1The present volume includes an account of two seasons of excavations that have taken place on the Bouffo or Kephali tou Ag. Antoniou at Sissi by the Belgian School at Athens, the first campaign from July 6th to July 27th, 2007, the second from July 14th to August 23rd, 2008. The excavations were directed by J. Driessen of the Université Catholique de Louvain, in collaboration with I. Schoep of the Katholieke Universiteit Leuven. We are extremely grateful to V. Apostolakou and V. Zografaki of the KD’ Ephoria for a splendid collaboration. S. Michou (2007) and L. Manousogiannaki (2008) represented the Ephoria during the work. Funding was provided by the Institute for Aegean Prehistory, the Université Catholique de Louvain, the Fonds National de la Recherche Scientifique, the Fonds voor Wetenschappelijk Onderzoek and a series of private sponsors to whom we are all very grateful – their names are mentioned explicitly on our website http://www.sarpedon.be.

2We are also grateful to the rectors of our universities, Prof. B. Coulie (UCL) and Prof. M. Vervenne (KUL) as well as Prof. R. Keunings (head Research Council UCL) and Prof. H. Bouillon (Dean of Arts, UCL) for their support. Site recording was greatly accelerated and facilitated in 2008 by the use of toughbooks, six site computers made available at low cost by Panasonic Belgium.We would also like to thank the Ecole française d’Athènes and its director, D. Mulliez, for putting at our disposal the apothikes of the dig house at Malia where our material is still stored. E. and K. Kivernitakis and A. and M. Moulara we thank for their Cretan hospitality (including the loan of a pick-up truck) and the mayor of the kinotita Vrachasiou, Mr. G. Sfirakis, and Mr. V. Prineas and M. Mamaloukos of the Bouffo Beach Bar, for making site work more agreeable. Mr. Christos Mamaloukos facilitated our life in general. We are also very grateful to Christope Gaston (EFA) for an excellent series of aerial and vertical telescopic cane photographs, to INSTAPEC and its director, Dr. T. Brogan, for the use of two site sieves, and to the IMS at Rethymnon and Dr. A. Sarris for the rent of a total station during both campaigns. Mr. P. Iossif, secretary of the Belgian School at Athens kindly took care of IKA and other administrative issues. In 2008, we were also the focus of a documentary of the French speaking Belgian television (RTBF: La Une, Au Quotidien: M. Ernst, J. Renard). Mrs. A. Debroux of the UCL is thanked for helping with various printing issues, Mrs. B. Vervaren (UCL) and Mrs. I. Lermuseau (UCL) for help with accounts.

3In toto about 35 archaeologists and students, mainly from the UCLouvain and the KULeuven but also from Brus-sels, Paris, Ghent and Saloniki Universities, as well as 12 local workmen, with Giorgios Metaxarakis as foreman, were involved in the two campaigns (see figs. 1.18 & 1.19). First Prof. C.J. Knappett (Un. of Toronto) and then Dr. C. Langohr (UCL-FNRS) directed the work in the apothiki, assisted by F. Liard. A. Aertssen (2007, 2008) and C. Delmotte (2008) took care of logistics. Soil sampling (dry and wet sieving) was done throughout through the care of Ms. S. Michou in 2007 whereas in 2008 Dr. Q. Letesson (UCL-FNRS) supervised the soil flotation of some 1500 litres of samples by P. Baulain, C. Digiovanni, V. Vanderlinden and C. Farnoux. In 2008, Dr. V. Isaakidou made a preliminary assessment of the environmental work and C. Baeteman helped in the geoarchaeology. Initial restoration of moveable objects (mostly pottery, a few metal pieces) was performed by Mr. C. Zachariou in 2007 whereas Alekos and Nearchos Nikakis expertly mended a series of vases and took care of the consolidation and removal of the lead amphora found in zone 5 during the 2008 campaign. All photography, including DVD–recording of site work, was done by P. Zintzen in 2007 and by H. Fiasse and L. Manoussogianaki in 2008. A special data base for the project was created by H. Fiasse, who also crafted our website. Objects were drawn by A. Rymenams, assisted by H. Joris (2008). The INSTAPEC photographer, C. Papanikolopoulos, came over to photograph objects (2008). We are grateful to H. Prokopiou for advice on stone tools and to our visitors (A. Farnoux, V. Zografaki, C.F. Macdonald, J.A. MacGillivray, S. Hemingway, L.H. Sackett, S. Muller, D. Panagiotopoulos, R. Docter, D. Viviers, K. Nowicki, S. Wallace, T.F. Cunningham, S. Soetens, H. Matthäus, D. Klontzas, J. Germain, M. Guterres, H. Bouillon, C & A. Dehollander, G. Markakis, MP M. Souladakis, the Zouraris family, Papamichalis, and a score of interested locals) for interesting observations and assistance.

1. Location and interest of the Site

4The Kephali hill is located about 500 m east of the present modern coastal village of Sissi, (Σίσι), itself about 4 km east of the palace of Malia (figs. 1.2 & 1.3). It is one of the villages that make up the present kinotita Vrachasiou (Eparchia Mirabello, nomos Lassithi) together with Vrachasi, Milatos, Epano Sissi and Latsida. The hill has an extent of about 3ha and is up to 22 m high. It slopes up gently from north to south, but its east, west and south edges present relatively steep ridges that have obvious defensive qualities. The river coming out the Selinari gorge has an outlet on both sides of the hill, now artificially turned into sandy beaches called Bouffo(s) to the west and Avlaki to the east. The summit plateau provides a fine view over the coast towards Malia east and Milatos west as well as the hinterland towards the Selinari gorge and Selena Mountains, the foothills of the Lassithi. Both coastal and inland roads passed here in ancient and more modern times, giving the site an obvious strategic importance (fig. 1.3). This and its proximity to the palatial site at Malia form two of the research issues especially directing the work. The location of the Kephali hill is of utmost importance since it controls="true" two land routes from west to east Crete that were undoubtedly important both during prehistoric and historical times.

1. The coastal route from Malia to Milatos: travelling from west to east, the hill stands directly across the coastal route connecting Milatos with the west and has been actually traced from the Malia palace up to the west side of the modern Sissi village (Müller 1991: 551), north of the Arkovouno hill.
2. The inland route from Malia to Neapolis passing the Selinari gorge. The Kephali hill is located exactly opposite the northern entrance of this gorge and the stream bed leading through the gorge abuts on either side of the Kephali.

5This may imply that both routes from the east actually converged at the Kephali, where they became one route that continued westwards. If this were the case, then the Kephali hill may have been a node in an interregional network and excavation will throw light on network dynamics in this part of the assumed ‘Malia State’ (Knappett 1999).

6Our work had also partly the nature of an emergency excavation. Despite the fencing of the site by the Ephoreia in 2005, both locals and tourists regularly damaged the enclosure and trespassed, the first mainly in their search for chorta and for pasturing sheep, the second using the site as a shortcut to reach the better beach at Kharkoma1. Trespassing and erosion resulted in revealing pottery fragments and human bones in connection to damaged architectural features in zone 1, the funerary area2 (cf. fig. 1.4).

2. Earlier Explorations

7The Kephali tou ag. Antoniou (after a small chapel immediately south of the hill) or Bouffo hill (after the west beach that received its name from the noise made by the waves into a submerged cave against the north side of the hill), has been known for some time in travel and archaeological literature. Sissi (as Sisi) is first mentioned in Venetian records of 1391 (Spanakis 1991: 714-715) and old Venetian (Boschini [1651], Coronelli [1696]) and Dutch maps (Mercator [1590]; J.B. Homann [1704]) indicate a place called Sissi or Sisi, sometimes east of Milato(s), sometimes west of it (Sfiroeras 1985, Melas 2005). It is possible that one cartographer was copying the fault of another but it may perhaps not be excluded altogether that the toponym travelled around within this limited area. Marinatos (1949, 11-16) at one time suggested that Milatos was the original name for Malia which eventually moved east. Sometimes another toponym, Ciccalari(a) is found on maps at the spot where Sissi is located (Spanakis 1991: 715, for some other references). Since the founding of the mountain village at Vrachasi against the Anavlochos – also first mentioned in 1391 (Spanakis 1991: 202) – Sissi served as a maritime outlet for its produce. Local informants mention that the Bouffo and Kharkoma beaches were used during Ottoman times by sailing boats and that a signalling post on the Anavlochos existed to warn Vrachasiotes of their arrival. Present topographical and sea current conditions are not favourable to the use of the Bouffo and Avlaki beaches as safe anchoring places for sea crafts. It is, however, not impossible that the site served as a secondary harbour in Minoan and historical times. A proper geomorphological diachronic reconstruction of the coast-line may elucidate this point in the future.

8Milatos, the village east of Sissi, receives a mention in the Iliad (2.647) as being one of the seven Cretan cities that sent contingents to the Trojan War. Sarpedon, brother of Minos and Rhadamanthys, is also generally associated with this area of the island since myth has it that both Minos and Sarpedon were in love with a boy called Milatos, who preferred Sarpedon (Herodotos I, 173). Minos, in his anger, exiled both and they ended up in Asia Minor, Milatos as founder of Miletus, Sarpedon as king of the Lycians (cf. Marinatos 1949: 16). It is usually thought that his son or grandson fought with the Trojans and, when killed by Patroclos, was carried away by Thanatos and Hipnos, a theme quite popular in later Greek vase painting.

9In 1865, Captain Spratt mentions the creek at Seese (Sissi) as well as the existence of a toponym helleniko which, according to the locals, indicated antiquities (Spratt 1865: 112).

10In the fifties, Paul Faure (pers. com.) visited the area during his explorations of the island, noting the presence of a large, natural cave east (spiliada) off the Bouffo – the same that was later used as a cattle mandra by a local cattle farmer Voidas and now used as a fancy restaurant by the Kalimera Kriti hotel. In one of his later publications, Faure remarks the following:

A 500 m au Nord de Kato Sisi (Mirabellou), en bordure de la mer, sur une terrasse rocheuse qui la domine d’une dizaine de mètres, dite Kremasma, ont été ramassées en 1929 un certain nombre de têtes humaines, très sommairement modelées dans l’argile, certaines pourvues d’une sorte de couronne (BCH 1929: 529). Mr. A. Dessenne qui les a étudiées en 1949 (BCH 1949, 307-315) les a justement attribuées à un enclose sacré, d’une vingtaine de mètres carrés, sans fondations visibles. Il proposait de les dater du MR 3 en raison des tessons de pithoi, d’amphores et de coupelles trouvés au voisinage. Mr Davaras, qui a visité le site en 1963 (Kr. Chr. 1963, 405), n’a identifié qu’un dépôt de tessons et de figurines d’époque subminoenne, ce qui paraît aussi vraisemblable étant donné la facture grossière de la plupart des visages. A 300 m à l’Est, au lieu-dit ormos Charkoma, la légende veut qu’on ait trouvé un chaudron de cuivre3. Ce petit mouillage est dominé par la butte de Kefali, au Sud, couverte de tessons et de murailles MR 3, semble-t-il (alt. 25 à 30 m). Dans l’angle oriental du mouillage, au lieu-dit Boufos ont été découverts en 1962 une figurine masculine d’argile à étui phallique et un vase trilobé (Kr.Chr. 1963, 399). Rappelons que Sisi a été un petit port très actif jusque vers 1914 (Faure 1967 : 142).

11Faure added that the name of the hill to the east, Kaminia, would also indicate an area of ancient tombs and that Sisi is a prehellenic name (Faure 1967: 142-143), something also underlined by Spanakis 1991: 715). It is sometimes connected to the city of Sis in Asia Minor and probably indicates a place where water rushes out (P. Faure per litteras 31.5.2007). The ritual deposit mentioned by Faure in the area to the west of the hill, at Kremasma, was found by Demargne in the 20’s (Demargne 1929: 529; Dessenne 1949: 307-315) and then dated to Minoan times. It was said to belong to a site at Agios Antonios where walls, sherds and galopetres were collected. Davaras (1963: 405; 1964: 442) found a further deposit of Subminoan and Protogeometric sherds and figurines within a built cist with associated wall to the west. A Protogeometric figurine handed in to the Iraklion Museum (Davaras 1964: 442) is also said to have been found here (cf. Tsipopoulou 2005: 67).

12As epimelete of antiquities, Kostas Davaras (1963: 405; 1964: 442) conducted some test trenches in the early sixties on the Kephali itself which resulted in the hill to be classified as an archaeological site. He especially explored the northwest terrace of the hill – our Zone 24 – and made a small sounding on top off the hill – our Zone 4.

13About the same time, the area was visited by Sinclair Hood, Peter Warren and Gerald Cadogan5. Peter Warren (per litteras) recalls:

On 17 February (1963) we explored the Sissi area and examined several sites (I have been consulting the records in my travel diary). The main site we explored was, wait for it, Aghios Antonios Kephali. My sketch map notes extensive megalithic walls, with a fine section on the west side of the hill, a small construction on the north, traces of walls on the east and south-east slopes, and «Minoan sherds everywhere». I noted it as «a large Minoan settlement, previously unrecognized as such. Kylix stems and rims among the pottery». We also examined other sites east and west of Kephali, including (1) a Minoan shrine! about 150m W of Kephali and 40m from the sea, on a small akrotiri. It was at Kremasma or Kremasta (I wrote both spellings, just one will be correct), I have some details; (2) a really good EM-LM site at Khalkoma on the north-east corner of a bay right on the coast c. 600m east of Sissi. It was being cut away by the sea and there was a wall projecting out of the north-west of the site, right at the sea. There was very good EM pottery.

14The pottery then collected is still kept in the Stratigraphical Museum at Knossos and includes especially EM wares.

15In the late eighties, a small round limestone column base was found on the hill and moved to Vrachasi where it was sitting next to the main road (at the Vamvasakis house) till quite recently. When a road was opened to the Kalimera Kriti hotel, the south tip of the hill was cut away and remains of vases including pithoi were seen by locals. The previous mayor, Giorgios Markakis (‘Markogiorgo’), pointed out where the original kalderim ran and some of it, against the south of the hill, still floats about 4 m above the present road surface. The modern road cutting still shows abundant remains, now protected by a fence. Klontzas (2009) gives the two Greek laws, responsible for preserving the site. The first, in 1976, the second in 1996.

ΦΕΚ 699 του 1976: «…χαρακτηρίζομεν ως αρχαιολογικό χώρο στην Ανατολική Κρήτη, τον γεωμετρικόν και ρωμαικόν οικισμόν παρά το Σείσι, και τον μινωικόν οικισμόν εις ύψωμα Μπούφος – Κρέμασμα»

ΦΕΚ 597 του 1996 «…για την αποτελεσματικότερη προστασία και ανάδειξη του αρχαιολογικού χώρου «Μπούφος» Σεισίου κοινότητας Βραχασίου, η περιοχή καθορίζεται ως αδόμητη Ζώνη Α απολύτου προστασίας και επιτρέπονται μόνον οι ανοικτές, με αβαθή άροση, καλλιέργειες και χωρίς χρήση μηχανικών μέσων και μόνον εφόσον έχουν την προηγούμενη έγκριση και επίβλεψη της αρμόδιας ΚΔ’ Εφορείας»

16From 1990 onwards, when co-directing the excavations of Quartier Nu at Malia, the Bouffo hill came to my attention as one of the major secondary settlements in the area. In 1993-94, I applied for a survey permit (together with Dr. J.-P. Crielaard of the Dutch Institute at Athens) which was declined, however, but preliminary explorations helped to document the nature of the remains. It is then that the cyclopean-like wall to the southwest of the hill was first noted. During the 90s too, a survey team of the Ecole française d’Athènes under the direction of Sylvie Müller also visited the area (cf. Blackman 1997: 110; Müller 1991: 551, fig. 10; Müller 1996). They mention that Hellenistic and Roman sherds were visible in the road cutting (including black glazed and terra sigillata sherds (cf. S. Müller in Blackman 1996: 110). At around the same time, the KD’ ephoreia conducted investigations at Kharkoma (including the identification of a partly submerged Prepalatial site east of the hill in the area of the Kalimera Kriti Hotel) which is probably where most of the sherds kept at Knossos came from. It may be mentioned that there are remains of a Venetian aquaduct in the hinterland of Sissi, close to the entrance of the Selinari gorge.

17Some lady members of the French School, including Marthe Oulié, toured the island in 1929 and left us one of the rare pictures of the original bridge in the Selinari gorge in the National Geographic Magazine of the same year (information courtesy Leonidas Klontzas) (Oulié & Jean-Brunhes 1929) (fig. 1.5).

18The Kephali hill was, according to some of our local informers, also a point chosen by occupying Italian forces during World War II to install some light machine guns6 and when Italians soldiers were killed offshore by the Nazis, some of the bodies apparently washed ashore and were buried locally, especially at Avlaki.

3. The 2007-2008 Campaigns

19The 2007 campaign started off with a geophysical and archaeological survey to detect the most promising archaeological areas. From the 9th to 12th of July 2007, a geophysical survey was carried out, under direction of C. Hübner (GGH - Solutions in Geosciences Geophysikalische Prospektion, Freiburg) assisted by M. Kampfer7. This covered a total surface of 5560 m² (0,556 ha), consisting of six separate topographical units (fig. 1.6). The GPR survey concentrated on the flat terraces of the hill since elsewhere exposed bedrock or vegetation made such work impossible. A georadar GSSI SIR 3000 with 200 MHz Antenna and kilometres counting wheel was used. The data were prepared for a GIS-environment and handed over to the archaeological team. As such, in connection with excavation results, structures can be inserted within subsurface time and depth lapses of c. 15 cm, 30 cm, 50 cm, 70 cm and 100 cm. In the 100 cm depth lapse most areas show the subsurface bedrock. In general, the georadar survey has given good results in surface proximate zones that need confirmation through further excavation.

20At the same time, P. Hacigüzeller, helped by L. Van Cauwenberghe and E. Mlokie in 2007 and by P. Baulain and A. MacGillivray in 2008, started a full topographical and architectural survey to inventory the existing remains and natural features with a total station8. All results are included in a GIS-data base to direct future work. It may be mentioned that a stray, large, plastered ashlar ammouda block was found on the south summit of the hill.

21We also organized a pick-up survey within the fenced area which resulted among others things in the retrieval of a sherd with a Linear A sign (fig. 1.8) and what is, for the moment, our earliest ceramic evidence, an EM IIA sherd, incised with dots and lines (fig. 1.9). Surface material, especially from the road cutting to the south, also includes LM II material (fig. 1.10). A fragmentary pierced loom weight with the remains of at least three very eroded seal impressions was also picked up (fig. 1.11)

22At the end of the 2007 and 2008 campaign, a full photographic coverage was made, including vertical telescopic cane and aerial kite photography by C. Gaston, then at the EFA. Figure 1.14 shows the north part of the hill in 2007 and figure 1.15 the entire hill at the end of 2008, before covering the trenches with geotextile and earth. Since we worked in extension rather than in depth, no walls as yet needed conservation. We invested in the repair of the fence and made a new entrance (with sign) to the south of the hill, away from the river and sea which had destroyed 50 m of the fence during the winter of 2007/2008.

23During the three weeks of excavation in 2007, four different areas of the hill were tested: the lowest terrace flanking the sea to the north (zone 1), where the funerary area is located, the middle terrace to the northwest (zone 2) where K. Davaras had revealed part of a rectangular building, the north summit of the hill (subdivided for practical reasons in zones 3 and 4) and the southwest slope of the summit (zone 5). In each case, the decision to open trenches was guided by the presence of standing architectural remains. In 2008, work continued in each of these zones (fig. 1.16). In addition, we opened a new trench on the east summit (zone 6) where there were impressive walls showing. During the 2008 campaign, we cleared some of the shrubbery hiding some of the architecture to facilitate planning. In the early nineties, part of a wall with large blocks had been noticed against the southwest foot of the hill but this had in the mean time been entirely overgrown. This was now cleared from overgrowth and turned into a massive cyclopean-looking wall (fig. 1.17), forming a corner against the southwest foot of the hill. The north-south wall is still about 2 m high and preserved for almost 10 meters. A Neopalatial pottery deposit was stratified against this wall but a fragment of a Roman amphora was also recovered. Since the slope here provides the easiest access to the hill, the construction of the walls may have been intended to serve as a defensive bastion. Future testing should clarify this. Finally, a 2 by 2 m trial trench (BG99) was opened in the valley east of the Kephali hill, within the archaeological zone. This forms part of a broader program of study aimed at characterising landscape changes around the hill during the Neolithic and Bronze Age (Jusseret, PhD in progress). This also implied a series of hand-operated auger cores taken by Jusseret in the valley southwest of the hill during the 2008 season, under the supervision of C. Baeteman (Belgian Geological Survey) and with the help of N. Kress (UCL).

4. Some Preliminary Results of the 2007-2008 Campaigns

24We may try to incorporate our results, discussed in detail in this volume, with what was known about the site from earlier explorations. First of all, it may be mentioned that Neolithic occupation has hitherto not been identified at the site but its presence is not unlikely considering the topographical and environmental conditions, with a low hill dominating a fertile, well-watered coastal valley. The closest Late Neolithic site thus far known is situated about 2 km to the southwest, on the south slope of the Arkovouno hill (S. Müller in Blackman 1996: 109).

25Occupation on and around the Bouffo may have started in EM I/II, probably first at Kharkoma, about 150 meters to the east where Hood and company found remains of a submerged settlement. The cemetery of this hamlet may already have been situated on the Kephali hill’s northeast lower flank where a series of rectangular structures can be made out with heavy concentrations of Prepalatial pottery and human bones. The remaining wall lines seem to imply at least a dozen House Tombs – a type known from settlements more to the east (Mochlos, Gournia, Palaikastro, Malia and now also Petras) but unusual in Central Crete (Soles 1992). Since the earliest House Tomb dates back to EM IIA and this is also the date of the earliest surface sherd collected (fig. 1.9), this may imply that some habitation existed on the hill from this moment onwards. Thus far, however, very few EM II sherds have been found elsewhere on the hill although the French School survey team collected sherds from all Bronze Age periods (Blackman 1997: 110). The cemetery continued to be used during the late Prepalatial Period (EM III/MM IA). This may also be the moment that larger parts of the summit of the hill were used for habitation since sherds were found in areas 4 and 5. In zone 4, they were found stratified against the lower course of a facade made of large blocks, in zone 5 within some rock-cut basins. The evidence for a potential major building during this period is still slight but would, if confirmed, be interesting in view of the presence of a palatial building at Malia from this period onwards.

26During the Protopalatial and Neopalatial periods9, the Kephali hill is assumed to have been the site of a small settlement that was a secondary site in the Malia State, in scale perhaps comparable to Myrtos-Pyrgos on the south coast. Protopalatial sherds are the most common on and around the hill (as well as on Kremasma to the west cf. Müller 1991: 551, n. 7), suggesting that, as at Malia, the Protopalatial site may have been the more extensive one. Excavations have not yet confirmed this, however, since as yet few Protopalatial remains have been found. The cemetery continued to be in use in MM II, at least for the deposition of a pithos burial as well as what may be the remains of other activities related to burial (in zone 1.1). It seems to have been abandoned afterwards. In zone 2, a layer of redeposited burnt Protopalatial destruction material may imply that the site suffered fire destruction. The reused Protopalatial column base found in zone 5 suggests a building of some standing. Future excavation may show to what degree the material culture of the site is similar to that of the main palace site, 4 km to the west, and inform us how its inhabitants related to those of the palatial site.

27Before excavation, Neopalatial sherds were relatively rare and limited to the road cutting made when the Kalimera Kriti Hotel was constructed (Blackman 1997: 110). Our excavations have shown that the Prepalatial structure on top of the hill (in zones 3 & 4) may have been patched up in LM I but the nature of the building still needs to be clarified. The presence of fine pottery and a foundation deposit may imply a building with some standing. During this period, at least one workshop, perhaps involved in textile production, was constructed on the middle terrace to the northwest, the so-called ‘Davaras Building’. There are as yet no traces for fire destruction within this period. Local ceramic production follows some types and fabrics already known at Malia but with enough originality and differences to show at least some cultural independence at this and other moments. The high degree of quality suggests the presence of social groups of a certain status and/or the participation of the Sissi population in at least some extra-ordinary networks.

28Like Malia, Sissi has LM II occupation which is rather surprising in view of its rarity elsewhere on the island but perhaps understandable in view of the strategic location of the hill. As yet, no clear layers or architecture can be connected to this phase, however.

29During the French School survey, postpalatial (LM III) chamber tombs were localised in the hill to the south of the Kephali (Blackman 1997: 110). These may belong to the habitation that existed on the Kephali hill since, within zones 3,4,5 and 6, all areas located on top of the summit, both LM IIIA and LM IIIB pottery is attested. As yet no evidence exists that the lower terraces were used during this period, apart from some pottery found in the cemetery. The actual terrace walls which delineate the west and north edges of the summit of the hill are now seen to hide the original outside walls of a major building, showing typical Minoan indentations on both sides. Where visible or cleared, these outside walls are built with large, sometimes dressed blocks of limestone or conglomerate, representing imposing facades. Despite the use of the hill for wheat growing and cattle pasture in more recent times and as a defensive position during the war – more than a dozen bullet cases and hand grenade fragments were collected in surface layers – the archaeological remains are well preserved to the west. At least during LM IIIB, the building was about 40 m long and 25 m wide, of irregular shape, as pressed between the top of the hill and an existing road to the southeast. It is very likely that this was the main construction on the site during this phase. The destruction was unexpected: although some rooms show traces of fire, most seem to have collapsed suddenly. Earthquake destruction seems most likely but remains hypothetical for the moment. It is not impossible that the buildings on the summit remained occupied slightly later than those at Malia: at first glance, the destruction material in zones 3 & 4 closely resembles that of Quartier Nu at Malia (Driessen & Farnoux 1994) but some vases seem a bit more advanced, dating the final destruction of the site to the mature LM IIIB period10.