Proceedings ofthe Danish Institute at Athens IV

Edited by Jonas Eiring and Jorgen Mejer © Copyright The Danish Institute at Athens, Athens 2004

The publication was sponsored by: The Danish Research Council for the Humanities Generalkonsul Gosta Enboms Fond.

Proceedings of the Danish Institute at Athens

General Editors: Jonas Eiring and Jorgen Mejer. Graphic design and production: George Geroulias, Press Line.

Printed in Greece on permanent paper.

ISBN: 87 7288 724 9

Distributed by: AARHUS UNIVERSITY PRESS Langelandsgade 177 DK-8200 Arhus N Fax (+45) 8942 5380

73 Lime Walk Headington, Oxford 0X3 7AD Fax (+44) 865 750 079

Box 511 Oakvill, Conn. 06779 Fax (+1)203 945 94 9468

Cover illustration: Finds from the Hellenistic grave at Chalkis, Aetolia. Photograph by Henrik Frost. Greek-Danish Excavations at Aetolian Chalkis 1999-2001. Third Preliminary Report

Edited by Soren Dietz, Lazaros Kolonas, Sanne Houby-Nielsen, Ioannis Moschos and Jonas Eiring

Contents

Preface Soren Dietz, Lazaros Kolonas and Michail Petropoulos

Excavations on the Hill of Hagia Triada Sanne Houby-Nielsen and Ioannis Moschos

The Coins from Hagia Triada II Georgia Z. Alexopoulou

Shellfish from the Excavations at Aetolian Chalkis Kaj Strand Pedersen

The Valley of Chalkis: Recent Rescue Excavations of a Bronze Age Site in GavroHmni Efy Saranti

The Cave of Hagios Nikolaos near Kato Vasiliki in Aetolia: Flint-technological Observations Lasse Sorensen

169

The Cave of Hagios Nikolaos near Kato Vasiliki in Aetolia: Flint- Technological Observations

Varassova (914 m) near the coastal vil Lasse Sorensen Introduction lage of Kato Vasiliki. The opening of the cave faces west and offers a good Information on prehistoric activity in view of the Gulf of Patras. Today, the mainland Greece is very scattered, and depth in this region of the gulf never Hagios Nikolaos is one of the first in exceeds 80 m, but during the periods vestigated caves in southern Aetolia. of low sea level until the last glacia- The basis of this paper is mixed lithic tion most of the gulf was land. In the material from the cave, found by dry transition between the Late Pleisto sieving random parts ofthe surface in cene and Early Holocene the sea level side the cave. Despite the un rose and the Gulf of Patras emerged.' favourable find circumstances, the The region around the gulf is well lithic material can still provide an in known for its large concentration of terpretative challenge. How do we in caves, due to the strong karstic lime terpret, with a hermeneutical ap stone found in the area.2 proach, the archaeological record from rock shelters or open-air sites, which Not many ofthose have been excavat through natural erosion or modern or ed but some lithic material was regis historic destructions, present very tered in the cave ofHagios Nikolaos in mixed deposits of lithic material? connection with the 1996 Chalkis field Some might say that the material is not survey.3 Hagios Nikolaos, named after useful, because it is impossible to date, the monastery built inside it, is a very even though it contains a vast amount large cave of approximately 400 ofresearch potential. The question still square metres, containing a large remains whether is it possible to use amount of prehistoric lithic material. different archaeological methods on The monastery was dated by the pot the lithic material from these sites. tery to the Byzantine period. During the construction of the monastery, the The Hagios Nikolaos cave is situated original cultural layers were damaged, on the western side of the mountain but it is possible that there are still

1C. Runnels eta!., in BSA Studies 3, 120-29; T.H. Van Andel and J.C. Shackleton, Journal of Field Archaeology 9 (1982), 445-54. 2 K.S. Petersen, in SPR, 269-75. 3A. Darlas, in BSA Studies 3, 303-10; Dietz and Kolonas, in FPR.

237 Fig. 1. Cave of Hagios Nikolaos. Photo Lasse Sorensen 2000.

r*>.

sealed deposits under the foundations. the cave. This caused certain problems Bedrock has not yet been reached in to the strategy of the survey. The nor any part ofthe disturbed areas. In 1999 mal procedure of a survey is for team and 2000 the Byzantine monastery and members to advance in a line, spaced its foundations were restored, which from 1 to 2 m, back and forth as if meant removal of deposits with a con ploughing a field, until the total sur siderable amount of lithic material. face is investigated. In that way it is Because the material from the 1996 possible to cluster the habitation and field survey consisted of approximate flint knapping areas. The strategy has ly 50 pieces of debitage, it was decid been used and revealed several impor ed to conduct further surveys in and tant prehistoric sites in Boeotia, around the cave in order to get a more around Nemea in the Argolis, in accurate picture of the lithic material Euboia, in the Grevena area, and final (Figs. 1-2). ly some of the many sites in Epirus (Asprochaliko, Kastritsa, Magalakkos, Field surveys Boila and Klithi). The sites in Epirus were discovered and explored by In 1999 and 2000, five field surveys British teams led by E. Higgs and G. were carried out around and inside the N. Bailey.4 cave. It was observed that the earlier disturbance from the monastery and In the case of Hagios Nikolaos, this the current restoration had spread lith survey method would be pointless ic material all over the surface inside since the lithics were not found in situ.

4Perles 2001, 24; T.H. Van Andel and C. Runnels, Beyond the Acropolis (Stanford 1987), 33ff.

238 Fig. 2. Interior of the cave with foundations of the Byzantine monastery. Photo Lasse Sorensen 2000.

The lithics were therefore collected Some burnt lithic material was also randomly all over the cave. A general found in the cave, indicating the pres problem concerning surveys is the fact ence of fireplaces or hearths. Organic that the eye focuses on large objects material inside the cave consisted of and that small ones, like small flakes bones and a large amount of land and tools, are often overlooked. The snails. The bones were most probably lithic material in the cave was there refuse from the Byzantine monastery, fore found by dry sieving the soil but it cannot be excluded that some of though a 4 mm mesh. This sieving the bones could have a much earlier method has proved its efficiency in date. The snail shells are particularly finding small stone objects in the de interesting, since they have been regis tailed survey at the Early Neolithic site tered in layers of other sites with of ^atalhoyiik in Turkey.5 Mesolithic and Palaeolithic material.7 This particular find situation, together By dry sieving a large amount of with coastal molluscs, occurs in caves worked lithic material was recovered from the southern part ofGreece, such from both inside and outside the cave, as Franchthi, Zaimis, Ulbrich, Sidari including some small flint chips from and Klisoura. In Klisoura and the cave. Experimental flint knapping Franchthi almost all the land snails be shows that such small chips occur in longed to the same species, Helix connection with flint knapping areas.6 Figulina, found in the Late Pleisto-

5D. Baird, in I. Hodder (ed.), On the surface: Catalhoyuk 1993-95, Catalhoyuk Project 1 (London 1996), 41-46; J. Conolly, ibid., 173-98. 6L. Johansen and D. Stapert, in N.J. Conard and C.J. Kind (eds.), Aktuelle Forschungen zum Mesolithikum, Urgeschichtliche Materialhefte 12 (Tubingen 1998), 29-41. 7Pedes, in BSA Studies 3, 314; Perles 2001; Koumouzelis et al. 2001, 532.

239 cene layers. All these sites were inter ber of technological subsystems, open preted by C. Pedes as indicating the to and related with other activities, and exploitation of diversified and patchy that technology can be viewed as a se environments, resulting in a diversi ries of optional solutions, linked to fied diet, with an apparent emphasis gether in the form of a decision-mak on the exploitation ofnumerous small- ing strategy, aimed at achieving specif sized species. This might also have ic objectives.9 The results from these been the situation for the prehistoric analyses make comparisons with lithic habitants in Hagios Nikolaos. Finally industries and settlement models from it is important to stress that the only other parts of Greece possible. In the pottery found in the cave was of following passages of this paper, the Byzantine date. The lack of earlier ce Hagios Nikolaos material will be ramics gives a terminus ante quern for analysed according to the same meth the lithics in the Early Neolithic peri ods. od. Is it possible to achieve an even more detailed date? Raw material studies

Some preliminary The predominant flint type, represent questions ing 98 % ofthe material, is radiolarite, which is dark red or almost chocolate- The find circumstances give rise to coloured flint. The quality is mainly some preliminary questions concern good and fine-grained, and radiolarite ing the lithic material. Is it realistic to is a very popular raw material, com recreate a picture of the original func mon in many Palaeolithic and tion and activities inside the cave, and Mesolithic assemblages in Greece and is it possible to give a preliminary date Anatolia.10 In certain cases, when the of the lithic material from a non-exca site is located near chert sources or far vated settlement just by analysing the away from the coast, radiolarite is rar lithic material from five surveys? er. The fact that there was found a large amount of worked radiolarite In order to answer the questions, it is material in the cave proves that there necessary to use archaeological meth was easy access to that type of raw ods such as the chaine operatoire, material. Although the exact prove which includes raw material studies, nance of the radiolarite remains under flint technology and flint typology. The investigation, the preliminary surveys advantage of chaine operatoire is the indicate that it is procured within a ra focus on human intentions, decisions, dius of approximately five kilometres choices and alternative pathways of from the site. It is possible that the in technological production as described habitants of the cave found the radio by Pedes.8 It is based on the presump larite after it had been washed out of tion that lithic industry is one ofa num the small stream, which runs down the

8Franchthi 3, 27-84. 9Tixier quoted in Franchthi 3, 26. 10 E. Adam, in BSA Studies 3, 267; M. Otte et ai, in BSA Studies 3, 82; Koumouzelis et al. 2001, 525; Pedes 2001, 202.

240 side ofthe mountain in the raining sea as can be inferred from the different son." A similar situation is observed at morphological types, whereas the ex Upper Palaeolithic settlements in otic materials are generally absent in Epirus such as Klithi, Megalakkos and the earlier stages of tool making, Boi'la.12 These sites rely almost exclu which is raw material acquisition and sively on locally available flint, which core production. This makes the radio is collected in the form of pebbles larite assemblage ideal for further from rivers, and in form of nodules studies in the chaine operatoire.13 from primary contexts or, more com monly, from secondary deposits. Flint technology

The remaining two per cent ofthe lith The lithic material from Hagios ic material from Hagios Nikolaos con Nikolaos showed evidence of four dif sists of a dark grey and very coarse ferent flint-knapping techniques. The grained flint, and a light yellow fine first technique, which is the hard direct grained flint. Exotic flints are present percussion, is a direct blow against the mainly in the form of blades and tools, raw flint nodule with a hammer stone and less commonly as cores. Although made of quartz. Two of these hammer these exotic raw materials are present stones were found inside the cave. The in relatively small quantities, their flakes produced from hard direct per presence could be interpreted either as cussion are relatively thick and have a by-product of visits to the site from clear scars on the surface. The primary further afield, or they could have been ends are heavily curved and often have introduced as part of a deliberate pro scars. It is the "oldest" technique, but curement strategy, aimed at comple is known in all prehistoric contexts, menting the deficiencies of the local being the basis for every flake or blade flint for producing longer blades, and production.14 for the production of tools made on small flakes and bladelets. The second technique is the soft direct percussion, which is a direct blow The difference between local and ex against the raw flint nodule with a soft otic raw materials is also mirrored in hammer stone made of chalk or a soft the technological debitage types, er material, e.g. antler, bone or hard when separated according to type of wood. This technique is also the start material. Local raw materials are gen ing point for many kinds of lithic pro erally present in all stages of tool pro ductions, but is often associated with duction (raw material acquisition, core Upper Palaeolithic assemblages. The production and exploitation, tool pro flakes produced in this particular tech duction, tool maintenance and discard) nique are often thin and flat, and fur-

11 Demoule and Pedes 1993, 358. 12 E. Adam in Klithi 2, 482ff; A. Sinclair in Klithi 2, 421; C. Roubet in Klithi 1, 132ff; W. Shawcross and N. Winder, in Klithi 1, 181 ff; E. Kotjabopoulou et al., in Klithi 2, 427-37. 13 L.G. Straus, 'The Role of Raw Materials in Lithic Assemblage Variability', Lithic Technology 3 (1980), 71; E. Adam, in Klithi 1997, 48If. 14 Inizan et al. 1999, 74; Andrefsky 1998, 115ff.

241 thermore the bulb of percussion is peared only 12,000 years ago in Japan, wide and flat. The lithic material from whereas, its invention can now be the cave shows a large concentration traced to the Sibero-Sino-Mongolian of flakes and some blades, which are region, about 25,000 years ago.17 knapped with this particular tech Pressure flaking requires the flint nique. The technique is also used in knapper to have particular technologi many lithic productions throughout cal skills, which makes this technique prehistory and is therefore not espe more difficult to master than other cially useful for dating the lithic mate ones. Instead ofresulting from a blow, rial.15 the microblades are pressed offthe mi- croblade core with hard pressure using The third technique is the soft elastic a pressure flaker. The core is fixed be indirect percussion, better known as tween the flint knapper's feet or can be punch-percussion, which, as the name fixed in a device made of wood. The implies, is an indirect blow against the pressure flaker is often pointed and flint core. Here an antler billet and a made of antler or bone. The accuracy hammer made of wood or antler is with which the pressure point can be used. The main products from this positioned leads to maximum preci technique are blades. The blades are sion and standardisation. This makes more than 5 cm long and have a char the recognition criteria on the cores acteristic lip at the proximal end. very typical. The cores must have a Some of the blades from the cave are very rectilinear and regularly parallel knapped in this technique, which arrises and the platforms can be corti emerges in Europe in the later part of cal, plain or prepared. There are two the Upper Palaeolithic and continues basic forms of cores, the pyramid and in Greece until the Middle Neolithic the parallelepiped. The pyramid cores period. The technique is therefore a do not appear in the Hagios Nikolaos better option of dating lithic material material but the parallelepiped do. than the two previous ones.16 They are described as flat, have a sin gle debitage surface or two successive The fourth technique used in the cave ones, and the technique could be ob is pressure flaking, which applies to served upon a number of microblade the manufacture of blades, especially cores from the cave.18 bladelets, only. Documentation of the technique has lately recently in In Europe, pressure flaking on micro- creased. The area of distribution in blade cores is characteristic of the cludes the Middle and Far East. Until Mesolithic technological tradition, but very recently, it seemed that pressure it can also occur in the transition be debitage of blades followed the indi tween the Upper Palaeolithic and rect percussion of blades, and ap Early Mesolithic periods. It is a tech-

15 Andrefsky 1998, 114ff; Inizan et al. 1999, 74. 16 Andrefsky 1998, 1Iff; Inizan et al. 1999, 76. 17 Inizan et al. 1999, 76ff. 18 J. Pelegrin, in J. Tixier (ed.), Prehistoire de la pierre taillee 2: Economie du debitage lami- naire (Paris 1984), 112; Demoule and Pedes 1993, 383; Inizian et al. 1999, 76ff.

242 nique ensuring the production of sequences or any significant area of straight microblades. The technology knapped debitage, which would allow was spread to a wide area in very short one to recognise a true operational time, something which must be related chaine in the Hagios Nikolaos materi to the efficiency of the backed blades, al, it was necessary for me to recreate which could be used as arrowheads.19 virtual operational chains, by analysing the various categories of Chaine operatoire cores, flakes and blades. Each lithic specimen was not only counted and The term chaine operatoire has been individually identified, but was widely used in French archaeology for clearly assigned to the different stages the last twenty years.20 The concept of of the knapper's sequence of opera chaine operatoire did not really come tions. The procedure allows the mate to the forefront in Greece until the ex rial to be regrouped and expicitly con cavation of a continuous stratigraphi- struct the various operational chains, cal sequence at Klithi in 1988, along which make up the industry as a with the definitive publication of the whole, and to identify more precisely Franchthi cave material.21 A detailed the knappers' objectives.24 The chaine method, such as the chaine operatoire, operatoire of the Hagios Nikolaos is not normally used on survey materi material involved the examination of al from sites in Greece or Europe.22 1861 artefacts, 176 of which are tools This is a shame, because the chaine or retouched pieces. It was hereby operatoire can squeeze the maximum possible to identify four main stages research potential out of a material. in the chaine operatoire (Table 1 and Sometimes it can even direct the re Fig. 3). searcher into a different perspective of the material. It can also be a structured Initial debitage, phases l-3:The raw guideline of how to analyse and de nodules were probably carried into the scribe a collection of material.23 cave to begin the flint knapping, using both a hard direct percussion and soft On the basis of the four different flint direct percussion. This corresponds techniques seen in the Hagios with the flake material and the small Nikolaos material it is now possible to chips found in the cave. Certain flake analyse the debitage using the chaine types indicate a preliminary prepara operatoire. Since there are no refitting tion of the cores' reduction sequences,

19 W, Andrefski, Lithics (Cambridge 1998), 194f; Inizianet al. 1999, 76ff. 20 E.g. J-M. Geneste, in J.P. Rigaud (ed.), La Grotte Vaufrey, Memoires de la Societe Prehistorique Francaise 19 (1989), 441-517; E. Boeda, in M. Otte (ed.), L'homme de Neandertal 4 (Liege 1986), 13-26; J. Pelegrin et al., in J. Tixier (ed.), Technologie Prehistorique (Paris 1988), 153. 21 C. Roubet, in Klithi 1, 125ff; Franchthi 3, 85ff; Franchthi 5, 2Iff. 22 C. Runnels et al., in BSA Studies 3, 127;A. Darlas, ibid., 303ff. 23 M.A. Dobres, Technology and Social Agency Outlining a Practice Frameworkfor Archaeology (Oxford 2000), 164ff. 24 C. Roubet in Klithi 1, 125ff

243 Retouched pieces cts Percentage

Endscrapers 33 18.8 Perforators 4 2.3 Burins 2 1.1 Scrapers 9 5.1 Backed bladelets 49 27.8 Flakes with retouche 35 19.9 Blades with retouche 12 6.8 Blades fragmented with retouche 31 17.6 Burned blade with retouche 1 0.6

Total retouched 176 100,0

Table 1. Retouched artefacts from the Hagios Nikolaos material.

Artefacts Percentage Unretouched

cores 76 4.5

Cores 19 1.1 Microblade cores 51 3,0 Burned cores 6 0.4

Debitage production 1192 70.7

Blade 38 2.2 Microblade 16 0.9 Flakes 743 44.9 Blade fragmented 139 8.2 Microblade fragmented 58 3.4 Burned flake 186 11,0 Burned blade fragmented 12 0.7

Debitage technical 417 24.7

Blade with cortex 7 0.4 Striking platform 7 0.4 Flake with cortex 29 1.7 Small flakes & flint chips 343 20.4 Hinge fractured pieces 8 0.5 Plunging terminated pieces 8 0.5 Crested blade 15 0.8

Total unretouched 1685 100,0

Table 2. Unretouched artefacts from the Hagios Nikolaos material.

244 Phase Action

Phase J Simple selection and Procurement of test knapping ofthe the raw nodules. raw nodules.

Phase 2 Preliminary shaping Preperation of ofthe core. the core. Knapping an initial flake coverd with cortex.

Phase 3 Shaping ofthe core. Preliminary blade Production ofblades production. which is totally or partly coverd with cortex

Phase 4 Production ofblads which Main blade is usedtoproduce other production. tools: Retouchedblades, scrapers,burins and perforators.

Phase 5 Thecore is reshapedby Secondary blade knapping crest blades of production. the core in oder toproduce blades. Because ofa hingefracture or a plunging termination.

Phase 6 The core is reusedinto Reworking the core. two different types of microblade cores. It couldalso be knapped into a smallflaking core.

Phase 7 The microblades is Production of knappedinto backed microblades bladelets type 1. from microblade cores type I.

Phase 8 The microblades is Production of knappedinto backed microblades bladelets type 2. from microblade cores type 2.

Fig. 3. The chaine operatoire of the Hagios Nikolaos material.

245 Fig.4. Artefacts from Hagios Nikolaos: 1-6,Microblade corestype 1.7-13, backed bladelets type 1.(All drawings by L. Sorensen).

246 for instance flakes covered with cor logical trends in European tex, and blades totally or almost total Mesolithic.26 ly covered by cortex. This reduction sequence is normal when the raw ma Final debitage, phase 5-6: Flake pro terial is found near the habitation site. duction is now primary and blade pro The same initial sequence is registered duction is secondary. The production on prehistoric sites all over Greece, sequence includes some reshaping of where the dominating flint resource is the core, due to hinge fractures, step local.25 fractures and plunging terminations. The actual reshaping of the core is Full debitage, phase 4: The lithic ma made by a series of core tablets. The terial reveals a preliminary and main core ends up with a multiform irregu production of especially blades. The lar shape, which has the size of an blades are of varying quality and are egg. The complete knapping of al often thick with a curved angle, which most all the cores meant that no raw makes a further production of tools material was wasted. Many small difficult. The exploitation continued cores are found, which are either until the core was exhausted, using cased away or discarded. Although soft indirect and soft direct percussion the flake production appears to be ir to maximise the full potential of the regular and heterogeneous, recent re core. The low frequency of long search proves otherwise. A similar blades in the debitage suggests either production of flakes is the discoid that very few cores started, large, or method, described by E. Boeda.27 The that many long blades have been re method has a wide time frame, as it moved from the site. The blade pro has been identified in lithic industries duction is a classical dating marker in from the Mousterian to the Neolithic. the Aurignacian, Epigravettian and The production consists of an inten Early Mesolithic blade production in tional chaine operatoire, but it has not Greece. Normally it should also be a yet been thoroughly identified in any valid marker in the Late Mesolithic lithic industry in Greece, although period in Greece but, as proven by there are signs of the reduction Pedes, the assemblage of lithic phase method at recently discussed VII Franchthi (ninth mellenium B.C.) Mousterian sites in Greece.28 is dominated by a flake production. The Mesolithic period in reveals dis Reuse debitage, phase 7-8 : The cores tinct technological traditions in could also at an earlier stage have been Greece, and the area could have been reused and re-knapped into microb cut off from the widespread techno lade cores, but from observations of

25 C. Roubet inKlithi 1, 132ff; W. Shawcross and N. Winder, ibid., 18Iff; E. Adam, inKlithi 2, 482ff; E. Kotjabopoulou, ibid., 427ff; A. Sinclair, ibid., 421; M. Otteet al., in BSA Studies 3, 82; E. Adam, ibid., 267; Koumouzelis et al. 2001, 525; Pedes 2001, 202. 26 Franchthi 3, 2Iff. 27 E. Boeda, Bulletin dela Societe Prehistorique Francaise 90 (1993), 392-404. 28 Kourtessi-Philippakis 1986, 150; D. Papagianni, Middle Palaeolithic Occupation and Technology in Northwestern Greece, BAR, IS 882 (Oxford 2000), 35ff.

247 ^> -^

Uo

<-h ^

On

Kj

\) ^

-K K» u*

Oo

^v

^> the microblade cores, it was possible Tool types to detect a difference in the technique. Microblade cores type 1, which ap Which tools did the occupants of the proximately have the size of a fist, Hagios Nikolaos cave produce from mostly are of a conical and one-poled the blades and can the assemblage shape, but are sometimes two-poled help dating the material? One must with a cylindrical shape. They are ir bear in mind that artefacts could be regular in shape and have many deep come mixed from trampling or a slow and wide scars, indicating a hard direct rate ofdeposition, and prehistoric arte percussion with the use of hammer facts are almost always mixed with stones made ofquartz. Another indica those of later or earlier periods.29 The tion is the many hinge fractures, which Hagios Nikolaos material has been to are seen on the sides of the core (Fig. tally mixed, which makes the separa 4.). Microblade cores type 2 are ofap tion of tools into different categories proximately the same size as type 1, practically impossible, and is further but have a high front and are often complicated by the lack ofcharacteris conical. The type has almost parallel, tic tools.30 Is it still possible to make a small and thin scars. The microblades, preliminary typological dating of the which are knapped from these microb tools? In order to answer this question, lade cores, represent a totally different it is necessary to look at the retouched technique than type 1. Pressure flaking pieces from the Hagios Nikolaos ma was probably used, but it cannot be terial. excluded that a soft direct percussion were used on some of the type 2 mi The analysis of the Hagios Nikolaos croblade cores (Fig. 5.). material includes 176 retouched pieces from a sample of 1861 pieces, e.g. end The chaine operatoire, and in particu scrapers, perforators, burins, compos lar the two different microblade ite tools, scrapers and backed pieces. cores, give rise to some questions. The tool inventory is dominated by Are we dealing with a single cultural backed bladelets (28 %). The total phase with only slight chronological scraper index is 24 % and is much variations or several, diachronic, higher than the burin index (1 %). End phases? Is there any methodological scrapers represent 19 % of the tool as approach that can confirm or invali semblage. There are no geometric mi date this initial hypothesis? How is crohms or shouldered pieces repre this hypothesis to be tested in the ab sented in the assemblage. Tools made sence of any stratigraphical reference on bladelets account c. 60 % of the sequence for the whole site? Those core inventory, corresponding to the are the key methodological chal large amount of microblade cores (51 lenges posed by the Hagios Nikolaos micro cores, type 1 and 2), which is material. the most dominant core type in the

29 C. Roubet, in Klithi 1, 126; E. Adam, in BSA Studies 3, 267; C. Perles, ibid., 312; Ead. 2001, 3 If; Koumouzelis et al. 2001, 523f. 30 C. Runnels, AJA 99 (1995), 702ff.

249 Hagios Nikolaos material (see Table backed bladelets and, nosed end scrap 2)." The production of the backed ers. These tool types dominate the clas bladelets was important, because it is sic Epigravettian assemblage. The dat presumed they were hafted either indi ing range of the Epigravettian corre vidually or in a group, and may have sponds to the stratigraphical gap be formed the cutting or piercing ele tween lithic phases III and IV in the ments of composite tools.31 Industries Franchthi Cave, C14-dated from 21,480 of this character have also been found +/-1,270 to 12,540 +/-180 years BP. The in the Balkans. The backed bladelets Late Epigravettian period at Franchthi could also have been used as mi- corresponds to the lithic phases IV to VI crolithic backed bladelets, which is a and is C14-dated from 12,540 +/-180 to common element in many Late Upper 10,260+/-110 BP.35 Palaeolithic assemblages from Europe, Southwest Asia and further There are also other C14 dates corre afield.32 sponding to the Epigravettian period. For instance at Asprochaliko 18,000 All these characteristic tool types, end +/-300 to 14,000 +/-600 BP and at scrapers, perforators, burins, compos Kastritsa 19,900 +/-370 to 13,400 +/- ite tools and scrapers are typical in 210 BP. Finally, there are the many several prehistoric assemblages such C14 dates from excavations in Epirus: as Aurignacian (lamelles Dufour), Klithi, 17,000 +/-400 to 10,420+/-150 C14-dated at Klisoura between 26,000 BP; Megalakkos, 16,100 +/-160 to and 22,500 BP, and at Franchti (lithic 15,410+/-210BP;Voidomatis, 11,100 phase I) c. 30,000 BP;33 Gravettian, +/-200 to 10,700 +/-200 BP, and Boila, C14-dated at Asprochaliko between 13,810 +/-130 to 10,190 +/-90 BP.36 26.100 +/-900 and 25.100 +/-700 BP, The C14 results cluster from c. 20,000 at Franchti (lithic phase II) between to 10,000 BP, which is the chronologi 22.330 +/-1270 and 21.480 +/-350 BP, cal time frame for the Epigravettian and at Kastritsa between 21.800 +/- period in Europe and Greece. The 470 and 20.200 +/-480 BP.34 Some of Epigravettian has a geological spread the lithic industries and tool types from Italy to the Balkans, Eastern could belong to either of these cate Europe, Greece and Anatolia.37 The gories. lithic industry in the Epigravettian pe riod is dominated by blade production, In the Hagios Nikolaos assemblage especially of backed bladelets, which there are two dominant tool types: corresponds to the chaine operatoire

31 P.V. Petersen, Flintfra Danmarks oldtid (Copenhagen 1993), 85. 32 Adam 1989, 252ff. 33 Franchthi 3, 89ff; Djindjian et al. 1999, 379; Koumouzelis et al. 2001, 522. 34 Franchthi 3, 98ff; Djindjian et al. 1999, 379f. 35 Franchthi 3, 108ff; Djindjian et al. 1999, 379ff; Koumouzelis et al. 2001, 534. 36 J. Gowlett et al, in Klithi 1, 27ff. 37 D. Mihailovic, in BSA Studies 3, 343ff; M. Otte et al., ibid., 82ff; D. Mihailovicand V. Dimitrijevic, in A.Thevenin (ed.), L'Europe des derniers Chasseurs. Epipaleolithique et me- solithique (Paris 1999), 391-398; Djindjian et al. 1999, 282ff; Mussi 2001, 286ff.

250 and the dominant tool types from A typological dating Hagios Nikolaos (Table. 1). The typological dating involves a dis Some of the tool types at Hagios cussion ofthe last two phases (Phase 7 Nikolaos, could also belong in the and 8) of the chaine operatoire analy Mesolithic period, which spans from c. sis of the material from Hagios 9,500 to 8,000 BP.38 The Mesolithic in Nikolaos. The fact that there are two Greece is a rarely documented period. different types of microblade cores in Mesolithic finds are more difficult to dicates that there must have been two identify than both earlier and later ma different types of microblades, which terial. The Mesolithic stone industry is again were reproduced into backed on the whole undiagnostic. The diag bladelets. The backed bladelets is a nostic microliths are most difficult to very important group, since the lithic spot on the surface. The period is also material can be typological dated from very short, since the Neolithic process the shape of these arrowheads. The came quickly to the Aegean area.39 As amount of flint in a backed bladelet is seen above, there is a vast amount of very small, and would therefore not dating possibilities from the different have been rational to reshape the ar tool types. A more detailed typological rowhead for reuse as another tool. dating could perhaps be provided by the backed bladelets. There were found two different types of backed blades in the cave. Both The production of backed bladelets in types had a retouche down along the the Hagios Nikolaos assemblage is side of the blade. Type 1 of backed seen not so much from the backed bladelet is thick and wide, the bladelet bladelets themselves, but from the has been knapped with a hard directly large amount of microblade cores, percussion and it was produced from suggesting an abundance of backed the first type ofmicroblade cores (Fig. bladelets. This can be related to sea 4). Type 2 of backed bladelet is thin, sonal hunting. One might expect peo narrow and has a straight back. These ple to arrive at the site with toolkits, bladelets are knapped either with already prepared for their first hunting pressure flaking or soft direct percus forays. As the various components of sion, which corresponds with the sec the toolkit became broken in the ond type of microblade cores (Fig. 5). course of use, they would have been The typological analysis ofthe backed discarded and replaced with locally bladelets suggests that there is a tech made elements. The production of nological change in time, but this backed bladelets would therefore rep needs to be validated by a wider range resent a major activity at the site. of comparable material, discussed in

38Perles2001,26. 39 Pedes 2001, 23.

251 what follows. The two different types ble-backed, and occasionally with a of backed bladelets has not only a dif transversal retouche in the form ofmi- ferent appearances, but also a different crolithic rectangles.44 The uppermost hafting, deduced from their different Aurignacian, layer III, has been CI 4- shapes (see below), again implying a dated between 21,720 +/-90 to 16,130 chronological discrepancy. It is very +/-40 BP. The interface between layers rare to observe two different arrow Ha and lib has been C14-dated to head types, made in two different flint 14,280 +/-90 BP. There is some uncer knapping techniques, in use at the tainty about these dates because of the same time. reservoir effect.45 The industrial layer of lib at Klisoura is Early The first type of backed bladelet has Epigravettian, and may be placed similarities with an early phase of chronologically between the lithic backed tool industries (Gravettian- phase III and phase IV at the Franchthi Early Epigravettian), identified at sev cave. Dated between 21,480 +/-350 eral sites in South-eastern Europe. In and 12,540+/-180 BP.46 the lower layers of Badanj,40 in layer IX at Asprochaliko, the early phases The results from the Klisoura cave in at Megalakkos,41 and lower layers at dicate that these backed bladelets have Kastritsa,42 and lithic phases II and III no parallels in western Greece, al from the Franchthi cave.43 These are though they are identified in the assemblages dated over a large time Hagios Nikolaos material. It differs span from the twenty-fifth to the from the shouldered point industries twelfth millennium B.C. from the lower layers at the Kastritsa Cave,47 as well as from the later indus Regional comparisons of the backed tries with backed bladelets, having a bladelets from the Klisoura cave, near straight back and often made with the Franchthi in Argolis, could reveal a microburin technique at sites such as more detailed typological dating ofthe Klithi.48 This description of the later backed bladelets from Hagios backed bladelets from Klithi corre Nikolaos, especially the backed sponds to the second type of backed bladelets from the uppermost bladelets in the Hagios Nikolaos mate Aurignacian unit (layer III) to the low rial. These backed bladelets belong to er Epigravettian (layer lib). These the Late Epigravettian period. bladelets were described as having a Especially the material from the upper concave blunted back, sometimes dou levels of Klithi (1000-series material)

40 R. Whallon, in BSA Studies 3, 332ff. 41 A. Sinclair, in Klithi 2, 415ff. 42 E.S. Higgs, ADelt 21 (1966), Chronika, 292-94; Bailey et al. 1983, 15ff. 43 Franchthi 3, 98ff. 44 Koumouzelis et al. 2001, 523. 45 Koumouzelis et al. 2001, 520ff. 46 Franchthi 3, 108ff; Djindjian et al. 1999, 379. 47 Adam 1989, 104ff; Bailey et al. 1983, 15ff. 48 C. Roubet, in Klithi 1, 155ff.

252 has not only similarities in knapping pezes, all manufactured with the mi- strategy, but the same dominance of croburin technique.^7 The same mi- backed bladelets, which is also ob croburin technique is seen in the late served in the Hagios Nikolaos materi backed bladelet industry in Klithi.58 In al.49 The backed pieces are similar to the Hagios Nikolaos material no mi- other Late Pleistocene sites in Greece, croburins or geometric microliths have Anatolia, the Balkan area,50 and Italy,51 been registered, something which e.g. the upper layers of Badanj,52 should argue against a relative dating Asprochaliko,53 Boila,54 Theopatra55 of the material in the Late and Franchthi, lithic phases IV to VI.56 Epigravettian or Early Mesolithic peri The dating of these backed pieces lies ods, although it is possible to make between the twelfth and the end ofthe backed bladelets without using the mi- ninth millennium B.C. The relatively croburin technique, just by making a late dating to the transition between retouche along the backside of the the Late Epigravettian and Early blade. This particular microlith pro Mesolithic periods can explain why duction can be detected both in Klithi the backed bladelets were produced by and Franchthi.59 In Europe, there is pressure flaking. As argued above, the normally an increased number of mi pressure flaking is a flint-technologi croliths produced from the microburin cal marker for Mesolithic assem technique during the Mesolithic peri blages. Although there is a slight prob od. The reverse occurs in Greece, lem with this assumption, since, along where virtually all microliths and mi- with this backed industry, there has croburins disappear during the Early been registered yet another microlithic Mesolithic period (ninth millennium production in the Franchthi Cave. In B.C., lithic phase VII at Franchthi). the Franchthi cave (lithic phases of V This particular phenomenon has been and VI) there was an, in typological registered at Franchthi,60 Sidari,61 and terms, already fully Mesolithic assem Maroula.62 The lack of microburins blage, with numerous geometric mi- could indicate an even later dating of croliths such as scalene bladelets, tri the Hagios Nikolaos material in the angles, segments and more rarely tra Late Mesolithic period, but the lithic

49 C. Roubet, in Klithi 1, 156. 50 M. Otte et al., in BSA Studies 3, 82ff. 51 Mussi2001,286ff 52 R. Whallon, in BSA Studies 3, 332ff. 53 Baileys al. 1983, 15ff. 54 E. Kotjabopoulou et al, in Klithi 2, 427ff 55 E. Adam, in BSA Studies 3 (1999), 267ff 56 Franchthi 3, 115ff. 57 Pedes 2001, 31. 58 C. Roubet, in Klithi 1, 155ff. 59 Franchthi 3, 115ff; C. Roubet, in Klithi 1, 155ff. 60Perles2001,31ff. 61 A. Sordinas, Balkan Studies 10 (1969), 403ff. 62 K. Honea, AJA 79 (1975), 278f.

253 industry in that period argues against to either Epigravettian or the transition the presumption. It is a crude flake in between Late Epigravettian and Early dustry, which does not occur in the Mesolithic, especially because of the Hagios Nikolaos assemblage. There pressure flaking used on the microb has also been observed a large amount lade cores. It cannot be excluded that of trapezes made on flakes in the Late there was a continuous settlement in Mesolithic, which are totally absent the cave, implying a relative dating, from the Hagios Nikolaos material.63 continuous through both the Very few Mesolithic sites are known Epigravettian and the Mesolithic peri in Greece, since they often are difficult ods. But it is essential not to forget that to identify. The Mesolithic period in it is a very mixed lithic material, Greece is dated by the Franchthi cave which most certainly will give some from the mid-tenth to the late ninth uncertainty in the typological dating. millennium BP - a very short period.64 Several Mesolithic assemblages have probably been confused and mixed Comparisons in the with Epigravettian material, because lithic industry they tend to be found at the same sites. Coastal sites are especially indicative Close interregional comparisons with of the period, as pointed out by the Hagios Nikolaos material are im Runnels in Epirus.65 Mesolithic mate possible, since no other sites have rial is in general difficult to recognise been investigated in the surrounding and differentiate from Upper area, although sites are known on the Palaeolithic artefacts, especially in the other side of the Gulf of Patras in absence of dated stratified deposits or North-western Peloponnese, e.g. diagnostic artefacts such as geometric Lakkopetra, Lapa, Kastro, Loutra, microliths.66 The typological dating of Amaliada, Mavri Myti, the Piros the backed bladelets from Hagios Valley, Kalamaki and Eleochori.67 All Nikolaos lies mainly in the of these sites, except for Mavri Myti, Epigravettian and in the transition be Amaliada, Kalamaki and Kastro, con tween Late Epigravettian and the tained a production of pebble tool in Early Mesolithic period. dustry (Lower Palaeolithic) or leval- lois flakes, which is a Middle Type 1 of the backed bladelets and the Palaeolithic (Mousterian) component. microblade core corresponds to a dat The pebble tool industry or levallois ing in the Early/Middle Epigravettian, method was not identified in the whereas type 2 ofthe backed bladelets Hagios Nikolaos material, which and the microblade core can be dated makes a comparison impossible.

63 Pedes 2001, 3Iff. 64 Demoule and Pedes 1993, 364ff. 65 C. Runnels et al, in BSA Studies 3, 126ff. 66 G.N. Bailey et al, in Klithi 2, 535. 67 Kourtessi-Philippakis 1986, 124ff; A. Darlas, in BSA Studies 3, 303ff; C. Pedes, Analecta Praehistorica Leidensia 31 (1999), 378ff.

254 Eleochori had an Aurignacian assem nally by the extraction of debitage. blage, unfortunately mixed with the This strategy is deduced from several Mousterian artefacts. In Amaliada, characteristics of the industry and the Kalamaki and Kastro a mixture of presence of chips produced during tools from different Upper Palaeolithic blade removals. The existence of par periods was identified, including ob tially crested pieces used to correct the sidian artefacts and pottery sherds. core front; the presence of opposed Furthermore, the sites are only briefly platform bi-directional cores indicates published, which makes a detailed maximum exploitation for the produc comparison difficult. No detailed in tion of straight blades and flakes, vestigations of sites from the which follows the longest axis of the Epigravettian or the Early Mesolithic nodules. The alteration of the flaking periods have taken place in that part of direction on the cores, indicates maxi Greece so, once again, we turn to the mum exploitation of the usable pieces two best published sites in Greece, viz. and points, to an economy ofraw mate Franchthi and Klithi. rial imposed not so much by the quan tity of the locally available flint as by There seems to be a long Epigravettian its quality. The majority of the cores sequence in the Hagios Nikolaos ma were abandoned, either when the plat terial, which allows us to pinpoint both forms were crushed, or when the core technological and morphological front was beyond repair. changes, evident in the interchange of different techniques, for the exploita These results from the upper levels of tion ofmicroblade cores and modifica Klithi did not allow Roubet and tion of microblades. But comparisons Kozlowski to identify a precise cultur in the microblade industry with other al affiliation, only to include them settlements during the Epigravettian within a broad Epigravettian group period are difficult, since they are not ing.69 The C14 dates from Klithi lie be defined as a group in Greece. tween 16,500 and 10,000 BP.70 The current interpretation of stratigraphical At Klithi (in the 1000-series material) data and C14 dates, suggests that little and Franchthi (lithic phase II to V) of the material can be assigned with there are overall similarities in the first confidence to a later date than c. six stages ofthe chaine operatoire, al 12,500 BP. Most of the assemblage re so seen in the Hagios Nikolaos materi covered in the excavation belonged to al.68 At both Klithi and Franchthi the the period from 16,500 to 13,500 BP, knapping strategy includes the cre but it cannot be excluded that some of ation of a striking platform, followed the material from the upper levels of by the removal of the flawed exterior Klithi could have a later date, c. 11,000 surface (cortex) of the nodules, and fi to 10,000 BP, which is on the transition

6S C. Roubet, in Klithi 1, 125ff; Franchthi 3, 98ff. 69 C. Roubet, m Klithi 1, 129ff. 70 Gowlett et al, in Klithi 1, 27ff.

255 between the Late Epigravetian and resources became an important eco Early Mesolitic periods.71 In the nomic factor during Mesolithic Francthi cave the results are more de times.72 There has furthermore been tailed, but the dating ofthe lithic phas access to many permanent natural re es II to V from the twenty-third or sources near the cave, for instance twenty-second to the early twelfth mil flint, but also fresh water from the lit lennium gives the same results as in the tle creeks running down the valley Klithi assemblage. This comparison near the cave. The site is well posi with the Klithi and Franchthi industries tioned for hunting, and the valley suggests overall similarities in the opens up both to the south and to the Epigravettian knapping strategy, also west. In this case, it would be logical seen in the Hagios Nikolaos material. to see the site in several roles. 1) As a The relative dating of the lithic materi hunting station suitably distanced al in Hagios Nikolaos suggests that the from prey that might otherwise be dis cave was settled some time during the turbed. 2) As a monitoring position for Epigravettian period, but the question observing animals passing through the is why this particular cave was chosen valley. 3) (Arguably) as a well-placed and why it was abandoned during the manufacturing and repair site, espe Mesolithic period. cially for projectile shafts.73 All these general observations are present in Settlement type many caves all over Europe and in Greece.74 When the cave was first occupied, some time in the Epigravettian or in The lithic evidence concerning the ac the Early Mesolithic period, the land tivities which took place inside the scape looked totally different from cave is very clear. The burnt flint, cor now. First ofall there was no sea, since responding to eleven per cent of the the water level was c. 200 m lower total assemblage, shows evidence of than today. But during the Late fireplaces, and the small chips indicate Epigravettian and the Early Mesolithic a massive flint knapping or dumping periods the water level rose. To take area inside the cave (Table 2). The up residence in the Hagios Nikolaos cave itself gave the habitants shelter cave would have many advantages in from the weather; here the hunters both periods. In the Epigravettian may have completed their cooking, the there would have been a good view of flint knapping, and other activities in the valley. This view was an advantage their everyday life. to the hunters, when they had to go on hunting expeditions in the area. In the On the basis of these considerations, I Mesolithic period there was easy ac would argue that the assemblage at cess to the coastal areas, where marine Hagios Nikolaos, with its predomi-

71 C. Roubet, in Klithi 1, 152. 72 Pedes 2001, 25ff. 73 L.R. Binford, American Antiquity 45 (1 , 4-20. 74 Djindjian et al 1999, 239ff.

256 nance of micro blade cores to produce and Early Mesolithic. The dates are in backed bladelets of local raw materi dicated by two different types of mi als, is not indicative of a specialised cro-blade cores and backed bladelets, hunting site, but of a habitation site. the final outcome when hard direct, Furthermore, the co-existence of local soft direct and flaking percussion is and exotic raw materials also consti used. Lack of comparable material tutes evidence of contact to a wider from the rest of the region rules out area. If the procurement of local raw any regional studies and characteris materials is considered part of a daily tics of the lithic material, something routine, the acquisition of non-local which makes these interpretationspre materials must indicate an organised liminary. Overall similarities in the plan, either for the procurement of the chaine operatoire from Klithi and material itself as the main objective, or Francthi have, however, been ob in combination with other activities - served. Finally, the varied number of hunting and investigation of resources tool types and activities, which have - as a complementary activity, or taken place inside the cave, such as through exchange. However, the exot flint knapping and hearths for cook ic raw materials could give the im ing, indicates a settlement type with pression of a small hunting party, permanent character. moving rapidly over the landscape and stopping briefly here and there to kill In spite of the unfavourable find cir an animal. But the fact that the major cumstances of the lithic material from ity ofthe pieces in this assemblage are the Hagios Nikolaos cave, it is still made of local raw material suggests possible to make a detailed analysis that the Hagios Nikolaos cave was pri using different archaeological meth marily used for making and repairing ods; raw material studies, flint technol tools and not simply for discarding ogy, chaine operatoire, flint typology, them. It also shows a diverse assem tool types and settlement models. blage used for a wider range of activi Prehistoric sites with chronologically ties, perhaps reflecting the day-to-day mixed material are often being rejected maintenance activities of people living from the archaeological statistics and in the cave. sometimes do not have the archaeolog ical value for an excavation, but they Concluding remarks can still provide us with some new knowledge. The results from the The analysis of the lithic material gave Hagios Nikolaos material prove that, the following results: local use of flint with a minimum of survey and dry resources and four different flint tech sieving, it is possible to collect, com niques, especially used for a systemat pare and analyse a rather large number ic blade production. Furthermore it is of lithics. This gives ideal powers of argued that there is a possible typolog observations, when dealing with a sur ical dating ofthe lithic material in two veying strategy with a lot ofrock shel phases: Early to Middle Epigravettian, ters or open-air sites, containing only and between the Late Epigravettian the lithic material.

257 References

E. Adam Franchthi 5 M. Mussi A Technological and C. Perles, Les industries Earliest Italy. An Overview of TypologicalAnalysis ofUpper lithiques taillees de Franchthi the Italian Palaeolithic and Palaeolithic Stone Industries of (Argolide, Grece) II: Mesolithic (New York 2001) Epirus, Northwestern Greece, Les industries du mesolithique BAR 512 (Oxford 1989) et du neolithique initial. C. Perles Excavations at Franchthi The Early Neolithic in Greece G.N. Bailey et al. Cave Greece 5 (Bloomington (Cambridge 2001) 'Asprochaliko and Kastritsa: 1990) Further Investigations ofthe M.L. Inizan et al. Palaeolithic Settlement and E- Technology and Terminology of conomy in Epirus (North-West Knapped Stone, Prehistoire de Greece)', Proceedings ofthe la pierre taillee 5 Prehistoric Society 49 (1983), (Paris 1999) 15-42

Klithi 1 BSA Studies 3 G.N. Bailey (ed.), Klithi. Palae G. Bailey et al. (eds.), The olithic Settlement and Quater Palaeolithic Archaeology of nary Environments in Northwest Greece and Adjacent Areas: Greece 1, Excavations and In- Proceedings ofthe ICOPAG tra-Site Analysis at Klithi (Cam Conference, Ioannina, Septem bridge 1997) ber 1994, British School at A- thens Studies 3 (1999) Klithi 2 G.N. Bailey (ed.), Klithi. Palae olithic Settlement and Quater P.J. Demoule and C. Pedes nary Environments in Northwest 'The Greek Neolithic, a New Greece 2, Klithi in its Local and Review', Journal of World Pre Regional Setting (Cambridge history 7 (1993), 355-416 1997)

F. Djindjian et al. M. Koumouzelis et al. Le paleolithique superieur en 'The Early Upper Palaeolithic Europe (Paris 1999) in Greece: The Excavations in the Klisoura Cave', Journal of Franchthi 3 Archaeological Science (2001), C. Perles, Les industries 515-39 lithiques taillees de Franchthi (Argolide, Grece) I: Presenta G. Kourtessi-Philippakis tion generale et industries Lepaleolithique de la Grece paleolithiques. Excavations at continental (Paris 1986) Franchthi Cave Greece 3 (Bloomington 1987)

258

A Prehistoric Settlement at Chania-Gavrolimni in Aetolia1

Efi Saranti Preparations, in late 2000, for the con both instances the prehistoric remains struction of a new town hall for the were found beneath later buildings. municipality of Chalkeia, at Gavro- Fragmentary remains have also been limni in Aetolia, led to the discovery excavated at Naupaktos.4 With such of significant remains of a prehistoric meagre evidence from Aetolia general settlement. The ensuing rescue exca ly, it is thought important to give a pre vation is still in progress at the time of liminary report of the excavation be writing. It has revealed a fairly exten fore its completion. sive and well preserved prehistoric settlement, flourishing in the Middle The prehistoric settlement at Chania and Late Helladic periods. is situated on the slopes of a low hill, referred to by the locals as Tapia (= The most interesting fact is that, with bastion), which is a foothill of the the exception of limited damage in the Kaliakouda mountain (height 250 upper levels from natural erosion and m). This place name is descriptive of modern activity, the settlement, which the site, which is naturally protected consists exclusively of prehistoric re and overlooks the only - in former mains, is basically undisturbed. It is days - available passage to the north one of very few settlements of MH-LH west, around the mountain masses of date in this area of Western Greece. Klokova and Varassova, as well as to Evidence elsewhere is restricted to the Gavrolimni valley, through which Thermos2 and Kato Vasiliki,3 (the hill communication with the northern in of Ag. Triadha, ancient Chalkis): in terior is possible (Fig.l).

11would like to express my warmest thanks to Mr. Ioannis Moschos for his help with this paper. The archaeologist Miss P. Staikou helped with much success in the early stages of the excavation. The excavation plan was drawn by Mr. I. Theocharopoulos. The first excavations at Thermos created much confusion about the dates of the buildings beneath and around the temple of Apollo, and the chronology of the MH period is still uncertain. See K. Romaios, 'Ex tou JiQo'iaxoQixoi) 0£Q(iou', ADelt 1 (1915), 224-79 and Id., "Eoeiwxi £V 08quxd\ ADelt2(1916), 179-89; R. Hope-Simpson, Mycenaean Greece, D75, 97; I. Papapostolou, 'Zn,Tr|uaTa xcov Meycxqq)v A xai B tou 08Q[iou', AE 1990, 191-200. 3FPR; SPR; and in this volume. 4G. Alexopoulou, 'Naupaktos', ADelt 49 (1994), Bl, 243-44; L. Kolonas and F. Saranti, 'Naupaktos', ADelt 49 (1994), Bl, 245; F. Saranti, 'H ccQxaia NcajJtaxxoc; xai r\ evgrnegy) 3T8QLOxr| xr|5. TojtOYQOKJHxd Se6ouiva xai JiQoa^axeg epeuvec;', Proceedings of the First Archaeological CongressofSouth-Western Greece, Patras 1996 (in press).

229 Fig. 1. View from the northeastern part ofTapia.

The valley of Gavrolimni is a fertile Gavrolimni. Signs of prehistoric habi area, watered by many streams and tation are even more extensive to the cultivated mainly with crops and north, reaching as far as the lowest olive-trees. Its name is interesting as slopes of Kaliakouda.7 its origin is explained either as raSgoXi/uvr] ('fish lake'), or Koljtqo- The hill of Tapia measures approxi Xlfivr\ ('boar lake') or Ka6ovgo- mately 110 m in height and is situated Aljuvn ('crab lake') and probably re north-west of the modern village of flects the memory of the area being Chania, at a distance of about three once a lake or swamp.5 kilometres from the sea and the bay of Kato Vasiliki. The slopes ofthe hill are As revealed by an investigation in the covered with pine-trees. The top is ob area prehistoric habitation is not re long and flat and carries no signs ofar stricted to the hill ofTapia, but extends tificial fortification, as far as is possible further eastwards. The construction of to discern through the thick vegetation a modern road several years ago re of shrubs. Buildings connected with vealed traces of buildings which are prehistoric pottery are traced on the visible on the south slope ofa low hill, SW part ofthe hill, approximately five about 350 m east ofTapia,6 east of the to six metres from the top. The entire road leading to the village of southern face of Tapia has been used

5The name seems to date from the Middle Ages: S.K. Alexandropoulou, Nona AiTcoAia, to odixo bixivo (Athens 1993), 60; G.E. Rapti, fJeQtyga^r] EJiagxiag Navjcaxtiag (New York 1955), 65-67. 6It is nameless according to the locals. 7Abundant MH pottery and traces of buildings are visible near the NE slope of Tapia, on the east side of a small stream that runs N-S, across the hill.

230 north passed at the level of the upper plateau. The famous inn,8 which gave its name to the village (Chania = inns), was actually located by that road, in line with our excavation site.

The excavated buildings are situated on the south slope of the hill. At this point, approx. 40 metres above sea level, the ground was levelled in the 1970s to form a terrace for a school, which was never built. Fortunately, the work caused only minor damages to the archaeological remains.

The initial aim of the excavation was to determine the extent and state of preservation of the settlement, on the site where the town hall was to be built, an area of about 500 m2. It was soon realised that the archaeological remains covered almost the entire site. It also became evident that they con tinue north, west and east, whereas it has yet to be determined whether they continued further to the south as well.

Fig. 2. Burial II. for quarrying and a great mass of Parts of at least four prehistoric build chipped stones is accumulated at the ings (A, B, C, D), evidently from two front. major phases, have so far been uncov ered. The building material is homoge As is evident from building remains nous, consisting of local sandstone and and pottery on the surface, the prehis shale (flysch). Three intra-mural burials toric settlement of Chania covered of infants were also excavated, two of primarily the lowest parts of the which were located in what was appar south-eastern and eastern slopes of ently an open area north of Building A the hill of Tapia. The layout of the (Fig. 2). settlement shows that the choice of the site depended mostly on the con The settlement was built over differ trol of the passage. A fact worth ent levels following the natural slope, noticing is that, prior to the construc the highest traced north of our exca tion of the national road at the foot of vation, at four to six metres above the the hill of Tapia, the road leading terrace level. Another level is found

Chani Skorda/Aaw Uxogdd.

231 Fig.3. Excavation plan Chania - Gavrolimni Aetolia - Eastern part of the excavation.

on the north side of the excavated of the ground. Extensive layers of area, at approximately half a metre fallen debris from this level were below the terrace level, where there found covering the north-west side of are visible traces of a building, or Building A. buildings, constructed with large, roughly rectangular blocks. The A short description of Buildings A and building(s) were obviously destroyed B will be attempted in spite of the ex at an early date by the natural erosion cavation being still in progress. They

232 Fig. 4. Building A from NW.

belong to the eastern part of the exca phase. Two possible entrances have vated area, where the excavation has been traced, both on the long sides of been concentrated in its second phase. the building (Figs 3,4).

Building A is a large rectangular (?) Building B is oriented exactly E-W, building resting on top of Building B without inclination, at a depth of 2,03 with an earth fill between them rang - 2,67 below the terrace level. Its east ing between 0,05 and 0,40 m. It is end appears to be apsidal, but it has yet found at a depth of 1,88 to 2,30 m un to be ascertained, since it lies beneath der the terrace level. It is oriented NW- the eastern room of Building A. Its SE and is of large dimensions, meas west end was most probably rectilin uring 15,40 by 5,50 m. Its east end ex ear, but is fragmentarily preserved. tends beyond the original limits of the The inner width ofthe building is 4,20 excavation grid and remains as yet un- m and its maximum internal length excavated. Rough flat stones, usually 8,80 m. Its walls were built of large set in two main rows, with smaller flatfish stones at the bottom and small stones filling the intervals, constitute and medium sized unworked stones, its walls. Mud is used as binding ma built without any order, on top. The terial. The external walls have a width maximum preserved height of the of 0,60-0,80 m, the internal being ap- walls is 0,37 m and their width is 0,60 prox. 0,45 m wide. Because of the m (Fig 5). ground inclination towards the south west, the walls of these sides are more On the south side of the excavated strongly built. The interior of the area we have uncovered parts of a building is divided into five rooms, but largely destroyed wall, constructed as until fully excavated it is uncertain a line of large rough sandstone blocks whether they all belong to the same on a layer of small unworked stones.

233 Fig. 5. Building B under building A, from NW.

It is preserved to a length of 9,00 m Kitchen ware of semi-coarse clay and seems to be a terrace-wall, sup forms the largest group found at all porting this level of the settlement. levels. The clay colour ranges from This is also indicated by the fact that pink to reddish yellow and contains the fill south of the wall is devoid of small white inclusions. Small voids sherds. Excavation below this level are visible on the surface. The clay is has as yet only begun, but a destruc unevenly fired, usually with a greyish tion level with large pieces of charred core. The ware is used for shapes of wood and burnt clay has been reached varying thickness, the most character at a depth of 2,65 to 3,28 under the istic being storage pithoi and large terrace level. pithoid jars with flat bases and trian gular handles. Six main categories of pottery can be discerned in the ceramic material con Matt-painted ware appears in signifi nected with the above remains. cant amounts in the level of Building B. The fabric is semi-coarse and cov Coarse handmade ware is present in ered with a matt greyish slip, often significant amounts at all excavated burnished. Very small silver mica in levels. It is unevenly fired, the surface clusions are sometimes noticeable. colour ranging from red to orange, and The decoration consists mainly of lin the core usually being grey or black. ear patterns, such as simple bands and The clay contains small stones, visible continuous solid zigzags.9 on the surface. At the level of Building B most of this pottery has a distinctly Grey minyan ware also makes its ap smoothed surface. pearance in the level of Building B,

9Clay colour 5YR 7,6, reddish yellow. Slip colour 10YR 8/2, very pale brown.

234 but in very limited quanitites. It is should, however, note that the excava characterised by the absence of well- tion, as well as the conservation of the smoothed surfaces and is reminiscent material, is still at an early stage. The ofArgive Minyan ware. sherds are often poorly preserved and the shapes are not easily reconstruct Fine ware. It is well fired and usually ed. There is, moreover, no established has a reddish yellow colour.10 The clay chronological system for this area. has a soft surface and is mostly well- tempered. Slip is not easily discerned. With these facts in mind, Buildings A and B seem to be dated at the end of Polychrome mainland ware appears in the MH period and the transition to small quantities in the level ofBuilding LH, whereas earlier habitation at the B. The colours are matt black and bur site cannot yet be excluded. We should nished reddish brown. The clay is also note that the entire fill above medium-tempered, containing black Building A had been removed before and white inclusions and small voids, excavation. On the contrary, in the as well as small amounts of silver mi western part of the excavated area, ca." where buildings appear at a greater depth, decorated pottery has been Finds, other than pottery, consist of found in the upper fill dating to the LH numerous flint chips and a couple of III (C?) period, at the time of the last flint cores, various stone implements habitation at the site. and tools (axes, grinders, whetstones, "lids"), clay spindle whorls and a large The excavation of the settlement of quantity of sea-shells and animal Chania can certainly contribute much bones found near and in the floor lev to our knowledge ofprehistoric habita els. A small number of boar tusks has tion in this part of western Greece, as also been found. Bone tools are rare well as ofAetolia in general, known to and metal is completely absent, except have been inhabited from an early for a lead clamp used on a repaired date. Traces of habitation in the area of sherd. Varassova dating from the Final Neolithic/Chalcolithic to the end of The preliminary study of the above Bronze Age have long been known material has made a few conclusions from Pangali,12 Ag. Triadha,13 Kryo- possible regarding the phases repre neri,14 and Ano Vasiliki.15 The settle sented by the excavated remains. We ment of Chania and the survey of its

10 Claycolour ranges between 7,5 YR 8/6, 7/6, reddish yellow, and 7,5 YR 8,4, pink. 11 Claycolour 7,5 YR 7/6, reddish yellow. Core colour 5YR 6/4, reddish yellow. Slip colour 5YR 6/4, reddish yellow. 12 S. Benton., 'The Ionian Islands', BSA 32 (1931-2), 239; Th. Mavridis and H. Alisoy, FPR, 272-79; M. Gazis, FPR, 280; T. Mavridis, SPR, 277-89. 13 E. Mastrokostas, 'Xcdxig', ADelt 22 (1967), B2, 320; Id., BCH 94 (1970), II, 1031. 14 S. Benton, 'The Ionian Islands', BSA 32 (1931-2), 239. 15 K Davaras, 'Neoi 6i;rrloi jre^exeig ex xng YY aQXaio^°YLXTl? Jteoidpeoeiac;', AAA III 3 (1970), 311-12; E. Mastrokostas, 'XcxXxig', ADelt 22 (1967), B2, 320.

235 Fig. 6. View of the hill of Hagia Triada from the hill of Tapia.

environment have proved that prehis logical system for this period in toric habitation in the area is far more Aetolia and to determine the relation extensive than formerly thought. The ship between this inland settlement and full excavation of the settlement will the one partially uncovered on the hill hopefully help to establish a chrono of Hagia Triada near the sea (Fig. 6).

236 Shellfish from the Excavations at Aetolian Chalkis

Kaj strand Introduction better-known seas, but this is far from Petersen' the case. Our former knowledge of the It has been argued that, among species Eastern Mediterranean is particularly in the northern Adriatic Sea, we should faulty. Zenetos has, however, recently also find temperate species character concluded that the marine bivalvia istic ofthe Eemian, or last interglacial, fauna is as rich in the Eastern period in northern waters (130,000- Mediterranean as it is in the Western,2 115,000 BP). The faunal assemblage which has been confirmed for the en is, however, much more varied than tire molluscan fauna by Delatmotte that known in northern waters in the and Vardala.3 hypsithermal interval of the Eemian period. Hence, when studying the fauna in the Ionian Sea, and the Gulf of Patras in Following the development ofa glacial particular, we are looking at a more period it has been shown that the varied fauna than what we find in Mediterranean indeed did have some northern waters. Therefore it has been northern species, such as Arctica is- a challenge for me to study the mol landica, Pecten islandicus and luscan material from the excavations Buccinum undatum. The Medi at Chalkis in Aetolia. terranean has been regarded as a faunal unit with a more varied fauna than that An overview of the geology of the of the northern ocean. However, fol coast at Chalkis was given in the lowing Ekman (1967), it has become Second Preliminary Report (2000). clear that the Straits of Gibraltar do not Here, the focus will be on the molluscs represent an important zoogeographi- sampled from the outset of the project cal boundary, and hence the in 1995 until 2000. In the final publi Mediterranean should not be regarded cation, material from the final excava as a distinct unit. tion season in 2001 will also be in cluded, and the whole body of materi Ekman goes on to say that the al will, as far as possible, be put in its Mediterranean should be one of the stratigraphical context.

1The Geological Survey ofDenmark and Greenland. 2A. Zenetos, Marine Bivalve Fauna in Greek waters. Variety, Distribution, Ecology and Zoogeography, Abstracts 12th International Malacol. Congress (Vigo 1995), 289-90. 1M. Delamotte and E. Vardala-Theodorou, Shells from the Greek Seas (Athens 1994).

215 During the six seasons covered in the Fig. 1. A cockle with an an thropogenic impact. Notice present report a total of 537 bags con the different look of the hole taining shellfish were recorded.4 The when compared with that of a predatory gastropod on number of specimens - and species - Figs. 3 and 11. varies greatly between the units.

The over-all dominating species is the cockle, which occasionally has an an thropogenic implication, when perfo rated by a hole and used as a necklace (Fig. 1). Among the 701 cockleshells The species in order recovered in the year 2000 season, 16 offrequency were pierced in this way. Other species, described in the list of mol The following descriptions are listed luscs below, may also have been used in the order of frequency of the as charms at times. species, with the intention of present-

Frequency Molluscan species epi infauna 448 Cerastoderma glaucum x shallow Acanthocandia tuberculata x shallow 75 Cerithium vulgatum 48 Hexaples trunculus 37 Pinna nobilis x shallow 48 Patella spp. 14 Spondylus gaederopus 13 Tonna galea 11 Tapes decussatus x deeper 11 Ocenebra erinaceus 10 Mactra glauca x deeper 8 Venus verricosa x shallow 7 Ostrea edulis 11 Trochidae 5 Glycymeris glycymeris x shallow 4 Pecten jacobaeus 3 Bolinus brandaris 3 Conus ventricosus 2 Callista chione x shallow 2 Aporrhais pespelicani x shallow 1 Area noae 1 Dosinia lupinus x deeper 1 Luria lurida

Table 1. Frequency

4 1995: 22; 1996: 46; 1997: 105; 1998: 101; 1999: 129; 2000: 134.

216 Fig. 2. The fisherman's wife crushing the shell of the com mon cerith in order to get the hermit crab used for bait.

'""

ing the habitat of each species, as well not really desirable for food, are as its culinary use, based on its role in found together with cockles in ar modern gastronomy. chaeological deposits. In Danish kitchen middens, oysters seem to Table 1 presents all the 22 marine have been the preferred fare, and species found, with an indication cockles are predominant only in peri whether it is an epifaunal element, i.e. ods when oysters are scarce.5 attached to a hard substrate, or an in- faunal element, burrowing in the Cerithium vulgatum (Brugiere, 1792) is sandy or muddy seabed. not subject to regular harvesting and rarely seen on the market. It is, however, Cerastoderma glaucum (Poiret, often used as bait for fishing. It ranges 1789). The Mediterranean cockle, in second among the species from the ex Greek xuSobvi (kydoni), lives in shal cavations at Chalkis. The shells are low water on sandy or muddy seabed. found on the rocky shore and on sandy Here, we may also include and muddy seabed. On the sandy seabed Acanthocardia tuberculata (L., offChalkis the shell is often occupied by 1758), which is more solid and thick- the hermit crab (Paguristes oculatus), shelled, but which lives on the same one of nearly 30 endemic species of de type of seabed and depth. When har capod crustaceans in the Mediterranean vested today, cockles are 'scooped and Atlantic region (Ekman 1967). The up' from the seabed. That was pre hermit crab from shells of Cerithium sumably the case in Antiquity too, vulgatum is used as bait at modern-day since many other species of molluscs, Kato Vasiliki (Fig. 2).

K..S. Petersen, Journal ofDanish Archaeology 5, 77-84.

217 The large number ofthis species found Fig. 3. Shells of the common cerith. To the left part of the in the archaeological deposits can, in hermit crab is seen otherwise my view, be explained by it uninten well protected in the shell - the two other shells both carry tionally being caught when the seabed the hole made by a predatory was trawled for cockles. Occasionally gastropod. the specimens have been attacked by a predatory gastropod boring the shell -c - - (Fig. 3).

Hexaples trunculus (L., 1758), in Greek JtOQ(j)iJQa, found from one to 100 metres ofdepth on muddy seabed. It is very common but the meat is infe rior to that of the purple dye murex. The species was used in Antiquity to produce a purple dye of lesser quality (Poppe and Goto 1991). The shells are sometimes inhabited by the hermit crab (Fig. 4), and, as with the previous species, its high frequency in archaeo logical deposits may be explained as

an accidental by-product of cockle Fig. 4. Hexaples trunculus can also be invaded by the hermit crab and be seen moving over fishing. the sandy seabed.

Pinna nobilis (L., 1758), or fan mus sel,6 lives in sand, mud or gravel, from the low tide mark to a depth of 60 me tres. Off the coast of Kato Vasiliki, Pinna is found in the fields of Posidonia oceanica (Fig. 5). Large specimens, up to 90 cm tall, are found at greater depth further out in the sea. The fourth most frequent species from the excavations is still used for its meat, as implied by its French name, jambonneau de mer. Its long and re sistant byssus (Fig. 6) was used in the textile industry as late as the nine teenth century (Poppe and Goto 1993). Fig. 5. Pinna nobilis "standing" in the field of Posidonia oceanica on 3-4 m of depth off Kato Patella, called JiorceAiSa in Greek, is Vasiliki. Fibrous sea-balls, formed of the shaggy remains of old leaves (T. Harris, The National History ofthe Mediterranean (London 1982), accumulate near the top of the shore represented by the spp. P. coerulae (L., as seen in the little bay at Pangali.

6K.S. Petersen, in SPR, fig. 52.

218 Fig. 6. The fan mussels from thorny oysters, was formerly very off the coast of Kato Vasiliki with part of the resistant common in the Mediterranean. For an byssus seen on the edge of the unknown reason, the colonies suffered shell to the left. The byssus is used anchoring the animal in a sharp decline in the 1980s (Poppe the sandy sea bottom. and Goto 1993). Occurring in some ar chaeological samples from Chalkis, its edible flesh must, as today, have been considered a delicacy in Antiquity (Fig. 9). The species belongs to the epifaunal on hard substrates and is in fra-littoral from six to 50 metres of 1758) and P. rustica (L., 1758), both depth, cemented to the rock with the common in the Mediterranean as epi- lower (right) valve. It is evident from faunal on rocky shores. The former is the archaeological specimens that found from the upper tidal mark to a these shells were removed deliberately few metres ofdepth, whereas the latter from their growing position, and not is restricted to the intertidal zone, accidentally with other species from which in the Mediterranean is very the level sandy seabed. narrow, measured in tens of centime tres.7 Specimens of both species are Tonna galea (L., 1758) lives at a depth rather common in the archaeological between 20 and 80 metres, on all kinds deposits, and both are found today on of seabed (Poppe and Goto 1991). rocky shores in the area of Kato Here is a species, which was not Vasiliki (figs 7-8). caught accidentally with the cockle but must have been fished for in its Spondylus gaederopus L., 1758. The own right. This large predatory mol family of the Spondilydae, also called lusc has been eaten around the Gulf of

Fig. 7. To the east of Hagia Triada the sandstone from the Flysh forms the foothill of Klokova forming the firm substrate for the epifauna, here dominated by Patella spp.

*!*»>

i mtr

7T. Harris, The National History ofthe Mediterranean (London 1982).

219 Patras until recently, but is now rarely Fig. 8. From the place shown on Fig. 7 a tiffin is easily seen at restaurants. gathered among the gregari ous limpets. One might also go for the Trochidae-topshells Tapes decussatus (L., 1758), in Greek which are very common here - one of the several mollusc species seen on Fig. 14 - but nearly all gastropods used as shellfish called a)

220 Fig. 11 The warty venus as the to a hard substrate on all types of name of Venus verrucosa runs in English has a sturdy shell seabed. Its popularity as a culinary where even a predatory gas tropod has given up as seen dish goes back to Antiquity8 with con by the unfinished hole in the notations of aphrodisiac qualities right valve. (Poppe and Goto 1993), and oyster fishing by diving is mentioned in Homer (//. XVI, 854-63):

And, like a diver, from the well-wrought car Headlong he plunged; and life forsook his limbs. Fig. 12. The hard substrate for O'er whom Patroclus thus with bitter jest: the oyster to fix upon might "Heaven! what agility! how deftly thrown be other specimens. Here the oyster themselves are covered That somersault! if only in the sea by fanworm (Polychaete) and Such feats he wrought, with him might few with post mortem borings of compete, Cliona celata - the boring Diving for oysters, if with such a plunge sponge. He left his boat, how rough soe'er the waves, As from his car he plunges to the ground: Troy can, it seems, accomplished tumblers boast. (trans. Edward, Earl of Derby, 1865)

The family Trochidae includes three species found at Chalkis: Monodonta turbinate {Born, 1780), Gibbuta divar icate (L., 1767) and Callistoma dredging. Being quite commonly zizyphinum (L., 1758). These play the found on the beach today and having a same role in the Mediterranean as do sturdy shell, which should be apt for the Littorinidae, living on rocky good preservation in soil, it is some shores, in the northern Atlantic. The what peculiar that it should be so rare frequency among the samples (Fig. among the species found during exca 13) was low, but these species are vation: only 8 specimens were regis common on the shore today (Fig. 14). tered (Fig. 11). Glycymeris glycymeris (L., 1758), in Ostrea edulis (L., 1758), the oyster - Greek xovxor| axL^aSa or 'thick in Greek ocrcoeo or OTQeiSi - is found shell', prefers sand and mud in shal all over the Mediterranean (Fig. 12). low water out to c. 80 metres from the The oysters live in shallow water fixed shore. In modern-day France it is cap-

8Fromthe studies of the shellfish from "Kokkenmoddinger" (Kitchen middens) at the Ertebolle settlement - locus classicus - in the Limfjord, Denmark it was concluded that the oyster only was eaten as a pleasant change in their diet: K. Strand Petersen, in L.-K. Konigsson (ed.) Nordic Late Quaternary Biology and Ecology, Stria 24, 221-26. From this pe riod - the Ertebolle lasting 1500 years from 5400 to 3900 B.C. - we cannot tell what kind of feeling thosepeople had but it has been suggested "that the chemical content of theseshells, in particular iron compounds, zinc, iodine and salts, rather than just their food value, was the in direct reason for the accumulation of the large shell middens": U. Mohl, 'Aggersund-boplad- sen zoologisk belyst', Kuml - Arbogfor Jysk Arkceologisk Selskab 1978, 57-75.

221 tured, like cockles, by dragnet. The Fig. 13. Here to the left the toothed winkle (Monodonta bittersweet clam is regarded as a deli turbinata) from the recent cacy in Europe, and its French name, shores of the Patras Bay is compared with one from the amande de mer, hints at its delicious excavations where the parallel taste when eaten cold. The low fre reddish transverse stripes have vanished. quency among the samples shows that it was not part of the staple diet in Antiquity, although it is common in the Mediterranean.9

PectenJacobaeus (L., 1758), in French Coquille Saint-Jacques de la Mediterranee - to be distinguished from de lAtlantique, which is Pecten maximus (L., 1758) - the shell famous as the pilgrim shell from Santiago de Compostela.10. Both are une des vedettes des menus de la mer (Quero and Vayne 1998, p. 100). They are found on sand and gravel seabeds at a depth between 25 and 250 metres (Poppe and Goto 1993). They are not found in shallow water, which may ex plain the paucity among the archaeo logical samples ofthis species, so pop ular today. implies that it was rather used for Fig. 14. Two toothed winkle "grassing" on the rocks in the food. The meat has a fine taste, and it surf zone. Bolinus brandaris (L., 1758) is com is the most popular gastropod fished in mon in the Mediterranean on sand and the Golfe du Lion, where it is served in mud seabeds between one and 200 its conspicuous shell and called escar- metres of depth (Fig. 15). Known as got de mer. the purple dye murex, it was the basic product in the manufacture of purple Conus ventricosus (Gmelin, 1791) is dye from Phoenician times (Poppe and found in shallow calm waters on rocky Goto 1991). As late as the time of shores (Poppe and Goto 1991, pi. 36, Christ we hear about Lydia "a seller of figs 5-7). This is the only species o. purple goods..." (Acts 16, 14). A huge Conus in the Mediterranean and there number ofrotten molluscs were neces fore also called Conus mediterraneus sary for the production of dye," so the (Hwass in Bruguiere, 1792). This frequency of the species at Chalkis species has not been caught for food

9S.P. Dance, Shells (London 1992). 10 In this way Pecten maximus is the one to be taken off Santiago de Compostela and given to the pilgrims on their stay in town on their pilgrimage to St. Jacques! 11 D.S. Reese, 'Palaikastro Shells and Bronze-Age Purple-dye Production in the Mediterranean Basin, BSA 82 (1987), 201-06.

222 Fig. 15. In the Levant, where Goto 1991). As in the case of the Phoenicians once lived, large piles of the purple dye Cerithium vulgatum, it may be pre murex are found along the sumed that shells are brought near coast - like old "Kokkenmoddinger", they are the shore by the hermit crab, a phe however left there as remnants nomenon also observed in of the purple dye production invented by the Phoenicians. Denmark.12 The shells, found on the shore or collected accidentally with the cockles, might have served as charms because of the shape, which "has caught the attention of fisher men and naturalists since ancient

Fig. 16. The venus shell times" (Poppe and Goto 1991, p. 116, (Callista chione) is character pi. 16, figs 8-11). This mollusc was istic by its horny layer cover ing the shell chestnut - brown described by Aristotle and Pliny,13 and shining which once made and it was appreciated as an hors our oldest grandchild saying d'oeuvre boiled in a lemon sauce. when asking him what to bring home this summer from The quantities registered from the ex Greece - the shiny mollusc, he cavations are however not large said. enough to make it likely to have been used for food at ancient Chalkis.

but, thanks to its varied shape and Area noae (L., 1758; Fig. 17) lives at colouration, it may have been used as tached by the byssus on rocks and oth charms. er hard substrata from the tidal zone to over 100 metres of depth, where direct Callista chione (L., 1758): the Greek sunlight is virtually absent (Poppe and name is \ieyahf\ ayy^o.ba, 'large Goto 1993). According to Lellak and shell'. This shiny mollusc lives in fine Cepicka (1975, p. 154), "les sujets and clean sand as a shallow infaunal el adultes preferent la zone superieure, ement (Fig. 16). Today it is common off bien eclairee, de la bande littorale"! Kato Vasiliki from shallow water to Although it is among the least frequent great depths. It is caught commercially species among the samples from by trawling and diving. Empty shells Chalkis, this mollusc, which is eaten are often found in the traces ofthe hunt raw, is now common along the shores ing octopus (Octopus vulgaris: Cuvier, of the Gulf of Patras. 1797). Dosinia lupinus (L., 1758) lives Aporrhais pespelicani (L., 1758) is deeply buried in pure sand and fine found from ten to 180 metres of gravel from the tidal zone to 200 me depth on muddy sand (Poppe and tres ofdepth. That fact may explain its

12 K.S. Petersen, Geology ofDenmark and Greenland Survey Bulletin 3 (2004). 13 Etymology: aporrhais (Aristotle), "spout-shell", from aporrheo, to flow away. It was only at the time of Linnaeus (1707-78) that the binomial system of scientific nomenclature was estab lished.

223 Fig. 17. Shells and other fau nal remains found along shore near to Hagia Triada: 1) Area noae, 2) Venus verricosa, 3) Cerithium vulgatum, 4) Hexaples trunculus, 5) Monodonta turbinata, 6) Sepia sp. and others.

low frequency among the samples but have been distinguished: Zebrina de- today it is used as seafood. trita (Miiller, 1774), Helicigona lapi- cida (L., 1758), Cepaea hortensis Luria lurida (L., 1758), with a biomet- (Muller, 1774), and Rumina decollata ric range from one to 60 metres of (L., 1758). depth, lives under stones in a sandy biotope. The species is one ofthe clas Nineteen specimens of Rumina decol sic Mediterranean shells (Poppe and lata were found. As implied by the Goto 1991), but because of its small name, it is characteristic by the deliber size, it was not used for food but rather ate truncation where the earlier formed as charms. It is among the Cypraeidae whorls are lost.14 A common Mediter whose shells are so beautiful that they ranean species it stills occurs in the were used in the Orient as money, northern coastal region of the Gulf of called 'kaurV (Lellak and Cepicka Patras. 1975). Helix pomatia (L., 1758) Fig. 18, with Finds ofterrestrial gastropods were al a living range in Central and South so recorded from the excavations, eastern Europe has been cultivated for among which the following species food since Roman times.15

14 M.P Kerney and R.A.D. Cameron, A Field Guide to the Land Snails ofBritain and North- West Europe (London 1979). 15 Helix pomatia was for instance introduced also to Denmark in the Middle Ages and was popular among the catholic as Lenten fare.

224 Fig. 18. The distribution of Helix pomatia is very much carrying evidence of boring by a affected by its culinary reputa predatory gastropod. The shells would tion. The occurrences in Denmark are regarded as in therefore have been picked up empty troduced by monks in the or occupied by the hermit crab, as of Middle Ages. ten is seen today (figs 2-3).

The many bivalves among the sam ples have one feature in common: as shallow to deeper infaunal species, their habitat is the sandy seabed (table 1). Venus verrucosa, Tapes decus sates, Pinna nobilis and Mactra coral- Una are all known as seafood. Among the bivalves as whole, thirteen fami lies covering 50 species from the Atlantic and the Mediterranean are recorded as edible. So it is more a cus The molluscan tomary attitude among the natives assemblages which determines what is accepted as seafood. All in all 22 species of marine mol luscs were recorded, as well as five Pinna nobilis, with a height which terrestrial gastropods, including Helix can reach nearly a metre, could be and Cepaea, which are still sold in called the king of shell fauna.16 The Greek markets. samples from the excavations are, however, all fragmentary, due to the The by far most numerous molluscs fragility of the shell when taken out from the Chalkis excavations is the of the water. Large specimens can cockle, found in great quantities in still be found at depths of four to sev shallow waters on sandy seabed. A en metres in the Gulf of Patras. shallow infaunal element, it was most probably dug out or trawled by net, as The Mytilidae are missing from the bi it is done today in southern Portugal, valves found at Chalkis. Mussels are where the local fishermen harvest otherwise very common and can form shellfish from the sandy seabed with 'bancs de moules' or grow on fishing dragnets. Such a method of harvesting tackles off Kato Vasiliki (Fig. 19). In would also explain the presence Greek, the Mytilus galloprovincialis among the samples of species, which (Lamarck, 1819) is called [ru&i cannot be regarded as comestibles, viz. KoovcrcavTivoimoXeooc; ('Constant small gastropods like Columbellidae inople mussel') or [icruoouudi ('black and Cerithium vulgatum. The latter is mussel'). An explanation as to why it is the most common gastropod species not found in the assemblage from from the excavations, its shell often Chalkis may be the fragility ofthe shell

K.S. Petersen, in SPR, fig. 52.

225 Fig. 19. The mussel is found also on the fishermen's sea trout ponds in great quantities, but it has not been recorded from the excavations.

when buried in soil. Another possible species seem to play the same role as reason is that it was not deemed suit do the Littorinidae in Northern able as food in ancient Chalkis. As it Europe. grows fixed to the hard substratum by the byssus, it would not be taken acci Among gastropods, the impressive dentally. Tonna galea, with a diameter of up to 25 cm, is an unpitiable predator, at Among the molluscs, we have species tacking starfish, echinoids and large known from rocky shores, such as bivalves. Not very common, it can still Patella spp. and Trochidae with gen be found on the menu. Living at era like Monodonta and Gibbula. The greater depths than most of the other latter is found in abundance along the species mentioned, it was caught by shore today. In some way these the population of ancient Chalkis.17

17 K.S. Petersen, in SPR, fig. 53.21

226 References

S. Ekman Zoogeography ofthe Sea (London 1967)

J. Lellak and A. Cepicka Coquillages. Atlas illustre (Paris 1975)

K.S. Petersen, 'Geological Investigations in the Area of Haghia Triadha', in SPR, Proceedings ofthe Danish Institute atAthens 3 (2000), 269-75

G.T. Poppe and V Goto European SeasheIIs I (Wiesbaden 1991)

G.T. Poppe and V. Goto European Seashells II (Wiesbaden 1993)

J.-C. Quero and J.-J. Vayne Lesfruits de la mer et plantes marines des pechesfrancaises (Paris 1998)

227

The Coins from Hagia Triada II

Georgia Z. Ninety-four coins came to light during League, Elis, Argos, Hermione and the Alexopoulou the excavations at Kato Vasiliki in Arcadian cities, Megalopolis, Cleitor 1999 and 2000.2 The catalogue is and Pheneus), Ionia (Miletus), and arranged according to the geographi Carthage. Apart from Euboean Chalkis cal order used in numismatic publica and Sicyon, none of the other areas was tions. The specimens of each group, represented in the excavations of the which come from the same mint, are previous period. We can observe that the catalogued in increasing order accord coins from Thessaly and the Pelopon ing to diameter and weight. Nineteen nese were circulated simultaneously. coins are in such a state of deteriora The specimens from those areas are lim tion as to be almost useless for study. ited (one to four patterns) with the ex ception of Corinth (eight patterns). As can be seen from Table I, all the coins are bronze except from four silver The Aetolian League is once again, specimens. Once again Aetolian League and quite naturally, the most frequent coinage prevails with a total of 37 ly represented mint. It has been possi pieces.3 There are also coins from ble to identify the following numis Amphipolis (Macedonia), from Thes- matic types: saly (Pelasgiotis: Attrax, Crannon, Eurea, and Phthiotis: Halus, Lamia, 1. Atalanta or Aetolia wearing kausia Peumata), from Epirus (Ambracia) and r. / Calydonian boar at bay r., above Acarnania (Argos Amphilochicum). AITQ (1 pattern) [cat. no. 10]. Olivier There are also issues from Locri Opuntii Picard dates this type at the end of the Phocis, with its federal coinage, and fourth century B.C.4 from Euboea (Chalkis, Eretria, Histi- aea). Specimens were found even from 2. Atalanta or Aetolia wearing kausia r. / the Peloponnese (Corinth, Phlius, spearhead 1., above AITQ, beneath Sicyon, Aegeira, Dyme, the Achaean AQN, on r. in the field grapes (7 pat-

1I wish to express my thanks to Dr. Lazaros Kolonas, General Director of Antiquities, who au thorized me to make this preliminary presentation of the coins found at Kato Vasiliki. I would also like to thank the conservator of the 6th Ephoreia, Despina Marinopoulou, and the Surveyor, Charalambos Marinopoulos, who produced the map. I also thank the philologist Sevi Gaitanidou who was kind enough to check the English text. 2In the present preliminary report are also included nineteen coins found duringthe excava tions of the previous year (1998), which were not discussed in SPR, since they had not been conserved at the time. 1Duringthe excavations of 1997 a total of 31 coins came to light, of whichtwenty wereof the Aetolian League, cf G. Z. Alexopoulou in SPR, 259-67. 4 Picard 1984,284.

189 terns) [nos. 11-17]. This type was, as the F. Scheu suggests the year 205 B.C. as previous type, issued at the end of the the beginning ofthe bronze issues ofthe fourth century B.C. according to Picard.5 Aetolian League, when Aetolia, exhaust ed by the wars against Philip V of 3. Young male head with wreath r. / Macedonia, went through a financial cri spearhead and jawbone of boar r.; sis and stopped issuing silver coins.10 K. above AITQ, beneath AQN on 1., in Liampi proposes an earlier dating of the the field grapes (8 patterns) and some bronze emissions in the first half of the times in the middle the monogram third century B.C., and claims that the cir culation ofthe bronze issues was parallel with the silver ones."

(15 patterns) [nos. 18-39]. Picard, hav The types 1 to 4 were examined in the ing conducted a thorough research of previous report.12 Type 5 is the latest is the coins from the Corycian Cave, dat sue according to SNG Cop,{3 whereas it ed this Aetolian type from 300/290 to is neither referred to by Head,14 nor in 191 B.C.6 The specimens of this type the catalogue of the British Museum.15 It represent a majority (22 patterns) of is the only type, which has the magis the total (37) ofthe coins issued by the trate's name, while the majority of Aetolian League and are represented Aetolian League coins does not carry in great variety, as it can also be seen names ofmagistrates. The weight ofthe in the BMC7 and SNG Cop.8 coins examined here varies from 2.2 to 6.0 gr., with the exception of no. 10, 4. Head ofAthena r. / Heracles standing which weighs 1.5 gr. According to with club and lion's skin; on r. AITQ, Picard they are "chalka" and "double on 1. AQN (6 patterns) [nos. 40-45]. chalka".16 The common bronze coin of Picard dates this type in the early sec Corinth with Pegasus / Trident is distin ond century B.C.9 guished in Pegasus r. (7 patterns) or 1. (1 pattern), with diameter from 11 to 14 5. Atalanta ofAetolia wearing kausia r. / mm. The issues with Pegasus on r. are owl standing (1 pattern) [no.46]. rare and generally considered early.17

5 Picard 1984, 284. 6 Picard 1984,284. 7BMC, Thessaly to Aetolia, 198-99. 8SNG Cop, Aetolia: nos. 28-34. 9 Picard 1984,284. 10 F. Scheu, 'Coinage systems ofAetolia', NC 140 (1960), 50. " K. Liampi, 'On the cronology of the bronze coinages of the Aetolian League and its mem bers (spearhead and jawbone types)', Agxaioyvcooi- 9, 1995-96 (1998), 83-103. 12 cf G. Z. Alexopoulou, SPR, 259. 13 SNGCop,Aetolia: no. 41. To be more specific, according to the catalogue the attribution of the coin to the Aetolian League is doubted. 14 Head, HN, p. 334-35. 15 BMC, Thessaly to Aetolia, 194-200. 16 Picard 1984,285. 17 N. Bookidis and J.E. Fischer, 'The Sanctuary of Demeter and Kore on Acrocorinth, Appendix: Coins', Hesperia 43, (1974), 294.

190 VVtPHIPOilS

•ATRAX •CRANNON

LAMIA.

,CHALKJS

| r," AEGEIRA ' 4nvur * siryoN* CORtNTHf* • ELIS CLEjTOR pmREi's*n,LU'* «AHGOS

\ MEGALOPOLIS

IMfFTM

f%$k^ -r •~>~ '**?***' ^~~^

,

TABLE 1 Greek Coins Mint Number of Date specimens 1. Amphipolis 1 AE c. 325 B.C. 2. Atrax (Pelasgiotis) 1 AE c. 300-196 B.C. 3. Crannon (Pelasgiotis) 1 AE c. 400-344 B.C. 4. Eurea (Pelasgiotis) 1 AE c. 344 B.C. 5. Halus (Phthiotis) 1 AE 3rd century B.C. 6. Lamia (Phthiotis) 1 AE c. 400-344 B.C. 7. Peumata (Phthiotis?) 1 AE c. 302-286 B.C. 8. Ambracia (Epicus) 1 AE c. 238-168 B.C. 9. Argos Amphilochicum 1 AE c. 350-250 B.C. 10. Aetolian League 37 AE c. 279-168 b.C. 11. Locri Opuntii (Epicnemidi) 2AE c. 338-300 B.C. 12. Phocis: Federal Coinage a. 4AE 371-357 B.C. and later b. 3AE c. 357-346 B.C. 13. Chalkis (Euboea) 1 AR c. 369-313 B.C. 14. Eretria (Euboea) 1 AE c. 369-313 B.C. and later 15. Histiaea (Euboea) 1 AE c. 369-338 B.C. 16. Corinth 8AE 4th-3rd centuries B.C. 17. Philius (Phliasia) a. 1 AE c. 400-360 B.C. b. 1 AE c. 370-280 B.C. 18. Sicyon a. 2AR 4th century B.C. b. 1 AE c. 330-200 B.C. c. 4AE c. 330-290 B.C. & 2nd century B.C. 19. Aegeira 1 AE c. 330 B.C. 20. Dyme 1 AE c. 350 B.C. 21. Achaean League 4AE before 251 B.C. 22. Elis a. 1 AR c. 363-343 B.C. b. 2AE c. 271 - 191 B.C. 23. Argos (Argolis) 3AE c. 4th-3rd centuries B.C. 24. Hermione (?) 1 AE c. 350-322 B.C. 25. Arcadia (struck at Megalopolis) 1 AR after c. 370 B.C. 26. Cleitor 1 AE c. 362-300 B.C. 27. Pheneus 1 AE after c. 362 B.C. 28. Miletus 1 AE c. 300-250 B.C. 29. Carthage (Zeugitanae) 1 AE c. 241-146 B.C. Byzantine Coin 30. Constantinopole 1 AE c. A.D. 568/9

192 CATALOGUE SNG Cop, no. 87. 7. Peumata (Phthiotis?), c. 302-286 B.C. Greek Coins Obv.: Head of nymph bound with oak-wreath, r., circle of dots.

Macedonia Rev.: The monogram 1. Amphipolis, c. 325 B.C. Obv.: Heracles in lion's skin, r. Rev.: Eagle r. standing on thunder around [IIEY] MA [TIQN]; in the bolt; AAEEA [NAPOY] camp on r. indistinct symbol. AE; hi; 17 mm; 3.4 gr.; j.tt.156. AE;h3; 14 mm; 1.5 gr.; N.I. 112. SNG Cop, no. 1025 SNG Cop, no. 198. Thessaly Epirus 2. Atrax (Pelasgiotis), c. 300-196 8. Ambracia, c. 238-168 B.C. B.C. Obv.: Head of Dione laureate and Obv.: Head ofApollo 1., laureate. veiled, r. Rev.: Horseman r. Rev.: Obelisk ofApollo within laurel- AE; h6; 17 mm; 6.6 gr.; j.71.141 wreath; on both sides of the obelisk SNG Cop, nos. 30-31 the letters BMC, Thessaly to Aetolia, 14, no. 2 AM 3. Crannon (Pelasgiotis), c. 400-344 BP B.C. AE; hl2; 17 mm; 5.0 gr.; N.I.67. Obv.: Head of Poseidon r., laureate. SNG Cop, no. 23. Rev.: Thessalian horseman r. BMC, Thessaly to Aetolia, 94, no. 5 AE;h3; 18 mm; 4.7 gr.; j.7r.l31. Acarnania SNG Cop, no. 39 9. Argos Amphilochicum, c. 350-250 4. Eurea (Pelasgiotis?), before c. 344 B.C. B.C. Obv.: Young male head, 1. Obv.: Female head facing. Rev.: Dog at bay r. Rev.: Vine-branch with grapes. AE;h2; 16 mm; 5.1 gr.; N.I.79. AE;hl2; 15 mm; 3.0 gr.; N.I. 170. SNG Cop, no. 319. SNG Cop, no. 49. Aetolia 5. Halus (Phthiotis), 3rd century B.C. 10. Aetolian League, c. 279-168 B.C. Obv.: Head of Zeus Laphystious 1., Obv.: Atalanta or Aetolia wearing diademed. kausia, r. Rev.: Phryxus clinging to ram, r., Rev.: Calydonian boar at bay, r.; with chlamys over the shoulders; be above AIT [Q]. neath AAEQ[N]. AE;hl2; 13 mm; 1.5 gr.; N.I.89. AE;hll; 13 mm; 2.0 gr.; N.I.92. SNG Cop, no. 21. SNG Cop, no. 64; BMC, Thessaly to BMC, Thessaly to Aetolia, 196, no. Aetolia, 13, no. 2; Head, HN, 295-96 27 6. Lamia (Phthiotis), c. 400-344 B.C. 11. Obv.: Atalanta or Aetoliawearing Obv.: Head ofAthena r. kausia, r. Rev.: Philoctetes standing and shoot Rev.: Spear-head 1.; above AITQ; be ing birds. neath AQN; on r. in the camp grapes. AE;hl2; 14 mm; 1.9 gr.; N.I.78. AE; h6; 19 mm; 4.7 gr.; N.I. 136.

193 SNG Cop, no. 24; BMC, Thessaly to Rev: Spear-head and jaw-bone of Aetolia, 197, no. 34 boar, r.; above AITQ; beneath AQN; 12. Similar to no. 11. on 1. in the camp grapes, in the mid AE;h9; 19 mm; 4.5 gr.; N.I.148. dle the monogram SNG Cop, no. 23. 13. Similar to no. 11. AE;h9; 18 mm; 5.1 gr.;N.I.142. SNG Cop, no. 22ff. AE;hl2;21 mm; 5.1 gr.;N.I.128. 14. Similar to no. 11. SNG Cop, no. 29; BMC, Thessaly to AE;h9; 17 mm; 4.1 gr.;N.I.163. Aetolia, 198, no. 51. SNG Cop, no. 23. 26. Similar to no. 25. 15. Similar to no. 11, but spear-head r. AE;h9; 14 mm; 2.7 gr.; N.I.83. AE;h3;20mm;5.7gr.;N.I.157. SNG Cop, no. 25. 27. Similar to no. 25. 16. Similar to no. 15. AE;h3; 19 mm; 6.8 gr.; N.I.91. AE;h9; 14 mm; 2.2 gr.; N.I. 158. 28. Similar to no. 25. SNG Cop, no. 25. AE;h6; 19 mm; 6.0 gr.; N.I. 143. 17. Similar to no. 15. 29. Similar to no. 25. AE;hl2; 13 mm; 3.0 gr.; N.I.86. AE;h6; 19 mm; 5.5 gr.; N.I.64. SNG Cop, no. 25. 30. Similar to no. 25. 18. Obv.: Young male-head laureate, r. AE;h3; 19 mm; 4.4 gr.; N.I.113. Rev.: Spear-head and jaw-bone of 31. Similar to no. 25. boar, r.; above AI [TQ], beneath AE;h2; 18 mm; 5.4 gr.; N.I.85. AQN; on 1. grapes. 32. Similar to no. 25. AE;h3;20mm;5.2gr.;N.I.106. AE;hl2; 18 mm; 5.0 gr.; N.I. 109. SNG Cop, no. 28; BMC, Thessaly to 33. Obv. and Rev.: as no. 25, but in Aetolia, 198, no. 4ff. stead of a monogram there are traces 19. As no. 18. of letters. AE;h8; 19 mm; 5.7 gr.; N.I.84. AE;h9; 19 mm; 4.7 gr.; N.I.81. SNG Cop, no. 28. SNG Cop, no. 31. 20. As no. 18. 34. Obv. and Rev.: as no. 25, but there AE;h3; 18 mm; 5.3 gr.; N.I.155. are no letters or monograms visible. SNG Cop, no. 28. AE;h3; 16 mm; 2.1 gr.;N.I.81. 21. As no. 18. SNG Cop, no. 32. AE;h6; 18 mm; 3.9gr.; N.I.118. 35. Similar to no. 34. SNG Cop, no. 28. AE;hl2; 15 mm; 2.4 gr.; N.1.63. 36. Similar to no. 34. 22. As no. 18. AE;hll; 17 mm; 4.8 gr.; N.1.117. AE; hl2; 15 mm; 2.4 gr.; N.I.129. SNG Cop, no. 28. 37. Similar to no. 34. 23. As no. 18. AE;h8; 15 mm; 2.3 gr.; N.I. 168. AE;h3; 17 mm; 4.5 gr.; N.I.I 11. 38. Similar to no. 34. AE;h2; 14 mm; 2.6 gr.; N.I. 102. SNG Cop, no. 28. 39. Similar to no. 34. 24. As no. 18. AE;h5; 17 mm; 4.3 gr.; N.I.77. AE;h3; 14 mm; 1.9 gr.; N.I.76. SNG Cop, no. 28. 40. Obv.: head ofAthena r. Rev.: Heracles standing with club and 25. Obv.: as no. 18.

194 lion's skin; on r. AITQ, on.l. AQN 51. Similar to no. 49. AE;h4; 19 mm; 6.0 gr.; N.I.68. AE;hl2; 13 mm; 1.9 gr.; N.I.130. SNG Cop, no. 35; BMC, Thessaly to SNG Cop, no. 115. Aetolia, 199, no. 64ff. 52. Similar to no. 49. 41. Similar to no. 40. AE;hl2; 13 mm; 1.6 gr.; N.I. 121. AE;h4; 18 mm; 4.8 gr.; N.I. 133. SNG Cop, no. 114. 42. Similar to no. 40. 53. Similar to no. 49. AE;hl; 18 mm; 3.7 gr.; N.I.73. AE;hl2; 13 mm; 4.5 gr.; N.I.105. 43. Similar to no. 40, but no legend is SNG Cop, no. 114. visible. 54. Similar to no. 49. AE;h3; 18 mm; 4.6 gr.; N.I.71. AE; hl2 (obverse); 12 mm; 1.9 gr.; SNG Cop, no. 37. N.I.115. 44. Similar to no. 40. SNG Cop, no. 114. AE;hl2; 18 mm; 4.4 gr.; N.I.62. 55. Federal Coinage, c. 357-346 B.C. SNG Cop, no. 37. Obv.: bull's head with fillets, facing. Rev.: vanished. 45. Similar to no. 40. AE; hl2 (obverse); 18 mm; 4.5 gr.; AE;h9; 18 mm; 3.8 gr.; N.I.132. N.I.147. SNG Cop, no. 39. SNG Cop, no. 127. 46. Obv.: Atalanta or Aetolia wearing Euboea kausia, r. 56. Chalcis, c. 369-313 B.C. Rev.: owl standing, traces of letters. Obv.: head of nymph Chalkis, 1.; AE;h6; 15 mm; 2.7 gr.; N.I. 124. above indistinct symbol. SNG Cop,no. 41. Rev.: flying eagle holding seprent, 1. Locris AE;hl2; 15 mm; 2.7 gr.; N.I.65. 47. Locri Opuntii (Epicnemidi), c. SNG Cop, no. 432ff (for the type); 338-300 B.C. Picard 1979, pi. I, no. 4g. Obv.: Head ofAthena, r. 57. Eretria, c. 369-313 B.C. and later. Rev.: Grapes; on 1. AOK [P]; on r. Obv.: bull standing, r. EnifKNA]. Rev.: Vine-branch with grapes; on 1. AE;hl2; 13 mm; 1.5 gr.; N.I.94. the letters E A. SNG Cop, no. 68-69. AE;hl2; 14 mm; 1.4 gr.; N.I.93. 48. Similar to no. 47. SNG Cop, no. 488. AE;hl2; 12 mm; 1.9 gr.; N.I.88. 58. Histiaea, c. 369-338 B.C. SNG Cop, no. 75. Obv.: head of Maenad wearing vine- Phocis wreath, r. 49. Federal Coinage, c. 371-357 B.C. Rev.: front part of bull, r.; above I2TI; and later. on 1.grapes. Obv.: head ofAthena. AE;hl2; 14 mm; 2.7 gr.; N.I.103. Rev.: O within laurel wreath. SNGCop, no. 514. AE;h3; 15 mm; 2.9 gr.; N.I. 161. Corinthia SNG Cop, no. 116. 59. Corinth, 4th-3rd centuries B.C. 50. Similar to no. 49, but on the re Obv.: Pegasus 1.; beneath Cp verse the initial letters OQ. Rev.: Trident decorated (upwards); on AE; hl2; 13 mm; 2.1 gr.; N.I.159. r. wreath (?). SNG Cop, no. 117. AE;h9; 14 mm; 2.4 gr.; N.I. 101. SNG Cop, no. 171.

195 60. Similar to no. 59. 70. Obv.: as no. 69, but no letters are AE;h6; 13 mm; 1.4 gr.; N.I. 108. visible. SNG Cop, no. 186. Rev.: Dove flying r. 61. Similar to no. 59, but on the re AR;h9; 15 mm; 1.6 gr.; N.I.95. verse the letter A is visible on r. of the SNG Cop, no. 64 and 65. 71. Sicyon, c. 330 - c. 200 B.C. camp. AE;h8; 13 mm; 1.3 gr.; N.I.72. Obv.: dove flying r. SNG Cop, no. 181. Rev.: 21 in wreath. 62. Similar to no. 59. AE;hl2; 13 mm; 1.9 gr.; N.I.87. AE;h9; 12 mm; 1.8 gr.; N.I.98. Warren, 37, no. A5, pi. 6. SNG Cop, no. 172. 72. Sicyon, c. 330 - c. 290 and 2nd 63. Similar to no. 59, but the trident century B.C. Obv.: dove 1. on the reverse is downwards. AE;h8; 12 mm; 1.2 gr.; N.I. 162. Rev.: EY in wreath. SNG Cop, no. 188. AE;hl2; 17 mm; 3.5 gr.; N.I.69. Warren, 44, no. 4a, pi. 7. 64. Similar to no. 59. AE;hl2; 11 mm; 2.2 gr.; N.I.97. 73. Obv.: as no. 71. SNG Cop, no. 169. Rev: laurel-wreath without letters. AE; h3; 16 mm; 3.2 gr.; N.I.100. 65. Similar to no. 59, but on the re Warren, 45, nos. 5a-c, pi. 7. verse on r. of the camp there is an in 74. Similar to no. 72. distinct object. AE;hl; 16 mm; 2.9 gr.; N.I. 120. AE;h8; 11 mm; 1.3 gr.; N.I.167. Warren, 44, no. 4a, pi. 7. SNG Cop, no. 178. 75. Similar to no. 72. 66. Obv.: Pegasus r. AE; h3; 16 mm; 2.3 gr.; N.I.140. Rev.: trident (downwards). Warren, 44, no. 1 (for the type), pi. 7. AE;h6; 11 mm; 1.4 gr.; N.I.74. 76. Aegeira, c. 330 B.C. SNG Cop, no. 188. Obv.: head ofAthena, r. Phliasia Rev.: front part of goat in olive wreath. 67. Phlius, c. 400-360 B.C. AE; hl2; 12 mm; 1.4 gr.; N.I.135. Obv.: butting bull, 1. SNG Cop, no. 127. Rev.: :0:. BMC, Pel., 17, nos. 1-2. AE;hl2; 12 mm; 1.5 gr.; NX127. Achaia SNG Cop, no. 10. 77. Dyme, c. 350 B.C. SMC, Pel, 34, no. 16 Obv.: head of Demeter, r. 68. Phlius, 370-280 B.C. Rev.: AY in wreath . Obv.: vanished. AE; h9; 14 mm; 2.0 gr.; N.I.172. Rev.: O within ivy-wreath. SNG Cop, no. 146. AE;hl2; 13 mm; 1.5 gr.; N.I.80. BMC, Pel., 21, no. 2. BMC, Pel., 34, no. 18. 78. Achaean League, before 251 B.C. Sicyonia Obv.: head of Zeus, r., laureate. 69. Sicyon, 4th century B.C. Rev.: Monogram Obv.: Chimaera 1.; below ZE. Rev.: dove flying 1. AR;h3; 16 mm; 1.8 gr.; N.I.I 10. x SNG Cop,no. 5\;BMC, Pel., 38, no. 30. in laurel-wreath

196 AE;hl; 15 mm; 3.4 gr.; N.I. 114. AE;hl2; 11 mm; 0.9 gr.; N.I.173. SNG Cop, no. 229. SNG Cop, no. 143. 79. Similar to no. 78. Arcadia AE;h3; 14 mm; 2.1 gr.;N.I.139. 89. Megalopolis, after c. 370 B.C. SNG Cop, no. 229. Obv.: head of young Pan, 1. 80. Similar to no. 78. Rev.: AR below syrinx. AE;hl2; 14 mm; 1.7 gr.; N.I.64. AR; h6; 0.9 mm; 0.3 gr.; N.I.99. SNG Cop, no. 229. SNG Cop, no. 197. 81. Similar to no. 78. BMC, Pel, 176. AE;hl2; 13 mm; 2.6 gr.; N.I.107. 90. Cleitor, c. 362-300 B.C. SNG Cop, no. 229. Obv.: Helios radiate facing. 82. Elis, c. 363-343B.C. Rev.: monogram KI Obv.: Head of Zeus laureate, r. AE;h6; 11 mm; 1.3 gr.; N.I.82. Rev.: Eagle standing on capital of ion SNG Cop, no. 229; BMC, Pel., 180, ic column; on 1. F and on r. A. no. 14. AR; hl2; 14 mm; 1.9 gr.; N.I.160. 91. Pheneus, after c. 362 B.C. SNG Cop, no. 387; Seltman, 60. Obv.: Dcmctcr wearing corm-wreath, 83. Elis, 271-191 B.C. 1. Obv.: Head of Zeus laureate, r. Rev.: caduceus; OE. Rev.: Thunderbolt; no letters visible. AE;h9; 17 mm; 3.1 gr.;N.I.134. AE;h6; 15 mm; 1.1 gr.; N.I.90. SNG Cop, no. 276; BMC, Pel., 195, Gardner 1879, 264, no.8. no. 17 84. Obv.: as no. 83, but the head on 1. 92. Miletus, c. 300-250 B.C. Rev.: FA in olive-wreath. Obv.: head ofApollo, r. laureate. Gardner 1879, 264, no. 3 (silver coin); Rev.: lion walking looking back. BMC, Pel., 73, no. 139. AE;h9; 13 mm; 1.4 gr.; N.I.122. Argolis SNG Cop, no. 985. 85. Argos, 4th-3rd centuries B.C. 93. Carthage Zeugitanae, c. 241-146 Obv.: head of wolf, r. B.C. Rev.: A (illegible). Obv.: Head of Tanit, 1. AE;h9; 13 mm; 1.9 gr.; N.I.I 19. Rev.: Horse standing r. in front of SNG Cop, nos. 62 and 77. palm-tree. 86. Obv.: as no. 85. AE; hl2; 16 mm; 2.0gr.; N.I.104. Rev: A. Below Corinthian helmet. Head, HN, 879 (for the type, coin of AE;h9; 12 mm; 2.0 gr.; N.I.I 16. electrum); Plant, no. 1215, 76. SNG Cop, no. 64; BMC, Pel., 144, no. Byzantine lOOff. 94. Constantinople 87. Similar to no.86, but on the re Justin II and Sophia, A.D. 568/9. verse beneath the letter A there are Obv.: [DNIVS] TIN [VS]. Justin and traces of illegible symbol. Sophia enthroned. AE;hl2; 12 mm; 1.9 gr.; N.I.96. Rev.: A SNG Cop, no. 62; BMC, Pel., 144, no. N 105. N 88. Hermione (?), c. 350-322 B.C. O Obv.: female head, r. AE; h6; 20 mm; 5.6 gr.; N.I.70. Rev.: traces of corn-wreath. A.R. Bellinger, Catalogue ofthe

197 Byzantine Coins in the Dumbarton cite only their inventory number, Oaks Collection and in the diameter and weight. Also due to Whittemore Collection, I (Washington the oxidized state ofthe coins and 1966), 214, no. 47b. to their normal wear and tear The following coins are so disin through handling, the weight cited tegrated and oxidized as to be il cannot be taken to represent the legible, for which reason we will original one.

95 AE h- 17mm 2.7 gr N.I. 171 96 AE h 15mm 3.7 gr N.I. 151 97 AE h 14mm 2.0 gr N.I. 169 98 AE h 14mm 1.2 gr N.I. 165 99 AE h 14mm 0.4 gr N.I. 138 100 AE h 13mm 2.6 gr N.I. 154 101 AE h 13mm 1.8 gr N.I. 75 102 AE h 13mm 1-7 gr N.I. 66 103 AE h 13mm 1.2 gr N.I. 174 104 AE h 12mm 1.6 gr N.I. 125 105 AE h 12mm 1.6 gr N.I. 125 106 AE h 12mm 1.3 gr N.I. 144 107 AE h 12mm 1.1 gr N.I. 145 108 AE h 12mm 0.6 gr N.I. 166 109 AE h 11mm 2.4 gr N.I. 152 110 AE h 11mm 1.3 gr N.I. 149 111 AE h 11mm 1.0 gr N.I. 150 112 AE h 11mm 0.7 gr N.I. 126 113 AE h 7mm 0.3 gr N.I. 137

198 5>

CO

> n 0

(D O o

01 o ho o to

f-ktfi <%••'* CO o

>

0 to o o to o On

D Additional information on and winter. Head believes that the the coins ofthe catalogue bronze coins ofthe city may be oftwo periods: 400-344 and 300-200 B.C. 1. Found 23-10-2000 in trench T30N, stratum 7, F00-1108. The coin is in very 6. Found 22-6-1998. Surface find. Poor condition. The weight is the same as good state ofpreservation. Its weight is that ofthe specimen in SNG Cop. less than the specimen of SNG Cop. I. Found 15-7-1999 in trench 027, stra 2. Found 19-7-2000 in trench T30, stra tum 5, F00-1042. The coin has a very tum 3, 2nd pass, area I., F99-3041. Our good state ofpreservation. It is heavier specimen is slightly lighter than the one compared to the specimen ofSNG Cop ofSNG Cop. Because ofits poor condi and the head ofApollo is turned 1. Fur tion the legend on the reverse is not vis thermore, our specimen can be com ible and there is an indistinct symbol. pared to the coin no. 2 of BMC Thes 8. Found 2-7-1998 in trench T6NA, saly, 14 (pi. 11,8). According to Head, stratum 5, F98-29. It is very well pre this type is copied from the coins of served and it weighs approximately the Philip of Macedon and is attributed to same as the specimen in SNG Cop. the city ofAtrax.

9. Found 22-6-1998. Surface find. The 3. Found 29-6-2000 in trench Tx92, coin is in bad condition and has ap stratum 2, F00-2014. It is moderately proximately the same weight as the one preserved and the letters on the reverse in SNG Cop. are not visible. It is slightly heavier than that in SNG Cop. 10. Found 12-7-1999 in trench 027, stratum 3, F99-3007. Very poor condi 4. Found 27-10-2000 in trench T2 9, stratum 5, F00-1142. Due to its bad tion; the second part [AQN] ofthe leg preservation the female head ofthe ob end AITQAQN is not visible. Weight smaller than the specimen in SNG Cop. verse and the legend on the reverse are not visible. It is lighter than the pattern II. Found 20-7-2000 in trench T30, in SNG Cop. stratum 6, F00-1062. Good condition. Smaller weight than the specimen in 5. Found 13-7-1999 in trench 027, stra SNG Cop. tum 3, area I, X-46. Its state ofpreser vation is excellent and it is lighter than 12. Found 12-7-2000 in trench Tx92, the specimen ofSNG Cop. In our spec stratum 5, FOO-2038. The coin is very imen the head of Zeus is turned 1., badly preserved and has approximately while on the ones in SNG Cop and the the same weight as the one in SNG Cop. BMC Thessaly to Aetolia are turned r. That in SNG Cop is dated in the third 13. Found 19-7-2000 in trench T30, century B.C., and BMC Thessaly gives FOO-1036. The weight is similar to the a date range from 300 to 190 B.C. As far one in SNG Cop. as the epithet of Laphystios is con cerned, it was a name of Zeus wor 14. Found 17-10-2000 in trench T27B, shipped at Halus as the god of storm stratum 6, FOO-1065. It is in poor con-

207 dition and its weight is slightly less than 22. Found 27-7-1999 in trench T27N, stra the one in SNG Cop. tum 5e, X-53. It weighs the same as the one in SNG Cop and, as in the previous 15. Found in trench 027, without date, case, the legend AQN is not visible. stratum 3, area I., F99-3005. It is very well preserved, it weighs a little more 23. Found 21-7-1999 in trench 027, than the specimen in SNG Cop. No stratum 3, area IV, F99-3074. It is in grapes on the reverse. very poor condition and weighs a bit less than the one in SNG Cop. Due to its 16. Found 24-10-2000 in trench T30, very bad condition there is no visible stratum 7, F00-1115. The coin weighs approximately the same as the one in the legend AITQAQN or the other symbols. SNG Cop. Well preserved and, as the previous one, has no grapes in the camp 24. Found 23-7-1998 in trench T6NA, on the reverse. stratum 5d., F98-5010. The coin is very 17. Found 13-7-1999 in trench 027, badly preserved and so no inscription is stratum 3, area I., F99-3021. Very poor visible on the reserve. Our specimen is state ofpreservation and on the reverse a bit lighter than the one in SNG Cop. the spear-head is on the r., while the oth 25. Found 19-7-2000 in trench T30, ers have it on the left. As the two pre FOO-1034. The coin is in very good state vious coins it bears no grapes in the of conservation but the reverse is dete camp on the reverse. It is heavier than riorated. It weighs less than the one in the one in SNG Cop. SNG Cop. 18. Found 13-7-1999 in trench 027, stra tum 3, area I., F99-3028. It is in very 26. Found 17-10-2000 in trench T28A, good condition. Its weight is heavier than stratum 5, F00-1064. It is in excellent the one in SNG Cop. Excellent features condition and has approximately the ofthe young male head on the obverse. same weight as the one in SNG Cop.

19. Found 12-9-1999 in trench 027, stra 27. Found 12-7-1999 in trench 027, tum 3, area I., 3604. This coin is also in stratum 3. As the previous it is in ex a very good state ofpreservation and is cellent condition and is heavier than the heavier than the one in SNG Cop. one in SNG Cop.

20. Found 16-10-2000 in trench T29, 28. Found 19-7-2000 in trench T30, stratum 5. It is in poor state and is heav FO0-1O35. It is in relatively good con ier than the one in SNG Cop. dition and it weighs slightly more than the one in SNG Cop. The figures ofthe 21. Found 27-7-1999 in trench T27N, reverse are in an incuse. stratum 5e, F99-1020. It weighs less than the one in SNG Cop. Worse state 29. Found 2-7-1998 in trench T6NA, of preservation on the reverse than on stratum 5e, F98-32. It has the same the observe. The second part ofthe leg weight as the one in SNG Cop. Fur end (AQN) is not visible on the lower thermore, as the previous, it bears an part ofthe camp on the reverse. incuse on the reverse.

208 30. Found 14-7-1999 in trench 027, is pronounced and no legend or symbol stratum 3, area I., F99-3036. It is in very is visible. It weighs less than the one in poor condition and the deterioration is SNG Cop. more pronounced on the reverse. Due to this, the legend is not visible and we 38. Found 22-7-1999 in trench T27B, can hardly distinguish the monogram. It stratum 4, F98-1015. It weighs less than is lighter than the one in SNG Cop. the one in SNG Cop.

31. Found 12-7-1999 in trench 027, 39. Found 24-6-1998 in trench Tx61, stratum 3, area I, F99-3009. As the pre stratum 2, F98-3005. Due to its deteri vious, it is deteriorated, almost the same oration the entire legend AITQAQN as the specimen in SNG Cop. is not very well visible. It weighs less than the one in SNG Cop. 32. Found 22-7-1999 in trench 026, stratum 3, area II, F99-3038. It is in 40. Found 25-6-1998 in trench T6, F98- poor condition and weighs less than the 4001. It is in fine condition and weighs one in SNG Cop. almost the same as the one in the cata logue ofSNG Cop. 33. Found 27-6-1999 in trench T6NA. It weighs almost the same as the one in 41. Found 4-7-2000 in trench T31, stra SNG Cop. On the reverse, the incuse is tum 2, F00-1007. It is in excellent con hardly distinguishable. The diameter dition and weighs less than the one in varies from 16 to 19 mm. SNG Cop. On the obverse the head of Athena is on the left side of the camp. 34. Found 23-10-2000 in trench T30N, On the reverse the second part of the stratum 7, F00-1102. It is in very poor legend is on the edge of the perimeter condition and a small part is missing. of the coin. Due to its remarkable deterioration no legend or monogram is visible on the 42. Found 3-7-1998 in trench T6NA, reverse. It weighs less than the speci stratum 5, F98-36. Deteriorated with no men in SNG Cop. letters of the legend visible. It weighs less than the one in SNG Cop. 35. Found 2-7-1998 in trench T6NA, stratum 5e, F98-19. As the previous it is 43. Found 1-7-1998 in trench TN6A, in poor condition and no symbols are stratum 6, F98-11. It is in very poor visible on the reverse. It weighs less state ofpreservation and weighs almost than the one in SNG Cop. the same as the one in SNG Cop.

36. Found 12-7-200 in trench 23, stra 44. Found 6-7-1998 in trench T6NA, tum 3. It is in a mediocre condition and stratum 5e, F98-5001. Its condition is on the reverse no monogram is visible, excellent and weighs the same as the while from the second part of the leg one in SNG Cop. end AQN only the letter A is visible. 45. Found 18-7-2000 in trench T29, 37. Found 26-10-200 in trench T29, stratum 4, F00-1032. It is in a mediocre stratum 5, F00-1140. The deterioration condition and weighs a bit less than the

209 one of SNG Cop. On the reverse is no 54. Found 13-7-1999 in trench 027, visible monogram. stratum 3, area I, X-39. It is very dete riorated and weighs slightly more than 46. Found 29-7-1999 in trench T27N, the one in SNG Cop. stratum 4d. Deteriorated and the legend is not visible on the reverse. It weighs 55. Found 20-7-2000 in trench T28, more than the one ofSNG Cop. stratum 4, F00-1063. The coin is very deteriorated and the figures on the re 47. Found 15-7-1999 in trench 027, verse have vanished. It weighs almost stratum 3, 2-1 pass, F99-3044. It is in the same as the one in SNG Cop. very good condition and has the same weight as the specimen no. 68 ofSNG 56. Found 1-7-1998 in trench T6NA, Cop and is lighter than no. 69 of the registered ad F98-15. Good state of same catalogue. preservation. On the reverse the in scription (XAA) is not visible. The 48. Found 12-7-1999 in trench 027, comparison in SNG Cop has the nymph stratum 3, area I, F99-3008. Almost ex on the obverse and the flying eagle on cellent state of preservation. Approxi the reverse r., while our coin has these mately the same weight as the one in figures on the 1. The indistinct symbol SNG Cop.On the reverse the inscription on the reverse probably is a cornucopia. EniKNA is hardly visible. According to Picard 1979, 17, this coin must be a drachma. Our specimen is 49. Found 23-10-2000 in trench T30A, very similar to his no. 4g., pi. I, which stratum 7, F00-1100. It is in very good weighs 3.65 gr. condition and is heavier than the one in SNG Cop. 57. Found 15-7-1999 in trench 027, stratum 3, area 1,2nd pass, X-45. Its con 50. Found 25-10-2000 in trench T30D, dition is mediocre and its weight is less stratum 8, FOO-1123. Its condition is than the one in SNG Cop. almost excellent and it weighs more than the one ofSNG Cop. 58. Found 15-7-1999 in trench 027, stra tum 3,2nd pass,areaI, F99-3039. Itscon 51. Found 19-7-2000 in trench T30, dition is excellent and it has almost the stratum 5, F00-1043. The same state of same weight as the one in SNG Cop. preservation as the previous and almost the same weight as the specimen in 59. Found 13-7-1999 in trench 027, SNG Cop. stratum 3, area I, X-38. It is in excel lent state and weighs more than the one 52. Found 15-7-1999 in trench 027, stra ofSNG Cop. The trident on the reverse tum 3,2nd pass, area I, F99-3052. Its con is placed 1. dition is mediocre and it weighs slightly less than the specimen in SNG Cop. 60. Found 13-7-1999 in trench 027, stra tum 3, area I, F.99-3031. Its condition is 53. Found 15-7-1999 in trench 027, mediocre. The surface on the reverse is stratum 3, 2nd pass, area I, F99-3047. smoothened by deterioration and as a re Very good condition, it has almost the sult no symbols or letters are visible. It same weight as the one in SNG Cop. weighs less than the one in SNG Cop.

210 61. Found 13-7-1998. Surface find. Its preservation and it weighs less than the state of conservation is fine and it one in SNG Cop. weighs less than the specimen in SNG Cop. Coins of this type in the BMC, 70. Found 13-7-1999 in trench 027, Corinth, 54-55 are dated 400-300 B.C. stratum 3, area I, F99-3025. Mediocre state ofpreservation. It weighs less than 62. Found 16-7-1999 in trench 027, the specimens no. 64 and 65 in SNG stratum 3, area 1,2nd pass, F99-3054. Its Cop, which are hemidrachms. condition is very bad and so the raffig- urations are hardly visible. It weighs 71. Found 12-7-1999 in trench 027, slightly more than the one in SNG Cop. stratum 3, area I, F99-3014. Its condi tion is very bad. 63. Found 23-10-2000 in trench T30N, stratum 7, F00-1101. As the previous 72. Found 10-7-1998 in trench 027, F98-4008. Excellent state of preserva one its condition is very bad and as a re tion. sult the symbols are not visible. Its weight is less than the one in SNG Cop. 73. Found 13-7-1999 in trench 027, stratum 3, area I, X-36. Its preservation 64. Found 15-7-1999 in trench 027, stra is bad and the initials ofthe magistrate tum 3,2nd pass, F99-3049. Itisbadly pre are not visible. served and the deterioration of the coin makes the surface smooth. It weighs a 74. Found 21-7-1999 in trench 027, bit more than the one of SNG Cop. stratum 3, area IV, F99-3073. As the previous it is deteriorated and the ini 65. Found 27-10-2000 in trench T27A, tials of the magistrate are not visible. stratum 7, without further indications. Its state of preservation is fine and it 75. Found 19-7-2000 in trench Tx93, has the same weight as the one in SNG F00-2055. It is in poor condition, espe Cop. The figure ofPegasus is placed on cially on the reverse and only the lau the r. ofthe camp. rel-wreath is visible.

66. Found 23-6-1998 in trench T6, sur 76. Found in trench T30, stratum 5. Its face find. Its condition is very poor and state of preservation is bad and it is it weighs slightly less than the one in lighter than the one in SNG Cop. SNG Cop. 77. Found 23-10-2000 in trench T30A, 67. Found 12-7-2000 in trench Tx92, stratum 7, FOO-1093. It is in a very poor stratum 5a, F00-2041. It is relatively condition and it weighs more than the well preserved and it weighs less than one in SNG Cop. the one in SNG Cop. 78. Found 9-7-1999 in trench N27, stra 68. Found 22-6-1998, without further tum 2, F99-2018. It is badly preserved, indications. Bad state of preservation. which makes it difficult to understand from which city it is emitted, but it must 69. Found 20-7-1999 in trench 027, be one of the League. Our specimen stratum 3, area II. It is in good state of weighs more than the one in SNG Cop.

211 79. Found 20-7-2000 in trench T30, stra tion is mediocre and it weighs more tum 5, F00-1059. As the previous it is in than the one in SNG Cop. a bad state ofpreservation and this spec imen must also belong to the first emis 87. Found 15-7-1999 in trench 027, sions ofthe League. It weighs more than stratum 3, 2nd pass, F99-3048. Due to the one in SNG Cop. deterioration the symbol beneath A on the reverse is not legible. It weighs more 80. Found 25-10-2000 in trench T30D, than the one in SNG Cop. stratum 8, F00-1124. It is deteriorated and also belongs to the first emissions of 88. Found 18-10-2000 in trench T28A, the League. It weighs almost the same stratum 6, F00-1072. Its state ofpreser as the one in SNG Cop. vation is very bad. This coin could be part ofthe mint ofHermione (Argolis), 81. Found 14-7-1999 in trench 027, as we can deduce from the features of stratum 3, area I, F99-3032. It is in poor the face on the obverse. It weighs less state of preservation and also must be than the one in SNG Cop. long to the first emissions ofthe League. 89. Found 15-7-1999 in trench 027, 82. Found 16-10-2000, stray find. Good stratum 3, 2nd pass, area I, FX-44. A state ofpreservation. It weighs less than small part is missing and its condition the one in SNG Cop. According to is mediocre. It weighs less than the one Gardner 1879, 246-49, this type is dat in SNG Cop. The type is dated 280-234 ed 365-362 B.C. (5th period). B.C. in BMC, Pel.

83. Found 12-7-1999 in trench 027, 90. Found 12-7-1999 in trench 027, stratum 3, F99-3018. The coin is very stratum 3, area I, F99-3019. Its preser deteriorated and no letters are visible vation is good and it has the same on the reverse. weight as the one in SNG Cop.

84. Found 15-7-1999 in trench 027, 91. Found 20-7-2000 in trench T30, stratum 3,2nd pass, area I, F99-3045. Its stratum 5, F00-1056. It is well pre condition is poor. The type of the re served and weighs slightly less than the verse is found only on silver specimens specimen in SNG Cop. dated 271-291 B.C., which are present ed in SNG Cop, as well as in Gardner 92. Found 15-7-1999 in trench 027, 1879,264, no. 3, and in the catalogue of stratum 3, 2nd pass, area V, F99-3037. BMC Pel The reverse is in a worse state ofpreser vation than the obverse. It weighs less 85. Found 13-7-1999 in trench 027, than the one in SNG Cop. stratum 3, area I. Its condition is bad. The reverse has completely vanished. 93. Found 15-7-1999 in trench 027, stra It weighs slightly more than nos. 62 and tum 3, 2nd pass, area I, F99-3050. Excel 77 in SNG Cop. lent state of preservation. According to Head this type ofcoins is mostly ofelec 86. Found 14-7-1999 in trench 027, tron. The female head of Tanit is a stratum 3, area I, F99-3033. Its condi Carthaginian form ofDemeter.

212 94. Found 9-7-1998 in trench 027, stra 105. Found 19-7-2000 in trench T30, tum 1, F98-4006. Its state ofpreserva stratum 5, FOO-1039. tion is mediocre. 106. Found 5-7-2000, stratum 2, F00- 95. Found 23-10-2000 in trench T30B, 1010. stratum 7, F00-1096. 107. Found 7-7-000 in trench T29, stra 96. Found 11-7-2000 in trench T30, tum 2, F00-1015. stratum 6, F00-1045.

97. Found 26-10-2000 in trench T31, 108. Found 25-10-2000 in trench T31, stratum 4, F00-1136. stratum 4, F00-1131.

98. Found 16-10-2000 in trench T27B, 109. Found 19-7-2000 in trench T30, stratum 6. F00-1046.

99. Found 18-7-2000 in trench Tx92, 110. Found 3-7-2000 in trench K29, stratum 7a. stratum 1, 3rd pass, FOO-5036.

100. Found 23-10-2000 in trench T30D, 111. Found 5-7-2000 in trench K29, stratum 7, F00-1107. stratum 1, 5th pass, F00-5042. 101. Found 21-7-1998 in trench 027, stratum 3, F98-4025. 112. Found 21-7-2000 in trench 027, stratum 3, area IV, F99-3072. 102. Found 13-7-1998 as a surface find. 113. Found 12-7-2000 in trench Tx92, 103. Found 24-10-2000 in trench T30N, stratum 5a. stratum 7, F00-1120.

104. Found 13-7-2000 in trench 027, stratum 3, area I, X-42.

213 References BMC,Thessaly to Aetolia Head,//N C.T Seltman P. Gardner, Catalogue ofGreek B.V. Head. Historia Numorum. A 'The Temple Coins ofOlympia', Coins, Thessaly toAetolia, 2nd ed. Manual ofGreekNumismatics, reprinted from Nomisma VIII, IX, (Bologna 1963) New and enlarged Edition (Ox XI ford 1911, repr. 1963) SNG, Cop BMC, Peloponnesus Sylloge Nummorum Graecorum, P. Gardner, Catalogue ofGreek Picard 1979 The Royal Collection ofCoins O. Picard, Chalcis et la Con Coins, Peloponnesus, 2nd ed. and Medals, Danish National federation Eubeenne. Etudede (Bologna 1963) Museum, 2nd ed. (New Jersey numismatique et d'histoire (IVe- 1982) lersiecle (Paris 1979) BMC, Corinth B.V. Head, Catalogue ofGreek J.A.W. Warren Picard 1984 Coins, Corinth, 2nd ed. (Bologna 'The Autonomous Bronze O. Picard, 'L'Antre Corycien IF, 1963) Coinage of Sicyon, part I', NC BCH Suppl IX (1984), 281-306 143(1983), 23-56, pi. 5-8

P. Gardner R. Plant 'The Coins ofElis',7VC 20 Greek Coin Types and Their (1879), 221-73, pi. XI-XVI Identification (London 1979)

214 Excavations on the Hill ofHagia Triada

Sanne Houby- The following pages contain a summa major archaeological horizons of the Nielsen and ry of the results of the excavations un site, and the conclusions are therefore loannis Moschos dertaken from 1999 to 2001 on the hill to be regarded as preliminary. ofHagia Triada. During the three sum mer seasons, excavations were contin Area I: Archaic and Clas ued from previous years in areas locat sical houses lining a road ed outside the Classical and Byzantine fortification wall (Fig. 1). During 1999-2000, three trenches (N26-N28) equivalent of 15.5m2 and Below follows an account ofthe exca one 12m long and 3m wide trench vations carried out in areas I-II while (Tx20-22, Tx91-93) were opened (Fig. a separate account of area III is under 1: area I; compare SPR, 235-238). All preparation. A thorough study remains in all, approximately 51.5m2 have to be carried out of the structures and been excavated in this area, in most the large amounts ofoften poorly pre parts down to the bedrock. served pottery, before the historical development of the site can be fully During these excavations, a series of understood. The account presented structures was uncovered below a huge here is no more than an outline of the stone-fill, which stemmed from the up-

Fig .1. Hill of Hagia Triada.

Areas of excavation __ v» 1999-2001 (I, II, III). Charalambos Marinopoulos. <> ft i 1 V \ IprQ^ + V j( irt" ^s. J..,. \ W l\ / 5g§F I %&; > / 4 ! t-rW tjt.i. Hi ^A

•f + "* T

175 Fig. 2. Archaic and Classical foundation walls lining a road (N26-N28). Classical walls running up towards the road in trenches Tx20-22 and Tx91 - 93. Charalambos Marinopoulos.

Fig. 3. View of the road lined per plateau ofthe hill. It was ascertained by Archaic and Classical that the structures belonged to Archaic foundation walls. and Classical buildings facing a road. Henrik Frost. Most of these structures and their asso ciated finds were found in a poor state m ':«« _ ***•* of preservation, but nevertheless point to the existence, already in Archaic times, of a wide public road, lined with <*0 buildings (Fig. 2). The road was in use at least until Classical times, when well-built houses, separated by narrow lanes, were laid out at right angles to the road. A more detailed description of the stratigraphical situation of the 15m towards the Northwest (Fig. 3). structures follows below, with some of Three dry-stone foundation walls the diagnostic finds, and their relation (ADQ, AFZ, AFN) were excavated to to each other in the area. wards the Northeast forming a slightly irregular row, which ran parallel to the The Road ACP wall at a distance of c. 5m. As will appear from the following de The large ashlar wall ACP, first dis scription, walls ADQ, AFN, AFZ covered in 1997, was seen to continue stemmed from buildings preceding the in both trenches N26 and N28. After road, but two of the walls (ADQ and clearance of the surface soil surround AFN) were reused as a support wall ing the excavation units, the wall for a fill, which served to level the could be followed for at least a further ground for a road.

176 N26 N27

Empty stone hole 53 Pottery ES Mortar I Stone Mi Small Stones S3 Tile E3 Rocks

Fig. 4. Profile drawing. NNW In excavation units N26-N28, two lm were heavily packed with sherds of side of trenches N26-N27. Anne Hooton. wide trenches were opened between handmade coarse ware.1 Diagnostic wall ACP and walls ADQ and AFN. sherds among these date to the EH- These small trial trenches were exca MH period.2A handful ofsherds, how vated down to bedrock, providing a ever, dated to the LG - EPC period, SSE and a NNW profile of what was among which characteristic bases to be recognized as a road (Fig. 4). from kotylai with ray decoration can Between the two trenches only the sur be mentioned. It is significant that so face level was excavated. As seen far only one later sherd has been iden from the profile drawing, the NNW tified from this level. It dated to the trench was characterized by stone- Late Archaic or Early Classical period packed levels towards the bottom and and was found in the uppermost part of levels with fewer stones in the upper level 5. part. Level 4 thinned out towards the The two lowermost levels, 5 and 6 on Northeast and, since it had been cut the profile drawing, became thicker to away by the surface level, did not wards the Northeast and, when exca reach up to wall AFN.3 It contained vated, clearly appeared to have been some medium-sized stones, pebbles thrown up against wall AFN in order and many sherds, especially in its up to make an even surface. Both levels per and western parts.4 It is important

1Level 5 is equal to AFL/A, AFL/B and AFL/C in the diary from 13.07.99, and AFL/B 1st pass from 14.07.99, and AFL/B 2nd pass from 15.07.99. Level 6 is equal toAFL/5B third pass indi aryfrom 16.07.99. The very large bags of sherds from these levels contradict the description and appearance of the levels inthe profile drawing according to which they only contain a few sherds. 2Reference is made in particular to bag 3023 from AFL/5b 2nd pass from 15.07.99. 3Equal toAFL/4 west 2nd pass in diary from 14.07.99, and AFL/4 eastfrom 13.07.99 and AFL/4 west 1st pass. 4These characteristics were clearly observed and noted during excavation (as shows the level ling overview), but were difficult to see in the stratigraphy Especially, the numerous sherds excavated in this level do not show well in the stratigraphy due to the fact that they were more numerous in the western part of the level.

177 to note that the pottery from this level sherds was noted. These were definite was completely different from that in ly absent from the level below (level 4). levels 5-6. Thus, contrary to the mate rial from the just described lowermost Level 2 was covered by a pavement, levels, the majority of the pottery now AFL, consisting of irregular slab- consisted of medium and fine-ware stones preserved along the ACP wall. body sherds, often carrying a fugitive, When excavating the pavement in the dull red slip. As we shall see below, trial-trenches, a few Early Classical this type of pottery soon proved to be black-figured sherds were found, one very characteristic of Early Archaic of which shows the back of a chariot Chalkis. Diagnostic sherds consisted and a mounting figure. on the one hand of a few LG-EPC red- banded kotyle and kalathos sherds, on Almost the same stratigraphical situa the other hand of a small number of tion was found in the SSE trench. kotyle bases with ray-decoration and body sherds with polychrome decora Again a stone-packed fill was found tion (plain ware with encircling bands on the sloping bedrock, resting in red, white and black or red-slipped against wall ADQ, the eastern exten sherds with encircling white bands). sion of wall AFN. As in the NNW The general absence offigure-decorat trench, the stone packing contained ed sherds and the presence ofdiagnos much pottery. Again, handmade tic LG-EPC sherds point to the early coarse-ware body sherds made up the seventh century. In the following we majority (80-90%) of the pottery. The will refer to this type of pottery as diagnostic sherds date in the EH and "seventh-century pottery" acknowl MH periods. One sherd may be edging the fact that a more thorough Geometric. Above the stone packing study is needed to date it more accu were levels containing fewer stones rately. and sherds, which ran up to wall ADQ but not over it, providing an Level 3 contained virtually no pottery ante quern date of the wall. Just like and should probably be understood as in the northwestern trench, these lev part of the foundation trench for ACP, els were characterized by a somewhat marked "7" on the profile drawing. higher percentage of LG-EPC sherds (notably red-banded kalathos and Level 2 was characterized by many kotyle sherds). Still, however, the stones, pebbles, and a large amount of coarse-ware material was in majority the just described "seventh-century pot (60-70 %). tery".5 It had been cut offby the surface level in the NNE, and consequently In the superseding level (AFL/4d), the does not reach wall AFN. In addition to proportions of the datable sherd mate the "seventh-century pottery", a small rial were reversed, with very few amount ofvery poorly preserved black- Prehistoric sherds. The "seventh-cen glazed sixth- or (early?) fifth-century tury pottery" dominated, and the diag-

5Equal to AFL/1, AFL/2 and AFL/3 in the diary from 06.07.99, 02.07.99, and 07.07.99.

178 nostic material consisted mostly of lower level than the road. The room body sherds with polychrome decora was excavated down to bedrock. A tion. hard-packed clay floor (AFU) with tiny pottery fragments was found in The levels above (AFL/3d, AFL/2d, the room. A rich amount of "seventh AFL/ld) were thinner but very rich in century BC" pottery was found in the sherds. The levels were only preserved level immediately above this floor. along the ACP wall, having eroded Only two sherds may date to the sixth down the hill towards the East when or fifth century but these stemmed the superstructure of the supporting from the uppermost part of the level walls collapsed. Again the familiar covering the floor. The floor AFU it body sherds with encircling bands in self was best preserved in the middle red, black and white, but also black- part of the room and produced several glazed sixth- or fifth-century sherds small bags of EH-MH and LG-EPC dominate. sherds, including characteristic Early Helladic T-rims, Grey Minyan, and The stratigraphy and finds indicate that Geometricfirnis sherds. the foundation walls ADQ and AFN were constructed at some time in the Below floor AFU, a thick level was seventh century and constituted the out excavated (AFU/2), which produced er walls of a row of small buildings, several large bags of pottery. which lay on a natural ridge of the slop Diagnostic sherds were EH and MH, ing hill. In order to create an even ter some LG-EPC, and some perhaps race in front ofthe buildings, to the SE, from a later date in the seventh centu a stone fill was thrown up against the ry, but generally the prehistoric mate walls. The fill contained material (EH rial outnumbered the Archaic. and MH sherds) from earlier settle ments. Apparently, the seventh-century In trench N28 to the southwest, exca houses were erected in an area, which vation among other things revealed an had not been used since the EH and MH earthen floor, ANA, and a pebble periods. In the Early Classical period, floor, AEG. The pebble floor was the ACP wall (see below) was erected, found to cover the foundation wall and the area between this wall and the AFZ, and extended into the area later old seventh-century walls received a to be used for a road. A foundation slab-stone pavement constituting part of wall, AFY, ran parallel to the ADR a road along this part ofthe hill. wall, and met the AFZ wall at a right angle, forming a room. Obviously, the The Archaic houses mud-brick superstructure of the room, formed by the AFZ and AFY founda In the trenches N26-N28 to the tion walls, collapsed or was torn down Northeast of the sequence of walls and covered by floors belonging to a ADQ, AFN and AFZ, two dry-stone larger, possibly open space (Fig. 2). foundation walls, ADR and AFR, were The floors lay at a level approximate excavated. They joined the ADQ wall ly 0.5 m higher than floor AFU in the at a right angle, forming a room at a neighbouring room.

179 Below these floors, a thick level Fig. 5. Loom weight found in situ next to a fireplace in trench (N28/3) connected with the AFZ and N28. AFY walls was excavated, which, Henrik Frost. apart from a few Prehistoric sherds, appeared to contain only seventh-cen tury sherds. In addition, a fireplace (ANC) was found in this level and close to it a complete skyphos and a lekane, a grinding stone, and five iden tical loom weights (Fig. 5).

In the stratum below (N28/4), beneath walls AFZ and AFY, a drain (AND) was excavated. All the pottery seems to date in the seventh century and, as mentioned earlier, the general lack of figure-decorated pottery points to a Fig. 6. Fragment of a small crater with encircling bands date in the early seventh century. A and a frieze of vertical lines. small krater rim- and body sherd is N28/4a 1st pass, below the shown in Fig. 6. floors ANA and AEG. LG-EPC. Henrik Frost.

The Classical houses House knownfrom the AM wall

In trench Tx93 a well-built dry-stone foundation wall, ANI, oriented NE- SW and running at a right angle up to wall ACP and the road AFL was exca vated. The wall was built directly on found both in the soil above this slab- the bedrock (Figs. 2, 7). Pottery in the stone floor and in the stratum below lowest levels on both sides of this wall the paving, which also went below the appears at the time of writing to pro ACP wall. Finally, the ACP wall must vide an ante quern date to no later than have been built when the upper levels the fifth century. of road AFL were constructed in the Early Classical period. The ACP wall House knownfrom walls ANB, The same levels went below the ACP ADP, ADN wall providing a post quern date in the fifth century BC for this wall. Further In Tx92 and Tx93, foundation walls indications for the date of wall ACP built of ashlar blocks (ANB, ADP, may be provided by the excavation of ADN) (Fig. 7) were excavated, which, a paved floor (only a few slabs were to judge from their quality, originally preserved) lying up against the ACP must have formed part of a well-con floor on its southern side. Sherds were structed building. The building is ori-

180 the building must be in the fourth or early third century BC. Excavations inside the building during previous seasons had produced pottery prelimi narily dated in the same period.

In conclusion, since this building is ori ented towards wall ACP and the road, it must be laterthan the EarlyClassical pe riod and, according to associated levels, it is earlier than, or contemporary with, the Late Classical/Early Hellenistic peri od.

Extension ofthe "acropolis wall"

In Tx91 an extension of the so-called "acropolis wall" (ACS) was excavated (see SPR, 23Iff, figs. 4, 9, 15, 28, 30, 31). This wall rested on a thick soil fill, at the bottom ofwhich a fragment of a worked stone (grinding stone?) was found. This was similar to a more Fig. 7. Classical walls running ented NE-SW, at a right angle from the up to the road in trenches complete example found in a clean Tx20-22, Tx91-93. Henrik road and the ACP wall. It is separated context of Hellenistic date in area III. Frost. by a narrow lane from the neighbour From the same fill came diagnostic ing house, known from wall ANI de pottery including fragments of two scribed above. The length ofthe build lamps datable in the second halfofthe ing is 10m, judging from wall ANB, fourth century or the third century BC. which in the SW terminated in a cor The structure ACS is therefore inter ner (ADP), and in the NE by the road. preted as a reparation of the Classical At least one cross-wall (ADN) divided "acropolis wall", which took place in the building into two rooms. Another the Late Classical or Early Hellenistic lane was found running along this period. building in a NW-SE direction along the foot of the acropolis, and thus ran There were many large limestone ash parallel to the road (AFL), a little lar blocks in the thick fill covering the higher up the slope. Classical houses and lanes. These blocks must stem from the upper The lowest levels found in situ direct courses of wall ACS, and perhaps the ly on the bedrock in the area outside "acropolis wall" itself. Pottery from the corner of the building (ADP), in the fill is preliminarily dated to no lat the southern part of trench Tx91 and er than the Early Hellenistic period. along the ANB-wall "in the lane", This indicates that, by this time, the demonstrate that the earliest date of houses had been abandoned and activ-

181 ity moved to the western side of the es (structures ARJ and AQY). In K29, hill instead (area III in Fig. 1). a patch of a hard-packed pebble floor, mixed with smaller stones, pottery, Area II: Bronze Age struc bones, and shells (ARK and ARM) tures, seventh- and sixth- went up to the line of the boulders on its eastern side. The sherds from this century settlement remains floor were MH in date. A poorly built and Early Hellenistic cist foundation wall (ARH) ran up towards tombs the boulders in K29 but may be older.

In K27 and K28, traces ofa wide foun At the foot of the hill to the West, four dation wall (AQU and AQN), appar trenches (4x4m) were opened, named ently forming a corner, ran above the K26-K29. Trenches K27 and K28 lie MH stone packing. The wall partially on each side of the trial trench Tx72, traversed the above-mentioned line of excavated in 1998 (see SPR, fig. 22; boulders at a different orientation here Figs. 1, 8a-f). Since the stratigra (NE-SW). It was built of large, un phy was comparatively undisturbed, dressed stone blocks and preserved to and the successive phases accordingly a length of 7.5m. A floor, consisting of easy to define, the finds and structures medium-sized stones, was excavated from these trenches will be summa to the west of this wall. rized in chronological order, starting with the earliest remains. Levels with Pottery from associated levels and Neolithic sherd material were excavat floors indicated a date in the MH peri ed in K28/29. od, whereas LHIIIC, and possibly LHIIIA-B, sherds were found in the Bronze Age structures soil above the structure. EH-MH evidence Settlement remains from the In K28 MH strata were excavated in seventh and early sixth cen a roughly 1x 1 m large trial-trench turies BC (Fig. 8a). The strata consisted mostly of stone packing, with diagnostic The main period represented in the ar sherds ranging in date from the EH to chaeological remains in the three MH, and of a type published in an trenches K27-29 appears, at the time of earlier report (SPR, figs. 10, 20). writing, to be the seventh and early sixth centuries BC (Fig. 8). A complex MH and possible LH structures was excavated, which consisted of comparatively well-preserved house A later phase, dating in the later part of walls, separated by lanes and open the Middle Helladic period, was com spaces, and a series of floors. Pottery paratively well defined and encoun from the earliest floor levels is from tered in three trenches (K27-29). A the LG-EPC period, while the material line of large boulders, oriented NNW- in the latest floors could be dated in the SSE, could be followed over a dis MC period. The settlement in this area tance of about 12m in all three trench ofthe hill is therefore more or less con-

182 temporary with the room complex pre ceding the road on the northeastern side ofthe hill described above (area I). Another similarity is the fact that the earliest rooms were built directly on top of the Bronze Age remains, as was K29 the case in trenches N26-N28. In fact, one of the earliest walls followed the orientation of the latest Bronze Age building (Fig. 9).

LG-EPC settlement structures

Two wall fragments in K29 rested di rectly on top of the MH-LH levels (AQM and wall ARF). The orientation of the walls suggested that they once met and constituted a corner in an L- shaped building. In that case, the CD Early Hellenistic building is likely to have had several

Fig. 8. General plan of trenches K.26-29 with cist tombs GR3 and GR4 in area II. rooms. Flat stones, probably from a Charalambos Marinopoulos. doorstep, were found on the suggested extension-line of wall AQM, a couple of metres to the West. This circum stance further supported the presump tion that walls AQM and ARF origi f- ^h + + nally met and constituted a corner (Fig. 10). Patches of a pebble-floor (ARL) were found in several areas to the south ofwall AQM. The outline of K29 AREA II this pebble-floor was clearly seen to stop at the presumed corner ofthe two h + + walls while conversely no sign of the Tx70 floor was found "inside" the supposed

Tx71 rooms. Thus, the pebble-floor points to

K28 the existence a large open space ex tending south and west ofan L-shaped K26 building. f \ + Tx72 Among the diagnostic sherds from this pebble-floor (ARL), LG-EPC kotyle bases with rays, and sherds of "eggshell" ware should be mentioned I + + + (Fig. 11). An interesting fragment of a horned acroterion also came from this Fig.9. Bronze Agestructures in trenches K27-K29 in areaII. Charalambos Marinopoulos. floor. Pottery from the strata associated

183 with wall ARF included LHIIIC sherds as well as the familiar LG-EPC kotyle h ---"\ + + bases mentioned earlier (Fig. 12). t \

In trenches K27-K28, a very strong K29 \ (a wall (the lower part of AEE), built of AREA II flattish, uncut sandstones and preserved h + + to a height of about 1 m and about 8 m Tx 70 „„--''' \ long, rested directly on the MH-LH levels. The wall is oriented NW-SE and Tx 71 ^---""\ ,- '"' K28 \ B\ followed the orientation of the Bronze Age wall AQU. The southern end of \ --""""" fl'V -"-"f"* \ K26 this wall is visible in the profile and ap h Tx72 \ ^- '"B \+ \ + pears to form a corner. In that case the \ \ \ \ 1 1 AEE wall once constituted the western, \ • 1 \ v 10*0*** outer wall of a NW-SE oriented build K27 \ \ ing (Fig. 13). \

Settlement remains in the h + + + PCperiod Fig. 10. LG-EPC settlement remains in trenches K27-29 in area II. The whole area (trenches K27-29) be Charalambos Marinopoulos. came the scene of an intense building activity very soon after - if not at the same time as - the erection of wall + + AEE. The rooms formed by walls ARF and AQM (mentioned above) fell out Mouse I of use and instead three or four houses were built, and open spaces defined and paved. Two of the houses were AREA II separated by a narrow lane with a + + drain. Several floors associated with Tx70 the walls could be defined, which su perseded each other over a height of Tx71 ,- approximately lm, and pointed to con ouse IV.,--"' tinuous settlement activity, including K26 minor rebuilding. Here, only the most v+ characteristic early and late floor levels are summarized. Walls from this peri od are shown in Fig. 11, and a general view of the trenches is shown in Fig. 14. + +

The earliest floor level lay around

2.30-2.5m above sea level, and about Fig. 11. Building remains dated in the PC period in trenches K27-29 in area II. 0.2-0.25m above the Bronze Age lev- Charalambos Marinopoulos.

184 els. Several floors were found on this

— + + level. The large house, known from \ \ wall AKK (Fig. 11), replaced the L- \

\ shaped building and its open yard. This new building had first a paved K29#S£0 AREA II \ floor consisting of stone-slabs (ARD House II \ at level 2.35m) and could be entered - + + lane^-jj^^^^ Tx 70 „„-'''"' \ from the lane. Wall ARF from the pre ceding period appeared to have been

room H||\| • p*-"""^ \ used as a doorstep. K28 \ House HI sg V-"'"' ' --'"' \ ^hThiium IV?...-'--''"" \ K26 In the narrow lane between walls AKL and AKK, the old pebble floor Tx 72 \ \ + >M \ \+ \ i \ \ (ARL) from the previous phase was \ Open area ^ \ \ replaced by a new pebble floor \ AKC \\ '""'""' (AQJ/1 on level 2.35m). K27 \

^g^ Walls AKL, AEE, and AEH (Fig. 11) h + + + constituted the outer walls of a house lining the narrow lane on its southern Fig. 12. Building remains dated to the EPC-LPC period in trenches K27-K29 in area II. side. An insubstantial foundation wall (AEI), consisting ofonly one course of flatfish stones divided the house into two rooms. The earliest and best-de fined floor level associated with these A\ + two rooms again lay directly on top of the MH-LH structures. In a corner of the easternmost room, twelve loom weights were found close to one anoth K29 er.

•fc + Tx70 Not much later, the room with the loom was equipped with a more sub \ stantial floor consisting of irregular K28 \ slab stones (AQZ and ARA). Three K26 terracotta spools were found on this + + level, the type of which has parallels Tx72 \ dating in the seventh and early sixth centuries.

In the open area between this building

+ + + and the one known from wall AKE in trench K27 (see Fig. 11), a level mixed with stones (AQR/1) was found, im

Fig. 13. Early Hellenistic cist tombs and other remains of the same period in area II. mediately above the MH-LH remains. Charalambos Marinopoulos. It probably served as a foundation for

185 the floor (AQO) above it. The level contained LHIIIC pottery mixed with LG sherds. Floor AQO itselfwas char acterized by tiny sherds, some char coal, and many mud-brick fragments. ,; : "-i*,^ti; Besides pottery, a fine bronze pin was found. The room probably defined by f.\-iy': the corner wall AQG in trench K29 was also built in this period. % In conclusion, the finds from this se ries of floors at c. 2.35m above sea

level point to a date in the PC period. \z

Settlement remains from ^

the MC period • \ 1

in trench K29, the house known from * wall AQG received a new slab-stone floor (AQL). A kotyle sherd with let (A^T), characterized by some peb Fig. 14. General view of Bronze Age structures and Archaic set ter-imitating decoration was found in bles, tiny sherds and hard-packed tlement remains in K27-K29. the level immediately above this floor. soil, shells and bones. From this floor Henrik Frost. came among other things two deco In the neighbouring house, known rated sherds, which still require a from wall AKK, and at approximate more precise dating. ly the same level, the plain ware jug with threefoil mouth shown in Fig. In the lane in between walls AKK and 15 was found. A sherd from a closed AKL, two terracotta sima fragments vase, decorated with a siren or bird, were found at level 2.66 m, just below is datable in the MC period, and a well-defined floor. In the house con came from the same level. On a stituted by walls AKL, AEH and AEE, slightly higher level in the same minor rebuilding took place. The di house, an earthen floor (BK/4) was viding wall AEI was given up, and the found to have replaced the earlier southern outer wall (AEH) was re pavement (ARD). From this floor placed by a new wall (AEF), making came a Middle Corinthian krater rim the house narrower. Perhaps this hap with lotus-palmette decoration (Fig. pened in order to enlarge the open area 16), and other diagnostic sherds and between walls AEF and AEK, which finds, likewise indicating a date in was now paved with large, neatly fit the MC period. The floor was soon ting slabs (AKC) (Fig. 17). This pave replaced by another earthen floor ment was found in the middle area of

186 Fig. 15. Jug (FO1-3002) found trench K27. However, since the upper on the floor in the room formed by wall AKK in trench courses of the southern part of wall K29. Henrik Frost. AEE had been taken down to the level of the pavement, the latter may origi nally have continued towards the East. In that case, both house AEE and house AKL, AEH and AEF were re built, in order to give way for a larger, open and finely paved area.

Archaic polychrome sherds were found at the level of this pavement, and a lekane sherd, and several py ramidal loom weights came from the level covering the pavement. Excavation of the pavement itself, however, produced among other things sherds from an open vase dec orated with rosettes and the forelegs of a panther, the style of which Fig. 16. Rim from crater with lotus and palmette decoration points to a date in the MC period (F00-5987). From floor in (Fig. 18). room defined by wall AKK in trench K29. Middle Corinthian. Henrik Frost. All in all, activities at this level in cluding the pavement of large open area seem to have taken in the MC pe riod. Fig. 17. Pavement consisting of stone slabs (AKC) in K27. To the right wall AKE, to the At a still higher level, around 3m left wall AEH, and in the background wall AEE, the above the sea, several earthen floors southern part of which has with sub-floors consisting of hard- been taken down to the level of the slabs in Antiquity. packed medium-sized stones were Henrik Frost. found. One such sub-floor (AQS) Fig. 18. Sherds from closed contained a sima-fragment and diag vase decorated with rosettes and forelegs of panther. From nostic pottery. In the lane in between the slabstone-pavement walls AKK and AKL, yet another (AKC) in trench K27. Middle Corinthian. Anne Hooton. earthen floor was found (AQA and AKM), characterized by the usual small pieces of charcoal, pebbles, sherds and tiny potsherds. Finally, in the strata covering wall AKK, a sherd with compass-drawn cable pattern was found (Fig. 19). However, more study is needed of the material from these later closed contexts before a date can be given.

187 Early Hellenistic tombs Fig. 19. Medium fine-ware sherd with compass drawn ca ble pattern (F00-5074) found After the last Archaic settlement in level covering wall AKK in trenches K28-K29. Henrik phase described above, the houses in Frost. trenches K26-29 appear to have been given up. The strata immediately on top of the Archaic horizon mainly contained mixed eroded material and no structures. By the Late Classical period, the area was turned into a cemetery for well-to-do inhabitants, something which had already been suspected in 1998 when the limestone elongated slab stones covered the structure AFJ in trench Tx71-72 ap eastern end of GR3, and a third slab peared (SPR, fig. 25), and it was con stone must originally have covered its firmed in the 1999 season when the western end. While GR4 was found trial trench was widened and com looted, the burials in GR3 were found pletely excavated (and called trench undisturbed. K26). Thus, the limestone structure proved to be one of two large, and Conclusions very finely constructed cist tombs. The construction of the tombs closely The first use of the Hill of Hagia resembles GR2 found during rescue Triada was a habitation in the excavations earlier in 1999 west ofthe Late/Final Neolithic period. Above Kato Vasiliki road at a distance of those early levels there was an impor about 2km from the Hill of Hagia tant Bronze Age settlement, which Triada (I. Moschos in SPR, 291-301). ceased to exist in the Dark Ages. In the Both tombs are published in a sepa late eighth or early seventh century, rate article by Jonas Eiring in this vol settlers laid the foundations of a new ume, and are therefore only very town. The comparatively well-pre briefly described here. served Archaic buildings with associ ated finds on the western and north The two cist tombs (GR3 and GR4), eastern side of the hill indicate a town approximately parallel to each other, of a considerable size, which must are oriented East to West (GR4 slight have come into existence during a rel ly more towards NNW-SSE). Only atively short period of time and which the lowermost courses of GR4 were was founded, in parts, directly above preserved while GR3 was undis the Bronze Age remains. The presence turbed. Both tombs are built of large of Early Hellenistic graves on top of limestone blocks, which were finely the Archaic settlement indicates that cut and smoothed on the inside. The the extension of the Archaic city may eastern side of the tombs was formed have been decidedly larger than the by huge, roughly square slabs. Two Hellenistic city. Preface

Soren Dietz, The Greek-Danish archaeological col The campaign in 1999 lasted from 28 Lazaros Kolonas and laboration at ancient Chalkis in June to 30 July. Excavations continued Michail Petropoulos Aetolia was carried out between 1995 on the middle terrace below the Acro and 2001, with the Danish Institute at polis, towards the northeast, where ar Athens and the Sixth Ephorate of chitectural remains from the Archaic Antiquities in Patras as responsible and Classical periods (trenches N26- participants. Seven campaigns of N28) previously had been attested. fieldwork included surveys, excava Excavations continued also further tions and geological examinations in east on the same terrace (towards the the area around the present coastal vil water), where foundations of a row of lage of Kato Vasiliki. The work was Hellenistic houses had been uncovered concentrated to the site of Pangali on during previous campaigns. The interi the eastern slope of Mount Varassova or walls of these houses were defined and, in particular, the small hill of by an earlier fortification wall, proba Hagia Triada on the coastal plain east bly of Classical date. ofthe village. Two preliminary reports have previously been published. The Already in 1998, trial trenches (Tx70- first covered the first two campaigns 74) had been opened on the plain west of 1995 and 1996, and was published ofthe hill, where the geological survey in 1998 in the Proceedings of the had attested a bay in Antiquity. Hence, Danish Institute at Athens II, 232-317 the area is near the ancient waterfront (= FPR), the second was published in and, presumably, the harbour men PDIA III (2000), 223-305 (= SPR) and tioned by Thucydides. A series of lev included the activities of the years els were excavated down to bedrock, 1997 and 1998. The present prelimi almost four metres below the present nary report, which is the third and the surface. The main groups of material last in the series, describes the work were given preliminary dates in the carried out at ancient Chalkis in 1999, Late Neolithic, Bronze Age and 2000 and 2001. Archaic periods. In 1999 three new trenches (K26-28) were opened, and a In addition to the preliminary descrip good stratigraphy of succeeding tion ofthe results from the three years, Archaic levels was uncovered. Two the final publication of the Hellenistic Hellenistic cist tombs were found in tombs excavated in 1999 is presented trench K26, one of which was undis separately in this volume (Eiring). turbed. The tombs were excavated Further results from the project are be during the 1999 campaign. In 2001 ex ing studied for final publication during cavations were restricted to the area the next four years. west of the hill (trenches K27-29).

171 The participants in the three cam Nielsen and Ioannis Moschos. paigns were the following: Pottery Registration: Jonas Eiring and Camilla Schneekloth Sorensen 1999: Geology: Kaj Strand Petersen Project leaders: Soren Dietz and Tiles: Claus Gronne Lazaros Kolonas Photographer: Henrik Frost Field Directors: Sanne Houby- Conservator: Leonidas Pavlatos Nielsen, Alexandra Nilsson, Ioannis Students: Pernille Bangsgaard, Peter Moschos and Michail Gazis. Rose, Julie Maria Mortensen, Pernille Surveyors: Charalambos and Marinos Andersen, Lasse Sorensen, Emma Marinopoulos. Ljung, Frederik Olsson, Daniel Pottery registration: Jonas Eiring, Sahlen, Annika Jeppsson Elizabeth Bollen and Hildegunn Logistics: Jane Berg Jorgensen and Borup. Petra Pakkanen Geologist: Kaj Strand Petersen Tiles: Claus Gronne 2001: Photographer: Henrik Frost In 2001 the main activities were con Conservator: Leonidas Pavlatos centrated to Ancient Calydon (spon Students: Peter Rose, Lasse Sorensen, sored by the Ny Carlsberg Pernille Bangsgaard Jensen, Julia Maria Mortensen, Trine Eltang, Sara Foundation). At Chalkis, Soren Dietz Birgitta Engblom, Pernille Andersen, and Michail Petropoulos were project Camilla Schneekloth Sorensen, Mette leaders with the participation of the Kjaer Schaldemose following students: Pernille Andersen, Logistics: Ann Thomas Emma Ljung and Mikkel Mayerhofer. Claus Gronne was responsible for the 2000: registration oftiles and Mette Project leaders: Soren Dietz and Mouritzen, architect, continued regis Michail Petropoulos. tration of the fortification walls on Field Directors: Sanne Houby- the Pangali.

Acknowledgements

The field work at Ancient Chalkis is a Gosta Enbom Foundation, which, as joint project between the Sixth during previous campaigns, covered Ephorate of Prehistoric and Classical the expenses of the Danish participa Antiquities in Patras and the Danish tion. The Council and Mayor of Institute at Athens. We are most grate ful to the Greek Ministry of Culture Gavrolimni kindly made the old for permission to carry out work at school in Kato Vasiliki available to the Chalkis and to the Consul General expedition for storage and research.

172