M. Franks 2012: 8-9, figs 8, 12.

5 

110

Kenneth Lapatin: Mercurial metrics

Figure 5. Hunting plate BnF 56.15, weight inscription on reverse

theoretical relationship to the pound was in some cases not certainly known.’6

DEO MERCVRIO KANNETONESSI GPROPERT SECVNDVSVSLM (To the god Mercury of Cannetonum, Gaius Propertius Secundus fulfilled his vow willingly as deserved.)

Calculating at 326 (the midpoint between 322 and 330) grams per Roman pound, the original 3 pounds, 4 ounces (or 3 1/3 pounds – the Roman pound having 12, not 16 ounces like the modern English pound) of Secundus’ plate would have originally weighed about 1087 grams. The 4 grams difference between this figure and Babelon’s certainly, approximately 0.4%, must fall into the allowable margin of error, especially given the damage to the object.

On the back of the plate its weight is neatly inscribed (Figure 5): P III :: (pondo libras tres tientem – three pounds, four ounces). Its current weight, in its damaged state, post-conservation in 2015, is 1,019 grams. According to Babelon it weighed 1,027 grams in 1916. Thus it has lost slightly less than 1% of its weight in the past century. What was its original weight, before it suffered rips and tears at the rim? Babelon calculated 1091.48 grams, using the widely accepted figure of 327.45 grams per Roman pound (Babelon 1916, 118). But this figure is far from universally agreed upon. As Kevin Buther and Matthew Ponting have demonstrated recently in The Metallurgy of Roman Silver Coinage, ‘…the exact weight of the Roman pound remains beyond our grasp. We can be fairly sure that it lies somewhere between about 322g and 330g, assuming that the standard was maintained precisely over a long period of time’ (Butcher and Ponting 2014: 93). The figure of 327.45 grams per Roman pound that is given in most text books is Augustus Böckh’s, and, as Philip Grierson has noted, was ‘based on groups of coins which may not all have been authentic, and whose

In what follows, I have tentatively calculated the possible Roman weights of several of the Berthouville objects by dividing their current, post-conservation weight, by both 326 and 327.5 grams per pound, the latter something akin to the traditional standard, but with less pretension of accuracy. The results are always quite close, and given the potential change from the original weight of the objects over the centuries, the differences are probably negligible.7 6  Böckh 1838; see also Hultsch 1882. Cf. Grierson 1964, quote from p. xii; Martin 1988; and now Butcher and Ponting 2014. 7  Cf. Grierson 1964: xiv: ‘there seems no sufficient justification for altering the accepted figures in terms of which so many calculations have been made, since although there is little positive to be said in

111

Wonders Lost and Found objects, despite losses over time and, in some cases, additions by modern restorers, strongly support the proposition that their weights were calculated to round numbers, thus facilitating conversion and exchange – and this was likely appreciated by divine recipients as well as humans. The Berthouville inscriptions reveal that the sanctuary was frequented by non-citizens as well as Romans, women as well as men, ex-slaves as well as the free born (see, e.g. Deniaux 2006; Lajoye 2008). A local woman, Germanissa the daughter of Viscarius, was a more generous donor to Mercury than Gaius Propertius Secundus. She offered two inscribed vessels to the god. Both are better preserved than Secundus’ plate, but not entirely undamaged: a slightly smaller plate with a gilt parrot (BnF 56.24, Figure 7) and a long-handled bowl apparently depicting Cybele and Mercury’s consort Rosmerta or mother Maia holding caduceus and cornucopia (BnF 56.25, Figure 8). These now weigh 746 and 895 grams, respectively, or a total of 1641 grams or just over 5 Roman pounds (5.03 at 326 grams per pound and 5.01 Roman pounds at 327.5 grams per pound). The three dedications of a certain Creticus, the son of Runas, (BnF 56.26-28, Figure 9) tell a similar tale: two long-handled bowls, undecorated except for bars along the interior rim and spirals at the joins to the handles, have surprisingly disparate weighs: 305 and 288 grams. (Babelon recorded 312 and 290.) A third bowl, without a handle, bears the same dedicatory inscription: MERCVR AVG CRETICVS RVNATIS DSO VSLM. It weighs 67 grams. The total of the three is 659 grams, or 2.02 Roman pounds at 326 (2.01 at 327.5).

Figure 6. Large Mercury statue BnF 56.1

Other vessels, such as a handled bowl with a satyr’s head (BnF 56.31, Figure 10) and a small amphora (BnF 56.35, Figure 1, right foreground) seem to weigh half a pound (today 167 and 166 grams respectively). Two beakers (BnF 56.37-38, Figure 1, left foreground) now weigh 157 and 170 grams, or, together as a pair, 327 grams: a Roman pound.

The large statuette of Mercury (BnF 56.1 – Figure 6) now weighs 2,722 grams, about 8 Roman pounds, 4 ounces,8 but the god has lost both his petasos and the moneybag he held in his right hand. They are evident in several depictions of the god in the same pose from the site (see Figure 1), and if the moneybag was plump, as is likely, this statue may have originally weighed 9 or perhaps even 10 pounds.

Remarkably for a hoard discovered in rural France, the Berthouville Treasure includes some the finest Roman silver to survive from anywhere. Technically and stylistically the most accomplished items were all dedicated by a single individual: Quintus Domitius Tutus. Of the nine vessels he dedicated to Mercury (BnF 56.4-12, Figure 2), six are paired cups and pitchers, while three are singletons. Of these, only one, a small simpulum (ladle) has iconography that is explicitly related to the god, while the others all bear imagery related to Greek mythology (the Trojan War and Omphale) or the world of Bacchus/Dionysos. These eight vessels are composed

Although we do not know the exact integer to divide into the weights of the objects in the Berthouville Treasure to determine their precise ancient weight equivalents, the preserved weights of many of the their favour – … criticism of how they were arrived at is fully justified – they cannot be far wrong.’ 8  The statuette’s current weight is 2.5% less than in 1916. The most recent conservation treatment included the removal of modern protective coatings as well as tarnish; earlier restoration treatments both before 1916 and in the 20th century removed heavy incrustations.

112

Kenneth Lapatin: Mercurial metrics

Figure 7. Germanissa plate BnF 56.24

Figure 8. Germanissa bowl BnF 56.25

113

Wonders Lost and Found

Figure 9. Dedications of Creticus BnF 56.226-28

Figure 10. Bowl with satyr head BnF 56.31

114

Kenneth Lapatin: Mercurial metrics

Figure 11. Dichterbecher BnF 56.13

Figure 12. Dichterbecher BnF 56.14

of multiple elements: beautifully adorned repoussé shells (crustae) attached to separately cast handles and feet, which were found separated from the vessels as their ancient solder had failed over the centuries and were reattached in Paris in the 1830s. Moreover, as Taurin had apparently avoided touching the objects with his bare hands, superstitiously fearing contact with the buried treasure, but had instead extracted it from the earth with his hoe, modern tears to the objects were added to any ancient damage.

(BnF 56.8-9, Figure 2, top left and right) now weigh 836 and 849 grams, about 2 1/2 Roman pounds each, or 5 pounds as a pair. Of Tutus’ singletons, the gilded modiolus (beaker) depicting scenes related to Corinth and Isthmia (BnF 56.10, Figure 2, center left) now weighs 461 grams, or slightly less than 1 1/2 Roman pounds, but it has suffered losses that are not visible in the photograph. It was likely originally part of a larger set, probably of four, although when the set was broken up we cannot know (see Lapatin 2014: 59-63). Likewise the phiale with Omphale in the omphalos (BnF 56.11, Figure 2, left) - perhaps an ancient pun - may originally have been paired with another, similar vessel, probably depicting Hercules (see G. Gautier in Lapatin 2014, 6367). It now weighs 893 grams, about 2¾ Roman pounds. The simpulum, much less elaborate in its technique and depicting Mercury (BnF 56.12, Figure 2, bottom center right), was likely made locally. It weighs 174 grams, or about half a Roman pound.

The so-called ‘Centaur scyphi’ (BnF 56.6-7, Figure 2, bottom left and right), whose inscribed inner liners were already mentioned, each bear dedicatory inscriptions on their undersides, but nothing explicitly signaling their total weight. Babelon recorded these at 1658 and 1637 grams. Following their recent conservation in Los Angeles they weigh 1654 and 1642 -- the reduction of the first likely owing to the removal of encrustations during cleaning treatments of the earlier twentieth century, and the increased weight of the second probably to the addition of modern restoration materials over the same period (including, for example, thick red wax used to re-adhere the detached shoulder of the young male centaur.) These weights, hardly coincidentally, are all quite close to 5 Roman pounds (1630 gr at 326; 1637.55 at 327.5).

Other vessels, however, do not resolve so easily. Several phiale with inserted emblemata are likely heavily ‘contaminated’ by modern solder and other conservation materials, for the repoussé emblemata and bowls were found separated and reattached in the 19th century. It is possible – and in some cases likely – that they were not properly matched (see Sánchez and Lansing Maish 2014: 120-123).

Other vessels dedicated by Tutus (and subsequently restored) also have weights that suggest that they originally constituted round numbers. The two pitchers bearing scenes from the end of the Trojan War (BnF 56.4-5, Figure 2, center right) now weigh 1045 and 1158 grams (the first being rather more damaged the second). The heavier pitcher is about 3 1/2 Roman pounds. The more damaged one may well have originally been so too. Meanwhile the pair of canthari with high relief masks and other Bacchic paraphernalia

Two of the most intriguing vessels from Berthouville, the so-called ‘Dichterbecher’ (BnF 56.13-14, Figure 1, center foreground, and Figures 11-12), a pair of canthari depicting paired male and female literary figures, fashioned in low relief and badly worn, now weigh 581 and 577 grams, or about 1¾ Roman pounds each. But given their damaged condition, they originally 115

Wonders Lost and Found weighed more. Once thought to be Hellenistic originals on account of their complex iconography, they are now considered Roman imitations, emulations, or even forgeries.9 It is possible that their original weights – and those of other objects in the treasure that do not conform so neatly to what we think we know about the Roman pound – might have been produced according to some other standard.10

Babelon, E. 1916. Le Trésor d’argenterie de Berthouville prés Bernay. Paris. Böckh, A. 1838. Metrologische Untersuchungen über Gewichte, Münzfüße und Maße des Altertums. Berlin. Butcher, K. and M. Ponting 2014. The Metallurgy of Roman Silver Coinage: From the Reform of Nero to the Reform of Trajan. Cambridge. Deniaux, E. 2006. Les dédicants du trésor du sanctuaire de Berthouville cité des Lexovii, in M. Dondin-Payre and M.-Th. Raepsaet-Charlier (eds) Sanctuaires, practiques cultuelles et territoires civiques dans l’Occident romain: 271-295. Brussels. Franks, H.M. 2012. Hunters, Heroes, Kings: The Frieze of Tomb II at Vergina. Princeton. Grierson, P. 1964. Weight and Coinage. Numismatic Chronicle 7.4: iii-xvii. Hultsch, F. 1971 Griechische und Römische Metrologie (Berlin 1882) reprint. Graz. Lajoye, P. 2008. Analyse sociale des donateurs du trésor de Berthouville (Eure), in Ralph Häussler (ed.) Romanisation et épigraphie. Études interdisciplinaires sur l’acculturation et l’identité dans l’Empire romain (Archéologie et Histoire Romaine 17): 127-32. Montagnac. Lapatin, K. (ed.) 2014. The Berthouville Silver Treasure and Roman Luxury. Los Angeles. Lapatin, K. 2015a. Luxus: The Sumptuous Arts of Greece and Rome. Los Angeles. Lapatin, K. 2015b. Some reflections on the Berthouville ‘Dichterbecher’, in P. Linant de Bellefonds, É. Prioux, and A. Rouveret (eds) D’Alexandre à Auguste: Dynamiques de la création dans les arts visuels  et  la poésie: 69-80. Rennes. Martin, M. 1988. Zum Gewicht des römischen Pfundes, in François Baratte (ed.) Argenterie romaine et byzantine: Actes de la Table Ronde Paris 11-13 octobre 1983: 211-25. Paris. Sánchez, E. and S. Lansing Maish 2014. The Hidden Lives of Ancient Objects: Conserving the Berthouville Treasure and Four Missoria, in Lapatin 2014: 107126. Vickers, M. 1990. Golden Greece: Relative Values, Minae and Temple Inventories. American Journal of Archaeology 94: 613-625. Vickers, M. 1992. The Metrology of Gold and Silver Plate in Classical Greece. The Economics of Cult in the Ancient Greek World, Boreas 21: 53-72. Vickers, M. 1995. Metrological Reflections: Attic, Hellenistic, Parthian and Sasanian Gold and Silver Plate. Studia Iranica 24: 163-185. Vickers, M. and D.W.J. Gill 1994. Artful Crafts: Ancient Greek Silverware and Pottery. Oxford.

Acknowledgements I am exceedingly grateful to Mathilde Avisseau-Broustet and Cécile Colonna, curators at the Cabinet des médailles for their most collegial collaboration throughout a multiyear conservation/exhibition project. Thanks are also due to past and current directors of the Cabinet, Michel Amandry and Frédérique Duyrat, and Bruno Racine, president of the Bibliothèque nationale de France (BnF). Numerous Getty Museum staff deserve thanks as well: the laborious conservation of the entire treasure was undertaken by Eduardo Sánchez and Susan Lansing Maish of the Antiquities Conservation Department and is described in their essay, Sanchez and Lansing Maish 2014. Photographs of the Treasure were expertly shot by Tahnee Cracchiola; X-rays by Jeffrey Maish. All have my genuine thanks. Following its exhibition at the Getty Villa from November 2014 to August 2015, the Treasure was displayed at the Fine Arts Museums of San Francisco, the Nelson-Atkins Art Museum in Kansas City, the Museum of Fine Arts, Houston, the Musée departmental Arles antiques, Arles, the Ny Carlsberg Glyptotek, Copenhagen, and the Institute for the Study of the Classical World, New York. An expanded French translation/edition of the catalogue, with somewhat different contents from Lapatin 2014, was produced after this essay was written: Le Luxe dans l’antiquité. Trésors de la bibliothéque nationale de France, ed. Mathilde Aviseau Broustet and Cécile Colonna (Arles/Gand 2017); and a smaller publication subsequently accompanied the exhibition in Copenhagen: High on Luxury: Lost Treasures from the Roman Empire, ed. Anne Marie Nielsen and Anna Wirenfeldt Minor (Copenhagen 2018). Bibliography Andronicos, M. 1984. Vergina: The Royal Tombs. Athens. Avisseau-Broustet M. et al. 2014 The Berthouville Treasure: A Discovery ‘As Marvelous as It Was Unexpected’, in Lapatin 2014: 7-68. On these paired vessels in general see, Lapatin 2015b.. I am grateful to Michael Vickers, who, with customary generosity, shared with me an earlier, unpublished version of his paper cited in note 1 above, in which he suggested that the Persian standard may have been (inadvertently?) adopted by the Romans, as well as the Greeks, through the traditional practice of craftsmen, for Roman authorities, being antipathetic towards eastern luxury, do not seem to have attempted to impose local weight standards that might have appeared to condone luxury. 9 

10 

116

The Erechtheion glass gems: classical innovation or Roman addition? Despina Ignatiadou To Professor Michael Vickers, who encouraged us to see beyond the evident was formulated already in the 19th century,1 and was revisited in recent times.2

The use of coloured gems on the Ionic capitals of the Erechtheion North Portico has been attested by travellers, mainly architects, who visited the Acropolis between the 17th and the 19th century. Their descriptions and watercolours (Figure 1) supply a glimpse of that polychrome decoration which was limited to the double guilloche on the echinus of the capitals (Figure 2; Stevens, Paton 1927: 85-86, fig. 52).

However, there are two issues which must be addressed. Firstly, the material the gems were made of, and secondly and, most important, the date at which they were placed on the monument. The descriptions3 and the gems

The gems themselves do not survive; nevertheless, their former existence is not disputed and furthermore it is thought to be a strong indication of the use of glass in architecture from the 5th century BC. The theory

In 1687 (the earliest testimony), Alessandro Locatelli mentioned semiprecious stones: …con dieci colonne d’ordine ionico nei capitelli, nelle quali vi erano incastrate

Figure 1. Drawing of the Erechtheion guilloche and gems by von Hallerstein 1811 (after Stern 1985, pl. 96.2). See Donaldson’s testimony in 1820. By the eminent glass historian E.M. Stern; see Stern 1983; Stern 1985; Stern 1986; Stern 1989; Stern 1999: 37-39, notes 82-84. 3  For the travellers’ testimonies see Stevens, Paton 1927: 550-631 and Stern 1983: 422-426 with full bibliography. 1  2 

117

Wonders Lost and Found

Figure 2. Ionic capital from the North Portico, Erechtheion, late 5th century BC (after Brouskari 1996, fig. 122).

Ionic columns of the Triple Temple in the Acropolis of Athens. They are of four different colours and were arranged in the order shown on the plate. He was the first to speak of the architectural use of glass in Classical Athens: The glass eyes also of the Ionic capitals of the tetrastyle portico in the Acropolis at Athens, prove that various materials were employed by the Athenians in the decoration of the exterior of their marble buildings.

pietre pretiose attrovandosi ancora alcunni frammenti di lapis lazuli, cornioli ed altre simili… In 1781-82, Louis François Sébastian Fauvel: des émaux coloriés dans l’śil de la volute et dans les entrelacs du chapiteau qui répondent ŕ la hauteur de cet śil. In 1784, Lois François Cassas: Les yeux des entrelacs qui en dehors sont creux, sont remplis dans l’intérieur de petites boules de verre bleu et jaune.

In 1820, John Lewis Wolfe made pencil drawings and noted: Cap. of column. In the eyes of plat band glass eyes remain composed of opaque vitreous substances –of bright colours- yellow, blue and purple. In one instance two of these eyes in angle. And also: The capitals are entitled to the same individual praise sat least in their present state for when the temple was perfect the glass and gilt ornaments about it must have been very injurious to the effect –filling up the channels between the volutes, which now give distinction to them …The platted band is very pretty.

In 1801 Giovanni Battista Lusieri (testimony recorded by Crippe and then Clarke): among the most delicate intertexture of the wreaths and foliage, small bronze nails, and bits of antique glass which had been fastened on to heighten the general delicacy and exquisite finishing of the work. In 1811, Carl Haller von Hallerstein: Ornement du dessous de l’ovolo au chapiteau/dans les trous oů restent des morceaux de verres peints, comme on se sert pour la mosaďque. Dans le 1er et 3me rang est la couleur des pierres bleu et noir dans le 2e rang jaune et pourpre.

In 1835-36, C. H. Bracebridge made a drawing. In 1847, Jacques Martin Tétaz was the last to refer to the gems by describing the last two preserved specimens as glass cylinders: La couronne d’ entrelacs au dessus des oves a tous ces centres en creux dans le marbre. Ces creux étaient remplis de petits cylindres d’émail de diverses couleurs, dont deux sont encore en place, fragmentés, l’un bleu, l’autre vert.

In 1819, Henry William Inwood: Front elevation of the angular capitals of the columns to the side portico…and the small circles in the plat ornament are cut very deep, and filled with different coloured stones or glass; a is black, the next adjoining very light blue, then black again alternately the first row; b is yellow, next very dark blue alternately, c is black, then very light blue, as a.

Today the cavities the gems were inserted in are empty. Donaldson brought fragments of four colours to the Royal Institute of British Architects, but those do not seem to survive,4 nor do those that were left behind.5 The gems were curved, circular elements,

In 1820, Thomas Leverton Donaldson saw the monument and painted the gems in a watercolour drawing accompanied by the caption: Triple Temple TLD. Disposition of the coloured glass beads within interlacing ornament on the capitals of the tetrastyle portico. The base was ornamented in the same manner. He also took portions of the glass eyes… from the torus between the volutes of the

Stern inquired, unsuccessfully, for the gems at the Royal Institute of British Architects, the British Museum, and the Institute of Archaeology, London. 5  The author inquired, unsuccessfully, for the gems at the Athens 4 

118

Despina Ignatiadou: The Erechtheion glass gems: classical innovation or Roman addition

Figure 3. Glass gems of the Graeco-Roman period (after Arveiller-Dulong, Nenna 2011, cat. no. 558).

and red as distinct entities were treated by Stern with scepticism, as she considers them to have been two versions of the same colour. She concluded that they both referred to carnelian given that it was impossible to make translucent red glass before the Roman period. Additionally, she reached the conclusion that green was identical with the light blue. The above-mentioned colours in fact appear rarely in the glass working of the 5th century BC, but they do exist (e.g. Lilibaki-Akamati et al. 2011: cat. no 356). The production of black glass is considered technically impossible in pre-Roman antiquity, but available evidence shows that any of the colourants could be used in a higher percentage to produce a very dark - appearing black - glass.

approximately 1 cm across. They were monochrome but were placed in three-colour combinations, possibly different on each capital. From the descriptions and pictorial representations we deduce the following colours: light blue, blue, yellow, purple, light brown, red, green and black. The inlaying of semiprecious stones was technically possible in the 5th century BC, as such stones had been in use long before that period, mainly in jewellery items. The trade of semi-precious stones could easily have met the demand for most of those polychrome combinations: turquoise for light blue, lapis lazuli for blue, amethyst (?) for purple, carnelian for light brown and red, malachite for green, and any very darkcoloured stone for black.

The making of coloured glass was therefore technically feasible even in the end of the 5th century BC, but the same does not apply for the making of glass gems.

Translucent and opaque glass had also been manufactured in most of those colours since the Bronze Age, to make vessels and beads. Especially in the Classical period, translucent blue or purple or green or reddish brown, as well as opaque light blue or turquoise or bright yellow glass was already used in the making and decorating of core-formed vessels and beads. Reservations on the use of red or green glass are justified. The occurrence of light brown (amber)

Glass gems and similar objects In the study of ancient glass the term glass gem applies to a circular plano-convex glass element, intended for insertion into various materials. Glass gems are usually coloured, albeit undecorated, and were made of translucent, transparent, or opaque glass (Figure 3).6 6  Translucent: material allowing light to pass; transparent: material permitting shapes to be seen through it; opaque: material that has been intentionally opacified (by the use of an additive), and not

Ephorate of Antiquities and at the National Archaeological Museum, Athens.

119

Wonders Lost and Found Their key feature is their colour, as they were made in imitation of semi-precious stones. They were manufactured by firing pieces of glass or crushed glass on a flat surface and they obtained their roughly circular shape without further manipulation, resulting naturally from the surface tension of the material.7 As a matter of fact, the use of glass gems in the 5th century BC is not supported by corresponding finds neither in the Classical nor in the Hellenistic period. Classical or Hellenistic finger rings with coloured undecorated glass gems are unknown. The glass gems of the time were always decorated and oval plano-convex with bevelled edge, to facilitate the inlaying process. They were of extremely meticulous workmanship achieved by the use of a mould (Ignatiadou 2013: 198210, 328-330. Ignatiadou 2009: fig. 3). Other jewellery items, e.g. wreaths, were not adorned with coloured inlays but with enamelling until the Early Hellenistic period, when reddish semi-precious stones were employed. The inlays of the late Classical period were of transparent colourless glass placed over gold foil to decorate furniture: cast plaques, eyes and palmette leaves (Ignatiadou 2001; 2007; 2013: 234-318, 330-333). The use of coloured (green and red) palmette leaves is extremely limited and attested only once, in the 4th century BC (Ignatiadou 2009: fig. 8).

Figure 4. Ceramic flask with glass gems, 2nd century AD (Metropolitan Museum of Art, inv. no 17.194.1891).

the Domus Transitoria are decorated with glass gems inlaid in plaster work (Baccheli et. al. 1998: 86, fig.3). Their use in the decoration of pottery is documented by several vases. Many of the 17 published vases adorned with glass gems were found in Rome where they were possibly manufactured.

The closest parallels to the Erechtheion gems would be the glass gaming counters in two or three translucent colours (blue, natural green, colourless) comprised in gaming sets of the 4th century BC. The earliest such find is a set from Sevaste, Pieria, circa 350 BC, with counters of blue and colourless glass (Ignatiadou 1996: fig. 3. Glass Cosmos 2010: cat. no 94). Those counters resemble the later, Roman, circular glass gems but are not used as inlays in architecture, in jewellery or elsewhere. The peak for the manufacture of glass gaming counters coincides with the late Classical to the early Roman period. To the traditional colours of the Classical gaming counters are then added more colours, mainly shades of yellow, brown, light blue, and some opaque colours.

They are dated to the later years of the Roman Republic or Imperial times (Figure 4) (Comfort 1960, Scott 1992). A key item associated to this group is a two-handled vase in the Bartholdt Collection. It is decorated with eighteen coloured glass gems, not inlayed but attached. They constitute part of the overall decoration of the vase engraved by parallel and intersecting lines (Inv. no 7624. See Lullies 1938: col. 465-466, fig. 42, with earlier bibliography). The dating of the Erechtheion gems Key issues for the discussion of the Erechtheion glass gems is their placement in cavities and the fact that on corresponding capitals that imitate those of the Erechtheion, carved bosses appear in the same positions. The cavities are 8-11 mm in diameter: thus indicating the approximate diameter of the gems. This dimension is extremely small, especially compared to the average diameter of the earliest comparable finds, the glass gaming counters of the 4th century BC, 1520 mm in diameter. The only known glass items of approximately this size, 10 mm across, are some gems of the late Hellenistic/Roman period, which were perhaps intended for the decoration of gold jewellery although this use is also questioned. Those small gems

Towards the end of that peak period, some of those glass elements were for the first time used as decoration in jewellery and as inlays. Thus, the use of coloured glass gems is considered a practice of the Roman times and flourished from the 1st century BC onwards. The use of glass gems in Roman architecture is attested in the Imperial period, in Rome where the arches in material that is too thick or too full of inclusions or concentrated colourants. 7  Lierke 2001. Oval and tear-shaped pieces are also found, occurring accidentally.

120

Despina Ignatiadou: The Erechtheion glass gems: classical innovation or Roman addition

Figure 5. Polychrome capital in the Macedonian Tomb of Euridice, third quarter of the 4th century BC (after Drougou, Saatsoglou-Paliadeli 2006, 185).

are usually 5-6 mm thick. If we consider that on the guilloche their convex part would protrude so as to create the impression of a boss, then the required depth for their fastening would not be more than 2 or 3 mm; it was unnecessary therefore to carve the twice as deep Erechtheion cavities (5-8 mm). Stern observed grains of silica or sand in the cavities and interpreted them as a possible additive to cement, consisting of plaster or asphalt (Stern 1985: 407-408); however, filling the deep cavity with cement would not enhance the fastening of the gems.

only chosen because it is most easily seen from a distance, but also for symbolic reasons, and because on early smaller luxury items, e.g. furniture, the inlays were made of amber. Its reddish colour influenced the painted rendering of similar elements that were thereafter painted red. When colourless glass eyes succeeded the amber ones the former were placed over a red painted background that remained visible under the transparent glass.8 Additionally, even a slight admixture of copper also tarnishes gold to a red colour; therefore a similar effect resulted when gilded bronze decorated (wooden or stone) architectural members.

The size of the gems on the guilloche would be inconveniently small even in the context of jewellery, let alone to be viewed from the distance imposed by the height of the columns. This was not a problem for the scholars of the 19th century who observed the columns from a loft 2.5-3 meters high, thus viewing the capitals from much closer to the eye level. Cockerell made a drawing of the loft (see Stevens, Paton 1927: fig. 223).

Gold attracts the eye even more than red does and was often used for the gilding of inlaid or relief decoration; e.g. the eyes of the volutes on the capitals in the Macedonian Tomb of Eurydice (Figure 5) (Drougou, Saatsoglou-Paliadeli 2006: 185). This is also attested on some exceptional objects decorated with a guilloche, e.g. on a faience pyxis of the 4th century BC, from Thessaloniki, decorated with gold studs (Figure 6).9 A building inscription of the late 5th century BC mentions the gilding of the ophthalmoi on the Erechtheion volutes: two leaves of gold were bought for gilding the two eyes of the column - two drachmas.10

The desired polychromy on Classical Greek architecture was implemented with bright colours. In that context the eyes and bosses were invariably coloured red or gold. A red colour traditionally adorns architectural members. Some examples include: the eyes of the volutes on a capital in Ancient Agora (Thompson and Wycherley 1972: 84 no. A 2972, 166); the eyes of an acroterion from the workshop of Phidias at Olympia (Kaltsas 1997: 63, fig. 76); the eyes of the volutes on the capitals in the Macedonian Tomb ‘of the Palmettes’ (Romiopoulou 1997: 31-32, fig. 26-27). Red was not

On the origin of the colourless glass inlays and the early use of amber see Ignatiadou 2002. 10 9  Ignatiadou, Lambrothanassi 2013: fig. 7a; see also figs 13-14, for the guilloches on a faience kalathos from Neapolis, Thessaloniki which were probably decorated with similar studs. 10  Kolbe 1901: 224-225. Stevens, Paton 1927: 396-397. For similar reasons gold foil is placed beneath the transparent glass eyes on funerary couches in the Macedonian tombs. 8 

121

Wonders Lost and Found inserts, the front bosses of which were gilded. Such impressive gilded elements would be visible even from a great height, but would also have been delicate and prone to destruction by rain or fire. It in the second half of the 1st century BC the Erechtheion was destroyed by fire (Stevens, Paton 1927: 478-479, n. 6), which could have brought about destruction of the original inserts. At that time the monopteros temple of Rome and Augustus was erected on the Acropolis. The new temple had capitals that imitate those of the North Portico of the Erechtheion and there the guilloche also preserves cavities (Figure 7). In the foundations of the Roman temple was actually found part of the original cornice of the Erechtheion, while plaster casts of the Erechtheion capitals travelled to Rome.12 It has been suggested that the imitation was the result of the employment of the same architect for both the Roman temple and for the Erechtheion repairs. The central beam of the North Portico roof, as well as the ceiling coffers on either side of it, were replaced. The entrance door of the North Portico was partly restored: the lintel of the door of was replaced but the original door jambs remained (Brouskari 1996: 162, 176-180).

Figure 6. Faience pyxis from Thessaloniki with gold studded guilloche, third quarter of the 4th century BC.

Several monuments of the 5th century BC feature cavities for inserted decoration: the eyes of the volutes on the capitals of the Acropolis Temple of Nike (Brouskari 1975: pl. 25); the eyes of the volutes on the half column in the house of Dionysos at Pella (Macaronas, Giouri 1989: 71, fig. 76); the eyes of the volutes on the capital from the archaic temple of ‘Therme’ in Thessaloniki (Vokotopoulou 1996: 27); on the altar of Zeus Agoraios in the Ancient Agora, the guilloche on the orthostates (Thompson, Wycherley 1972: 161); in the temple of Nemesis at Rhamnus (on the guilloche probably associated with the base of the Nemesis statue – the cavities are deep and seemingly intended for the insertion of a cylindrical element rather than a boss);11 the Erechtheion North Portico door itself features rosettes with cavities (Brouskari 1996: 180, pl. 125).

It is likely that during those repair works glass gems were placed in the cavities left empty after the destruction of the original inserts. The solution was chosen according to the contemporary fashion of inlaying coloured glass gems, and also by necessity, as it was neither feasible nor desirable to replace entire capitals. The contrast between original and replaced parts is explicitly shown on the entrance door of the North Portico: the relief rosettes decorating the original door jambs have empty central cavities, which were obviously intended to receive decorative inserted bosses. On the replacement lintel similar rosettes were carved but without cavities (Figure 8). The Erechtheion gems remain to date an isolated example of embellishing Classical architecture with coloured glass gems. Rather than its acceptance as a Classical innovation, current research in the history of glass permits its attribution to an inspired Roman architect.

In other guilloches, e.g. on the base of the same Erechtheion columns (Brouskari 1996: pl. 123), there are bosses instead of cavities; we can therefore conclude that some boss-like element was meant to be inserted in the cavities, wherever the latter existed.

Acknowledgements

It is therefore rather probable that the cavities on the Erechtheion guilloche were initially intended for gilded elements. Those could have been bronze, although no remains of corroded bronze seem to survive in the cavities. We cannot exclude the possibility of wooden

The author wishes to thank K. Manteli and E. Leka, for reading the text and suggesting valuable improvements.

11  Despinis 1971: 71. Kallipolitis 1978: 81-82, pl. 25. Pétracos 1981: 216, fig. 12. Stern 1989.

12  For the cornice see Stevens, Paton 1927: 479. For the casts see Kockel 1991.

122

Despina Ignatiadou: The Erechtheion glass gems: classical innovation or Roman addition

Figure 7. Ionic capital from the temple of Rome and Augustus, late 1st century BC.

Figure 8. Door of the North Portico, Erechtheion; the door jambs are late 5th century BC, the lintel is late 1st century BC (after Brouskari 1996, fig. 125).

123

Wonders Lost and Found Bibliography

Lullies, R. 1938. Neuerwerbungen der Antikensammlungen in München. Jahrbuch des Deutschen Archäologischen Instituts / Archäologischer Anzeiger 53: 420-466. Μακαρόνας, Χ., Ε. Γιούρη 1989. Οι οικίες Αρπαγής της Ελένης και Διονύσου της Πέλλας. Athens. Pétracos, B. 1981. La base de la Némésis d’Agoracrite. Bulletin de Correspondance Hellénique 105: 227-253. Romiopoulou, K. 1997. Lefkadia, Ancient Mieza. Athens. Scott, R.T. 1992. A new fragment of ‘Serpent Ware’ from Cosa. Journal of Glass Studies 34: 158-159. Stern, E.M. 1988. Glass and gold in classical Greek architecture 12th International Congress on Classical Archaeology 3 (Athens 1983): 304-308, pl. 64. Stern, E.M. 1985 Die Kapitelle der Nordhalle des Erechtheion. Athenische Mitteilungen 100: 405-426, pls. 91-96. Stern, E.M. 1986. Das Haus des Erechtheus. Boreas 9: 5164, pl. 4. Stern, E.M. 1989 Colored glass inlays in architectural ornament: Athens and Rhamnous. American Journal of Archaeology 93: 254. Stern, E.M. 1999. Ancient glass in Athenian temple treasures. Journal of Glass Studies 41: 19-50. Stevens, G.P. et al., J.M. Paton (ed.) 1927. The Erechtheum, measured, drawn and restored. Cambridge MA. Thompson, H.A. and R. E. Wycherley 1972. The Agora of Athens, the history, shape and uses of an ancient city center (The Athenian Agora 14). Princeton. Vokotopoulou, J. 1966. Guide to the Archaeological Museum of Thessaloniki. Athens.

Arveiller-Dulong, V. and M.-D. Nenna 2011. Les verres antiques 3. Paris. [Brouskari] Μπρούσκαρη, Μ. Σ. 1975. Τα μνημεία της Ακροπόλεως. Athens. [Brouskari] Μπρούσκαρη, Μ. Σ. 1996. Τα μνημεία της Ακρόπολης. Athens. Comfort, H. 1960. Roman ceramic-and-glass vases at Heidelberg and New York. American Journal of Archaeology 64: 273, pl. 75. [Despoinis] Δεσπίνης, Γ.1971. Η Νέμεση του Αγοράκριτου. Athens. Drougou, S. and C. Saatsoglou-Paliadeli 2006. Vergina. The land and its history. Athens. Ignatiadou, D. (sc. ed.) 2010. Glass Cosmos (exhibition catalogue, Archaeological Museum of Thessaloniki). Thessaloniki. [Ignatidou] Ιγνατιάδου, Δ. 1999. Ελληνιστικό επιτραπέζιο παιχνίδι με γυάλινους πεσσούς. Ancient Macedonia 6 (1996): 508-522. Ignatiadou, D. 2001. Glass and gold on Macedonian funerary couches’ Annales de l’Association internationale pour l’histoire du verre 15: 4-7. Ignatiadou, D. 2007. Amber and glass inlays. A chronological sequence’ Ancient Macedonia 7: 473483. Ignatiadou, D. 2007. Le verre incolore, élément du décor polychrome du mobilier funéraire de Macédoine, in S. Descamps-Lequime (ed.) Peinture et couleur dans le monde grec antique: 219-227. Paris. Ignatiadou, D. 2009. A haematinon bowl from Pydna. Annales de l’Association internationale pour l’histoire du verre 18: 69-74. Ignatiadou, D. 2013 Colorless glass for the elite in ancient Macedonia. Thessaloniki. (in Greek, with English summary). Ignatiadou, D. and E. Lambrothanassi 2013. A glass kotyle and a faience pyxis from Thessaloniki. Journal of Glass Studies 55: 35-52. [Kallipolitis] Καλλιπολίτης, Β. Γ. 1978. Η βάση του αγάλματος της Ραμνουσίας Νέμεσης. Archaiologike Ephemeris: 1-90, pls 1-32. Kockel, V. 1991. Überlegungen zu einer Kopie der ionischen Kapitelle der Erechtheion-Nord-Halle. Archäologischer Anzeiger 106: 281-285. Kolbe, W. 1901. Die Bauurkunde des Erechtheion vom Jahre 408/7. Athenische Mitteilungen 26: 223-240. Lierke, R. 2002. With ‘trial and error’ through ancient glass technology. Hyalos-Vitrum-Glass (1st International Conference, Rhodes 2001): 181-186. Athens 2002. Lilibaki-Akamati, M., I. M. Akamatis, An. Chrysostomou, P. Chrysostomou 2011. The Archaeological Museum of Pella. Athens. http://www.latsis-foundation.org/ eng/electronic-library/the-museum-cycle/thearchaeological-museum-of-pella 124

Carp from the Danube delta? Notes on an unusual gold-glass in the Wilshere Collection Susan Walker Introduction

in this respect: while it is clear that AN2007.40 was until 1862 in the Recupero Collection, AN2007.23 has no documented earlier history.3

In 2007, three years before he retired, Michael Vickers registered a significant acquisition for the Department of Antiquities at the Ashmolean Museum: the core items from the Wilshere Collection of gold-glass, sarcophagi and inscriptions bequeathed by Charles Wilshere to Pusey House, Oxford by Deed of Trust in 1895. For much of his career Michael had cared for these objects, then on longterm loan to the Ashmolean, where he displayed them in the Beazley Gallery and the Byzantine corridor, and engaged in scholarly dialogue about them with academic visitors and students. Michael also gave highly informative and entertaining lectures on Charles Wilshere, an exceptionally well-educated Hertfordshire landowner with a passion for the liturgical interests of the Oxford Movement. Based on his lectures, Michael published a paper introducing the collection (Vickers 2011).

Manufacture and chemical composition of the glass fragment AN2007.23 is an irregularly broken fragment of sandwich gold-glass measuring 5.5 by 4.3cm. As it now survives the fragment comprises four distinct pieces of glass. One forms the base layer 3mm thick, onto which the gold leaf was stuck; the other pieces form a double layer of protection 1.7mm thick over the gold leaf (Figures 1-3). The gold leaf was sealed in place by controlled reheating of the base and upper layers. Usually the presence of a double protective layer indicates dissatisfaction of the glassmaker with a first attempt to form the upper part of the vessel (normally a shallow bowl), the separately blown base of which was decorated with the gold leaf (Howells 2013: 114-8; 2015: 49). However, as this fragment is completely flat with no discernible base ring, it was apparently not designed as a vessel. In this instance the double upper layer may have been intended as additional protection for the gold leaf. At an unknown date, but probably in the fourth or fifth century AD, the uppermost protective layer was broken and repaired with a fourth piece of glass 1.65mm thick (see below).

In recent years it has been my task to compile a catalogue of the Wilshere Collection. Archival research in the Vatican Library has revealed much about Wilshere’s motives and collecting habits, while analytical research in Oxford and Shrivenham has produced new information on the materials of the collection. To mark Michael’s substantial contribution to the Ashmolean and his interest in the Wilshere Collection, I offer a short discussion of a particularly puzzling fragment of goldglass, accessioned at the Ashmolean as AN2007.23.1 The fragment is included in the recently published catalogue of the collection, which also features introductory essays and catalogue entries by Sean Leatherbury and David Rini, and appendices describing the analysis of the glass by Andrew Shortland, Julian Henderson and Patrick Degryse (Walker 2017: 149-150, no.16).

This fragment was purchased by Wilshere, probably in Rome, at an unknown date, but most likely during the documented period of formation of his collection between 1865 and approximately 1885. The piece is not mentioned in his correspondence with De Rossi.2 Morey’s confusion of this fragment with AN2007.40 in the footnotes to his entry and the plates is misleading

Each of these glass layers, including the repair, has a distinct chemical signature indicating the agent used to decolour the raw glass. Roman glass is chemically remarkably consistent, suggesting a major production of raw glass, principally located around an arc of coast in the south-eastern Mediterranean stretching from Syria to Egypt (Freestone et al. 2000, Gorin-Rosen 2000, and Nenna et al. 2000). Much vessel and window glass was recycled after breakage, and there was a significant trade in cullet. Roman glassmakers supplying the luxurious end of the market used decolourants to clarify the glass by removing the elements of iron in the sand that during the primary heating process turned ‘natural’ Roman glass a distinctive shade of blue-green.

= Morey 1959: 63, no. 376. In the footnotes to his catalogue entry and Plate 32 AN2007.23 is conflated with AN2007.40. 2  See note 1 above and, on the formation of Wilshere’s collection, Walker 2017: 19-69.

3  On Recupero, see Olivier-Poli 1825, Garrucci 1864 and Walker 2017: 51-54. Garrucci 1864: 8 and pl. 12 clearly refers to AN2007.40, which is also listed by Vopel 1899: 111 no .441.

Collection history

1 

125

Wonders Lost and Found

Figure 1. AN2007.23. The gold-leaf decoration seen from below, with the letters reversed. Ashmolean Museum, University of Oxford.

Figure 2. AN2007.23. The weathered upper surface of the fragment, with repair (right). Ashmolean Museum, University of Oxford.

fragments come from the graves of Christians. Some decoration of the gold leaf concerns stories of personal salvation from the Old and New Testaments, illustrating the commendatio animae, an ancient prayer for the commendation of the soul of the deceased. As the Church became institutionally more powerful, saints and martyrs, including popes of particular relevance to Rome, came to dominate the gold-leaf repertoire (Leatherbury in Walker 2017: 113-125; Walker et al., 2018: 380 for the rise of Saints Peter and Paul). The chemical compositions of the separate elements of our fragment, though inconsistent with each other, nonetheless fall within the spectrum of gold-glass vessels decoloured with manganese alone (Figure 4).4 Figure 3. Backlit view of the fragment from above, showing triple layers of glass and the gold-leaf net appearing lower left. Dana Norris, Department of Conservation, Ashmolean Museum, University of Oxford.

However, the inner layer contains a detectable amount of antimony, suggesting that the raw glass used for the topmost covering layer contained at least some recycled cullet. While the inner and outer layers lie close to each other in manganese content, the repair contains significantly more manganese. Nonetheless all three readings fall within the overall results for the Wilshere Collection, indicating that the fragment is chemically consistent and most likely contemporary with the vessels from the catacombs.

The decolouring process was undertaken as part of the primary production of raw glass, before the blocks or chunks of raw glass were exported to Europe (Jackson 2005). Here they were blown or cast into finished products and, in the case of gold-glass, decorated with gold leaf sandwiched between layers of glass. Though gold-glass is known from sites in Italy and from the Rhine-Danube frontier (especially Cologne), only Rome was a significant producer of gold-glass, notably in the fourth century AD, when Christianity was established (Howells 2015: 53-58). Although pagan and Jewish gold-glass was produced, most of the surviving vessel

By comparison with similarly decoloured Roman glass from archaeologically stratified contexts, gold- glass products decoloured with manganese may be dated 4  Shortland in Walker 2017: 224-226, Appendix 4a. The lower of the two covering layers was not accessible for portable XRF analysis. See Meek in Howells 2015: 30-40 for comparable examples in the collections of the British Museum.

126

Susan Walker: Carp from the Danube delta? Notes on an unusual gold-glass in the Wilshere Collection

Figure 4. Manganese (horizontal axis) and antimony XRF readings for AN2007.23, shown as blue lozenges within red circles. Andrew Shortland, Cranfield University.

Beneath the fish are three letters, most likely PEV (PEU) if read in Latin, or RHEU if read in Greek. However, the reading of the last letter is uncertain: its upper part has the curved inner profile suggestive of a late antique V, though it also has a short lower tail suggestive of a Greek upsilon, Y; there is insufficient space between the outer lines of the letter and below the line of the first two letters to read X, though the lower edges of all three letters coincide with the lower edge of the fragment, making it impossible to be certain of the reading.

within the middle decades of the fourth century AD.5 This, the largest group within the Wilshere Collection, contains the widest range of products. Particularly close to the chemical composition of the repair made to the fragment is AN2007.31, like AN2007.23 a sherd from a vessel outside the normal range of shallow bowls from the catacombs. The pictorial decoration in gold-leaf The fragment is decorated in gold leaf with added coloured enamel, now brownish but once perhaps red. Depicted is a scene of fishing, more easily seen from the outside of the object, where the image and text appear in reverse (Figure 1). The directions given in the description below thus require reversal to recover the original orientation of the scene and text.

To the right of the letters appears a fan-shaped object, centrally divided with diagonal lines running from the central spine. The shape of the object strongly suggests a bag-shaped net, its poles meeting at the closed end to allow the fisherman (unseen on this fragment) to operate it like a pair of tongs. Another object lies between the head of the upper fish, shown swimming to the left, and the tail of the lower. The object is pointed at the right end, with a hook at the left; the pointed element is surely a barb. Both net and barbed hook - the line to which the latter was attached is not visible - are directed at catching the upper fish, while the lower fish swims away from the hook and the net, the prongs of the net’s supporting poles passing over its body. Both fish are large and plump, their fleshy bodies covered with scales. Though lacking dorsal fins, in body, head shape and scale formation they resemble carp (Figure 5).

Two fish dominate the scene, but the principal subject of the narrative is the process of catching the upper of the two. The fish swim below a double curved line, perhaps simply marking the edge of the scene but, given that it only appears above them, more likely representing the curve of a landscape feature, a bay, harbour, pool or tank.

For comparable results from stratified finds, see Jackson 2005 and Foster and Jackson 2010.

5 

127

Wonders Lost and Found

Figure 5. Cyprinus carpio (carp) recorded in 1879, compared with the fish depicted on AN2007.23. Wikimedia Commons and Ashmolean Museum, University of Oxford.

To the right of the letters is a curved blob of gold leaf, with a distinct central fold, perhaps representing a rocky surface.

personification of the Nile riding an elephant in a ritual procession, accompanied by cupids and musicians. The cupids sail and fish in a second panel paired with the first. A third panel set below the first shows a woman in purple robes with a gilded raised stephane or diadem; of regal appearance and seated on the shore, she instructs a servant bearing white cloths in the water. In the foreground, to the right of the panel, beyond various scenes of fishing with large, weighted nets from boats, an agile young fisherman catches a large fish with bagshaped net and line from a rocky islet; he and his seated elderly companion are shown at larger scale than most of the other figures (Figure 6).

Evidently only part of the scene is preserved in the surviving fragment. No double line appears below the break, and the possible representation of rock in the lower right corner, beside the last letter of the inscription, suggests that the landscape was originally more extensive. If a fisherman had been included, the figure would need to be placed to the lower right of the surviving fragment. Reconstructing the context of the scene

The theme of the entire panel is perhaps a fishing contest taking place in the Nile delta, in which the figures shown at larger scale are the principal protagonists. In his Life of Antony, 29, 3, Plutarch recounts a scene of competitive fishing between Antony and Cleopatra: the queen instructs her servants to plant a salted Pontic fish for the hapless Antony to catch to guffaws of laughter from an invited audience. ‘Leave the fishing-rod, General,’ said the wily queen, ‘to us poor sovereigns of Pharos and Canopus; your game is cities, provinces and kingdoms.’

Both the barbed hook and the distinctive bag-shaped net shown on the gold-glass are, today as in antiquity, principally used for catching single large fish from the shore, not for catching large numbers of fish from a boat.6 They may both be operated at the same time by one fisherman, if exceptionally agile and strong. An impressive demonstration of such agility, strength and skill appears on the third of four panels of a mosaic pavement found in the bath suite of the Villa of the Nile.7 Located on the southern Mediterranean shore to the north of the ancient harbour of Lepcis Magna, Tripolitania, the villa was named by its Italian excavators after the principal subject of the mosaic pavement, the management of the Nile and subsequent enjoyment of its riches by the Ptolemaic rulers of Egypt. Such wealth (Greek, tryphe) could only be achieved through appropriate worship of the river’s tutelary deities, here represented in the first (upper left) panel by a

Comparable scenes on late Roman gold-glass Paired fish swimming in opposed directions feature on late Roman gold-glass, and have often been understood to express a Christian affiliation (Dölger 1922-43). Within the Wilshere Collection the base of a chalice (AN2007.39) contains a brief text set within a fishshaped frame between the two fish, which circle around it, the scene encircled by a cabled border.

Gallant 1985: 13-14 describes types of net and trap, after Oppian 3, 73-90 and Aelian 12,43, made of branches from small bushes such as Spanish broom or cypress, and specialised varieties of hook and barb, all used in low-productivity fishing of solitary, in-shore species. See further Becker-Nielsen 2005, 89. 7  Aurigemma 1960: 49 with pl. 97. For the discovery: Guidi 1930. 6 

This text may be read from a mid-20th century photograph (Figure 7) in the archive of the department of Antiquities, Ashmolean Museum as SPE (Latin: ‘in 128

Susan Walker: Carp from the Danube delta? Notes on an unusual gold-glass in the Wilshere Collection

Figure 6. Detail of the fishing contest, right end of the third mosaic panel from the bath suite of the Villa of the Nile (Tripoli Museum). Photo: Philip Kenrick.

Figure 7. AN2007.39: two fish swim around a text. Department of Antiquities archive, Ashmolean Museum, University of Oxford.

129

Wonders Lost and Found may not be identified, though the stance of their feet suggests that they might have been at prayer. They are evidently not actively fishing; indeed the act of fishing itself does not appear to feature in overtly Christian scenes of fourth-century date, despite the biblical references to Christ inviting his disciples to progress from a humdrum existence as Galilean fishermen to become fishers of men; the long established symbolic use of fish to express Christian affiliation, and plenty of textual and written evidence for the consumption of fish in miraculous feasts and not least at Roman funerary banquets (Dölger 1922-43). Surviving scenes of fishing in third and fourth-century Roman art appear to have more to do with lifestyle than religious affiliation: a third-century marble sarcophagus from Brignoles (Var, south-east France) shows a fisherman catching a dolphin in the presence of a bust of Sol: the figure is juxtaposed with that of a ‘Good Shepherd’ carrying a ram in a mountainous landscape (Christern-Briesenick 2003: 103 no.197, pl. 52, 1-5). Between the two rustics stand a well-dressed woman at prayer and a man seated to read. This is one of a number of sarcophagi juxtaposing the spiritual and intellectual with the bucolic life, all pursuits indicative of personal fulfilment (Zanker 1995: 284-5 with fig.154). The context is not unequivocally Christian despite the presence of the praying woman and the ‘Good Shepherd’. The act of fishing is shown in some fourth-century engraved glass made in Rome, in which fishermen appear as isolated figures, the vignettes often defined by decorated borders. However, there is no religious content in the decoration, which seems to have been inspired, like the Tripolitanian mosaic, by Nilotic scenes. None of the engraved glass bowls shows the type of bag-net and barbed hook depicted on the fragment of gold-glass (Paolucci 2002: 18-19, figs. 9-10). Similarly, fishermen appearing in the mid-fourth century glass opus sectile panels from Kenchreai were set within harbour scenes juxtaposed with a Nilotic, marshland landscape. Like the fisherman in the mosaic from Lepcis Magna, these figures wear round hats and are shown at large scale. They fish with rod and line; one carries a net of similar design to that of AN2007.23, but without any trace of the tong-like frame.8 Thus, our fragment is to my knowledge unique within the late antique gold-glass repertoire in showing the process of fishing and is most closely comparable to Nilotic scenes in other media.

Figure 8. AN2007.40. Fragment of the base of a large, oval dish: four fish swim in a pool below the feet of people probably shown at prayer. Ashmolean Museum, University of Oxford.

hope’), but only the E is now clear (Walker 2017: 175, no.36). These fish too could be carp. This glass was too constrained, apparently by a modern repair, to be chemically analysed; from the reconstructed vessel form it is likely to be early in the series, in which case it was probably decoloured with antimony alone. A single fish appears on a small fragment of greenish glass in the Vatican (Morey 1959: 37, no.198). Though in Morey’s corpus the photograph was reversed and illustrated upside-down, the form of the lettering SP flanking a cross set within a circle (possibly a bread bun was intended) and that of the fish with dorsal fin do not match fourth-century representations of fish, and Morey’s doubts concerning the antiquity of this piece may be sustained. Another glass in the Wilshere Collection (AN2007.40) shows four fish swimming in a turquoise pool (Figure 8). Unfortunately, the lower edge of this fragment has been chipped away, so no base ring has survived; nonetheless the concave upper surface suggests that it comes from a shallow bowl or dish. If the band of gold leaf below the fish represents the edge of a central medallion of such a vessel, the vessel would have been very large, possibly an oval serving plate for fish. Like AN2007.23, this fragment belongs to the group of gold-glass decoloured with manganese. The most prominent fish may be identified as a grouper and a carp, with two smaller fish below. People stand above the fish at the edge of the pool; only their feet are preserved, with no text, so they

The gold-leaf text As with the imagery, the surviving text does not match that on any gold-glass vessel. These were mostly written in formulae, comprising Greek injunctions 8  Ibrahim, Scranton and Brill 1976: see drawing of panel no.22, 1.2.B for the comparable net.

130

Susan Walker: Carp from the Danube delta? Notes on an unusual gold-glass in the Wilshere Collection

written in Latin script, urging the deceased to drink and to live [‘for ever’ is usually understood]: PIE! ZESES! Often the text was extended to urge the deceased to live in peace, sometimes specified as with his or her spouse and family.9 Such extended text was expressed in Latin. In many cases personal names are included, but no hitherto recorded injunction, extended good wish or personal name matches the letters on this fragment, nor do the letters coincide with any of the Greek names of fish listed by Oppian, or their Latin counterparts, unless we suppose that the last letter was an error for R, and the fish are to be identified as perch (percae), which they do not visually resemble.10 From the letter shapes and serifs, and the likely orientation of the glass, I suggest the text is to be read in Latin as PEV (PEU). The text is unlikely to refer to the eastern Nile delta city of Pelusium, by the fourth century AD already remote from the sea, its channel silted. Pelusium was better known for its military significance and for the cultivation of flax in the surrounding fields than for fishing and Nilotic festivals (Smith 1872: 372-3 s.v. Pelusium). Moreover the text would need to be read as PE[L]V, the absence of the L interpreted as an error by the craftsman. However it must be admitted that it is not unusual for texts on gold-glasses to include errors, for example AN2007.26, where DIGNITAS is rendered DIGNTIAS.

and seduction of both women and young boys, for fish was a symbol of gourmandism, which was itself equated with an uncontrolled appetite for sex. Fish was used to point up the division between gods and mortals, and to characterise the extreme tastes of oriental despots (Davidson 1993: 60-64). By contrast, the principal role of fishermen in Hellenistic and Roman art was to illustrate the lives of the poor: the fishermen apparently enjoying royal patronage in the mosaic of the Villa of the Nile are an exception, though the Nile Delta often provides the setting for images of poor fishermen.11 Fishermen were sometimes paired with shepherds, their rustic pursuits set within a bucolic landscape (see above). The personalised lids of sarcophagi made in Rome in the early fourth century often show funerary feasts at which a large fish is served on a platter;12 this was also a subject of catacomb painting, but such vignettes became rare as, alongside clerical disapproval of such feasts, the cult of Christian martyrs gained strength in the later fourth century with papal patronage.13 Carp in Greece and Rome Though carp bones were found in excavations at prehistoric Sitagroi and classical burial mounds in eastern Greek Macedonia, carp was unfamiliar in classical Athens, where saltwater fish was preferred. However, by the close of the first century AD the satirical poet Martial noted that the northern bank of the Danube was ‘iam nostra ripa’ (now our bank), and the river’s waters had been enslaved following a treaty made by the emperor Domitian (Martial 5.3, 6, 10). The most highly regarded carp were to be found in the Danube delta, where rights of ownership and management of the waterway and its banks were strongly contested in Roman times. Historic, official gifts of rights of access and exploitation to the local population were long remembered and cited in later disputes.14 The Romans bred carp in the Danube for transport live in tanks around the Mediterranean, a practice which continued into the early sixth century AD, as reported by Cassiodorus (Variae, 12, 4), noting its appearance at the table of Theodoric, Ostrogothic King of Rome, who was born in Pannonia and raised in Constantinople, both within reach of the carp’s natural habitat (Hoffmann 2006: 161-7).

Morey’s contention that the lettering is possibly Greek, of mixed upper and lower case, and of questionable antiquity is undermined by his mis-orientation of the fragment. Indeed, the forms of the letters with serifs are entirely consistent with mid-later fourth-century Latin script on gold-glass vessels (Lega 2012) and thus with the date suggested by the chemical composition of the fragment. The question remains as to how the text is to be completed. If the fish have been correctly identified as carp, and the location of the scene as a delta-like landscape, then the Peuce channel of the Danube delta might provide a suitable solution, allowing the text to be restored as PEV[CE]. Fish in Greek and Roman culture Lavish and louche funerary feasts, usually featuring a large fish as the main dish, were strongly criticised by fourth-century Christian bishops (Rébillard 2009: 144 with n. 18). The association of fish with conspicuous consumption already had a long history (Davidson 1993: 53-5). Indeed, in classical and later Greek literature fish was often associated with extravagance and class division. Fish is often served in tales of power

Bayer 1983: 180 (pathetic aspect of fishermen), 184 (links to Nilotic iconography). See also note 21. 12  Within the Wilshere Collection, AN2007.45 and 46, Walker 2017: 81, Fig. 36 and 178-179, nos. 38-39; see Koch 2000: 25 for the end of the portrayal of funerary feasts on sarcophagi. 13  For painted scenes of banquets within the catacombs, see Fiocchi Nicolai, Bisconti and Mazzoleni 2008: 110-113. Walker 2017: 80-81 offers evidence of the move to representations of saints and martyrs. 14  Unpublished seminar by J.J. Wilkes, Oxford University Epigraphic Workshop, 29/10/2012; see CIL III no.781 (Berlin 1876); Latyschev 1885: v.1, no.3; Latyschev 1916, v.1, no.4; Oliver 1965; Inscripţile din Scythia Minor no.67 (Text A), no. 68 (Text B), no. 69 (unattributed fragment). 11 

9  Some scholars have concluded that such gold-glass vessels were commissioned as wedding presents, but marriage also forms a significant element within Roman funerary iconography: Leatherbury in Walker 2017: 115-116. 10  Oppian, Halieutica, Loeb edition index pp. 520-522. For a late thirdcentury AD representation of a perch in glass mosaic, see Oliver 2001.

131

Wonders Lost and Found The original function of the fragment

of use, the object of which the fragment formed a part was broken, and this fragment was kept for reuse in a Christian context, to show the two fish as a symbol of Christian allegiance. Any fisherman once present was discarded, the surviving tackle becoming peripheral and insignificant. The repair made to the original upper surface of the glass may be associated with such secondary use of the broken fragment. Similar reuse of a larger, painted glass panel is known from the Catacomb of Ponzianus on the Via Salaria Nova, where a segment of a scene of birds plucking fruit had been cut to a perfect roundel 35.3cm in diameter.19

As was noted by Marlia Mango in her unpublished catalogue of the Wilshere Collection begun in the 1980s, the fragment most likely does not come from the normal range of gold-glass vessels - shallow bowls and chalices - as it is completely flat, with no trace of any base ring surviving on the lower layer. On account of the weathering visible on the outer of the two sheets making up the upper layer (which she believed to be the outer, or lower layer), Mango suggested that the fragment belonged to a window. Window glass (specularis) is listed as a product in its own right, valued at half the sum per pound of ‘natural’ vessel glass, in the Edict on Maximum Prices issued on behalf of the Emperor Diocletian in AD 301, and studies of workshops suggest that it was made separately of locally obtainable glass that was not decoloured, as all the elements of our fragment were.15 However, the windows of prestigious buildings, including early Christian basilicas, might include vignettes of coloured glass.16 In a contemporary pagan context, figured panels of coloured mosaic glass (opus sectile), were discovered packed in crates in the courtyard of a harbour-side building in Kenchreai, the port of ancient Corinth; apparently intended as a decorative wall-frieze, they are thought to have been shipped from Egypt.17 Smaller panels of glass opus sectile commissioned as wall decoration are known from Corinth and Rimini (Oliver 2001 [Corinth]; Ortalli 2000 [Rimini]). The surviving elements of narrative indicate that the Wilshere fragment originally formed part of a larger scene, its iconography unmatched in the context of gold-glass vessels. If it did not form part of a window or architectural frieze, it could have served as inlay, perhaps of a pond in which carp were kept. It may not be entirely excluded that the gold-leaf decoration is part of a pictorial map: a fragment of flat transparent glass, apparently without a protective layer, made in Cologne and now in Bonn, was decorated in gold-leaf with a pictorial representation of a city and its harbour.18

Conclusion This unusual fragment of gold-glass illustrates the iconographical and functional range of the medium, particularly notable within the group of gold-glasses decoloured with manganese produced at Rome in the middle decades of the fourth century AD. The fragment also demonstrates that gold-glass could be repurposed functionally and iconographically from a secular story of conspicuous consumption to a Christian funerary context. Such versatility apparently occurred in a short space of time, in a climate of clerical criticism of excessive expenditure on private funerals and lack of focus by wealthy Christians on saints, martyrs and symbols of piety and allegiance. The fragment may also illustrate the extraordinary gastronomic range of the later Roman empire, in which carp originally bred in the Danube Delta were especially prized at the dinner tables of Rome. Bibliography Agnello, G. 1957. Il Museo Biscari di Catania nella storia della cultura illuministica italiana del settecento. Archivio storico per la Sicilia orientale 10: 142-159. Aurigemma, S. 1960. L’Italia in Africa. Le scoperte archeologiche. Tripolitania I. I monumenti d’arte decorative. Parte 1.1 Mosaici. Rome. Bayer, E. 1983. Fischerbilder in der hellenistischen Plastik. Berlin. Becker-Nielsen, T. 2005. The technology and productivity of ancient sea fishing, in T. BeckerNielsen (ed.) Ancient Fishing and Fish Processing in the Black Sea Region: 83-95. Aarhus. Brill, R.H. 1996. The Kenchreai Panels revisited. Corning Museum of Glass Newsletter (summer): 1-2. Corning, N.Y. Christern-Briesenick, B. 2003. Repertorium der Christlichantiken Sarkophage III. Frankreich, Algerien, Tunisien. Mainz. Davidson, J. 1993. Fish, Sex and Revolution in Athens. Classical Quarterly 43: 53-66.

Two contexts for the gold-glass This fragment of gold-glass thus originally formed part of a secular commission, most likely marking the appreciation of tasty fish from a distant, designated origin and/or the activity of fishing rather than any sense of Christian piety. Indeed it does seem likely that any Christian, funerary context was secondary to the original purpose of the glass. In the process of change Stern 1999: 462, Table 1 for the Price Edict; 464-6 for a useful discussion of the manufacture of ancient window-glass. 16  Dellacqua 2003: 101-2 discusses Prudentius’ description of the windows of the Theodosian church of San Paolo fuori le mura at Rome. 17  See n. 21 above and Zanker 1995: 325-6, fig.177 (Homer); Brill 1996; Oliver 2001. 18  Morey 1959: 69, no. 429, with pl. 35 = Landesmuseum Bonn no. LXVIII; Fremersdorf 1967: 200-1, pl. 281. 15 

19  Vattuone 2010: 7 with fig.1 reproducing a watercolour of the panel, which had been left in situ, by Gregorio Mariani.

132

Susan Walker: Carp from the Danube delta? Notes on an unusual gold-glass in the Wilshere Collection

Dellacqua, F. 2003. Illuminando colorat’. La vetrata tra l’età tardo imperial e l’alto medoevo: le fonti, l’archeologia. Spoleto. Dölger, F.J. 1922-43. Ichthus: Das Fisch-Symbol in frühchristlicher Zeit: Ichthus als Kürzung der Namen Jesu: Iesous Christos Theou Uios Soter. Münster in Westfalen. Entwistle, C. and L. James (eds) 2013. New light on old glass; recent research on Byzantine mosaics and glass. London. Fiocchi Nicolai, V., F. Bisconti and D. Mazzoleni 2008. The Christian Catacombs of Rome. History, Decoration, Inscriptions (3rd edition in English translated by Cristina Carla Stella and Lori-Ann Touchette). Regensburg. Foster, H.E. and C. Jackson 2010. The composition of late Romano-British colourless vessel glass: glass production and consumption. Journal of Archaeological Science 37: 3068-80. Freestone, I., Y. Gorin-Rosen and M.J. Hughes 2000. Primary glass from Israel and the production of glass in late antiquity, in Nenna: 65-84. Fremersdorf, F. 1967. Die römischen Gläser mit Schliff, Bemalung und Goldauflagen aus Köln (Die Denkmäler des römischen Köln 8). Cologne. Gallant, T. 1985. A Fisherman’s Tale. Ghent. Garrucci, R. 1864. Descrizione dei vetri ornati di figure in oro appartenenti al sig. Tommaso Capobianchi, negoziante d’antichità in Via del Babuino n. 152. Rome. Gorin-Rosen, Y. 2000. The ancient glass industry in Israel: summary of the finds and new discoveries, in Nenna: 49-64. Guidi, G. 1933. La Villa del Nilo. Africa Italiana 5:1-56. Hoffmann, R.C. 2006. Der Karpfen (Cyprinus carpio Lat.). Der lange Weg eines ‘Fremdlings’ in die Schweiz, in Plogmann: 161-167. Howells, D.T. 2013. Making late antique gold-glass, in Entwistle and James: 112-20. Howells, D.T. 2015. A Catalogue of the Late Antique Gold Glass in the British Museum (British Museum Research Publication 198). London. Ibrahim, L., R. Scranton and R. Brill 1976. Kenchreai, Eastern Port of Corinth 2. The Panels of Opus Sectile in Glass. Leiden. Jackson, C. 2005. Making colourless glass in the Roman period. Archaeometry 47.4: 763-80. Koch, G. 2000. Frühchristliche Sarkophage. Munich. Latyschev, V. (ed.) 1916. Inscriptiones antiquae orae septentrionalis Pontis Euxini Graecae et Latinae (St Petersburg 1885) 2nd edition of Volume 1. Lega, C. 2012. Il corredo epigrafico dei vetri dorati: novità e considerazioni. Sylloge Epigraphica Barcinonensis (SEBarc) 10: 263-286. Morey, C.R. 1959. The gold-glass collection of the Vatican Library, with additional catalogues of other gold-glass collections. Vatican City. Nenna, M.-D. 2000. La route du verre: ateliers primaires et secondaires du second millénaire avant J.-C. au moyen

age (Association française pour l’archéologie du verre. Encontres 12, 1997, Lyon, France). Lyon. Nenna, M.-D., M. Picon and M. Vichy 2000. Ateliers primaires et secondaires en Égypte à l’époque gréco-romaine, in Nenna: 97-112. Oliver, A. 2001. A glass opus sectile panel from Corinth. Hesperia 70: 349-63. Oliver, J.H. 1965.Texts A and B of the Horostheia Dossier at Istros. Greek, Roman and Byzantine Studies 6: 143156. Olivier-Poli, G. 1825. Continuazione al nuovo dizionario istorico degli uomini che si sono renduti piu’ celebri per talenti, virtù, scelleratazze, errori, ecc., la quale abbraccia il periodo degli ultimi 40 anni dell’era volgare 7. Naples 1. Ortalli, J. 2000. Rimini: la domus ‘del Chirurgo’, in Aemilia. La cultura romana in Emilia-Romagna dal III secolo a.c. all’età costantiniana: 513-26. Venice. Paolucci, F. 2002. L’arte del vetro inciso a Roma nel IV. secolo. Florence. Plogmann H.H. (ed.) 2006. Fisch und Fischer aus zwei Jahrtausenden (Forschungen in Augst 39) Augst. Rébillard, E. 2009. The Care of the Dead in Late Antiquity (Translated by Elizabeth Trapnell Rawlings and Jeanine Routier-Pucci). Cornell. Smith, W. 1872. A Dictionary of Greek and Roman Geography 2nd edn. London. Stern, E.M. 1999. Roman Glass-blowing in a Cultural Context. American Journal of Archaeology 103: 441-484. Stiaffini, D. and S. Ciappi (eds) 2010. Trame di Luce. Vetri da finestra e vetrate dall’età Romana al Novecento. Pisa, 12-14 novembre 2004. Milan. Vattuone, L. 2010. Vetri da finestra e vetrate a Roma e in Vaticano: venti secoli di storia, in Stiaffini and Ciappi: 7-16. Vickers, M. 2011. The Wilshere Collection of early Christian and Jewish antiquities in the Ashmolean Museum, Oxford, in Miscellanea Emilio Marin sexagenario dicata. Kačić, Acta provinciae SS. Redemptoris Ordinis Fratrum Minorum in Croatia 42: 605-14. Vopel, H. 1899. Altchristlichen Goldgläser: ein Beitrag zur altchristlichen Kunst- und Kulturgeschichte. Freiburg im Breisgau. Walker, S. (ed.) 2017. Saints and Salvation. The Wilshere collection of gold-glass, sarcophagi and inscriptions from Rome and Southern Italy. Oxford. Walker, S., A. Shortland and J Henderson 2018. Patterns in production: the Wilshere Collection of gold-glass examined, in D. Rosenow, M. Phelps, A. Meek and I. Freestone (eds) Things that Travelled. Mediterranean Glass in the First Millennium CE: 368-383. London. Available free on-line at www.ucl.ac.uk/ucl-press). Zanker, P. 1995. The Mask of Socrates. The Image of the Intellectual in Antiquity (translated by Alan Shapiro). Berkeley and Oxford.

133

Mediterranean drinking habits in Roman Britain: celery-flavoured wine prepared in an Iron Age bronze strainer Eberhard W. Sauer, Mark Robinson and Graham Morgan1 EWS first met Michael Vickers when he was a graduate student and running the Oxford University Archaeological Society excavations at Alchester, a Roman invasion-era fortress that developed into a town, 15km NNE of Oxford. This project created a huge quantity of finds, and Michael very kindly helped us to store and examine them on the premises of the Ashmolean Museum. More than a decade later, it was again with Michael’s help that EWS was able to embark upon a new project in frontier territory, Dariali Gorge in Georgia. It thus seems appropriate to offer a contribution to his Festschrift that focuses on one of our most fascinating finds at Alchester and on wine, for which Georgia has been famous for millennia (Vickers 2008: 41-2) and a beverage suitable for a celebration, even if the celery-flavour may perhaps not be to everybody’s taste.1

and, presumably, those of the Roman fortress which preceded it, it seems likely that the ditch was covered in this area to facilitate movement along the defences. It is conceivable that the remains of beamslots, which respect the drainage ditch, may have formed part of the support for the postulated wooden cover. If so, this could explain the non-recovery of the vessel, probably accidentally washed into a covered drain. The fact that the three shallow beamslots all come to an end within centimetres of the ditch, without being cut by it, renders it in any case unlikely that they are earlier. While we should allow for the possibility that veterans continued to employ familiar construction techniques after the withdrawal of the army, there is so far no other evidence for civilian-period beamslots at Alchester, thus adding strength to an attribution of the slots and the ditch to the military phase (AD 43/44-c. 55/65).

A strainer flushed into a covered drain at Alchester Roman Fortress

A Late Iron Age work of art The strainer was found in what must until recently have been anaerobic conditions. Parts of it were covered in patina, but others looked like a shiny bronze vessel produced yesterday. The whole was lifted more or less en bloc, with most of the object coming into the lab at the University of Leicester in one block of soil. Examination, during the micro-excavation process, showed that there was variable preservation. The main bowl was severely flattened and fragmented, with some areas showing little corrosion but others with complete mineralisation. Apart from the thin and brittle sheet of the bowl, there was also the spout, rim, suspension ring and two feet (Figures 3-9).

Archaeological excavations in 2002 unearthed the remains of a late Iron-Age work of art, a strainer with a fish-head spout. The vessel was found in Trench 32, in the lower fill of a west-east flowing drainage ditch within the area of the Claudian military base, just southeast of the west gate of the later town (Figures 1-2). It was found squashed, but largely complete, some 60 to 130 mm higher than the deepest point of the ditch, on its south slope immediately on top of the natural gravel the ditch was cut into (making the en-bloc recovery difficult). This suggests that it became trapped in this position at a very early date and while the drainage ditch was still functioning. The pottery from the fill of the ditch suggests that it had been entirely filled up in the later first century or, at the very latest, early second. There were not enough datable finds to establish an independent date for its bottom fill. However, as the feature is forming a T-junction with a deeper northsouth running ditch, it must be contemporary or later. The pottery assemblage and coins from the bottom of the latter point towards a mid-first century construction date (Cooper and Wells 2020; Sauer 2020), perhaps soon after the foundation of the fortress in AD 43/44 (Sauer 2005; 2020). Considering its location near the western limits of the later Roman town

The spout, ring and feet were generally more corroded than the sheet of the bowl. The bowl required little cleaning, most of the burial deposit being easy to remove with a water and alcohol solution. After cleaning, the bowl was stabilised in benzotriazole solution and lacquered with Incralac. The fittings were however more difficult, the hard, and fairly massive corrosion, particularly on the spout and ring, requiring the use of “Airbrasive” cleaning. No sign of any deposits were found on the bowl, but the spout and perforated strainer did have traces of substances, which were analysed separately. During the cleaning various features were uncovered. Traces of solder were found where the spout had been soldered onto the bowl, as in case of other strainers,

1 

Authorship: conservation and technical analysis (GM), archaeobotany (MR), excavations, archaeological and literary evidence (EWS)

134

Sauer, Robinson and Morgan: Mediterranean drinking habits in Roman Britain

Figure 1. Alchester in the military era.

135

Wonders Lost and Found

Figure 2. Trench 32 with military-period features and the findspot of the wine strainer.

Figure 3. The fish-head spout.

Figure 4. The fish-head spout in profile.

136

Sauer, Robinson and Morgan: Mediterranean drinking habits in Roman Britain

Figure 5. Side view of the strainer with position of potential lid, drawn by Vanda Morton (scale: 1:2).

Figure 6. Front view of the strainer, drawn by Vanda Morton (scale: 1:2).

Figure 7. Back view of the spout with fragments of the sieve that would have covered all of the spout, drawn by Vanda Morton (scale: 1:2).

notably that from Felmersham (Kennett 1970: 86; cf. May 1971: 256), over the perforated sieve area, and where the feet had been attached to the base. The integration of the sieve into the bowl at the juncture with the spout (Figure 7), as in case of the Łęg Piekarski specimen (Megaw 1963: 29 fig. 2, pl. 13a; Jope 2000b: pl. 169b), marks it out as a purpose-made strainer, as opposed to the bowls from Welwyn Garden City in Hertfordshire, the strainer from the ‘doctor’s burial’ at Stanway (Crummy et al. 2007: 222-3 with fig. 113-14) and two strainers from a hoard at Kingston Deverill in Wiltshire with their strainer panels merely slotted into

position (Stead 1967: 25, pl. V; Jope 2000b: pl. 307b-c; Worrell 2006: 460-2). The diameter of the punched holes ranged from 1.0 to 1.8 mm. Various marks were found on both the inside and outside of the bowl. Concentric lines around the circumference of the bowl allowed the reconstruction of the fragments although it was too flattened for the three-dimensional shape to be restored. The maximum diameter of the bowl was 248 mm, the inner edge of the rim measured 229 mm across, and it was 70 mm high (Figures 5-9) – dimensions within a few percent of its equivalent from Łęg Piekarski (Megaw 1963: 27-8 with fig. 1). Similar to other strainer 137

Wonders Lost and Found

Figure 8. The strainer seen from above, drawn by Vanda Morton (scale: 1:2).

equidistant to, the strainer’s axis of symmetry, ensuring that the vessel remained stable when standing or when the ring was used to tip the vessel to pour out the liquid through the spout. This is a fascinating use of geometry where three scratched concentric circles position the feet. Intersecting arcs based on the radius length of the inner circle formed petal-like pointers, trisecting the circumference. Placing the inner circle just apart from the feet would prevent solder from obscuring the pointers during positioning (observations by Vanda Morton and Graham Morgan). A strainer bowl found in the ‘doctor’s burial’ at Stanway had also three peltashaped feet of different shape (Crummy et al. 2007: 222 fig. 113, 224).

bowls, there is a suspension ring opposite the spout for lifting or tilting the vessel, which would have facilitated pouring out its contents. At the base, the vessel was merely 0.25 mm thick. Two pelta-shaped feet were attached to the base; traces of solder show where the third had been; the corroded state of the solder suggests that it had been lost before deposition. Faint engraved petal-like features within three concentric rings on the outside face of the base (Figure 9) will have served as guidelines to accurately position the feet, rather than as a decorative geometric pattern (cf. Jope 2000b: pl. 166k). Using dividers, three intersecting arcs were incised, creating three petallike shapes whose axes of symmetry divide the inner circle into three segments of 120 degrees each. Each petal points precisely to the mid-point of each foot. The centre line of the rear foot was in alignment with the centre of the ring attachment and the centre line of the spout. The outer and middle circles were used to position the outer and inner edges of each foot, resulting in a perfectly symmetrical arrangement of the three feet. The rear foot was on, and the two front feet

Concentric rings on the inside of the bowl (Figure 8), with a central indentation, showed that the bowl had been spun from the bronze sheet. Microprobe analysis showed that the very thin sheet was not gilt, but a bronze, containing 13.1% tin, 86.8% copper with traces,