Original article SreNd isotopic analysis of glass from Sagalassos (SW Turkey) Patrick Degryse * , Jens Schneider, Veerle Lauwers, Dieter Brems Section Geology, Centre for Archaeological Sciences, K.U. Leuven, Celestijnenlaan 200E, bus 2408, BE-3001 Leuven, Belgium Received 30 March 2008; accepted 29 July 2008 Abstract The Nd isotopic composition of an ancient glass is typical of the (heavy) non-quartz fraction in the silica raw material. The Sr isotopic composition is considered typical for the lime component in the raw material. In the case of natron glass, this is mostly attributed to shell or limestone in the silica sand. In this study, 1st to 5th and 6th to 7th century AD glass from Sagalassos is compared to the typical signature of 4th to 8th century AD primary glass with a known production location in Egypt and Syro-Palestine. The results suggest a difference in glass supply to Sagalassos between both periods, and that glass was produced in the eastern and western Mediterranean area at least before the 6th century AD. Ó 2008 Elsevier Masson SAS. All rights reserved. Keywords: Archaeometry; Mineral resources; Natron glass; Neodymium; Radiogenic isotopes; Roman; Sand; Strontium 1. Introduction Natron glass was the predominant type of ancient glass in the Mediterranean and Europe from the middle of the first millennium BC through to the ninth century AD [1]. The main raw materials used in its production were mineral soda and quartz sands rich in lime. Natron glass was traded throughout the known world in the form of raw glass or chunks, to be remelted and formed elsewhere [2]. Hence, primary work- shops made raw glass and were distinct from secondary workshops that shaped glass. This division of production implies that glass compositions reflect the primary source, rather than the secondary workshops. Thus, while for ceramics elemental analysis is routinely employed for provenance determination of raw materials, in glass studies this approach is insufficient [2]. It is clear from excavation that large quantities of raw natron glass were made in Egypt and Syro-Palestine in the fourth to eighth century AD [2]. Suggestions towards the existence of similar units elsewhere in the early Roman Empire can be found with the ancient author Pliny. He writes in his Natural History (the only surviving written account of Roman glass making) that besides Egyptian soda and Levan- tine sands, sands from the coast of Italy and from the Gallic and Spanish provinces were used. Using SreNd isotopes as an analytical technique, it was shown that indeed in the Western Roman Empire, at least from the first until the third century AD, primary production of glass must have taken place [3]. 2. Research aims Shaping or secondary production of glass is known to have been carried out at Sagalassos, in modern-day Turkey [4]. Through isotopic analysis of vessel glass from different strati- graphic layers of the excavations in the ancient city, variations through time in the import of raw glass (or readily-made objects) from Egypt or the Levant or other production sites in the west into the local glass workshop(s) were investigated. 3. Materials and methods 24 samples of vessel glass from well-dated contexts were analysed for their SreNd isotopic composition [3,5]. 12 samples were dated to the first to fifth century AD, 12 samples were dated to the sixth to seventh century AD. For isotope * Corresponding author. Tel.: þ32 16326460; fax: þ32 16322980. E-mail addresses: patrick.degryse@geo.kuleuven.be (P. Degryse), jens. schneider@geo.kuleuven.be (J. Schneider), veerle.lauwers@arts.kuleuven.be (V. Lauwers), dieter.brems@geo.kuleuven.be (D. Brems). 1296-2074/$ - see front matter Ó 2008 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.culher.2008.07.002 Available online at www.sciencedirect.com Journal of Cultural Heritage 9 (2008) e47ee49