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. substantial foundation of a major northlsouth Excavations in the Southwest Complex (Grid wall has also been unearthed here, well dated by squares H 16, H 17, I 15, I 16, I 17 and cleaning a coin of Cleopatra VII and Ptolemy XV Caesar G 16 and G 17) exposed a sei-ies of Hellenistic (47-44 B.C.). Of the 694 sherds retrieved from walls and a courtyard between them (Connelly the North Central Complex, 20.20% are Chalcol- 2002, 256-63, fig. 8). This area yielded just over ithic in date, a marked increase over the numbers 3500 sherds of which a mere 0.2% were Chalcol- encountered on the west end of the island. ithic in date. To the northwest, on sheer cliffs and As we move further to the east, the density of largely collapsed into the sea, are the foundations Chalcolithic material ever increases. Two cisterns of West Building (Grid squares 0 8, P 8, Q 9, R are visible in the aerial photo (Fig. 2) which 81, which seems to represent the Hellenistic shows the broad semicirculas paved impluvium shrine structure that served as the centre of wor- of the Hellenistic water tank at centre and, just to ship on Geronisos during the 1st century B.C. the northwest of it, the polygonal platform and (Connelly 2002, 264-66, fig. 25). Just over 1000 square tank of the early Byzantine cistern (Con- sherds were recovered from this area of which nelly and Wilson 2002, figs 2, 25). The Hellenis- just 2.5% represents Chalcolithic material. tic cistern (Cistern 1) yielded some 282 sherds of The density of Chalcolithic pottery increases which just one was Chalcolithic in date. The dramatically as we turn to the trenches excavated trench dug beside it (P 41-P 42nlQ 41-Q 42s) in the centre of the island. The largest area produced a total of 774 sherds of which 2.45% opened up to date, that of the Central South represented Chalcolithic wares. Cistern 2, which Complex (Grid squares N 22, N 23, N 24, N 25, dates to the Early Byzantine period, yielded just 0 22,O 23,O 24,O 25, P 23 and P 24) measures 78 sherds of which only one was Chalcolithic. roughly 20 X 20 metres (Connelly and M+ynar- The trenches surrounding it (S 39, S 37-38nlT czyk 2002, 299-308 and Connelly 2005, forth- 37-38s, V 37elV 38w, V 40eN 41w, T 40, T 41, coming). It comprises several rooms, roughly S 40, S 41w baulk and R 41dS 41s. check) pro- square in shape and a fairly large open courtyasd. duced 6199 of which 35.36% are Chalcolithic. JOAN BRETON CONNELLY and CAROLE McCARTNEY

Of the 641 8 sherds collected from trenches at While the course of Hellenistic ashlars may the eastei-n end of the island (R 46, R 49, R 501R represent part of the island's original circuit wall, 511s 507 P 5011, Q 50s) the majority of some the function of the later building that rests atop it 53.50% are Chalcolithic in date (Figs 2-4).It remains uncertain. Voussoir blocks found within should be said that the condition of the Chalcol- the structure indicate that it was barrel-vaulted. A ithic material recovered from this area is superi- bronze belt buckle (MB.97.01) found in trench R or to that found elsewhere on the island. This 5OlR 511s 50 is decorated with a Christian cross may suggest that this material is found in its orig- as are two stone pithos lids found just to the west inal deposition location while artifacts recovered of the structure. This material may suggest, but to the west, particularly from the Central South certainly does not confirm, a religious associa- Complex, may be re-deposited and even reused. tion for the building. The structure may have A circular pit, carved into a trampled floor in been related in some way to the community of Grid Square R 49, was deliberately filled with a clergy serving the three Early Byzantine basilicas carefully selected assembly of tools and a fig- across the water on the mainland. It must be said7 urine, then sealed with very large stones (Figs 5- however, that East Building could have had a completely non-religious function. Indeed, the 9). This significant feature will be discussed northlsouth orientation of the structui-e would below in Section 1 S. argue against its interpretation as a chapel. 1.5 Deposit in Pit, Grid Square R 49 At the start of our work, the remains of rub- ble foundation sunning along a northlsouth axis, Far and away, the greatest density of Chalcol- roughly parallel to East Building, was partially ithic material has been found at the eastei-n end of visible. This had been exposed by Hadjisavvas' the island, that which is closest to and visible slit trench (Fig. 3, and at the left of Fig. 4) dug in from the mainland. Today, the most conspicuous 1982 along a northlsouth line just to the west of feature in this area is a rectangular structure, ori- East Building. In 1997, we opened a 4x4111. ented along a northlsouth axis at the very tip to trench that straddled this wall and designated it as the island (Figs 2-4). First excavated by Had- grid square R 49 (Fig. 4, top left and Fig. 5). This jisavvas in 1982, the structure was designated as trench has yielded the richest of the Chalcolithic East Building. In 1997, the New York University finds excavated to date on Geronisos. team re-examined the building and sunk new At an elevation of 20.87 (level 3) many frag- trenches along the exterior of its western wall. ments of glass and pottery dating to the 6111 cen- East Building exhibits three different types of tury A.D. were unearthed as well as two large masonry reflecting three different phases in its stone pithos lids7 each incised with a Christian construction: Hellenistic7 Eai-ly Byzantine, and cross (St.97.16-17). It seems that this area served ByzantinelPost Byzantine renovation. The foun- as a storage facility, perhaps associated with East dation coui-se is comprised of large Hellenistic Building. Directly below this Early Byzantine ashlar blocks. Above them, a course of much floor level, we encountered a layer containing smaller, rectangular cut stones set together with quantities of Chalcolithic chipped stone and pot- slate and tiles represents the second construction tery (level 4). The underlying level 5 produced phase of the building. Material retrieved from Chalcolithic pottery as well as very large quanti- trench R 501R 511s 50 established a 6111 century ties of chipped stone found together with nuin- A.D. date for this period of construction. The bers of ground stone tools (St.97.21, 24, 27, 28, uppermost course of masonry is comprised of 66, 68, 69) some of which can be seen in Fig. 9. small irregular stones set in cement mortar i-epre- It is of interest that no Hellenistic material was senting the third and most recent phase in the uneai-thed in this area and that the Byzantine building's history. level rested directly atop the Chalcolithic layei-. THE CHALCOLITHIC OCCUPATION OF GERONISOS ISLAND 23

At an elevation of 20.44-20.42m., a trampled reused it for their own purposes. The x and y reddish earth floor was reached and designated as coordinates of every piece of chipped stone exca- level 6.1. This floor produced much pottery, vated from Geronisos have been plotted and ele- chipped stone, ground stone tools (St.97.14, 36- vations have been taken for each artifact. This has 39, 40, 45, 88) and a chalcedony bead, St.97.43 enabled us to see the distribution of Chalcolithic (Fig. 35). Up from under this floor level and pro- chipped stone across Hellenistic floor levels truding to an elevation of 20.66m., a concentra- island-wide. tion of large stones, including two boulder-sized Certain rooms show floors on which chipped rocks, was encountered at the south of the trench stone was distributed in what appear to be work- (Fig. 5). Removal of these stones revealed the ing areas. In grid square N 22, for example, some lines of a circular pit cut neatly through the floor thirteen pieces of chipped stone were recovered level (Figs. 6-8). The pit was roughly circular in including one high quality blade (St.93.31). Six shape, measuring some 1.10m. at its largest diam- pieces were found in an interior room clustered eter and dug to a depth of ca 0.20m. It was filled in close proximity to a Dressel 1 A amphora neck with a deposit of ash and soft ashy earth, desig- that served as a pot stand (P.93.27) and a Cypri- nated as levelldeposit 6.2. It is significant that while the earth found within the pit showed quan- ote Sigillata jug (P.93.53). Three small worked tities of ash and some heat-cracked stones, no blocks placed roughly in a circle around this area, charcoal was found. A fire at a separate location and significant quantities of ash, may define a seems to have produced the ash which was delib- work space. The grouping of artifacts is sugges- erately imported to the pit along with many fist- tive of food preparation in which Chalcolithic sized stones. chipped stone recovered by Hellenistic inhabi- tants seems to have been utilized. In contrast, The pit contained (Fig. 10) pottery (one medi- neighbouring grid squares N 23 and 0 23 yield- um-sized bag of sherds was recovered), chipped ed just one piece of chipped stone each, perhaps stone and a jasper chip (St.97.59 found at eleva- tion 20.26). It also contained a stone bowl frag- suggesting that rooms in these squares served ment (St.97.58, elevation 20.22), a stone slab/ quite different functions. plate fragment (St.97.56), a pecked stone frag- Squares N 24 and 0 24 showed an extraordi- ment St.97.42 (elevation 20.36), and a stone nary density of chipped stone with over 60 pieces (St.97.84). An assemblage of stone tools was recovered from the two trenches combined. Many recovered from the deposit including grinder1 of these stones were found on the Hellenistic hammerstones (St. 97.64 and St.97.65, elevation floor level in close proximity to pots, ash and 20.37), a grinderlpolisher (St.97.62), an axe other artifacts. Stone tools were also found at (St.97.44, elevation 20.39), and what may be a Hellenistic levels within these trenches (chisel, rubber (St.97.63). At the very lowest level of the St.93.17 and St.93.24). Two of the rubble plat- pit (elevation 20.21) a stone figurine was recov- forms that rise above floor level in N 2410 24 ered (St.97.55, Figs 37-38). show Chalcolithic tools built into the construc- tions. Clearly, the Hellenistic builders were 1.6 Secondary use of Chalcolithic mater- encountering Chalcolithic material each time they ial during Hellenistic period sunk a trench and they used what they found, both Significant quantities of chipped stone have for building material as well as for utilitarian pur- been recovered from Hellenistic floor levels poses. within the Central South Complex of the island (Connelly 2005). It appears that the Hellenistic 2.1 Introduction inhabitants of the island discovered chipped The results presented below are preliminary, stone in the course of their building activities and part of the ongoing research into the prehistoric 24 JOAN BRETON CONNELLY and CAROLE McCARTNEY

occupation of Geronisos island.2 Initial results with difficulties, the term is retained in the pre- from the analysis of each of the main artifact cat- sent report for the purpose of documenting the egories, ceramics, ground stone and chipped stone relative chronology of the site (cf. Bolger 1988, are outlined, providing evidence that demon- 290; Baird 1991; Bolger and Shiels 2003). strates the Early Chalcolithic character of the Future analysis of the ceramic assemblage from island's prehistoric occupation. Much of the ana- Geronisos will refine the categories presented lytical research to date has focused on the eastern below through attribute analysis that will attempt end of the island where in situ Chalcolithic occu- to better understand the relationships between the pation deposits have been excavated (Figs 2-4). various fabric types, vessel shapes and the sur- The character of the material culture is consid- face treatments utilized. This analysis will, at the ered in detail, providing evidence for inter-site same time, permit the large body of abraded comparisons while demonstrating the signifi- sherd material to be more fully documented in cance of Geronisos island to the interpretation of terms of the general ceramic technology. the Cypriot Early Chalcolithic. Ceramic wares and Relative Chronology On-going research will seek to further refine the artifact descriptions outlined below in order The wares documented in the present report to better understand the technologies employed are based on the definitions provided by Bolger and their implications for inter-site comparisons. (1 998a, 95- 100) for the Kissonerga-Mosfilia The nature of the distribution of Chalcolithic assemblage. At present three major wares can be material across the island will be documented in documented at Geronisos, including the domi- detail and considered in terms of its relationship nant Glossy Burnished ware (GBW), a smaller to later occupation of the island. Future excava- proportion of Red-on-White Painted ware (RW), tion of prehistoric deposits will focus on provid- and Coarse ware (CW). Accompanying these pri- ing a more extensive exposure of in situ Chalcol- mary ceramic types is a small number of sherds, ithic deposits and in understanding preliminary which represent a variation of Red Monochrome indications of contextual variability. Such infor- along with one Combed ware sherd and a single mation will refine our understanding of the nature sherd with a blackened exterior not identified to of the Chalcolithic occupation on Geronisos type (Table 1). This trio of dominant ceramic Island and in relation to other Early Chalcolithic types (GBW, RW and CW) as well as the pres- sites across Cyprus. ence of a few Combed ware and Red Mono- chrome sherds has been used to define the Early 2.2 Ceramics Chalcolithic elsewhere and provides a relative The ceramic assemblage collected from both date (c. 3,800-3,500 B.C.) for the prehistoric occupation of Geronisos Island (Peltenburg 1987, excavated and surface contexts on Geronisos pro- 53-54, 1998, 12-16; see also Bolger 1998a, 102- vides a variety of evidence relevant to the date 107). and character of the Chalcolithic occupation on the island. Much of the sample analyzed to date Analysis of vessel shape is still at a prelimi- (62.94%) is unfortunately heavily abraded (Table nary stage and will be refined with future analy- 1). Preliminary analysis of the c. 113 of sherds with sis. At present it is possible to indicate a number preserved diagnostic surfaces provides informa- tion relevant to the recognition of a number of 'wares' that characterize the Geronisos Chalcol- 2. 1 would like to express my appreciation to Dr Joan Connelly ithic occupation (Tables 1-2, Figs 12- 17). While for inviting me to study the Chalcolithic material from Gero- nisos Island, and to thank Dr Diane Bolger for comments on it is fully recognized that the definition of wares an earlier version of this paper. All opinions expressed in the for prehistoric hand-made ceramics is fraught text of Section 2, however, are my responsibility alone. THE CHALCOLITHIC OCCUPATlON OF GERONISOS ISLAND 25 of formal details relevant to the discussion of core, with a uniform fine white grit, hard with a ceramic types and relative chronology. Both the bricky fracture

GBW and RW clearly demonstrate the use of shapes - platters, bowls, spouted platters1 both open and closed vessels at the site, while bowls, flasks only open vessels are represented by CW (Table Glossy Burnished wase, first defined by Pel- 1). Plain rims demonstrate the presence of both tenburg (1979, 33-34) for the assemblage collect- flaring straight-sided platters as well as smaller ed from the site of Kissonerga-Mylo~~tlzkia,has bowls for both GBW and RW, with GBW mani- become the one of the most significant type fos- festing the occasional spouted version. These sils defining the Early Chalcolithic in western vessels typically displayed flat bases with one Cyprus. Described by Bolger (1998a, 108) as a GBW example exhibiting a slight flange. Closed 'true ware' in terms of its consistency of surface forms dominate the sherd counts for both GBW treatment, fabric and vessel morphology, this and RW, more heavily so in the case of RW. ceramic type is highly consistent between the Prevalent among the closed sherds are a signifi- excavated assemblages of Kissonerga-Myloutlzkia, cant number of neck fragments that demonstrate Kissonesga-Musfilia, and now Geronisos (Bolger the common occurrence of flasks (again most 1998a, 95-95; Bolger and Shiels 2003). The frequently with RW). The bowl forms clearly fabrics and surface treatment of the Geronisos derive from the Ceramic Neolithic with the flask GBW closely correspond with the descriptions representing the only new vessel form introduced provided for the Lempa Cluster sites, while pre- in the Early Chalcolithic (Bolger 1991, 85, liminary indications of vessel form demonstrate 1998a, 107). A similar prevalence of flasks was the presence of the most characteristic Early noted for the Early Chalcolithic at Kalavasos- Chalcolithic shapes and elements (Fig. 12). Pampo~desand Kalavasos-Agio~~s,though the According to the survey reports, GBW is pre- need at Geronisos for the storage of liquid was sent in the predominantly Middle Chalcolithic more clearly vital since there are no water sources ceramic assemblage of Miliou-Rodoudes provid- on the island (Clarke and Todd 1993; Baisd in ing an inland extension to presence of this domi- press; Connelly and Wilson 2002). CW is exclu- nant ceramic type, while recent survey collection sively dominated by vertical sided flat-based shows GBW from Agios Kononos in the Akamas flanged trays with one example exhibiting a asea (Stewart 1981, 42-46; Bolger pers. comm.). small lug. Other diagnostic forms including Glossy Burnished ware shows only a residual pointed and omphalos bases are represented, but presence in the early Middle Chalcolithic (3A), is unfortunately occur only among the abraded gone by sub-period (3B) at Kissonerga-Mosfilia, fragments (Fig. 17). and is absent from Period 1 at Lempa-Laklzous Glossy Burnished ware (GBW) (Peltenburg 1993, 12; Bolger 1998a, 102). A sim- ilar fabric lacking the diagnostic glossy surface 1 su$ace treatment - red paint applied directly treatment (perhaps due to surface abrasion) was onto an unslipped surface, heavily burnished to a represented at Early Chalcolithic Maa-Palaiokas- high lustre, with burnishing strokes frequently tro, which is dominated instead by Red mono- visible chrome painted ceramics (Peltenburg 1985, 14, fabrics - A: a well-levigated orangey-pink table 1; Stewart 1985,240-50; Bolger 1988,291- colour typically with a grey core, fine rounded 2). Ceramic parallels at the start of the Chalco- white, light grey and sometimes red grit, hard lithic elsewhere, notably at Kalavasos-Agio~~s, with a bricky fracture show highly burnished surfaces used somewhat - B: a well-levigated very pale brown to differently on both Red monochrome and Red- pink colour, more occasional grey sandwiched on-White painted ceramics particularly in phase 26 JOAN BRETON CONNELLY and CAROLE McCARTNEY

lb, while corresponding vessel shapes are promi- the origin of the Early Chalcolithic tradition has nent especially in phase 2 (Kromholz 198 1, 21- been demonstrated to emerge from the preceding 22; Peltenburg 1987,54; Baird in press). The par- Late Neolithic (Peltenburg 1980, 3-4, 1983, 10; ticular character of GBW as reflected in the sig- Bolger, 1985, 33; Peltenburg and Spanou 1999, nificance of the prominence of monochrome pot- 19). Parallels exhibited between the final Agios tery in the west, however, has been described as Epktitios-Vtysi RW style and that of Kalavasos- a regional feature of the Early Chalcolithic tradi- Agio~~s(phase la) allow the Late Neolithic to tion belonging to the Pafos region (Peltenburg Early Chalcolithic transition to be placed near the 1980, 3; Baird in press). Geronisos island, there- start of the 4th millennium B.C. (Bolger 1987, 84, fore, clearly fits within that regional milieu and 1991, 84; Peltenburg and Spanou 1999, 17-19; belongs within the Early Chalcolithic landscape Baird in press). In the west RW is comparatively documented by the Lempa Cluster sites. less frequent, but demonstrates the same range of linear motifs as well as the use of a multiple Red-on-White ware (RW) brush technique that define the Early Chalcolith- surface treatment - cream to pinky coloured ic RW tradition elsewhere (Bolger 1991, 84; Pel- slip, at times relatively thick and chalky, some- tenburg and Spanou 1999, 17; Bolger and Shiels times self-slipped, painted with a thin red-brown 2003; Baird in press). The Geronisos RW mate- paint in broad and narrow lines, block coloured rial clearly belongs to this Early Chalcolithic areas and the occasional lattice linear tradition.

fabrics - C: a well-levigated orangy-pink Many of the sherds counted as RW in the colour typically with a grey core, fine rounded Geronisos assemblage were defined on the basis white or light grey grit (closely resembles fabic of the presence of a creamy coloured slip because A in terms of colour and temper) comparatively few RW sherds from the island had painted motifs preserved. This sherdage - D: a well-levigated light yellow to cream colour, sometimes with grey core, homogeneous could be defined as 'plain white' (cf. Bolger and 20031, but because these sherds were fine grey grit, relatively soft with a smooth Shiels indistinguishable from those exhibiting RW rounded fracture (resembles standard Neolithic painted motifs in terms of both fabric and slip, fabric A at Kissonerga, Bolger 1998a, 97) they were grouped together in the present report. shapes - bowls, platters(?), flasks The use of a slip on Geronisos RW sherds is a Though GBW may be the most discrete ware feature prevalent during the Early Chalcolithic in and clearly dominates at Geronisos, it is with the the Pafos region at sites like Kissonerga-Mosfilia painted ceramics that 1-egional aspects of the (Bolger 1998a, 96). In contrast, the use of a slip Easly Chalcolithic across Cyprus as well as the was comparatively rare at Kalavasos-Agious transition from the Late Neolithic have been most prior to phase 2 where it was used with burnish- well documented. In contrast to the northern ing for decorative effect (Ksomholz 1981,29-30; Ceramic Neolithic tradition, Paralimni-Nissia, Baird 1991, 23-24, in press). This feature con- and perhaps the Late Neolithic of the Pafos dis- firms the island's close relationship with other trict, a low percentage of RW sherds (relative to sites of the Pafos district, exemplified by those of monochrome) is a characteristic of Early Chal- the Lempa Cluster. colithic ceramic assemblages (Kromholz 1981, For the majority of RW sherds only small 29; Baird 1985, 15-18; Bolger 1991; Flourentzos traces of the red painted pigment have survived. 1997, 4; Peltenburg 1998, 8). The limited number of Geronisos RW sherds It is with the shift from curvilinear to linear exhibiting preserved painted motifs shows the painted decoration at Agios Epiktitos-V~sithat use of parallel lines and broad bands as well as THE CHALCOLITHIC OCCUPATION OF GERONISOS ISLAND

block painted areas and the presence of the lattice Coarse ware (CW) motif (Figs 14- 16). Such a combination of motifs surface treatment - typically untreated, but suggests that the RW of Geronisos would be most sometimes exhibits a thin red wash or paint satisfactorily placed under the description of inside and out or a thick creamy rough plaster Red-on-White band and line ware or RWBL like slip (Bolger 1998a, 96). This variety of RW is char- acteristic of period 2 at Kissonerga-Musfilia and fabric - E: coarse dai-k brown to black Kissonerga-Mylu~~thkia,tho~~gh the Geronisos crumbly fabric with large coarse chopped veg- range of painted motifs is far more limited etable filler and a coarse rounded grey to black (Bolger 1998a, 109; Bolgei- and Shiels 2003). and more occasionally red grit. Interestingly, the Geronisos RW appears most slzapes - tray compatable with the description provided by The significance of the Coarse ware trays in Bolger for the Maa-Palaiukastro RW (Bolger the Geronisos ceramic assemblage is two-fold 1988, 290-291). RW at both of these small Early (Fig. 13). Not only is this vessel type highly diag- Chalcolithic 'homesteads' exhibits a relatively nostic of the Early Chalcolithic, but the very frag- soft creamy slip and a limited range of motifs ile nature of CW trays suggests that they must suggesting that the degree of vai-iability for this have been produced in sit^^ (Bolger 1998a, 98). ware within any particular site may, in part, be a The relatively frequent inclusion of this ceramic function of site size and corresponding socio- type in the Geronisos assemblage, therefore, economic complexity. Bolger (1991, 85) has implies that occupation on the island was more inferred a chronological distinction linking the RW of Maa-Palaiulcastro and Kalavasos-Agiu~~s than a mere camping expedition and that at least phase 1 (a) with the Late phase at Agios Epikti- some ceramics were being produced on the tos-Vi-ysi in contrast to greater similarities in the island. The variety of surface treatments shown RW of Kissonerga-Musfilia and Kissonerga- by the Geronisos CW sherds is typical of this Mylo~ithlciato that of Kalavasos-Agiu~~sphases ware seen at Maa-Palaiukasti-u again providing a 1b and 2 (see also Baird in press). The RW from link been these two small Early Chalcolithic sites Geronosis may be similarly early, but promi- (Bolger 1988, 291). The same surface treatments nence of GWB at Geronisos, in contrast to Maa- were also noted at Kalavasos-Agious, while the Palaiokastru, implies that the former is more presence of red painted surfaces on Coarse ware closely related to the Kissonerga Early Chalcol- fabric has also recently been noted at Kissonerga- ithic assemblages. The apparent similarity of Mylu~~thkia(Kromholz 1981, 22; Bolger and both the Geronisos and Maa-Palaikastro RW to Shiels 2003). The u-shaped openings charac- the earlier Late Neolithic variety (as distinct from terizing CW trays at Kissonerga-Musfilia, Kisson- the orange-pink (slip) of the Mylu~~thkiavarient erga-Mylu~~thkiaand Kalavasos-Agiuus, howev- also shown at Kissonerga-Musfilia) may, there- er, have not yet been identified at Geronisos fore, suggest a slower rate of change occurring at (Baird 1991, 22; Bolger 1998a, 98; Bolger and smaller sites (Stewart 1983, 245; Bolger 1998a, Shiels 2003). The absence of this particular CW 96). Since the Kissonerga-Mylu~~tlzkiaRW has form also from Maa-Palaiokastro assemblage more recently been described as exhibiting a buff (if this site represents a small homestead as inter- or whitish slip (Bolger and Shiels 20031, it preted by the excavator) suggests that u-shaped appears that this distinction, whether between the openings in CW trays may be functionally relat- smaller and more substantial sites or across time, ed to site size, being absent from smaller scale may be limited to the vasiety of motifs rather than occupations (Bolger 1988, 293; Thomas 1988, motif as well as slip character. 283). 28 JOAN BRETON CONNELLY and CAROLE McCARTNEY

Red monochrome ware (Rm) Intra-site variability surface tl-eatment - unslipped but painted A detailed analysis of context has not yet with red pigment been carried out, but preliminary ceramic analy- fabrics - F: somewhat crumbly dark red- sis results demonstrate interesting possibilities of brown fabric with medium to large angular red intra-site variation that require further investiga- and black grit, hard with a rough bricky fracture tion (Table 2). Square R 49 (Figs 2-31> which (resembles Kissonerga fabric E, Bolger 1998a, intersects Hadjisavvas's north/south slit trench of 97). 1982, demonstrates a ceramic profile heavily dominated by GBW. A clear contrast is demon- shapes - unknown, few sherds strated by square R 46 (Figs 2-8) where RW was A small number of sherds in the Geronisos somewhat more prevalent than GBW. The latter assemblage exhibiting red painted but unslipped ceramic type is again most prominent in square Q surfaces and a dark red-brown fabric resemble 50NlR 50s and sub-square Q 50S, but at a level material defined as Red monochrome (Rm) and significantly below that of R 49. RW in Q 5ONlR fabric E at Kissonerga-Mosfilia and RM Kisson- 50s and Q 50S, as in square R 46, is again well erga-Mylouthkia (Bolger 1998a, 96-97; Bolger represented. These data suggest that the use of and Shiels 2003). Bolger (1998b, 132) has RW was prominent in the area underlying the documented the presence of a limited amount of East building, while the monochrome GBW was RM in the Kissonerga-Mosfilia period 2 assem- more heavily used further to the west. The degree blage, but noted that this ceramic type, though to which this circumstance represents a true clearly introduced at this time is as yet poorly 'functional' contrast in these different contexts of understood. RM ceramics are characteristic of the Geronisos Chalcolithic occupation, or an the Middle Chalcolithic phases and at Kissoner- accident of surface treatment preservation will ga-Mosfilia demonstrating variability in terms of need to be explored with future excavation as fabric that has been used to distinguish the peri- well as stratigraphic and fabric analyses of the od 3 sub-phases 3A and 3B (Bolger 1998a, 103). collected sherdage. The small sample of Red monochrome sherds at Square P 50N stands apart with its unusually Geronisos provides an additional example illus- high percentage of CW. While not quantified, the trating the presence of Rm pottery within the CW of the adjacent squares Q 50s and Q 50NlR Early Chalcolithic. 50s seemed comparatively poorly preserved. The use of CW trays (probably for food preparation), Combed ware (Cb) therefore, appears to be a particular feature asso- su$ace treatment - pink coloured slip with ciated with the Chalcolithic occupation abutting thin red paint combed with parallel wavy lines the southeastern edge of the island. At Lempa- fabric - D (the same as RW and the standard Lakkous (Period 2) CW trays were similarly iso- Neolithic fabric A at Kissonerga, (Bolger 1998a, lated in external areas (Peltenburg 1985, 321). Additional excavation in this area is required to 97)). better understand this observation. shapes - unknown, single sherd 2.3 Ground stone ? - unidentified ware (?) The ground stone assemblage from Gero- suqace treatment - slipped? with a black bur- nisos Island is still not fully documented in nished surface on the sherd exterior numerical detail and so is not tabulated separate- fabric - C (a misfired RW?) ly in the present report. A number of ground shapes - unknown, single sherd stone materials from both the centre of in situ THE CHALCOLITHIC OCCUPATION OF GERONISOS ISLAND 29

Chalcolithic occupation at square R 49 as well as a number of cutting tools including axes, adzes, artifacts collected from the south-central part of chisels and flaked tools. The axes, in particular, the island have been analyzed and are summa- exhibit features such as their small size (less than rized below. The character of the overall assem- 12cm. long), rounded butts and limited type vari- blage, significant tool types and preliminary ety that suggest an Early Chalcolithic date cosre- inter-site comparisons are provided, that further sponding to the ceramic assemblage (Elliott 1983, define the nature of the prehistoric occupation of 14, 1985, 274). The axes also show polishing the island. limited to the blade end with the rest of the tool The ground stone industry of Geronisos exhibiting only pecking, an early feat~lredemon- Island is typical of the Chalcolithic, while a num- strating continuity with the preceding Neolithic ber of features clearly demonstrate the early period. In light of the discussion of RW ceramics above, it is interesting to note that axes from character of the material (Fig. 13). Both volcanic Maa-Palaiokasfro, like those of Lempa-Lakko~~s and calcareous rocks were utilized on the island period 1, show a greater degree of facetting and of Geronisos as elsewhere in Chalcolithic Cyprus. overall extent of surface grinding. These features It is worth noting, however, that the heavier vol- were interpreted as indicating a slightly later date canic materials, in particular, are not native to the than those of Kissonerga-Mylouthkia (Thomas island's geology but must have been carried 1988,282). In contrast to the axes, chisels includ- (laboriously in the case of the larger querns) onto ing cigar-shaped examples were polished more the island. Calcarenites are immediately avail- extensively and finished to a much higher degree. able in local Pleistocene terrace deposits of the Ovoid flaked tools and small, broad well-made island, but all other stone materials must have adzes, provide additional features derived from been collected from the mainland environment. Neolithic tradition that are typical of the Early Volcanic stones are typically available as rolled Chalcolithic, elsewhere as at Geronisos (Elliott pebbles from beach localities, while limestone 1983, 15-16, 1985, 274). These tool types are and serpentenite are available from the Mamonia typical at Kissonerga-Mylo~~thkia,the impover- complex that covers most of the Pafos region. ished period 2 assemblage from Kissonerga- The presence of the distinctive Akamas red sand- Mosfilia, Kalavasos-Agio~~s,Maa-Palaiokastro, stone in the assemblage demonstrates knowledge and Lempa-Lakkous period 1 (Elliott 1983, 14-16, of the raw materials available in that area (G. 1985, 274, 1998a, 180; South 1985,76; Thomas Petrides pers. comm.). Broadly speaking, the 1988, 279; see also Jackson 2003). While not yet Geronisos environment in terms of raw material quantified at Geronisos, such tools represent availability is the same as that for the Lempa frequent occurrences suggesting that this assem- Cluster sites (Elliott 1983,35, fig. 8). The fact that blage also reflects the typical prevalence (c. 30- such raw materials were brought to the island 36%) of cutting tools in other Early Chalcolithic helps to demonstrate the periodic (probably sea- assemblages (Elliott 1983, 12-18; South 1985, sonal) nature of the Early Chalcolithic occupa- 76-78; Thomas 1988, 279-281). tion on Geronisos Island and suggests familiarity with large areas of the mainland environment. Abrasion tools Numerically more prominent are tools asso- Cutting tools ciated with various types of abrasion (Figs 21-24 In terms of tool type, the ground stone indus- and 30-33). Abrasion tools similarly dominate (c. try from Geronisos is, perhaps not surprisingly, 50%) other Early Chalcolithic assemblages dominated by smaller hand-held objects (Figs around Cyprus and imply a significant amount of 18-20 and 26-29). Of the artifacts analyzed to food preparation activity on Geronisos Island date, tools made primarily from diabase comprise (Elliott 1983, 1998a, 180, 1998b, 193; South 30 JOAN BRETON CONNELLY and CAROLE McCARTNEY

1985, 76-78; Thomas 1988, 279-281). Tool types clay or plastering (Peltenburg 1980, 5-6; Bolger include numerous grinders and rubber fragments, 1985, 29, 1988, 292). Similarly, the burnishers a few querns, one cup-marked stone, hammer- could have been employed in the production of stones, pounders, polishing and burnishing stones, GBW pottery (cf. Bolger 1998a, 108). Other pecking stones and pestles. The relative paucity tools described for the Kissonerga-Mylouthkia of pestles at Geronisos and their simple conical assemblage, in particular, include an association form (inherited from the Neolithic) is typical of between the crushing of red ochre and cup- other Early Chalcolithic assemblages (Elliott marked stones, and the presence of heat-fractured 1983, 21, 1998b, 193). The composition and rel- tools found in non-hearth contexts (Peltenburg ative frequency of ground stone items on Gei-o- 1980, 5-6; Elliott 1983, 24). The presence of a nisos, like the ceramic assemblage, characterize number of small lumps of red ochre from Gero- the Chalcolithic occupation of the island as more nisos confirms the processing of this material on than just a fleeting visit. The variety of wild the island. plants available on the island of Geronisos today comprises 95% of the mainland species, while Vessels similar wild plant resources were recorded at Two small irregular bowl fragments made of Kissonerga-Mosfilia (Murray 1998,222; Connel- calcareous materials were collected from the in ly 2002). The range of food plants available on situ deposits of square R 49. These pieces repre- the island are likely to have been exploited in sent the only examples of stone vessels currently antiquity adding to any food stuffs that may have known in the prehistoric ground stone assem- been carried from the mainland. blage on Geronisos. Missing from the assem- Beyond food preparation, several lines of evi- blage is the array of stone bowls, basins and other dence suggest manufacturing activities were car- containers accounting for c. 11-16% of other ried out using ground stone tools. Like the Early Early Chalcolithic sites such as Kissonerga- and Middle Chalcolithic ground stone assem- Mylouthkia, Kalavasos-Agious and Maa-Palaio- blages belonging to the Lempa Cluster, tools kastro (Elliott 1983, 24-27; South 1985, 55; from Geronisos were frequently used or re-used Thomas 1988, 279-28 1). At Kissonerga-Mosfilia for more than one type of activity, for example period 2 where the cutting tools are so rare, stone hammerstonelgrinders, hammerstone/burnishers, vessels comprise over 113 of the ground stone pounderlgrinder, grinderlpolishers or rubber1 assemblage (Elliott 1998a, 180, 1998b, 193). The grinders. Such non-purpose designed or expedi- limited domestic vessel repertoire from Gero- ent tools are especially typical of the Kissonerga- nisos, particularly of larger items, is perhaps Mylouthkia and Lempa-Lakkous assemblages indicative of the logistic difficulties faced in (Peltenburg 1980, 5-6, 1981, 30, 1985, 320-321; transporting finished stone objects up to the top Elliott 1985, 275). While this characterization is of the island, but fails to explain why the native not unique to the 1st half of the 4th millennium calcarenite was not more heavily exploited. While B.C. it does appear to epitomize the Early Chal- this difference may indicate a functional differ- colithic prior to the development of more spe- entiation in the activities performed at the site, a cialized tool types in the Middle Chalcolithic larger exposure of the prehistoric deposits on the (Peltenburg 1998, 243). island is required to confirm or deny this appar- One interesting feature of the assemblage is a ent anomaly. At present, judging from the types small number of grindinglpolishing implements of ceramic vessel recovered and the general exhibiting a thin mudlclay coating over much of paucity of ground stone vessels, it appears that their surfaces. This encrustation often exhibits storage of things other than liquids was of a rela- multiple striations and may be the result of man- tively low priority. This low storage capacity fea- ufacturing activities, perhaps the processing of ture of the Geronisos ceramic and ground stone THE CHALCOLITHIC OCCUPATION OF GERONISOS ISLAND 3 1 assemblages, in light of the current absence of The final non-utilitarian item is a stone fig- deep bell-shaped storage pits characteristic of urine made on a sedimentary rock (Figs 37-38). Early Chalcolithic sites elsewhere, corresponds This stylized anthropomorphic object is highly to the suggested temporary (seasonal) nature of characteristic of the Chalcolithic period during visits to the island. which stone 'idols' of chalk, calcarenite, ceramic and (most recognizably) picrolite proliferate. The Non-utilitarian and other items Geronisos example is crudely modeled providing Included in the Geronisos ground stone a highly stylized outline of the human form with- assemblage evaluated to date, are a few objects of out reference to gender. The piece was roughly personal adornment, possible structural detail, worked and damaged, perhaps deliberately, prior and a single figurine. Two examples of stone to deposition (see below). A similar single stone beads, which have Late Neolithic precedents, are figurine was recorded for period 2 at Kissonerga- represented in the form of unfinished bead blanks Mosfilia as well as at Maa-Palaiokastro, which (Dikaios 1961, 201). These artifacts belong to a provide close parallels to the Geronisos example relatively large barrel-shaped bead type both of (Thomas 1988, 282; Goring 1998, 149, table 2). which were made of bright greenlgrey serpenti- A single stone figurine from Kalavasos-Agious nite (Fig. 25). Similar shaped stone beads were is accompanied by more prolific ceramic exam- recorded at Kissonerga-MosfiUa and add to the ples, while at Kissonerga-Mylo~~tlzkiafigurines of variety of ornamental objects belonging to Early stone, including picrolite, and ceramic are abun- Chalcolithic sites (Peltenburg 1998, pi. 37.1). At dant (South 1985, 67-68; Peltenburg et al. 2003). Geronisos, these objects appear to be attempts to The presence of a singular stone figurine of the utilize a material exhibiting the colour of picro- hand-held type defined by Goring (1998, 150) life, but lacking its softer more easily carved appears to be characteristic of many Early Chal- characteristics. Picrolite is represented by a sin- colithic sites. This feature perhaps built upon a gle small (1.3X0.5cm.) carefully shaped and Late Neolithic tradition as suggested by Sotira- highly polished biconical pendant (Fig. 34). The Teppes was substantially altered at Kissonerga- extreme rarity of this stone at Geronisos (and its Mylouthkia, where the prominent use of picrolite absence at Maa-Palaiokastro) suggests that this (more typically seen in Middle Chalcolithic fig- valued material was more easily obtained by urines) appears to have initiated (Dikaios 196 1, inhabitants of the larger Early Chalcolithic sites 20 1-202; Peltenburg 199 1a, 1 14, 2003). like Kissonerga-Mylouthkia or Kalavasos-Agious prior to the development of more intensive trade Two other objects that deserve mention are networks in the Middle Chalcolithic (South 1985, fragments of flat plates, pot-stands or post-pads 73; Thomas 1988; Peltenburg 199 la, 2003). A one of which is found in situ in square 0 22 (Fig. second elongated biconical pendant preform 11). These objects have potential parallels at both (Fig. 35) has clear parallels at Kalavasos-Agious Kissonerga-Mosfilia period 2 and Kalavasos- and within the Lempa cluster (Peltenburg 1983, Agious (Todd 1982, 67; Elliott 1998b, 193). The 10; South 1985, 73). The example from Kalava- possibility that these artifacts may be structurally sos-Agious is made from carnelian, possibly related is of significance to the interpretation of imported from Egypt, while the Geronisos exam- context outlined below. ple (Fig. 35) appears to be made of local orange translucent chalcedony (Todd 1982, 55; South 2.4 Chipped stone 1985, 73). One additional ornament, a perforated The Geronisos chipped stone assemblage is flat pebble 'pendant' made of a siliceous materi- relatively small (n=737), but proportionate to the al is somewhat singular due to its unpolished limited extent of the Chalcolithic exposures form (Fig. 36). (Table 3). In general, the assemblage is heavily 32 JOAN BRETON CONNELLY and CAROLE McCARTNEY fragmented and exhibits the characteristics of same access to the distant 'Moni' stone sources severe raw material scarcity (see Fig. 9 for a sam- as the knappers from these large Lempa Cluster ple of chipped stone from trench R 49). This sites. The exploitation of this particular chert dominating feature of the Geronisos chipped variety on Geronisos like presence of picrolite, a stone assemblage dictated the core technology, material derived from the Kouris river, show the the heavy degree of core and tool exhaustion, and inhabitants of the island to be actively exploiting the most prevalent tool types. Thus, while the resources from a wide area across south-west assemblage can be usefully compared with other Cyprus (see McCartney in press and references contemporary assemblages on Cyprus, its island there-in). context makes it somewhat unique, and demon- The final point that must be mentioned strates a distinct ch&e ophtoire. regarding raw material, is the presence of a single piece of obsidian, excavated from square T 40 in Raw material utilization the area of the cisterns. This heavily abraided Though restricted in terms of raw material artifact (a medial prismatic bladelet segment) is abundance, the knappers of Geronisos were not technologically inconsistent with the dominant impoverished in terms of material quality (Table Chalcolithic character of the assemblage, SUE- 4). Over 213rds of the assemblage (c. 71%) was gesting, instead, an Aceramic Neolithic date. The made on good to high quality chert types domi- piece exhibits unidirectional dorsal scars and was nated by 'Lefkara-translucent' chert along with probably produced with a pressure technique in the distinctive black to dark grey 'Moni' chert keeping with other Aceramic Neolithic examples and the very high quality 'Translucent' variety. such as those from Kissonerga-Myloutlzkia The use of poorer quality 'Lefkara-dense translu- (McCartney 2003, see also below). It should be cent' and particularly chalcedony or the rare noted, however, that a single obsidian artifact jasper was facilitated by the specific core tech- was also recovered from floor 1 of pit 15/35 nology employed (see below). Though the Maa-Palaiokastro, a site lacking any evidence of assemblage is composed of a variety of Lefkara Neolithic activity (Thomas 1988,282). One piece chert types available of the Pafos region, it is sig- of obsidian, interpreted as being residual from nificant that the more moderate chert type most the Aceramic Neolithic, was also collected from widely available in the local mainland environ- a period 2 context at Kissonerga-Myloutlzkia ment ("Lefkara-basal') was not heavily exploited (McCartney 2003; see also McCastney 1998,260 on the island. This chert type has been observed for a discussion of obsidian from Chalcolithic in the area above the Avgas and Aspros rivers contexts at Kissonerga-Mosfilia). immediately north of the site. Instead, the preva- lence of the 'Moni' variety, typical of the Xeros Technology and Diarizos rivers in eastern Pafos, suggests that Once on the island, the inhabitants of Gero- materials were collected from further afield. nisos adapted their knapping technology to fit the Notably, the Geronisos raw material profile par- restricted raw material availability. The extreme- allels those of the Kissonerga-Mosflia and ly intensive core technology is characterized by a Kissonerga-Mylouthkia assemblages in terms of dominant use of the bipolar-on-anvil technique the use of high quality 'Translucent' and 'Moni' and heavily exhausted flake cores. Splintered chests (though the later assemblages both show a pieces exhibiting small flake and bladelet or spa11 greater use of 'Lefkasa-basal' and less 'Lekara- scars demonstrate an effort to maximize raw translucent' chert than seen at Geronisos) material exploitation. Notably, splintered pieces (McCartney 1998, 2003). The knappers from and bipolar products occur frequently on the high Geronisos, therefore, appear to have had the quality raw materials noted above. Other cores THE CHALCOLITHIC OCCUPATION OF GERONISOS ISLAND 77 show the use of the discoidal change-of-orienta- denticulates, glossed pieces, notches, perforators, tion, single platform and irregular core types for pieces esquilEes, retouched pieces, scrapers, and the production of flakes. Broad plain and facetted truncations (Table 5). Of these, burins and mar- butts and a small number of simple platform reju- ginally retouched flakes are the most prevalent venation flakes further define this simple core corresponding well with the prominence of burins, technology. The presence of cores, core-trim- in particular, in the Kissonerga-Mylo~dzkia, ming elements, unworked flake blanks and debris Kissonerga-Mosflia and Kalavasos-Agious Early clearly demonstrate the production of tools on the Chalcolithic assemblages (McCartney 1998, island. Frequently, larger flakes (cores-on-flakes) 2003; Betts 2004). Characteristic Chalcolithic were used as cores for smaller flakes. One feature flake scrapers are represented by heavily reduced of this intensive industry shows the use of side- fragments including examples made on 'Moni' blow truncations to generate segments from larg- chert, the preferred material type for scraper pro- er pieces. Such technological features are common duction, particularly prominent in the subsequent in other Chalcolithic assemblages, including those Middle Chalcolithic (McCartney 1998, 2003, in of the Lempa Cluster, but were not used as inten- press b). sively elsewhere (McCartney 1998, 2003). The final tool that should be noted is a single symmetrical point tang recovered from a surface Tools context near square R 46 on the eastern end of the The re-use of tools for secondary tool func- island, which resembles the Byblos type except tions or as core material also characterizes the for the presence of a small notch on the left later- material paucity of the island's chipped stone al edge. The artifact exhibits a light gloss on both assemblage. Heavily currated tools saw their use lateral edges indicating that it was re-utilized in lives atypically extended and were reduced to antiquity (possibly accounting for the notch) pre- often quite small residual fragments. The promi- sumably by the Chalcolithic inhabitants of the nence of simple utilized flakes is typically viewed island from whom quality chert was valued at as indicative of 'ad-hoc' tool manufacture in a such a premium. This piece, made on a bidirec- resource-plentiful environment. On Geronisos tional prismatic blade segment of high quality such informal tools demonstrate the final stage of 'Moni' chert is consistent with point tangs docu- very intensive raw material exploitation on the mented for the Early Aceramic Neolithic assem- island. In this particular context, even the utilized blage from Kissonerga-Myloutlzlcia, though the tools were probably used multiple times. Thus, material used is more typical of the subsequent while tool re-utilization and utililized flakes are Middle Aceramic phase (McCartney 2001; Pel- indicative of Chalcolithic assemblages elsewhere tenburg 2001; McCartney 2003). The point in the Pafos region, they are particularly marked tang, together with the obsidian bladelet, imply at Geronisos (McCartney 2003). In addition, the that Geronisos Island was visited during the Ace- prominence of utilized pieces in this coastal ramie Neolithic prior to the more extended occu- assemblage suggests that such tools are not indica- pation of the island during the Early Chalcolithic. tive of a coastal versus inland dichotomy, but a feature of the particular context of this assemblage Chipped stone intra-site varaibility (contra De'Annibale 1999). Assemblage context, In spite of the heavy recycled nature of the site size and place within a regional site hierarchy chipped stone artifacts belonging to the Gero- are factors of equal if not greater importance to nisos assemblage, intra-site variability is appar- geographical location (Bolger, McCartney and ent in excavation squares with high chipped stone Peltenburg in press). densities (Table 6). At the eastern end of the Other tool classes represented in the Gero- island, the concentration of 'in situ' Chalcolithic nisos assemblage include burins, backed pieces, occupation in square R 49 shows an unusually 34 JOAN BRETON CONNELLY and CAROLE McCARTNEY high percentage of tools and a relatively low per- of Chalcolithic artifacts. These density peaks rep- centage of cores and core-trimming elements. resent clear areas of Chalcolithic occupation, These figures along with a lower than average though all excavation squares except for R 49 amount of blanks and debris suggests less core demonstrate varying levels of disturbance by later reduction than tool utilization activity in the area occupants of the island. Other excavation squares of Square R 49. In contrast, in the area of squares where artifact densities were lower appear to be S 40 and T 41 between cisterns 1 and 2 the per- more indicative of the disturbance activity by later centage of tools is considerably lower and the inhabitants of the island or erosion adding more to percentages of debris and blanks high. A some- our understanding of artifact variability than the what higher than average percentage of cores and nature of the Chalcolithic occupation. core-trimming elements in square T 41, in partic- The Chalcolithic material from the Central ular, suggest a greater focus on core reduction in South Complex was unfortunately heavily dis- this area. Square T 40 shows a higher percentage turbed by subsequent Hellenistic building activi- of tools and lower percentage of debris and blanks ty. This area provides evidence of artifact vari- than either S 40 or T 41, but is still at a level ability, but little concerning the nature of the below that of square R 49 to the east. In the south- occupation. In contrast, Chalcolithic materials central area of the island a pattern of greater core recovered between cisterns 1 and 2 demonstrate reduction resembling that of the cistern area is not only high densities of artifacts, but evidence apparent in square 0 24, while that of square N 25 of 'mud-brick' and ashy fill in square T 41 sug- shows a pattern similar to square T 40. gestive of possible temporary structural debris. Similar evidence of pis6 or daub has been recov- 2.5 Site Character ered from other sites like Kalavasos-Agious and In spite of the fleeting evidence of an Ace- used to define the lightly-built structures that ramie Neolithic visit to Geronisos, the prehistoric characterize the Early Chalcolithic (Todd 1985, activity on the island can be defined as a single 86). At Kissonerga-Mosfilia (period 2) and Kis- period occupation belonging to the Early Chal- sonerga-Mylo~~tlzlziaconsistent post hole evi- colithic. This characterization can be confidently dence associated with structural mud indicates made on the basis of the ceramic, ground stone the presence of timber framed structures set with- and chipped stone artifacts, which all indicate a in shallow hollow depressions (Peltenburg 1998, relative date placing the site in the first half of the 24, 2003). The flat stone plateslpost-pads from 4th millennium B.C. In spite of the consistency of 0 22 could be related to the use of light timber the various Chalcolithic artifact assemblages, one structures, but the lack of post-holes in the area priority of future excavation will be to obtain makes such evidence tenuous. The potential for material for carbon-14 in order to establish an similar timber framed structures on Geronisos, absolute chronology for the site. therefore, requires further investigation. At present square R 49 provides the clearest Chalcolithic contexts and features parallels with other Early Chalcolithic sites pro- The Chalcolithic contexts have been outlined viding the only in situ features excavated thus far in detail in Section 1 of this report. Major areas of on the island. Though the single stone figurine Chalcolithic occupation, however, are briefly con- clearly represents an ideological artifact it is the sidered here in terms of artifact density. Three context of this figurine that is most suggestive of areas, namely, the eastern end of the island around 'ritual' activity on the island during the Chalco- square R 49, between cisterns 1 and 2 in squares S lithic. Two aspects clearly distinguish this feature, 40, T 40 and T 4 1, and the south-central area of the namely, that the figurine was clearly damaged island particularly along the eastern side in squares (chipped) prior to discard and it was recovered N 24-25 and 0 24-25 provide the highest densities from the base of a pit cut into a trampled earth THE CHALCOLITHIC OCCUPATION OF GERONISOS ISLAND 35

surface. The figurine, along with a number of erga-Myloztthkia, Kissonerga-Mosfilia, surface other stone objects, was deliberately deposited in collected Miliou-Rodoudes and Agios Kononas the pit and 'sealed' with a number of irregular and now Geronisos Island. A parallel reliance on stones heaped on the top of the pit fill (see section stone figurines at Geronsisos, Maa-Palaiokastro 1, Figs 5-8). Parallel examples show a single fig- and Kissonerga-Mosfilia (period 2) shows a con- urine deposited in the upper fill of pit J (unit 67) trast to the more exclusive use of ceramic fig- at Maa-Palaiokastro with KM 35 18 placed in pit urines at Kalavasos-Agious, with Kissonerga- 1657 at Kissonerga-Mosfilia (Thomas 1988,272; Mylouthlcia representing the main intersection of Miles and Peltenburg 1998, 84). At Kalavasos- these different material preferences. Agious both the single stone figurine and the Divergence is also apparent in the individual numerous ceramic examples were found in the sec- character of many of the objects of personal ondary fills of various pit contexts (Todd 1991, 8). ornament, though this sort of difference may have These Early Chalcolithic examples of figurines deposited in pit contexts, including the most more to do with aspects of site hierarchy than evi- recent example from Geronisos, provide a pre- dence of regional variability. Other features lude to the extra ordinary deposit of KM 1015 including the lack of u-shaped openings in CW from Kissonerga-Mosfilia in the subsequent Mid- trays and an impoverished RW motif repertoire dle Chalcolithic (Peltenburg 199 lb). suggest a distinction between smaller sites like Geronisos and Maa-Palaiokastro and the exten- General perspectives sive establishments of Kissonerga and Kalava- The Early Chalcolithic phase in Cypriot pre- sos-Agious. These are differences which also history once labeled simply as the 'transition' need to be tested against the possibility of earlier from the Late Neolithic to the Chalcolithic has and later sub-phases in the Early Chalcolithic (cf. remained difficult to interpret due to an impover- Peltenburg 2003, 257-260). ished record (Peltenburg 1983, 11- 12, 1993, 12). In terms of site character, Early Chalcolithic The latter deficiencies are accounted for by the sites have been interpreted as demonstrating evi- negative features and conditions of erosion defin- dence of a mobile population for whom tempo- ing the period (ibid, 2003, p. 261). As a single rary lightly-built structures were sufficient shel- period Early Chalcolithic site, the addition of ter and who stored their possessions and com- detailed information from Geronisos Island con- modities in a variety of pits and other underground tributes significantly to our understanding of features (Peltenburg 1987, 54, 1998,240). If cor- regional perspectives and material culture change rect, the large pit complex sites like Kalavasos- through the period. The material culture from the Agious, Kissonerga-Mylouthlcia and Kissonerga- Chalcolithic occupation on the island of Gero- Mosfilia, which have been interpreted in terms of nisos is in many ways very typical of the Early specialized activities including artifact manufac- Chalcolithic phase, thereby adding to the general ture and communal grain storage, may provide picture of cultural uniformity during this phase evidence of central sites to which the occupants (Bolger 1987, 73). As suggested by Peltenburg of small periphery sites like Geronisos returned (1983, 12) however, there are divergences within seasonally (Todd 1981, 67; Peltenburg 1980, 5, this overall picture of uniformity, suggesting, 1998, 240-24 1). Alternatively, from evidence instead, that regional variability does not require showing the development of timber buildings complete homogeneity of material culture (Pel- to structures built of stone, Peltenburg (2003, tenburg 1987, 56). That a regional component 270-274) has recently suggested that the latter (even if of degree rather than of kind) does exist ephemeral structures and pit complexes represent in the Early Chalcolithic is suggested by the new settlement foundations. In this model, the close correspondence of GBW between Kisson- later stone-built architecture represents a consol- 36 JOAN BRETON CONNELLY and CAROLE McCARTNEY idation phase of such newly established settle- similarities between the more distant Erimi-Pam- ments. This model, however, cannot apply to sin- poules and Kissonerga-Mosfilia assemblages in gle phase Geronisos where the settlement, though contrast to differences between the latter and near it could be viewed as a new settlement failure, is neighbour Lempa-Lakkous may perpetuate this more likely to represent an example of one of the Ceramic Neolithic social pattern (Bolger 1991, many varied 'lifeways' pursued during the Early 90-92). While the restricted range of RW motifs Chalcolithic (ibid. 273-274). The unique charac- at Geronisos may be related to this kind of social ter of the occupation on Geronisos island, more dynamic, other factors such as chronology, site than other Early Chalcolithic sites, requires an size, function, and length of occupation need to interpretation that views at least part of the pop- be explored further. ulation as relatively (perhaps seasonally) mobile. The extreme site location of the Early Chal- Economic evidence is still forthcoming from colithic occupation on Geronisos provides an Geronisos with faunal and flotation data still to added dimension in terms of site type dynamics be processed though deer is notably present in not faced elsewhere. These dynamics are clearly the faunal sample (Croft pers. comm.). The abun- represented, for example, in the heavily currated dance of deer in earlier phases of the Chalcolith- character of the chipped stone assemblage. Other ic elsewhere has been cited as evidence for wide problems concerning the nature of the occupation territorial access and/or contact (Peltenburg 1Wla, 108). Access to a wide variety of stone similarly revolve around resource availability, raw materials on Geronisos clearly supports the water not being the least among the required suggestion of extensive territory or contacts. It is commodities. The occupation of resource-poor this kind of economic feature that would have Geronisos Island, particularly in light of the rela- stimulated contact in a mobile society rather than tively low Early Chalcolithic site density, is both fragmented it (contra Peltenburg 1987, 54). non-essential and comparatively impractical. Indeed, recent evidence of ceramic diversity in Beyond the type of general site hierarchy rela- the Late Neolithic of Cyprus suggests that near tionships suggested above, and considering the neighbours demonstrated more divergent patterns nature of the features recorded to date, one of RW motif variability than sites located at a underlying motive for this unique Early Chalcol- greater distance within any particular region ithic occupation may be as a prehistoric prelude (Clarke 2001, 73-78). Chalcolithic motifs demon- to the highly symbolic occupations of the subse- strate a significant degree of 'eclecticism', but quent Hellenistic, and Byzantine periods. THE CHALCOLITHIC OCCUPATION OF GERONISOS ISLAND 37

Ware Open Closed Body Rim Base Lug Spout Neck Total GBW 172 236 30 1 65 6 0 3 26 809 % 21.26 29.17 37.21 8.03 0.74 0.00 0.37 3.21 99.99 RW 39 82 121 4 0 0 0 23 269 % 14.50 30.48 44.98 1.49 0.00 0.00 0.00 8.55 100.0 CW 15 0 127 17 17 1 0 0 177 % 8.47 0.00 71.75 9.61 9.61 0.56 0.00 0.00 100.0 Rm 4 5 6 0 1 0 1 0 17 % 23.53 29.41 35.29 0.00 5.88 0.00 5.88 0.00 99.99 Cb 1 0 0 0 0 0 0 0 1 % 100.0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.0 9 1 0 0 0 0 0 0 0 1 % 100.0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.0 ABR 0 264 1752 78 21 0 2 47 2164 % 0.00 12.20 80.96 3.60 0.97 0.00 0.09 2.17 99.99 TOTAL 232 5 87 2307 164 45 1 6 96 3438

Table 1. Preliminary sherd counts and percentages by ware.

Square GBW RW CW Rm Cb 7 Total R49 57 1 88 9 1 6 0 0 756 % 75.53 11.64 12.04 0.79 0.00 0.00 100.0 R46 70 74 2 1 2 0 0 167 % 41.92 44.31 12.57 1.20 0.00 0.00 100.0 Q50S 7 1 50 29 6 1 0 157 % 45.22 31.85 18.47 3.82 0.64 0.00 100.0 Q50NIR50S 62 46 5 1 0 1 115 % 53.91 40.00 4.35 0.87 0.00 0.87 100.00 P50N 33 10 28 2 0 0 73 % 45.21 13.70 38.36 2.74 0.00 0.00 100.01

Table 2. Preliminary intra-site variability of ceramic wares: counts and percentages. 38 JOAN BRETON CONNELLY and CAROLE McCARTNEY

Area Cores CTE Flakes Blades BL Debris ChipsTools Total East-end 9 5 15 0 0 46 5 57 137 % 6.57 3.65 10.95 0.00 0.00 33.58 3.65 41.61 100.01 Cisterns 18 3 34 0 2 89 15 74 235 % 7.66 1.28 14.47 0.00 0.85 37.87 6.38 31.49 100.0 Central 25 14 50 3 6 102 13 88 301 % 8.3 1 4.65 16.61 1.OO 1.99 33.89 4.32 29.24 100.0 North 3 0 5 0 0 22 0 27 57 % 5.26 0.00 8.77 0.00 0.00 38.60 0.00 47.37 100.0 Sanctuary 0 1 0 0 0 5 0 2 8 % 0.00 12.50 0.00 0.00 0.00 62.50 0.00 25.00 100.0 Total 55 23 104 3 8 264 33 248 738 type % 7.45 3.12 14.09 0.41 1.08 35.77 4.47 33.60 99.99

Table 3. Preliminary intra-site chipped stone category counts and percentages (CTE = core-trim- ming elements, BL = bladelets, Debris = all blank fragments and chunks).

Chert Type Number Percentage Translucent 45 11.17 Lefakra-translucent 173 42.93 Lefkara-dense translucent 30 7.44 Lefkara-basal 33 8.19 Moni 68 16.87 Chalcedony 34 8.44 Jasper 8 1.99 Obsidian 1 0.25 Other 11 2.73 , Table 4. Chipped Stone raw material utilization (based on a sample of 403 artifacts). THE CHALCOLITHIC OCCUPATION OF GERONISOS ISLAND 39

- Tool Class East End Cisterns Central North Sanctuary TOTAL Backed 1 2 0 0 0 3 % 1.75 2.70 0.00 0.00 0.00 1.21 Burin 4 8 7 1 0 20 % 7.02 10.81 7.95 3.70 0.00 8.06 Denticulate 1 1 2 0 0 4 % 1.75 1.35 2.30 0.00 0.00 1.61 Glossed 2 5 1 0 0 8 % 3.5 1 6.76 1.14 0.00 0.00 3.23 Notched 1 1 10 3 0 15 % 1.75 1.35 11.36 11.11 0.00 6.05 Multi-tool 3 0 0 0 0 3 % 5.26 0.00 0.00 0.00 0.00 1.21 Perforator 1 0 2 0 0 3 % 1.75 0.00 2.30 0.00 0.00 1.21 Pike esquillee 1 2 0 2 0 5 % 1.75 2.70 0.00 2.7 1 0.00 2.02 Point 1 0 0 0 0 1 % 1.75 0.00 0.00 0.00 0.00 0.40 Retouched 1 5 14 2 0 22 % 1.75 6.76 15.91 7.14 0.00 8.87 Scraper 3 3 6 0 0 12 % 5.26 4.05 6.8 0.00 0.00 4.84 Truncation 1 1 0 2 0 4 % 1.75 1.75 0.00 7.41 0.00 1.6 1 Utilized 33 43 40 16 2 134 % 57.89 58.11 45.45 59.26 100.0 54.03 tool resharpening 4 3 6 1 0 14 % 7.14 4.05 6.8 3.70 0.00 5.63

Table 5. Tool class counts and percentages for each excavation area and the total tool sample. 40 JOAN BRETON CONNELLY and CAROLE McCARTNEY

Square CoredCTE Debris/Blanks Tools R49 6.15 46.15 47.69

Table 6. Preliminary chipped stone intra-site variability. THE CHALCOLITHIC OCCUPATION OF GERONISOS ISLAND

BIBLIOGRAPHY

BAIRD,D. 1985: "Survey in Peyia Village Tei-ri- historic West: Lempa Archaeological Pro- tory, Paphos, 1983," RDAC, 340-349. ject Western Cyprus Survey" in Archaeo- - 199 1: "Independent Variables? A Flexible logical Field Survey in Cyprus: Past histo- Classification of Late Neolithic and Chal- ry, Future Potentials (University of Cyprus, colithic Pottery," pp. 21-27 in J.A. Barlow, Nicosia). D.L. Bolger and B. King (eds) Cypriot BOLGER,D. and J. SHIELS(2003): "The Pottery," Ceramics: Reading the Prehistoric Record Chapter 13 in E. Peltenburg, Excavations (University Museum Symposium Series at Kissonerga-Mylouthkia 1977-1995, vol. 11, Philadelphia). LAP vol. 111.1 (SIMA LXX: 4) (Goteborg). - (in press): "Ceramics" in I. Todd (ed.) Exca- CLARKE,J. 2001: "Style and Society in Ceramic vations at Kalavassos-Agious,VVP.8 (SIMA Neolithic Cyprus," Levant 33, 65-80. vol. LXXI: 8) (Goteborg). CLARKE,J. and I. A. TODD1993: "The Field Survey BETTS,A. (2004): "The Chipped Stone" in I. of Kalavassos-Pampoules," RDAC, 1 1-28. Todd (ed.) Excavations at Kalavassos- CONNELLY,J. B. 1996: "Yeronisos: Sanctuary of Agious, VVP.8 (SIMA vol. LXXI: 8) (Gote- Apollo," The Explorers Club Journal 74.1, borg). 14-18. BOLGER,D. 1985: "From Typology to Ethnology: - 2002: "Excavations on Geronisos (1990- Techniques of the Erimi Potters", RDAC, 1997): First Report," RDAC 245-268. 22-35. - 2005: "Excavations on Geronisos (1990- 1987: "Is There a Western Cyprus? A Chal- 1997): Second Report," RDAC. colithic Viewpoint," pp. 69-79 in D.W. CONNELLY,J. B. and A. I. WILSON2002: "Hel- Rupp (ed.) Western Cyprus Connections lenistic and Byzantine Cisterns on Gero- (SIMA vol. LXXVII) (Goteborg). nisos," with mortar analysis by C. Doherty, 1988: "Chalcolithic Maa: The Pottery," pp. Report of the RDAC Cyprus, 269-292. 390-400 in V. Karageorghis and M. Demas CONNELLY,J. B. and J. MLYNARCZYK2002: "Ter- (eds.) Excavations at Maa-Palaiokastro racotta Oil Lamps from Geronisos and 1979- 1986 (Department of Antiquities of their Contexts," RDAC 293-3 16. Cyprus, Nicosia). D'ANNIBALE,C. 1999: 'The Chipped Stone Indus- 1991: "The Evolution of the Chalcolithic try', pp. 52-55 in D. Rupp (et. al.) "Agios Painted Style," BASOR 2821283, 81-93. Savvas tis Karonis Monastery (Pafos Dis- 1998a: "The Pottery," pp. 93-132 in E. Pel- trict, Cyprus): 1994- 1995 Field Season and tenburg (ed.) Excavations at Kissonerga- 1992- 1995 Artifact Analyses", BASOR Mosfilia 1979-1992, LAP vol. II.IA (SIMA 316, 23-88. vol. LXX: 2) (Goteborg). DIKAIOS,P. 1961 : Sotira (University of Pennsyl- 1998b: "Pottery Archive Report," pp. 121- vania, Philadelphia). 169 in E. Peltenburg (ed.) Excavations at ELLIOTT,C. 1983: "Kissonerga-Mylouthkia: an Kissoneea-Mosfilia 1979-1992, LAP vol. outline of the ground stone industry," Lev- II.IB (Part 1) (University of Edinburgh ant 15, 11-37. Occasional Paper 19, Edinburgh). - 1985: "The Ground Stone Industry," Chap- BOLGER,D.; C. MCCARTNEYand E. PELTENBURG ter. 5.6 in E. Peltenburg (ed.) Excavations (in press): "Regional Interaction in the Pre- at Lempa-Lakkous 1976-1983, LAP I 42 JOAN BRETON CONNELLY and CAROLE McCARTNEY

(SIMA LXX: 1) (Goteborg). KROMHOLZ,S. 1981: "A Preliminary Report on - 1998a: "The Ground Stone Industry," the Earlier Prehistoric Ceramics of the Chapter 7 in E. Peltenburg (ed.) Excava- Vasilikos Valley," pp. 17-55 in J. Biers and tions at Kissonerga-Mosfilia 1979-1992, D. Soren (eds) Studies in Cypriote Archae- LAP vol. II.IA (SIMA vol. LXX: 2) (Gote- ology (University of California Monograph borg). XVIII) (Los Angeles). 1998b: "Ground Stone tools," Chapter 19 MCCARTNEY,C. 1998: "Chipped Stone Report," in E. Peltenburg (ed.) Excavations at Kis- pp. 249-293 in E. Peltenburg (ed.) Excava- sonerga-Mosfilia 1979-1992, LAP vol. tions at Kissonerga-Mosfilia 1979-1992, KIB (Part 1) (University of Edinburgh LAP vol. II.IB (Part 2) (University of Edin- Occasional Paper 19, Edinburgh). burgh Occasional Paper 19, Edinburg). FLOURENTZOS,P. 1997: "Excavations at the - 2001: "The Chipped Stone Assemblage Neolithic Site of Paralimni, A Preliminary from Tenta (Cyprus), Cultural and Chrono- Report," RDAC, 1-10. logical Implications," pp. 427-436 in I. Caneva, C. Lemorini, D. Zampetti and P. GOODWIN,J. C. 1984: An Historical Toponymy of Biagi (eds) Beyond Tools: Redefining the Cyprus (4th edition, vol. I), 1691. PPN Lithic Assemblages of the Levant, GORING,E. 1998: "Figurines, Figurine Fragments, Studies in Early Near Eastern Production, Phalli, Possibly Figurative Worked and Subsistence and Environment 9 (Berlin), Unworked Stones, Unidentifiable Worked ex oriente. Stone and Pottery Fragments," Chapter 6 in - 2003: "The Chipped Stone," Chapters 2 E. Peltenburg (ed.) Excavations at Kisson- and 17 in E. Peltenburg, Excavations at erga-Mosfilia 1979-1992, LAP vol. ILIA Kissonerga-Myloutlzkia 1977-1995, LAP (SIMA vol. LXX: 2) (Goteborg). vol. 111.1 (SIMA LXX: 4) (Goteborg). GUNNIS,R. 1936: Historic Cyprus. A guide to its - in press: "Prehistoric Chert Use in Cy- towns and villages, monasteries and cas- prus", in C. Delage (ed.) Geology and tles (Nicosia), 380. Archaeology of Siliceous Rocks in the Near HELD, S. 0. 1992: Pleistocene Fauna and East: Availability, Characterization and Holocene Humans: A Gazeteer of Paleon- Prehistoric Exploitation. tological and Early Arclzeological Sties on MILES,D. and E. PELTENBURG1998: "Catalogue Cyprus (Jonsered). of Pits," Chapter 15.5 in E. Peltenburg (ed.) HOGARTH,D. G. 1889: Devia Cypria. Notes of Excavations at Kissonerga-Mosfilia 1979- an archaeological journey in Cyprus in 1992, LAP vol. II.IB (Part 1) (University of 1888 (London), 12. Edinburgh Occasional Paper 19, Edin- JACKSON,A. (2003): "The Ground Stone In- burg). dustry," Chaper 15 in E. Peltenburg, Exca- MURRAY,M. A. 1998: "Archaeobotanical Report," vations at Kissonerga-Myloutlzkia 1977- Chapter 11 in E. Peltenburg (ed.) Excava- 1995, LAP vol. 111.1 (SIMA LXX: 4) tions at Kissonerga-Mosfilia 1979-1992, (Goteborg). LAP vol. II.IA (SIMA vol. LXX: 2) (Gote- KARAGEORGHIS,V. 1983: "An Archaeological borg). Survey and Trial Excavations on the Small PELTENBURG,E. 1979: "Lempa Archaeological Island 'Geronisos' off the Pafos Coast", Project, Cyprus 1976-77: Preliminary Annual Report of the Department of Antiq- Report", Levant 11, 9-45. uities for the Year 1982 (Nicosia). - 1980: "Lempa Archaeological Project, THE CHALCOLITHIC OCCUPATION OF GERONISOS ISLAND 43

Cyprus 1978: Preliminary Report", Levant PETRIDES,G. and M. HARALAMBOUS,1981: Yero- 12, 1-21. nisos Island Geological Geotechnical - 1983: "The Prehistory of West Cyprus: Report (Ministry of Agriculture and Natur- al Resources, Geological Survey Depart- Ktima Lowlands Investigations 1979- ment) (Nicosia). 1982", RDAC, 9-55. SOUTH,A. 1985: "Figurines and Other Objects - 1985: Excavations at Lempa-Lakkous from Kalavassos-Agious," Levant 17, 65- 1976-1983, LAP I (SIMA LXX: 1) (Gote- 79. borgj. STEWART,J 198 1: "A Brief Guide to the Pottery - 1987: "A Late Prehistoric Pottery Sequence Wares of Miliou Rodoudes," 42-46 in E. for Western Cyprus," pp. 53-64 in D.W. Peltenburg (et. al.) "Lempa Archaeological Rupp (ed.) Western Cyprus Connections Project, Cyprus 1978: Preliminary (SIMA vol. LXXVII) (Goteborg). Report", Levant 13, 28-50. - 199 1a: "Local Exchange in Prehistoric - 1985: "Discussion of Areas I and I1 Ceram- Cyprus: An Initial Assessment of Picro- ics," Chapter 5.5 in Excavations at Lempa- lite," BASOR 282, 107-126. Lakkous 1976-1983, LAP I (SIMA LXX: - 199 1b: "Lempa Archaeological Project 1) (Goteborg). 112.1A Ceremonial Area at Kissonerga" THOMAS,G. 1988: "Part 1: The Maa Chalcolithic (SIMA LXX: 3) (Goteborg). Excavations", 267-289 in V. Karageorghis - 1993: "Settlement Discontinuity and and M. Demas (eds) Excavations at Maa- Palaiokastro 1979-1986 (Department of Resistance to Complexity in Cyprus, c. Antiquities of Cyprus, Nicosia). 4,500-2,500 BCE", BASOR 292, 9-23. TODD,I. 198 1 : "Current Research in the Vasilikos - 1998: Excavations at Kissonerga-Mosfilia Valley," pp. 57-68 in J. Reade (ed.) Chal- 1979-1992, LAP vol. II.IA (SIMA vol. colithic Cyprus and Western Asia (Lon- LXX: 2) (Goteborg). don). - 2001: "Neolithic Dispersals from the Lev- - 1982: "Vasilikos Valley Project: Fourth antine Corridor: a Mediterranean Perspec- Preliminary Report, 1979-1980", JFA 9, tive," Levant 33, 35-64. 35-77. - 2003: Excavations at Kissonerga- - 1985: "Excavations in the Vasilikos Valley," Mylouthkia 1977-1995, LAP vol. 111.1 pp. 8 1-91, in V. Karageorghis (ed.) Archae- (SIMA LXX: 4) (Goteborg). ology in Cyprus 1960-1985 (Nicosia). PELTENBURG,E. and S. SPANOU1999: "Neolithic - 1991: "The Vasilikos Valley and the Chal- Painted Pottery from Agios Epiktitos- colithic Period in Cyprus," BASOR 2821 Vrysi," RDAC, 13-34. 283, 3-16. JOAN BRETON CONNELLY and CAROLE McCARTNEY

OLavaoxacpkg TOU navsn~crc11j.uo~nig Nkag Yopxqg kxouv cpkpei OTIIV emcpdveia oqpav~txd nphipa XaXxoXiOixd evpipa~ao~pnep~Xap~avo~ikvcov xepaptxiig, Xetaoj-ikvcov XiOivcov epya- Xeicov xoa XiQivcov Xeni6cov TO onoia xa0opi£,o'upta eviaia nepio6o xa~oixqo~lg~OU xpovoko- YE~TOLOTO ~tphT0[.UOO 'C?'lgT~T~QTT)~ ~L~LET~U~ 7G.X. AUTO TO UXLV.O OV~~~X~ELOTIIV %u~'UTE@~ xa~avoi1qTWV ~ontxhvyvcoptoj.idrcov ~qgn~pioxijg xai ~igo~oth6eig noXi~io~txkg aXXaykg xa9 O?L~ 6idpxe~ame; ~;€@Lo~oH L~L~LT&QOTI~T~ Tllg %UTO~~I~OI]~ 0Ti1 r&p0v1100 +&i'Cai OTO yeyovog OTL ~~~Xdxio~ovpkpog nX116'uopo'u n1g IIphij.i~lgXaXxoXt0txfig pe~axtv~i~a~nard 6iao~iipa~a(iocog enoxiaxd). Yndpxovv ~piaxDpia av~~npoocone~~~xdei611 xepapix~ig, n~co~io~cog ~11gS~tXnviig S~iX@co~iig KOO oe LUKQOTEQ~IavaXoyia ~i1gEp~Opfic; ndvco oe AEWOxai 11 ~paxeiaaxoopq~q xepa~iixfi. 2'a'i)~oi)gTOV~ Tunow ~i~topoi+ieva npoo9Sooqie kva ptxpo~epoapt0po Gpauo~id~covxepaptxiig ~OUav~~npoocone'uo~v na.~akXayfi TOU EQuQQoV Movoxpcopo'u xai KTEVLOTOW~'uno~. IIapopol~g,TO. epyaXeia OCTO Xetaoj~kvilnstQa kxouv nphipa xa~ax~il~io~txd.Eivai XUQLCOC; narao~evaqkvamo 6ia~dq xai oz)~,ixepiXa[.ipdvo'uvepyaXeia xonfig oncog aEivsg, oxendpvia, opiXeg xai Xem6oei61i epyakeia 000 enioqg xai epyaXeia aXkoparog now UJIEQS~OUVaptOpq~txd oncog 01 ~LUXO~ETQ~~,x~i~ou/i)Xoi 611p11~(~axhv, cqVQeg, xonavot, o~ikpoo~iipeg,opiXeg xai yo~6oxkpia.H napovda EQIJGQOTJipappi~ii d TOV Andpa ~E~~VELOTLfioav yvho~egTWV nph~ov uXhv 6ta9koipcov oe exeiq qvnepioxr). OL6idcpopot ~11noiTW epyakeicov ~no6etxv6o'uvTOGO TUV napaoxeufi ~pocpipcov000 xca xa~aoxet)ao~ixhv6paoqpio~ipv. Opiopkva ~~yaXe1,a~poxiopa- ~ogI o~tkphpa~og eivai e~uoq~opkva~ie otphq n'tiXo11 pe noXXanXkg pap6hoeig o~qven~cpuve~a. AUTOiocog UJTOVOE~T~V ene&pyaoia apyiXow 11 y11ipov xai niGavov va ano6eixvi)ei T~Vnapaywyii xepaptxfig OTO vqoi TIIS I?ep~vfp~'u. rikpav TW 6110 TQLTCOV TII~av'k'koy~ig neXexq~hv ne~phv elvai xapcopkveg ano ei611 xepa- TOXLOOU xaX4g pkxpt ~'lpqXijgnoioqxag. En~xpato~ho 6iacpav1ig xepa~oXi0ogAeuxdpcov xai o ~iavpoI yxpi£,o xepato'kiQog Mov~ig.H pydX11 noixtXia ~Q~TCOVuXhv ano nk~paUXOVOEL enacpkg e~peiagkx~ao11g. 'Eva pova6ixo xop@~tuaXh6oug o1)t6iavo'u (~piip.np~opa~~xo'u Xen16i01) peockoii peyk9ovg) eivca Locog xa~dXotnomo T~V Axepa~iix~j NeoXiOtxii nepio6o. Avdpeoa ma qooconw& avcixei~ievacrupn~ptXap~dvov~at pia 11~iid1ig ydv~pa x'uXtv6p~xo'u oxfipa~ogmo oepnev~ivi1, kva 6txcovixo nepian~oan6 n~xpoXt90,kva ~LXCOVLKOmpian~o an6 xaXxq6ovio Xi90 xai kva enine6o nepiax~0/(3o~oaXo. STUVAva~oXixfi dxpq TOU vqoiou pp60ilxe kva av6~conopo~cpoJI~TQIVO et6hXi0 mno9e~~kvo os kva xvxX~xoXdxxo oxap~ikvooe 6dne60 ~qgXakxoXtOiwig ne~io6ou.2/co ei6hXio cpaive~aiOTL OKOJU~~~~tp~~~fi@i~x& <11pid JtQiv ax0 TIlV €v~J~oOâ‚TOU p60a OTO T£(p~h6&xh~i,CX xai Ti5 pay~opkvqGOTO cpcozid nk~pec;JWU pe~acp6~6qxavano ak'ko6 xai oxendo~qxavan6 ndvco pe peydXeg X~TQE~.Mati pe TO ei6hXio Pp6911xav xepaptxfi, XiGiveg Xeni6eg, 9pa'uopa xaXa£,ia 0pdopa moJT~TQLV~ xoIJna, repdxio adnk~pivq nXdxa / nidxo, [wXone~peg/ ocp'Upeg, OTLX~W- ~~ipag,ma a&va xoa Svag ~ptnTripag. To ei6hXio xai TO. ~noXotnaav~ixeipeva ZOU oxonipa TOJ~OO~T~~G~~U~pa

Fig. 1. Aerial view of Geronisos from west, Maniki rocks visible at south.

Fig. 2. Aerial view of east end of Geronisos, from west. East Building, trenches R 49 and R 46. 46 JOAN BRETON CONNELLY and CAROLE McCARTNEY

Fig. 3. Aerial view of east end of Geronisos, from south. Fig. 4. Aerial view of East Building and R 49, East Building, trenches R 49 and R 46.

Fig. 5. Trench R 49 with boulders sealing pit, from north. Fig. 6. Trench R 49 showing pit, from north,

Fig. 7. Trench R 49 showing pit, close up from east. Fig. 8. Trench R 49, with pit cleaned out, from north. THE CHALCOLITHIC OCCUPATION OF GERONISOS ISLAND 47

Fig. 9. Objects found from levels 5 and 6.1, R 49.

Fig. 11. Stone post-pad shown in situ in Fig. 10. Objects found in ritual deposit of R 49. 0 22, Central South Sector. JOAN BRETON CONNELLY and CAROLE McCARTNEY

Fig. 12. Glossy Burnished bowl fragment, P.97.18. Fig. 13. Course ware tray, P.97.08.

Fig. 15. Red-on-White sherd, P.97.07

Fig. 14. Red-on-White sherd, P.96.32.

Fig. 16. Red-on-Whiie sherd, P.97.05.

Fig. 18. Axe,

Fig. 17. Pointed Base (abraded), P.97.06.

Fig. 19. Axe, St.96.41. THE CHALCOLITHIC OCCUPATION OF GERONISOS ISLAND

Fig. 22. Quei-n, St.92.13

Fig. 20. Chisel, St.93.61.

Fig. 21. Cupped stone, St.96.40.

Fig. 23. Haminerstonelgrinder, St.93.78

Fig. 25. Bead, St.93.25.

Fig. 24. Grinder, St.97.49. JOAN BRETON CONNELLY and CAROLE McCARTNEY

Fig. 27. Adze, St.97.25.

Fig. 26. Axe, St.97.36.

Fig. 32. Cupped stone, St.93.14.

Fig. 3 1. Pestle, St.97.39

Fig. 29. Chisel, Fig. 33. R~~bberlcup-marked stone, St.97.3 1. Fig. 30. Grinder, St.97.37. st.97.45.

Fig. 34. Biconical picrolite pendant, St.96

Fig. 35. Biconical pendant, St.97.43.

Fig. 28. Chisel, Fig. 36. Pebble Fig. 37. Stone figurine (front), Fig. 38. Stone figurine (back), St.97.01. 'pendant', St.97.57. St.97.55. St.97.55.