Analysis of early Iron age glass beads (8th to 7th c. BC) from the Tsaishi necropolis (Georgia)

• This paper shows the importance of trace elements such as B, Li and REE for the study of early glass production and trade. • The data challenges the assumption that high Ce glass were exclusive to Italian assemblages of the 9 th -8 th centuries BC. • They establish similarities between bird-shaped glass beads from Tsaishi (Georgia) & Cuma (Italy), both being high Ce glass. • It confirms the presence of a new group of glasses with high Mg K P contents as early as the beginning of the 8 th century BC. • It suggests that Colchis was inserted in a broad glass beads distribution network both to the south and the west. We present the analytical results obtained by LA-ICP-MS from a set of 38 glass beads of various colours dating from the 8th to the first half of the 7th century BCE, found at Tsaishi cemetery, located in the central area of historical Colchis (Georgia). Through the comparison with published analytical data of coeval, earlier and later glass samples, the aim of this study is to investigate the provenance of Tsaishi assemblages in order to highlight the cultural influences and trading networks of the Colchis community between the 8th and the 7th centuries BC. Tsaishi glass beads predate previously studied glass beads from Brili necropolis, located in the mountainous part of Colchis (North-west Georgia). In contrast to the compositions of the Brili glasses, the measurements of the Tsaishi glass beads do not allow for the identification of the well-established Iron Age compositional groups. Furthermore, in contrast to the other Early Iron Age sites, they do not appear to indicate the presence of typical Egyptian or Syro-Palestinian natron glass productions. Although our assemblage shows considerable compositional variability, it highlights the existence of a new group of glass recently identified in Italy and characterised by a positive cerium anomaly, as well as a group of high boron and lithium glasses. However, despite the large variation in contents for minor elements, colourless and purple glass are all part of the same group. Comparison with other Early Iron Age glass studies points to a fairly complex situation, probably due to small scale production of raw glass and glass beads, the multitude of production sites and the diversity of recipes used by glass workers at the beginning of the first millennium B.C.