JOAN BRETON CONNELLY AND CAROLE McCARTNEY LAND Reprinted from the Report of the Department of Antiquities, Cyprus 2004 halcolithic ation of Geronisos Joan Breton Connelly and Carole McCartney 1.1. Introduction 1983). This campaign established a Chalcolithic date for the prehistoric material that was uneasthed Rising from the waters just off the coast of including pottery, stone tools and chipped stone. western Cyprus, between Pafos and the Akamas, The character of this Chalcolithic occupation was the island of Geronisos has long been a tempting not, however, defined in any detail at this time. but formidable destination for visitors from the mainland (Fig. 1). Geronisos was originally part In 1990, New York University was awarded a of a coastal headland and was probably attached license to explore and excavate Geronisos Island to the mainland as late as the early Holocene (c. and its surrounding waters under the direction of 12,000-8,000 B.C., Held 1992, 139). It is clear, the first author, who has written Section 1 of this however, that it was already an island when report.' To date, one archaeological survey sea- humans first visited and deposited artifacts here son (1 990), two ecological survey seasons (1 990, during the Early Chalcolithic period (c. 3,800- 1992), six excavation seasons (1 992- 1994, 1996- 1997, 2004) and four study seasons (1995, 2000, 3,500 B.C.). 2001, 2002) have been completed. Three major Steep cliffs, strong currents and the absence periods of occupation have been established for of water have discouraged a11 but the most intre- Geronisos, each separated by long periods of pid travellers. The first published accounts of vis- abandonment. These include the Early Chalcol- its to the islet are late. D. G. Hogarth described ithic, the Late Hellenistic, and the Early Byzantine his visit to Agios Georgios tis Pegeias on the (Connelly 1995, Connelly 2002, Connelly and mainland in 1888, but did not venture out to the Wilson 2002, Connelly and Mtynarczyk 2002, island itself (Hogarth 1889, 12). He relied on the Connelly 2005). report of one Captain Thompson who told him of In 1996, the second author, who has written the presence of cisterns on Geronisos that resem- Section 2 of this report, began her study of the bled those seen on the mainland. Alfred West- Geronisos chipped stone. In 1997 she expanded holm travelled out to the island in 1933 and was her work to include examination of the ground the first to report the presence of what he called stone retrieved from the island and, since 2000, Neolithic finds, including stone implements, flint and pottery (Goodwin 1984, 169 I). Westholm found this material spread across an area of some I. We thank the Department of Antiquities of Cyprus and its 100 X 70m. Rupest Gunnis may be refen-ing to the Directors under whom we have been licensed to excavate, Westholm survey material when he describes the including Dr Athanasios Papageorghiou, Dr Demos Christou, Dr Sophocles Hadjisavvas and Dr Pavlos Floin-entzos. We "Neolithic flints and pottery" found along the thank New York University and the Friends of Yeronisos who southern side of Geronisos in his 1936 guidebook have generously financed our work, especially James Ott- to Cyprus (Gunnis 1936). away, Jr., Salvatore S. Ranieri, William J. Murray, Carl S. Forsythe I11 and the de Coizart Perpetual Charitable Trust. The island remained unexcavated until 1982 We also thank the John D. and Catherine T. MacArthur Foun- when Sophocles Hadjisavvas undertook five dation for its support of Prof. Connelly's work during the years of excavation and study, 1996-2001. Drawings are by weeks of excavations on behalf of the Depart- Julia B~~rdaj~wiczand photographs are by Benjamin Fraker ment of Antiquities of Cyprus (Karageorghis and J.B. Connelly. 20 JOAN BRETON CONNELLY md CAROLE McCARTNEY she has undertaken systematic study of the Gero- 1.3 Stratigraphic contexts nisos ceramics as well. This undertaking repre- Geronisos has a fairly shallow deposit of earth sents the first time that the Chalcolithic material and fill above bedrock, ranging from Just 20cm.- from Geronisos has been comprehensively exam- lm. and averaging roughly 50cm. in thickness ined. Results of this study establish more than across the island as a whole. The Chalcolithic Just a seasonal presence on the island but rather a material is most often associated with a level of ma~orestablishment with significant ties to main- deep redlpurple earth that lies immediately above land sites and with wide-ranging implications for bedrock. Sometimes nicknamed 'Pleistocene Red' our understanding of the Early Chalcolithic peri- this earth represents the terra rosa that lies above od in western Cyprus. bedrock itself and with the secondary limestone Section 1 of this report presents the geology? formation of kaflalla. In a few cases7Chalcolith- stratigraphic contexts and distribution of the pre- ic pottery was found deposited directly upon historic material across the island. Section 2 pre- bedrock, suggesting that earth had been cleared sents the character of the artifacts recovered, their away during this period of occupation. context and the nature of the Chalcolithic occu- We also have evidence that the Chalcolithic pation that they define. levels were disturbed during construction activity dating to the Hellenistic period. It is clear that 1.2 Geology some Chalcolithic material? particularly chipped The geology of Geronisos is dominated by stone, was reused by the Hellenistic inhabitants of two primary formations: a hard, calcarenite level the island and re-deposited on Hellenistic floors. of Pleistocene marine terraces at top, ro~~ghly2.5 This will be discussed in detail in Section 1.6. to 9m. in thickness, capping a soft7 marl core The stratigraphic deposition of historical lev- below (Petrides and Charalambos 198 1). The cal- els on Geronisos is fairly straightforward. For the carenite horizon is the younger of the two geo- most part, the sequence remains the same across logical formations and is made up mostly of shell the island: topsoil, fill? late Hellenistic destruc- fragments and microfossi1s cemented by the tion level, Hellenistic occupation level, Hellenis- agent CaC03. At the surface of the island, sec- tic construction level, Chalcolithic level, terra ondary CaC03 brought by capillary action con- rosa, bedrock (Connelly 2002? 255> fig. 9). In tributes to development of a thin hard layer of certain locations, we have Early Byzantine (5th- secondary limestone known locally as kajkalla. 7111 century A.D.) construction and occupation The underlying marl, greenish grey when levels (Connelly and Wilson 2002) as well as evi- fresh and ranging from hard to very stiff, is a fis- dence for some Byzantine (13111 century A.D.) sured? silty clay sequence that occurs in bands presence. varying in thickness from 20-4Ocm. (Petrides and What is striking about the stratigraphy of Charalambos 1981). This marl is very s~~sceptible Geronisos is the very close proximity of occupa- to erosion by wave action that has created large tion levels separated by millenia of abandon- overhangs on the island where the calcarenite ment. In trench R 49 for example, stone pithos crust survives and the marl core has been lost. lids of Byzantine date were found in situ ~~1st Great boulders of calcarenite have fallen from the 20cm. above a good Chalcolithic floor level into surface of the island creating an effective wave which a pit was deliberately dug (Figs 5-8). Else- breaker (Fig. 1). Though the island has lost a sig- where on the island, good Hellenistic floors lie nificant amount of its original surface area7 the directly atop Chalcolithic levels? one of which naturally formed breakwater of boulders has con- preserves a stone post-pad in situ right beside a siderably reduced further destructive action of plaster setting bed for Hellenistic ashlar blocks wave erosion. (Fig. 11). THE CHALCOLITHIC OCCUPATION OF GERONISOS ISLAND 2 1 1.4 Distribution of Chalcolithic rnaterial The Complex seems to represent the domestic To date, some 38 grid squares have been quarter of the Hellenistic sanctuary and is com- excavated across Geronisos, most measuring plete with evidence for food preparation and dis- 4X4m. (Connelly 2002, fig. 7). In several cases tribution. Some 19,142 sherds were collected associated baulks have been removed, opening from the Central South Complex of which 7.59% 5x5 metre squares. The total excavated area are Chalcolithic in date. It should be said, how- comes to roughly 650 square m. This total ever, that Chalcolithic levels were reached only includes some nine grid squares at the west end in a limited number of probe trenches in this area, of the island, some ten excavated trenches in the much of which was excavated only down to Hel- Central South Complex (Connelly 2004, Fig. 31, lenistic floor levels. There is a strong possibility four trenches in the Central Nosth Complex and that the Chalcolithic occupation in this area is ten grid squares in and around the cisterns towards much more robust that these numbers suggest. the east end of the island (Connelly and Wilson The percentage of Chalcolithic material to 2002, fig. 25). Five trenches have been excavated the overall sherd count rises markedly when we at the east end of the island. Distribution of Chal- turn directly to the north. Here, the Central North colithic finds across these trenches shows a very Complex (Grid squares R 23111s 23s, T 21, T distinct contrast from the western end of the 24nlU 24s, V 24) preserves a circular enclosure island, which yielded very little Chalcolithic of some 13m. in diameter ringed by a simple rnaterial, to the eastern end where the distribution two-course rubble wall (Connelly 2002, fig. 7). A of prehistoric material is moderate to dense.