Institutionen för arkeologi och antik historia

The Late Bronze Age Sanctuary at Ayios Iakovos: Dhima Revisited

Adam Lindqvist

Masteruppsats 45 hp i Antikens kultur och samhällsliv VT 2017 Handledare: Michael Lindblom

Abstract

Lindqvist, A. 2017. Ett återbesök i den sena bronsålders helgedomen i Ayios Iakovos: Dhima

År 1929 grävde den Svenska Cypernexpeditionen ut en helgedom daterad till den tidiga delen av Sencypriotisk II (1400-1340/1315), namngiven Ayios Iakovos: Dhima. Kring ett terrakotta-kar återfann arkeologerna flera värdefulla och exotiska föremål. Sedan dess har platsen och dess fynd tolkats på många olika sätt, utan någon egentlig klarhet. Genom att göra en systematisk studie över det hittills opublicerade skärvmaterialet har nya slutsatser om platsen kunnat läggas fram. Tidigare tolkningar om ett kronologiskt gap under Sencypriotisk I kan nu ifrågasättas. Det finns belägg för ett kontinuerligt bruk från Mellancypriotisk III fram tills platsen övergavs under Sencypriotisk II. Den stora mängden slutna kärl, förknippade med transport av väldoftande oljor och salvor, vittnar om de aktiviteter som en gång företogs på platsen. Dessutom visar närvaron av typiska rituella dryckeskärl ett av de tidigaste exemplen på utvecklingen av Cypriotisk rituell tradition, nu separerad från de tidigare starka banden till gravriter.

i Acknowledgements

I would first like to thank my supervisor, Dr Michael Lindblom of the Department of archaeology and ancient history at Uppsala University. Without his invaluable guidance and persistent help, this thesis would not have been possible.

I would also like to extend my thanks to the staff of the Mediterranean Museum in Stockholm, especially Dr Christian Mühlenbock, Dr Eva Myhrdal and Emma Andersson for their assistance in accessing the material that makes up the core of this study.

Finally, I must express my gratitude to my parents and my partner, for providing me with unfailing support and continuous encouragement throughout my years of study and through the process of researching and writing this thesis. This accomplishment would not have been possible without them. Thank you.

Adam Lindqvist

ii CONTENTS ABSTRACT ...... I 1. INTRODUCTION ...... 6 1.1 PURPOSE OF THESIS AND RELEVANT RESEARCH QUESTIONS ...... 6 1.2 APPROACH ...... 7 1.3 RE-EXAMINING AN OLD EXCAVATION – PROBLEMS AND POSSIBILITIES ...... 7 2. THE SWEDISH EXPEDITION ...... 9 3. WHAT MAKES A SANCTUARY? ...... 10 3.1 THE EXCAVATION AND IDENTIFICATION OF AYIOS IAKOVOS: DHIMA ...... 10 3.1.1 The excavation ...... 11 3.1.2 The documentation ...... 12 3.1.3 The stratigraphy ...... 13 3.1.4 The architecture ...... 15 3.1.5 The material culture ...... 16 3.2 AYIOS IAKOVOS IN TIME AND SPACE...... 20 4. PRESENTATION OF THE PREVIOUSLY UNSTUDIED MATERIAL ...... 23 4.1 THE CERAMIC ASSEMBLAGE AT AYIOS IAKOVOS: DHIMA ...... 23 4.2.1 Middle Bronze Age ...... 24 4.2.2 Late Bronze Age ...... 30 4.2 COMMENTS ON THE POTTERY ...... 41 5. RE-CONTEXTUALIZING AYIOS IAKOVOS ...... 57 5.1 PREVIOUS INTERPRETATIONS OF AYIOS IAKOVOS: DHIMA IN A NEW LIGHT...... 57 5.1.1 Dhima as a dual deity site ...... 57 5.1.2 Dhima as a typical Cypriot Sanctuary ...... 59 5.1.3 Dhima as a short-lived sanctuary and funerary site ...... 61 5.1.4 Dhima as a last resort during times of distress ...... 63 5.1.5 Dhima as an eastern Mediterranean sanctuary ...... 65 5.2 DHIMA IN THE LIGHT OF NEWLY STUDIED POTTERY ...... 67 5.2.1 The construction of the floor and the chronology of the sanctuary ...... 67 5.2.2 An enclosure ...... 68 5.2.3 Two areas in the sanctuary...... 68 5.2.4 The pottery and its function during the LCII period...... 68 5.3 WHAT DOES THE RE-CONTEXTUALISATION MEAN FOR THE UNDERSTANDING OF EARLY LCII CULT?...... 70 BIBLIOGRAPHY ...... 72 LIST OF ILLUSTRATIONS ...... 77 LIST OF TABLES ...... 77 LIST OF PLATES ...... 77

iii Table 1. General terminology and Abbreviations. Pottery Wares Chronology Base Ring BR Early Cypriot EC Black Slip BS Middle Cypriot MC Cypro-Mycenaean Cypr. Myc. Late Cypriot LC Grey Ware GW Late Helladic LH Plain Ware handmade PWHM Plain Ware wheelmade PWWM Red Lustrous RL Red-on-Black RoB Red Polished RP Red Slip RS White Painted handmade WPHM White Painted wheelmade WPWM White Porous WPo White Slip WS

Table 2. Relative and (approx.) absolute chronology of Cyprus and the Aegean (after Crewe 2007, and Mountjoy 1986, ). Middle-Late Cypriot Periods Mainland Aegean periods MCI (1950-1850) MCII (1850-1750) MCIII (1750-1650) LCIA (1650-1550) LHI (1550-1500) LCIB (1550-1400) LHIIA (1500-1450) LHIIB (1450-1400) LCIIA (1400-1375) LHIIIA1 (1400-1350) LCIIB (1375-1340/15) LHIIIA2 (1350-1300) LCIIC (1340/1315-1200) LHIIIB1 (1300-1225) LHIIIB2 (1225-1175) LCIIIA (1200-1100) LHIIIC Early (1175-1130) LHIIIC Middle (1130-1090) LCIIIB (1100-1050) LHIIIC Late (1090-1060)

iv

Figure 1. Map of Cyprus with important Late Cypriot sites (from Knapp 2008a).

v 1. Introduction

In 1999, Jennifer Webb defined the Late Bronze Age Sanctuary at Ayios Iakovos: Dhima as an “extramural cult locale possibly associated with mortuary ceremonial (of which Ayios Iakovos Dhima is the only example and might be unique)”. However, the interpretations of this site have been many and widely differing. All primarily based on the publication by the Swedish Cyprus Expedition, a report that neglected a most of the pottery and focused only on the complete finds. The numerous gold objects, the well-preserved ceramic vessels and the many imports have been puzzling to many. It is cited as one of the best examples of a largely undisturbed Cypriot sanctuary and is usually included in studies on early Cypriot cult or ceramic studies on the unusually large number of Red Lustrous arm vessels.1 But beyond that, it is only mentioned in passing as a cultic locale found in the Cypriot hinterlands.2 The sherd material from the sanctuary has never been thoroughly studied and it is time to update Dhima with a century worth of archaeological progress in Cypriot prehistory.

1.1 Purpose of thesis and relevant research questions The aim of this thesis is to revisit the sanctuary at Ayios Iakovos: Dhima, to gain a better understanding of its role in the Cypriot ritual tradition. Focusing on the unpublished pottery, this study will re-evaluate previous interpretations of Dhima to determine which interpretations are, and which are not, supported. With a more thorough understanding of the locale, this study will then contextualise Dhima within the broader ritual landscape of Cyprus during the end of the Middle Cypriot and beginning of the Late Cypriot period. On an initial level, the aim is to answer some of the basic questions about the sanctuary, such as: What constitutes the complete ceramic assemblage at Ayios Iakovos: Dhima? Does the additional sherd material support the dating reached by the previously published vessels and is there, as has been suggested, a hiatus at the sanctuary between Middle Cypriot (MC) III and Late Cypriot (LC) II? Is there a pattern to which types of vessels were found? Beyond these initial observations, this thesis will look deeper into the activities conducted at Dhima, attempting to establish a further understanding of events that occurred: What kinds of activities were performed at Dhima? What type of sanctuary is Dhima and how, if at all, does it relate to other Cypriot sanctuaries? The questions will then be summed up and brought into context to try and further understand Ayios Iakovos: Dhima as a Cypriot sanctuary in a Cypriot context.

1 See Wright 1992; Eriksson 1993; Webb 1992; Webb 1999; Hitchcock 2011 among others. 2 See Wijngaarden 2002, Steel 2004a and Knapp 2008.

6 1.2 Approach To be able to understand this site, it has been necessary not only to re-evaluate the published record but also to do a more comprehensive study of the unpublished material. All previous studies of Ayios Iakovos: Dhima have been based on three main sources: Einar Gjerstad’s account of the excavation published in 1933, the main excavation report written by Erik Sjöqvist, published in 1934 and Erik Sjöqvist’s dissertation, published in 1940. Left out of these reports are roughly 2700 sherds, which were collected and placed in storage without further study. This thesis will systematically re-examine all sherds available at the Mediterranean Museum in Stockholm. It will then integrate the assemblage with the already published finds to determine if and how it changes the interpretation of the Dhima sanctuary. In addition to the ceramic assemblage itself, original documentation has been used where available, old plans and sections have been redrawn for this thesis in order to get as close as possible to the original data. Initially a full recount of the Ayios Iakovos: Dhima excavation will be presented, based on the available documentation. In addition to the published record, this presentation will utilise as much of the original documentation as possible. This includes a more detailed look at the stratigraphy of the site, a description of all constructions as well as a more thorough presentation of the different concentrations of finds published by the Swedish Cyprus Expedition in 1934. Following will be the presentation and main analysis of the unstudied pottery of Dhima. The first part will deal with the specific wares and sherds, working with shapes and distribution within the Dhima sanctuary. Each ware will be contextualised within the Cypriot material culture as much as possible and important sherds and shapes will be further discussed. The determination of shapes will be done to establish if there is a pattern in certain types of vessels and if these can be linked to any activities conducted at Dhima. This will be followed by a discussion of the entire assemblage and the interpretations that can be drawn, both general and specific. With the material culture at Dhima established, previous studies3 of the sanctuary will be evaluated separately in light of the newly studied material. Interpretations and theories will be tested to determine the ritual context of Ayios Iakovos: Dhima. With each previous major study considered, Dhima will be reinterpreted, based on new observations about the ceramic assemblage and how it related to the previously published material. Dhima will then be brought into a larger Cypriot context. To best facilitate this and in order to determine its role in the Cypriote ritual landscape, Dhima should be compared to other contemporary sanctuaries. While not entirely contemporary, two sanctuaries, Myrtou-Pighades and Athienou Bamboulari tis Koukounninas are commonly described together with Dhima as the first proper sanctuaries in Cyprus. 4 The comparative analysis will establish common characteristics of the material assemblage, as well as, probable activities conducted at each site, in order to give a clearer perspective on the early development of Cypriot cult.

1.3 Re-examining an old excavation – problems and possibilities Approaching an archaeological material almost 100 years after its excavation comes with many problems. The primary focus of this study is the sherds from the excavation at Ayios Iakovos: Dhima that were never published. The 2700 sherds were shipped to from Cyprus at the end of the expedition in 1932. Since then, they

3 Specifically studies by Sjöqvist 1940; Wright 1992; Webb 1999; Collard 2008 and Hitchcock 2011. 4 See Wijngaarden 2002, Collard 2011.

7 have been left largely unstudied for the past 80 years, except a study by Paul Åström in 1972, where the author mentioned some of the better-preserved pieces. His focus was, however, mainly on the Mycenaean pottery and the results are recorded in his very thorough work on the pottery of the Late Bronze Age in Cyprus.5 This study aims to bring the last remaining sherds and the site into context and to present the results in a more transparent way. The ceramic material is studied in its entirety and classified to the best of the author’s ability, based on their fabric, production, decoration and shape. During the examination, individual diagnostic sherds, such as rims, handles and bases, as well as distinct types of decoration, were catalogued. These sherds were examined more thoroughly and more fully described in Table 3. Any sherds determined to be from a later period than LC period have been noted and separated fro the study.6 As with any selection, there are limitations. In this case, there are two main constraints. The first is the availability of representative samples of different wares and shapes. Particular wares or shapes might not have diagnostic pieces preserved, and its information would, therefore, be lost in the analysis. A way to prevent this is to be less strict about which sherds to consider diagnostic, but the analysis would undoubtedly be less precise. Secondly, the selection itself is subject to the author’s reasoning of what is worth recording. The aim of this study is to have a sample that is as diverse as possible. Therefore, as wide a range as possible of different wares and shapes has been catalogued. Additionally, vessels, which frequently occur in the material, in a range of sizes, are described in full but illustrated in a representative way. In such cases, the best-preserved sherds have been drawn. Given the very simple stratigraphy at Dhima, it is not possible to determine secure, dateable layers succeeding each other. As is shown in Fig. 6, the only clear marker in the stratigraphy that separates the earliest part of the sanctuary from the latter is the floor. The floor defines the closing of the MCIII-pit and is itself defined by the finds above.7 The excavators did not separate the sherds by context, but by arbitrary levels, each measuring roughly 20cm.8 Due to the superficial nature of the sanctuary and the evidence of farming, it is likely that sherds have been moved and mixed at the upmost level of 0–20cm. The 20–40cm level seem more secure since the plough layer only appear to reach c. 20cm down across the site. The well-preserved finds around the terracotta basin indicate as much. Due to this stratigraphic and archaeological situation, any interpretation has to be made from the relative depth, site distribution and the previously established function, if any, of vessels of similar ware and shape. The sherds were recorded from the grid system used during excavation, each measuring 2x2m (see Fig. 2). Sherds from about half of the grids are present in the preserved material, of which there is a decent spread across central and outlying grids, as well as a few grids outside the floor area. It is uncertain what happed to the sherds of the grids that are not present in the Mediterranean Museum collections. No doubt there were sherds in these grids as well. Presumably, they were either discarded during excavation or left in storage in Cyprus. Nevertheless, the available material should still offer a representative picture of the full ceramic assemblage at Dhima.

5 Åström 1972b. 6 There is evidence from later activity in the area from both Archaic/Classical, Roman, and later periods in the form of sporadic sherds. Sjöqvist considered the stone constructions seen in Grid A4 to be Classical/Byzantine (see Fig. 2 and 6, Profile III). 7 Gjerstad et al. 1934a, 360–361. 8 The only exception is the context recorded as ”On the first floor”, abbreviated in Table 3 as 1stFl. Unlike the name suggest, there is no second floor. There is only one floor layer discussed throughout this study.

8 2. The Swedish Cyprus Expedition

The Swedish Cyprus Expedition has a unique beginning, being more of a chance than a planned venture. It began in August 1922, when the Swedish archaeologist Axel W. Persson was travelling on a train between Hungary and Yugoslavia. On board this train he met Luki Z. Perides, the Swedish Consul in Cyprus, who wanted to borrow 15 pounds, to which Axel W. Persson agreed. This small loan started a chain of events, which ultimately lead to the Swedish Cyprus Expedition. After the short meeting on the train, Persson’s assistant, Einar Gjerstad, was invited to Cyprus to meet with Perides. Gjerstad arrived in Cyprus in 1923 and he started studying excavated material, surveying the island and participating in excavations. Gjerstad laid the foundation for future archaeological research in Cyprus with his thesis ”Studies on Prehistoric Cyprus”, published in 1926. During this time, plans for a large archaeological expedition started to form. In September 1927, the three Swedish archaeologists Einar Gjerstad, Alfred Westholm and Erik Sjöqvist, accompanied by the architect John Lindros, departed for Cyprus. They were to be known as the Swedish Cyprus Expedition (SCE). During the following four years, until March 1931, the expedition excavated up to 25 different sites on the island, covering a chronological span of almost 10.000 years, from the Chalcolithic period to Roman times. A wide range of different sites was excavated, such as tomb, sanctuaries, settlements and a Roman theatre. One of the most well known sites was the open-air sanctuary at Ayia Irini, where over 2000 terracotta sculptures were found in situ around an altar.9 The aim of the expedition was to create a complete chronological sequence over the early history of Cyprus; this was mainly accomplished through the excavation of almost 375 tombs. The information gained in these burials was then applied to further understand the chronology of the other types of sites excavated.10 Part of the agreement between Cyprus and the Swedish Cyprus Expedition was that a portion of the excavated finds would be given to Sweden. When the expedition was over, more than 10.000 artefacts (65% of the finds) were shipped to Sweden, as well as 5.000 boxes filled with sherds.11 Pierides changed Cypriot law to make the agreement possible, allowing such a large amount of finds to be shipped abroad. The finds arrived in Sweden in 1932 and were subsequently published in twelve volumes titled The Swedish Cyprus Expedition I-IV:3 between 1934 and 1972. The collection housed in the Mediterranean Museum in Stockholm still constitutes the largest collection of Cypriot antiquities outside the island. While much has been published, there are still years of research waiting to be done.

9 Styrenius 1994, 7–9. 10 Winbladh 1992, 7. 11 Winbladh 1992, 7.

9 3. What makes a sanctuary?

The interpretation of Ayios Iakovos: Dhima as a sanctuary is one of the few things that have not been questioned by later studies. However, given the isolated location of the site, the limited architecture and the characteristics of the material culture, any other interpretation seems unlikely. There are no indications of any domestic activities, such as a roofed living area or household artefacts. Likewise, none of the finds suggests activities related to production or farming. The lack of any roofed structure contradicts communal long-term storage and the absence of any human remains speaks against a burial. Therefore, the most plausible identification remains a ritual locale.

Figure 2. Plan of Ayios Iakovos: Dhima. The three profile sections (I-III) in the plan are illustrated separately in Fig. 6 Redrawn from the original.

10 3.1 The excavation and identification of Ayios Iakovos: Dhima

3.1.1 The excavation From June to August 1929 the Swedish Cyprus Expedition excavated the necropolis at Ayios Iakovos: Melia, containing burials dating to the Middle and Late Cypriot period. Hoping to find a settlement connected to these burials, the surrounding area was surveyed. A local villager from Mandres informed them that he had found artefacts while ploughing his fields. The area was examined, and excavations began in July on one of a series of small hilltops, called Dhima, located between the villages Ayios Iakovos and Mandres. Initially, a single test trench was made across the hilltop to the bedrock, at a depth of c. 80cm. Gjerstad describes the initial excavation as sensational: “After 10 minutes of excavation, the gold started to drizzle out of the soil, only 20cm from the surface”.12 The area was then systematically excavated in a grid system from (See Fig. 2. The X-axis is labelled A–F and the Y-axis 1–5, each grid measuring 2x2m. Some grids have been added to the profiles that extend beyond Fig. 2). The excavations of the hilltop were carried out for the full month of July in 1929.

Figure 3. View of the sanctuary at Ayios Iakovos: Dhima from the west. The site was identified as a sanctuary during the excavation, to which the field drawings give witness.13 In the final publication, it is noted: “There can be no doubt that the excavated area was once a cult place.” This interpretation was based on the architecture, the numerous rich finds and “above all” the Red Lustrous arm vessels, which at the time were assumed to be incense burners.14 The boundaries of the site were determined by the extent of a floor layer made from a mix of small stones and limestone. The floor had the rough shape of a circle

12 Gjerstad 1933, 97. 13 All original drawings are marked ”The Bronze Age Sanctuary”. 14 Gjerstad et al. 1934a, 359.

11 measuring c. 10m in diameter. In the centre of the floor, a single wall separates the area in two, an eastern and a western half. The eastern half contains two large stone platforms, one c. 2.4m in diameter and the other c. 1.2m in diameter. Underneath the floor level of the eastern half, a circular pit was found. The contents of the pit and the fact that the floor covers it suggest, as mentioned by Sjöqvist that it is an earlier feature on the site. It will henceforth be referred to as the “MCIII-pit.” In the western half, there was a terracotta basin, placed in a pit, which had been cut into the bedrock. It was around this basin that the majority of the finds were excavated. Southeast of this basin was a small pit, trapezoid in shape, containing a few artefacts of bronze. These are the only architectural remains. It has, however, also been theorised that the area was once fenced in. Sjöqvist, who supported his arguments with pictographic evidence of cylinder seals, initially argued for the presence of the enclosure.15 Later studies have agreed that there must have been some enclosure surrounding the site, built from a perishable material.

Figure 4. View of the sanctuary at Ayios Iakovos: Dhima from the north.

3.1.2 The documentation The results and finds were all published in the first volume of the Swedish Cyprus Expedition, Finds and Results of the Excavations in Cyprus 1927-1931, vol.1. Much of the original documentation is also preserved in the collections of the Mediterranean Museum in Stockholm. Original photographs, field drawings and many legal documents surrounding the excavation of Ayios Iakovos are still available for study.16 Unfortunately, the excavation diaries kept by Sjöqvist are missing for Ayios Iakovos: Dhima. Without them, the circumstances and early thoughts during the excavation will be more problematic to understand. The remaining documentation, however, can

15 Gjerstad et al. 1934a, 359; Sjöqvist 1940, 3. 16 All photographs relating to the excavation of Ayios Iakovos: Dhima has been included in this study. See Figs. 3–5, 7–8, 10 and 12.

12 be of value in the revisit to this sanctuary. Several original drawings, made by John Lindros between the 17th and 19th of July in 1929 offers information not made available in the final publication. For the purpose of this study, each original plan and profile have been redrawn (see Figs. 2, 6 and 9 and 11).

3.1.3 The stratigraphy The stratigraphy of the sanctuary is illustrated in the publication through three drawing of the sections.17 The sequence consists of four stratigraphic layers. These are not described in much detail in the publication, but the original field drawings provide some additional information in this respect (for all redrawn profiles, see Fig. 6). Sjöqvist describes the stratigraphy of Dhima as simple. While the published stratigraphy follows this sentiment, the original drawings are somewhat more detailed. The first layer (Layer 1) is described as “plough layer.” It covers the entire site and reaches beyond the excavated area. Following is a layer described as “undisturbed soil,” (Layer 2). Like Layer 1, it covers most of the excavated site, except the far southeast part, where it thins out roughly halfway through Grid F4, where the floor ends. Just west, in Profile II, it continues beyond Grid F2 and the floor area. In both Profile II and III Layer 2 continues beyond the limits of the site. Directly underneath the second layer is the limestone floor (Layer 3). This layer is limited to the B–F and 1–5 grid system. The two stone platforms, the wall, the terracotta basin and the trapezoid pit cut this layer. The drawings indicate that the floor is roughly even in thickness, and slopes to the east/southeast. The layer is noticeably lower on the eastern side of the wall, compared to the western side. Below the floor is a “sterile

Figure 5. View of the MCIII pit, small platform and central wall. Photo taken from the north.

17 Gjerstad et al. 1934a, 55.

13

III (fromIII top to bottom). Redrawn from originals. – . Profiles I 6

Figure

14 sand deposit” (Layer 4). It is entirely covered by the floor, except in profile III, where it extends beyond the site, and connects with Layer 1 beyond the floor. 18 All architectural features on the site cut into this layer and the bedrock underneath. The published drawing of the MCIII-pit is more or less identical to the original. The stratigraphic sequence consists of four layers: a layer of stones, followed by a soil layer containing a lot of pottery (presumably the majority of the sherds described in the publication). Below this is the layer of charcoal, which is noted by Sjöqvist. He claims that the pit was “filled with ash and potsherds.”19 The drawing of the pit (Fig. 6, Profile III), however, only indicates a thin layer marked as “charcoal”. At the very bottom is a “soil layer”. Similar to the MCIII-pit, the original and the published drawings of the terracotta basin are similar. The basin contained three separate stratigraphic layers. The top layer is possibly the same as Layer 2, “Undisturbed soil,” but none of the two drawings is clear if this is the case. Neither the drawing nor the publication has any description of the two lower layers in the basin. The middle layer, on top of which the RL Spindle bottle (No. 40) was found, looks to be an accumulation layer given the sloping sides. The basin is often described as being filled with ash and animal bones, something that must have been cited from Gjerstad. In Sekler och dagar, he mentions that the terracotta basin was ”filled to the brim with ash and burned animal bones, several votive gifts of gold and copper and ceramic vessels…”20 Sjöqvist, who was the excavator, does not mention this type of content either in the publication or his dissertation Problems of the Late Cypriot Bronze Age. This is curious since Sjöqvist makes a note of the charcoal in the MCIII- pit. On the original drawing, there is a faint indication that the thin, bottom layer has a separate legend from the layers above. It is similar to the legend used for the charcoal layer inside the MCIII-pit. If that is the case, it could indicate that there was a faint layer of ash at the very bottom of the basin. The trapezoid pit seems to contain only a single layer, marked as “rock-hard petrified sand” on the original drawing.

3.1.4 The architecture The two stone platforms on the eastern side of the sanctuary were constructed of rocks and soil.21 They were founded on the bedrock and measure c. 0.95 in height. The large platform is c. 2.4m in diameter, while the small measures c. 1.2m. Sjöqvist describes the floor as running up against and abutting the two platforms.22 These two platforms have been interpreted as altars, possibly for up to two deities.23 Cultic items were probably placed upon them for display.24 No objects or ash were found around the platforms, however. Since such remains are usually found at or around altars in Cyprus, the absence is striking.25 Sjöqvist explains the lack of objects “by the fact that the altars were only covered by a very thin layer of earth, which was entirely turned up by the plough.”26 The central wall, constructed from rubble, measures c. 0.4m wide, 4m long and has a preserved height of c. 0.5m. The wall runs in a north–northwest to south– southeast direction. As seen in Fig. 2, it does not cross the entire floor area. The small trapezoid pit has been cut partly into the bedrock. The MCIII-pit south of the platforms contained both ash and potsherds and was covered by rough stones. The

18 Sjöqvist describes this layer as a ”levelling stratum of sandy earth”. Gjerstad et al. 1934a, 356. 19 Gjerstad et al. 1934a, 356. 20 Gjerstad 1933, 99. 21 Gjerstad et al. 1934a, 356. 22 Gjerstad et al. 1934a, 356. 23 Sjöqvist 1940, 3. 24 Webb 1999, 171. 25 Webb 1999, 177. 26 Gjerstad et al. 1934a, 359.

15

Figure 7. The architecture of Dhima. Photo taken from the south. sherds found in this pit were studied by Sjöqvist and published in 1934.27 All sherds were dated to the Middle Cypriot period and concluded to predate the floor and the finds above it. For this reason, this pit is usually excluded from the discussions of the site and is only mentioned as evidence of activity during the MC period at Dhima.28 The terracotta basin is placed in a shallow, rock-cut pit and measures 1,28x0,63x0,47m. It is preserved to a height of 0,33m above the floor. The bottom consists of a single piece of conglomerate.29 There is no comment on whether the upper rim is completely preserved, but it looks quite damaged from Figs. 8 and 10.

3.1.5 The material culture While there were not many finds in total, compared to other Cypriot sanctuaries, several of them are striking, mainly for their valuable and exotic nature. All in all, 59 objects were inventoried and described in the publication.30 All finds were made on the western side of the sanctuary. The finds can roughly be divided into seven groups, according to their location: inside the terracotta basin, around the terracotta basin, in the trapezoid pit, Grid D3, Grid C3, the concentration of arm vessels in Grid D1 and stray finds. There are a few discrepancies between the published and original drawing regarding the finds, which will be noted in each case below. While the discrepancies might not have any direct impact on the present study, it seems appropriate to make a note of them. Several objects were recorded under the same inventory number, which is why a single number may appear two or more times. Any reference to the published

27 Gjerstad et al. 1934a, 360. 28 See for example Wright 1992 and Webb 1999. 29 Gjerstad et al. 1934a, 356–358. The conglomerate bottom is also visible in Fig. 6, Profile I. 30 For a full description of each item, please see Gjerstad et al. 1934a, 357–358.

16 plan refers to Plan XIII in the publication,31 while the original drawing refers to Fig. 9 or 11 The terracotta basin: The finds connected to the basin have been separated by whether they were found inside and around the basin. The main reason for this is to establish if there is a pattern to where certain items were placed, if possible. The interior of the basin contains the most varied amount of find of all groups listed. Two ceramic vessels are recorded within the basin, a “Levanto-Helladic” amphora (No. 31) and Red Lustrous Spindle bottle (No. 40). There was also a multi-coloured glass bottle (No. 32), shaped to imitate a Base Ring jug.

Figure 8. Excavation photo of the terracotta basin in Grid D2. Photo from the north.

There were several finds of metal inside the basin, including fragments of a gold leaf (No. 15), a silver funnel (No. 16), bronze tweezers (No. 17, 33), arrowhead (No. 35), a small lion figurine of bronze (No. 36), a bronze chain (No. 17) as well as some lumps of iron (No. 17). The basin also contained the only shell (No 34) and the only stone objects, a triangular axe (No. 13) and a marble cylinder (No. 34). A pierced plaque (No. 14) is the only find made out of bone and could perhaps be the shuttle from a loom. Lastly, a Babylonian cylinder seal (No. 12) of haematite was also found.32

31 Gjerstad et al. 1934a, plan XIII. 32 For a full description, see Gjerstad et al. 1934a, 357.

17

Figure 9. Terracotta basin, detail plan, Grids E2–E3 & D2–D3. Drawn from the original. Most large ceramic vessels were placed around the terracotta basin. This includes the White Slip crater (No. 18), a “Levanto-Helladic” amphora (No. 19), a Base Ring jug (No. 29), a “Levanto-Helladic” jug (No. 30), a “Levanto-Helladic” crater (No. 44), a Plain White wheelmade crater (No. 45) and a Plain White wheelmade jug (No. 46). Most of these vessels seem to have been too large to place inside the basin and were instead placed around the exterior. Apart from the pottery, three additional artefacts were recorded around the basin: an alabaster piriform jar (No. 53), fragments of a gold leaf (No. 20) and a haematite cylinder seal (No. 2833). The trapezoid pit: The only finds associated with this pit are metal, primarily of bronze. Four bronze pikes (Nos. 43, 48–50) were found inside the pit, according to the published drawing. No. 43 is, however, not present on the original drawing. Two bronze arrowheads (No. 51) were located inside the pit on the original drawing, while placed northeast of the terracotta basin on the published drawing. Lastly, two gold object, a gold-leaf diadem (No. 37) and fragments of gold leaf (No. 38) can be found on the west side of the pit on the original drawing. On the published drawing, however, these items are located between the terracotta basin and the arm vessel- concentration. Grid D3: A concentration of silver and gold objects was found in Grid D3, east of the terracotta basin. Except for items 3–4 and 27, no objects were marked on the original drawing. This group consists almost exclusively of gold jewellery and silver objects. The gold objects include three finger rings (Nos. 2, 8 and 11), 17 pearls, seven pomegranate and ten dates (No. 4 and No. 27), two pairs of earrings (Nos. 5 and 9), three pairs of hair-rings (Nos. 6–7 and 10). The silver objects consist primarily of funnels. A total of nine small silver funnels were recorded (Nos. 24–26) as well as a bracelet (No. 1) and a ring (No. 23). Lastly, two bronze arrowheads (No. 21–22) are included in this group, as well as a pierced haematite cylinder seal (No. 3),

33 For a full description of the motif, see Gjerstad et al. 1934a, 357.

18 which was used as a bead in a necklace together with the gold pearls.

Figure 10. The excavated terracotta basin in Grids D2–E2. Taken from the south. Grid C3: Two objects, not recorded on the original drawing, were found in Grid C3, a plain haematite cylinder (No. 41) and a bronze arrowhead (No. 42). The Arm vessel concentration: This group consists of several Red Lustrous arm vessels. Four (Nos. 54–57) were registered on the original drawing, and two more (Nos. 58–59) were only marked on the published drawing. The original drawing (Fig. 11) indicates a concentration of broken pottery in this area, consisting of both arm vessels and at least three piriform jars.

Stray finds: Two artefacts Figure 11. Red Lustrous Arm Vessel concentration were recorded without an apparent in Grid D1. Drawn from the original. relationship to any above features or concentrations. One is a gold earring (No. 47), which is recorded north of

19 the trapezoid pit. This artefact was not marked on the original drawing. It could perhaps be grouped with the gold/silver artefacts in D3. Additionally, a single glass bead (No. 39) was found south of the terracotta basin.

Figure 12. Excavation photo of the arm vessel concentration in Grid D1. Photo from the north.

3.2 Ayios Iakovos in time and space. During the end of the MC period, and the beginning of the LC period Cyprus underwent fundamental social changes. A pattern of settlements emerged along the coasts of Cyprus alongside an influx of foreign goods previously not found on the island.34 It was through these coastal centres that foreign goods started to penetrate the Cypriot material culture, alongside an increased export of Cypriot copper and crafts.35 This increased connectivity with the surrounding cultures began altering the ritual nature of Cyprus. Since the Early Bronze Age there had existed a communal tradition of consumption. During the Early and Middle Cypriot period, this consumption was directly related burials and deceased members of the community. Artefacts associated with the consumption of liquids and numerous offerings of

34 Voskos and Knapp 2008, 663. 35 Knapp 2008, 136.

20 animals are commonly found in tombs from these periods.36 In the EC period, the activities were conducted outside the burials, before final deposit inside the tombs.37 This behaviour changed somewhat in the MC period when the final deposits seem to have been made either in pits near the burial, as at Palaeskoutella, or on top of the burial, as at Toumba tou Skourou.38 The dead served a central part in the daily life throughout the entire Cypriot Bronze Age, with burials located close to the settlements.39 During the urbanisation of the LC period, this changed. In the emerging coastal centres, burials and their associated traditions were moved from an extra-urban, communal context into the private households. Tombs were located near, or even inside the house. In the hinterlands there was more continuity of the earlier extra-mural burial tradition. 40 However, both areas see the emergence of luxurious imports in the burial gifts repertoire. Tombs became the central stage for wealth and status display.41 From the end of the LCII period and onwards, this tradition of social display declined, especially in coastal centres. Tombs became less lavish and with fewer items. However, alongside the “privatisation” of burial, the communal tradition of consumption was gradually separated from the funerary context into a ritual tradition of its own.42 During this period of urbanisation, ritual activities see a shift from a funerary context to a specialised location, a sanctuary.43 Cypriot sanctuaries during the LC period are distinguished by the exclusive material culture, such as the osteological material, figurines, altars and other prestige goods. However, as Webb has noted, there is a great deal of variety between different ritual locales.44 Very few sanctuaries have been found from these early stages of the LC period. Dhima is still, after almost 100 years, considered the earliest example. There are, however, two other sanctuaries that are commonly grouped together with Dhima as the first ritual locales on the Island: Myrtou-Pighades and Athienou Bamboulari tis Koukounninas. Both of these sanctuaries have revealed a material culture oriented towards ritual activities. Myrtou-Pighades was an intramural complex of cult buildings located on northwestern Cyprus. Here, both BR cups and WS bowls are well attested in the material. There were also a wide range of Mycenaean vessels, both open and closed, as well as a rhyton.45 While there are many different shapes noted, Mycenaean vessels only make up a fraction of the assemblage (38 recorded sherds of the total assemblage) and it was most likely not introduced before the end of LCIIB/ LHIIIA2 period.46 BR cups and WS are more numerous and were most likely used for liquid consumption.47 There is also some presence of RL ware at Myrtou, but it does not seem to have been a major part of the assemblage. The recorded shapes are the

36 Keswani 2012, 316. 37 Keswani 2012, 316. 38 Gjerstad et al. 1934a, 419, 423, 429; Keswani 2012, 317 39 Keswani 2012, 317. 40 Keswani 2004, 140–141. 41 Keswani 2004, 142 42 This is especially evident given that evidence of feasting has been found both at sanctuaries during the LCII period, and settlement context towards the end of the LCII and beginning of the LCIIIA period, such at, Kalavasos-Ayios Dhimitrios and Kouklia-Evreti where feasting deposits have been found in wells, Steel 1998, 290–291. 43 Collard 2011, 31. 44 Webb 1999, 397. She has identified five different categories of sanctuaries, each with its own function. However, not all are contemporary. 45 Catling 1957, 42–48. 46 Catling 1957, 46. 47 Catling notes 66 BR cups and almost 500 WS sherds, Catling 1957, 38, 41.

21 Lentoid flask and Spindle bottle.48 Lastly, Myrtou shows evidence of a wide range of Plain Ware vessel, such as bowls, kraters and jugs. 49 In addition to the ritual rhyton, there was also the finding of a bull figurine, several miniature vessels and a large amount of zoo-osteological material.50 Athienou Bamboulari tis Koukounninas was a large, open court sanctuary, located on central Cyprus, east of the Troodos Mountains. The assemblage at Athienou show many similarities with Myrtou-Pighades; there is a wide range of LHIIIB1 Mycenaean jugs, juglets, bowls, cups and jars recorded, but the actual amount of vessels are few.51 Instead, there are numerous BR cups and some WS bowls.52 However, there was no RL ware found at Athienou. An ivory rhyton was also located.53 The most remarkable find at Athienou was the presence of thousands of miniature votive vessels, imitations of both local and imported wares.54 There were also several pits with ash and bones noted at the courtyard at Athienou, suggesting the occurrence of feasting.55 Both Myrtou and Athienou show evidence of an established ritual material culture with rhyta, miniature vessels and bull figurines. All are items without a clear practical function. Lastly, there has been much zoo-osteological material noted at both sanctuaries, suggesting communal feasting as a central part of the activities conducted. 56 The dating of the Mycenaean vessels at both Athienou and Myrtou (LHIIIA2-LHIIIB) indicated that Mycenaean imports had gained a strong presence in the Cypriot ceramic assemblage. It was not until the LCIIB period that such a wide range of Mycenaean vessels was integrated into the Cypriot repertoire.57

48 Catling 1957, 35. 49 Catling 1957, 48–56. 50 Webb 1999, 36–37, 250–251. 51 Dothan and Ben-Tor 1983, 46–49. 52 Dothan and Ben-Tor 1983, 33–37, 41. 53 Dothan and Ben-Tor 1983, 123. 54 Dothan and Ben-Tor 1983, 53–110. 55 Dothan and Ben-Tor 1983, 11–12 56 Dothan and Ben-Tor 1983, 11 and Webb 1999 Webb 1999, 250-251. 57 Steel 1998, 286.

22 4. Presentation of the previously unstudied material

4.1 The ceramic assemblage at Ayios Iakovos: Dhima Below will be a presentation and analysis of the complete ceramic assemblage of Ayios Iakovos: Dhima. A short overview and research history for each ware will be given, followed by a summary of the already published ceramic material of the site and the material from this study. This will be followed by an overview of the distribution of each ware on the site, with particular note being made of concentrations, depths and joining sherds (primarily from different grids). Lastly, each ware will be given a short discussion on possible interpretations and chronology of the catalogued sherds. This will all be concluded in a longer discussion focusing on general trends and notes on the complete assemblage, and how this affects the interpretation of the site. A short note on the terminology; any mention of sherds studied refers to the more general analysis incorporating all sherds from Dhima. Comments regarding catalogued sherds refer to selected material presented in the catalogue (Table 3). Any bold number, e.g. 153, corresponds to a post in the table as well as an illustration, if applicable (see Plates I–VI). While the illustration refers to a single sherd for each post, there are in some cases several sherds from the same vessel or of similar shape catalogued together. A short, horizontal line below the last sherd separates these groups. Joining sherds in different grids have been recorded separately to illustrate later movement of pottery.

Categories of Table 3, Catalogued sherds. ID – Catalogue number. Grid – The Grid in which the sherd was found. Depth – At what depth below the ground surface the sherd was found. H/W – Whether the sherd is handmade (H) or wheelmade (W). Frag. – Fragment, which part of the vessel. Ri – Rim. Sp – Spout. Ne – Neck. Sh – Shoulder. Ha – Handle. Bo – Body. Rb/Db/Fb/Pb/Udb – Ring-/Disc-/Flat-/Pointed-/Upraised base. Ø – Estimated diameter. Joins – A record of which catalogued sherds are joining. Munsell – Colour record: Fabric: Exterior (E), interior (I) and core (C). Slip: Exterior (ES) and interior (IS). Decoration: Exterior (ED) and interior (ID).

23 4.2.1 Middle Bronze Age

Black Slip Description: Black Slip (BS) ware, together with Red Slip, is one of the characteristic wares of the MC period.58 It is a handmade ware, with a black slip covering the surface. The ware earliest production is dated to the MCI period and is commonly used as an indication of the start of the MC period. 59 It continues to be used throughout the entire MC period, as well as in the LCI period, at which time a wheelmade variant is introduced. The handmade version is discontinued, while wheelmade BS continues to be produced through the LC period. 60 The most common shapes are the shallow bowls, jugs, juglets and tankards, with jugs and juglets becoming more popular during the LCI period.61 The ware was produced all over Cyprus but does have some minor regional differences, such as decorations or variations in fabric.62 General comments: The Swedish Cyprus Expedition found 96 sherds of BS ware within the MCIII-pit.63 Additionally, Åström marks Ayios Iakovos: Dhima as a site where “other fragmentary jugs or juglets” of BS IV ware had been noted.64 In this thesis, 168 sherds of BS ware were studied: 121 of closed vessels, 31 open and 16 handles (most probably from closed vessels). Eight sherds have been catalogued, 29– 36. Most of these, 29–34 consist of bowls. While only one sherd, 35, show direct evidence of a spout, it is possible that several of the bowls were, in fact, spouted. A single sherd, 36, also belongs to a jug. All catalogued sherds are handmade, except 31, which is wheelmade. Distribution: The majority of the sherds studied, 96, were located in Grid E5 46 sherds had no recorded depth, 42 at were located at a depth of 20–40cm and eight at 40–60cm. Two other concentrations can be noted. One in F5, consisting of 19 sherds located “on the first floor” and one in D0, consisting of 15 sherds located at a depth of 0–20cm. The remaining 38 sherds are spread across the site on both sides of the wall, at both 0–20cm and 20–40cm. The only exception is eight sherds in Grid D5, which were located at a depth of 60–80cm. Discussion: Bowls: 29 is a bowl with an upraised handle, similar in shape to both RS and RoB bowls 116, 213 and 247. The handle is similar to the one published by Åström 1972a, fig XXIV.4, but with a less acute inclination. 31 is the only wheelmade catalogued sherd of BS ware. It is a wide, shallow bowl and must date to LCI or later. 35 is a spouted bowl, very similar in shape to 259. Jug 36 is of a similar shape as 265, also without a visible handle. Like RS ware, most catalogued sherds are handmade and therefore dating to the MCIII–LCI period. Three sherds are recorded above the floor level (30 is on the first floor and 32–33 at 0–20cm). 34–36 are all found at 20–40cm and could, therefore, have been found both above and below the floor. Unlike RS ware, which had an equal ratio of open and closed shapes; the BS pottery is mainly closed. Roughly 80% of the BS material comes from closed vessels, most likely jugs. Sjöqvist concluded that the MCIII-pit contained 96 sherds of BS ware. In this study, a total of 96 sherds were studied in Grid E5. Unlike the case with the RS sherds, only about 46 sherds of BS are recorded without depth in Grid E5, with an additional 42

58 Crewe 2004, 78; Spigelman 2015, 469. 59 Crewe 2004, 78. 60 Crewe 2004, 78–79; Spigelman 2015, 470. 61 Crewe 2004, 79. 62 Crewe 2004, 79. 63 Gjerstad et al. 1934a, 360. 64 Åström 1972b, 78.

24 recorded at a depth of 20–40cm. It seems likely that most, if not all, of the BS sherds in E5 must have come from the pit.

Plain Ware Traditionally, most plain wares in Cypriot assemblages have been classified as “Plain White ware”. Unlike the suggestive name, Plain White (PW) pottery does not have to be white to be classified as such. It is rather a collective term for a broad range of ceramics with different coloured fabrics and slips. As such, the ware can be covered in a white slip, but it is not a pre-requisite.65 Due to the variety of colours of the fabric, it is common to just classify this ware as “Plain Ware”. The introduction of Plain Ware happened quite late in Cyprus. It was not until the end of the MCIII period that the ware was first introduced, then as a handmade, regional phenomenon on eastern Cyprus, called Plain White Handmade (PWHM). During the LCI period it started to appear island-wide.66 It diversifies itself from the other local wares of the time, such as RP/RS/BS/RoB due to the lack of a surface treatment. Åström and Jones place the first occurrences of Plain White Wheelmade (PWWM) during the MC period, but later studies have shown that it was first produced slightly later, during the LCI period.67 It is accepted that the handmade Plain Ware was introduced through the increased Levantine contacts during the MCIII period.68 It was most likely introduced as a part of consumption and feasting rituals, rather than for the ware itself.69 During the LCII, most of the large vessels associated with ritual consumption of liquids, such as kraters and jugs, were produced in Plain Ware. However, cups for individual consumption were gradually replaced with local wares, such as WS and BR bowls, and later on Mycenaean vessels.70 It also occurs in burial, possibly as lower class burial gift, given its low production cost.71 Being such a broad category of pottery, there are bound to be stray sherds, which are perhaps better placed in another ware, finding its way into this category. This is mainly an issue for the PWWM ware, and later counterparts, as it can be difficult to separate wall fragments chronologically. Additionally, as Crewe has pointed out, the distinction between hand- and wheelmade sherds is not always clear, and a margin for error must be cautioned.72

Plain White Handmade General comment: A PWHM oval jug was found and published from the initial excavations of the site. Additionally, 110 sherds from PWHM ware were studied from the contents of the MCIII-pit, but not commented on further.73 Åström also mentions the jug in his study from 1972.74 Beyond that, there has been no study on the PWHM ware from Dhima. In total, 244 sherds of PWHM-ware have been studied here, 193 closed, 43 open, five handles and three unknown. 37 sherds have been further catalogued (113–149). While a majority of the sherds seem to belong to closed vessel, most of the rims belong to open vessel (29/38 rims). It is possible that some of the

65 See Åström 1972b, 225–259 for a full overview; Crewe 2009; Jacobs 2009; Keswani 1991. 66 Crewe 2009, 79. 67 Åström 1972b, 232; Jones 1986, 542; Crewe 2007, 213; Crewe 2009, 79, 86. Crewe also suggested the occurrence of PWHM ware during the MCII period, but notes that this is uncertain. 68 See Catling 1957, 59; Crewe 2009, 80. 69 Crewe 2009, 80. 70 Crewe 2016, 128. 71 Keswani 1991, 99. 72 Crewe 2016, 117. 73 Gjerstad et al. 1934a, 358, 360. 74 Åström 1972b, 229.

25 wall sherds have been misidentified since the distinction between open and closed can be hard to differentiate at times. Of the 38 rims, most are unique shapes. All in all, 18 different shapes can be seen, showing a significant variation. The open vessels, 113– 134, consisting mainly of large bowls, while the closed vessels, 135–149, consist of jars and jugs. Distribution: The majority of the PWHM ware has been found on the eastern side of the site. Only 22 sherds were found on the west side, all of them at a depth of 0–20cm. In the eastern half, the sherds were mainly concentrated in Grids D5 (37 sherds), E5 (96 sherds) and F5 (32 sherds). All but ten sherds on the eastern side were found below a depth of 20cm and 84 were found below 40cm. 57 sherds are without recorded depth, but the box (246) contained mostly wares associated with the MCIII- Pit, suggesting them to be from below the floor level. There were a few joins noted across the grids: 116 and 117 (D5 and E5). 126–129 were spread across E2, E5, B5/E5 and F5, showing lots of movement. Discussion: Open shapes: The open vessels, 113–134, are mostly composed of large bowls, possibly some kraters, of varying shapes, several with decorations on the rim or the exterior wall (116, 119–120). With the exception of 120,75 no clear parallels have been noted, even for the more elaborate rims. Crewe lists large bowls to be a shape indicative of the LCI–LCIIA period.76 Jars: Like the open shapes, there is a wide range of shapes of PWHM jars. At least ten different ones can be discerned (135–148). These rims are less elaborate compared to the open shapes. 145–148 are all flat-based sherds, which is characteristic of pithos. This shape was developed during MCIII and produced through LCI77 142–144 are all common decoration for PWHM pithos. There are several parallels found around Cyprus, and the most common one seems to be the rope pattern of sherd 144.78 Hult classifies a sherd similar to 142 as a “late” ware, but seeing as the sherd was found at a depth of 60– 80cm, a late date is unlikely in this case. The fact that there are several parallels in the MCIII/LCI material supports this. 79 Crewe classifies this a strict MCIII type of decoration, which would make more sense in this context.80 Jugs: A single PWHM rim shows signs of a handle, 149. Unfortunately, too little remains of this sherd for a proper illustration and analysis. There are not many clear parallels of PWHM, most likely due to its regional and non-standardized nature, which resulted in a wide range of shapes, especially during the MCIII.81 It seems likely that all sherds found in the western part of the site can be explained due to post-depositional events. They are all found in the topsoil, indicating disturbance. The fact that PWHM is found at all depths illustrates that the ware was part of the assemblage at Dhima both before and after the construction of the floor. Given the long timespan of PWHM, this is not surprising. Unlike BS and RS, PWHM continued to be produced into the LCII period alongside the wheelmade variant, PWWM ware.82 However, considering the preference for wheelmade pottery in the Karpas region from LCI and onwards, most PWHM sherds from Dhima are probably from MCIII/LCI.83

75 Which is comparable to Catling 1957, 50, fig. 21.222. However, this was not decorated. 76 Crewe 2016, 118, table 5.2. 77 Crewe 2016, 123. 78 See Åström 1966, 193–206 for a corpus of relief decorations. Cf. Åström 1966, fig. 160–161, 209– 213; Hult 1992 fig. 14.7, 8, 9, 25, fig. 19.1, 6, 8 and Crewe 2009, fig. 3.2–3 for parallels. 79 Hult 1992, 71 and fig. 28.27. 80 Crewe 2016, 121, fig. 5.2.A. 81 Crewe 2007, 216. 82 Crewe 2004, 43; 2007, 216. 83 Crewe 2007, 225.

26 Red-on-Black Description: During the Middle Cypriot period, the so-called Red-on-Black (RoB) ware is one of the more common ware types. It consists of handmade pottery covered in a black slip with patterns red lines drawn across the body. The ware comes in various shapes, most commonly as jugs and bowls. Bowls are often decorated with groups of long or short vertical lines from the rim, while horizontal and diagonal lines occur more frequently on closed vessels. The lines are usually made with either a brush or a comb-like instrument. 84 A study of this ware has determined its chronological range between the MCII period and LCI period. Popham marks LCIA as the latest period.85 However, more recent studies have concluded that the RoB pottery was probably produced during the full LCI period. Spiegelman concludes that LCIIA marks the end of RoB production.86 Horowitz, however, argues that it might have survived in some areas until the late LCII period.87 The concentration of RoB pottery in northeastern Cyprus, in particular on the Karpas peninsula has lead to the conclusion that it was produced in this area. The geographical range is roughly defined as Phlamoudhi in the west and Enkomi in the south.88 Smith has suggested Phlamoudhi-Melissa as a likely production site of this ware. 89 The Light-on-Dark sherds noted by Sjöqvist in the original publication have been included in this ware, due to the similar production techniques and decoration patterns. General comments of the assemblage at Dhima. In the original publication, 50 sherds were found in the MCIII-pit.90 During this study, 112 sherds of RoB ware have been studied, 85 open, 26 closed and one handle. From these, 20 have been further catalogued, 215–234. The open vessels are mostly from bowls. One catalogued sherd, 230, is from a spouted bowl. Like the BS ware, almost all sherds are of similar shape, and it is possible that more than one bowl was spouted. Four sherds, 231–234, are from jugs. Due to its close relation to both RS, BS and Red-on-Red pottery, there are undoubtedly mixes between these wares, especially since the slip and decoration can change from red to black due to the firing process.91 There are also examples of what Hult calls “ghost lines”, where the decoration and slip underneath have been lost, such as 220 or 224.92 Distribution: The majority of the RoB sherds, 101 sherds, are recorded east of the wall. 92 sherds were found in Grid E5 and F5. 53 sherds are found in Grid E5. 35 of these are found at a depth of 20–40cm and 18 are without depth. Most likely, these are all from the MCIII pit. In F5, 29 sherds were found on a level that is described as “on the first floor”. In regards to the rest of the sherds, nine are recorded at a depth of 60–80 in Grid D5, which would put them below the floor level. The remaining 17 sherds are spread across the site at both 0–20cm and 20–40cm. Almost all of the closed sherds were found in E5, while the open sherds were somewhat more spread out between E5 and F5. There does not seem to have been much movement of the RoB, with no discovered joins between the grids. Discussion: Bowls: 215–216 are the only bowls with a carinated rim. 226 is very similar to sherds in Hult’s material,93 while 227 is most likely a hemispherical

84 Åström 1972a 117. 85 Popham 1972b, 700. 86 Spigelman 2015, 519. 87 Broadly in the Karpas and northeastern Mesaoria region, and specifically in Melissa. Horowitz 2008, 77; Merillees 2008, 118. 88 Merillees 2008, 116; Spigelman 2015, 305–309. 89 Smith 2008, 62. 90 Gjerstad et al. 1934a, 360. 91 Spigelman 2015, 410. 92 Hult 1992, 64. 93 Hult 1992, fig. 23.1.

27 bowl with marked rim.94 Both have several vertical, parallel red lines from the rim and downwards on both interior and exterior. Most sherds (217–229) are of bowls of similar shape or a slight variation of the same shape. At least one (230) is a spouted bowl, but it is possible that other sherds are part of such bowls as well, especially considering the amount of spouted bowls/spouts in the RS and BS material. In general, the decoration pattern is the same for the bowls: groups of long or short vertical parallel lines, usually around ten lines, spread across the rim on both interior and exterior side. The colour is mostly red, but sometimes of a more pink or whitish colour. There is some variation of the technique used to draw the lines. 221 has the usual vertical lines on the interior. The lines are, however, drawn in several short strokes, making them uneven between both ends and the middle of each stroke. The short, horizontal lines on the exterior are whiter in colour, possibly due to a discoloration of the regular red lines. 223 almost looks to be more BoR than RoB on the interior, due to the very thin parts of black slip showing, and the fact that the red lines seem to expand into a slip on the interior side. It is possible that this is a sherd with inverted colours. The exterior has the same pattern, but with more clear RoB decoration. 226 is the only sherd with a preserved attachment for a horizontal ring handle, most likely of wishbone shape. 229 has a suspension hole just below the rim. 230 is a spouted bowl, with only part of the spout preserved. The interior shows groups of vertical, parallel lines, but is very fragmented. The exterior side has the standard groups of long, vertical parallel lines. In addition, there are several rows, not always parallel, of diagonal, parallel lines in-between the vertical groups.95 Jugs: Jugs are also noted in the material (231–234), in three different shapes. 231 is a complete rim and neck down to the shoulder, but show no traces of a handle. It is somewhat broken along the rim, and it is possible that an attachment has been lost. It also has a noticeably shorter neck than the vessels presented by Åström, perhaps suggesting a smaller vessel. The decoration consists of twelve sets short horizontal, parallel red lines along the neck and twelve vertical parallel lines starting at the base of the neck.96 232 is also missing an attachment of a handle at the rim, but the shape would suggest a large globular jug.97 The decoration on this sherd has almost completely disappeared. 233–234 are of similar shape with a wide band handle from the rim to, presumably, the shoulder.98 Very little is preserved of the neck of 234, but the handle has sets of horizontal parallel red lines. 233 has, similar to 231, several sets of short horizontal, parallel red lines on the neck, and similar sets of diagonal, parallel lines along the rim. There has been no clear chronological distinction between the shapes and different decoration pattern made so far, but there is no evidence for any of the decoration pattern with less than 7–10 lines, nor does any sherd have a wavy line pattern as found in Åström.99

Red Polished Description: Red Polished (RP) ware is one of the more characteristic Cypriot wares during the EC and MC period. It is a handmade ware with a glossy red polished surface treatment, hence the name.100 The earliest occurrence of this ware is during

94 Cf. Åström 1972b, fig. XXXI.8 and Hult 1992, fig. 22.51. 95 Very similar to the pattern on Åström 1972b, fig. XXXIII.10. 96 This pattern is similar to Åström 1972b, XXXIII.10, but with 2/3 shorter neck, and no handle. 97 Cf. Åström 1972b, fig. XXIV.8. 98 Cf. Åström 1972a fig. XXXIII.6–7. 99 Cf. Åström 1972a, figs. XXXI–XXXVI. 100 Davit et al. 2014, 3.

28 the EC period, and it continues to be used throughout the entire MC period and the very early stages of the LC period.101 General comments: From the initial publication, three sherds of RP ware were found in the MCIII-pit. 102 During this study, only a single RP was studied and catalogued, 235. Distribution: 235 was found in Grid D5, at a depth of 60–80cm, suggesting that it is not part of the MCIII-pit assemblage. Discussion: 235 is a flat base from an open vessel, most likely a bowl. While RP ware is produced to some extent in the early LC period, it is not very common. It is by far more popular in the MC period. The recorded depth of this sherd would support MC date for this sherd, since 60–80cm would be well below the floor level.

Red Slip Description: Red slip (RS) ware, like Red slip, was initially produced during the MCI period. Like RS the ware is named after its surface treatment, a red/dark brown slip. Due to the similarities in fabric and production, these two wares are usually grouped together.103 For a full description and chronology, see Black Slip above. General comments: In the original publication 27 sherds were identified as RS ware from the contents of the MCIII-pit.104 In this study, 174 sherds of RS ware have been studied, 86 open, 78 closed and ten unknown. From these, 31 have been further catalogued, 236–266. The open vessels are either from bowls or spouted bowls. There are probably several sherds from spouted bowls, but with the spout missing. There is also a wide range of shapes within the bowls, with eight different types catalogued (both with and without spouts), and one type of jug. The ratio between open and closed vessels is more or less 50/50. Distribution: The biggest concentration of RS ware can be found in Grid E5, containing 71 sherds, including 239–240 and 265. 33 of these sherds were recorded at a depth of 20–40cm, 8 at 40–60cm and 30 without recorded depth. The second concentration of RS sherds can be found in Grid D0, outside the floor area, and it consists of 26 sherds, including 238, 255–257 and 263, at a recorded depth of 0– 20cm. The remaining sherds are spread across most grids, with slightly higher amounts in A5, D5 and F5 (each containing more than 10 sherds). There also seems to have been some movement of RS sherds based on joining sherds between grids, 244 and 245 join and are located in E3 and A5. Similarly 259 and 260 also join, being located in Grids B5/E5 and E3. Discussion: Cups and bowls: 236–258 are all various open shapes, mostly bowls. 237 is a wide, shallow bowl with a sharp, inward rim. 240 most likely belong to a wide bowl, a reoccurring shape in both BS and RoB ware.105 251 and 259 (similar shape, but with spout) are both very similar to 240, but with a less acute inclination. 243 is from a bowl of similar size, but with the rim turned inwards. Overall these shapes correspond with established shapes of RS/BS ware from the MC period. 236 and 244 are possibly small cups, due to the smaller diameter. Although wheelmade, a close parallel seem to be the wishbone handled cups catalogued by Åström.106 246 stands out, and seem to be a very shallow plate or a lid. Unfortunately, very little of the rim is preserved. It is possible that this sherd is of a later date. Jugs: Sherd 265

101 Crewe 2004, 88–89; Spigelman 2015, 38. 102 Gjerstad et al. 1934a, 360. 103 See, for example, Crewe 2004, 78–80. 104 Gjerstad et al. 1934a, 360. 105 Cf. Åström 1972a, fig. XXIV.21. 106 See especially Åström 1972a, fig. XXI.3 or possibly fig. XXIV.15.

29 originates from a globular jar. There is no sign of a handle, but a similar shape found at Myrthou-Pighades show a handle further down the neck.107 The majority of the catalogued sherds are handmade (237–243, 247–254, 256–265), and should therefore be dated to the MCIII–LCI period. Several of these sherds were located above the floor level (247, 249–250 and 258 on the first floor and 238–239, 254, 256–257 at 0–20cm). Sherds 237, 240, 248, 251–253, 259–261 and 264 are all recorded at a depth of 20–40cm. While this is probably above the floor level in many cases, it is somewhat more ambiguous since the floor is roughly at a depth of 40cm throughout the sanctuary. The few, catalogued wheelmade sherds (236, 243–346, 255 and 266) must all date to the LCI period, possibly slightly later. Almost all are, however found in the uppermost layers at Dhima. The 30 sherds without a recorded depth are most likely the from the MCIII-pit assemblage, considering that the pit is roughly 40cm below the surface. This means that the sherds recorded at 20–40cm were located above the floor level.

White Painted Handmade Description: White Painted Handmade ware (WPHM) is commonly found in MC contexts, but it does occasionally overlap with the beginning of the LC period. It is not as common as RP/RS/BS/RoB ware108 General comments: Only two sherds of this ware have been identified, both from closed vessels. None were further catalogued. It is possible that more WPHM sherds exist in the Dhima assemblage. Wear and tear on the painted decorations would make any sherd indistinguishable from ordinary plain sherds. Distribution: The two sherds were located in D4 and D5 respectively, at a depth of 20–40cm and on the first floor. Discussion: Not much can be said about the presence of this ware other than that the sherd found on the first floor must date to either late MCIII or LCI, since the ware was no longer produced in the LCII period.109

4.2.2 Late Bronze Age

Base Ring Description: Base Ring (BR) ware is, alongside White Slip ware one of the most characteristic types of pottery of the Late Cypriot society. The ware is handmade and commonly identified by its very thin, almost metallic fabric of a grey or dark brown colour. It is also common to have a dark, burnished slip and is consistent in its use of ring-bases.110 A wide range of shapes was produced, such as bowls, jugs, jars and flasks, among others. Two of the most characteristic shapes are the juglet and the carinated bowl with an upraised wishbone handle.111 The Carinated bowls often occur in ritual contexts and are believed to have served a central role in the offering of liquids.112 BR ware was also used to make animal-shaped vessels, such as horses and bulls.

107 Catling 1957, 27, fig. 14.19; Åström 1972a, fig. XXVII.4. 108 Campbell Gagné 2012, 33. 109 Campbell Gagné 2012, 33. 110 Åström 1972b, 137. 111 For the full range of shapes and variations, see Åström 1972b, 137–198. 112 Steel 2004a, 177; Webb 1999, 189. Webb also notes that the Y-shaped bowl is one of the most common types to find also in tombs and settlements.

30 While the classification of Proto BR (PBR), BR I and BR II is still commonly used today, Vaughan made an attempt to rectify this classification.113 She deemed it out-dated, given modern methods of analysis and problematic, since it is largely based on tomb assemblages. She also criticises the fact that there are several examples of vessels that fit two of the types.114 Instead, she proposes a different classification system, with the main determinative feature being the different kinds of slip: Metallic Slip, Red Burnished, Matte Slip and Uncoated. Vaughan does admit, however, that a complete reclassification of previously published vessels would be impossible, and notes that her new classification format would rather serve future projects, than older.115 Even after this suggestion, the older types of PBR, BR I and BR II have remained. This established system will be used in this study as well, since its purpose is not to reclassify the established typology system. General comments: A single BR jug was found and published in the excavation report. It was placed just outside the terracotta basin, on the northwestern side. The jug was classified as BR II.116 In addition to the jug, Åström notes the presence of BR sherds from Grid D6 as Y-shaped bowls,117 most likely 10–11, 14 and 21–25. In total, 89 sherds of BR ware have been examined in this study, 45 closed, 30 open, five handles (four from open vessels and one from a closed) and nine unknown. From these, 26 sherds have been catalogued, 2–28. Most, 2–26, are parts from carinated Y-shaped bowls. Five different variations of such bowls have been identified. Additionally, two sherds, 26–27, are ring bases from jugs. Distribution: The studied sherds are concentrated in three grids. 34 sherds were found in Grid D3 (including 3–4, 8, 12–13, 16–17 and 19), at a depth of 20– 40cm. 27 were found in Grid E3 (including 5–7 and 20), at a depth of 0–20cm. The third concentration is in Grid D6, where 20 sherds (including 10–11, 14, 21–25) were registered at a depth of 20–40cm. Eight sherds were found in Grid D0 (including 2, 9, 15 and 26–28), at a depth of 0–20cm. Lastly, a single sherd, 18, was catalogued from A5. There are a few joins spread across the site: 2–4 (D0 and D3) and 20–21 (E3– D6). With the exception of the Grid D6 concentration (and 18), all sherds are located on the western side of the wall. Discussion: 2–24 all consist of rims from carinated bowls of five different shapes, all fairly similar with some minor variation. Sherds 7–10 are similar to sherds in Hult’s 2002, fig. 1.9, 14–24 shares similarities with Åström 1972b, fig. LII.4. 23– 24 joins with 22 and are part of an upraised handle, belonging to a carinated bowl. 25–26 are Rbs of a carinated bowl. 26 is similar to Åström 1972b, fig. LII.4 and Hult 2002, fig. 1.4. 27–28 belong to closed vessels, probably jugs. 27 is similar to Åström 1972b, fig. LIII.6, while 28 is probably Åström 1972b, fig. LII.10. The distinction between BR I and II is not always clear, due to the small size of the sherds in the Dhima assemblage. However, the carinated bowls seem to mostly belong to BR II, given their flat angle below the rim and the absence of any relief decorations. The bases of the closed vessels also have parallels in the BRII material. On the other hand, there has been no trace of the white paint, which is usually indicative of BR II ware. However, there is no direct evidence for any BR I vessels. A classification of BR II would date the material to LCIIA and somewhat later.

113 See Vaughan 1991. 114 Vaughan 1991, 119. 115 Vaughan 1991, 125–128. 116 Gjerstad et al. 1934a, 357. 117 Åström 1972b, 177.

31 Cypro-Mycenaean Description: The term “Cypro-Mycenaean” usually only refers to certain shapes within the locally produced pottery, more specifically FS47, identified as an explicit eastern Mediterranean shape.118 In this study, all pottery that resembles Mycenaean will be classified as Cypro-Mycenaean (Cypr. Myc.) for simplicity. While most sherds are clearly Mycenaean in origin, and some clearly Cypriot, there is a range of sherds in between, which could befall to either category, due to ware and tear from formation processes. This general classification also allows for a more general overview of the material since the sherds can be treated as a united group. Since all sherds in this group seem to originate from one of two shapes, it is more useful to discuss the shape, and possible contents, rather than its provenience. Mycenaean imports to Cyprus are commonly divided into three stages. During the LCI period, only small amounts of Mycenaean pottery was imported. The amounts gradually increased towards the end of the LCI and beginning of LCII period, before significantly increasing during the end of the LCII period. Mycenaean pottery is most often found in non-settlement contexts, such as burial and ritual settings.119 Especially popular were closed vessels, either wide mouth piriform jars or “slow pouring-vessels” such as stirrup jars. It is believed that such “slow pouring- vessels” were imported for their contents. But vessels associated with drinking, such as cups and kraters were also imported in large amounts, especially during the late LCII period.120 Social consumption of liquids was already an established custom in Cyprus during the LCII period. It is therefore believed that the Mycenaean drinking vessels were incorporated into the practice, rather than imported alongside it.121 General comments: Sjöqvist recorded four Mycenaean vessels: a krater (FS7), two piriform jars (FS44 and FS47) and a juglet (FS113). All vessels were found around or inside the terracotta basin.122 All were dated as LHIIIA2 vessels.123 Åström also comments on the presence of the stirrup jars, as well as three piriform jars.124 In this study, 327 sherds have been classified as Cypr. Myc. 304 from closed vessels and 21 handles, all of them most likely from closed vessels as well, and two unknown sherds. From these, 80 sherds have been further catalogued, 38–109. It is clear that all sherds from this ware come from similar shapes, predominantly piriform jars. Almost all sherds show some form of line and band decorations along the body, with individual variations in shoulder decoration. Distribution: There are three primary concentrations of Cypr. Myc. pottery at Dhima. The vast majority was found in Grid E3, consisting of 152 sherds (including 46, 50, 55, 56–60, 68, 71–72, 78, 84, 87, 93, 94, 96 and 101–104). All of these are at a recorded depth of 0–20cm, meaning that there has probably been some disturbance. The second largest concentration is in Grid D3, consisting of 93 sherds (including 38, 44, 48–49, 52–53, 64–67, 69–70, 74–76, 80–81, 83, 85–86, 88, 90–92, 98 and 108– 109). All at a recorded depth of 20–40cm. Considering their proximity to the terracotta basin, it is likely that these vessels were originally placed somewhere in its vicinity. The last concentration is in Grid D0, consisting of 44 sherds (including 39– 42, 45, 47, 51, 54, 61–62, 73, 79, 95, 97 and 106–107). The remaining 38 sherds are spread across Grids A5, B5, D4, D6, E2, E4 and E5. There seems to have been considerable movement of sherds, since several joins can be found across the site: 38– 39 (D3 and D0), 40 and 84 (D0 and E3), 44–45, (D3 and D0), 46–48 (E3, D0 and

118 Åström 1972b, 290. 119 Steel 1998, 286. 120 Steel 1998, 286. 121 Steel 1998, 292. 122 Gjerstad et al. 1934a, 357–358. 123 Webb 1999, 31. 124 Åström 1972b, 305, 348.

32 D3), 53–54 (D3, D0), 67–68 and 99 (D3, E3 and D6), 79–81 (D0 and D3), 82–83 (E4 and D3) and 96–98 (E3, D0 and D3). With the single exception of one sherd in Grid E5 (at a depth of 40–60cm), all Cypr. Myc. sherds are recorded at a depth of 20–40cm or above. Discussion: 125 Jars: sherd 38 is most likely a FS44, given the sloping, decorated rim.126 The rounded, sloping rim of 46 indicates a FS31.127 However, this shape does not have horizontal handles. It is, therefore, more likely to be a FS44 as well, dating to the LHIIIA1.128 The horizontal rim of sherd 41 closely compares to FS 39, indicating a LHIIIA2 date.129 53 is a FS45 or possibly an alabastron, given the triangular rim. However, there has been no evidence of any alabastra in the material. Either way, it dates to the LHIIIA2 period.130 Sherd 59 belongs to FS47, which is believed to have been produced in Cyprus. The rather straight rim separates it from other piriform shapes during the LHIII period.131 82+83 have a decoration of FM 57, net pattern. While occurring in both LHIIIA1 and LHIIIA2, it is less common on LHIIIA2 vessels 132 Sherd 88 also has a net pattern decoration. But instead of paint, each line consists of small dotted impressions. 86–87 both have a FM53 decoration, a horizontal wavy line. This motif has been particularly common on piriform jars in the LHIIIA2 period.133 The scale pattern decoration found on 78, 84–85 most closely match the pattern found during the LHIIIA1 period, without a singular dot within each scale usually found in the LHIIIA2 period.134 Both 90 and 99 seem to belong to FS31, a LHIIIA1 small piriform jar.135 All other bases, 89, 91, 94 and 106–107 can belong to both LHIIIA1 and LHIIIA2 shapes. 108 and 109 are from stirrup jars, most likely FS166, a tall conical-piriform shape. 136 A vessel with almost the same decoration pattern was found at Ayios Andronikos. This vessel is classified as LHIIIB2. However, the shape bears a closer resemblance to FS166 with a less globular body, which dated to LHIIIA2. Since the unvoluted flower motif is also used during LHIIIA2, it seems like a more fitting match.137 All catalogued sherds are from either piriform jars or stirrup jars. While it can be difficult to differentiate between the lower body of stirrup and piriform jars, as well as between the upper body of alabastra, pyxides and piriform jars,138 the total assemblage found at Dhima indicates an almost exclusive presence of piriform jars. There has been no evidence of bases from alabastra or pyxides in the studied material. It would seem reasonable to assume that even if not all sherds can be securely interpreted as piriform jars, it is likely that most are. It is believed that the piriform jars contained a “thick, viscous oil or unguent”, while the stirrup jars probably contained a more pourable substance. 139 Such contents have previously been

125 The dating for Mycenaean vessels is based on the Aegean system, rather than the Cypriot one. LHIIIA1 roughly corresponds with LCIIA and LHIIIA2 with LCIIB. 126 Cf Mountjoy 1986, 56, fig. 62. 127 Cf Mountjoy 1986, 55, fig. 61. 128 Compare Mountjoy 1986, 55, fig. 61 & 56, fig. 62. 129 Cf. Mountjoy 1986, 71, fig. 80. 130 Cf Mountjoy 1986, 72, fig. 81 & 73, fig. 83. 131 See Mountjoy 1986, 91–133. Compare with Hankey 1967, fig. 9.6248; Åström 1972b, 290, 302 132 While occurring in both LHIIIA1 and LHIIIA2, it is less common during on LHIIIA2 vessels. Mountjoy 1986, 68. 133 Mountjoy 1986, 70. 134 Compare FM 70 during LHIIIA1 and LHIIIA2, Mountjoy 1986, 52 and 58. The scale pattern is also more uncommon in LHIIIA2 shapes. 135 Compare to Mountjoy 1986, 55, fig 61.2 and 61.10. 136 Cf. Mountjoy 1986, 78, fig. 91.1. & Nicolaou and Nicoulaou 1989, Plate XIV.331. 137 Compare to Mountjoy 1986, 78, fig. 91. 138 Leonard 1981, 93–94. 139 Leonard 1981, 94.

33 connected with libation rites.140 But whether this content was used for libation, or if it was manually extracted for smearing is difficult to say. The piriform jar also occurs as a burial gift during the LC period, but unlike the pictorial krater and the RL arm vessel, it is much more widespread. It does occur in elite tombs as well, but then among a wider range of Mycenaean shapes.141 The piriform shape is more common than Mycenaean cups and krater, possibly due to its importance in funerary rites.142 The presence of both LHIIIA1 (38, 46, 78, 84–85, 90, 99) and LHIIIA2 (41, 53, 59, 86–87, 108–109) pottery suggests that Mycenaean imports were brought to Dhima during a prolonged time, perhaps through the entire LCIIA and early LCIIB period. That is, assuming the LHIIIA2 vessels were only introduced at the very end of the activities at Dhima (which is likely considering the preserved LHIIIA2 vessels around the basin). Additionally, if the FS7 Mycenaean krater should be dated to LHIIIA1 instead of LHIIIA2, as Mountjoy suggests, it indicates that this vessel stayed in use for a long time at Dhima.143 The joining sherds between separate grids are also noteworthy. While this to some extent is due to formation processes (in particular between Grids D3 and E3), it might also indicate that material was actively moved from the central area and placed outside the floor area. On a final note, Mycenaean pottery usually only makes up a small fraction of the total ceramic assemblages in Cypriot contexts (1–2% before including plain wares).144 At Dhima, the Cypr. Myc. sherds make out 17% of the LCII assemblage, even after the inclusion of plain wares, indicating its dominant presence in the material.

Grey Ware Description: Grey ware has a problematic history in Cyprus. It has long been argued both how to define this ware, as well as what to call it.145 It is not an indigenous Cypriot ware, but an import, most likely from Anatolia, where the finds are numerous. For this reason, it has been argued that such grey wares found in Cyprus should be named “Anatolian Grey Ware”.146 The ware comes in a wide range of shapes, both open and closed, some even imitating Mycenaean shapes. The most common shapes are, however, bowls and krater.147 It is mainly found in Cyprus in the very late LCIIC and LCIIIA period as part of trading conducted with the Aegean.148 General comments: Three sherds with a distinctly grey fabric and slip have been classified here simply as “Grey Ware”. All sherds are from open vessels and have been catalogued, 110–112. Distribution. The sherds are spread across the site, in B5 (110, 20–40cm), D3 (111, 20–40cm) and D5 (112, 60–80cm). There are no joins. The depth suggests an early placement on the site, especially of 112, being found below the floor level. Discussion: All three sherds (110–112) come from small bowls. There is little more that can be said about these sherds, especially since so little of the rims remain. Allen argues that the Anatolian Grey Ware cannot be found in Cyprus before the 13th century.149 This is problematic in regards to Dhima since the site was abandoned at that time. However, there is no history of any ware with a similar grey fabric

140 Steel 2004a, 75. 141 Steel 2004a, 77–78. 142 Wijngaarden 2002, 153. 143 Åström 1972b, 290 notes this as a FS7, which originally was considered a LHIIIA2 krater. However, Mountjoy (1986, 61) suggest that FS7 should instead be dated to LHIIIA1. 144 Steel 1998, 286. 145 Allen 1991, 151. 146 Allen 1991, 154. 147 Allen 1991, 152. 148 Allen 1991, 153. 149 Allen 1991, 151.

34 produced in Cyprus during the Bronze Age. So if this ware is not Anatolian Grey Ware, it is unclear what its provenience is. Perhaps a local, small-scale production?

Plain White Wheelmade Description: For a more detailed history of the Plain White Wheelmade (PWWM) ware, see Plain Wares, above. General Comments: During the initial excavations, a complete PWWM krater was found next to the terracotta basin. This was also included in Åström’s study.150 In regards to the sherd assemblage at Dhima, PWWM constitutes the single largest ware- type of the site, with 872 studied sherds, 796 closed, 39 open, 24 unknown and 13 handles. However, it is also likely that sherds from different periods have made their way into this group, especially considering the presence of Roman amphora sherds. From the 872 studied sherds, 45 have been further catalogued, 150–194. The majority of the material seems to be of closed vessels, but most of the rims belong to open vessels, 36 open rims to 18 closed. The open vessels consist of bowls of various sizes and kraters. Closed vessels are either from jars or jugs. The numerous amounts of closed wall fragments might be explained as coming from large storage vessels or simply the ambiguity of the sherds themselves. Distribution: Very few sherds are found at the lower depths of the site, none are securely placed below the floor level. 55 sherds are documented in the ambiguous depths of B5 (40–100cm). The vast majority, 813 sherds, were documented above the floor level, 431 were documented at a depth of 20–40cm and 382 sherds were founds at 0–20cm. At 20–40cm, much like with PWHM, there seems to be a concentration on the east side of the wall (especially in Grids A6, B5, D4, D6, E5), with most sherds in D6 (135 sherds), which is beyond the floor. There is, however, also a large concentration in D3, of 112 sherds, suggesting at least some presence of PWWM in the western part. At 0–20cm, there is a greater spread of the ware, much like with PWHM. There are several concentrations, in Grids A5, B1, D0, E3 and E4, suggesting a substantial movement of the pottery, possibly due to farming. In the northern parts, in the A and B grids, there is the possibility that a majority of the sherds are of a later date, considering the Classical/Byzantine stone construction found there. The concentrations at Grid D0 and Grids E3–4 probably belong to the late Bronze Age material, considering their proximity to lower level concentrations. It is possible that at least the E3–4 sherds are a displacement of the D3 concentration. Another clear indication of the spreading of the pottery is the fact that several sherds from different grids have found joins: 152–156 (D0, D3, E2, E3 and E4). The sherd at the greatest depth is 153, at 20–40cm, possibly suggesting an original position. 159–160 (B5 and E3), sherds 178–179 (E4 and D0), 162–163 (D3 and E2), 171–173 (D0, D5 and F5 at vastly different depths, from 60–80cm, 1stFL to 0–20cm), 183–184 (A5 and D0) and 86–188 (A6, E2 and E4). Discussion: Open shapes: 150–160 are all small and shallow bowls, with and without carination. 161–162, 164–165, 166–168, 171 and 174–175 are all from large bowls or kraters, with various types of rims. Several of these shapes can be found in the Myrtou-Pighades material, 161,151165,152 167,153 168,154 271,155 and 174.156 164

150 Gjerstad et al. 1934a, 358; Åström 1972b, 243. 151 Catling 1957, fig. 22.294. 152 Catling 1957, fig. 22.245. 153 Catling 1957, fig. 22.293. 154 Catling 1957, fig. 21.260. 155 Catling 1957, fig. 22.283. 156 Possibly Catling 1957, fig. 22.297.

35 has a close parallel at Enkomi.157 No rims or walls of the catalogued PWWM sherds have any trace of decorations. 170 is possibly from a later date, given the unusual shape of the rim. Unfortunately, very little is preserved. Jars: 180–183 and 186 are storage jars of varying sizes. 186 is possibly pithos. The pointed base of 193 is most likely from a Canaanite jar.158 This type of ware is usually classified as a separate ware, but since this is the only sherd from such a vessel, it is catalogued here. PW kraters and small bowl shapes are very common in the LCI period, representing the main equipment for consumption. During the initial phases of the late Cypriot periods there was a high variation of PWWM shapes in Cyprus, with regionalism being considered the unifying factor between different areas.159 At this time the number of shapes was numerous.160 The material at Dhima reflects this with at least 25 different catalogued shapes. It was not until the late LCII period that a certain level of standardisation was established. 161 With the decrease of PWWM shapes, the small bowls for personal consumption were gradually replaced with BR, WS and Mycenaean wares.162

Red Lustrous wheelmade ware Description: The Red Lustrous wheelmade (RL) ware is one of the more distinct wares found in Cyprus. It comes in a very limited number of shapes, mostly closed and with a homogeneous fabric.163 The most popular shapes are the Lentoid flask and the Spindle bottle. One of the unique shapes is the so-called ”arm vessel”, named due to its resemblance to a forearm holding a small cup in its hand. While the ware is usually represented in most types of context, it is most commonly found in burial and, to a certain extent, other ritual contexts.164 There have been numerous vessels and sherds found around the Mediterranean, and several studies, including those by the Swedish Cyprus Expedition, initially interpreted RL as a Syrian import, and thus treated as evidence of foreign contacts. 165 One of the boxes from this study, containing mainly RL sherds, was labelled as “Syrian Ware”. However, later studies have concluded that RL ware was most likely produced in Cyprus, although Anatolian production of the ware has not been completely ruled out.166 Kozal especially argues for a possible Anatolian origin, based on recent finds of RL ware in Anatolia. The ware was produced between LCIA:2 – LCIIIA:1, spanning roughly 300 years. General comments: In the original publication, the only recorded vessels of RL ware was a Spindle bottle found inside the terracotta basin, and the unique finding of five arm-shaped vessels found together in Grid D1.167 Åström commented on some of the more distinct RL sherds. He notes “Fragments of pilgrims flasks, mainly of type VIIa”, 168 the presence of the six published arm vessels and an additional “five

157 Crewe 2004, 353, fig. A1.12.12. 158 See Catling 1957, fig. 23.32. 159 See Crewe 2009; Jacobs 2009. 160 Crewe 2009, 79. 161 Crewe 2016, 129; Jacobs 2009, 91. 162 Crewe 2007, 229–230. 163 See Eriksson 1993 for a complete overview of the ware, its geographic spread and its characteristics. 164 Eriksson 1993, 144. 165 See Gjerstad 1926; Sjöqvist 1940; Merillees 1963. 166 Eriksson 1993, 149; Knappett et al. 2005, 49; Kozal 2016, 53. 167 Gjerstad et al. 1934a, 359 + Plate XIII. Eriksson (1993, 221) classifies the published spindle bottle as a type VIa1c. 168 Åström 1972b, 205, 207.

36 fragments, possibly representing less than five vessels”.169 Åström also comments on most of the recorded potmarks, which are discussed below. In this analysis, 526 sherds of RL ware have been studied, 476 closed, 28 unknown, 13 open and nine handles. All handles are from closed vessels, probably Lentoid flasks. From these, 20 sherds have been further catalogued, 195–214. Distribution: The 526 RL sherds were primarily found in four concentrations. The biggest concentration of 246 sherds was in Grid E3 at a depth of 0–20cm. It consists of mostly fragmentary pieces of arm vessels. The second largest concentration, of 139 sherds, is unfortunately without grid or depth, and also consists mostly fragmented of arm vessel sherds. These sherds are much better preserved than the ones from E3, and it is likely that this is a collection of the “finer specimens” from the site, and therefore missing its proper context. The third largest concentration is in Grid D3, at a depth of 20–40cm with 88 sherds. There are some arm vessel fragments here as well, but not as many as in the other concentrations. Mostly it is sherds from Lentoid flasks, bowls and Spindle bottles. The final concentration consists of 39 sherds, and is from Grid D0, at a depth of 0–20cm. The three concentrations of RL ware suggest a clear presence in D3, E3 and D0. Additionally, there are several sherds belonging to the same vessel spread across these three grids. 195–197 all belong to the same vessel, and are found in both D3 (20–40cm) and E3 (0–20cm). 204–206 also join and are spread between D0 (0–20cm) and D3 (0–20cm). This means that there has been a movement of pottery between these grids. The wide distribution of arm vessel sherds also makes it difficult to assess if all arm vessels were originally placed in Grid D1. It is convenient to assume that most sherds from box 248 (the 139 sherds without context) probably came from D1, but this cannot be confirmed. The sherds from D0 are also close to this context. The 246 sherds at E3 and the 88 at D3 are more difficult to explain, and it seems logical that they have not moved very far, suggesting a presence of RL vessels in the central, western area of the site. Discussion: Bowls: Three sherds of a RL bowl, 195–197, have been catalogued. All three sherds belong to the same vessel, a type IAb bowl. According to Eriksson, the RL bowl is an uncommon shape and might have been quite valued in the LC society. It shares similarities with a PWWM sherd, 161. It has been suggested that the RL bowl does borrow its shape from PWWM vessels.170 The only complete examples of this type are from a tomb of uncertain date in Kazaphani, making a more precise date of introduction difficult to assess.171 Lentoid flasks: Seven rims have been catalogued, 198–206, two joining, 198– 199. All have a very thin diameter of 2–3cm. Sherds 200–203 are most likely from Lentoid flasks, while 198–199 are probably from a Lentoid flask, but may also be from a Spindle bottle.172 Additionally, necks, 205, five handles, 207–210 and two body sherds, 209–210, have also been catalogued. The handles have all been discussed by Åström, but some additional comments are needed. Åström is somewhat cautious with the typology of the handles, labelling them as ”VIIa”, ”VII(a?)” or ”VII, probably a”.173 Since there is no evidence of any other type of flask at the site, it seems reasonable that all handles belong to type VIIa flasks. The Lentoid flask is one of the more common RL shapes and has not been found in contexts earlier than LCIIA:1. Eriksson also notes that the shape is Anatolian, where similar types of vessels have previously been found.174 Potmarks:175The potmarks on 206 and 210

169 Åström 1972b, 205. 170 Eriksson 1993, 21. 171 Eriksson 1993, 21. 172 For 200–201, compare Eriksson 1993, fig. 5a (VIA1:612) and fig. 5b (1276). 173 Åström 1972b, 207. 174 Eriksson 1993, 25.

37 were identified by Åström as mark 31b. However, both seem to be more like mark 31a. Eriksson completely ignores 31b and settles for 31a as a single mark, 31. Åström identified the mark 207 as mark 29, but this seems to be either a mistake or a misprint, since it is quite clearly mark 28. The mark on 209 is identified by Åström as mark 19, which is not used by Eriksson, and seems to be a variation of 18. In either case, it is more likely to be mark 17, due to the single vertical line, crossed by a horizontal line. Mark 18, while similar to 17, consists of three separate lines, two vertical, separated by a horizontal. There is an additional horizontal line on the lower left of the mark, possibly why it was identified as mark 19 by Åström. Similar to 209, Åström identified the mark of 208 as mark 18. Again, this is more likely mark 17, for the same reason as above.176 Spindle Bottles: Åström identified 214 as a type VIA Spindle bottle, but it can most likely be narrowed down further to a VIa1b, given the shape of the base.177 Like the Lentoid flasks, the Spindle bottles belong to the most common shape of the RL repertoire, and have been found in contexts ranging from LCIA:2 – LCIIA:1178 The mark on 214 is not discussed by Åström, but is most likely a mark 13. Arm vessels: RL arm vessels are divided into two types, VIIIa and VIIIb. Type VIIIa is characterised as a shorter vessel, usually with the hand holding the cup on its side. The VIIIb type is usually longer, and the hand is placed underneath the cup. While the sherds of the arm vessels are all very fragmented, they seem to belong to the more common type VIIIb. The arm vessels are, according to Eriksson, a later addition to the repertoire, first produced around LCIIA:1. They have been found all over the Mediterranean and are usually discovered in ritual areas. Eriksson identifies Boghazköy in Anatolia as the site with most arm vessel fragments, having a total of 92 catalogued sherds.179 In this study, at least ten sherds have been identified as part of an arm vessel cup, with two sherds joining. This means there could be as many as nine additional vessels to the ones published in 1934. However, Åström’s caution is perhaps advised, as the sherds are all very worn and could very well represent fewer than nine vessels. It is also possible that some sherds belong to one of the almost complete vessels that are missing their hands. Even without an exact number, it seems like the Ayios Iakovos assemblage is more numerous than previously assumed in regards to both complete and fragmented arm vessels. The VIa1b Spindle bottle is one of the earliest shapes of RL ware, introduced during the LCIA:2 period. The type VIIAa Lentoid flask makes its introduction during the LCIIA:1 period, at which time the VIa1b Spindle bottle is the predominant shape. During the peak of the ware, LCIIA2, the arm vessel is introduced. Eriksson does not comment on the introduction of the IAb bowl, but the IAa bowl is also introduced during LCIIA:2.180 All in all, it seems like the RL assemblage at Ayios Iakovos predominately originates from LCIIA:2. This is at the height of the wares popularity, with the common Lentoid flask and Spindle bottles present, as well as the new, and more rare, shapes such as the arm vessel and bowl.

175 For Potmarks, see Eriksson 1993, 146, fig 41–42 (which is a more elaborated version of Åström 1972b, 207, fig. 42). 176 Additionally, Åström notes the context as E 30–20, it should be E3 0–20. This occurs on several posts, and is most likely a printing error. 177 Compare to Eriksson 1993, 22, fig. 5a. There is also another misprint in Åström 1972b, the context says ”DS 20–40”, it should be D3 20–40. 178 Eriksson 1993, 25. 179 Eriksson 1993, 27, Cat. No.1059–1151. 180 Eriksson 1993, 31, 35, 40 and 174–175.

38 White Painted Wheelmade Description: The White Painted wheelmade Ware (WPWM) is another one of the more distinct Cypriot pottery types. It shares many of the characteristics of the PWWM ware regarding fabric and slip, while also being decorated with a black or brown paint in various patterns. The ware occurs in a wide range of shapes, such as bowls, jugs, tankards and flasks, with the most common shape being jugs and tankards. Like other Cypriot wares, it is categorised with Roman numerals from I– VII, with VI being the LC type. This was initially a chronological division, but more recent studies have concluded that regionalism is also an important factor to consider for differences in ware and decoration.181 General comments: Two sherds of WPWM ware have been studied, both from open vessels. One of these has been further catalogued, 269. It is a small rim of a bowl with a black band decoration along the rim. Distribution: The sherd was found in Grid E3 at a depth of 0–20cm, making its original placement difficult to assess. Discussion: The closest parallel is found in Åström 1972, among one of the PWWM bowls.182 The most interesting aspect about these sherds is the fact that there are so few of this otherwise quite popular ware.

White Porous Description: This ware, while perhaps easy to categorise as a PWWM, does not quite fit the criteria. The fabric is much more porous and of distinctly different colour than the other wares. It is a very pale yellow, almost white colour. Without a clear parallel in the Cypriot assemblage, this study has labelled these sherds as “White porous” (WPo). General comments: A total of 12 sherds of this ware has been studied, both parts of a rim, handle, body and a disc base. Two sherds were catalogued, 264–265. It is likely that all twelve sherds belong to the same vessel. The shape is a piriform jar. Distribution: All sherds were found in Grid E3 at a depth of 0–20cm. Discussion: 264 and 265 seem to have been a piriform jar. It is of the same shape as 59, FS47. It should therefore also belong to same timeframe, LCII. While it is possible that this vessel was once decorated, there remains no trace of it now. Given the lack of decoration and the distinct fabric, it is likely that this vessel is a local imitation of a piriform jar.

White Slip Description: White Slip (WS) pottery is another characteristic ware of the LC repertoire. It was widely produced from c. LCIA1 until LCIIC, and to a certain extent up until LCIIIB2. Vessels of this ware have been found not only in Cyprus but also all around the eastern Mediterranean. It is easily identified by the thick, white, or light coloured slip, which covers the grey or dark brown fabric. The exterior is decorated with a dark brown/black or red paint. This ware is always handmade. The most common shape is the thin-walled hemispherical bowl with a singular, horizontal wishbone handle, usually called a ”milk bowl”.183 What is most distinct about the WS bowls is its long-lived lack of change. Other shapes, however, are also known, most notably the krater and tankard. These shapes are based on the hemispherical bowl, onto which a lip is added to create the krater or tankard. Like the bowls, these vessels were probably used for consumption of food and drinks. It is a ware that has been

181 See Eriksson 2009. 182 Åström 1972b, LXI.13. 183 Karageorghis 2001, 9; Popham 1972a 431; Popham 1972b, 700–701.

39 found in almost all types of contexts, ranging from settlements to burials and the common denominator seems to be its association with the consumption of food and liquids.184 The ware was initially described by Gjerstad and later divided into two types, WS I and WS II, by Petrie. While Petrie’s classification still remains more or less the same, later researchers have improved and broadened his types. Sjöqvist gave a more detailed description on the distinctions between the two types. 185 This identification was broadened in 1962 by M.R. Popham to include the newly identified early stages of White Slip, later called Proto White Slip.186 In 1972, Popham further divided the ware into five types, Proto White Slip, White Slip I, Transitional White Slip I–II, White Slip IIA and White Slip II. The main difference between these types is the quality and composition of the decorations. Popham’s study is still today the standard work for White Slip studies. Later studies have also concluded that the difference between the types can be found not only in the decoration but also in the manufacturing process. Aloupi, Perdikatsis & Lekka concluded that there is a clear distinction between WS I and WS II in regards to the components of the decoration and the slip.187 General comments: During the excavations at Dhima, an almost complete WS krater was found on the north side of the terracotta basin. This was the only WS find in the published report.188 In total, 48 sherds of WS ware have been studied, all from open vessels. From these, eight have been further catalogued, 270–277. Several sherds are joining or from the same vessels. Distribution: A majority of the WS (24 sherds) was found in Grid D3, which is just northeast of the terracotta basin, suggesting a concentrated presence close to the other finds. All 24 sherds were recorded at a depth of 20–40cm. Five additional sherds were found in E3, three at a depth of 0–20cm and two at a depth of 20–40cm. This makes at least two sherds likely to be related to the sherds in Grid D3. A second concentration was found in Grid D0, suggesting an association with the arm vessel deposit in Grid D1. However, all 16 sherds were found in the 0–20cm bracket, so it is possible that they have been moved from their initial location. Since some of these sherds match sherds found in Grid D3, the latter suggestion would seem more probable. Lastly, three sherds were found in Grid D6, at a depth of 20–40cm. This is beyond the limit of the site, and their placement can be due to one of many reasons. The relatively low number suggests a displacement, rather than a deposit. With the exceptions of the sherds in Grid D6, all sherds were found on the western part of the site. There is some evidence of a disturbance of the WS pottery. 270–272 are all joining and found in D3 (20–40cm) and D0 (0–20cm). Discussion: The complete krater was initially identified as WS I, but is classified as WS II at the Mediterranean Museum in Stockholm. This is a more likely classification based on the horizontal “hooked lozenges” pattern around the belly.189 The different decorations on 271, 273 and 274 suggest at least three vessels, possibly four. The decoration indicates that all vessels are either of WS IIA (273)190 or WS II (270–272 + 275–277) 191 type. This would be somewhat later than the already published WS I krater and dates the ware to LCIIA:2. The krater has, as previously

184 South and Steel 2001, 68. 185 Sjöqvist 1940, 43–50. 186 Appendix in Åström and Wright 1962. 187 Aloupi, Perdikatsis, and Lekka 2001, 24. 188 Gjerstad et al. 1934a, 357. 189 Gjerstad et al. 1934a, 357; Sesam (1997). 190 See Popham 1972a, 434, fig. 46g.5 & 445 fig. 51.6 for comparison. 191 See Popham 1972a, 450, fig. 54.7 and 54.9 for comparison.

40 mentioned, been re-classified to WS II, which would make it contemporary with the other sherds. The presence of WS II also suggests an early date of LCIIA1.192

4.2 Comments on the pottery This discussion will summarise some observations that can be made about the ceramic assemblage at Dhima, focusing on wares, shapes, recorded depth and distribution. On a general note, the MCIII/LCI wares are evenly distributed between the different wares, with the majority being PWHM (35%), followed by BS (25%), RS (24%) and RoB (16%). The exception is the WPHM and RP sherds, which both comprise of less than 1% of the total MCIII/LCI material.193 The majority of the vessels are closed, with roughly 37% open vessels.194 There is more diversification in the LCII period. The PWWM makes up for almost half of the material in this period (47%). The other two big groups are RL (28%) and Cypr. Myc. (17%). The remaining wares, BR (5%), WS (2%), WPo (1%), WPWM (>1%) and GW (>1%) make up less than 10% of the material combined. 195 During the LCII period, closed shapes dominate the material, with only 7% of the shapes being open.196 With the PW wares, it is likely that there is a certain degree of overlap between MCIII/LCI/LCII, due to the wares coexistence, especially for PWWM, which might range from LCI-LCII in some cases. Table 4 illustrates certain concentrations of the various wares. RS, BS, RoB and PWHM are primarily found on the eastern side of the sanctuary, with a sporadic presence on the western side. BR, Cypr. Myc., RL, WS, on the other hand, are mainly found on the western side, around the basin, with only some presence on the eastern side. The only ware that is present throughout the site is the PWWM ware. It is also evident that sherds from all periods were found in the outskirts of the floor area. This is interesting since there are numerous joining sherds between the central Grids E3 and D3 and grids outside the floor area, such as Grids A5, B5, D0 and D6. Taken on its own it might not seem like much more than the result of formation processes, such as agricultural activities. However, combined with the “simple” stratigraphy, it is plausible to argue that this is evidence of cleaning of the floor area. Old vessels, either broken or whole, were continuously removed from the sanctuary to make room for new vessels. If the floor was routinely kept clean from sherds and waste, it might not have left a clear stratigraphic trace. It would explain why sherds from so many grids apart find joins. It is unfortunate not more material from these outlying grids could be included (there surely must have been more). Whether it is because they were not excavated or that the pottery simply was not collected is impossible to say. But the few preserved grids that are left show that the sherds outside the sanctuary might be just as important to understand it as the sherds found within. Table 5 gives an overview of the vertical distribution. It shows that most sherds recorded at the lowest level at Dhima belong to wares dateable to the MCIII/LCI period, which was to be expected. Similarly, the LCII pottery is more or less contained within the 0–20cm or 20–40cm levels, corresponding to the floor level and above. It also shows that BS (including 30, 32–33), PWHM, RoB (including 216– 218, 224, 228 and 230–231) and RS (47, 249–250, 254, 256–258, 263 and 266)

192 Gjerstad et al. 1934a, 357; Popham 1972b, 700–701. 193 See Fig. 13. 194 See Fig. 14. 195 See Fig. 15. 196 See Fig. 16. It is likely that some of the PWWM sherds are wrongly classified as closed. It is no always clear when it comes to this ware.

41 pottery was intermixed with the upper levels, even above floor level, although in few numbers. What becomes increasingly clear is that the chronological sequence must be examined. Dhima was concluded to have had a MCIII occupation (The MCIII-pit), followed by a LCI hiatus and a revival during the LCII period (with the construction of the floor, wall, platforms and terracotta basin. The reason behind this chronology was the lack of finds dating to the LCI period above the floor level.197 However, as is evident from BS, RoB and RS sherds in Table 5, there is enough evidence dating to the MCIII/LCI period found above the floor level at Dhima to suggest that this hiatus must be questioned. Since no PWWM pottery can be associated with the contents of the MCIII-pit, it is likely that this initial dating of MCIII can be kept, especially considering 142, which is from a typical MCIII pithos and was found below the floor. PWWM did not enter the Cypriot ceramic repertoire until the LCI period.198 This means that the floor must be constructed sometime during the LCI period, given the presence of BS (including 30, 32–33), RoB (including 216–218, 224, 228 and 230– 231) and RS (including 47, 249–250, 254, 256–258, 263 and 266) found above it. Since the floor runs up against the constructions, 199 it suggests that they were constructed before the floor, in the LCI period. The activities continued during the LCII period, as is evident by the presence of LCII wares, such as BR, Cypr. Myc., RL, and WS. The Mycenaean sherds offer some further insight into the LCII chronology. FS 31 (90, 99-105) and 44 (38-40, 44-48) are most commonly dated to the LHIIIA1 period, while FS 39 (41-43) and 45 (49-54) and FS 166 (108-109) are dated to the LHIIIA2 period.200 This indicates that imported vessels were brought to the sanctuary during a prolonged period of time, up until the LCIIB period. If the sanctuary was routinely cleared of items, as suggested above, it explains the interpretation of a hiatus. Only the last votives were studied at any length in the publication and the stray sherds were of little interest. It is also possible to discuss potential activities conducted at the sanctuary with a bit more certainty than before. The presence of a Mycenaean krater (No. 44), several PWWM sherds possibly originating from kraters (161–162, 164–165, 166– 168, 171 and 174–175), PWWM bowls (150–160 and 178), BR cups (2–26) and WS bowls (270–277) suggest that some form of consumption occurred. During the LCI period, PW was considered to be the standard equipment for consumption, both regarding larger vessels, as well as smaller bowls for personal use. This gradually changed during the LCII period, when first BR and WS bowls/cups and later Mycenaean cups started to replace PW.201 With this in mind, perhaps what we see at Dhima is part of this change, with the presence of both PW cups and bowls, as well as BR and WS bowls. The absence of any Mycenaean cups at Dhima suggests that such shapes had not yet been introduced in the broader consumption repertoire. Instead, BR and WS bowls must have been used, possibly alongside PW bowls. Furthermore, since there is no secure record of osteological material, it seems like this consumption was limited to liquids. Even more evident than consumption is the abundance of closed shapes at Dhima during the LCII period. There is no doubt some misinterpreted sherds, due to the ambiguous nature of the PWWM ware, but nevertheless, the vast majority (93%) of sherds are from closed vessels. Most notable are the many piriform jars (38–107 and 268–269). The piriform jar is among the most popular Mycenaean vessels in

197 Gjerstad et al. 1934a, 361. 198 Crewe 2007, 218. 199 Gjerstad et al. 1934a, 356. 200 Mountjoy 1986, 202. 201 Crewe 2016, 128.

42 Cyprus, imported for its precious contents.202 Almost all of the Cypr. Myc. sherds at Dhima can be attributed to this shape.203 There is some variation, illustrated by FS 31 (90, 99-105), 39 (41-43), 44 (38-40, 44-48), 45 (49-54), but is seems clear that their content must have been important. The two catalogued stirrup jars (FS166, 108–109) and almost all sherds of RL ware (including 198–214) are also closely connected to the transportation of oils and perfumes.204 Vessels of these types, and their contents, have been connected with libation rites in Cyprus, originally in mortuary settings but from the LCII period and onwards also in sanctuaries.205 In later sanctuaries, such as Myrtou-Pighades and Athienou Bamboulari tis Koukounninas, the rhyton seems to have been established as the go-to libation vessel.206 However, there is no evidence of a rhyton at Dhima. Instead, the vessels most clearly connected with libation are the LR arm vessels.

202 Steel 1998, 286. 203 Only 11 sherds can securely be categorized as stirrup jars. Originating from two vessels. 204 Knappett et al. 2005, 49; Steel 1998, 286. 205 Steel 1998, 75; 77. 206 Webb 1999, 200–201.

43 Table 3. Catalogued sherds.

ID Grid Depth H/W Frag. Ø Joins Munsell Black-on-Red 1 E2 0–20 H Bo - - 7.5YR 6/6 (E) 10YR 6/4 (I) 10YR7/3 (S) Base Ring (BR) Carinated Bowls 2 D0 0–20 H Ri 16 2, 3, 4 7.5YR 7/6 (E/I) 7.5YR 5/1 (C) 7.5YR 6/6 (ES/IS) 3 D3 20–40 H Ri 16 2, 3, 4 7.5YR 7/6 (E/I) 7.5YR 5/1 (C) 7.5YR 6/6 (ES/IS) 4 D3 20–40 H Ri 16 2, 3, 4 7.5YR 7/6 (E/I) 7.5YR 5/1 (C) 7.5YR 6/6 (ES/IS) 5 E3 0–20 H Ri 14 - 7.5YR 6/6 (E/I) 7.5YR 4/1 5YR 7.5YR5/1 (ES) 6/6 (IS) 6 E3 0–20 H Ri + Ha c. 24 - 5YR 5/6 (E/I) 10YR 4/1 GLEY1 4/ (ES) 7.5YR 4/2 (IS) 7 E3 0–20 H Ri - - 2.5YR 4/6 (E/I) 10YR 4/1 (C) 10YR 5/2 (ES) 5YR 6/8 (IS) 8 D3 20–40 H Ri + Ha - - Too thin (E/I) 10YR 5/2 (C) 10R 4/1 (ES) 9 D0 0–20 H Ri + Ha - - Too thin (E/I) 10YR 5/2 (C) 7.5YR 6/6 (ES) 5YR 6/5 (IS) 10 D6 20–40 H Ri 18 10, 11 7.5YR 6/4 (E/I) 2.5YR 4/6 (C) 2.5Y 4/1 (ES) 10YR 4/1 (IS) 11 D6 20–40 H Ri 18 10, 11 7.5YR 6/4 (E/I) 2.5YR 4/6 (C) 2.5Y 4/1 (ES) 10YR 4/1 (IS) 44 12 D3 20–40 H Ri 16 12, 13 5YR 5/6 (E/I) 5YR 5/6 (C 7.5YR 4/3 (ES/IS) 13 D3 20–40 H Ri 16 12, 13 5YR 5/6 (E/I) 5YR 5/6 (C 7.5YR 4/3 (ES/IS) 14 D6 20–40 H Ri 17 - GLEY1 4/ (E/I/C) 7.5YR 6/4 (ES/IS) 15 D0 0–20 H Ri 20 - GLEY1 4/ (E/I/C) 7.5YR 5/1 (ES/IS) 16 D3 20–40 H Ri 18 16, 17 5YR 5/6 (E) 2.5YR 4/1 (I) Too thin (C) 2.5YR 4/6 (ES) GLEY1 2.5/ (IS) 17 D3 20–40 H Ri 18 16, 17 5YR 5/6 (E) 2.5YR 4/1 (I) Too thin (C) 2.5YR 4/6 (ES) GLEY1 2.5/ (IS) 18 A5 40-100 H Ri - - Too think (E/I) 2.5YR 5/1 (C) 2.5YR 3/1 (ES/IS) 19 D3 20–40 H Ri - - 5YR 6/6 (E/I) 2.5YR 5/1 (C) 7.5YR 4/1 (ES) Too worn (IS) 20 E3 0–20 H Ri c. 18 20, 21 10R 4/6 (E/I) 7.5YR (5/4) (C) 7.5YR 4/2 (ES) 10YR 4/1 (IS) 21 D6 20–40 H Ri c. 18 20, 21 10R 4/6 (E/I) 7.5YR (5/4) (C) 7.5YR 4/2 (ES) 10YR 4/1 (IS) 22 D6 20–40 H Ri - 22, 23, 24 5YR 4/4 (E) GLEY2 4/1 (I) GLEY1 4/ (ES/IS) 23 D6 20–40 H Ha - 22, 23, 24 5YR 4/4 (E) GLEY2 4/1 (I) GLEY1 4/ (ES/IS) 24 D6 20–40 H Ha - 22, 23, 24 5YR 4/4 (E) GLEY2 4/1 (I) GLEY1 4/ (ES/IS) 25 D6 20–40 H Rb 6 - GLEY1 5/ (E/I/C) 10YR 6/4 (ES) 26 D0 0–20 H Rb 6 - 2.5Y 6/1 (E/I/C) 5YR 6/6 (ES) GLEY1 3/ IS) Jugs 27 D6 20–40 H Rb 6 - 5YR 5/6 (E) GLEY1 5/ (I) GLEY 1 3/ (ES)

ID Grid Depth H/W Frag. Ø Joins Munsell 28 D6 20–40 H Rb 7 - 7.5YR 5/6 (E) GLEY1 3/ (I) 7.5YR 5/2 (C) 5YR 6/6 (ES) Black Slip (BS) Bowls 29 E5 - H Ri+Ha 30 29, 30 10YR 6/4 (E/I) 7.5YR 6/6 (C) 5YR 3/1 (ES/IS) 30 F5 1stFL H Ri+Ha 30 29, 30 7.5YR 6/4 (E/I/C) 5YR 3/1 (ES/IS) 31 B5 0–20 W Ri 20 - 7.5YR 5/4 (E/I/C) 2.5YR 3/1 (ES/IS) 32 D0 0–20 H Ri 18 - 5YR 6/6 (E/I) GLEY1 4/ (C) GLEY1 3/ (ES/IS) 33 E3 0–20 H Ri 18 - 7.5YR 5/4) E/I/C) 10YR 5/2 (ES/IS) 34 B3 20–40 H Ri c. 14 - 2.5Y 7/3 (E/I) 10YR 6/4 (C) GLEY1 3/ (ES/IS) Spouted Bowls 35 B5/E5 20–40 H Ri+Sp - - 5YR 5/6 (E/I/C) 5YR 3/1 (ES/IS) Jugs 36 B3 20–40 H Ri 8 - 2.5YR 4/2 (E) 2.5YR 4/6 (I/C) Cooking Ware (CW) Jars 45 37 A5 40–100 W Ri 14 - 2.5YR 4/2 (E) 2.5YR 4/6 (I/C) Cypro-Mycenaean (CM) Piriform Jars 38 D3 20–40 W Ri 8 38, 39 5YR 6/6 (EI/C) 10YR 8/4 (ES/IS) 5YR 5/6 (ED/IS) 39 D0 0–20 W Ri 8 38, 39 5YR 6/6 (EI/C) 10YR 8/4 (ES/IS) 5YR 5/6 (ED/IS) 40 D0 0–20 W Ri 6 40, 84 10YR 7/4 (E/I) 5YR 6/8 (C) 10YR 7/4 (ES) 5YR 4/2 (ED/IS) 41 D0 0–20 W Ri 7 - 2.5Y 6/4 (E/I/C) 10YR 7/4 (ES/IS) 10YR 2/1 (ED/ID) 42 D0 0–20 W Ri 8 - 5Y 7/3 (E/I/C) 10YR 8/3 (ES/IS) 10YR 3/1 (ED/ID) 43 D6 20–40 W Ri 8 - 10YR 6/3 (E/I/C) 2.5Y 7/4 (ES/IS) GLEY1 2.5 (ED/ID) 44 D3 20–40 W Ri 8 44, 45? 2.5Y 7/3 (E/I/C) 2.5Y 8/3 (ES/IS) GLEY1 3/ (ED/ID) 45 D0 0–20 W Ri 8 44, 45? 2.5Y 7/3 (E/I/C) 2.5Y 8/3 (ES/IS) GLEY1 3/ (ED/ID) 46 E3 0–20 W Ri + Bo 8 46, 47, 48 2.5Y 6/3 (E/I) 7.5YR 7/6 (C) 2.5Y 8/3 (ES/IS) 7.5YR 3/1 (ID) 47 D0 0–20 W Ri 8 46, 47, 48 2.5Y 6/3 (E/I) 7.5YR 7/6 (C) 2.5Y 8/3 (ES/IS) 7.5YR 3/1 (ID) 48 D3 20–40 W Ri 8 46, 47, 48 2.5Y 6/3 (E/I) 7.5YR 7/6 (C) 2.5Y 8/3 (ES/IS) 7.5YR 3/1 (ID) 49 D3 20–40 W Ri 8 - 2.5Y 7/3 (E/I) 7.5YR 6/6 (C) 10YR 7/4 (ES/IS) 50 E3 0–20 W Ri 10 - 10YR 8/4 (E/I) 10YR 6/4 (C) Too worn (ES/ED) 51 D0 0–20 W Ri - 10YR 6/4 (E/I) 7.5YR 6/6 (C) 10YR 7/3 (ES/IS) Too worn (ED)

ID Grid Depth H/W Frag. Ø Joins Munsell 52 D3 20–40 W Ri 8 - 2.5Y 7/3 (E/I) 7.5YR 7/6 (C) 10YR 8/3 (ES/IS) 7.5YR 3/1 (ED/ID) 53 D3 20–40 W Ri 6 53, 54? 10YR 7/4 (E/I/C) 2.5Y 7/6 (ES/IS) 10YR 3/1 (ED/ID) 54 D0 0–20 W Ri 6 53, 54? 10YR 7/4 (E/I/C) 2.5Y 7/6 (ES/IS) 10YR 3/1 (ED/ID) 55 E3 0–20 W Ri 8 - 10YR 7/4 (E/I) 10YR 6/4 (C) 10YR 8/4 (ES) 10YR 3/1 (ID) 56 E3 0–20 W Ri c. 7-8 - 10YR 7/4 (E/I/C/ES/IS) 57 E3 0–20 W Ri 8 - 10YR 8/4 (E/I) 7.5YR 8/6 (C) Too worn (ES/ED) 58 E3 0–20 W Ri 8 - 10YR 8/4 (E/I) 7.5YR 5/6 (C) Too worn (ES) 5YR 5/4 (ED) 59 E3 0–20 W Ri 7 59, 60, 101? 10YR 8/6 (E) 5YR 6/8 (I) 10YR 6/4 (C) Too worn (ES/ED) 60 E3 0–20 W Ri 7 59, 60, 101? 10YR 8/6 (E) 5YR 6/8 (I) 10YR 6/4 (C) Too worn (ES/ED) 61 D0 0–20 W Ri - 10YR 6/4 (E/I/C) 10YR 7/4 (ES/IS) Too worn (ED/ID) 62 D0 0–20 W Ri 6 - 10YR 7/4 (E/I/C) 10YR 6/4 (ED) 63 E4 0–20 W Ri 6 - 10YR 6/4 (E/I/C) GLEY1 2.5/ (ED/ID) 64 D3 20–40 W Ri 8 - 7.5YR 7/4 (E/I/C) 10YR 7/4 (ES/IS) 10YR 6/6 (ED) 7.5YR 3/1 (ID) 65 D3 20–40 W Ri 8 - 10YR 8/4 (E/I/C) 10YR 8/3 (ES/IS) 7.5YR 6/6 (ED) 66 D3 20–40 W Ri 6 - 10YR 6/4 (E) 7.5YR 6/6 (I) 10YR 4/1 (ED) 46 67 D3 20–40 W Ri 7 67, 68, 99 2.5Y 7/3 (E/I/C) 10YR 8/3 (ES/IS) GLEY2 2.5/1 (ED) 68 E3 0–20 W Ri + Db c. 10 (Ri), 5 (Db) 67, 68, 99 7.5YR 8/4 (E/I) 7.5YR 8/6 (C) 10YR 7/6 (ES) 7.5YR 6/6 (ED) 69 D3 20–40 W Ri 8 - 10YR 7/6 (E/I/C) 10YR 8/4 (ES/IS) Too worn (ED) 70 D3 20–40 W Ri 8 - 7.5YR 7/4 (E/I) 2.5Y 7/3 (C) 10YR 8/4 (ES/IS) 7.5YR 5/3 (ED) 71 E3 0–20 W Ri c. 10 - 7.5YR 7/6 (E/I) 5YR 5/6 (C) 72 E3 0–20 W Ri - - 7.5YR 5/6 (E/I) 10YR 6/4 (C) 10YR 7/4 (ES/IS) 73 D0 0–20 W Ri 7 - 7.5YR 6/6 (E/I) 5YR 5/6 (C) 74 D3 20–40 W Ri 9 - 10YR 8/3 (E/I) 7.5YR 8/4 (C) 75 D3 20–40 W Ri 7 - 7.5YR 6/6 (E/I/C) 7.5YR 7/6 (ES/IS) 76 D3 20–40 W Ri 8 - 7.5YR 7/8 (E/I/C) 10YR 8/3 (ES/IS) 77 D5 60–80 W Ri 8 - 10YR 8/4 (E/I/C) 10YR 8/3 (ES/IS) 78 E3 0–20 W Ri + Bo 7 - 10YR 7/4 (E/I/C/ES) GLEY1 3/ /ED) 79 D0 0–20 W Ri 6 79, 80, 81 10YR 7/4 /E/I/C/ES) 7.5YR 4/2 + 5YR 5/8 (ED) 80 D3 20–40 W Bo - 79, 80, 81 10YR 7/4 /E/I/C/ES) 7.5YR 4/2 + 5YR 5/8 (ED) 81 D3 20–40 W Ha - 79, 80, 81 10YR 7/4 /E/I/C/ES) 7.5YR 4/2 + 5YR 5/8 (ED) 82 E4 0–20 W Bo - 82, 83 10YR 7/4 (E/I) 10YR 6/4 (C) 10YR 8/3 (ES) GLEY1 3/ (ED) 83 D3 20–40 W Bo - 82, 83 10YR 7/4 (E/I/C/IS) 10YR 8/3 (ES) 10YR 3/2 (ED)

ID Grid Depth H/W Frag. Ø Joins Munsell 84 E3 0–20 W Bo - 40, 84 10YR 7/4 (E/I) 5YR 6/8 (C) 10YR 7/4 (ES) 5YR 4/2 (ED/IS) 85 D3 20–40 W Bo - - 5YR 7/6 (E) 10YR 7/4 (I) 2.5Y 7/3 (C) 10YR 7/4 (ES/IS) 7.5YR 4/3 (ED) 86 D3 20–40 W Bo - - 2.5Y 7/3 (E/I/C) 2.5Y 7/4 (ES/IS) 10YR 3/1 (ED) GLEY1 3/ (ID) 87 E3 0–20 W Bo - - 10YR 7/4 (E/I/ES) 10YR 7/3 (C) GLEY1 2.5 (ED) 88 D3 20–40 W Bo - - 2.5Y 7/3 (E/I/C) 10YR 8/3 (ES/IS) 89 D3 20–40 W Rb 3,6 - 5YR 5/6 (E/I) Too worn (C/ES) 2.5Y 5/8 (ED) 90 D3 20–40 W Rb 4 - 10YR 7/4 (E/I/C/ES) GLEY1 3/ (ED) 91 D3 20–40 W Udb 5 - 7.5YR 7/6 (E/I/C) 10YR 7/4 (ES) 7.5YR 5/6 (ED) 92 D3 20–40 W Udb 5 - 10YR 8/4 (E/I/ES) 7.5YR 7/4 (C) 7.5YR 5/6 (ED) 93 E3 0–20 W Udb 4 - 10YR 7/4 (E/I/C) Too worn (ED) 94 E3 0–20 W Rb 4 - 7.5YR 7/4 (E/I) 2.5Y 7/3 (C) 19YR 7/4 (ES/IS) 7.5YR 3/2 (ED) 95 D0 0–20 W Rb 5 - 2.5Y 7/3 (E/I) 2.5Y 6/4 (C) 2.5Y 7/3 (ES) 2.5Y 5/1 (ED) 96 E3 0–20 W Rb 4 96, 97, 98 2.5Y 7/4 (E/I) 2.5Y 8/3 (ES 2.5Y 3/1 (ED) 97 D0 0–20 W Rb 4 96, 97, 98 2.5Y 7/4 (E/I) 2.5Y 8/3 (ES 2.5Y 3/1 (ED) 98 D3 20–40 W Rb 4 96, 97, 98 2.5Y 7/4 (E/I) 2.5Y 8/3 (ES 2.5Y 3/1 (ED) 47 99 D6 20–40 W Db 5 67, 68, 99 10YR 7/4 (E) 7.5YR 7/4 (I) 10YR 8/4 (ES) GLEY1 3/ (ED) 100 D6 20–40 W Db 4,5 - 10YR 6/4 (E/I/C) 10YR 7/3 (ES) GLEY1 2.5/ (ED) 101 E3 0–20 W Db 4 59, 60, 101? 7.5YR 6/6 (E/I) 10YR 7/4 (C) 102 E3 0–20 W Db 4 - 7.5YR 6/6 (E/I) 2.5Y 7/3 (C) 2.5Y 5/6 (ED) 103 E3 0–20 W Db 4 - 10YR 7/4 (E/I/ES) 10YR 6/4 (C) GLEY1 2.5/ (ED) 104 E3 0–20 W Db 4 - 7.5YR 7/6 (E/I) 10YR 7/4 (C) Too worn (ES) 7.5YR 6/6 (ED) 105 B1 0–20 W Db 4,5 - 10YR 7/4 (E/I/C) 10YR 8/3 (ES/IS) 106 D0 0–20 W Rb 5 - 10YR 7/4 (E/I) 10YR 6/4 (C) 10YR 2/1 (ED) 107 D0 0–20 W Rb 4,5 - 2.5Y 7/4 (E/I) 7.5YR 6/6 (C) 10YR 7/4 (ES) 5YR 3/1 (ED) Stirrup jars 108 D3 20–40 W Ri 4 (Ri), 3-4 (Di), 7 (Rb) - 7.5YR 7/6 (E) 10YR 8/4 (I) 7.5YR 7/6 (ES) 2.5YR 4/6 (ED) 109 D3 20–40 W Di - 5YR 7/6 (E/I) 2.5Y 7/3 (C) 10YR 7/4 (ES) 2.5YR 4/8 (ED) Grey Ware (GW) Bowls 110 B5 20–40 W Ri c. 20 - 7.5YR 4/1 (E/I) 7.5YR 5/1 (C) 111 D3 20–40 W Ri c. 16 - 7.5YR 5/1 (E/I/C/ES/IS)

ID Grid Depth H/W Frag. Ø Joins Munsell 112 D5 60–80 W Ri 14 - 2.5Y 4/1 (E/I) 5YR 4/4 (C) GLEY 1 4/ (ES/IS) Plain White Handmade (PWHM) Bowls 113 D0 0–20 H Ri - - 5YR 5/4 (E/I) Too thin (C) 114 E5 - H Ri 23 - 10YR 6/4 (E/I/C) 10YR 7/4 (ES/IS) 115 D5 60–80 H Ri 12 - 2.5Y 5/4 (E/I/C) 10YR 8/3 (ES/IS) 116 D5 60–80 H Ri 33 116, 117 10YR 6/3 (E/I/C) 2.5Y 7/4 (ES/IS) 117 E5 - H Ri 33 116, 117 10YR 6/3 (E/I/C) 2.5Y 7/4 (ES/IS) 118 E5 - H Ri 22 - 7.5YR 6/6 (E/I/C) 10YR 8/4 (ES/IS) 119 D5 60–80 H Ri 22 - 5YR 7/6 (E/I) 10YR 8/3 (C) 2.5Y 8/2 (ES/IS) 120 D5 60–80 H Ri - - 10Yr 7/4 (E/I/C) 2.5Y 8/2 (ES/IS) 121 D5 60–80 H Ri 18 - 10YR 5/4 (E/I/C) 10YR 8/3 (ES/IS) 122 D5 60–80 H Ri c. 28 122, 123 7.5YR 5/6 (E/I/C) 10YR 8/4 (ES) 7.5YR 7/6 (IS) 123 D0 0–20 H Ri c. 28 122, 123 7.5YR 6/6 (E/I) 7.5YR 6/4 (C) 124 D4 20–40 H Ri 28 - 7.5YR 6/4 (E/I/C) 48 125 B5/E5 20–40 H Ri 43 - 2.5YR 7/6 (E/I) 10YR 6/4 (C) 10YR 8/3 (ES/IS) 126 E5 40–60 H Ri 30 126, 127, 128, 129 10YR 6/4 (E/I) 10YR 6/3 (C) 10YR 7/4(ES/IS) 127 E2 0–20 H Ri 30 126, 127, 128, 129 10YR 6/4 (E/I) 10YR 6/3 (C) 10YR 7/4(ES/IS) 128 B5/E5 20–40 H Ri 30 126, 127, 128, 129 10YR 6/4 (E/I) 10YR 6/3 (C) 10YR 7/4(ES/IS) 129 F5 1stFL H Ri 30 126, 127, 128, 129 10YR 6/4 (E/I) 10YR 6/3 (C) 10YR 7/4(ES/IS) 130 D5 60–80 H Ri 30 - 10YR 6/4 (E/I) 10YR 6/2 (C) 10YR 7/4 (ES/IS) 131 B3 20–40 H Ri 22 - 10YR 6/4 (E/I) 10YR 6/3 (C) 10YR 8/3 (ES/IS) 132 B5/E5 20–40 H Ri 38 - 7.5YR 6/4 (E/I/C) 10YR 8/3 (ES/IS) 133 F5 1stFL H Ri+Ha 42 133, 134 10YR 6/4 (E/I/C) 10YR 8/3 (ES/IS) 134 F5 1stFL H Ri 42 133, 134 10YR 6/4 (E/I/C) 10YR 8/3 (ES/IS) Jars 135 B5 20–40 H Ri c. 30 - 5YR 5/6 (E/I/C) 10YR 8/3 (ES/IS) 136 E5 - H Ri 32 - 10YR 8/4 (E/I) 10YR 6/3 (C) 10YR 8/3 (ES/IS) 137 D6 20–40 H Ri 10 - 2.5Y 7/2 (E/I) 2.5Y 6/1 (C) 138 D4 20–40 H Ri 11 - 7.5YR 8/6 (E/I) 7.5YR 6/4 (C) 2.5Y 8/3 (ES/IS) 139 D4 20–40 H Ri 14 - 10YR 5/4 (E/I) Too thin (C) 10YR 7/4 (ES/IS)

ID Grid Depth H/W Frag. Ø Joins Munsell 140 D4 20–40 H Ri 16 - 5YR 5/6 (E/I) 7.5YR 6/3 (C) 10YR 8/3 (ES) 141 E2 0–20 H Ri 29,5 - 7.5YR 6/6 (E/I/C) 2.5Y 8/2 (ES/IS) 142 D5 60–80 H Bo - - 10YR 6/4 (E/I) 5YR 7/4 (C) 10YR 8/2 (ES) 7.5YR 8/4 (IS) 143 E5 - H Bo - - 7.5YR 6/6 (E/I) 10YR 7/2 (C) 10YR 7/4 (ES/IS) 144 B3 20–40 H Bo - - 7.5YR 6/4 (E) 10YR 8/4 (I) 10YR 7/4 (ES) 145 E5 40–60 H Fb 19 - 10YR 6/3 (E/I) 10YR 6/1 (C) 10YR 7/4 (ES/IS) 146 B3 20–40 H Fb 22 - 2.5Y 7/3 (E/I) 5Y 6/3 (C) 147 B5/E5 20–40 H Fb 14 - 10YR 5/3 (E/I) 10YR 6/2 (C) 10YR 8/2 /ES/IS) 148 F5 1stFL H Fb c. 20 - 10YR 6/4 (E) 10YR 7/3 (I) 10YR 8/2 (ES) Jugs 149 B3 20–40 W Ri+Ha - - 10YR 7/3 (E/I) 2.5Y 4/2 (C) Plain White Wheelmade (PWWM) Bowls 150 B5 20–40 W Ri c. 14 - 10YR 6/4 (E) 5YR 7/6 (I) 10YR 8/4 (ES/IS) 151 D3 20–40 W Ri - - 10YR 6/4 (E/I/C) 2.5YR 7/4 (ES/IS) 49 152 E2 0–20 W Ri 16 152, 153, 154, 155, 156 7.5YR 7/4 (E/I/C) 2.5Y 8/2 (ES/IS) 153 D3 20–40 W Ri 16 152, 153, 154, 155, 156 7.5YR 7/4 (E/I/C) 2.5Y 8/2 (ES/IS) 154 E4 0–20 W Ri 16 152, 153, 154, 155, 156 7.5YR 7/4 (E/I/C) 2.5Y 8/2 (ES/IS) 155 D0 0–20 W Ri 16 152, 153, 154, 155, 156 7.5YR 7/4 (E/I/C) 2.5Y 8/2 (ES/IS) 156 E3 0–20 W Ri 16 152, 153, 154, 155, 156 7.5YR 7/4 (E/I/C) 2.5Y 8/2 (ES/IS) 157 D0 0–20 W Ri c. 17 - 2.5YR 5/6 (E/I/C) 2.5Y 8/3 (ES) 10YR 7/4 (IS) 158 D3 20–40 W Ri 24 159, 160 7.5YR 7/4 (E/I/C) 7.5YR 7/6 (ES) 10YR 7/4 (IS) 159 B5 20–40 W Ri 24 159, 160 10YR 6/2 (E 5YR 7/6 (I) 2.5Y 8/2 (ES/IS) 160 A5 0–20 W Ri c. 12 - 5YR 7/6 (E/I/C) 10YR 8/3 (ES) 161 D4 20–40 W Ri 26 - Too thin (E/I) 7.5YR 6/6 (C) 10YR 7/4 (ES/IS) 162 E2 0–20 W Ri 22 162, 163 10YR 6/4 (E/I/C) 2.5Y 7/3 (ES/IS) 163 D3 20–40 W Ri 22 162, 163 10YR 6/4 (E/I/C) 2.5Y 7/3 (ES/IS) 164 D6 20–40 W Ri 32 - 5YR 7/6 (E/I/C) 10YR 8/3 (ES/IS 165 D4 20–40 W Ri 36 - 10YR 5/2 (E/I) 10YR 5/3 (C) 10YR 7/3 (ES/IS) 166 D6 20–40 W Ri 34 - 7.5YR 6/6 (E/I) 10YR 5/4 (C) 2.5Y 8/3 (ES/IS) 167 E2 0–20 W Ri 54 - 10YR 5/4 (E/I/C) 10YR 8/4 (ES/IS)

ID Grid Depth H/W Frag. Ø Joins Munsell 168 B5 20–40 W Ri 36 - 10YR 6/6 (E/I/C) 10YR 8/3 (ES/IS) 169 A3 20–40 W Ri 39 - 7.5YR 6/4 (E/I) 10YR 5/3 (C) 170 B5 0–20 W Ri? 40 - 10YR 8/4 (E/I) 5YR 7/6 (C) 171 F5 1stFL W Ri 42 171, 172, 173 2.5Y 6/2 (E/I) 7.5YR 6/4 (C) 25Y 7/4 (ES/IS) 172 D0 0–20 H Ri 42 171, 172, 173 2.5Y 6/2 (E/I) 7.5YR 6/4 (C) 25Y 7/4 (ES/IS) 173 D5 60–80 W Ri 42 171, 172, 173 2.5Y 6/2 (E/I) 7.5YR 6/4 (C) 25Y 7/4 (ES/IS) 174 B1 0–20 W Ri 16 - 7.5YR 6/4 (E/I/C) 2.5Y 7/3 (ES/IS) 175 D5 60–80 W Ri 20 - 7.5YR 7/3 (E) 10YR 6/4 (I) Too thin (C) 176 E3 0–20 W Ri c. 16 - 7.5YR 6/6 (E/I/C) 10YR 8/3 (ES/IS) 177 D4 20–40 W Ri 32 - 10YR 8/4 (E/I 7.5YR 6/6 (C) Carinated Bowls 178 E4 0–20 W Ri 16 178, 179 5YR 6/4 (E/I/C) 2.5Y 8/3 (ES/IS) 179 D0 0–20 W Ri 16 178, 179 5YR 6/4 (E/I/C) 2.5Y 8/3 (ES/IS) Jugs and jars 180 D0 0–20 W Ri 8 - 10YR 7/4 (E/I) 7.5YR 6/6 (C) 50 181 B1 0–20 W Ri 11 - 7.5YR 6/4 (E/I) 10YR 7/4 (C) 2.5Y 7/3 (ES/IS) 182 E4 0–20 W Ri 13 - 10YR 7/4 (E/I) 7.5YR 6/6 (C) 183 A5 0–20 W Ri 14 183, 184 7.5YR 6/6 (E/I) 7.5YR 5/6 (C) 10YR 6/6 (ES) 184 D0 0–20 W Ri 14 183, 184 7.5YR 6/6 (E/I) 7.5YR 5/6 (C) 10YR 6/6 (ES) 185 D3 20–40 W Ri 18 - 10YR 7/4 (E/I/C) 186 A6 20–40 W Ri 18 186, 187, 188 5YR 6/4 (E/I/C) 2.5Y 8/3 (ES/IS) 187 E4 0–20 W Ri 18 186, 187, 188 5YR 6/4 (E/I/C) 2.5Y 8/3 (ES/IS) 188 E2 0–20 W Ri 18 186, 187, 188 5YR 6/4 (E/I/C) 2.5Y 8/3 (ES/IS) 189 F5 1stFL W Ri+Ha 9 - 7.5YR 7/4 (E/I) 7.5YR 6/6 (C) Unknown, closed 190 A3 20–40 W - 20 - 7.5YR 6/4 (E/I/C) 2.5Y 8/3 (ES/IS) 191 E2 0–20 W Rb 11 - 5YR 7/6 (E/I/C) 10YR 8/3 (ES) 192 D6 20–40 W Rb 11 - 7.5YR 7/4 (E/I) 5YR 6/6 (C) 2.5Y 8/2 (ES) 193 D6 20–40 W Pb 3,5 - 2.5Y 7/3 (E) 10YR 7/4 (I) 2.5Y 6/2 (C) Unknown, open 194 D6 20–40 W Rb 6 - 7.5YR 6/4 (E/I/C) 10YR 8/4 (ES/IS)

ID Grid Depth H/W Frag. Ø Joins Munsell Red Lustrous (RL) Bowls 195 E3 0–20 W Ri 24 195, 196, 197 5YR 5/8 (E/I/C) 5YR 6/8 (ES/IS) 196 D3 20–40 W Ri 24 195, 196, 197 5YR 5/8 (E/I/C) 5YR 6/8 (ES/IS) 197 E3 0–20 W Ri 24 195, 196, 197 5YR 5/8 (E/I/C) 5YR 6/8 (ES/IS) Lentoid flasks 198 D6 20–40 W Ri 2,5 198, 199 Too thin (E) 5YR 7/6 (I) GLEY1 5/ (C) 5YR 7/6 (ES) 199 D3 20–40 W Ri 2,5 198, 199 Too thin (E) 5YR 7/6 (I) GLEY1 5/ (C) 5YR 7/6 (ES) 200 E3 0–20 W Ri 2 - 5YR 5/8 (E/I/C) Too worn (ES) 201 D3 20–40 W Ri 2 - 5YR 7/6 (E/I/C) 202 E3 0–20 W Ri 3 - 5YR 5/8 (E/I/C) 5YR 7/6 (ES/IS) 203 E3 0–20 W Ri 3 - 5YR 5/8 (E/I/C) 5YR 5/6 (ES) 204 D0 0–20 W Ri 2,5 204, 205, 206 2.5YR 5/8 (E/I/C) 5YR 6/6 (ES) 205 D0 0–20 W Ne+Ha - 204, 205, 206 2.5YR 5/8 (E/I/C) 5YR 6/6 (ES) 206 E3 0–20 W Ha - 204, 205, 206 2.5YR 5/8 (E/I/C) 5YR 6/6 (ES) 51 207 D0 0–20 W Ha - - 2.5YR 7/8 (E/I/C) 2.5YR 5/6 (ES) 208 E3 0–20 W Ha - - 5YR 5/8 (E/I/C) 5YR 7/6 (ES) 209 D0 0–20 W Ha - - 5YR 6/8 (E/I) 5YR 5/8 (C) 5YR 6/8 (ES) 210 D0 0–20 W Ha - - 5YR 5/8 (E/I/C) 5YR 6/6 (ES) 211 D0 0–20 W Bo - - 5YR 5/8 (E/I/C) 2.5YR 5/6 (ES) 212 D0 0–20 W Bo - - 5YR 7/6 (E/I/C) Armvessels 213 E3 0–20 W Ri - - 7.5YR 6/6 /E/I/C) Too worn (ES/IS) Spindle Bottles 214 D3 20–40 W Rb 4 - 5YR 7/6 (E/I) 5YR 6/6 (C) Red-on-Black (RoB) Carinated Bowls 215 B3 20–40 W? Ri 18 215, 216 7.5YR 5/4 (E/I/C) GLEY1 2.5/ (ES/IS) 2.5YR 5/3 (ED/ID) 216 F5 1stFL W? Ri 18 215, 216 Too thin (E/I) 10YR 4/4 (C) GLEY1 3/ (ES/IS) Too worn (ED/ID) Bowls 217 F5 1stFL H Ri c. 24 - 10YR 6/4 (E/I) Too thin (C) 7.5YR 4/1 (ES) 5YR 4/2 (IS) 7.5YR 8/3 (ED) 7.5YR 8/2 (ID) 218 D0 0–20 H Ri 28 - 10YR 8/3 (E/I) 10YR 6/6 (C) GLEY1 3/ (ES/IS) 5YR 5/6 (ED)

ID Grid Depth H/W Frag. Ø Joins Munsell 219 D5 60–80 H Ri 12 - 10YR 6/4 (E/I/C) GLEY 13/ (ES) 5YR 4/4 (IS) Too worn (ED) 2.5YR 4/3 (ID) 220 E5 - H Ri 34 - 7.5YR 6/6 /E/I/C) 5YR 5/4 (ES) GLEY1 3/ (IS) Too worn (ED) 7.5YR 8/2 (ID) 221 E5 - H Ri 20 - 10YR 7/4 (E/I) 10YR 7/3 (C) GLEY1 3/ (ES/IS) 2.5YR 5/6 (ED/ID) 222 D3 20–40 H Ri - - 10YR 7/4 (E/I) 7.5YR 7/4 (C) 2.5YR 4/4 (ES) 7.5YR 4/1 (IS) 7.5YR 7/2 (ED/ID) 223 B3 20–40 H Ri c. 28 - 7.5YR 6/6 (E/I/C) 7.5YR 4/ (ES) 5YR 2.5/1 (IS) 2.5YR 5/4 (ED/ID) 224 F5 1stFL H Ri - - 10YR 7/4 (E/I/C) GLEY1 4/ (ES) 2.5YR 4/1 /(IS) 10YR 8/3 (ID) 225 B3 20–40 H Ri c. 12 - 7.5YR 6/6 (E/I/C) 2.5YR 4/1 (ES) 5YR 4/1 (IS) 2.5YR 5/4 (ID) 226 D5 60–80 H Ri+Ha - - 10YR 5/4 (E/I/C) GLEY1 3/ (ES) 5YR 4/2 (IS) 2.5YR 4/3 (ED/ID) 227 B3 20–40 H Ri c. 28 - 10YR 8/3 (E/I) 7.5YR 7/4 (C) Too worn (ES) 7.5YR 4/ (IS) Too worn (ED) 5YR 6/4 (D) 228 F5 1stFL H Ri 30 - 7.5YR 6/4 (E/I/C) 7.5YR 4/1 (ES) GLEY1 4/ (IS) 7.5YR 7/4 (ED) 7.5YR 7/6 (ID) 229 D5 60–80 H Ri+Ha - - 7.5YR 5/6 (E/I/C) GLEY1 4/ (ES) GLEY1 3/ (IS) Too worn (ED/ID) Spouted bowls 230 F5 1stFL H Ri+Sp 32 - 7.5YR 6/4 (E/I/C) 2.5YR 4/1 (ES) GLEY1 4/ (IS) 5YR 6/4 (ED) 2.5YR 6/4 (ID) Jugs 231 F5 1stFL H Ri 5 - 7.5YR 6/6 (E) 10YR 6/4 (I) 7.5YR 4/1 (ES/IS) 10YR 8/3 (ED) 52 232 D5 60–80 H Ri 8 - 10YR 8/3 (E/I) 10YR 6/4 (C) Too worn (ES/ED) 233 B3 20–40 H Ri+Ha c. 14 - 7.5YR 5/6 (E/I/C) 7.5YR 4/1 (ES) 2.5YR 5/4 (ED) 234 B3 20–40 H Ri+Ha - - 10YR 6/4 (E/I) 7.5YR 6/4 (C) GLEY1 3/ (ES/IS) Too worn (ED) Red Polished (RP) 235 D5 60–80 H Fb 6 - 10YR 6/4 (E/I) 2.5Y 6/2 (C) 10YR 6/6 (ES/IS) Red Slip (RS) Bowls 236 A5 40–100 W Ri c. 12 - 7.5YR 6/6 (E/I/C) 10R 5/6 (ES) 2.5YR 5/8 (IS) 237 D3 20–40 H Ri 12 - 7.5YR 7/4 (E/I/C) 2.5YR 4/4 (ES) 5YR 5/4 (IS) 238 D0 0–20 H Ri - - 5YR 5/8 (E/I) 7.5YR 5/6 (C) 2.5YR 4/3 (ES) 5YR 3/1 (IS) 239 E3 0–20 H Ri c. 20 - 7.5YR 6/6 (E/I/C) 5YR 5/4 (ES) GLEY1 3/ (IS) 240 B3 20–40 H Ri c. 30 - 7.5YR 5/6 (E/I/C) 5YR 5/6 (ES/IS) 241 E5 - H Ri 9 - 10YR 6/4 (E/I/C) 7.5YR 5/6 (ES) 2.5YR 5/6 (IS) 242 E5 - H Ri - - 10YR 8/3 (E/I) 7.5YR 7/4 (C) Too worn (ES) 7.5YR 5/6 (IS) 243 B5 0–20 W Ri 24 - 5YR 6/6 (E/I/C) 5YR 5/6 (ES) 2.5YR 4/3 (IS) 244 E3 0–20 W? Ri 12 244, 245 7.5YR 6/4 (E/I/C) 5YR 5/4 (ES) 10R 5/6 (IS) 245 A5 0–20 W? Ri 12 244, 245 7.5YR 6/4 (E/I/C) 5YR 5/4 (ES) 10R 5/6 (IS)

ID Grid Depth H/W Frag. Ø Joins Munsell 246 B5 20–40 W? Ri c. 22 - 7.5YR 6/6 (E/I/C) 2.5YR 5/8 (ES) 5YR 5/6 (IS) 247 F5 1stFL H Ri - - 7.5YR 6/6 (E/I) 10YR 6/4 (C) 2.5YR 5/6 (ES) 7.5YR 5/6 (IS) 248 D3 20–40 H Ri - - 7.5YR 5/6 (E/I/C) 5YR 5/6 (ES) 2.5YR 5/6 (IS) 249 F5 1stFL H Ri 30 249, 250 5YR 5/6 (E/I/C) 5YR 5/4 (ES/IS) 250 F5 1stFL H Ri 30 249, 250 5YR 5/6 (E/I/C) 5YR 5/4 (ES/IS) 251 B3 20–40 H Ri 26 - 7.5YR 6/6 (E/I/C) 10R 5/6 (ES) Too worn (IS) 252 B3 20–40 H Ri - - 7.5YR 6/6 /E/I/C) 10R 5/6 (ES) 5YR 7/6 (IS) 253 B5 20–40 H Ri c. 24 - 10YR 6/6 (E/I) 7.5YR 6/4 (C) 2.5YR 5/4 (ES) 2.5YR 4/6 (IS) 254 A5 0–20 H Ri - - 7.5YR 6/6 (E/I/C) Too worn (ES) 5YR 6/8 (IS) 255 D0 0–20 W? Ri c. 22 - 10YR 5/4 (E/I) 10YR 5/2 (C) 2.5YR 4/8 (ES) 2.5YR 5/4 (IS) 256 D0 0–20 H Ri c. 20-30 256, 257 7.5YR 6/6 (E/I/C) 5YR 6/6 (ES) Too worn (IS) 257 D0 0–20 H Ri c. 20-30 256, 257 7.5YR 6/6 (E/I/C) 5YR 6/6 (ES) Too worn (IS) 258 F5 1stFL H Ri - - 5YR 6/6 (E/I/C) Too worn (ES) 5YR 6/6 (IS) Spouted Bowls 259 B5/E5 20–40 H Ri+Sp c. 28-30 259, 260 7.5YR 6/4 (E/I/C) 5YR 6/8 (ES) 5YR 5/6 (IS) 53 260 B3 20–40 H Ri c. 28-30 259, 260 7.5YR 6/4 (E/I/C) Too worn (ES) 5YR 5/6 (IS) 261 B3 20–40 H Ri+Sp - - 7.5YR 6/6 (E/I/C) 5YR 5/6 (ES) 10R 5/6 (IS) 262 D5 60–80 H Ri+Sp - - 7.5YR 4/4 (E/I/C) 2.5YR 4/6 (ES) 10R 5/6 (IS) 263 D0 0–20 H Sp - - 7.5YR 6/6 (E/I) 10YR 6/4 (C) Too worn (ES) 2.5YR 4/6 (IS) 264 B3 20–40 H Sp - - 7.5YR 6/6 (E/I/C) 7.5YR 5/4 (ES) 2.5YR 5/6 (IS) Jugs 265 E5 40–60 H Ri 9 - 10YR 8/3 (E) 2.5Y 8/2 (I) 10YR 6/6 (C) Too worn (ES) 5YR 4/2 (IS) Unknown 266 E3 0–20 W Bo - - 7.5YR 6/6 (E) 10YR 6/3 (I) 7.5YR 7/4 (ES) White Painted Wheelmade (WPWM) 267 E3 0–20 W Ri 12 - 10YR 7/4 (E/I) 7.5YR 6/4 (ED) White Porous (WPo) Piriform jar 268 E3 0–20 W Ri 10 267, 268 10 YR 8/3 (E/I/C) 269 E3 0–20 W Db 5 267, 268 10 YR 8/3 (E/I/C)

ID Grid Depth H/W Frag. Ø Joins Munsell White Slip (WS) Bowls 270 D0 0–20 H Ri 20 270, 271, 272 10YR 5/3 (E) GLEY 1 6/1 (I/C) 2.5Y 7/4 (ES/IS) 10YR 4/3 (ED) 271 D3 20–40 H Ri+Ha 20 270, 271, 272 10YR 5/3 (E) GLEY 1 6/1 (I/C) 2.5Y 7/4 (ES/IS) 10YR 4/3 (ED) 272 D0 0–20 H Ri 20 270, 271, 272 10YR 5/3 (E) GLEY1 6/1 (I/C) 2.5Y 7/4 (ES/IS) 10YR 4/3 (ED) 273 D0 0–20 H Ri 17 - 10YR 4/3 (E) 2.5Y 5/1 (I) 2.5Y 7/4 (ES) 2.5Y 6/8 (IS) 2.5Y 4/1 (ED) 274 D3 20–40 H Ri 18 - 10YR 3/2 (E) GLEY1 7/ (I) 10YR 6/4 (C) 2.5Y 7/4 (ES/IS) 10YR 7/3 (ED) 275 D0 0–20 H Ri c. 24 - 10YR 4/2 (E) GLEY 1 5/1 (I) 2.5Y 7/4 (ES/IS) 10YR 3/1 (ED) 276 D0 0–20 H Ri - - Too thin (E) GLEY 6/1 (I/C) 10YR 8/3 (ES/IS) 10YR 4/1 (ED) 277 E3 0–20 H Ha - - 10YR 4/2 (E/I/C) 10YR 7/4 (ES/IS) 10YR 3/1 (ED)

54

Figure 13. Ratio of MCIII/LCI wares. Figure 14. Open/closed ratio of MCIII/LCI wares.

55

Figure 15. Ratio of LCII wares. Figure 16. Open/closed ratio of LCII wares

Table 4. Horizontal distribution of sherds. Grid No Grid Ware A3 A5 A6 B1 B3 B5 B5/E5 D0 D3 D4 D5 D6 E2 E3 E4 E5 F5 Total MCIII/LCI BS 1 4 1 4 15 8 5 8 1 1 4 1 96 19 168 RS 1 14 6 3 26 9 1 12 8 4 7 1 71 11 174 PWHM 1 7 22 11 16 37 7 9 2 4 96 32 244 RoB 1 7 7 2 9 2 53 31 112 WPHM 1 1 2 LCII BR 1 8 34 20 27 1 91 Cypr. 12 1 44 93 1 7 1 152 5 1 10 327 Myc. GW 1 1 1 3 PWWM 44 106 27 77 14 77 82 112 32 3 135 11 70 61 20 1 872 RL 1 1 2 39 88 1 2 246 5 139 524 WPWM 1 1 WPo 12 12

56 WS 16 24 5 45 Total 46 140 27 77 15 95 36 248 371 57 70 180 28 525 77 338 95 150 2575

Table 5. Vertical distribution of sherds. Depth MCIII/LCI LCII BS PWHM RoB RS WPHM BR Cypr. Myc. GW PWWM RL WPWM WPo WS Total 0–20 25 32 10 50 34 213 382 293 12 19 1070 20–40 61 71 44 60 1 55 101 2 431 92 26 944 1stFL 19 32 31 1 1 84 40–60 8 15 8 1 32 60–80 8 37 9 12 1 3 70 40–100 1 3 1 2 55 62 No Depth 46 57 18 41 1 10 139 1 313 Total 168 244 112 174 2 91 327 3 872 524 1 12 45 2575

5. Re-contextualizing Ayios Iakovos

5.1 Previous interpretations of Ayios Iakovos: Dhima in a new light

5.1.1 Dhima as a dual deity site The interpretation of Ayios Iakovos: Dhima by the Swedish Cyprus Expedition can be separated into two accounts. The first is the description by Gjerstad in his book Sekler och dagar, published in 1933.207 This book was a popular account of the expedition to Cyprus and its four-year journey. The text contains the first record of the excavation and interpretation of the Dhima Sanctuary. 208 The second account is the official publication of the site, written by Sjöqvist, in 1934 and his dissertation Problem of the Late Cypriot Bronze Age, published in 1940.209 It is in these three texts that most of the later research is grounded. When Ayios Iakovos: Dhima was first excavated it was immediately interpreted as a sanctuary. Sjöqvist notes that the “type of architecture, the rich finds, and above all the presence of the incense-burners, give satisfactory evidence in the respect.”210 Presumably, the architecture is referring to the two stone platforms, which were interpreted as altars. 211 Sjöqvist further suggests that they represent a dual divinity cult at Dhima.212 While admitting that the precise nature of the deities cannot be determined, he suggested a mother goddess and consort, similar to Magna Mater–Attis, Ishtar–Tammuz and Isis–Osiris cults.213 Due to the destruction by farming and the absence of finds, the purpose of the platforms is problematic to understand. In his description of the excavation of Dhima, Gjerstad notes that the terracotta basin in the western part of the sanctuary was overflowing with burned animal bones and ashes, which he proposed was used as an offering pit.214 Sjöqvist, however, does not mention that the basin contained either bones or ashes. Sjöqvist does note that there was ash in the MCIII-pit. It seems likely that Gjerstad was exaggerating in this case. Sjöqvist also discussed the spatiality of the site. He argues that the entrance must have been from the northwest since steep slopes limited the southern and western sides. The eastern side was considered the “inner sanctum” and he believed it was unlikely to have served as an entrance to the sanctuary. Access to the eastern side was instead achieved on the south side of the sanctuary, where the wall ended. The large platform would have blocked access on the north side.215 With this flow of movement established, Sjöqvist suggested that the worshippers would first access the

207 Gjerstad 1933. 208 Gjerstad 1933, 96–102. 209 Gjerstad et al. 1934a, 355–361; Sjöqvist 1940, 2–4. 210 Gjerstad et al. 1934a, 359. 211 Gjerstad et al. 1934a, 359; Sjöqvist 1940, 3. 212 Sjöqvist 1940, 3. 213 Sjöqvist 1940, 3. 214 Gjerstad 1933, 99. 215 Gjerstad et al. 1934a, 359. See also Fig. 2.

57 terracotta basin, deposit the votive gifts and then proceed to the inner sanctum. While later researchers have had no objections to the movement pattern, it is based on the assumption that the eastern side is private or restricted, as opposed to a more public western side. Sjöqvist originally suggested the idea that Dhima was fenced off with perishable material.216 Later studies have not disputed this interpretation, despite the absence of physical evidence. Sjöqvist refers to cylinder seals found in Cyprus and argues that since there are pictographic representations of cultic scenes with details of an enclosure, such a feature must then have existed at Dhima as well. However, the enclosure he describes, a fence of poles and twigs with two posts serving as a gate would probably have left some traces of postholes in the ground.217 Lastly, the date of the sanctuary was based on two main find-groups: The pit, which only contained sherds from the MCIII-period, sealed by the floor, and the finds excavated above floor level, mostly around the terracotta basin. All finds and pottery found on the floor were dated to the LCII period, and Sjöqvist notes that nothing could be securely dated to the LCI period, and therefore Dhima was not in use during this period.218 In this re-examination of Dhima, some of these conclusions must be questioned. The terracotta basin. While later studies have referenced Gjerstad’s record of the overfilled terracotta basin, it must be stressed that Sjöqvist never mentioned this in either of his discussions on Dhima. The drawings done by Westholm do not support Gjerstad either. On the original section of the basin (see Profile I) there are three distinct layers in the basin. While there is no description of these layers, the top two share a similar legend and appear to be accumulation layers. The bottom layer shares the same legend as the charcoal layer in the MCIII-pit. If anything, it seems like there was only ash at the very bottom of the basin. It would seem like the basin was either cleaned out or only contained a small accumulation of material by the time the layers above were formed. This is noteworthy because faunal remains are common in other Cypriot sanctuaries and usually linked to both animal sacrifice as well as post- sacrificial consumption.219 The platforms. The fact that no finds were made on or around the platforms is explained by disturbance of agricultural activity.220 While possible, it seems unlikely that all traces of objects placed on top of the platforms would have completely disappeared. This study has concluded that while there were no small finds on the eastern side, there was a presence of LCII pottery. Although few, there is also evidence of Cypr. Myc. piriform jars (27 sherds in total, including 77, 82 and 99– 100), BR cups (21 sherds in total, including 10–11, 14, 21–25), PWWM (135 sherds in total, including 164 and 166) and RL (ten sherds in total, including 198) wares. The majority of these were found in grid D6, outside the floor area. The original drawing of the section (see Profile I) indicates a small layer of accumulated soil, abutting the smaller platform at the western edge of grid D6. This layer could be an indication of LCII activities close to one of the platforms and is the only concentration of clear LCII pottery on the eastern side of the wall. There is also a concentration of PWWM (77 sherds in total, including 159, 168 and 170) pottery in Grid B5, presumably in the east part of the grid. Grid D0 is just outside the floor area and it is one of the richest grids in terms of pottery. The material consists chiefly of BR (eight sherds, including 2, 9, 15 and 26), Cypr. Myc. (44 sherds, including 39–42, 45, 47, 51, 54, 61–62, 73, 79, 95, 97 and 106–107), PWWM (82 sherds, including 155, 157, 172, 160, 180,

216 Gjerstad et al. 1934a, 359. 217 Sjöqvist 1940, 3. 218 Gjerstad et al. 1934a, 361. 219 Webb 1999, 252. 220 Gjerstad et al. 1934a, 359–360.

58 184), RL (39 sherds, including 204–205, 207 and 209–212) and WS (16 sherds, including 270, 272–273 and 275–276). Grid B5, D6 and D0 are all outside the floor area, but still contain concentration of LCII pottery. These three Grids could indicate a habit to regularly cleaning the floor, so that sherd material was swept away to the edge of the sanctuary. If this was the case, these are the only faint remains of the activities performed around the platforms, in contrast to other Cypriot sanctuaries where finds can be closely tied to the altars.221 Chronology. While there are no complete vessels that date to the LCI period, there is enough evidence in the sherd material to suggest that there was some activity at Dhima during the LCI period. In this study, several sherds of RoB as well as handmade RS and BS have been found at stratigraphic levels, which are clearly above the floor level (the hard separator of the MCIII and LCII period at Dhima). These layers are either marked as 0–20cm (indicating depth) or as “on the first floor.” Since the 0–20cm-layer roughly corresponds with the plough layer it is no doubt mixed. However, "on the first floor" would seem to indicate just that, excavated above the floor layer. Around a third of all RoB sherds, 41 sherds, have been recorded as either “on the first floor” (31 studied sherds, including 216–217, 224, 228 and 230–231) or “0–20cm” (10 studied sherds, including 218). Similarly, nine RS handmade sherds (247, 249–250, 254, 256–258, 263 and 266) and three BS handmade sherds (30, 32– 33) have been found in these two layers. There are also RoB, RS and BS handmade sherds found within the depth of "20–40cm". However, there is the possibility that sherds found at this depth can be from both below and above the floor since both LCII sherds and partial contents of the MCIII-pit are recorded within this layer. The RoB, RS and BS handmade wares all disappear during the end of the LCI period. The fact that there is a clear presence of these wares above the floor level means that the dating of the floor should be pushed back to the LCI period. Sjöqvist concludes that Dhima must have been of some importance, due to its rich finds, but admits that there are difficulties interpreting the place, due to the limited architectural remains. 222 While there is little evidence that disputes the possibility of a dual deity at Dhima, there is also little confirming it, beyond the two platforms. Some of the original interpretations by Sjöqvist are reasonable, while others, like the chronology, can be questioned by the MCIII/LCI sherds found above floor level. Gjerstad’s account is more problematic and suffers from a literary flare to excite the reader, rather than describing the sanctuary accurately. While surely this is innocent, it has created problems in modern research, since later studies have taken Gjerstad’s claims as facts.

5.1.2 Dhima as a typical Cypriot Sanctuary In 1992 George Wright discussed the similarities between Cypriot, Levantine and Aegean rural sanctuaries. His general summary of Cypriot rural sanctuaries characterises them as closely connected with nature. As such, rather than sanctifying the land with the sanctuary, the sanctuary was placed on an already sacred place. Similarly, he argues that a sacred tree, and possibly a sacred stone, represented the connection with nature. He also notes that the altar was the focal point of any cultic activity. Additionally, he argues that while often placed on hilltops, Cypriot rural sanctuaries are never found on mountain peaks, such as sanctuaries found on Crete.223 Lastly, Wright argues that Cypriot rural sanctuaries are characteristically not found close to major settlement centres, but rather spread on the countryside.224

221 Compare with Myrtou-Pighades. 222 Gjerstad et al. 1934a, 3–4. 223 Wright 1992, 275–276. 224 Wright 1992, 275.

59 Wright applies this version of a Cypriot rural sanctuary to Dhima by claiming that it “manifests the essential elements of enclosed sacred space in natural surrounding.”225 While doing so, he agrees with Sjöqvist that Dhima was enclosed. He also adds that the sanctuary had a sacred tree and that one of the platforms served as a base for the deity’s symbol, while the other was the main altar. He explains the terracotta basin as an eschara, a sunken trough for ctonic rites and that the wall divided the space into a public and a sacred area. 226 Wright makes no comment regarding the finds at Dhima; his focus is on the landscape, architecture and the presence of a sacred tree and stone. Given the above description, Wright argues that Dhima is in no way special; it is a rather typical sanctuary. Using his model of a Cypriot sanctuary, Wright argues that this type of rural, hilltop sanctuary is comparable to the Levantine bamah, or “high place,” as described in the Bible.227 A bamah is explained as a hilltop sanctuary, containing a central altar, a tree and an aniconic stone.228 Wright concludes that this is directly comparable to the situation in Cyprus.229 Similarly, he argues that the same type of sanctuary, found in a rural setting with a sacred tree, was the norm in the Aegean.230 Wrights interpretation is problematic regarding Dhima and its connection to the Levant and the Aegean. His claim that there were a tree, an enclosure and a sacred stone on one of the platforms is not backed by the archaeological record, and seems more to be what Wright supposes should be there. At other rural sanctuaries in Cyprus, such as Ayia Irini, there is at least some archaeological evidence for an enclosure, the presence of a tree or a sacred stone.231 However, no such evidence exists at Dhima. While Wrights assumptions of Dhima are problematic, it is still possible to look at his conclusions that are supported by the archaeological material. Dhima was not a peak sanctuary; it was not even placed on the highest hill in the area.232 So Wright is perhaps right in his interpretation that this trait was not as important for the location of the sanctuary. Similarly, if the construction of a sanctuary made the land sacred, it would have been more practical to place the sanctuary close to a settlement. As it is now, Dhima is located on its own, far away from any known settlement. Therefore, part of Wrights argument is plausible; the landscape and nature were sacred and the place was chosen for this reason. This offers not only explanation as to why Dhima is found in such a remote place, but also an insight into the religious concepts of the late Middle and early Late Cypriot period. Furthermore, the argument that could be made is that Dhima bears some resemblance to the Bamah. According to Wright, a central part of a Bamah would be the altar; it is even translated as “a sanctuary centred around a sacred platform.”233 If Dhima were a Bamah, it would mean that the platform(s) were the focal point of the site. The few finds in connection with the platforms may speak against this interpretation. However, even if Levantine sites inspired a rural sanctuary like Dhima, it does not mean that the full concept of the Bamah was included in Cyprus. It could be a partial import, similar to how other cultural influences have been imported to

225 Wright 1992, 269. 226 Wright 1992, 269. The use of eschara rather than larnax for the basin is unusual, and perhaps an indication by Wrights of a cultic function, rather than a functional similarity to a bathtub. The basin is usually referred to as a larnax. 227 Wright 1992, 277–278. 228 Wright 1992, 279. Bamah is suggested to mean hilltops, high place, altar or sanctuary with a central altar. 229 Wright 1992, 279. 230 Wright 1992, 279–281. 231 See Gjerstad et al. 1934b, 642-824. 232 See title page. 233 Wright 1992, 278.

60 Cyprus (such as a suggested introduction of feasting and the Plain White wares). However, there is little evidence to back this up beyond speculation. What can be said is that the only thing in common between Dhima and a Bamah is that it was located on a hilltop. Wrights classification of Dhima as both as a typical rural Cypriot sanctuary and as an eastern Mediterranean sanctuary is therefore problematic. His arguments are based on speculation and not reflected in the material, especially since he does not discuss the material culture. However, his interpretation that the land itself is sacred and that the location for Dhima or any other Cypriot sanctuary was chosen because of it, offers a new perspective on how to approach the geographic location of sanctuaries in Cyprus. No doubt this changes during the later parts of the LC period, when the intra-mural sanctuaries appear at the larger coastal centres such as Enkomi and Kition.

5.1.3 Dhima as a short-lived sanctuary and funerary site Jennifer Webb conducted one of the more thorough studies of Dhima in 1999.234 In her general study of ritual architecture in Cyprus, she includes Dhima due to its interpretation as a sanctuary. Unlike the studies done by Collard and Hitchcock, this is a more general survey. Webb identifies Dhima as a short-lived sanctuary, where the constructions, activities, and abandonment happened shortly after one another. She also argues that the sanctuary was connected to funerary rites. Webb argues that the minimal efforts of the constructions would explain what she calls a “discrepancy” between the poor architecture and the rich finds.235 She notes that the preservation of the finds and their apparent in situ placement is not found in any other Cypriot sanctuary. Other sites are usually of long, continuous use and with evidence of destruction, reconstruction and removal of the votives. This does not appear to be the case at Dhima. Instead, Webb argues that the finds and their in situ preservation indicate a deliberate destruction of the site, as well as a “single event”-ritual. Because of the well-preserved objects, she thinks that the site was not only abandoned after its use but also covered.236 While Webb is correct in the fact that most published objects are remarkably well preserved, it seems like quite an undertaking to first create a flat (albeit slanting) floor, construct one very large rubble platform and a second smaller one, a rubble wall, and a terracotta basin, and two pits, all founded in cut bedrock. And then, after this single use, cover the entire place, to preserve the objects. While not as monumental as other cultic sites in Cyprus, it is still a substantial undertaking to make for a single occasion. This event might have taken place during several days, but nonetheless, the architectural effort is perhaps not as minimal as suggested. It should also be noted that the chronology and relationship between the floor, the wall and the platforms are only partially clear. Sjöqvist originally concluded that the architecture is contemporary with the floor and that only the pit containing the MCIII sherds predates the floor.237 The only stratigraphic comment in the original publication is that the floor runs up against the wall and the platforms, indicating that the walls were built first in the construction process. It is possible that the architecture was constructed either long before the floor, or directly prior. In either case, as discussed above, the dating of the floor should be pushed back to the LCI period, due to the presence of LCI pottery found above floor level. The interpretation of Dhima as a single event site is also problematic considering some of the photographs taken at the excavation. Figs. 8 and 10 show that

234 Webb 1999. 235 Webb 1999, 33. 236 Webb 1999, 34. 237 Gjerstad et al. 1934a, 356, 359–360.

61 several of the vessels found around the basin were placed at different heights, possibly suggesting deposition at different times. Additionally, the amount of pottery studied in this study gives witness to the fact that, while an ample amount of finds were preserved, there is a sizeable number of broken vessels spread around the site. As previously noted, the pottery in the Mediterranean museum’s storage is most likely not the entire assemblage, pottery is missing from several grids, and it seems unlikely that sherds would not be present in some form all over the area, especially considering the agricultural activities in the topmost layer. As such, the broken pottery can be explained in two ways: 1) There was no hiatus at Dhima, and the broken pottery simply remains from a longer period of activities. 2) All pottery is contemporary from a single event, and the sherds originate from vessels broken during the covering of the site. Of these possibilities, the first seems the most likely scenario. It is supported by the presence of pottery dated to the MCIII/LCI period found intermixed with the LCII pottery and the photographs of the finds around the terracotta basin from the excavation. The second possibility does not explain the pottery found close to the floor, nor why some, and not other, vessels were so pristinely preserved. Likewise, the fact that sherds 30, 32–33, 213, 216–217, 224, 228, 230–231, 247, 249–250, 254, 256–258, 263 and 266 all date to a period before LCII and after the construction of the floor, this cannot be the case. Therefore, it is unlikely that Dhima was merely constructed to serve for a single occasion and the MCIII/LCI pottery found above the floor indicates a longer continuity at Dhima before the LCII period. A prolonged use of the sanctuary is further reinforced by the presence of both LHIIIA1 (38, 46, 78, 84–85, 90, 99) and LHIIIA2 (41, 53, 59, 86–87, 108–109) vessels in the assemblage. Additionally, if the complete krater (No. 44) is re-dated to LHIIIA1, it indicates a continuous use of the vessel until the abandonment of the sanctuary. Lastly, the pottery found in Grids B5, D0 and D6 could be explained as evidence of regular cleaning of the sanctuary. Especially considering the many joining sherds between the central Grids (D3 and E3) and the grids outside the floor area (B5, D0 and D6). If these sherds are the remains of cleaning, it suggests that Dhima was in continuous use during the LCI and early LCII period.

Dhima as a funerary site Beyond being a single event site, Webb suggests that Dhima was a sanctuary connected with funerary rites. This is partly based on the connection between the finds and their function in other Cypriot contexts. Webb argues for the parallels between finds from Dhima and other burial deposits in Cyprus, most notably in regards to the numerous RL arm vessels.238 As mentioned in this study, the possibility of up to 10–11 different arm vessels at Dhima is unparalleled anywhere else.239 The fact that they are primarily found in funerary deposits does support a funerary connection at Dhima. Additionally, Webb compares the many silver funnels located at Dhima to gold funnels found in burials across Cyprus.240 However, funnels are a rarity and only occasionally appear among the burial gifts.241 Other leading arguments are the proximity to burials in the area around Dhima and similarities between the terracotta basin and the rock-cut pits filled with ash and bones found at Korovia Palaeskoutella. 242 However, given Gjerstad’s dubious account of the terracotta basin at Dhima, this parallel should be approached

238 Webb 1999, 34. 239 See Eriksson 1993 for an overview of contexts containing RL arm vessels. 240 Webb 1999, 33. 241 See Keswani 2004, tables 5.8–5.13. 242 Gjerstad et al. 1934a, 419, 423, 429; Webb 1999, 34.

62 cautiously. It seems unlikely that there is enough evidence to support a clear parallel between the basin at Dhima and the pits at Palaeskoutella. As for the proximity to the burials at Ayios Iakovos: Melia, Ayios Iakovos, and Kakotri, it is true that they are in the general area, c. 1km to the east. Interestingly enough, Collard uses the same argument of geographical placement to deduce that Dhima is not a funerary site, considering it to be too far away from the burials.243 This shows that the geographical distance and connection between two sites can be differently interpreted, depending on what is considered close and far away. In either case, both Melia and Kakotri are more than one kilometre to the east of Dhima and obscured from sight. The tombs at Ayios Iakovos are closer, halfway between Dhima and Kakotri, and could have had a connection with Dhima. But without any direct link between Dhima and the burials, any connection is, in the end, speculative. Likewise, there are no clear parallels between the burial gifts at either site and Dhima.244 In regards to the two platforms, Webb groups them together with her “Stone Tables.” She compares them to similar tables found in Kition, and Ayia Irini. She notes that such installations have frequently been found with concentrations of vessels and other objects in their vicinity, and were primarily used as repositories for food and drink offerings, votives and other cult equipment.245 There is little evidence for this at Dhima since only sporadic sherds were found close to the platforms. There is the small concentration of LCII pottery in Grid D6, as noted above. While this group is possibly connected to the smaller platform, it seems more likely that the platforms at Dhima did not serve the same purpose as stone tables at other sites. All in all, Webb concludes that Dhima is a one-of-a-kind archaeological site rather than a typical example of an open air, rural sanctuary. She labels Dhima as a “specialized extramural cult locale, possibly associated with mortuary ceremonial (Ayios Iakovos: Dhima is the only example and may be unique).”246 As discussed, this interpretation is somewhat problematic, both in regards to representing a singular event as well as the connection to a funerary cult.

5.1.4 Dhima as a last resort during times of distress David Collard offers one of the more recent and more thorough studies of Ayios Iakovos: Dhima.247 He does this through a rather unusual ”narrative” approach, in which he incorporates the site, its architecture, the archaeological material and the surrounding landscape in an attempt to recreate the activities and rituals that took place at Dhima. The initial presentation of both the site and the artefacts is rather comprehensive and is primarily based on the original publication and Webb’s study from 1999. Collard also puts more attention to the plans, sections and photographs published in the original report.248 While he agrees with previous interpretations that this is a cultic site,249 he disputes the interpretation that the site was used for funerary rituals, mainly due to the absence of a contemporary burial near the site and the differences in material culture. To Collard, the burials at Ayios Iakovos, Kakotri and Melia are too far away. He does, however, admit the possibility of funerary rituals in a non-funerary setting.250 Instead, he interprets the spear butts and arrowheads as

243 Collard 2008, 33–34. 244 Eriksson 1993, 258–261. No arm vessels, and few sherds of RL ware at all, found in any of the nearby burial areas. There are some Mycenaean vessels in the burial assemblages (See Gjerstad et al 1934, 302–355), but not the distinct pattern towards piriform jars found at Dhima. 245 Webb 1999, 171. 246 Webb 1999, 296–297. 247 Collard 2008, 28–40. 248 Collard 2008, 28–31. 249 See Gjerstad et al. 1934a, Wright 1992, Webb 1999. 250 Collard 2008, 33–34.

63 evidence of a ritual that centred “on hunting, warfare or initiation of males.”251 He does accept Webb’s interpretation that the LCII activities at Dhima were from a single event, and uses this to support his theory that Dhima was used during the LCII period as a desperate response to an unstable period. He argues that the mass burials in some of the tombs of Melia, dating to the LCI period, are indicative of a time of distress in the area around Ayios Iakovos. The rich finds at Dhima do therefore reflect the inhabitant’s desperation to accept parting with such valuable objects as offerings.252 Collard’s study is also the first to suggest a different chronology of the architecture at Dhima. Unlike Webb and Sjöqvist, who both see the architecture as a pure LCII construction, Collard believes it to be remains of an earlier cult. The architectural remains, the platforms and the wall, are seen as an ancestral link.253 He argues that while the architecture may have been from the earlier phases of the site, they seem to be incorporated into the LCII event. In his narrative approach, Collards suggests a procession rite, leading up the hill to Dhima, which at the time was enclosed with a wooden fence. Like previous interpretations, he argues that the site was entered from the northwest. Given the lack of finds near the platforms, Collard suggests that the eastern area was not the focus of the ritual; it rather acted as a reminder and an ancestral bond to the sanctuary. Instead, the terracotta basin was the centrepiece of the cult. A person would have been seated inside the basin, being anointed by the others with the contents of the pottery found around the basin. At some point, personal items, such as the weapons and jewellery, would have been deposited inside and around the basin as well. He explains the placement of the RL arm vessels at the edge of the sanctuary as a way to dispose of them after usage. Collard argues that the wall was part of an earlier phase and adopts earlier interpretations that it was an important border between the “outer” sanctuary on the west side and the “inner” sanctuary on the east. The ritual in the basin served as a purification ritual before entering the eastern part of the site. 254 After the anointment and procession into the eastern half, Collard suggests that a sacrifice of animals and a feast took place, based on Gjerstad’s record that the basin was filled with ash and bones, in addition to the pottery connected with feasting found around the basin (such as the krater, No. 40). 255 Collard also claims that the site was deliberately destroyed after the ceremony by fire, which he explains with the thin, even dark layer directly above the plaster floor of the place.256 This would have been done to preserve the rich votive gifts deposited at the site. In conclusion, Collard argues that the Dhima site was a place for male initiation during a time of distress in the area. It centred on an anointment ritual connected to an ancestor cult and feasting activity where the terracotta basin served as a gathering point both before and after rituals in the eastern part took place.257 He offers a thorough and vivid record of the archaeological finds at Dhima. Some of his interpretations, such as the chronology of the architecture, are not discussed elsewhere. However, he does draw some puzzling conclusions from the original publication. Most strikingly is his interpretation that the site was burned after the ritual. Looking at the same section drawings, and the original drawings (Profile 1–III), there is nothing to suggest a destruction layer covering the site. The only reasonable explanation is that Collard misinterprets the dark outline of the floor as a destruction layer. Likewise, the relation between the floor and the wall is problematic. The

251 Collard 2008, 34. 252 Collard 2008, 34–35. 253 Collard 2008, 31. 254 Collard 2008, 35–37. 255 Collard 2008, 37–38. 256 Collard 2008, 38. 257 Collard 2008, 38–39.

64 sections (both published and in original, Fig. 6, Profile I–III) give the impression that the wall was almost covered by the floor, while the photographs (see Figs. 3 and 7) indicated that the wall is about the same height as the larger platform, and most likely visible during the LCII period. Collard does change his opinion in his narrative interpretations and includes the wall as a boundary in his ritual. His suggestion about an ancestral tribute during the rites, while interesting, raises some questions: if the platforms were the main focus of the rituals after the anointment/initiation, why leave all the votive gifts on the western side? If the initiation ritual was not the central part of the latter ritual, why were every piece of feasting equipment and votive gift left on the “outside”? Another interesting question raised by Collard is the location of the hearth used to produce the “ash and bones” inside the basin, as recorded by Gjerstad. While the amount inside the basin can be questioned, it is a valid question as to how and where it was produced. The pottery present in this study supports Collard’s interpretation that oils and unguents were an important part of the rituals at Dhima. The numerous RL arm vessels (nine studied sherds, including 213), Lentoid flasks (198–212), Cypr. Myc. piriform (38–107) and stirrup jars (108–109) are all believed to have been primarily used as transport vessels. Especially the Mycenaean piriform jars were very common across Cyprus and believed to be imported from the Aegean specifically for their content.258 However, it is difficult to assess exactly how these contents were utilised at Dhima, whether to anoint a person seated in the tub, to pour inside the basin as a libation or merely as votive gifts, placed around the basin. It seems like the overarching research question of Collard’s study, whether the Sea people introduced bathtubs to Cyprus, clouds his interpretation of Dhima. There is little evidence for the distress, turmoil and destruction of Collard’s narrative at Dhima.

5.1.5 Dhima as an eastern Mediterranean sanctuary In 2011 Hitchcock published a study on Cypriot rural sanctuaries in a larger Mediterranean context, with a focus on shared features between Cypriot sanctuaries and similar sites in the Aegean and the Levant. The most prominent categories she discusses are the occurrences of floors, sacred trees and stones, association with triads, traces of offering, bovine imagery and feasting debris. Within this type of sanctuaries, she includes the “high place” or bamah, also discussed by Wright above.259 A central focus of her study is placed on the use of raised platforms, which she claims functioned as display platforms for cult objects.260 Compared to Hitchcock’s descriptions of extra-urban sanctuaries, Dhima does not seem to match. As argued above in connection with Wrights interpretation of the sacred tree and stone, there is no evidence for such features at Dhima. Any interpretation that such objects or installations existed should be carefully approached. Similarly, there is no evidence of a triad setting in the limited architectural remains. Neither is there any bovine imagery. There is no pictographic evidence on the pottery or fragments of figurines found either during the original publication or in this study. In fact, there is only one example of animal imagery, the bronze lion found inside the basin.261 This leaves the floor, traces of offering and feasting equipment.

258 Steel 1998, 286. 259 Hitchcock 2011, 510. 260 Hitchcock 2011, 511. 261 See Gjerstad et al. 1934a, 358.

65 While Gjerstad interpreted the ash and bones in the basin as an offering, Hitchcock sees it as evidence of consumption and feasting.262 The legitimacy of the ash- and bone-contents of the basin has been questioned, but it could be argued that some consumption took place at Dhima. However, the presence of the Mycenaean krater (No. 44), BR cups (2–26) and WS bowls (270–277) indicates a consumption of liquids, rather than eating. There is also the absence of any animal remains, something that is commonly found at other sanctuary sites connected to feasting. 263 The Mycenaean krater served as a central part in Cypriot consumption activities and BR cups were the go-to drinking vessel during the early LC period, before Mycenaean open vessels replaced them.264 While WS bowls have been concluded to have a wider use, the consumption of liquids is known to have occurred.265 Without being able to specify the precise nature of the consumption done at Dhima, there seems to be a preference to liquid consumption and with it possibly libation. On a more precise nature, Hitchcock compares Dhima to the so-called “Bull Site”, due to similarities with regards to the architectural layout.266 The bull site was an Iron Age I extra-urban sanctuary found on the summit of a ridge in the northern part of the Samaria hills. It is dated to the 12th century BC and located far from any neighbouring settlements. 267 The floor area consisted of a roughly 21m x 23m elliptical area, surrounded by a wall constructed of large stones. The encircling wall turns into the site on the eastern side, possibly as the remains of an entrance. In the southeastern part of the sanctuary was a large rectangular stone, interpreted as an Māșșēbâh, or standing stone, possibly used as an altar. Mazar also speculates on the presence of a sacred tree, although there does not seem to be any physical evidence of such a feature.268 Only a few finds were recovered, most consistent of pottery, but also part of a ceramic incense burner, a bronze handle, a few animal bones and the eponymous bull figurine.269 The bull figurine was found on the western side of the sanctuary, and all finds were located around the Māșșēbâh.270 Some of these characteristics are also found at Dhima. The remote location on a hilltop, the round shape and the partitioning wall are all features found at Dhima. However, there are also differences. There are no figurines at Dhima, and no connection with bovine imagery, which Mazar links to a cult of either Ba´al or Yahweh.271 The few pottery sherds recorded at the Bull Site include mostly cooking vessels and open shapes, which differs from the predominantly closed, transport vessels found at Dhima (most notably the many piriform jars 38–107, Red lustrous container vessels 198–214). Hitchcock’s note of incense burners at Dhima is also problematic. While the original publication documented the RL arm vessels as incense burners, it has since been concluded that they were more likely libation vessels.272 Lastly, given the 12th-century date of the bull site, Dhima predates it by three centuries, giving the two sites not just a geographical distance, but a chronological one as well. This is an aspect not discussed by Hitchcock at all. In the end, while there are similarities between Dhima and Hitchcock’s rural sanctuaries, they seem more coincidental than actual.

262 Hitchcock 2011, 517–518. 263 Knapp 2008, 165. See also Webb 1999, 250–252 for a summary of faunal remains at sanctuaries. 264 Steel 2004b, 293. 265 Beck et al. 2004, 18. 266 Hitchcock 2011, 517. 267 Mazar 1982, 33, 36. 268 Mazar 1982, 34–35. 269 Mazar 1982, 35–36. 270 Mazar 1982, 27, 35. 271 Mazar 1982, 40. 272 See Eriksson 1993, 144 and Knappett et al. 2005, 49.

66 5.2 Dhima in the light of newly studied pottery

5.2.1 The construction of the floor and the chronology of the sanctuary Previous studies have accepted the chronology of Dhima as established by Sjöqvist in 1934. However, there is enough evidence in this study to question both the dating of the sanctuary, as well as the construction of the platforms, wall and floor. Most of the ceramic material and its recorded stratigraphic depth support previous interpretations of Dhima. Table 5 shows that pottery dating to the LCII period is almost exclusively found in the upper layers, while the lower layers primarily consist of pottery dated to the MC period. However, due to the presence of RoB, BS handmade and RS handmade pottery found above the floor level, the construction of the floor must be adjusted. 41 sherds of RoB were found above the floor, either on a depth of 0–20cm or “on the first floor.” Seven sherds from this group are catalogued (216–218, 224, 228 and 230–231). Additionally, 44 sherds were recorded as being found between 20– 40cm. However, the fact that almost all of them were found in Grid E5 most likely indicate that they were part of the MCIII-pit fill, and therefore below the floor. RoB pottery is commonly dated to the MCIII and LCI period.273 Ten catalogued BS and RS handmade sherds (30, 32, 33, 247, 249–250, 254, 256–258) were also found either on the first floor or at a depth of 0–20cm. While both RS and BS pottery continue to be in use from the MC to the LC period, handmade production of these wares did not continue during the LCII period,274 but was exclusively produced on the wheel. Like the presence of RoB, this suggests activity at Dhima also in the LCI period. The presence of MCIII–LCI RoB, RS and BS sherds above the floor indicates one of two things: 1. There was a hiatus at Dhima, but the floor was constructed earlier than previously suggested, before the end of LCI. 2. The sanctuary was in continued use after the closure of the MCIII pit and a floor was constructed before the end of LCI. Both options push the dating of the floor to the LCI period and suggest a longer period of use at Dhima, either from the MCIII period or at least from the LCI period onwards. This contradicts the interpretation of a hiatus during the LCI period, as well as the interpretation of Dhima as a “single-event” site. It does, however, not necessarily mean that the sanctuary was in continuous use from the construction of the MCIII-pit to the abandonment in the LCII period. But it does suggest that the floor was constructed in LCI and that there was activity during this period as well. Since the floor runs up against the wall and platforms at the sanctuary, it follows that the construction of these installations must be pushed back to the LCI period as well. This leaves only a small window of time for the hiatus to take place. Because of the few LCI finds and the simple stratigraphy, it is difficult to determine the hiatus further with any certainty. But, if objects at Dhima were routinely removed and deposited outside the sanctuary area, it might have removed much of the stratigraphic evidence from the earliest use of the floor. Since only the direct vicinity of the site was excavated, any nearby deposits might have been overlooked. The pottery in Grids B5, D0 and D6 could be interpreted as evidence of such cleaning. Especially considering the joining sherds between these grids and the central area of the sanctuary. There is also the factor of the steep sides of the hill; much of the earlier, cleaned out material simply have eroded away. With the narrow timeframe and the possibility of routine cleaning, it is then perhaps reasonable to argue that the hiatus did, in fact, not take place at all.

273 Crewe 2004, 88; Spigelman 2015 519. 274 Popham 1972b, 700–701, Crewe 2007, 213, table 2.

67 5.2.2 An enclosure While there may have been an enclosure surrounding Dhima, there is no archaeological evidence supporting its existence. If it was constructed entirely from perishable material, without being founded in the ground, it would not have left any traces. In either case, its alleged existence can neither be proved, nor disproved, at this point. Most likely, its interpretation is based on the presence of enclosures at other ritual sites, such as Ayia Irini (which shows evidence of an earth enclosure) or the Vounous bowl (which has been interpreted as a model of a sanctuary similar to Dhima).275 However, neither of these are contemporary with Dhima and there is no record of an enclosure at either Myrtou-Pighades or Athienou Bamboulari tis Koukounninas.

5.2.3 Two areas in the sanctuary. It is clear that there was a distinction between the eastern and western side of the sanctuary. While this has previously been established by the architecture, the platforms, the basin and the fact that all published small finds were found on the western side, it is also supported by the distribution of the pottery. The wall serves as a clear boundary between the eastern and western side of Dhima and it separates concentrations of earlier and later pottery. On the eastern side, the majority of RoB, RS, BS and PWHM ware, usually dated to the MC and LCI periods, is found (see Table 4). Most sherds are from the MCIII-pit, but there are occurrences in other grids and at other depths as well, specifically Grid F5. Additionally, only sporadic sherds of the characteristically LCII wares, BR (21 sherds, including 10–11, 14, 18, 21–25, 27– 28), Cypr. Myc. (21 sherds, including 43, 63, 77, 82 and 99–100), RL (five sherds, including 198) and three not catalogued WS sherds, are found on the eastern side (see Table 2). The exception is the minor concentration of LCII pottery recorded in Grid D6, containing BR (20 sherds, 10, 11, 14, 21–25 27–28), Cypr. Myc., (seven sherds, including 43, 99–100), RL (two sherds, including198) and the three WS sherds. On the western side the situation is the opposite, as shown in Table 2. There are large concentrations of Cypr. Myc. (290 sherds, including 38–42, 44–62, 64–76, 78–81, 83–98, 101–109), BR (69 sherds, including 2–9, 12–13, 15–17, 19–20 and 26), RL (374 sherds, including 195–197 and 199–214) and WS pottery (45 sherds, including 270–277), with only sporadic occurrences of RoB (11 sherds, including 215, 218, 222–223, 225, 227, 233–234) RS (237–240, 248, 251–252, 256–257, 260–264 and 266) and BS (32–34 and 36). While the handmade vessels of RS are limited to the MC and LCI period, wheelmade production continues during the LCII period and has been recorded on both sides of the wall (See 236, 243–246 and 255). The division of the pottery introduces the question how the eastern side was used during the LCII period. Perhaps the eastern side, the wall and the platforms were part of the earliest stages of the cult, which was then altered when the basin was installed, and the rites were moved to the western side. Or perhaps the eastern side was merely cleaned after use, as indicated by the small concentration in Grid D6, outside the floor area. In any case, the final deposition of the votive gifts during the LCII period occurred in and around the terracotta basin.

5.2.4 The pottery and its function during the LCII period. Several of the studies discussed above have described feasting and consumption as a central part of the ritual activities conducted at Dhima during the LCII period. This has primarily been based on Gjerstad’s description of the terracotta

275 Wright 1992, 271; Steel 2013.

68 basin, as well as the presence of the Mycenaean krater. Towards the end of the LCI and beginning of the LCII period, imported Mycenaean pottery became the standard equipment used for such consumption activities, successively outcompeting the local wares.276 But even so, Mycenaean pottery commonly consists of only a few percent at other Cypriot sites.277 The abnormal amount of Cypr. Myc. ware (17% of the LCII assemblage) at Dhima is therefore noticeable. However, none of these sherds can be associated with consumption. Instead, the studied Cypr. Myc. assemblage consists of only two shapes, piriform (FS31, 39, 44, 45 and 47) and stirrup jars (FS166). This could be explained by the presence of BR cups and WS bowls, and absence of Mycenaean cups. Before the import of Mycenaean pottery took off during the LCIIB period, these shapes were commonly used in drinking contexts.278 The absence of bones, the presence of BR cups, WS bowls and the Mycenaean krater, therefore, makes the consumption of liquids at Dhima seem very likely. The act of ritual drinking itself was not new to Cyprus in the Late Bronze Age, and not something that was introduced with Mycenaean pottery. The imported vessels were instead incorporated into an existing custom and quickly became the standard equipment.279 The vast majority of the LCII sherds recorded at Dhima have been identified as closed (93%). This suggests that a central part of the activities at Dhima included the storage of goods. Most of these sherds belong to unidentified plain wares, but several catalogued sherds are worth mentioning. The almost exclusive presence of Cypr. Myc. piriform jars is striking. Possibly as many as 56 vessels are recorded (c. 300 sherds, including 38–107). While this type of container is the most commonly found Mycenaean vessel in Cyprus, it is usually mixed with other shapes.280 These vessels were primarily imported to Cyprus for their contents of oils and unguents.281 Similarly, the RL Lentoid flasks and arm vessels are believed to have contained perfumed oils.282 Like Cypr. Myc. sherds, the RL sherds make up a substantial part of the LCII assemblage, c. 28%. This means that almost 50% of the entire LCII assemblage consists of small container vessels, filled with precious oils and unguent. This content must have served a significant purpose at Dhima. As mentioned above, Collard’s suggestion of using such oils to anoint a person in the basin would explain the need for such contents. It is also possible that the oils were used for libation offerings inside the basin, which is what Webb associated a high amount of closed vessels with.283 It is possible that RL arm vessels could have functioned as both containers and libation vessels.284 Eriksson also links the silver funnels at Dhima to the arm vessels and libation sacrifices.285 Additionally, besides being the standard drinking cup, BR cups have been suggested to function as a libation vessel as well, in sanctuary contexts.286 As such, there is enough evidence in the ceramic assemblage to suggest at least two activities conducted at Dhima. The first is liquid consumption, an activity with a long history in Cyprus and which can be closely connected with the pottery assemblage found at Dhima. The second is the use of perfumed oils and unguents. Whether the oils were used for anointment or libation is difficult to say. However,

276 See Steel 1998, 2004b, Knapp 2008, 257. 277 Steel 1998, 286. 278 Especially in regards to the BR Y-shaped cups, which has been identified as a preferred drinking vessel in Cyprus (Steel 1998, 293; Steel 2004b, 292). 279 Steel 1998, 292. 280 Steel 1998, 286; Wijngaarden 2002, 153. 281 Steel 1998, 286; Steel 2004a, 73. 282 Eriksson 1993, 143–144; Knappett et al. 2005, 49. 283 Webb 1999, 189. 284 Eriksson 2008a, 304. 285 Eriksson 2008, 304. 286 Vaughan 1991, 124.

69 libation sacrifices are more established than the use of oil for anointment, especially in sanctuary contexts.287 Perhaps the unidentified layer at the bottom of the terracotta basin is simply the residue of such libation sacrifices?

5.3 What does the recontextualisation mean for the understanding of early LCII cult? As shown above, Dhima can be interpreted in many ways. Nevertheless, based on the material from this study, it is clear that some interpretations and conclusions are more plausible than others. With this in mind, it is time to recontextualise Dhima in a larger Cypriot setting. Several sanctuaries have been excavated in Cyprus, but only a handful are dated to the LCIIA and LCIIB periods.288 Dhima is most commonly discussed together with two, more or less, contemporary, sanctuaries: Myrtou- Pighades and Athienou Bamboulari tis Koukounninas. Both of these sanctuaries were in use during the LCII period.289 Looking at the pottery assemblages, there is a certain level of homogeneity between the three sanctuaries. All sites have produced pottery typically used for liquid consumption, BR and WS bowls and PWWM and Mycenaean kraters. However, this is more or less where the similarities end. Both Myrtou and Athienou show a wider range of Mycenaean shapes, whereas Dhima has a very limited range. There is also the sheer amount of both Cypr. Mycenaean and RL sherds at Dhima, which is unparalleled at both other sanctuaries. The preference for closed vessels at Dhima is also noteworthy, suggesting the importance of the contents of the containers, whereas, at Myrtou-Pighades, the majority of the shapes were open. 290 The importance of feasting is likewise established at both Myrtou and Athienou, in the form of large osteological assemblages. Osteological remains found in ritual contexts have been linked to both animal sacrifice and ritual consumption.291 The absence of such remains at Dhima illustrates a difference in consumption behaviour. It could be argued that any osteological remains at Dhima were cleaned alongside the pottery, but if this was the case, some of it should have been noted along the edges of the sanctuary. This absence of faunal remains is puzzling since one of the most characteristic elements of Cypriot sanctuaries is the presence of osteological material. Lastly, there is the situation with ritual equipment. The only artefacts at Dhima that have any clear connection with a ritual behaviour are the RL arm vessels. Both Myrtou and Athienou show evidence of a more developed ritual tradition. Considering the updated dating of Dhima, this is perhaps not so surprising. Before the earliest sanctuaries, communal consumption of liquids and food was a central part of Cypriot mortuary rites.292 Even if Dhima is not directly linked to funerary cult, or shows evidence of feasting, it is possible that certain aspects of a long, established tradition of communal consumption were extracted from its original funerary setting into a general ritual context. With this in mind, perhaps Dhima should not be viewed as a unique phenomenon as much as an early regional cult locale with roots in the funerary traditions from the end of the MC period. During the early stages of the LC period major social changes swept over Cyprus and the material culture was

287 Steel 2004, 296–297. 288 See Webb 1999 for a full overview. 289 Myrtou is dated to the LCIIA-LCIIC period, and Athienou to the LCIA-LCIIIA period. However, the earliest phases of Athienou are not necessarily ritual in nature during the LCI/early LCII period, Dothan and Ben-Tor 1983, 139. Collard 2011, 288, 292, table 6.5. 290 Webb 1999, 188–191, and table 2. 291 Webb 1999, 250–252. 292 Steel 2004b, 287–288.

70 primarily regional until the LCII period.293 It is not unlikely that ritual locales, such as Dhima, but also Myrtou and Athienou, would show evidence of regionalism. The fact is that neither of these sites is entirely similar to one another. Webb classifies all three of these sanctuaries as serving a different purpose. 294 Nevertheless, there is homogeneity in some certain aspects. In conclusion, since Dhima does not quite correspond with other contemporary Cypriot sanctuaries, its interpretation as a unique cult locale would appear to be solidified. There are common elements shared with other Cypriot sanctuaries, both with regards to material culture and conducted activities, but they do not seem to extend beyond a communal consumption of liquids. Nevertheless, instead of interpreting Dhima as an entirely isolated phenomenon, it should perhaps be viewed as part of the early development of Cypriot ritual behaviour, as it got separated from a funerary context into a more distinct ritual tradition. If anything, this study has shown that there is much to be gained by revisiting old excavations and unstudied material. By looking at the entire ceramic assemblage at Dhima, it has been possible to re-evaluate previous interpretations as well as to draw new conclusions both in regards to the continuity, chronology and function of the sanctuary at Ayios Iakovos: Dhima.

293 Especially on the Karpas peninsula, which is believed to have been on the frontier with the increased external contacts during this period. 294 Webb 1999, 297.

71 Bibliography

Allen, S.H. 1991. ‘Late Bronze Age Grey Wares in Cyprus’, in Cypriot Ceramics: Reading the Prehistoric Record. Proceeding of an international conference held at the University Museum of Archaeology and Anthropology in October 1989 (University Museum Symposium Series, II), eds. J.A. Barlow, D.L. Bolger, and B. Kling, Philadelphia, 151–68.

Aloupi, E., V. Perdikatsis, and A. Lekka 2001. ‘Assessment of the White Slip Classification Scheme based on Physical and Chemical Aspects of the Technique’, in The White SLip ware of Late Bronze Age Cyprus: Proceedings International Conference Organized by the Anastasios G. Leventis Foundation, Nicosia in Honour of Malcom Wiener, Nicosia 29th-30th October 1997. (Contributions to the chronology of the Eastern Mediterranean, XX), ed. V. Karageorghis, 15–26.

Åström, P. 1966. Excavations at Kalopsidha and Ayios Iakovos in Cyprus (SIMA, II), .

--- ed., 1972a. The Middle Cypriot Bronze Age (SwCyprusExp, IV:1b), Lund.

--- ed., 1972b. The Late Cypriote Bronze Age: architecture and pottery (SwCyprusExp, IV:1c), Lund.

Åström, P. and G.R.H. Wright 1962. ‘Two Bronze Age Tombs at Dhenia in Cyprus’, OpAth:IV, 225–303.

Beck, C.W., E. C. Stout, K.M. Wovkulich, V. Karageorghis & E. Aloupi 2004. ‘The Uses of Cypriot White-Slip Ware Inferred from Organic Residue Analysis’, Ägypten und Levante / Egypt and the Levant 14, 13–43.

Campbell Gagné, L.A. 2012. ‘Middle Cypriot White Painted Ware: A Study of Pottery Production and Distribution in Middle Bronze Age Cyprus’, University of Toronto.

Catling, H.W. 1957. ‘The Bronze Age Pottery’, in Myrtou-Pigadhes: a Late Bronze Age Sanctuary in Cyprus), ed. J.d.P. Taylor, Oxford, 26–59.

Collard, D. 2008. Funcation and Ethnicity: 'Bathtubs' from Late Bronze Age Cyprus, ed. P. Åström (SIMA-PB, 171), Sävedalen.

--- 2011. ‘Altered States of Consciousness and Ritual in Late Bronze Age Cyprus’, University of Nottingham.

Crewe, L. 2004. Social Complexity and Cermic Technology on Late Bronze Age Cyprus: The New Evidence from Enkomi, Edinburgh.

72 --- 2007. ‘Sophistication in Simplicity: The First Production of Wheelmade Pottery on Late Bronze Age Cyprus’, JMA 20:2, 209–38.

--- 2009. ‘Regionalism and the First Appearence of Plain White Handmade Ware in the Middle Cypriot Bronze Age’, in The Formation of Cyprus in the 2nd Millenium B.C. Studies in Regionalism during the Middle and Late Bronze Age. Proceedings of a workshop held at the 4th Cyprological Congress, May 2nd 2008, Lefkosia, Cyprus (Contributions to the Chronology of the Eastern Mediterranean, XX), ed. I. Heinz, Wien, 79–90.

--- 2016. ‘Plain Wares and Urban Identities in Late Bronze Age Cyprus’, in Plain Pottery Traditions of the Eastern Mediterranean and Near East), ed. C. Glatz, London, 115–34.

Davit, P., F. Turco, S. Coluccia, L Operti, F. Chelazzi & L Bombardieri 2014. ‘Technological and Compositional Characterization of Red Polished Ware from the Bronze Age Kouris Valley’, MAA XX:X, 2–17.

Dothan, T. and A. Ben-Tor 1983. ‘Excavations at Athienou, Cyprus 1971-1972’, Qedem 16, 1–147.

Eriksson, K.O. 1993. Red Lustrous Wheel-Made Ware (SIMA, CIII), Jonsered.

--- 2008a. ‘Feasting as Part of the Multiculturalism of Late Bronze Age Cyprus’, in Dais - The Aegean Feast. Proceeding of the 12th International Aegean Conference/University of Melbourne, Centre for CLassics and Archaeology, 25-29 March 2008 (Aegaeum, 29:29), eds. L.A. Hitchcock, R. KLaffineur, and J. Crowley, 2008, 297-307.

--- 2009. ‘Regionalism and Island-wide Analysis: some Observations regarding White Painted V and VI wares from Middle Cypriot III/Late Cypriot I Tombs from the Northwest Region of Cyprus.’, in The formation of Cyprus in teh 2nd Millenum B.C. Studies in Regionalism during the Middle and Late Bronze Ages, Proceedings of a workshop held at the 4th Cyprological Congress, May 2nd 2008, Lefkosia, Cyprus. (Contributions to the Chronology of the Eastern Mediterranean, XX), ed. I. Hein, Wien, 49-64.

Gjerstad, E. 1926. ‘Studies on Prehistoric Cyprus’, Uppsala University.

--- 1933. Sekler och dagar: med svenskarna på Cypern 1927-1931, Stockholm.

Gjerstad, E., J. Lindros, E. Sjöqvist & A. Westholm 1934a. Finds and Results of the Excavations in Cyprus 1927-1931 (SwCyprusExp, I:1), Stockholm.

--- 1934b. Finds and Results of the Excavations in Cyprus 1927-1931 (SwCyprusExp, II:1), Stockholm.

Hankey, V. 1967. ‘Mycenaean Pottery in the Middle East: Note on Finds since 1951’, BSA 62, 107–47.

Hitchcock, L. 2011. ‘Cypriot Rural Sanctuaries in Their Larger Eastern Mediterranean Context’, in The IV. International Cyprological Congress. Proceedings of the IV- International Cyprological Congress, ed. A. Demetriou, Lefkosia, 509–20.

73 Horowitz, M.T. 2008. ‘Phlamoudhi- Vounari: A Multi-Function Site in Cyprus’, in Views from Phlamoudhi, Cyprus (AASOR, 63), ed. J.S. Smith, Boston, 69–86.

Hult, G. 1992. Nitovikla reconsidered (Medelhavsmuseet memoir, 8), Stockholm.

Jacobs, A. 2009. ‘Considering Ceramic Variability on Late Bronze Age Cyprus. A Case-study: The Plain Vessels of Alassa Pano-Mandilaris’, in The Formation of Cyprus in the 2nd Millenium B.C. Studies in Regionalism during the Middle and Late Bronze Age. Proceedings of a workshop held at the 4th Cyprological Congress, May 2nd 2008, Lefkosia, Cyprus. (Contributions to the Chronology of the Eastern Mediterranean, XX)Wien, 91–106.

Jones, R.E. 1986. Greek and Cypriot Pottery. A Review of Scientific Studies, Athens.

Karageorghis, V. 2001. ‘Why White SLip?’, in The White SLip ware of Late Bronze Age Cyprus: Proceedings International Conference Organized by the Anastasios G. Leventis Foundation, Nicosia in Honour of Malcom Wiener, Nicosia 29th-30th October 1997. (Contributions to the chronology of the Eastern Mediterranean, XX), ed. V. Karageorghis, 9–14.

Keswani, P. 1991. ‘A Preliminary Investigation of System of Ceramic Production and Distribution in Cyprus During the Late Bronze Age’, in Cypriots Ceramics: Reading the Prehistoric Record. Proceeding of an international conference held at the University Museum of Archaeology and Anthropology in October 1989 (University Museum Symposium Series, II), eds. J.A. Barlow, D.L. Bolger, and B. Kling, Philadelphia, 97–118.

--- 2012. ‘Mortuary Practices and Burial Cults in Cyprus from the Bronze Age Through the Early Iron Age’, in Parallel Lives: Ancient Island Societies in Crete and CYprus 20), eds. G. Cadogan, et al., London, 313–30.

Keswani, P.S. 2004. Mortuary Ritual and Society in Bronze Age Cyprus (MMA), London.

Knapp, B., A. 2009. ‘Monumental Architecture, Identity and Memory’, in Bronze Age Architectural Traditions in the East Mediterranean: Diffusion and Diversity (Gasteig, Munich, 7-8 May, 2008). Proceedings of the Symposium ‘Bronze Age Architectural Traditions in the Eastern Mediterranean: Diffusion and Diversity’, 07.-08. 05. 2008 in Munich Weilheim, 47–59.

--- 2008b. Prehistoric and Protohistoric Cyprus: Identity, Insularity, and Conncetivity, Oxford.

Knappett, C., V. Kilikoglou, V. Steele & B. Stern 2005. ‘The Circulation and Consumption of Red Lustrous Wheelmade Ware: Petrographic, Chemical and Residue Analysis’, AnatSt 55, 25–59.

Kozal, E. 2016. ‘Cypro-Anatolian Connections in the 2nd Millenium BC’, in The Northern Face of Cyprus: New Studies in Cypriot Archaeology and Art History), eds. L. Summerer and H. Kaba, Istanbul, 51–64.

Leonard, A. 1981. ‘Considerations of Morphological Variation in the Mycenaean Pottery from the Southeastern Mediterranean’, BASOR 241, 87–101.

74 Mazar, A. 1982. ‘The "Bull Site": An Iron Age I Open Cult Place’, BASOR 247, 27– 42.

Merillees, R., S. 1963. ‘Bronze Age Spindle Bottles from the Levant’, OpAth:IV, 187–96.

--- 2008. ‘Phlamoudhi Vounari and Melissa: 35 Years and 35 Centuries Later’, in Views from Phlamoudhi, Cyprus (AASOR, 63), ed. J.S. Smith, Boston, 111– 20.

Mountjoy, P. 1986. Mycenaean Decorated Pottery: A Guide to Identification (SIMA, LXXIII), Göteborg.

Nicolaou, I. and K. Nicoulaou 1989. Kazaphani: A Middle/Late Cypriot Tomb at Kazaphani - Ayios Andronikos: T 2A, B, Nicosia.

Popham, M.R. 1972a. ‘White Slip Ware’, in The Late Cypriote Bronze Age: architecture and pottery (SwCyprusExp, IV:1c), ed. P. Åström, Lund, 431–71.

--- 1972b. ‘A Note on the Relative Chronology of White Slip Ware’, in The Late Cypriote Bronze Age: other arts and crafts (SwCyprusExp, IV:1d), eds. P. Åström and L. Åström, Lund, 699–705.

Sesam, 1997. A.J. B.A.S. 018 :: Crater. Medelhavsmuseet .

Sjöqvist, E. 1940. ‘Problems of the late Cypriot bronze age’, Uppsala University.

Smith, J.S. 2008. ‘Settlement to Sanctuary at PhlamoudhiMelissa’, in Views from Phlamoudhi, Cyprus (AASOR, Boston, 45–68.

South, A. and L. Steel 2001. ‘The White Slip Ware of Late Bronze Age Cyprus: Proceedings of an International Conference Organized by the Anastasios G. Leventis Foundation, Nicosia, in Honour of Malcom Wiener, Nicosia 29th– 30th October 1998’, in The White Slip Ware of Late Bronze Age Cyprus. ed. V. Karageoghis, Vienna, 65–74.

Spigelman, M.D. 2015. ‘Red-on-Black Style Ceramics: Production and Trade in Bronze Age Cyprus’, New York Unviersity.

Steel, L. 1998. ‘The Social Impact of Mycenaean Imported Pottery in Cyprus’, BSA 93, 285–96.

--- 2004a. ‘A reappraisal of the distribution, context and function of Mycenaean Pottery in Cyprus’, in La Céramique mycénienne de l'Égée au Levant. Hommage à Vronwy Hankey. Table ronde internationale, Maison de l'Orient et de la Méditerranée, 20 mars 1999 (Travaux de la Maison de l'Orient et de la Méditerranée, 41), eds. J. Balensi, J.-Y. Monchambert, and S. Müller Celka, Lyon, 69–85.

--- 2004b. ‘A Goodly Feast... A Cup of Mellow Wine: Feasting in Bronze Age Cyprus’, Hesperia 73:2, 281–300.

75 --- 2013. ‘The Social World of Early-Middle Bronze Age Cyprus: Rethinking the Vounous Bowl’, JMA 26:1, 51–73.

Styrenius, C.-G. 1994. ‘From the Swedish Cyprus Expedition to the Medelhavsmuseet of Today’, in The Swedish Cyprus Expedition. The Living Past (Medelhavsmuseet Memoir, 9), ed. E. Rysted, Stockholm, 7–17.

Vaughan, S.J. 1991. ‘Material and Technical Classification of Base Ring Ware: A New Fabric Typology’, in Cypriot Ceramics: Reading the Prehistoric Record. Proceeding of an international conference held at the University Museum of Archaeology and Anthropology in October 1989 (University Museum Symposium Series, II), eds. J.A. Barlow, D.L. Bolger, and B. Kling, Philadelphia, 119–30.

Voskos, I. and B. Knapp, A. 2008. ‘Cyprus at the End of the Late Bronze Age: Crisis and Colonization or Continuity and Hybridization’, AJA 112:4, 659–84

Webb, J.M. 1992. ‘Funerary Ideology in Bronze Age Cyprus: Towards the Recognition and Analysis of Cypriote Ritual Data’, in Studies in honour of Vassos Karageorghis 54-55), ed. G.C. Ioannides, Leucõsia, 87–100.

--- 1999. Ritual Architecture, Iconography and Practice in the Late Cypriot Bronze Age (SIMA-PB, 75), Jonsered.

Wijngaarden, G.J.V. 2002. Use and Appreciation of Mycenaean Pottery in the Levant, Cyprus and Italy (1600-1200 BC) (Amsterdam Archaeological Studies, 8), Amsterdam.

Winbladh, M.-L. 1992. Cypern: möte mellan kulturer: om Cyperns arkeologi och cypernsamlingarna i Medelhavsmuseet (Medelhavsmuseet skrifter, 18), Stockholm.

Wright, G.R.H. 1992. ‘The Cypriot Rural Sanctuary: An Illuminating Document in Comparative Religion’, in phier ma sto Baso aragi rgē: Studies in honour of Vassos Karageorghis 54-55), ed. G.C. Ioannides, Leucõsia, 269–83.

76 List of Illustrations Figure 1. Map of Cyprus with important Late Cypriot sites (from Knapp 2008a) ...... v Figure 2. Plan of Ayios Iakovos: Dhima (Coutesy of the Mediterranean Museum. Redrawn digitally by the author) ...... 10 Figure 3. View of Ayios Iakovos: Dhima from the west (Courtesy of the Mediterranean Museum, ID C02546.) ...... 11 Figure 4. View of Ayios Iakovos: Dhima from the north (Courtesy of the Mediterranean Museum, ID C01418) ...... 12 Figure 5. View of the MCIII pit, small platform and central wall. Photo taken from the north (Courtesy of the Mediterranean Museum, ID C01416) ...... 13 Figure 6. Profiles I–III (Coutesy of the Mediterranean Museum. Redrawn digitally by the author). .... 14 Figure 7. The architecture of Ayios Iakovos: Dhima. Photo taken from the south (Courtesy of the Mediterranean Museum, ID C01460)...... 16 Figure 8. Excavation photo of the terracotta basin in Grid D2. Photo from the north (Courtesy of the Mediterranean Museum, ID C01419)...... 17 Figure 9. Terracotta basin, detail plan, Grids E2–E3 & D2–D3 (Coutesy of the Mediterranean Museum. Redrawn digitally by the author) ...... 18 Figure 10. The excavated terracotta basin in Grids D3–E2. Taken from the south. (Courtesy of the Mediterranean Museum, ID C1417)...... 19 Figure 11. Red Lustrous Arm Vessel concentration in Grid D1 (Coutesy of the Mediterranean Museum. Redrawn digitally by the author) ...... 19 Figure 12. Excavation photo of the arm vessel concentration in Grid D1. Photo from the north (Courtesy of the Mediterranean Museum, ID C01420) ...... 20 Figure 13. Ratio of MCIII/LCI wares (by the author) ...... 55 Figure 14. Open/closed ratio of MCIII/LCI wares (by the author) ...... 55 Figure 15. Ratio of LCII wares (by the author) ...... 55 Figure 16. Open/closed ratio of LCII wares (by the author) ...... 55

List of Tables Table 1. General terminology and Abbreviations...... iv Table 2. Relative and (approx.) absolute chronology of Cyprus and the Aegean (after Crewe 2007, and Mountjoy 1986, )...... iv Table 3. Catalogued sherds...... 44 Table 4. Horizontal distribution of sherds...... 56 Table 5. Vertical distribution of sherds...... 56

List of Plates Plate I (by the author)...... 78 Plate II (by the author) ...... 79 Plate III (by the author) ...... 80 Plate IV (by the author) ...... 81 Plate V (by the author) ...... 82 Plate VI (by the author) ...... 83

77

Plate I.

78

Plate II.

79

Plate III.

80

Plate IV.

81

Plate V.

82

Plate VI.