STUDIES IN ARCHAEOLOGICAL SCIENCES Isotopes in Vitreous Materials Patrick Degryse, Julian Henderson, Greg Hodgins (Eds) Leuven University Press Isotopes in Vitreous Materials Reprint from: Isotopes in Vitreous Materials - ISBN 978 90 5867 690 0 - Leuven University Press 2009 isotopes_030209.indd 1 12 feb 2009 15:12 Studies in Archaeological Sciences 1 The series Studies in Archaeological Sciences presents state-of-the-art methodological, technical or material science contributions to Archaeological Sciences. The series aims to reconstruct the integrated story of human and material culture through time and testiies to the necessity of inter- and multidisciplinary research in cultural heritage studies. Editor-in-Chief Prof. Patrick Degryse, Centre for Archaeological Sciences, K.U.Leuven, Belgium Editorial Board Prof. Ian Freestone, Cardif Department of Archaeology, Cardif University, United Kingdom Prof. Carl Knappett, Department of Art, University of Toronto, Canada Dr. Andrew Shortland, Centre for Archaeological and Forensic Analysis, Cranield University, United Kingdom Prof. Manuel Sintubin, Department of Earth & Environmental Sciences, K.U.Leuven, Belgium Prof. Marc Waelkens, Centre for Archaeological Sciences, K.U.Leuven, Belgium Reprint from: Isotopes in Vitreous Materials - ISBN 978 90 5867 690 0 - Leuven University Press 2009 isotopes_030209.indd 2 12 feb 2009 15:12 Isotopes in Vitreous Materials Edited by Patrick Degryse, Julian Henderson and Greg Hodgins Leuven University Press Reprint from: Isotopes in Vitreous Materials - ISBN 978 90 5867 690 0 - Leuven University Press 2009 isotopes_030209.indd 3 12 feb 2009 15:12 © 2009 by Leuven University Press / Presses Universitaires de Louvain / Universitaire Pers Leuven. Minderbroedersstraat 4, B-3000 Leuven (Belgium). All rights reserved. Except in those cases expressly determined by law, no part of this publication may be multiplied, saved in an automated dataile or made public in any way whatsoever without the express prior written consent of the publishers. ISBN 978 90 5867 690 0 D / 2009 / 1869 / 1 NUR: 682-971 Lay-out: Friedemann BVBA (Hasselt) Cover: Jurgen Leemans Reprint from: Isotopes in Vitreous Materials - ISBN 978 90 5867 690 0 - Leuven University Press 2009 isotopes_030209.indd 4 12 feb 2009 15:12 Neodymium and strontium isotopes in the provenance determination of primary natron glass production Patrick Degryse, Jens Schneider, Veerle Lauwers, Julian Henderson, Bernard Van Daele, Marleen Martens, Hans (D.J.) Huisman, David De Muynck, Philippe Muchez Introduction The great majority of ancient glass was chemically based upon silica luxed with soda or potash. The earliest known glass was found in Late Bronze Age Mesopotamia and Egypt. It was a soda-lime silica glass, and this type predominated across Western Asia and the Mediterranean right up to the modern period (Freestone 2006). Chemically, ancient soda-lime-silica glass falls into two categories (Sayre and Smith 1961): (1) plant ash glass, combining a plant ash with quartz pebbles, and (2) natron glass, combining soda-rich mineral matter with quartz sand. Natron glass was the predominant type of ancient glass in the Mediterranean and Europe from the middle of the irst millennium BC until the 9th century AD (Henderson 1989, Freestone et al. 2002a, Henderson 2003, Shortland 2004). Work by Foy et al. (2003) suggests that there are likely to have been around 10 major glass groups in the Mediterranean and Western European region between the 1st and 9th centuries AD. Before that time plant ash glass was also produced, mainly in Egypt and Mesopotamia. Throughout the Mediterranean and Europe, however, using plant ashes as a lux became dominant practice only from the 9th century onwards (Henderson et al. 2004, Freestone 2006). Initially it was assumed that glass was made in the same workshops where the vessels, windows etc. were being formed. However, the discovery of raw glass in the form of ingots in the Late Bronze Age (Nicholson et al. 1997, Rehren and Pusch 1997) and as lumps of glass (chunks) in the Roman and early medieval periods (Foy et al. 2000) suggests the export of glass chunks as an economic commodity. Primary workshops which made the raw glass were, in many cases, clearly distinct from the secondary workshops which shaped the glass vessels. A single primary workshop could then supply many secondary workshops over a large geographical area (Gorin-Rosen 2000, Nenna et al. 2000). However, it is necessary still to be cautious about applying such models too rigidly to ancient economies, and to some extent the separation of glassmaking and glassworking activities may have been dependant on the scale of production. For example, in an (inland) urban Reprint from: Isotopes in Vitreous Materials - ISBN 978 90 5867 690 0 - Leuven University Press 2009 isotopes_030209.indd 53 12 feb 2009 15:13 54 Patrick Degryse, Jens Schneider et al. environment why would primary and secondary glass production necessarily be separated physically by any great distance? If a glassblowing workshop was set up close to a primary glassmaking furnace, then clearly raw glass could be supplied directly to the local glassblowers. Indeed evidence of this was discovered during the excavation of a 9th century AD Islamic extra-mural industrial complex at alRaqqa, northern Syria (Henderson et al. 2005a). This is not to say that raw glass manufactured at al-Raqqa was not also exported to other glassworking centres. For late Roman glass production, theories are centred on two models (Jackson et al. 2003). The irst states that contemporary natron glass production was divided between a relatively small number of workshops which made raw glass and a large number of secondary workshops which fabricated vessels (Freestone 2006). It is clear from excavation that large quantities of natron glass were being made from its mineral raw materials in a relatively limited number of primary glass production centres mainly in Egypt in the 1st to 3rd centuries AD and Syro-Palestine in the 4th to 8th centuries AD (Brill 1988, 1999, Freestone et al. 2000, 2002a, Picon and Vichy 2003). Suggestions that similar units existed in the Levant in early Roman times have only recently been proven, with the discovery of early Roman primary glassmaking furnaces in Beirut, Lebanon (Kouwatli et al. 2008). It has been argued that Roman blue-green glass and later glass produced in the Levant are suficiently similar for it to be likely that Roman glass was made there (Nenna et al. 1997, Picon and Vichy 2003, Foy et al. 2003), although archaeological and scienctiic evidence is dificult to interpret (Baxter et al. 2005). Some authors have suggested that early Roman primary production may have taken place elsewhere (Leslie et al. 2006, Jackson et al. 2003). Moreover, the second model of late Roman glass production proposes the existence of local glassmaking and -working centres (Wedepohl et al., 2003). Also, there is evidence which supports the manufacture of primary glass in Roman Europe. The ancient author, Pliny the Elder, writing before 79 AD, indicates in his Natural History (Hist. Nat XXXVI, 194) that sands from the coast of Italy between Cumae and Literno near Naples and the ‘Spanish and Gaulish provinces’ were also used (Freestone et al. in press). This, however, has never been conirmed by excavations, although the suitability of some of the sands explicitly described by Pliny has been suggested (Silvestri et al. 2006). The concept of a division of production leads to a very different interpretation of analytical data, so that glass compositions relect predominantly the primary glassmaking sources, rather than the secondary workshops in which the objects were made (e.g. Nenna et al. 1997; Foy et al. 2000; Freestone et al. 2000, 2002b). This model has signiicant implications for the study of ancient glass production based upon the chemical analysis of glass artefacts. While for several decades claybased ceramics have been routinely subjected to elemental analysis to determine Reprint from: Isotopes in Vitreous Materials - ISBN 978 90 5867 690 0 - Leuven University Press 2009 isotopes_030209.indd 54 12 feb 2009 15:13 Neodymium and strontium isotopes in provenance determination 55 provenance, the application of these methods to archaeological glass has thus proved far less tractable (Freestone 2006). The combined effects of the mixing of primary resources and the recycling of glass can stymie attempts to identify the origin of glass raw materials based upon elemental analysis alone. Glass provenancing GLASS PROVENANCING AND ELEMENTAL ANALYSIS A great deal of effort has gone into the major element analysis of glass (Brill 1999). For the most part, this has not led to meaningful groupings with respect to the geographical origin of the mineral resources. For example, all Roman glass was found to be relatively homogeneous natron glass with little variation in major element composition (Freestone 2006). Though signiicant advances have been made, progress toward an understanding of the exploitation of raw materials, technology and trade through main and trace element analysis remains limited (Freestone 2006). Most progress has been made in studying trace elements like lime, iron, magnesium and alumina, as they can be related to the concentrations of speciic minerals (feldspars, micas and clays) in the glassmaking sand. Trace elements in glass have been exploited to separate compositional groups, and the implication has been made that individual objects with these trace element signatures were produced from the same ‘batch’ (Freestone 2006). However, the presence of elevated transition metals has indicated that scrap glass, including small quantities of coloured glass, may have been incorporated into a batch, pointing to ‘recycling’ material, and this complicates the picture (Henderson 1993, Jackson 1997). Studies by Freestone et al. (2000, 2002b) and Aerts et al. (2003) have used trace elements as speciic indicators of the origin of glass raw materials. Huisman et al. (in press) used trace element composition to source decolorants (Sb) used in the production of roman colourless glass. GLASS PROVENANCING AND ISOTOPES Recent studies (Wedepohl and Baumann 2000, Freestone et al. 2003, Henderson et al. 2005, Degryse and Schneider 2008, Degryse et al. 2005, 2006a and b, Freestone et al. in press, Henderson et al. in press) have shown that the use of radiogenic isotope systems, speciically for strontium (Sr) and neodymium (Nd), has led to the development of new approaches in the provenancing of primary glass, even after its transformation or recycling in secondary workshops. Reprint from: Isotopes in Vitreous Materials - ISBN 978 90 5867 690 0 - Leuven University Press 2009 isotopes_030209.indd 55 12 feb 2009 15:13 56 Patrick Degryse, Jens Schneider et al. Sr in ancient glass is mainly incorporated with the lime-bearing material, being shell, limestone or plant ash (Wedepohl and Baumann 2000). Where the lime in glass was derived from Holocene sea shell, the Sr isotopic composition of the glass relects that of modern seawater (Wedepohl and Baumann 2000). Where the lime was derived from ‘geologically aged’ limestone, the signature of the glass relects that of the limestone, possibly modiied by diagenesis (Freestone et al. 2003, Henderson et al. 2005). Nd in glass is likely to have originated from the heavy mineral content of the sand raw material used. Rare earth element (REE) patterns have been suggested before as a means of distinguishing sand raw material sources (Freestone et al. 2002b). Nd isotopes are used as an indicator of the provenance of detrital sediments in a range of sedimentary basin types (Banner 2004). The Nd isotopic composition of the earth’s crust shows a wide variation, from εNd -45 to +12, but sediments tend to be homogenized so that the sedimentary loads of most of the world’s major rivers and airborne dusts vary between εNd of -16 and -3 (Goldstein et al. 1984). This is the range within which many glassmaking sands are likely to fall (Freestone et al. in press). Due to its geological and geographical variability Nd offers great potential in tracing the origins of primary glass production in ancient times. Moreover, the effect of recycling on the Nd composition of a glass batch does not seem to be signiicant (since there are no high Nd glasses which could modify the base composition of the glass, unlike e.g. lead isotopes affected by high lead glasses in re-melting), nor is the effect of adding colorants or opaciiers (Freestone et al. 2005). Though largely unexplored, Nd isotopes show great promise for addressing hypotheses regarding the primary production of glass in the Roman-Byzantine world. A irst example is given in the provenance determination of 4th to 8th century AD glass from Syro-Palestine and Egypt (Freestone et al. in press). Levantinetype glass of that era has a Nile-dominated Mediterranean 143Nd/144Nd signature, lower Nd content, and a high 87Sr/86Sr signature close to the Holocene seawater composition. Contemporary HIMT-type glass is made up of a mixture of a Levantine-type glass and an end member with a Nile-dominated Mediterranean 143 Nd/144Nd signature, higher Nd content and a low 87Sr/86Sr signature. The similarity of Levantine and HIMT glass in terms of 143Nd/144Nd signature (values between ε = -6.0 to –5.1), and the fact that these values are similar to Nile-dominated sediments (Weldeab et al. 2002, Stanley et al. 2003), strongly suggest that HIMT glass comes from an area extending from the Nile delta northwards to the Levant (Freestone et al. in press). A second study investigated the primary provenance of 1st to 3rd century AD natron vessel glass (Degryse and Schneider 2008). Different sand raw materials used for primary glass production in this period were distinguished and Reprint from: Isotopes in Vitreous Materials - ISBN 978 90 5867 690 0 - Leuven University Press 2009 isotopes_030209.indd 56 12 feb 2009 15:13 Neodymium and strontium isotopes in provenance determination 57 characterized by combined Sr and Nd isotopic analyses. Again, a Nile-dominated eastern Mediterranean Nd signature (higher than -6.0 ε Nd) characterized some glass, but a different Nd signature (lower than -7.0 ε Nd) was determined for a large number of samples, suggesting a primary production location in the western Mediterranean or north-western Europe. In this way, strontium and neodymium isotopes proved that Pliny’s writings were correct: primary glass production was not exclusive to the Levant or Egypt in early Roman days: other factories of raw glass, probably in the Western Roman Empire were in play. In this study, the primary provenance of Roman-Byzantine natron vessel glass from different sites in the eastern and western Roman Empire is investigated from the perspective of main elemental versus isotopic analysis. These isotope data obtained from the glass samples are compared with the main element data and the known signatures of primary production centres in the eastern Mediterranean. Methodology SAMPLING Samples were obtained from several locations in the Roman Empire through cooperation with the VIOE (Vlaams Instituut voor Onroerend Erfgoed - excavation at Tienen), the Rijksdienst voor Oudheidkundig Bodemonderzoek (the Netherlands – excavations at Bocholtz and Maastricht), the excavation at Kelemantia (Slovakia) and at Sagalassos (Turkey). A series of 47 glass samples were selected for both main element and Sr and Nd isotope analysis. Most samples represent free-blown vessel glass, but pressed or slumped plates were also analysed; various colours were selected by eye. Sample dates were determined by stratigraphical association. CHEMICAL ANALYSIS For main element analysis, samples were fused with a LiBO2 lux and then dissolved in 1N HNO3. Silicon, aluminium, iron, magnesium, calcium, titanium and phosphorus were determined by atomic emission spectrometry (AES) on a Spectrojet III spectrometer. Sodium and potassium contents were obtained from the same solutions by atomic absorption spectrometry (AAS) on a Varian Techtron AA6 spectrometer. Accuracy for both AAS and AES is better than 2%. Analytical precision at the 95% conidence level determined by replicate analysis was better than 0.5%. Detection limits were at the ppm level for both AAS and AES, but concentrations were expressed at the 0.01 % level. Data accuracy was evaluated Reprint from: Isotopes in Vitreous Materials - ISBN 978 90 5867 690 0 - Leuven University Press 2009 isotopes_030209.indd 57 12 feb 2009 15:13 58 Patrick Degryse, Jens Schneider et al. by analysis of the international standards Basalt BN-01 and GSJ-JB-1, Granite GN-02 and NIM-G, Lujavrite NIM-L, Feldspar NBS-99a and Gabbro MRG-1. For isotope analysis, samples were weighed into Telon screw-top beakers and dissolved in a 3:1 mixture of 22 M HF and 14 M HNO3 on a hot plate. Solutions were dried and dissolved in aqua regia. Aliquots of these solutions were spiked with a highly enriched 84Sr and 150Nd tracer for separate concentration analyses by isotope dilution, whereas unspiked aliquots were used for determination of isotope ratios. For separation of Sr and Nd from the same sample solutions sequential extraction methods developed by Pin et al. (1994) were utilized and slightly modiied. Sr and REE were separated using 2 M HNO3 using coupled miniaturized Telon columns containing 50 µl of EICHROM Sr and TRU resin, respectively, and eluted with deionized H2O. For separation of Nd, the REE cut was further passed through a column containing 2 ml EICHROM Ln resin. For this, the column was washed with 5.5 ml 0.25 M HCl after adding the sample. Nd was then stripped off using 4 ml 0.25 M HCl. All measurements were performed on a six-collector FINNIGAN MAT 262 thermal ionization mass spectrometer (TIMS) running in static multicollection mode. Sr isotopic ratios were normalized to 86Sr/88Sr = 0.1194, Nd isotopic ratios were normalized to 146Nd/144Nd = 0.7219. Repeated static measurements of the NBS 987 standard over the duration of the study yielded an average 87Sr/86Sr ratio of 0.71025 ± 0.00002 (2σ, n=22). Repeated measurements of the La Jolla Nd standard yielded 143Nd/144Nd = 0.511848 ± 0.000009 (2σ, n = 8). Total procedural blanks (n=6) did not exceed 30 pg Sr and 50 pg Nd and were found to be negligible. Archaeological context SAGALASSOS It has already been suggested that early Byzantine (6th-7th century AD) blue raw glass from Sagalassos was imported from several production sites in the Levant (Degryse et al. 2005, 2006a), whereas HIMT raw glass from Sagalassos corresponded very well to previously described material (Freestone et al. 2005), of which the primary production site is placed in Egypt (Freestone et al. in press). Conversely, early to late Roman glass at Sagalassos shows a distinctive major element composition, suggesting a different raw material mixture and possible different origin (Degryse et al. 2006b). The samples are representative of the common colour varieties of window and free-blown vessel glass from Sagalassos. The chronology was determined by stratigraphical association with Sagalassos red Reprint from: Isotopes in Vitreous Materials - ISBN 978 90 5867 690 0 - Leuven University Press 2009 isotopes_030209.indd 58 12 feb 2009 15:13 Neodymium and strontium isotopes in provenance determination 59 slip ware (Poblome 1999) and Sagalassos common wares (Degeest 2000). Glass from three distinct periods was sampled: imperial (1st-3rd century AD), late Roman (4th-irst half 5th century AD) and early Byzantine (second half 5th to 7th century AD). MAASTRICHT Sample Ma3a was retrieved from a grave in the Scharnweg in 1986. Besides pottery, several glass objects such as beakers and bowls were recovered. Typochronologically all the material can be dated to the irst half of the third century AD (Panhuysen and Dijkman 1987, p.212 and afb.11). Sample M5a was excavated in 1983 under the Hotel Derlon, in layers assigned to the second quarter of the 5th century AD (Dijkman 1993, Fig. 9-C1 and D8). KELEMANTIA The Roman auxiliary fort of Iža (Kelemantia) in Slovakia is situated about 4 km east of the conluence of the rivers Waag and Danube. A double ditch was uncovered, together with the remains of more than eleven barracks. The remains of the earth-and-timber fortiication all belong to one single construction phase dating between 175 and 179 AD. This secured dating was possible thanks to the discovery of several coins and terra sigillata pottery. Comparing the data obtained with historical texts made it possible to link the fort of Kelemantia with the Marcomannic Wars, waged between the Germanic Marcomanni and Quadi and emperor Marcus Aurelius’ troops. The wooden construction was laid to waste by German attackers in 179 AD or was dismantled, abandoned and set on ire by the Roman forces themselves when they left. A few years later, under Emperor Commodus’ rule, a stone castellum was built on exactly the same spot. This stone camp was occupied until the end of the reign of Valentinianus I in 375 AD, when that emperor died at Brigetio and the barbarians invaded the frontier zone. During the excavations substantial amounts of all kinds of material were found, including many glass fragments belonging to different kinds of glass objects like bottles, bowls, windows and pearls. Samples KEL 82, KEL 229, KEL 299 and KEL 300 belong to excavation layers of the earth-and-timber camp, dated to 175-179 AD. Sample KEL 234 comes from excavation layers in the castellum and was dated to the 3rd century AD. Reprint from: Isotopes in Vitreous Materials - ISBN 978 90 5867 690 0 - Leuven University Press 2009 isotopes_030209.indd 59 12 feb 2009 15:13 60 Patrick Degryse, Jens Schneider et al. BOCHOLTZ In 2003, a stone sarcophagus was found in Bocholtz (the Netherlands) as a part of an underground burial chamber, close to a known Roman villa (de Groot 2006). The chamber was dated to the last quarter of the 2nd to the irst quarter of the 3rd century AD. Glass grave goods were identiied and sampled for analysis. Sample BO 106 is free-blown colourless plate with a greenish tinge, Isings type 42b (Isings 1957), dated to the 2nd century AD. Sample BO 109 is a free-blown colourless cylindrical bottle, Isings type 51b (Isings 1957), dated to the late 2nd – early 3rd century AD. Sample BO 123 is a colourless cast or slumped small bowl. TIENEN The small Roman town of Tienen is situated in Belgium, and was part of the Roman civitas Tungrorum. It was founded during the reign of Claudius on the road from Cologne to Boulogne. Large-scale excavations in the periphery of the town revealed numerous pottery kilns, traces of iron workings and bronze casting and glass production. In 2001, a glass furnace dated to the 2nd century AD was excavated there (Cosyns and Martens 2002–2003). All samples analysed here belong to this context. They are samples of free-blown blue (aqua) vessel glass. Determinable pieces are fragments of Isings type 50 (square bottles) or Isings type 3 (ribbed bowls). Results The analytical data from this study are given in Table 2.1. Sr-Nd isotopic results for the 1st–3rd century glass are taken from Degryse and Schneider (2008). Major element analyses are expressed as weight %; Sr and Nd isotopic compositions are expressed as ratios. The ratio 143Nd/144Nd is also expressed as ε Nd, a parameter which indicates the isotopic composition of the sample, relative to a theoretical primordial composition. All glass can be characterized as low-magnesia, soda-lime-silica glasses (Henderson 2000). All samples can be identiied as natron glass. However, samples KEL2 and SAG573 have elevated MgO, K2O and P2O5 contents. This suggests that this early Roman sample is not a pure natron-based glass, but that plant ashes may have been used as a lux, or mixed with natron glass. However, the high Al2O3 and Na2O contents of this sample are not in concordance with ‘standard’ compositions of such glass. The blue and green glass in this study is naturally Reprint from: Isotopes in Vitreous Materials - ISBN 978 90 5867 690 0 - Leuven University Press 2009 isotopes_030209.indd 60 12 feb 2009 15:13 Neodymium and strontium isotopes in provenance determination 61 coloured by the presence of Fe2O3, the colourless glass is decoloured with Sb (e.g. Degryse et al. 2005). The Sr isotopic signature of most of the 1st–3rd century, 4th-5th century and 6thth 7 century AD natron glass shows a composition near to that of the modern-day seawater (between 0.7087 and 0.7091 for 87Sr/86Sr). Some of the 1st–3rd century AD glass has a signiicantly lower 87Sr/86Sr composition (between 0.7075 and 0.7086), while one 1st–3rd century AD glass sample has a signiicantly higher Sr signature (0.7096) and one 4th–5th century AD sample has an entirely different, much higher Sr isotopic signature (0.7255). The 6th–7th century AD HIMT glass has a lower Sr isotopic signature (between 0.7078 and 0.7085), as already reported by Freestone et al. (2005). The plant ash glass sample SAG 573 has a Sr isotopic composition of 0.7086 for 87Sr/86Sr, while the plant ash glass sample KEL 2 has a Sr isotopic composition of 0.7090. The Nd isotopic data of the 1st–3rd and 4th–5th century AD natron glass show a wide range in composition, varying between 0.512511 and 0.511974 for 143 Nd/144Nd, between -2.5 and –13.0 for ε Nd. The plant ash glass sample SAG 573 has an Nd isotopic composition of 0.51229 for 143Nd/144Nd, -6.7 for ε Nd, while the plant ash glass sample KEL 2 has an Nd isotopic composition of 0.51226 for 143 Nd/144Nd, -7.3 for ε Nd. The Nd isotopic data of the 6th–7th century AD natron glass vary much less, between 0.512408 and 0.512345 for 143Nd/144Nd, between -4.5 and –5.7 for ε Nd, with one exceptional sample of 0.512180 for 143Nd/144Nd, -8.9 for ε Nd. Discussion The blue (aqua) and green glass analysed has not been deliberately coloured or opaciied, thus there has been no contamination of the primary raw materials of the base glasses with materials from other sources. The decoloriser in colourless glass is Sb. It is unlikely, however, that this constituent would contribute signiicantly to the Sr-Nd balance of the glass. All glass analysed was imported to the respective site either as raw glass from primary production centres located outside the territory of the town (e.g. Tienen, Sagalassos) or as inished objects (possible for all sites). The spread in major element composition of the natron glass suggests that different silica raw materials may have been used for several individuals (Fig. 2.1). Reprint from: Isotopes in Vitreous Materials - ISBN 978 90 5867 690 0 - Leuven University Press 2009 isotopes_030209.indd 61 12 feb 2009 15:13 62 Patrick Degryse, Jens Schneider et al. Sample 1st-3rd AD Maastricht Ma 3 a Tienen Tie 11 Tie 12 Tie 17 Tie 24 Tie 35 Tie 37 Tie 41 Tie 45 Tie 48 Tie 49 Tie 50 Bocholtz Bo 106 Bo 109 Bo 123 Kelemantia Kel 1 - 82/91 Kel 2 - 229/06 Kel 3 - 229/88 Kel 4 - 234/88 Kel 5 - 300/06 Sagalassos Sag 575 Sag 717 Sag 718 Sag 574 Sag 709 Sag 573 Sag 722 Sag 721 Sag 723 Sag 724 4th-5th AD Sagalassos Sag 579 Sag 580 Sag 713 5th-7th AD Sagalassos Sag H54 Sag 714 Sag 583 Sag 589 SA04VL8A JP 16 JP 28 Sag 588 SA04VL8B Sag 586 SA00JP25B SA04VL4 Maastricht Ma 5 b Date Colour Nd/ Nd 143 144 2s e Nd 87 Sr/86Sr irst half 3rd AD blue 0,512343 0,000013 -5.7 0,70913 2nd AD 2nd AD 2nd AD 2nd AD 2nd AD 2nd AD 2nd AD 2nd AD 2nd AD 2nd AD 2nd AD blue blue blue blue blue blue blue blue blue blue blue 0,512511 0,512267 0,512378 0,512376 0,512219 0,512083 0,512337 0,512174 0,512262 0,512249 0,512362 0,000009 0,000009 0,000010 0,000013 0,000009 0,000006 0,000009 0,000008 0,000005 0,000010 0,000008 -2,5 -7.2 -5,1 -5.1 -8.2 -10.8 -5.9 -9.1 -7.3 -7.6 -5,4 0,70893 0,70899 0,70902 0,70902 0,70886 0,70891 0,70901 0,70904 0,70896 0,70759 0,70898 last quarter 2nd AD late 2nd - early 3rd AD late 2nd - early 3rd AD colourless colourless colourless 0,512296 0,512298 0,512291 0,000008 0,000008 0,000009 -6,7 -6,6 -6,8 0,70905 0,70903 0,70906 175-179 AD 175-179 AD 175-179 AD 175-179 AD 3rd AD colourless blue green colourless colourless 0,512325 0,512266 0,512325 0,512177 0,512336 0,000010 0,000012 0,000009 0,000011 0,000010 -6.1 -7.3 -6.1 -9.0 -5.9 0,70904 0,70901 0,70877 0,70966 0,70894 1st-3rd AD 1st-3rd AD 1st-3rd AD 1st-3rd AD 1st-3rd AD 1st-3rd AD 1st-3rd AD 1st-3rd AD 1st-3rd AD 1st-3rd AD blue blue blue colourless colourless green green green green green 0,512430 0,512410 0,512291 0,512460 0,512308 0,512294 nd 0,512392 0,512425 0,512374 0,000002 0,000002 0,000005 0,000002 0,000006 0,000005 nd 0,000013 0,000007 0,000006 -4,0 -4,4 -6,8 -3,4 -6,4 -6,7 nd -4,8 -4,1 -5,1 0,70894 0,70879 0,70882 0,70905 0,70910 0,70865 0,70886 0,70880 0,70857 0,70901 4th-5th AD 4th-5th AD 4th-5th AD colourless colourless colourless 0,512352 0,511974 0,512387 0,000002 0,000002 0,000011 -5,6 -13,0 -4,9 0,70907 0,72548 0,70881 5th-7th AD 5th-7th AD 5th-7th AD 5th-7th AD 5th-7th AD 5th-7th AD 5th-7th AD 5th-7th AD 5th-7th AD 5th-7th AD 5th-7th AD 5th-7th AD blue blue blue blue blue Co-blue Co-blue colourless purple HIMT HIMT HIMT 0,512408 0,512406 0,512385 0,512381 0,512345 0,512383 0,512382 0,512420 0,512389 0,512373 0,512355 0,512353 0,000002 0,000002 0,000009 0,000009 0,000009 0,000007 0,000009 0,000002 0,000005 0,000009 0,000009 0,000009 -4,5 -4,5 -5,0 -5,0 -5,7 -5,0 -5,0 -4,3 -4,9 -5,2 -5,6 -5,6 0,70895 0,70887 0,70881 0,70896 0,70886 0,70889 0,70908 0,70895 0,70874 0,70849 0,70782 0,70848 sec quarter 5th AD blue 0,512180 0,000009 -8,9 0,70876 Table 2.1 Analytical data of the glass studied (nd: not determined) Reprint from: Isotopes in Vitreous Materials - ISBN 978 90 5867 690 0 - Leuven University Press 2009 isotopes_030209.indd 62 12 feb 2009 15:13 63 Neodymium and strontium isotopes in provenance determination 2s SiO2 % Al2O3 % FeO % Na2O % K 2O % CaO % MgO % MnO % TiO2 % P2O5 % Total % 0,00001 66,20 2,39 0,94 18,75 0,58 6,34 1,17 1,25 0,25 0,01 97,88 0,00001 0,00002 0,00001 0,00002 0,00001 0,00002 0,00001 0,00001 0,00001 0,00001 0,00001 69,17 68,72 70,17 71,16 69,65 70,06 69,95 69,36 69,23 66,91 70,14 2,60 2,85 2,48 1,88 2,68 2,97 2,90 2,68 2,92 3,41 2,69 0,26 0,15 0,05 0,51 0,43 0,26 0,26 0,51 0,05 0,05 0,53 18,93 17,28 17,55 19,45 17,24 12,33 16,56 19,42 16,45 15,32 17,13 0,53 0,53 0,66 0,51 0,72 0,78 0,71 0,57 0,63 1,12 0,69 8,41 9,35 7,29 5,95 8,30 8,83 8,12 6,79 8,80 10,25 8,40 0,52 0,58 0,45 0,42 0,56 0,75 0,51 0,52 0,49 0,50 0,45 0,42 0,41 0,42 0,01 0,98 0,24 0,40 0,25 0,36 0,37 0,36 0,02 0,04 0,06 0,05 0,04 0,06 0,04 0,08 0,04 0,06 0,04 0,17 0,17 0,20 0,05 0,25 0,25 0,21 0,14 0,20 0,05 0,20 101,01 100,07 99,31 99,98 100,84 96,50 99,65 100,31 99,16 98,03 100,62 0,00001 0,00002 0,00002 66,12 71,00 71,40 2,16 1,89 1,93 0,53 0,33 0,32 19,61 20,35 14,82 0,47 0,43 0,33 5,77 5,42 5,69 0,57 0,34 0,38 0,02 0,01 0,02 0,15 0,08 0,08 0,06 0,04 0,04 95,44 99,90 95,00 0,00001 0,00002 0,00002 0,00002 0,00001 71,48 66,23 65,73 70,22 71,51 1,89 2,02 1,90 1,82 1,89 0,30 0,51 0,31 0,25 0,28 18,09 14,34 19,24 19,04 18,32 0,34 3,64 0,49 0,40 0,41 4,74 8,53 5,64 5,08 4,67 0,36 1,04 0,45 0,32 0,31 0,03 0,30 0,19 0,01 0,01 0,04 0,07 0,08 0,08 0,04 0,01 0,08 0,01 0,01 0,01 97,28 96,76 94,04 97,23 97,45 0,00001 0,00001 0,00002 0,00001 0,00002 0,00003 0,00001 0,00001 0,00001 0,00001 69,86 69,31 68,11 71,77 71,35 66,37 73,48 69,11 69,77 71,30 2,17 2,21 1,87 1,55 1,70 2,41 1,65 2,59 1,93 1,72 0,54 0,53 0,85 0,36 0,42 1,33 0,36 0,51 0,54 0,45 16,87 16,03 17,68 18,38 17,60 16,43 15,71 15,28 17,30 17,22 0,69 0,60 0,90 0,35 0,47 1,02 0,53 0,57 0,64 0,42 7,31 7,82 7,28 5,01 6,06 7,65 5,94 8,03 7,01 6,32 0,57 0,58 0,90 0,42 0,42 2,34 0,34 0,53 0,57 0,55 0,49 0,92 0,29 0,02 0,02 0,59 0,03 1,52 0,35 0,12 0,10 0,09 0,14 0,09 0,10 0,20 0,09 0,09 0,10 0,10 0,13 0,14 0,15 0,06 0,05 0,34 0,15 0,14 0,16 0,12 98,73 98,23 98,17 98,01 98,19 98,68 98,28 98,37 98,37 98,32 0,00001 0,00001 0,00001 70,92 69,15 66,09 1,87 1,70 1,78 0,62 0,51 1,12 17,71 19,07 19,33 0,57 0,43 0,29 5,97 6,64 7,57 0,62 0,66 0,81 0,03 0,03 1,04 0,11 0,10 0,14 0,05 0,05 0,13 98,47 98,34 98,30 0,00002 0,00002 0,00001 0,00001 0,00001 0,00001 0,00002 0,00002 0,00001 0,00001 0,00001 0,00001 69,34 67,41 66,84 70,94 65,40 68,00 65,37 70,01 64,30 62,38 63,76 63,82 2,95 2,58 2,69 2,42 1,39 1,75 2,34 2,15 1,35 2,58 3,18 2,22 0,74 0,78 0,83 0,48 1,77 1,01 2,22 0,46 2,06 1,87 3,77 2,40 14,78 15,40 15,33 16,04 17,52 19,20 18,50 17,12 17,55 20,41 15,96 18,10 0,82 0,73 0,76 0,74 0,39 0,54 0,50 0,61 0,38 0,38 0,44 0,75 9,09 9,06 9,42 7,12 8,27 7,07 7,73 6,87 8,31 6,04 5,63 7,12 0,68 0,80 0,91 0,54 0,62 0,52 0,88 0,50 0,57 1,12 1,29 1,21 0,22 0,37 0,37 0,04 2,98 0,03 0,48 0,54 4,33 2,72 1,84 3,14 0,10 0,12 0,13 0,10 0,09 0,01 0,25 0,09 0,07 0,62 0,59 0,42 0,13 0,12 0,12 0,14 nd nd nd 0,06 nd 0,08 0,13 0,11 98,84 97,37 97,40 98,56 98,43 98,13 97,91 98,41 98,92 98,20 96,59 99,29 0,00001 65,35 2,22 0,66 20,34 0,52 6,39 0,71 1,12 0,17 0,01 97,49 Reprint from: Isotopes in Vitreous Materials - ISBN 978 90 5867 690 0 - Leuven University Press 2009 isotopes_030209.indd 63 12 feb 2009 15:13 64 Patrick Degryse, Jens Schneider et al. 12 10 CaO 8 6 4 2 Maastricht Tienen Bocholtz Kelemantia 0 0 0,5 Sagalassos1 early Roman late Roman Sagalassos 1,5 2 2,5 3 3,5 4 Al2O3 byz sagalassos byz Maastricht Fig. 2.1 CaO-Al2O3 biplot of the glass analysed in Degryse et al. (this volume) A great deal of the 1st–5th century AD natron glass from the sites studied is distinct from the glass of 4th–8th century primary production centres in the Levant and Egypt. Compositions for comparison with our own analyses were taken from Freestone et al. (2000), Nenna et al. (2000), Freestone (2006) and Freestone et al. (2005). The glass from Sagalassos, Maastricht, Bocholtz and Kelemantia and some of the glass from Tienen dated to the irst half of the 5th century AD has in general higher Na2O and lower Al2O3 and CaO contents than the Levantine I, Levantine II and Egyptian II groups. Also, it has lower MgO and SiO2 contents than the Levantine I and II glass and higher MgO and K2O contents than the Egyptian II glass. In comparison to the Egyptian I group, the early and late Roman glass of Sagalassos has higher SiO2, CaO and K2O contents and lower Al2O3 and Na2O contents. The glass from Tienen on the high end of the Al2O3-CaO diagram, however, shows a good correspondence with the Levantine I glass group. The Reprint from: Isotopes in Vitreous Materials - ISBN 978 90 5867 690 0 - Leuven University Press 2009 isotopes_030209.indd 64 12 feb 2009 15:13 Neodymium and strontium isotopes in provenance determination 65 early to late Roman natron glass from the sites studied is, however, very similar in overall composition to early Roman material from all over the empire as deined by Nenna et al. (1997), Aerts et al. (2000) and Freestone et al. (2005). The possible 1st–3rd century AD plant ash glasses KEL2 and SAG 573 have a major element composition which does not correspond to known plant ash glasses, though there are few data for such Roman ‘plant ash’ glasses. In view of the elevated but not very high content of K2O and the high content of Na2O and Al2O3 in this sample, it is probable that this vessel was produced from a mixture of natron and plant ash glass. Similar levels of potassium oxide have been found in early Roman glasses from Fishbourne (Henderson in press). Remarkably, the compostion of KEL2 resembles that of the Egyptian II glass group. The Sagalassos glass dated to the second half of the ifth up to the seventh century AD corresponds well to that from the known primary production sites of that time. The HIMT glass from Sagalassos is identical to the HIMT group described by Freestone (2006), while most of the other glass from Sagalassos corresponds well to the Levantine I group. The glass from Maastricht, however, does not correspond to the Levantine I group, and has a composition identical to the 1st–3rd century Roman glass. Remarkably, samples SA04VL8A & B correspond well to the Egypt II group (e.g. Freestone 2006). Raw glass from the 4th–8th century AD primary production sites in Egypt and the Levant has already been analysed for its Sr and Nd isotopic composition (Freestone et al. in press). Strontium is considered a proxy for the lime-rich component(s) in the glass raw materials (Freestone et al. 2003). In Levantine samples, the 87Sr/86Sr signature close to the modern day marine signature of 0.7092 indicated the use of shell as a lime source in the glass (Wedepohl and Baumann 2000, Freestone et al. 2003). This shell was a natural inclusion in the beach sand of the Levantine coast, which was used to manufacture the glasses (Brill 1988). The lower 87Sr/86Sr signature of the Egyptian samples pointed to either the use of limestone (Freestone et al. 2003) or the inluence of other minerals in the sand (Degryse et al. 2005, Freestone et al. 2005) relatively low in radiogenic strontium. The low variation in 143Nd/144Nd for Levantine and HIMT (Egyptian) primary glass, with values between ε = -6.0 to –5.1, was consistent with the values given for Nile-dominated sediments in the Eastern Mediterranean (Weldeab et al. 2002, Stanley et al. 2003). This range of Sr and Nd isotopic values is repeated in most of the 5th–7th century AD glass in this study, conirming its eastern Mediterranean origin. In this way, the main element and isotopic data concur. However, both techniques are complementary and can give different information. Where main elements are less likely to be able to distinguish, for instance, between the use of different sands along the coast of Syro-Palestine (the so-called Levantine I group), Reprint from: Isotopes in Vitreous Materials - ISBN 978 90 5867 690 0 - Leuven University Press 2009 isotopes_030209.indd 65 12 feb 2009 15:13 66 Patrick Degryse, Jens Schneider et al. the variation in Nd isotopic signature may be more revealing. Conversely, the range in Nd signatures will be similar for the Egyptian or Levantine origin of a glass, while the main element composition of this glass will clearly distinguish Egyptian and Syro-Palestinian sources. The Sr isotopic signature of such glass will indicate the difference in lime source used to make the glass, the low Sr isotope values seem to be indicative of an Egyptian glass origin, while higher values, close to the modern-day oceanic signature, seem to be typical for but not exclusive to a Levantine or Syro-Palestinian origin. In this respect, samples VL8A and VL8B from Sagalassos are remarkable, as their major element composition points to the Egyptian II group, their Nd signature is indicative of an eastern Mediterranean origin, but their Sr signature is close to that of modern-day sea shell, probably identifying the lime source as such. Some 1st–3rd century AD glass from Sagalassos, Tienen, Bocholtz and Kelemantia has a Sr-Nd isotopic composition identical or very similar to the signature of the known 4th–8th century AD primary production locations in the Levant and Egypt. As mentioned before, the discovery of early Roman glass furnaces in Beirut shows that Early Roman primary glass production took place in the eastern Mediterranean, although not necessarily in the same geographical area as the aforementioned primary glass units, especially for samples with an Nd isotopic signature between –4.4 and -2.5 ε Nd. Such variation in type/composition and geographical location of sands used along e.g. the coast of Syro-Palestine for primary production could therefore be the explanation for the varying major element chemistry between early Roman and later glass, as suggested in previous studies (Nenna et al. 1997, Picon and Vichy 2003, Foy et al. 2003). The Sr signature of this glass is very homogeneous, between 0.70877 and 0.70905 87Sr/86Sr. This is nearly identical to the Sr signature of the sands and is due to shell as a lime source of the glass Conversely, some glass samples from Maastricht, Tienen, Bocholtz and Kelemantia clearly have an exotic Sr-Nd isotopic composition, not corresponding to sediment signatures from the eastern Mediterranean basin. It is clear from the study of e.g. Goldstein et al. (1984), Grousset et al. (1988) and Weldeab et al. (2002) that the Sr and Nd ratios of sediments in the Mediterranean vary signiicantly. Sediments in the east-west axis of the Mediterranean range from -10.1 ε Nd at Gibraltar to -3.3 ε Nd at the mouth of the river Nile, with a maximum of +4.6 ε Nd of the Graeco–Turkish coast. Samples with an isotopic signature of between -6.4 and -10.8 ε Nd are not consistent with any sediment in the eastern Mediterranean but correspond well to the range in isotopic values of beach and deep-sea sediments from the western Mediterranean, from the Italian peninsula to the French and Spanish coast and from north-western Europe (Degryse and Reprint from: Isotopes in Vitreous Materials - ISBN 978 90 5867 690 0 - Leuven University Press 2009 isotopes_030209.indd 66 12 feb 2009 15:13 Neodymium and strontium isotopes in provenance determination 67 Schneider 2008). The primary production location of this glass is therefore most likely to lie in the western Roman Empire. The different major element composition of the 1st–3rd century vessel glass as compared to the typical composition of the known 4th–8th century primary producers indicates that the glass has an entirely different primary origin, and is not just a variation in composition of sands in the same geographical area. Also unusual here is the 2nd century AD glass from Tienen, where a larger part of the glass samples correspond in their major element composition to the 4th–8th century AD Levantine I group. The Nd isotopic signature of these samples does not indicate an eastern Mediterranean origin, but places their primary production location in the western Mediterranean or north-western Europe (Degryse and Schneider 2008). In this example it is dificult to suggest the origins of the glass raw materials using the major element databases currently available. The Sr signature of most of this glass is very homogeneous, close to the modernday oceanic composition and likely to be indicative of the use of shell as a lime source in the glass. Some samples show a truly exotic Sr-Nd signature. Sample TIE 49 has a signature of -7.6 ε Nd and 0.70759 87Sr/86Sr. This is consistent with the Nd signature of Egyptian sands (Degryse and Schneider 2008) and the earlier analysis of early-Byzantine/Islamic Egyptian glass (Freestone et al. in press, Degryse et al. 2006a). This could suggest that the glass originated in Egypt. The major element composition of the sample distinguishes it from all other samples and early Roman glass. Sample KEL 234/88 has a signature of -9.0 ε Nd and 0.70966 87Sr/86Sr. The Nd signature of this glass sample suggests an origin in the western Roman Empire (Degryse and Schneider 2008). The Sr signature points to the use of a lime source other than shell or limestone, with an Sr signature relatively higher in radiogenic strontium than the modern seawater composition. The main element composition of this sample is identical to the main early Roman glass group. The 4th–5th century AD glass from Sagalassos on the one hand has a Sr-Nd isotopic composition identical to the signature of the known 4th–8th century AD primary production locations in the Levant and Egypt. Its main element composition, however, is closer to the early Roman glass group than the Levantine I group. One sample is quite remarkable, with a very exceptional signature of -13.0 ε Nd and 0.7254 87Sr/86Sr. Sediments dominated by input from wind-blown Saharan dusts show a typical isotopic composition with ε Nd between -12 to -13.5 and 87Sr/86Sr around 0.725 (Goldstein et al. 1984). It is tempting to assign the primary origin of this glass on this basis to North Africa, though on the basis of this one analysis this is speculative. The main element composition of this sample is typical of early Roman glass. Reprint from: Isotopes in Vitreous Materials - ISBN 978 90 5867 690 0 - Leuven University Press 2009 isotopes_030209.indd 67 12 feb 2009 15:13 68 Patrick Degryse, Jens Schneider et al. Conclusion Neodymium and strontium isotopes are clearly useful for tracing the origin of primary glass. Nd is characteristic of the mineral fraction other than quartz in the silica raw material, while Sr is in most cases characteristic of the lime component, either attributed to the sand raw material or as a separate constituent in the form of shell. These isotopes do not supplant main element analyses and both techniques discussed here should be regarded as complementary. In summary: Eastern Mediterranean 4th–8th century AD primary glass has a Nile-dominated Mediterranean Nd signature (higher than -6.0 ε Nd), SyroPalestinian glass has a sea shell-dominated Sr isotopic signature (close to 0.7092), and (Egyptian) HIMT glass has lower Sr isotopic values (as low as 0.7075). In general, lower 87Sr/86Sr values may be indicative of an Egyptian origin for glass (see also Freestone et al. 2003). In this period, groups and geographical origins deined by main element analysis (especially Levantine I and HIMT glass) concur well with groups and origins deined on the basis of isotopic data. Assigning the primary origin of 1st–3rd century AD glass appears not to be as straightforward as for the later period of natron glass production. Some 1st–3rd century AD glass has a Nile-dominated Mediterranean Nd signature (higher than -6.0 ε Nd), pointing to an Eastern Mediterranean origin. This suggests that the glass may have come from primary glassmaking sites in Egypt or in the Levant (Kouwatli et al. 2008). Glass with a different Nd signature (lower than -7.0 ε Nd) has also been identiied. This signature locates primary production in the western Mediterranean or north-western Europe (Degryse and Schneider 2008). Moreover, it has also been suggested that some 4th–5th century AD glass may have a North African origin, using Saharan sands. With the current data available, such a mismatch between major element characterisation and the results from Sr and Nd isotopes is dificult to interpret. For example, glass from 2nd century AD Tienen, has a major element composition identical to that of Levantine I glass (but chronologically produced at least two centuries earlier), and has Nd signatures excluding the use of eastern Mediterranean sands. It is unclear, however, how commonly primary glass from outside the eastern Mediterranean was used and on what scale ‘western’ glass was produced and traded. Acknowledgements This research was supported through a Fellowship of the Alexander von Humboldt Foundation awarded to P. Degryse. This research is also supported by the Reprint from: Isotopes in Vitreous Materials - ISBN 978 90 5867 690 0 - Leuven University Press 2009 isotopes_030209.indd 68 12 feb 2009 15:13 Neodymium and strontium isotopes in provenance determination 69 Interuniversity Attraction Poles Programme - Belgian Science Policy (IUAP VI). The text also presents results of GOA 2007/02 (Onderzoeksfonds K.U.Leuven, Research Fund K.U.Leuven) and of FWO projects no. G.0421.06, G.0585.06 and KAN2006 1.5.004.06N. References A. Aerts, B. Velde, K. Janssens, W. Dijkman, 2003, Change in silica sources in Roman and post-Roman glass, Spectrochimica Acta part B, 58, 659-667. J.L. Banner, 2004, Radiogenic isotopes: systematics and applications to earth surface processes and chemical stratigraphy, Earth Science Reviews, 65, 141-194. R. H. Brill, 1988, Scientiic investigations of the Jalame glass and related inds, in: G.D. Weinberg (ed.) Excavations at Jalame. Site of a glass factory in Late Roman Palestine, Missouri Press, 257-294. R.H. Brill, 1999, Chemical analyses of early glasses, Corning Museum of Glass. R.H. Brill, 2006, Strontium isotope analysis of historical glasses and some related materials: a progress report, paper presented at the 17th international conference of the Association Internationale de l’Histoire de Verre, 4-8 September 2006, Antwerp. R.H. Brill, J.M. Wampler, 1965, Isotope studies of ancient lead, American Journal of Archaeology, 71, 63-77. W.H. Burke, R.E. Denison, E.A. Hetherington, R.B. Koepnick, H.F. Nelson, J.B. Otto, 1982, Variation of seawater 87Sr/86Sr throughout Phanerozoic time, Geology, 10, 516519. P. Cosyns, M. Martens, 2002-03, Un four de verrier Romain du deuxième siècle à Tirlemont (Belgique). Bulletin de l’association française pour l’archéologie du verre 2002-03, 34-37. T. de Groot, 2006, Resultaten van de opgraving van een Romeins tumulusgraf in Bocholtz (Simpelveld), Rapportage Archeologische Monumentenzorg, 127, Rijksdienst voor Oudheidkundig Bodemonderzoek. P. Degryse, Ph. Muchez, J. Naud, M. Waelkens, M., 2003, Iron production at the Roman to Byzantine city of Sagalassos: an archaeometrical case study, in: Proceedings of the International Conference on Archaeometallurgy in Europe, 24 to 26 September 2003, Milan, Italy, 133-142. P. Degryse, J. Schneider, J. Poblome, Ph. Muchez, U. Haack, M. Waelkens, 2005, Geochemical study of Roman to Byzantine Glass from Sagalassos, Southwest Turkey, Journal of Archaeological Science, 32, 287-299. P. Degryse, J. Schneider, U. Haack, V. Lauwers, J. Poblome, M. Waelkens, Ph. Muchez, 2006a, Evidence for glass ‘recycling’ using Pb and Sr isotopic ratios and Sr-mixing lines: the case of early Byzantine Sagalassos, Journal of Archaeological Science, 33, 494-501. P. Degryse, J. Schneider, V. Lauwers, 2006b, Sr and Nd isotopic provenance determination of ancient glass, paper presented at the 17th international conference of the Association Internationale de l’Histoire de Verre, 4-8 September 2006, Antwerp. Reprint from: Isotopes in Vitreous Materials - ISBN 978 90 5867 690 0 - Leuven University Press 2009 isotopes_030209.indd 69 12 feb 2009 15:13 70 Patrick Degryse, Jens Schneider et al. P. Degryse, J. Schneider, 2008, Pliny the Elder and Sr-Nd isotopes: tracing the provenance of raw materials for Roman glass production, Journal of Archaeological Science, 35, 1993-2000. P. Degryse, J. Schneider, N. Kellens, M. Waelkens, Ph. Muchez, 2007, Tracing the Resources of Iron Working at Ancient Sagalassos (SW Turkey): a Combined Lead and Strontium Isotope Study on Iron Artefacts and Ores, Archaeometry, 49, 75-86. W. Dijkman, 1993, La terre sigillée décorée à la molette à motifs chrétiens dans la stratigraphie maastrichtoise (Pays-Bas) et dans le nord-ouest de l’Europe, in: Gallia, 49, 129-172. nd G. Faure, 1986, Principles of isotope geology, 2 ed, John Wiley and Sons. G. Faure, 2001, Origin of igneous rocks, the isotopic evidence, Springer. D. Foy, M. Vichy, M. Picon, 2000, Lingots de verre en Méditerrané occidentale, in : Annales du 14th congrès de l’Association pour l’Histoire du Verre, AIHV, Amsterdam, 51-57. D. Foy, M. Picon, M. Vichy, V. Thirion-Merle, 2003, Charactérisation des verres de la in de l’Antiquité en Mediterranée occidentale: l’émergence de nouveaux courants commerciaux, in : D. Foy, M.D. Nenna (eds.) Echanges et Commerce du verre dans le Monde Antique, Editions Monique Mergoil, 41-85. I. C. Freestone, Y. Gorin-Rosen, M. J. Hughes, 2000, Primary glass from Israel and the production of glass in Late Antiquity and the Early Islamic period, in: M.D. Nenna (ed.) La route du verre. Ateliers primaires et secondaires du second millénaire avant J.C. au Moyen Age, Travaux de la Maison de l’Orient Méditerranéen, 33, TMO, 6582. I.C. Freestone, R. Greenwood, M. Thirlwall, 2002a, Byzantine and early Islamic glassmaking in the eastern Mediterranean: production and distribution of primary glass, in: Kordas, G. (ed.), Proceedings of the 1st international conference on HyalosVitrum-Glass. History, technology and restoration of glass in the Hellenic world, 167174. I.C. Freestone, M. Ponting, J. Hughes, 2002b, The origins of Byzantine glass from Maroni Petrera, Cyprus, Archaeometry 44, 257-272. I.C. Freestone, K. A. Leslie, M. Thirlwall, Y. Gorin-Rosen, 2003, Strontium isotopes in the investigation of early glass production: Byzantine and early Islamic glass from the Near East, Archaeometry, 45, 19-32. I.C. Freestone, S. Wolf, M. Thirlwall, 2005, The production of HIMT glass: elemental and isotopic evidence, in: Proceedings of the 16th Congress of the Association Internationale pour l’Histoire du Verre, London, 153-157. I.C. Freestone, 2006, Glass production in Late Antiquity and the Early Islamic period: a geochemical perspective, in: M. Maggetti, B. Messiga (eds) Geomaterials in Cultural Heritage, Geological Society of London Special Publication, 201-216. I.C. Freestone, P. Degryse, J. Shepherd, Y. Gorin-Rosen, J. Schneider, in press, Neodymium and Strontium Isotopes Indicate a Near Eastern Origin for Late Roman Glass in London, Antiquity. S.L. Goldstein, R.K. O’Nions, P.J. Hamilton, 1984, A Sm-Nd isotopic study of atmospheric dusts and particulates from major river systems, Earth and Planetary Science Letters, 70, 221-236. Reprint from: Isotopes in Vitreous Materials - ISBN 978 90 5867 690 0 - Leuven University Press 2009 isotopes_030209.indd 70 12 feb 2009 15:13 Neodymium and strontium isotopes in provenance determination 71 J. Henderson, 1989, The scientiic analysis of ancient glass and its archaeological interpretation, in: J. Henderson (ed) Scientiic Analysis in Archaeology, Oxford University Committee for Archaeology, 30–62. J. Henderson, 1993, Aspects of early medieval glass production in Britain, in: Proceedings of the 12th Congress of the International Association of the History of Glass, 26-31. J. Henderson, in press, The provenance of archaeological plant ash glasses, in: A.J. Shortland, Th. Rehren, I.C. Freestone (eds) From mines to microscope - Studies in honour of Mike Tite, University College London Press. J. Henderson, D. McPhail, 2004, Radical changes in Islamic glass technology: evidence for conservatism and experimentation with new glass recipes from early and middle Islamic Raqqa, Syria, Archaeometry, 46, 439-468. J. Henderson, J.A. Evans, H.J. Sloane, M.J. Leng, C. Doherty, 2005, The use of oxygen, strontium and lead isotopes to provenance ancient glasses in the Middle East, Journal of Archaeological Science, 32, 665-673. J. Henderson, K. Challis, S. O’Hara, S. McLoughlin, A.Gardner, G. Priestnall, 2005a, Experiment and innovation: early Islamic industry at al-Raqqa, Syria, Antiquity, 79, 130–145. J. Henderson, J. Evans, Y.Barkoudah, in press, The roots of provenance: glass, plants and isotopes in the Islamic Middle East, Antiquity. C. Pin, D. Briot, C. Bassin, F. Poitrasson, 1994, Concomitant separation of strontium and samarium-neodymium for isotopic analysis in silicate samples, based on speciic extraction chromatography, Analytica Chimica Acta, 298, 209-217. D.J. Huisman, T. de Groot, S. Pols, B.J. van Os, P. Degryse, in press, Compositional variation in Roman colourless glass objects from the Bocholtz burial (The Netherlands), Archaeometry. C. Isings, 1957, Roman glass from dated inds, Archaeologica Traiectina II. C. Jackson, 1997, From Roman to early medieval glasses. Many happy returns or a new birth?, in: Proceedings of the 13th Congress of the International Association for the History of Glass, 289-302. R. Jackson, R. Burchill, D. Dungworth, C. Mortimer, 2005, Excavations on the site of Sir Abraham Elton’s glassworks, Cheese Lane, Bristol, Post-Medieval Archaeology, 39, 92-132. C.M. Jackson, L. Joyner, C.A. Booth, P.M. Day, E.C.W. Wager, V. Kilikoglou, 2003, Roman Glass-Making At Coppergate York? Analytical Evidence For The Nature Of Production, Archaeometry, 45, 435-456. I. Kouwatli, H.H. Curvers, B. Sturt, Y. Sablerolles, J. Henderson, P. Reynolds, 2008, A pottery and glass production site in Beirut (015), BAAL, 10. K.A. Leslie, I.C. Freestone, D. Lowry, M. Thirlwall, 2006, Provenance and technology of near Eastern glass: oxygen isotopes by laser luorination as a compliment to Sr, Archaeometry, 48, 253-270. M.D. Nenna, M. Vichy, M. Picon, 1997, L’Atelier de verrier de Lyon, du Ier siècle après J.-C., et l’origine des verres “Romains”, Revue d’Archèomètrie, 21, 81-87. M.D. Nenna, M. Picon, M. Vichy, 2000, Ateliers primaires et secondaires en Égypte à l’époque gréco-romaine, in: M.D. Nenna (ed.) La route du verre. Ateliers Primaires et Secondaires du Second Millénaire av. J.-C. au Moyen Âge, Travaux de la Maison de l’Orient Méditerranéen, 33, 97-112. Reprint from: Isotopes in Vitreous Materials - ISBN 978 90 5867 690 0 - Leuven University Press 2009 isotopes_030209.indd 71 12 feb 2009 15:13 72 Patrick Degryse, Jens Schneider et al. M.D. Nenna, 2003, Les ateliers égyptiens à l’époque gréco-romaine, in: D. Foy (ed.) Coeur de verre. Production et diffusion du verre antique, Gollion, 32-33. P.T. Nicholson, C.M. Jackson, K.M. Trott, 1997, The Ulu Burun glass ingots, cylindrical vessels and Egyptian glass, Journal of Egyptian Archaeology, 83, 143-153. T. Panhuysen, W. Dijkman, 1987, Romeinse vondsten van de Scharnweg, in: Publications de la Société Archéologique et Historique dans le Limbourg, CXXIII, 212-216. M. Picon, M. Vichy, 2003, D’Orient en Occident: l’origine du verre à l’époque romaine et durant le haut Moyen Âge, in: D. Foy, M.D. Nenna (eds.) Echanges et commerce du verre dans le monde antique, Monographies Instrumentum, 24, éditions Monique Mergoil, 17-31. M. Pollard, in press, What a long strange trip it’s been: lead isotopes and Archaeology, in: A.J. Shortland, Th. Rehren, I.C. Freestone (eds.) From mines to microscope - Studies in honour of Mike Tite, University College London Press. Th. Rehren, E.B. Pusch, 1997, New Kingdom glass- melting crucibles from QantirPiramesses, Journal of Egyptian Archaeology, 83, 127-141. E.V. Sayre, R.V. Smith, 1961, Compositional categories of ancient glass, Science, 133, 1824-1826. A. J. Shortland, 2004, Evaporites of the Wadi Natrun: seasonal and annual variation and its implication for ancient exploitation, Archaeometry, 46, 497-516. A.J. Shortland, M.S. Tite, I. Ewart, 2006, Ancient exploitation and use of cobalt alums from western oases of Egypt, Archaeometry, 48, 153-168. A. Silvestri, G. Molin, G. Salviulo, R. Schievenin, 2006, Sand for Roman glass production: an experimental and philological study on source of supply, Archaeometry, 48, 415432. J.D. Stanley, M.D. Krom, R.A. Cliff, J.C. Woodward, 2003, Nile Flow Failure at the end of the Old Kingdom, Egypt: Strontium Isotopic evidence, Geoarchaeology, 18, 395-402. M. Walton, 2005. A materials chemistry investigation of archaeological lead glazes. Doctorate Thesis, University of Oxford, Linacre College. K.H. Wedepohl, A. Baumann, 2000, The use of marine molluscan shells in the Roman glass and local raw glass production in the Eifel area (Western Germany), Naturwissenschaften, 87, 129-132. K.H. Wedepohl, W. Gaitzsch, A.B. Follmann-Schulz, 2003, Glassmaking and glassworking in six Roman factories in the Hambach Forest, Germany, in: Annales of the 15th Congress of the Association Internationale de l’Histoire de Verre, New York, 53-55. S. Weldeab, K.C. Emeis, C. Hemleben, W. Siebel, 2002, Provenance of lithogenic surface sediments and pathways of riverine suspended matter in the eastern Mediterranean Sea: evidence from 143Nd/144Nd and 87Sr/86Sr ratios, Chemical Geology, 186, 139-149. S. Wolf, S. Stos, R. Mason, M.S. Tite, 2003, Lead isotope analyses of Islamic pottery glazes from Fustat, Egypt, Archaeometry, 45, 405-420. Reprint from: Isotopes in Vitreous Materials - ISBN 978 90 5867 690 0 - Leuven University Press 2009 isotopes_030209.indd 72 12 feb 2009 15:13