Acknowledgements

Thank you to the staff of the Piraeus Museum in Greece for their time and assistance accessing the Aegina-Kolonna collections; to the anonymous reviewers whose comments strengthened this paper, for their thoughtful consideration; to the German Archaeological Institute-Athens; and to the staff of the Geological Sciences SEM Laboratory at the University of Wisconsin-Madison. Funding was provided by Dr. Pieniążek’s Marie Curie Intra-European Fellowship (Patterns of trade and consumption of valuable objects in the northern Aegean area in the 2^nd millennium BC) and by the University of Wisconsin-Madison Departments of Anthropology and Classics.

1The 3^rd millennium BCE is a period characterised by the development of early urban civilisations throughout the Old World marked by increasing long distance trade and cultural interaction of other kinds (e.g., migration, settlement, and conquest), demonstrating the cultivation of interregional relationships (Kenoyer 1998; Aruz 2003). This situation is well-documented in both the historical and archaeological records of the major regional centres of the 3^rd millennium world: Egypt, Mesopotamia and the Indus Valley (Kenoyer 1998, 2003b; Allen 2003; Aruz 2003). Yet other interaction zones of the 3^rd millennium world have also yielded plentiful data for an increasing intensity of exchange and cultural interrelations, especially trade, during this period (Ramstorf 2015). Ancient societies such as those in the 3^rd millennium Southern Levant, the Persian Gulf, Central Anatolia and the Aegean region were distinct entities, yet nevertheless were closely linked to the great powers of the age. They participated in interconnected systems of economy and ideology, appropriating and syncretising foreign cultural elements into their own local milieus (Aruz 2003). Such areas were often defined by their participation in these social, political, and economic interaction networks. Small finds are useful, if often overlooked, indicators of these cultural relationships. Beads, in particular, were idiosyncratically produced in different ways in different regions, dependent on the learned production methodology of certain traditions of crafting. These differences in chaîne opératoire can be discerned and isolated when collections are viewed comparatively. It is possible to identify products found in one location that were affiliated with different regional craft traditions and thereby reconstruct links between the culture of origin and the place of discovery. Therefore, the stone beads from the citadel of Aegina-Kolonna in Greece provide an excellent case study in the lively systems of 3^rd millennium exchange connecting the wider Near East, local identity negotiation, and social value as expressed in material culture.

2In this paper, we examine the twenty hard stone beads that were found in a hoard at the site of Aegina-Kolonna. The nineteen carnelian beads and one of rock crystal were recovered along with gold and silver object and published by C. Reinholdt (2003a, 2003b, 2008). One of these carnelian beads displayed a tell-tale sign of extra-regional manufacture: a white eye design, commonly referred to as “etched” represents a bleaching technique considered to be diagnostic of Indus Valley civilisation craft traditions (Kenoyer 1997, 1998, 2000, 2003a, 2003b, 2008; Inizan 2000; Roux and Matarasso 2000; Reinholdt 2003a; Vidale 2004). If one Indus-related bead was present in the hoard, the question remained: could there have been others? By reanalysing the corpus, we sought to identify the regional origins of beads from Aegina-Kolonna using high-resolution analytical techniques, including stylistic, mineralogical, and morphometric comparisons and drill hole analyses including taper/mm calculation and scanning electron microscope (SEM) investigation of drill hole impressions. On the basis of quantifiable stylistic, technological, and morphometric similarities it was possible to compare the Aegina-Kolonna beads with beads from the Indus region to determine if they reflect the same or similar cultural traditions of crafting, or what can be called a workshop tradition (Chakrabarti 1990; Vidale et al. 1992; Roux et al. 1995; Kenoyer 1997: 267 and 270-274; 1998: 97-98 and 160-2; 2003a: 16-17; 2008: 21-22 and 25-26; 2017a: 130-140; 2017c: 416-417 and 422-425; Vidale 2004; Ludvik et al. 2014, 2015; Jamison 2017; Ludvik 2018). Using this approach, we identify fourteen additional Indus-style carnelian beads in the Aegina-Kolonna hoard and suggest the shared origin of two to three of the four others in the same workshop tradition.

Archaeological context

3The mid-3^rd millennium BCE fortified centre at Aegina-Kolonna was excavated by the University of Salzburg under C. Reinholdt in the late 1990s/early 2000s. Aegina-Kolonna is located on the Greek island of Aegina in the Saronic Gulf, about 30 km (ca. 19 miles) south by southwest of Piraeus, the celebrated port of the Greek city-state of Athens. The settlement at Kolonna hill dates from Final Neolithic (Kolonna I) but the first monumental buildings were erected during late Early Helladic II period (Kolonna II and III, ca. 2500‒2200 BCE; Reinholdt 2008; Berger and Gauss 2016). From the end of the 3^rd millennium BCE (Early Helladic III, Kolonna V) until at least the Shaft Grave period (Middle Helladic III‒Late Helladic I, ca. 1700‒1600 BCE, Kolonna X) Aegina-Kolonna was a fortified citadel (Reinholdt 2008). The settlement continued until Late Helladic IIIB with substantial re-occupation of the entire area during the later Archaic and Classical periods (Reinholdt 2003a; Jalkotzy 2009; Berger and Gauss 2016). The site in the late 3^rd millennium BCE was, “an impressive and well-fortified settlement…[whose] fortifications merit comparison with the walls of Troy” (Reinholdt 2003a: 260). The architecture of Kolonna II‒III belongs to the tradition of the so-called “corridor houses,” a sub-phase of Early Helladic II characterised by the appearance of this type of large structure at sites like Lerna, Thebes, or Troy, similar in general plan to Anatolian Megaron-type buildings and, “associated with a new, more centralised form of government for localities on the Greek mainland” (Reinholdt 2003b: 255). Monumental architecture, impressive fortifications combined with concentrations of luxury goods such as the hoard discussed in detail below, provide evidence for some form of hierarchical, centralised socio-political organisation, or “elites,” at the site. The inhabitants of Aegina-Kolonna, with a well-defended centre also well-positioned to take advantage of sea trade with the wider Aegean region, likely had relatively intensive contact with folk in Anatolia, displayed in shared ceramic and jewellery types and similar architecture and governmental (?) buildings, especially well-compared to the situation at contemporary Troy II-III in western Anatolia (Reinholdt 2003b; Berger and Gauss 2016). The comparisons with 3^rd millennium Troy are further worthy of note, as that site was a wealthy and powerful trading centre during the late 3^rd millennium BCE with far-reaching material connections (Ludvik et al. 2015). The 3^rd millennium elite inhabitants of Troy were themselves well-connected with the interregional interaction systems that linked their home region of western Anatolia to other Anatolian cultures, the wider Aegean, Mesopotamia and beyond. It is demonstrated both by finds coming from settlement contexts as well as several hoarded treasures (Bass 1970; Ludvik et al. 2015; Blum 2016; Pieniążek 2016b; Sazcı 2016). Gold objects coming from these contexts in Troy provide some good parallels to finds from Aegina-Kolonna (Treister 1996). This suggests that elites at Aegina-Kolonna were participating in the same cultural sphere of interaction as those at Troy; it is not surprising to find similarities in the types of stone beads found at both sites.

4Turning then to the specific context of the stone beads analysed, these artefacts were discovered in a jewellery hoard from what seems to have been Aegina-Kolonna’s “elite” precinct. This collection of high-value material was found within Aegina-Kolonna’s citadel, dating to the Greek Early Bronze Age III (Early Helladic III period, local architectural phase V, final 3^rd millennium BCE; Reinholdt 2008; Berger and Gauss 2016). The hoard consisted of 52 artefacts made from very precious materials, notably gold, silver, and hard stone (Reinholdt 2003a: 260). In addition to the 20 hard stone beads, another 171 beads of faience were recovered plus 13 faience fragments (Reinholdt 2008: 110). All of these items were stored in a nested bundle found buried beneath the floor of House 19 (Reinholdt 2008: 68-71). It was deposited most probably shortly before the destruction of the Kolonna V settlement, therefore around 2100 BCE. Reinholdt observed that “the hoard is obviously a collection of reused material. The disk- and ring-shaped beads appear to have originally belonged to larger pieces of jewellery. The components of the hoard are [also] older than the time of its deposit; they date to the Early Bronze Age II” (Reinholdt 2003a: 260). While most of the items fit comfortably in the jewellery corpuses of the Aegean and Anatolian world, evidence for long-distance cultural interaction was documented in the hoard. For example, a gold discoid pendant with a horizontal string tube was highlighted. The singular features of its shape, decoration and technology of production show parallels with various objects known Aegean, Anatolia, and Mesopotamia, but the combination of those features is seemingly unique (Reinholdt 2003a: 261; 2008: 29-45).

5A total of twenty hard stone beads were found in this hoard (fig. 1). These beads were apparently strung together as a single necklace on a very fine silver wire, interspersed with silver tubular beads and faience beads (Reinholdt 2008: 16, 52-54, 117 and 127). Of the twenty beads, one was rock crystal and the rest were made of carnelian. As carnelian probably does not naturally occur on the Greek mainland or in the Aegean isles, the carnelian beads or at least the raw materials were obtained from more distant regions where good quality carnelian is available, such as Anatolia, Iran, and Arabia (e.g., Oman, Yemen, Saudi Arabia; Moorey 1994; Hausleiter 2011), Egypt (Moorey 1994; Harrell 2012) or the Indus region (e.g., Gujarat, among others; Trivedi 1964; Law 2011). Hard stone beads, especially of carnelian, seem to have been highly valued in Antiquity for their intrinsic beauty, magical/medicinal properties, exotic status as exchange items, and the skilled workmanship displayed in their creation (Helms 1993: 4-5; Kenoyer 2003a; Frenez et al. 2016; Pieniążek 2016a; Frenez 2018; Ludvik 2018). Ancient Sumerian texts, for example, explicitly refer to carnelian as both “precious and desirable” (Ludvik 2018: 279). Given the high value of carnelian, it is not surprising to find the beads in association with objects made of gold and silver. Together, then, the contents of the jewellery hoard at Aegina represent materials of high social value and wealth, and also reflect the need to protect these materials for recycling or reuse through burial.

6These beads were originally published in the Austrian expedition’s report on the hoard from Aegina-Kolonna (Reinholdt 2008: 16, 52-54, 117 and 127). Measurements were taken, shapes were generally described, and some interregional stylistic parallels were noted. Reinholdt identified one “etched” (bleached) carnelian barrel bead at Aegina-Kolonna, correctly taking this to be evidence that the bead was derived from the Indus Valley craft tradition. This was based on the decorative “etching” (bleaching) technique immediately recognisable on its exterior surface. The bead, here termed AK-1, was “decorated with three groups of two concentric circles, a design particularly well known from Kish in Mesopotamia” (Collins in Reinholdt 2003a: 261; see also Reinholdt 2008: 54). This decorative technique is indeed closely linked to Indus workshop traditions of bead manufacture and decoration (Kenoyer 1998; Prabhakar 2018). Additional observations were made in Collins’ assessment about the processes that shaped this bead to its present form, significant for their implications for its regional origin: “Because the string sides of the bead have been recut, which probably removed part of the white pattern, [and given the close parallels in shape and decoration] it is possible that the bead was imported from Mesopotamia and adapted for use in local jewellery” (Collins in Reinholdt 2003a: 261). This is consistent with other finds in the hoard. The gold disk pendant mentioned above is perhaps the best example of such adaptation. Collins also suggested that one other long barrel carnelian bead, here labelled AK-2, seemed, “vaguely reminiscent of Indus long biconical beads” (Collins in Reinholdt 2003a: 261).

7In 2014, Ludvik and Pieniążek were given permission by the curators of the Piraeus Museum in Greece to reanalyse the Aegina-Kolonna collection at the museum (fig. 1). The main goal of this study was to use high-resolution analyses to investigate the technology of manufacture of the beads including tool use and technique, as well as links between the Indus and other regions, methods used successfully in Mesopotamia (Kenoyer 1997; 2003b) and Oman (Kenoyer and Frenez 2018a; Frenez 2018; Kenoyer and Frenez 2018b).

Methodology

8The twenty beads from Aegina-Kolonna were examined according to methods commonly employed by Kenoyer and those trained by him around the world. A full discussion of this methodology can be found elsewhere (Kenoyer 2017b; Ludvik 2018), so it will only be summarised here. After photography, the beads were seriated into stylistic types based on measurements taken for all major bead metrics (fig. 2) and external appearance using Kenoyer’s simplified version of H. Beck’s (1928) bead typology, which describes beads by their length relative to thickness (“long” versus “short”) and the appearance of their longitudinal section in profile (i.e., “barrel”, “bicone”, etc.).

9Two metrics, in particular, have been consistently shown by previous studies to provide useful summaries of morphology and proportions: length (end to end) to width (maximum width at centre) ratios compared to average drill hole diameters (Ludvik 2018). After this, each bead’s raw material was described generally based on colour and inclusions visible according to the protocol used by Kenoyer at Harappa, and exterior surface modifications (chipping, grinding, polishing, etc.) were recorded to reconstruct the shaping processes that led to the production of a finished bead (Kenoyer 2003a, 2017a, 2017b).

10After cleaning with distilled water, the drill hole impressions were taken of nineteen of these beads; AK-17 was not sampled given the small silver tubular bead stuck in one of its perforation holes, though it was carefully examined. Impressions were made using 3M vinyl polysiloxane, a nondestructive silicon putty that produces fine, accurate casts of the internal surfaces of each drill hole (Kenoyer 2017c). Using the drill hole impression, it is possible to observe diagnostic features of drilling and bead use through traces of abrasion and polish from wear. Two different tests were done. For one, the impressions were measured under a digital microscope and the amount of drill hole taper per millimetre of drilling depth was calculated (= width _initial – width _final / distance). The resulting taper/mm values are diagnostic of specific types of drill raw materials and abrasives, since different drill bits wear at different, quantifiable rates depending on the material (Ludvik 2018). This may vary based on the amount of use wear on an individual drill, but material trends are evident nonetheless (Ludvik 2018). A separate set of impressions were examined at both low (20x and 50x) and high (100x and 300x) magnification under a scanning electron microscope (SEM). High-resolution SEM images were compared to images taken from experimentally drilled beads made by Kenoyer and those whose methods of drilling were known, resulting in a confident identification of drill type and method (Kenoyer 1997, 2003a, 2008, 2017b, 2017c).

11Each stage of our project methodology is designed to provide a holistic quantitative and qualitative description of bead stylistic, morphometric, and technological features. Using these data, it was then possible to identify and group beads based on those specific features that mark them as products of the same idiosyncratic production process. The main variables are similar raw material selection, shared styles/forms (i.e., shape), closely-matching metric proportions (i.e., size), similar production sequences (chaîne opératoire), and similar patterns of tool use. Such beads are considered products of the same traditions of crafting, or even physical workshops, according to the workshop tradition model advanced by the authors and their colleagues for beads and other materials such as steatite seals (Kenoyer 2008; Jamison 2017; Ludvik 2018).

Production processes and workshop traditions

12Our analyses indicate that most of the hard stone beads from the Aegina-Kolonna hoard shared stylistic, morphometric, and technological characteristics to such an extent that the different types identified should nonetheless be considered products of the same workshop tradition of manufacture. In terms of stylistic results, a total of eighteen out of the twenty beads were truncated biconical barrels, either long (n=9) or short (n=9). The remaining two beads (AK-5 and AK-18; fig. 1), one made of carnelian and one of rock crystal, were very short barrels. In terms of mineralogy and appearance, the nineteen carnelian beads were all made from nearly identical raw material: slightly translucent deep red-orange carnelian, with slight variation in colour due to fluctuations in heat-treatment temperature. Some beads were deeper and more brownish red/purple in colour, while others were more a vibrant reddish-orange. However, the very high quality of each carnelian bead, slight translucence with few to no impurities, internal cracks, or other quartzite inclusions, suggests a common mineralogical origin for the beads, and an exceptional one at that. This mineralogical evidence suggests a careful selection of carnelian by the artisans involved, who appear to have preferred the same type and quality of carnelian, and used them for the execution of several related bead forms. These carnelian varieties exhibited rich colour, slight translucence, and few to no internal flaws—characteristics of ornamental stones that are often considered beautiful and representative of superlative quality in workshops today (Kenoyer 1998, 2003a, 2017a, 2017b) and potentially in the past (Ludvik 2018). Such high-quality sources of carnelian are most common at Gujarat in western India (Trivedi 1964; Inizan 2000; Roux and Matarasso 2000; Law 2011; Golani 2013), but have also been reported in Egypt (Moorey 1994; Harrell 2012).

13Examination of external shaping procedures indicated that all three forms of beads identified were produced using nearly the same sequence of production (chaîne opératoire). The nineteen carnelian beads were all chipped into shape, with flakes removed from a central ridge, then finely ground with some striae still visible in areas, and well-polished with a high luster. The carnelian very short barrel bead ends were chipped and coarsely ground with striae visible and it does not appear to have been shortened or reground during Antiquity. The rock crystal bead was also chipped, ground finely, and well-polished with chipped, ground, and worn ends, also not shortened or reground. Biconical barrel bead ends were extremely smooth and lustrous, with evidence for breakage (either accidental or deliberate, with the result of shorter beads than originally planned) and subsequent regrinding at slight angles. These reground sections were polished further through extensive use after repair or modification. This is itself indicative of a long use life for the beads; the Aegina-Kolonna beads seem to have been worn and used or reused for a considerable period of time prior to their final burial in the hoard. Drill hole wear indications also support this, both interior and exterior. All SEM images of beads from Aegina-Kolonna have signs of polishing from wear on their ends. All interior drilled surfaces except that of AK-10 also displayed similar evidence. Wear on the exterior bead ends was present clearly on all samples, visible under SEM in the form of highly polished, smooth surfaces perpendicular to the drilling channels. Figure 3 shows four examples of worn bead surfaces, interior and exterior. For the biconical beads in particular, this suggests the reworking and grinding of originally longer beads, which were subsequently worn and rubbed against one another while strung for a fairly long period of time. On the interior, wear was most clearly visible on the surfaces where drill tips from the two perforation channels joined together. Apart from AK-10, which featured a fresh and jagged surface along the entire drill section, all Aegina beads displayed at least some areas of smoothed polish in these chipped regions. This is interpreted as evidence for moderate to heavy string wear on what was originally a rough chipped surface at the joining of two drill holes. String wear on the drill channels proper was difficult to identify on beads perforated with stone drill bits, since the surface was so highly polished to begin with. However, four of the five beads that were pecked or drilled with abrasives displayed moderate to heavy interior polish from wear that smoothed sections of their perforation holes. The entire corpus therefore seems to have been used and actively worn well prior to deposit in this hoard. In all, these stylistic, external shaping, and mineralogical considerations together suggest that, despite slight differences in shape, these beads seem to have shared major characteristics, including likely derivation from the same high-quality carnelian raw material sources and the same or very similar exterior shaping, grinding, and polishing procedures.

Photos G.E. Ludvik and J.M. Kenoyer

Morphometric Sub-groups 1-3

14The similarities between beads found in the Aegina hoard were further investigated by the comparison of specific morphometric proportions (i.e., length to width ratios and drill hole diameters). By combining this with stylistic and shaping evidence, we identified three morphometric sub-groups within the Aegina-Kolonna bead corpus:

• Group 1: group of three beads with small length to width ratios and very large perforations, two of which were very short barrel beads, one of which was a short biconical barrel in shape—closely related to beads of Group 2.
• Group 2: group of long biconical barrel beads with a quite variable length but similar drill hole sizes.
• Group 3: group of smaller biconical barrel beads, shorter than Group 2 with much less varied proportions and the same clustered drill hole sizes as Group 2.

15SEM and taper/mm analyses indicated that different perforation technologies were utilised on these beads and that some of the beads morphometrically similar were nonetheless drilled with different methods. These same analyses, however, revealed that fifteen beads of the morphometric Sub-groups 2 and 3 were all drilled with the same type of drill, used the same way.

Technological Sub-groups A-C

16As a result of drill hole examination, three different drilling technologies were identified: A) pecking (n=3), B) emery abrasives and a solid copper drill (n=2), and C) Indus-style constricted cylindrical stone (probably “ernestite”) drills (n=15), with some longer and larger (C1) and some smaller and shorter (C2). These identifications were made by comparison of SEM images to those of known identity identified by J.M. Kenoyer (1997, 2017a, 2017b, 2017c) and comparing drill hole taper/mm values determined for Aegina-Kolonna beads and those from other regions, namely the corpus found in the Southern Levant (Ludvik 2018). Figures 4-6 display these SEM images taken at 25x, 50x, 100x, and 300x magnification sequentially. For experimental discussions and in-depth protocols, see Kenoyer’s detailed treatment the subject (2017c).

Technological Sub-group A

17SEM imaging revealed three beads from Aegina-Kolonna were perforated via pecking with a stone percussor (i.e., chiseling through the stone by striking a stone tool). This technology has been experimentally replicated by J.M. Kenoyer. His studies have suggested pecking was accomplished with chipped stone points repeatedly struck against bead blanks to produce the deep, broad, conical perforation holes noted in the archaeological record (Kenoyer 2017c). The pecked beads from Aegina-Kolonna were AK-5, AK-10, and AK-18 (the only rock crystal bead in the corpus). AK-10 was stylistically a short barrel bead, not as short as the other two and, apart from its perforation method, looked virtually indistinguishable from the others of that form (Group 2). AK-5 was also very similar in overall morphology to the short biconical barrel beads, its perforation and smaller length to width ratio excepted. The pecking technique utilised on these three beads was the same, suggesting the use of similar tools. This left relatively narrow, concave perforation channels and jagged internal surfaces. In Ludvik’s (2018) analyses of late 3^rd millennium carnelian beads in the Southern Levant, this was one of the more common variants of the technique, especially noted on beads from sites in northern Syro-Palestine made of similar high-quality carnelian material and which, much like these from Aegina-Kolonna, consistently co-occurred with beads drilled with constricted cylindrical stone drills.

18Figure 4a-b displays SEM images of two of the three pecked beads. Note the coarseness of the internal surface and the narrow, concave channel. Based on Kenoyer’s extensive studies (1997, 1998, 2003a, 2017a, 2017b, 2017c), this is considered diagnostic of the use of a stone percussor (perhaps of flint or jasper), struck repeatedly to break through the carnelian.

Technological Sub-group B

19A second technology SEM imaging indicated was solid copper drilling with a hard abrasive that appears to be similar to emery abrasives documented through experimental drilling by J.M. Kenoyer. Figure 5 depicts the SEM images of bead AK-16, the sole example of this drilling method whose impression was accessible; bead AK-17 was also probably perforated in this manner, though the silver tubular bead stuck in one end made it impractical to acquire a drill hole impression. In addition to its precise stylistic and close morphometric similarities, visual examination of the internal section of AK-17 confirmed a morphology similar enough to AK-16 to merit classification as a second example of emery abrasive use. The two drill sections of AK-16 were long tapered cylindrical in shape, with drilling proceeding from both sides, though almost entirely from one side. The two channels met off centre at a sharp angle. This tapering drill hole suggests the use of a solid metal drill, probably copper or bronze, which was used with an abrasive slurry. Drilling with hard abrasives is documented from beads at Indus sites such as Harappa and Dholavira (Kenoyer 1997: 272; 2017a: 140), but the precise identification of the abrasive is still not determined. The most common abrasive in the Eastern Mediterranean was emery, a very hard, powdered form of corundum that would provide significant cutting power during drilling when used with a soft metal drill. The use of emery abrasives has been documented from at least the early 3^rd millennium BCE onwards through the Roman/Byzantine period in the Near Eastern world (Ludvik 2018). The relatively coarse abrading visible on the interior surface at higher magnification is diagnostic of emery abrasives, though very heavy string wear seems to have polished the surface and smoothed the abraded areas considerably, implying a long use life. Apart from the use of a different drill and abrasive, this bead was stylistically, morphometrically, and mineralogically indistinguishable from other long biconical barrel beads at Aegina-Kolonna. Despite the difference in drilling technique, it seems likely that AK-16 and probably AK-17 were made in the same workshop tradition as the other biconical barrel beads from Aegina-Kolonna. Further studies of the drill taper measurements for abrasive drilling with a much larger sample are needed to determine if the abrasive drilling can be linked to Indus technology or Mediterranean/Anatolian technology.

Technological Sub-group C

20The remaining fifteen beads were all drilled with Indus-style constricted cylindrical stone drills, probably made of the unique raw material called “ernestite” only known from the Indus Valley region (Kenoyer and Vidale 1992). The extremely low taper/mm values, as well as the straightness of the drilling channels, initially supported this, and SEM analyses confirm the Indus-related character of these Aegina beads. The use of constricted cylindrical stone drills has been discussed in detail elsewhere (Reade 1979; Kenoyer and Vidale 1992; Kenoyer 1997, 1998, 2000, 2003a, 2003b, 2008, 2017b; Inizan 2000; Roux and Matarasso 2000; Vidale 2004; Ludvik et al. 2014, 2015; Frenez 2018: 391-392; Ludvik 2018) and this technology has been documented as highly diagnostic of the craft traditions of the Indus Valley civilisation during the Harappan period (2600-1900 BCE). Beads made with these drills were widely circulated throughout the 3^rd millennium world, from Central Asia to the Southern Levant, as far west as Aegina-Kolonna and in the Shaft Graves of Mycenae (n=2), currently under investigation by the authors. Such beads may have been produced within the Indus proper, but also may have been produced to suit local demand and in slightly different forms in Indus-derived craft workshops in Mesopotamia (Kenoyer 1997: 272).

21Interior drill hole sections of these Indus-style beads were straight cylindrical in shape, proceeding either from one direction with the remainder of the carnelian popped through or drilled from both sides with the same drills. All exhibited diagnostic signs of Indus-style constricted cylindrical stone drills: extremely straight, highly polished, smooth interior surfaces with smooth-ridged undulations and slight offsets exposing chipped sections from drill realignment, and constricted-shaped channels at the tips where drilling terminated. Excellent parallels in style, technology, raw materials, and proportions come from Indus-style carnelian beads from: Troy, the Southern Levantine cemeteries at Tell el-Ajjul, Moza’Illit, Silwan, Kedesh, Gezer, and Azor, as well as at Mesopotamian sites like Kish, Mari, and the Royal Cemetery at Ur, and others in the Persian Gulf region, such as at Bid-Bid, Oman, and in the Indus Valley proper, such as Harappa, Mohenjo-daro, Dholavira, and Chanhu-daro (Reade 1979; Kenoyer 1997: 267 and 270-274; 1998: 97-98, 143 and 160-162; 2003a: 16-17; 2008: 21-22 and 25-26; 2017a: 130-140; 2017c: 416-417 and 422-425; Ludvik et al. 2015; Frenez 2018: 390-2; Ludvik 2018). These Indus-style bead characteristics were present in a paradigmatic form on all fifteen examples from Aegina. These beads appear to be products of an Indus-related workshop tradition. The close similarity between the Aegina-Kolonna beads and Indus-style beads found in Mesopotamia and the Southern Levant, lead us to suggest that they may have been produced in Mesopotamian workshops operating with Indus-derived tools, technologies, styles, and methods. Figure 6a-d depicts SEM images of four Indus-style beads from Aegina: two long barrel (AK-2 and AK-7) and two short barrel (AK-1, the “etched” bead, and AK-21). Note the very straight sides, the smooth and highly polished surfaces, and the constricting section at the terminus of drilling.

22As indicated by the SEM studies, three different techniques of drilling and drill raw materials were identified by taper/mm analysis. While comprehensive studies of regional differences in tapered drilling have not yet been completed, this is part of the long-term study of drill technology we are conducting. Here, the taper/mm values are used only to refer to differences in overall drill type and are not used to examine regional differences within types; this data is currently being compiled. The values calculated for drill hole taper/mm are displayed in Table 1. Note how closely these match values calculated for beads whose drilling agents have been determined previously: Aegina Group A = 1.589, pecking technique = 1.702; Group B = 0.368, solid copper with emery abrasives = 0.333; Group C1 = 0.029, constricted cylindrical drilling I (long) = 0.031; Group C2 = 0.051, constricted cylindrical drilling II (short) = 0.060. It is therefore reasonable to consider the Aegina-Kolonna beads to have been perforated with these same types of drills.

Discussion and conclusions

23Our reanalysis of the hard stone beads from the small jewellery hoard at Aegina-Kolonna indicated the use of a standard set of bead stylistic, morphometric, and mineralogical characteristics for most, if not all, twenty beads, despite differences in drilling method. We have demonstrated that three different technologies (pecking, solid copper with hard [emery?] abrasives, and constricted cylindrical stone drills) were employed to create three morphometric varieties. Combined, these parameters were used to suggest the presence of three techno-stylistic groups of hard stone beads in the Aegina-Kolonna hoard:

• Group I consists of two very short barrel beads that were pecked (i.e., combined sub-groups 1 and A);
• Group II consists of nine long biconical barrel carnelian beads, seven of which were drilled with Indus-style constricted cylindrical stone drills, two of which were drilled with solid copper and hard (emery?) abrasives (combined Sub-groups 2 and B/C1);
• Group III consists of nine short biconical barrel carnelian beads, eight of which were drilled using Indus technologies (combined Sub-groups 3 and C2) and one of which was pecked using the same method present on the pecked beads of technological Sub-group 1.

24Results of this appear in Figure 7. Note the distinct clusters of values and their correlation with different techniques from SEM and taper/mm analysis.

25Using this evidence, we have suggested regional workshop tradition identifications for the twenty hard stone beads at Aegina-Kolonna. Based on the results of our study, Groups II and III are identified as Indus-style and thus as products of an Indus-related workshop tradition, either in operation in the Indus heartland or abroad in the form of Indus-trained and equipped craftspersons, perhaps in Mesopotamia. At the very least, our study has demonstrated that fifteen beads from Aegina display the tool marks characteristic of highly diagnostic Indus-style constricted cylindrical drilling with “ernestite” stone drill bits. These were very similar to each other (and to other Indus-style carnelian beads identified elsewhere, for that matter) in overall chaîne opératoire, especially morphometric shape, style, raw materials, and type, size and morphology of the drills employed. According to the workshop tradition model, then, the closely shared features demonstrated among the fifteen Aegina beads perforated with Indus-style constricted cylindrical drills strongly suggest shared origins in a single workshop tradition and, possibly, the same physical workshop.

26The five beads that were not drilled using Indus-style constricted cylindrical drills nevertheless seem to have shared origins in the same workshop tradition as those that were; given the very close similarities between these beads and Indus-style beads in all features but drill type, it seems reasonable to postulate that all three groups may have shared a common origin. They may represent the contents of a single necklace originally produced in an Indus-related workshop, albeit one in which craftspersons utilised more than one drilling technique. All five demonstrated close parallels to beads frequently co-occurring on Indus-derived necklaces in the Southern Levant, not only in style and proportions but also in raw material and perforation methodology. At both Tell el-Ajjul and Silwan cemeteries in the Southern Levant, dated to the same time period as Aegina-Kolonna (ca. 2200 BCE), for example, the same variety of narrow, concave pecking and nearly indistinguishable shapes and raw materials were used to produced pecked beads that provide the best parallels to Group I. Figure 8 compares these beads.

Photo G.E. Ludvik

27Parallels are so close that it is possible such beads may have all derived from one and the same Indus-related workshop tradition; it is possible that products of the same Indus-style workshop have been identified in Greece and Syro-Palestine. The morphology of the Aegina emery-drilled perforation sections closely matched those found alongside Indus-style beads at Motza Illit and at Tell el-Ajjul. Parallels aside, the quantitative evidence from the Aegina-Kolonna collection alone suggests these five beads may likewise have originated in the same workshop tradition. For three beads in particular the only noticeable difference from Indus-style beads was in drill method (i.e., the two emery-drilled beads and the one pecked bead), and both of these methods are known in the Indus craft repertoire. Given the close shape, proportion, style, and raw material of the other pecked carnelian bead (AK-5, compare to AK-4 and AK-21), it is therefore possible that all nineteen of the carnelian beads may have been Indus in style, produced in the same workshop tradition. Since the rock crystal bead was perforated in precisely the same manner as the other two pecked beads, this suggests that it was likely made with the same type of tool as the others. This raises the possibility that it was produced in the same workshop tradition despite its difference in appearance, or was derived from a separate tradition of manufacture, but one using the same types and shapes of tools as the other two. It is therefore possible that all of the beads from this hoard may have been traded as a single necklace, acquired from a workshop producing Indus style beads with multiple tools and techniques. However, we cannot rule out the possibility that some of the beads from Groups A and B did come from other workshops/areas. This is most conceivable with the rock crystal bead, which has a different shape and raw material than the carnelian beads. Rock crystal was available in the Aegean region and was worked during the same general time period; many objects and raw material pieces are known for example from Troy I‒III (Pieniążek 2016b: 139-143).

28While earlier studies had already proposed that one bead from Aegina-Kolonna was linked to the Indus (Reinholdt 2003a: 260), our analyses have identified the presence of many more Indus-style beads in the same assemblage than were previously known. We do not know if the hoard belonged to the original necklace owner or someone who had collected a large amount of valuable goods that were hastily buried during a period of instability. However, Indus-style carnelian beads in circulation during the 3^rd millennium were probably highly valued and may have held particularly important economic, social, or symbolic-ritual value. The presence of the beads at the site indicate that the original owner would have had access to the interaction networks bringing such goods to Aegina from production centres in South Asia or Indus-related workshops in Mesopotamia. This study also shows the utility and urgent need of re-examining stone beads from earlier collections to better understand the larger pattern of interaction between ancient Greece and Asia as a whole.