Herding cats – Roman to Late Antique glass groups from Bubastis, northern Egypt (Rosenow &amp; Rehren 2014, JAS 49)

Thilo Rehren

Eighty-seven glass fragments from Roman and Late Antique layers at Tell Basta/Bubastis in the Eastern Nile Delta were typologically evaluated and chemically analysed to determine chronological and compositional patterns of glass use at this important Egyptian city, and how this relates to larger pattern of glass production and consumption in the ﬁrst half of the ﬁrst millennium AD. Bubastis is situated in geographical proximity to Alexandria, an important seaport, and at the same time close to the raw glass production areas in the Wadi Natrun and Sinai peninsula. This paper reports the ﬁrst substantial set of compositional data of Roman to Late Antique glass from a settlement in northern Egypt, ﬁlling an important gap in our knowledge of glass consumption pattern in the ﬁrst half of the ﬁrst millennium AD. The glass from Bubastis falls into several compositional groups known already from elsewhere in the Roman and Late Antique world, including antimony- and manganese-decoloured glass and two varieties of HIMT glass. Changes in glass composition over more than 500 years are in line with earlier observations concerning changes in prevalence of these glass groups. However, compositional groups known to dominate archaeological glass assemblages elsewhere, such as Roman blue/green during the earlier part of the period under study, or Levantine I in the later period, are notably absent. For the later period, this is probably due to the proximity of Tell Basta to the suspected production region of HIMT glass in northern Sinai/Egypt. By analogy, this might indicate that the earlier Roman blue/green glass has a production origin further away from the Delta than the decolourised glasses prevailing in Bubastis. A particular vessel type, small-volume thick-walled dark green unguentaria, is made of probably Egyptian plant ash glass, indicating the existence of a specialised glassmaker during the early ﬁrst millennium AD.

Journal of Archaeological Science 49 (2014) 170e184 Contents lists available at ScienceDirect Journal of Archaeological Science journal homepage: http://www.elsevier.com/locate/jas Herding cats e Roman to Late Antique glass groups from Bubastis, northern Egypt D. Rosenow a, *, Th. Rehren b a b UCL Institute of Archaeology, UK UCL Qatar, Qatar a r t i c l e i n f o a b s t r a c t Article history: Received 22 July 2013 Received in revised form 29 April 2014 Accepted 30 April 2014 Available online xxx Eighty-seven glass fragments from Roman and Late Antique layers at Tell Basta/Bubastis in the Eastern Nile Delta were typologically evaluated and chemically analysed to determine chronological and compositional patterns of glass use at this important Egyptian city, and how this relates to larger pattern of glass production and consumption in the ﬁrst half of the ﬁrst millennium AD. Bubastis is situated in geographical proximity to Alexandria, an important seaport, and at the same time close to the raw glass production areas in the Wadi Natrun and Sinai peninsula. This paper reports the ﬁrst substantial set of compositional data of Roman to Late Antique glass from a settlement in northern Egypt, ﬁlling an important gap in our knowledge of glass consumption pattern in the ﬁrst half of the ﬁrst millennium AD. The glass from Bubastis falls into several compositional groups known already from elsewhere in the Roman and Late Antique world, including antimony- and manganese-decoloured glass and two varieties of HIMT glass. Changes in glass composition over more than 500 years are in line with earlier observations concerning changes in prevalence of these glass groups. However, compositional groups known to dominate archaeological glass assemblages elsewhere, such as Roman blue/green during the earlier part of the period under study, or Levantine I in the later period, are notably absent. For the later period, this is probably due to the proximity of Tell Basta to the suspected production region of HIMT glass in northern Sinai/Egypt. By analogy, this might indicate that the earlier Roman blue/green glass has a production origin further away from the Delta than the decolourised glasses prevailing in Bubastis. A particular vessel type, small-volume thick-walled dark green unguentaria, is made of probably Egyptian plant ash glass, indicating the existence of a specialised glassmaker during the early ﬁrst millennium AD. Ó 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/3.0/). Keywords: Glass Chemical composition Roman Late Antique Egypt 1. Introduction The composition of Hellenistic to Byzantine glass is characterised by a surprising degree of fundamental similarity and consistency over more than a thousand years (Sayre and Smith, 1961), which may be explained at least in part by a combination of faithfully maintained traditional recipes using tightly controlled raw materials, and partly by the self-governing behaviour of the melt-forming soda-lime-silica system (Rehren, 2000; Tanimoto and Rehren, 2008). Within this broad homogeneity, however, there are well-developed and long recognized speciﬁc compositional groups, characterised by their minor oxide and trace element contents. It is generally assumed that the minor oxide and trace element contents of ancient glass reﬂect the composition of the sand used in its * Corresponding author. E-mail address: d.rosenow@ucl.ac.uk (D. Rosenow). production (e.g., Freestone, 2006), while the soda levels are determined by batch recipes. For the ﬁrst four centuries AD several ‘Roman’ glass groups have been established, mostly through the analysis of samples from Italy and the Northern provinces (e.g. Jackson, 2005; Silvestri et al., 2008; Foy et al., 2003). The most common glass there is naturally blue/green coloured, with no intentional additives to manipulate its colour; this natural colour is due to the iron impurities in the sand and the prevailing redox conditions in the glassmaking furnace. It is often referred to as ‘aqua’, to distinguish it from glass intentionally coloured blue or green through the addition of metal oxides. Colourless Roman glass is characterised by the addition either of antimony or manganese oxide to counter-act the colouring effect of iron oxide, or a combination of both oxides (Jackson, 2005; Silvestri et al., 2008; Foster and Jackson, 2010). Antimony-decoloured glass is typically dated earlier than manganese-decoloured glass; substantial data sets have been published, among others, by Paynter (2006) for glass from Britain and Silvestri et al. (2008) from a ship wreck in the http://dx.doi.org/10.1016/j.jas.2014.04.025 0305-4403/Ó 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/3.0/). D. Rosenow, Th. Rehren / Journal of Archaeological Science 49 (2014) 170e184 171 Fig. 1. The Egyptian Nile Delta showing the position of Bubastis, the Wadi Natrun, Alexandria, Cairo and Sinai peninsula. northern Adria. Roman blue/green glass is generally considered to be of Levantine origin (Nenna et al., 1997), while decoloured glass is linked to a production in northern Egypt (Nenna, 2007). For the mid to late ﬁrst millennium AD, ﬁve main compositional glass groups have been identiﬁed, mostly through analysis of glasses from the eastern Mediterranean region. These include Egyptian I and II, Levantine I and II and HIMT glass (Freestone et al., 2005). The ﬁrst four groups can be associated with raw glass production centres in Egypt (Wadi Natrun and Ashmunein; Nenna, 2007) and the Levant (bay of Haifa and Bet Eli’ezer; Gorin-Rosen, 2000), respectively. The production region of HIMT glass cannot be located precisely, but is thought to be in northern Egypt, possibly the northern coast of the Sinai (Foy et al., 2003; Freestone et al., 2005). Levantine and HIMT glass has been discovered at numerous sites and regions throughout the Roman Empire, while published evidence for Egyptian I glass is relatively rare outside Egypt. Signiﬁcantly, the major compositional groups have distinct chronological ranges, indicating that each production site only had a limited period of activity, spanning a few centuries. According to Freestone et al. (2000) and Freestone et al. (2005), HIMT was mostly in circulation from the late fourth to the sixth centuries AD, Levantine I during the fourth to seventh centuries AD, Levantine II during the seventh to eighth centuries AD. Egyptian I was in use from an as yet unknown start date up to the eighth century AD, while Egyptian II was the predominant glass in the Levant during the eighth and ninth centuries AD (Freestone et al., 2000). Very little is known, however, about the relative proportions of these various glass groups in northern Egypt, the heartland of early glass making, restricting our ability to discuss the organisation of 1st millennium AD glass making and consumption. The link between regionally different sand compositions and the minor oxide and trace element content of ancient glass provides a promising tool to explore the relationship between production origin and regions of glass consumption. Two competing models have been put forward, supporting either a more localised or dispersed system of raw glass production (Wedepohl and Baumann, 2000; Baxter et al., 2005; Degryse and Schneider, 2008; Silvestri et al., 2008; Foster and Jackson, 2010) or a more centralised one (Foy and Jezegou, 1997; Foy et al., 2000; Freestone et al., 2000; Picon and Vichy, 2003; Paynter, 2006; Nenna, 2007). The ﬁrst model assumes the existence of a range of regional primary glass production centres, not exclusive to the Levant or Egypt, but also including sites in e.g. Italy (Silvestri et al., 2008) or the northern provinces (Wedepohl and Baumann, 2000; Jackson, 2005). The centralised system on the other hand supports the idea of raw glass production on a large scale at only a small number of locations at any one time. From the 4th century AD onward there is good archaeological and compositional evidence for a strongly centralised production of glass in large scale factories on the Eastern Mediterranean shores, both in Egypt and the Levant (GorinRosen, 2000; Freestone et al., 2000; Picon and Vichy, 2003; Nenna, 2007), from where the raw glass would then have been sent as chunks to secondary glass working furnaces across the Empire for artefact production serving local or regional markets. It is less clear, however, whether this system also holds for the ﬁrst three centuries AD, and this study aims to throw some light on this issue. 2. Introduction to the site The ancient city of Bubastis (Tell Basta) is an Egyptian site of major historical and cultural signiﬁcance, with continuous occupation ranging from the Old Kingdom (2686-2160 BC) to Late Antiquity (6th century AD). It is situated on the south-eastern edge of Zagazig in the Eastern Nile Delta (Fig. 1), and is best known for its temple dedicated to the Egyptian cat goddess Bastet. Its visible remains date to the Third Intermediate Period (1069e664 BC) and Late Period (664e343 BC) (all dates following Shaw, 2003). Bubastis still played a signiﬁcant cultic role during Early Ptolemaic times, but at some point after the late 3rd century BC the temple collapsed, 172 D. Rosenow, Th. Rehren / Journal of Archaeological Science 49 (2014) 170e184 Fig. 2. The ancient site of Bubastis. probably due to an earthquake. In the aftermath, and certainly during Late Roman times, it was used as a quarry, and its signiﬁcance as a cultic centre seems to have waned. However, the city continued to be well integrated into the Roman world, as indicated by the presence of numerous imported ceramic vessels. The Tell Basta Project1 spent the last six years exploring the area east of the temple, where e following Herodotus’ description of the ancient city e the settlement of Bubastis was situated. This area of approximately 40 ha remains completely unexcavated. A survey in 2008 revealed a large number of objects dating to the GraecoRoman Period, including numerous glass fragments of the Roman and Late Antique periods. Earlier excavations in the zone connecting the temple and the settlement (so-called Area A) brought to light remains of a Roman monument (Habachi, 1957, 93e94). To the east and south, remains of domestic and semi-ofﬁcial buildings of red bricks were unearthed, clearly connected to the Roman ediﬁce. These building remains can be ascribed to a period when the temple, after its collapse, had not been in use anymore as a cult place. Only the contexts closest to the Roman limestone monument revealed glass ﬁnds, probably belonging to a period of subsequent use or reuse of the temple. The amount of glass discovered in deposits further away is negligible, and probably represents ﬁnds that have been accidentally moved during the last centuries; due to their uncertain archaeological origin these are not included in this study. 2.1. Glass at Bubastis 1 The Tell Basta Project is a German-British-Egyptian Joint Mission and directed by Eva Lange. Fig. 3. Base fragment of an oval dish demonstrating the average colour of HIMT glass from Bubastis. About 2500 glass fragments have so far been recorded at Tell Basta. The pieces studied here originate primarily from three D. Rosenow, Th. Rehren / Journal of Archaeological Science 49 (2014) 170e184 173 Some glass vessels are decorated by single wheel-engraved lines or bands and/or ornaments of applied blobs of blue glass, pinched elements, indents, single applied threads of the same or a different colour below the rim, or incised horizontal lines. One fragment displays facet-cut circular impressions. The dating of the glass vessels used for this study is based on typology, since most glass was retrieved from disturbed contexts. According to parallels with dated ﬁnds from Roman and Late Antique Egypt, such as Mons Porphyrites (Bailey, 2007), Kom elDikka (Kucharczyk, 2004, 2006, 2010), Quseir al-Qadim/Myos Hormos (Meyer, 1992; Peacock, 2011), Bagawat (Nenna, 2010; Hill and Nenna, 2003), Ismant el-Kharab (Marchini, 1999), several sites in the Eastern Desert (Brun, 2003a,b, 2011), Karanis (Harden, 1936), Kom el-Nana (Faiers, 2013) or Tebtynis (Nenna, 2000; Foy, 2001), the Bubastis corpus roughly covers a time between the ﬁrst century BC to the sixth century AD. The majority of the material dates to the second to fourth centuries AD. 3. Materials and methods Fig. 4. Rim fragment demonstrating the average colour of Sb decol glass from Bubastis. contexts: the unexplored area east of the temple, the area connecting the temple and the settlement (Area A), and the entrance court of the sanctuary (Fig. 2). All glass fragments discovered during the 2008 survey are surface ﬁnds and have no speciﬁc archaeological contexts. The overwhelming majority of glass ﬁnds, approximately 90%, are from 10 grid squares (10 10 m) in Area A, most of them coming from the uppermost layers covering or surrounding the limestone monument and red-brick buildings. They are associated with other artefacts and pottery dating from the 3rd century BC to Late Antiquity. About 10% of the ﬁnds have been excavated from the entrance court of the temple of Bastet. These fragments derive also from heavily disturbed contexts, with some associated ceramic ﬁnds in deeper deposits dating from the New Kingdom to the 5th century AD. All glass ﬁnds from Tell Basta were studied, recorded, drawn and typologically compared to published parallels; the results of this will be published elsewhere. Fragments include pieces of lamps, beakers, bowls, plates, cups, jugs, bottles, jars, ﬂasks, goblets, oval dishes, and small and large containers. In addition, intensively coloured bracelets, beads and counters/gaming pieces were recovered. The majority of the assemblage belongs to vessel types representing utilitarian ware for daily use. This is consistent with the ceramic evidence from Area A, indicating that at this time the temple of Bastet was no longer used as an active place of worship. Luxury glass is thus scarce, with just a few fragments of milleﬁori glass dishes, facet-cut colourless glass or indented beakers. Only very few pieces can be related to secondary production processes, such as wasters, moils or chunks. A few fragments represent mould cast vessels such as cast ribbed bowls, plates or bowls made of milleﬁori glass, or rims and bases from cast bowls and plates. However, free blown glass is by far the most dominant; some vessels are mould blown. The most usual colour within the Bubastis glass is yellowish green to deep olive green to brown, in varying shades and intensities, particularly for the typologically later material (Fig. 3). Among the earlier ﬁnds, however, many sherds are pale bluish-greenish (‘aqua’) to colourless. Some pieces are amber, a few ﬁnds are blue, purple or red. Due to the moist environment of the Egyptian Nile delta, corrosion is affecting the majority of glasses, and more so the earlier ﬁnds (Fig. 4); the formation of dark brown or whitish crusts can obscure the original colour and transparency of the glass. The main aims of this project are to learn more about the economic position of Bubastis during the Roman and Late Antique periods, and to improve our understanding of the distribution of speciﬁc glass compositions in space and time, particularly for Egypt. The project also aims to test the utility of portable XRF analysis to assign glass fragments to speciﬁc compositional groups, in order to be able to analyse large assemblages such as this one on site to quantify the relative importance of each glass group over time, while minimising the need for more invasive and timeconsuming laboratory-based analyses. Detailed results of this will be presented elsewhere, as this topic is outside the remit of this paper. 3.1. The analysed assemblage Eighty-seven glass vessel fragments were selected for quantitative analysis, including mould cast, free-blown and mould blown vessels (Table 1). Due to their state of preservation, a few samples could not be typologically identiﬁed. In this case, sampling was motivated by the colour of the glass. The majority of the samples are of pale greenish colour, yellow-greenish, green or colourless. Two samples are intentionally coloured blue, four samples are light blue. There is one sample each of pale purple, red-brown, burgundy and brownish pink colour. The low capacity unguentaria are made of dark green or bluish-green ‘emerald’ glass. Fragments selected for analysis are thought to reﬂect the whole range of glass vessel types, manufacturing methods and decoration techniques, over the entire period of time when glass ﬁnds are attested in Bubastis. With less than ﬁve percent of all fragments analysed, and covering a period of more than 500 years, this is necessarily only a pilot study, and the relative proportions of samples reported here are not representative of the types and compositions constituting the total glass assemblage. In particular the early glasses are over-represented in the analysed corpus, with nearly all cast glass and a signiﬁcant proportion of the visually identiﬁed antimony-decoloured glass analysed. In contrast, the vast abundance of late glass, typically of olive green to brown colour, is under-represented among the analyses, even though these pieces make up about two thirds of all analysed material. 3.2. Data generation and handling EPMA was done on polished cross sections using a JEOL JXA 8100 with three spectrometers. The instrument was operated at 20 kV with a beam current of 6 nA and count times of 10 s on the 174 D. Rosenow, Th. Rehren / Journal of Archaeological Science 49 (2014) 170e184 Table 1 Catalogue of analysed samples. Mn decoloured Mn 01 TB3a- -Z/3.SCH.1-G009 Skyphos Handle Mould cast Mn 02 Mn 03 Mn 04 TB2a-OPQ/2006-G001 TB1a-T/8-G001 TB2b-W/3.SCH.1-G008 Ribbed bowl Bowl Beaker Wall Rim Rim Mould cast Mould cast Free blown Mn 05 TB2b-X/3.SCH.1-G004 Bowl Rim Mould cast Mn 06 TB3b-X/4.SCH.1-G025 Ribbed bowl Rim Mould cast Sb decoloured Sb 01 TB2b-X/4.SCH.1-G033 Bowl or plate Base Technique Mould cast Sb 02 TB1a-D/12.2-G001 Flask? Sb 03 TB3a- -Z/3.SCH.1-G056 Plate Base with folded (tubular) ring Base Free blown Sb 04 TB1b-X/2.AbH-G022 Bowl Rim Mould cast Colourless-light green Colourless Sb 05 TB3a- -Z/3.BEF.1-G003 Bowl Base Free blown Colourless Sb 06 TBXIV-OPQ-G010 Plate Base Mould cast Sb 07 TB3a-X/4.SCH.1-G023 Wall Free blown Sb 08 TB3a-Y/3.SCH.1-G060 Facet-cut beaker? Beaker? Base Free blown Sb 09 TB3a-X/2.TS.SCH.1-G012 Flask? Free blown Sb 10 TB2b-X/3.SCH.1-G029 Free blown Colourless 15 Sb 11 TB3a- -Z/3.SCH.1-G050 Beaker, goblet, sprinkler? Aryballos Base with folded (tubular) ring Base with pinched feet Rim Colourless-light green Light greencolourless Colourless-light green Colourless-light green Free blown Colourless 5 Sb 12 TB3b-Y/4.SCH.1-G042 Waster Base Free blown Sb 13 TB1a-T/8-G018 Flask? Free blown 103 Isings type 104a 1./2. AD Sb 14 TBXIV-D/7-G001 Cup? Free blown Colourless 103 Isings type 37 1./2. AD Sb 15 TB2a-X/2.AbH-G007 Plate or bowl? Base with folded (tubular) ring Base with folded (tubular) ring Base Light greencolourless Colourless Free blown (?) TB2a-X/3.SCH.3-G007 Bottle/ﬂask 72 Peacock, 2011, 74, Fig. 7.13.158 Isings type 102b 1.-3. AD Sb 16 Colourless-light green Colourless 3. AD Sb 17 TB1b-W/2.SCH.1-G008 48 Isings type 80 1.-3. AD Sb 18 10 1.-4. AD Sb 19 Bailey, 2007, 259, Fig. 8.17.142 Bailey, 2007, 258, Fig. 8.16.135 Bailey, 2007, 259, Fig. 8.17.141 Bailey, 2007, 259, Fig. 8.17.138 Bailey, 2007, 261, Fig. 8.18.155 Bailey, 2007, 262, Figs. 8.19, 165 1./2. AD Sternini 1999, 99, Fig. 9.119 Harden, 1936, pl. XIX, 739 4./5. AD Free blown Bowl Rim with applied thread Base Mould cast TB2b-X/3.SCH.1-G011 Small container Base Free blown TB1a-Survey-G024 Small container Rim Mid-capacity unguentarium Mid-capacity unguentarium Low-capacity unguentarium Storage or transport container Plant ash PA 01 TB3b- -Z/3.SCH.1-G001 PA 02 TBXX-G20093 PA 03 TB1a-Survey-G017 PA 04 TB1b-W/2.SCH.1-G006 Weak HIMT wH 01 TB3a-Y/3.SCH.1-G024 wH 02 TB2a-X/2.SCH.2-G019 wH 03 TB3a- -Z/3.SCH.1-G073 wH 04 TB3a-Survey-G003 Free blown Colourless-light purple Aqua Colourless Colourless Colourless-light green Dark blue Colourless-light green Colourless-light green 1 Isings type 55 1. AD 7 4 5 Isings type 3a Meyer, 1992, pl. 2.26 Peacock, 2011, 67, Fig. 7.7.77 Jennings 2006, 35, Fig. 2.6.5 Isings type 3b 50 BCe130 AD 1./2. AD 1./2. AD Isings type 80 Late 1.-mid 3. AD Isings type 104a Mid 1.-mid 3. AD Bailey, 2007, 254, Fig. 8.14.15 Peacock, 2011, 69, Fig. 7.9.106 Brun 2003b, 384, Fig. 8.6 Bailey, 2007, 238, Fig. 8.2.19 Peacock, 2011, 65, Fig. 7.5.62 Bailey, 2007, 247, Fig. 8.8.71 Isings type 104a Late 1.elate 2. AD 4 7 48 103 42 3 103 48 1 37 103 Brun, 2011, 239, Fig. 271.136 Bailey, 2007, 256, Fig. 8.15.118 50 BCe50 AD 50 BCe130 AD Late 1.e175 AD 2. half 2. AD 1./2. AD? 1./2. AD? late 1.-2. AD 1./2. AD? 2.-6. AD 1./2. AD? unknown 42 Free blown Colourless-light green Colourless-light green Blue 33 Base Free blown Green-turquois 2 Base Free blown Green 2 Base and body Free blown Dark green 4 Ridged handle Free blown Turquois 6 Beaker, jug, goblet, ﬂask? Bottle, jug, ﬂask, beaker? Base with applied rings Rim with applied thread Free blown Colourless 39 Free blown 72 Bottle, jug, ﬂask, beaker? Bottle, beaker or ﬂask? Base, pinched feet Base, pinched feet, wall indented Free blown Wall (sampled part) light green, ring blue Colourless 15 Free blown Green 15 Harden, 1936, pl. XIX, 682 Harden, 1936, pl. XIX, 682 1./2. AD 1./2. AD 1./2. AD 1.-5. AD ab 3. AD 2.-6. AD 1.-4. AD 175 D. Rosenow, Th. Rehren / Journal of Archaeological Science 49 (2014) 170e184 Table 1 (continued ) wH 05 wH 06 TBXX-G20087a TB1b-W/2.SCH.1-G019 wH 07 TB3a- -Z/3.SCH.1-G002 Lamp Beaker, jug, goblet, ﬂask? Lamp or beaker Base (pointed) Base with applied rings Base (conical hollow) Rim (tubular) Free blown Free blown wH 08 TB3a-Y/3.SCH.1-G002 Bowl wH 09 TB1a-Survey-G012 Lamp wH 10 TBXX-G20049 Lamp Base (with solid stem) Base (pointed) wH 11 TB2a-X/2.AbH-G037 wH 12 TB3a-Z/3.SCH.1-G027 Beaker, jug, goblet, ﬂask? Small container Base with applied rings Base wH 13 TB2a-X/3.SCH.5.BEF.1-G001 Plate or bowl wH 14 TB2b-W/3.SCH.2-G005 wH 15 TB2b-W/2.AbH-G010 Bottle, jug, ﬂask? Aryballos Base (high footring) Rim with applied thread Handle wH 16 TB3a-Y/3.SCH.1-G008 Bottle or jug? wH 17 TB3a-Y/3.SCH.1-G042 wH 18 TBXIV-S/2-G013 Bottle, ﬂask or jug? Goblet Base with applied rings Base with applied ring Base and stem wH 19 TB3a- -Z/3.SCH.1-G001 Goblet, beaker or ﬂask? Base with applied ring Free blown wH 20 wH 21 TB1b-OPQ-G009 TB1b-OPQ-G006 TB3a-Z/3.SCH.1-G024 Rim, tubular Base, pinched feet Wall Free blown Free blown wH 22 Oval dish Bottle, ﬂask, beaker or jug? Dish made of mosaic glass wH 23 TBXIV-G14011 neck and rim free blown wH 24 TBXIV-OPQ-G005 ﬂask/toilet bottle? Beaker, jug, goblet? Base with applied ring Free blown wH 25 wH 26 TB2a-X/2.AbH-G020 TB1b-W/2.SCH.5-G006 wH 27 TB3b-X/4.Steg-G011 wH 28 Free blown Free blown Free blown Free blown Free blown Free blown Free blown Free blown Free blown Free blown Free blown Free blown Light greencolourless Light olive green Yellowish green Unknown, corroded Light green 6 39 14 25 9 6 39 20 Colourless-light green Light green 120 Unknown, corroded Light olive green Wall light green, ring blue Light green 5 wall colourlesslight green, base ring blue (sampled part) Light blue Light green 72 39 37 6 37 17 15 Base Base Free blown Free blown Base (high footring) Stem Free blown TB3a-X/4.SCH.1-G027 Flask/bottle Base indented beaker Jug, ﬂask, bowl? Bowl Free blown Purple-red 17 wH 29 TB3a- -Z/3.SCH.1-G028 Cup or bowl Rim (inturned) Free blown Light blue 35 HIMT H 01 TB3b-V/3.SCH.1-G159 Wall Mould blown TB2b-X/3.BEF.1-G005 Base (pointed) Free blown H 03 TB2b-X/3.BEF.1-G010 ???? Wall Mould blown H 04 TB2b-X/3.SCH.1-G040 Wall Mould blown H 05 TB2a-M/1.SCH.1-G015 Bowl or beaker? ??? Unknown, corroded Light greencolourless Colourless-light green Green 33 H 02 Bowl or beaker? Lamp H 06 TB2b-X/3.SCH.3-G009 H 07 TB1a-W/2.AbH-G007 H 08 TB2b-X/3.SCH.1-G016 H 09 H 10 TBXX-G20041 TB2b-W/3.SCH.3-G008 H 11 TBXX-G20008b Hemispherical bowl or cup Oval dish Bowl or drinking vessel Lamp H 12 TB2a-W/3.SCH.4-G001 Bowl or ﬂask? H 13 TBXX-G20026 Lamp or beaker Colourless Isings type 106d Tatton-Brown 1984, 206, Fig. 68.103 Harden, 1936, pl. XVI, 457 Marchini 1999, 80, Fig. 3 b Jennings 2006, 146, Fig. 6.20.11-13 Isings type 106d From 4. AD 4./5. AD Tatton-Brown 1984, 206, Fig. 68.103 Harden, 1936, pl. XX, 799 Harden, 1936, pl. XII.83/130 Harden, 1936, pl. XIX, 712 Isings type 61 4./5. AD Tatton-Brown 1984, 206, Fig. 68.103 Keller 2006, Tafel 21.g Harden, 1936, pl. XVI, 482 Bailey, 1998, pl. 93.Y72 Isings type 97b Harden, 1936, pl. XIX, 682 3 Green (sampled) and yellow Colourless-light green Wall colourless, base ring blue (sampled part) Colourless Unknown, corroded Reddish brown Conical lamp or beaker Stemmed bowl Cast Light green Light greencolourless Light green 5 37 70 25 120 6 1./2. AD From 4. AD From 4. AD 1.-3. AD From 4. AD From 3. AD Late 1.-7. AD 4./5. AD 4./5. AD 5.-7. AD 4./5. AD 3.-5. AD 2.-6. AD 1.-5. AD Nenna, 2010, 210, Fig. 10.34 Hill/Nenna, 2001, 91, Fig. 4.4 1.-3. AD Isings type 133 Harden, 1936, pl. XV, 376 Harden, 1936, pl. XIV.274 Harden, 1936, pl. XV, 360 Nenna, 2000, 23, Fig. 9.4 1.-4. AD 1.-4. AD Harden, 1936, pl. XIII, 217 Isings type 106d 4./5.AD 33 25 From 4. AD 4./5. AD 4.-7. AD 4./5. AD From 4. AD From 4. AD 2.-4.AD Harden, 1936, pl. XIII, 189, pl. XV, 409 2 Unknown Wall (with cut decoration) Rim Free blown Green 90 Isings type 106d From 4. AD Stem Free blown Green 17 4./5. AD Rim Free blown 20 Base Base (high footring) Base (with twisted blob) Base (high footring, wavy) Base (conical hollow) Free blown Free blown Light greencolourless Olive green Light green Harden, 1936, pl. XV, 358 Isings type 96 Free blown Light green 6 Free blown Olive green 2 Free blown Green 17 120 14 Unknown Isings type 97b Harden, 1936, pl. XV, 360 Kucharczyk, 2006, 48, Fig. 1.4 Harden, 1936, pl. XIX, 672 Harden, 1936, pl. XVI, 457 From 4. AD 3.-5. AD From 4. AD From 4. AD From 4. AD From 4. AD (continued on next page) 176 D. Rosenow, Th. Rehren / Journal of Archaeological Science 49 (2014) 170e184 Table 1 (continued ) H 14 TB3a-Y/3.SCH.1-G039 Bowl H 15 TB3a-Y/3.SCH.1-G033 Bowl H 16 TB1a-V/2.AbH-G189 H 17 TB2b-OPQ-G014 Transport or storage container Bowl H 18 H 19 TB3b-V/3.SCH.1-G172 TB2a-X/2.SCH.1-G054 H 20 TB3a-Y/3.SCH.1-G093 H 21 TB3a-Y/3.SCH.1-G017 Bowl Conical lamp or beaker Beaker/jug/ ﬂask? Cup or bowl H 22 TB1a-T/8-G003 Cup or bow H 23 TB1b-W2.SCH.3-G006 Flask H 24 TBXX-G20008c Bowl H 25 TB1b-X/2.AbH-G119 Bottle H 26 H 27 TB1a-V/2-G005 TB3a-X/2.TS.SCH.3-G001 H 28 ukn TB2a-X/2.AbH-G047 TB1a-D/11.SCH.3-G004 Rim (horizontal) Rim (edge going up) Handle (ridged) Free blown Olive green 4 Free blown Light green 10 Light green 6 Rim Optical blown Olive green 4 Rim Rim Free blown Free blown Pinkish brown Green 1 90 Base with applied rings Rim (strongly everted) Rim (horizontal) Neck Free blown Green 39 Free blown Light olive green Green 15 Light greenyellowish Yellowish green 26 Free blown Free blown Free blown Base with folded (tubular) ring Wall Mould blown Amphora? Conical lamp or beaker? Handle (ridged) Wall Free blown Bowl Stemmed goblet Rim (tubular) Stem Free blown Free blown peak position and 5 s on the background positions. Soda loss during analysis was prevented by scanning the beam over the area visible at 800 times magniﬁcation. Table 2 reports the results of measurements of Corning A and B reference glasses analysed together with the Bubastis samples. For most oxides the measured values are within 5% of the published values; however, alumina and phosphorus oxide were consistently analysed lower than the published values, while antimony oxide was measured higher by about one third (20 and 40% in Corning A and B, respectively) of the published value (Brill 1999). No adjustment has been made for these systematic deviations in the reported data in Table 3. Concentrations of antimony oxide need to be treated with caution, and values below 0.3 wt% are not reported; these may well reﬂect analytical error rather than real presence of this compound. The glasses were sorted into compositional groups based on minor oxide concentrations. Typical values for published glass groups of the oxides of aluminium, calcium, titanium, manganese and antimony informed a ﬁrst allocation of the newly-analysed 4 Harden, 1936, pl. XII, 130 Nenna, 2000, 23, Fig. 9.2 From 4. AD From 4. AD Harden 1936, pl. XXIV, 256 See drawing Isings type 106d Tatton-Brown 1984, 206, Fig. 68.103 Tatton-Brown 1994, 283, Fig. 15.1.5 Weinberg 1988, 52, Figs. 4-12.95 Isings type 133 From 4. AD Mid 4.-mid 5. AD From 4. AD 4./5. AD From 4. AD 4.-6. AD Unknown 103 Harden, 1936, pl. XIV, 245 4. AD Olive green 25 Harden, 1936, pl. XIX, 700 and 701 4. AD Green Wall light green (sampled part), blue blob Olive green Aqua 6 18 7 6 4.-5. AD From 4. AD Kucharczyk, 2006, 48, Fig. 1.15 Harden, 1936, pl. XII, 89 Sternini 1999, 95, Fig. 6.67 4.-7. AD From 4. AD samples to the known groups; this was then further reﬁned by checking the levels of the remaining minor oxides for consistency with those typically found in the established compositional groups, re-allocating samples as necessary to obtain a subjective best ﬁt. All but one sample were thus allocated to speciﬁc compositional groups. 4. Results Eighty-three of the eighty-seven samples are mineral-natron based soda-lime-silica glasses (Table 3), while four fragments appear to be made from plant ash glass. Among the mineral natron glasses, four main groups dominate: Manganese-decoloured (6), antimony-decoloured (19, including one coloured blue by cobalt), weak HIMT (29) and strong HIMT (28). A single pale-coloured stemmed goblet cannot be attributed to any of these groups, but stands compositionally alone. The main groups are presented below in chronological order as listed in Table 1, with the plant-ash Table 2 Comparison of published compositions for Corning A and B (Brill 1999: 544) and the average values of 7 measurements of Corning A and B during the course of the analysis of the Tell Basta samples. The precision of the analyses is indicated by the standard deviation among the seven individual analyses for each of the Corning glasses, while the accuracy is expressed by the deviation of the analysed value Ca from the published composition Cp. This d rel% value is calculated using the formula (CaeCp)/Cp*100. SiO2 Na2O CaO K2O Cor A published Cor A aver (n ¼ 7) StdDev d rel% 66.56 67.08 0.61 0.8 14.30 14.16 0.11 1.0 5.03 4.92 0.03 2.2 2.87 2.78 0.03 3.1 Cor B published Cor B aver (n ¼ 7) StdDev d rel% 61.55 62.18 0.47 1.0 17.00 17.08 0.29 0.5 8.56 8.43 0.07 1.5 1.00 0.99 0.03 1.0 MgO Al2O3 Fe2O3 TiO2 Sb2O5 MnO CuO P2O5 Cl SO3 2.66 2.59 0.03 2.6 1.00 0.91 0.02 9.0 1.09 1.03 0.02 5.5 0.79 0.78 0.01 1.3 1.76 2.12 0.04 20.5 1.00 0.97 0.00 3.0 1.17 1.19 0.04 1.7 0.13 0.09 0.01 30.8 0.09 0.01 0.15 0.01 1.03 1.00 0.01 2.9 4.36 4.15 0.10 4.8 0.34 0.33 0.03 2.9 0.09 0.09 0.01 1.1 0.46 0.64 0.01 39.1 0.25 0.21 0.03 16.0 2.66 2.70 0.05 1.5 0.82 0.69 0.03 15.9 0.20 0.17 0.01 15.0 0.54 0.54 0.02 0.0 Analytical total 99.53 99.87 0.67 99.98 100.04 0.83 177 D. Rosenow, Th. Rehren / Journal of Archaeological Science 49 (2014) 170e184 Table 3 EPMA analyses of glass samples from Bubastis, data in weight percent. Tin, cobalt and lead were analysed for, but not found at levels above 300m (500m for lead). Mn Mn Mn Mn Mn Mn 01 02 03 04 05 06 66.6 69.2 66.8 68.5 67.7 67.6 16.4 15.1 16.6 14.5 15.6 17.8 8.10 7.28 8.68 7.62 7.43 9.21 0.52 0.65 0.53 0.45 0.87 0.59 0.53 0.42 0.65 0.60 0.94 0.55 2.03 2.13 2.13 2.49 2.25 2.44 0.27 0.28 0.33 0.32 0.34 1.17 0.05 0.04 0.05 0.06 0.05 0.04 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 0.86 0.46 1.13 1.30 1.14 1.11 <0.03 0.03 0.03 <0.03 <0.03 0.07 0.08 0.11 0.07 0.06 0.09 0.10 0.68 0.96 1.13 0.92 0.82 0.83 0.32 0.14 0.18 0.21 0.19 0.29 96.7 96.8 98.4 97.0 97.7 100.9 Average StDev 67.7 1.0 16.0 1.2 8.05 0.76 0.60 0.15 0.62 0.18 2.25 0.18 0.31 0.35 0.05 0.00 <0.03 1.00 0.30 0.02 0.03 0.08 0.02 0.89 0.15 0.22 0.07 97.9 1.6 Sb Sb Sb Sb Sb Sb Sb Sb Sb Sb Sb Sb Sb Sb Sb Sb Sb Sb Sb 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 71.3 69.1 69.9 69.2 70.8 68.5 69.5 67.7 66.1 65.9 68.6 66.1 68.1 68.3 66.3 66.0 65.9 71.5 71.3 17.9 17.6 16.5 17.3 16.0 18.1 17.7 17.1 17.5 18.0 18.5 17.6 16.8 16.9 17.4 18.4 17.1 16.0 14.2 5.11 5.96 5.67 6.44 5.67 5.90 6.92 6.23 6.58 8.27 6.47 6.87 7.83 7.34 8.55 8.42 8.22 5.81 7.67 0.42 0.45 0.41 0.46 0.46 0.68 0.55 0.65 0.54 0.39 0.51 0.54 0.54 0.68 0.53 0.58 0.51 0.46 0.61 0.48 0.44 0.48 0.50 0.43 0.50 0.49 0.60 0.59 0.60 0.58 0.61 0.67 0.58 0.79 0.76 0.71 1.05 0.43 1.91 1.81 1.88 1.77 1.86 1.87 1.97 1.98 1.98 1.91 2.08 2.04 2.15 2.21 2.09 2.23 2.31 2.30 2.50 0.33 0.31 0.34 0.37 0.36 0.40 0.41 0.52 0.49 0.48 0.50 0.53 0.51 0.64 0.62 0.58 0.62 0.61 1.31 0.04 0.06 0.07 0.08 0.08 0.07 0.06 0.06 0.07 0.09 0.10 0.10 0.10 0.10 0.10 0.11 0.11 0.12 0.06 0.8 0.5 0.5 0.8 0.6 0.7 0.9 0.7 0.7 0.4 0.8 0.7 0.7 1.1 0.5 0.5 0.6 0.8 0.8 <0.03 <0.03 <0.03 0.04 <0.03 0.03 <0.03 0.07 <0.03 <0.03 0.03 0.03 <0.03 <0.03 <0.03 <0.03 <0.03 0.03 0.05 <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 0.05 <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 0.03 0.21 0.03 0.03 0.03 0.03 0.02 0.04 0.04 0.07 0.05 0.03 0.05 0.06 0.04 0.04 0.05 0.05 0.04 0.03 0.06 1.23 1.11 1.06 0.96 1.01 1.04 1.08 1.02 1.05 0.93 1.03 1.07 0.98 0.87 0.83 0.89 1.02 0.97 0.46 0.24 0.26 0.21 0.27 0.20 0.24 0.33 0.25 0.31 0.35 0.30 0.33 0.32 0.29 0.34 0.34 0.31 0.21 0.38 99.7 97.6 97.0 98.2 97.4 98.1 99.9 96.9 95.9 97.3 99.5 96.6 98.8 99.1 98.2 98.9 97.4 99.8 100.2 Average StDev 68.4 2.0 17.2 1.0 6.84 1.07 0.53 0.09 0.59 0.15 2.0 0.20 0.48 0.22 0.08 0.02 0.7 0.2 0.02 0.02 0.01 0.05 0.04 0.01 0.98 0.16 0.29 0.05 98.2 1.3 PA PA PA PA 01 02 03 04 63.4 65.2 62.3 64.7 18.5 16.9 18.2 14.3 6.59 6.05 7.00 8.85 1.15 1.49 1.34 1.75 1.44 3.39 2.53 2.30 2.09 1.76 1.97 1.85 0.81 1.05 1.17 1.09 0.12 0.16 0.16 0.11 0.6 <0.3 0.5 0.5 0.47 1.55 1.05 0.22 <0.03 <0.03 <0.03 0.03 0.37 1.10 0.73 0.59 0.99 1.07 1.10 1.00 0.29 0.21 0.26 0.20 96.8 100.2 98.3 97.4 Average StDev 63.9 1.3 17.0 1.9 7.12 1.22 1.43 0.25 2.41 0.80 1.92 0.14 1.03 0.15 0.14 0.03 0.5 0.2 0.82 0.60 0.01 0.01 0.70 0.31 1.04 0.05 0.24 0.04 98.2 1.5 wH wH wH wH wH wH wH wH wH wH wH wH wH wH wH wH wH wH wH wH wH wH wH wH wH wH wH wH wH 70.7 69.1 69.4 68.3 67.0 65.2 68.3 65.2 65.6 66.3 66.3 65.9 65.4 68.7 67.2 65.8 66.0 65.0 62.4 66.4 66.6 61.3 66.1 67.9 64.2 63.2 66.0 63.1 64.7 17.0 17.9 18.2 18.2 20.0 18.8 18.8 17.2 18.4 17.6 16.9 18.7 17.6 17.9 16.9 17.1 16.7 17.9 17.1 17.6 18.0 18.0 17.8 17.4 17.9 17.4 18.2 17.7 19.7 5.77 5.32 6.04 7.45 6.86 7.57 6.25 7.93 7.47 8.79 7.26 5.93 7.25 5.62 7.57 8.28 7.19 7.56 5.88 8.52 6.50 8.06 7.48 4.63 8.88 9.73 6.55 7.01 7.04 0.42 0.41 0.38 0.40 0.46 0.52 0.50 0.57 0.51 0.71 0.55 0.49 0.52 0.49 0.68 0.65 0.43 0.69 0.52 0.81 0.57 0.69 0.54 0.47 0.55 0.61 0.41 0.65 0.56 0.54 0.68 0.86 0.82 0.76 0.81 0.87 0.86 0.97 1.13 0.98 0.82 0.84 0.68 0.98 1.02 1.14 1.04 0.98 1.02 0.79 1.37 0.85 0.97 1.17 1.20 0.95 1.01 0.85 1.77 1.99 2.10 2.04 1.97 2.32 2.23 2.11 2.01 2.44 2.24 2.20 1.96 2.19 2.40 2.30 2.66 2.38 2.20 2.39 2.53 2.62 2.53 2.96 2.82 2.82 2.26 2.43 2.34 0.49 0.66 0.56 0.60 0.58 0.76 0.81 0.76 0.70 0.98 0.81 0.74 0.61 0.77 0.91 0.73 1.08 0.95 0.88 0.81 0.92 1.15 0.74 1.00 0.88 1.00 0.91 1.11 1.14 0.09 0.12 0.09 0.09 0.10 0.11 0.11 0.11 0.12 0.12 0.12 0.12 0.13 0.13 0.14 0.14 0.15 0.15 0.15 0.16 0.17 0.17 0.18 0.18 0.19 0.20 0.22 0.25 0.26 <0.3 <0.3 0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 0.3 <0.3 0.3 <0.3 0.4 0.4 0.8 <0.3 0.6 0.6 <0.3 <0.3 <0.3 1.16 1.26 0.54 0.45 1.15 0.49 0.94 1.36 1.28 1.42 1.25 0.98 1.77 1.65 1.69 1.59 0.27 1.33 0.83 0.92 0.23 0.14 0.01 1.17 0.58 0.60 2.01 2.43 0.89 <0.03 <0.03 0.03 0.03 <0.03 <0.03 <0.03 <0.03 <0.03 0.04 0.51 <0.03 <0.03 <0.03 <0.03 0.03 2.78 <0.03 5.88 <0.03 <0.03 2.40 <0.03 0.55 0.03 <0.03 0.03 <0.03 <0.03 0.02 0.02 0.02 0.07 0.04 0.05 0.05 0.10 0.07 0.14 0.10 0.07 0.07 0.13 0.07 0.07 0.06 0.09 0.08 0.16 0.09 0.20 0.03 0.04 0.08 0.11 0.04 0.07 0.10 1.01 1.20 1.07 1.27 0.97 1.27 1.23 0.84 1.04 0.89 1.08 1.11 0.86 1.12 0.80 0.78 1.00 0.84 0.95 0.74 0.99 1.14 1.03 1.02 0.99 1.05 1.04 0.95 0.89 0.24 0.17 0.23 0.21 0.40 0.30 0.27 0.42 0.32 0.38 0.20 0.28 0.33 0.24 0.35 0.32 0.26 0.36 0.34 0.36 0.37 0.31 0.29 0.15 0.28 0.32 0.22 0.28 0.40 99.2 98.8 99.7 99.9 100.3 98.2 100.4 97.5 98.6 101.1 98.3 97.4 97.4 99.7 99.8 99.0 100.1 98.5 98.6 100.2 98.0 99.3 98.4 98.5 99.1 98.8 99.0 97.2 98.9 66.1 2.1 17.9 0.8 7.12 1.16 0.54 0.11 0.93 0.17 2.32 0.28 0.83 0.18 0.15 0.04 0.08 0.04 1.01 0.14 0.30 0.07 99.0 1.0 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 Average StDev 1.05 0.58 (continued on next page) 178 D. Rosenow, Th. Rehren / Journal of Archaeological Science 49 (2014) 170e184 Table 3 (continued ) SiO2 Na2O CaO K2O MgO Al2O3 FeO TiO2 Sb2O5 CuO P2O5 Cl SO3 Analytical total 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 66.1 63.9 64.9 64.4 68.4 67.8 63.8 63.0 64.8 65.8 62.9 64.8 63.3 62.4 65.6 63.7 66.1 66.2 66.6 63.9 62.9 64.7 63.3 62.7 66.2 63.4 65.7 65.9 19.3 18.5 19.7 18.2 17.6 17.4 19.0 19.8 18.2 16.0 18.5 17.0 19.6 18.4 17.8 17.8 16.8 18.9 16.6 17.5 19.9 18.4 19.0 18.1 18.0 18.5 18.0 15.8 5.95 6.05 6.08 5.60 5.32 5.40 5.50 4.89 5.43 5.51 5.56 6.05 5.74 6.65 5.60 6.48 6.34 5.29 4.97 6.00 5.91 5.57 5.94 6.26 4.93 5.79 5.37 5.46 0.44 0.41 0.42 0.28 0.40 0.39 0.36 0.33 0.34 0.44 0.42 0.43 0.36 0.37 0.43 0.55 0.45 0.49 0.42 0.51 0.40 0.47 0.49 0.40 0.44 0.41 0.45 0.39 1.04 0.99 1.04 1.07 0.83 0.81 0.99 0.97 0.78 1.18 1.06 1.01 1.06 1.11 1.00 1.08 1.02 0.71 1.20 1.07 1.12 1.17 1.19 1.18 1.02 1.27 0.96 0.95 2.34 2.33 2.40 2.45 2.57 2.53 2.69 2.65 2.61 2.69 2.69 2.57 2.67 2.64 3.00 2.63 2.62 2.45 3.13 2.70 2.77 2.97 3.01 2.78 2.82 2.96 2.98 2.88 1.07 1.12 1.15 1.16 1.31 1.37 1.37 1.43 1.49 1.50 1.51 1.55 1.56 1.58 1.58 1.71 1.71 1.73 1.77 1.78 1.78 1.79 2.56 2.87 3.01 3.01 3.27 3.76 0.30 0.33 0.32 0.41 0.40 0.37 0.51 0.54 0.47 0.46 0.45 0.64 0.53 0.41 0.62 0.43 0.53 0.31 0.72 0.78 0.60 0.53 0.52 0.58 0.31 0.55 0.67 0.61 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 1.85 1.80 1.91 2.07 1.98 1.52 2.30 2.25 2.25 2.45 2.05 2.54 2.17 2.41 2.01 1.95 2.74 2.13 2.77 1.86 1.96 2.13 1.49 2.05 0.90 1.50 0.96 2.18 0.05 <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 0.03 <0.03 <0.03 <0.03 0.03 <0.03 0.03 0.04 0.03 0.04 <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 0.03 <0.03 <0.03 0.05 0.04 0.05 0.04 0.05 0.05 0.03 0.05 0.04 0.03 0.05 0.04 0.06 0.06 0.04 0.09 0.05 0.08 0.04 0.05 0.06 0.05 0.09 0.09 0.11 0.08 0.14 0.19 1.17 1.11 1.14 1.23 1.03 1.06 1.08 1.07 1.07 0.92 1.03 0.84 1.14 0.97 0.81 0.87 0.87 0.98 1.00 0.86 1.06 0.99 0.83 0.86 1.07 1.01 0.91 0.89 0.25 0.24 0.28 0.15 0.19 0.22 0.25 0.27 0.25 0.21 0.22 0.26 0.23 0.24 0.27 0.31 0.24 0.32 0.18 0.29 0.23 0.23 0.45 0.29 0.26 0.24 0.26 0.18 100.0 96.9 99.4 97.1 100.1 98.9 97.8 97.3 98.3 97.4 96.5 97.9 98.5 97.2 98.9 97.9 99.5 99.6 99.5 97.3 98.7 99.1 98.9 98.4 99.1 98.7 99.7 99.4 Average StDev 64.8 1.6 18.1 1.1 5.70 0.45 0.42 0.06 1.03 0.13 2.70 0.21 1.84 0.71 0.50 0.13 <0.3 2.01 0.44 0.02 0.01 0.06 0.04 0.99 0.11 0.25 0.06 98.5 1.0 ukn 72.0 16.3 4.95 0.60 0.55 2.82 0.47 0.07 <0.3 0.01 0.02 0.06 0.90 0.13 98.9 H H H H H H H H H H H H H H H H H H H H H H H H H H H H based glass group set between the antimony- and manganesedecoloured and the two HIMT groups, respectively. The six manganese-decoloured glasses are all but one from cast vessels. They have between 0.5 and 1.3 wt% manganese oxide. Compared to the antimony-decoloured glass, they have higher calcium oxide, alumina and phosphorus, and signiﬁcantly lower levels of iron oxide and titania. The antimony-decoloured glass has between half and one percent antimony oxide, and relatively low levels of calcium oxide Fig. 5. Plot of the four natron groups from Bubastis. Note the good match of HIMT glass from Bubastis with published HIMT analyses, and the absence of glass with a Wadi Natrun signature. Levantine I glass is also not represented. Graph based on Freestone et al. (2005). MnO (5e8.5 wt%) and alumina (typically around 2 wt%, reaching up to 2.5 wt%). Potash, magnesia and iron oxide are all around half of a percent, and titania from 0.05 to 0.11 wt%. This closely matches data published by Paynter (2006), Silvestri (2008) and Schibille (2011) for contemporary antimony-decoloured glass found in Britain, northern Italy and Albania. Four glasses have higher potash (1.2e1.8 wt% K2O), magnesia (1.4e3.4 wt% MgO), and very high phosphorus oxide (0.4e1.1 wt% P2O5) compared to both the early and later glasses. These elevated values are the reason for interpreting the glass as plant-ash based, in line with arguments developed ﬁrst by Brill (1970) for Egyptian Late Bronze Age glass. Alternatively, the elevated levels could originate from contamination by fuel ash during extended periods of heating (Paynter, 2008; Rehren et al., 2010: 75e76, Schibille, 2011: 2940); further research is necessary to understand this issue better. The ‘weak HIMT’ group has from 0.5 to 1.1 wt% iron oxide, from 0.1 to 0.2 wt% titania, and typically between 0.5 and 2 wt% manganese oxide. Its calcium oxide content ranges from about 6 to about 9 wt%, and alumina ranges from 2 to 3 wt%. Potash concentrations are around one half of a percent, while magnesia is as high as iron oxide e around one percent by weight. This compositional pattern clearly differs from typical HIMT glass; in particular, the calcium oxide levels are too high by comparison, and show a slight positive correlation with alumina (Fig. 5) not normally seen in HIMT glass. Despite a basic similarity in composition, the colour of some of these samples does not always match the olive green tint of typical HIMT glass. The ﬁnal large group among the analysed samples is made from unambiguous HIMT glass. Iron oxide in this ‘strong’ HIMT group ranges from 1 to more than 3 wt%, manganese oxide from 1.5 to 2.5 wt%, and titania from 0.3 to nearly 0.8 wt%. Calcium oxide levels D. Rosenow, Th. Rehren / Journal of Archaeological Science 49 (2014) 170e184 179 Fig. 6. Glass vessel fragments from Bubastis. Scale 1:2 (Drawings: Daniela Rosenow, digitalisation: Mandy Mamedow). are relatively narrowly set between 5 and 6.5 wt%, while alumina ranges from 2.3 to more than 3 wt%. Potash is present at just under half a percent, while magnesia levels fall closely around 1 percent; all these values are fully compatible with published HIMT analyses from other assemblages elsewhere (Mirti et al., 1993; Freestone, 1994; see particularly the discussion of weak and strong HIMT glass from Britain and France in Foster and Jackson, 2009: 193e4). One sample did not meet our criteria to be allocated to one of the groups above; it is compositionally closest to the Sb-decoloured glass, but has no antimony above the detection limit, and deviates also in its content in lime (too low) and alumina (too high). We therefore left it unassigned and labelled it ukn for unknown. 5. Discussion The compositional groups identiﬁed among the Tell Basta samples are discussed below in chronological order. The manganese-decoloured, antimony-decoloured and the plant-ash based glasses fall almost all into the ﬁrst to third centuries AD. The weak HIMT glasses potentially overlap with these early groups and continue into the ﬁfth century AD, while the strong HIMT glasses are all from the fourth century or later. It has to be stressed here that the dating of individual samples is done purely on typological grounds, often with rather long run times of some types stretching over several centuries, and not on stratigraphic or other evidence that would date speciﬁc ﬁnds more narrowly. 5.1. Early decoloured glass The early date of the six manganese-decoloured glasses is noteworthy, as is the fact that ﬁve of the six are from mould-cast vessels (see Table 1, Mn 01-06, with primary typological reference). Two fragments belong to cast ribbed bowls, one represents the handle of a skyphos and two fragments are rims of cast hemispherical grooved bowls (see Fig. 6.3). The blown fragment belongs to a beaker with wheel-cut horizontal grooves. The early use of 180 D. Rosenow, Th. Rehren / Journal of Archaeological Science 49 (2014) 170e184 which can therefore not be linked directly to the Wadi Natrun production sites. Elsewhere in the Roman world, decoloured glass, particularly by antimony, is often seen in more high-class products, and being made of purer raw materials and the presumably expensive antimony (Jackson, 2005; Paynter, 2006; Nenna, 2007; Silvestri et al., 2008; Foster and Jackson, 2010). This does not apply for the Bubastis material, where antimony-decoloured glass seems to predominate heavily over manganese-decoloured glass. It is remarkable that the analysed assemblage does not contain any naturally-coloured aqua or blue-green glass, despite the fact that during the ﬁrst three centuries AD in many parts of the Roman Empire this ‘Roman blue/green’ glass was most commonly used for vessels (Paynter, 2006; Silvestri, 2008). 5.2. Plant-ash glass Fig. 7. Comparison of the four plant ash glasses from Bubastis to typical compositional ﬁelds of other glasses. The best match is with Egyptian LBA glasses (data from Smirniou and Rehren, 2013, and references therein). Graph based on Freestone (2006). manganese-decoloured glass is consistent with the occurrence of the same glass among Hellenistic assemblages in the eastern Mediterranean (e.g. Connolly et al., 2012), while elsewhere in Europe, manganese-decoloured glass only appears much later, as in Southern France (Foy et al., 2000: 54e56, corresponding groupe 3) or Italy (Silvestri et al., 2008), where manganese seems to have replaced antimony as a decolourant only at the end of the second/ start of the third century AD, or in Roman Britain, where the use of manganese as a decolourant also does not seem to start before the fourth century AD (Foster and Jackson, 2010). One of the six analysed fragments (Mn 06) is dark blue coloured by cobalt (0.05 wt%) and copper oxides; this sample has also much higher iron oxide content than the others in this group. This is consistent with the observation from contemporary Pergamon (Rehren et al., 2014), where dark blue glasses are also coloured by a combination of cobalt, copper and iron oxides. This colorant has some similarity with the cobalt-blue colorant used earlier in New Kingdom Egypt (Smirniou and Rehren, 2013), and may indicate a continuity of its exploration well into the ﬁrst millennium AD. Antimony-decoloured glass predominates at Roman Bubastis. It has been used for mould cast vessels, vessels with a folded/ tubular base ring (see Fig. 6.1), an applied foot ring (see Fig. 6.2) or pinched feet elements, a wall fragment with cut circular facettes, the rim of an aryballos, possibly an indented beaker with a thick base, a small container and a bottle or ﬂask with ﬂaring rim and applied thread, all of which e apart from one waster (Sb 12) e have typological parallels in the ﬁrst three centuries AD (see Table 1, Sb 01-19, with primary typological reference). Comparable analytical data exists from the eastern Mediterranean, e.g. from Petra (Schibille et al., 2012) or Pergamon (Rehren et al., 2014) where the majority of glass decoloured by antimony can be dated to the ﬁrst and second centuries AD. Elsewhere, as in Thamusida (Morocco: Gliozzo et al., 2013) and Roman Britain, these glasses are dating to the ﬁrst to third (Jackson, 2005; Paynter, 2006), or even to the fourth centuries AD (Foster and Jackson, 2010). Signiﬁcantly, antimony-decoloured glass has been identiﬁed by Picon et al. (2008) at the Wadi Natrun primary production installations in Zakik, Beni Salame and Bir Hooker, which date to the ﬁrst and second centuries AD. In particular their group wnc has close similarities to the Bubastis glass; it has, however, signiﬁcantly higher amount of soda compared to the Bubastis material Particular mention has to be made of the plant-ash based glasses (see Table 1, PA 01-04, with primary typological reference). All dark green-turquois translucent low-capacity, thick walled unguentaria fall into this group; three of these are reported here, while one sherd represents a ridged handle, probably from a relatively large and also thick walled transport or storage container. Large quantities of such unguentaria are also known from other Roman-period sites in Egypt (such as in Tuna, M. Flossmann pers. com.), and we hope to analyse these in the near future. The only published comparable contemporary glass vessels from Egypt and apparently made of plant ash glass are those deriving from Wadi Natrun (Picon et al., 2008) and several kohl ﬂacons and unguentaria, dating to the second and third centuries AD, mentioned among the glass ﬁnds in the Louvre collection (Arveiller-Dulong and Nenna, 2005; Nenna et al., 2005). Further examples of this composition have been published from Britain and France (Jackson et al., 2009) as well as Italy (Gallo et al., 2013) and Albania (Schibille, 2011). Plant ash glass is very rare in the eastern Mediterranean during the ﬁrst millennium BC and the ﬁrst half of the ﬁrst millennium AD. Having dominated Egyptian and Mesopotamian glassmaking during the second millennium BC, it is replaced in Egypt and the eastern Mediterranean around 1000 BC by mineral-natron based glass (Schlick-Nolte and Werthmann, 2003). However, plant-ash based glassmaking persisted in the Sasanian Empire, to the east of the Euphrates, from where it may have found its way back into the west following the collapse of natron supply in the eighth or ninth century AD (Whitehouse, 2002; Shortland et al., 2006). Islamic-period glassmaking appears to be concentrated in Syro-Palestine, with little if any Egyptian plant-ash glass making known from this period (Freestone et al., 2009). Against this traditional narrative, Picon et al. (2008) in their study of Roman glass from the Wadi Natrun report four plant-ash glass ﬁnds which they link to a local Egyptian production. A comparison of their analyses with ours, however, shows signiﬁcant differences. The Wadi Natrun plant ash glasses are very rich in lime (10e16 wt%), alumina (4e7 wt%) and iron oxide (2e3 wt %), and very low in soda (9.5e12.5 wt%). This composition is very unusual and does not resemble the plant ash glass analyses from Bubastis, or elsewhere. We therefore do not link the Bubastis samples to a production from Wadi Natrun. The geographical origin of the glass used to produce these vessels is of considerable interest, as the presence of several such vessels in Tell Basta could imply that the Roman town was still engaged in long-distance trade if the glass were indeed coming from east of the Euphrates. To discuss this, we can look at both the composition of the ﬂux and that of the sand used to make this glass. There appears to be a tendency that the levels of potash and magnesia in plant ash glasses follow a broad trend of increasing levels from Syria eastwards (Freestone, 2006). Using this criterion D. Rosenow, Th. Rehren / Journal of Archaeological Science 49 (2014) 170e184 suggests that the relevant values found in the Tell Basta assemblage are far too low to assume an import of these vessels from the Sasanian Empire; they are also lower than the levels found in most Islamic glass from Syria (Fig. 7). The closest match in their alkali composition for PA 02-04 is with the relatively low-potash New Kingdom cobalt-blue glasses found both in Egypt and Mycenaean Greece (Smirniou and Rehren, 2013) for which production evidence has been found in Amarna (Smirniou and Rehren, 2011). This suggests that these Roman-period plant ash glasses might also have been made in Egypt. The elevated concentrations of sand-derived components in these glasses, particularly iron oxide, titania and manganese oxide, are also more consistent with glasses typically linked to an Egyptian origin than a Levantine one. However, the phosphorus content of these glasses exceeds that of most other glasses from the Bronze Age and Classical Antiquity, reaching more than one percent by weight in one sample. It is hoped that our ongoing research on vessels of this type across Egypt will shed more light on the chemical composition and geographic and chronological distribution of this intriguing glass group. For now, we note that only these particular vessels were made consistently using plant ash glass rather than mineral-natron based glass. Their existence strongly indicates that production of plantash glass persisted in Egypt a millennium after the introduction of mineral-natron based glassmaking, and at least half a millennium before the re-introduction of plant ash-based glassmaking in the Islamic period, as already observed by Picon and co-workers (2008). It may be signiﬁcant that particularly low-capacity unguentaria and kohl ﬂacons seem to fall into this glass category, raising the question whether plant ash glass was produced specifically for vessels carrying low-volume high-value goods. 5.3. HIMT glass HIMT glass is widespread between the fourth and sixth centuries AD throughout the whole Mediterranean and Europe. After its introduction it soon dominated the Roman and Late Antique glass industry, and at least outside the Levant became more popular than the contemporary Levantine glasses (Freestone et al., 2002b). It is therefore not surprising that almost two thirds of the samples reported in Table 1 are HIMT glass, divided equally between a ‘weak HIMT’ and a ‘strong HIMT’ group. The weak HIMT group (see Table 1, wH 01-29, with primary typological reference) consists of an oval dish, a wall fragment of an mosaic glass vessel, an aryballos, a stemmed goblet, a stemmed bowl, an indented beaker, fragments of ﬂasks (see Fig. 6.7), base fragments of vessels with pinched feet, base fragments of vessels with single applied base rings (see Fig. 6.6) and multiple applied base rings, lamps with a pointed base, a conical hollow base or manufactured with a solid stem (see Fig. 6.5), vessel bases with a high footring, ﬂaring rims of bottles or jugs with an applied thread and rims of cups with tubular or upgoing rims (see Fig. 6.4 and 6.8). Some of these are only faintlycoloured, appearing almost colourless when thin-walled, and at least one third of the analysed objects in this group can be dated by typology to the ﬁrst three centuries AD. However, the majority of this group, and all of the strong HIMT glasses, are later ﬁnds, starting mostly in the fourth century AD, and are of dark green to brown colour (see Table 1, H 01-28, with primary typological reference). Identiﬁable objects include conical lamps or hemispherical bowls with cracked-off rims and pointed bases or a solid blob, a vessel base with multiple applied base rings, cups or bowls with various rim shapes (see Figs. 6.9, 6.10, 6.11 and 6.12), one oval dish, vessel bases with a high footring or a folded/tubular base, a stemmed bowl, four wall fragments of mould blown vessels, one wall fragment of a lamp with an applied blue blob, two fragments of 181 ridged handles, possibly deriving from transport or storage containers and the neck of a ﬂask. Despite its importance, HIMT glass has not been well deﬁned compositionally (Foster and Jackson, 2009: 193). It is generally accepted that in addition to the eponymous higher concentrations in iron, manganese and titania it also has elevated levels of magnesia as well as zirconium, chromium, barium and other trace elements, and typically a good positive correlation between alumina, iron oxide, and most of the other characteristic elements. In contrast, lime levels are normally relatively constant and scatter around 6 wt% CaO, regardless of alumina concentrations (e.g. Fig. 5 in Freestone et al., 2002a,b). The increase in recent years in published data for HIMT glasses has resulted in the identiﬁcation of considerable compositional variability within this group (see e.g. Gallo et al., 2014; Schibille, 2011: group WD2), including the presence of HIMT glass without manganese (dubbed HIT glass: Rehren and Cholakova, 2010), and of ever more extreme concentrations of some of the characteristic elements. On the other hand, considerable uncertainty exists regarding the lower end of acceptable HIMT compositions or, in other words, how little iron, manganese and titania can a glass have and still be called HIMT? Here is not the place to discuss this, but sufﬁce it to say that Foster and Jackson (2009) consider a group of glass with an average of 0.6 wt% FeO, 0.1 wt% TiO2 and 1 wt% MnO still as (weak) HIMT. We adopt the concept of weak HIMT here, even though Foster and Jackson’s (2009) weak HIMT group is compositionally different from the Tell Basta weak HIMT. In some aspects, the weak HIMT glass from Tell Basta forms a continuum with the strong HIMT group (Figs. 8 and 9). However, it differs from typical HIMT glass in its higher calcium oxide content (Fig. 5). We interpret the existence of weak and strong HIMT glass to indicate the use of two only broadly similar sand sources, possibly in geographical proximity, but probably producing glass independently of each other as indicated by the chronological and compositional differences between the two groups. Compositionally closest to the weak HIMT group, including the higher lime levels and despite some subtle differences in the alumina and iron oxide ratios, is a set of glasses from northern Europe (Saxon I: Freestone et al., 2008), dating from 400 to 550 AD. Our identiﬁcation of glass of this composition as a major group in Egypt supports Freestone’s assumption that the Saxon glass was an import, and that its elevated levels of HIMT indicators did not result from the repeated recycling of earlier Roman glass following the end of Roman rule in northern Europe. On current Fig. 8. Weak HIMT glass plots between the decoloured glasses and HIMT glass. HIMT glass is split into two groups, with six samples having higher iron oxide relative to titania than the majority of the HIMT glass. 182 D. Rosenow, Th. Rehren / Journal of Archaeological Science 49 (2014) 170e184 Fig. 9. Both weak HIMT and HIMT show a positive correlation between alumina and iron oxide, in contrast to Levantine I glass. The correlation is more pronounced for HIMT glass. Note also the group of six HIMT samples with excess iron oxide. The two decoloured glasses with high iron oxide content are coloured by cobalt, which is associated with increased iron oxide. evidence one can suspect that it is of Egyptian rather than Levantine origin; however, more work is clearly needed to better understand the HIMT glass family in all its compositional complexity. Types found among this glass group include vessels made of mosaic glass, indented beakers, vessels with pinched feet, oval dishes, stemmed goblets, and an aryballos. Their dating covers a relatively wide time span, ranging from the ﬁrst to the seventh centuries AD, and includes a number of intentionally coloured samples. The strong HIMT group matches compositionally published data for HIMT glass; noteworthy here is the emergence of a small subgroup characterised by excess iron oxide relative to titania and alumina compared to the bulk of this group (Figs. 8 and 9); this has been observed elsewhere before (Rehren and Cholakova, 2010) and again underlines the compositional complexity of this glass group, as well as its super-regional importance for much of Late Antiquity. 5.4. Absence of Egypt I and II glass compositions In the introduction we mentioned the glass groups Egypt I and II; Egypt II is chronologically outside the frame of our study, and its absence from our data hence not surprising. In contrast, the absence of Egypt I glass was unexpected. It is linked to production in the Wadi Natrun since it shares some characteristics with the composition of glass ﬁnds from primary glassmaking installations there published by Picon et al. (2008), such as very low levels of lime and rather high soda levels. These primary glassmaking furnaces are dated to the ﬁrst two centuries of the ﬁrst millennium AD, contemporary with much of the glass from Bubastis. It is remarkable then that none of the glasses analysed to date from Tell Basta show this very characteristic Wadi Natrun signature, despite the relative proximity of the two sites. 6. Conclusion The aim of this paper was to provide a ﬁrst insight into glass supply and consumption at an Egyptian town between the ﬁrst century BC and the end of the sixth century AD. Situated in the Eastern Nile Delta, Bubastis can be expected to have been well integrated into the Roman trade e as indicated by the contemporary ceramic ﬁnds from the city. On the basis of compositional analysis, the glass discovered here falls into ﬁve main groups, four of which are well known from elsewhere. In the ﬁrst three centuries of the ﬁrst millennium AD, manganese and antimony-decoloured glass compositions dominate, while a previously little known highphosphorus plant-ash glass was used for unguentaria. The later part of the assemblage consists of two different HIMT glass groups, one relatively low and one rather high in iron, manganese and titanium oxides. With the exception possibly of the manganesedecoloured glass, contributing less than 10% of the analysed sample and certainly even less of the entire assemblage, none of the material appears to be imported from outside Egypt, painting a picture in contrast to what the ceramic indicates. Three observations are of particular interest and underscore the wider signiﬁcance of this data set. First: Manganese-decoloured and antimony-decoloured glass is evident here as elsewhere, with nearly identical base glass compositions and levels of additives as seen elsewhere in the Roman Empire. For Bubastis, this conﬁrms that the town was integrated into the wider trade network of the Roman and Late Antique world. The data is consistent with a model of a limited number of primary glass producers serving super-regional markets, spanning the entire Roman world, and beyond, from Britain and France to Turkey and Egypt. There is, however, no evidence so far for the presence of the faintly-coloured Roman blue/green glass, which during this time is so wide-spread in the Northern Provinces. In contrast, in Bubastis the antimony-decoloured glass is not a glass chosen only for high-status objects, but appears during the early period to be the predominant glass type, apart from a few possibly imported and relatively early samples of manganese-decoloured glass. The absence of Roman b/g glass could indicate that this particular group was not produced in Egypt, while the Syro-Palestine area has been mentioned repeatedly in this context (e.g. Nenna, 2000; Foy et al., 2003; Gliozzo et al., 2013). The existence of plant-ash glass in this early period, probably made regionally, is intriguing, particularly with its close association to a particular vessel type. Second: Only HIMT glasses were used at Bubastis from the fourth century AD onward, with two co-existing sub-groups recognisable by their different levels of diagnostic elements. No evidence for the use of Levantine glass has been found, despite those glasses dominating archaeological assemblages in current-day Israel and Jordan, where in turn HIMT glass is rather rare (Kato et al., 2009; Rehren et al., 2010). This suggests a strong element of regional preference in glass consumption, most likely based on proximity to the primary producer. This, in turn, could indicate that the earlier Roman blue/green glass was also made in the Levant, and hence not used in Bubastis. The chronological and compositional difference between the two HIMT groups, and the difference in composition between these two and some other HIMT assemblages reported elsewhere, indicates that there is signiﬁcant systematic variation within the HIMT group. This could indicate that there were a number of contemporary and consecutive glassmaking installations active using similar but different sand sources, within the broader HIMT deﬁnitions. Also, weak HIMT glass is compositionally very close to Anglo-Saxon I (Freestone et al., 2008), Frankish German (Wedepohl et al., 1997) and Merovingian French (Velde, 1990) glass dating to the 5th and 6th centuries AD, conﬁrming that fresh raw glass from Egypt was still reaching Europe after the collapse following the departure of the Roman army. It is hoped that a comprehensive study of the HIMT glass family and its sub-groups will shed more light on this, even though trying to order the increasing numbers of analyses seem to make the picture more complex rather than clearer e a bit like herding cats. Third: The complete absence at Bubastis of low-lime glass compositions typically linked to the Wadi Natrun, such as the D. Rosenow, Th. Rehren / Journal of Archaeological Science 49 (2014) 170e184 primary production remains reported by Picon et al. (2008) or Freestone’s Egyptian I glass based on analyses by Gratuze and Barrandon (1990) is remarkable, given that the city is located relatively close to the Wadi Natrun. This underscores how little we really know about this glass group, and its signiﬁcance relative to the other, better-known groups. The material presented here currently comprises the only available substantial data set for glass compositions from Roman and late Antique Egypt, making it difﬁcult to generalise our observations beyond the statements just made. Clearly, more analytical data is required to support a more reﬁned discussion about glass supply and consumption in Roman and Late Antique Egypt; some of this is currently underway as part of the Marie Curie project Glass in Late Antiquity e Science and Society, of which this is the ﬁrst outcome. Acknowledgements A debt of gratitude is due to the Egyptian Ministry of State for Antiquities, for permitting the work reported here to be undertaken. We are grateful to Professor Ian Freestone, who ﬁrst pointed out the close similarity of our ‘weak HIMT’ glass composition to the Anglo-Saxon I glasses analysed by him, and for inspiring and informative discussions throughout e even where we did not always follow his suggestions. Marie-Dominique Nenna provided help with the dating of the weak HIMT group fragments, for which we are most thankful. Several funding bodies supported the underlying excavations of the Tell Basta Project: The German Research Foundation, the German Ministry of Foreign Affairs and the Egypt Exploration Society generously funded our ﬁeldwork at site. The glass related research is supported by a current Marie Curie IntraEuropean Fellowship within the 7th European Community Programme (project number 298127, FP7-PEOPLE-2011-IEF, to D.R.). Initial funding for this project was provided by the Fritz Thyssen Foundation, establishing the collaboration between the two authors at the UCL Institute of Archaeology. Comments by three anonymous reviewers are greatly appreciated; any remaining errors are ours. References Arveiller-Dulong, V., Nenna, M.-D., 2005. Les verres antiques du musée du Louvre II, Vaisselle et contenants du Ier siècle au début du VIIe siècle après J.-C., Paris. Baxter, M.J., Cool, H.E.M., Jackson, C.M., 2005. Further studies in the compositional variability of colourless Romano-British vessel glass. Archaeometry 47, 47e68. Bailey, D., 2007. Glass. In: Maxﬁeld, V.A., Peacock, D. (Eds.), The Roman Imperial Quarries Survey and Excavations at Mons Porphyrites 1994e1998, The Excavations, London, vol. II, pp. 233e266. Brun, J.-P., 2003a. Les objets en verre. In: Cuvigny, H. (Ed.), La route de Myos Hormos, L’armée romaine dans le désert oriental d’Égypte, Cairo, pp. 515e537. Brun, J.-P., 2003b. Le verre dans le désert Oriental d’Égypte. Contextes datés. In: Foy, D., Nenna, M.-D. (Eds.), Échanges et commerce du verre dans le monde antique. Montagnac, pp. 377e387. Brun, J.-P., 2011. Les objets en verre. In: Cuvigny, H. (Ed.), Didymoi. Une garnison romaine dans le désert Oriental d’Égypte, 1. Les fouilles et le matériel, Cairo, pp. 215e241. Connolly, P., Rehren, Th, Doulgeri-Intzesiloglou, A., Arachoviti, P., 2012. The Hellenistic Glass of Pherai, Thessaly. Annales du 18e Congrès de l’Association Internationale pour l’Histoire du Verre, Thessaloniki, pp. 91e97, 2009. Degryse, P., Schneider, J., 2008. Pliny the Elder and SreNd isotopes: tracing the provenance of raw materials for Roman glass production. J. Archaeol. Sci. 35, 1993e2000. Faiers, J., 2013. Late Roman glassware and pottery from Amarna and related studies. Egypt Explor. Soc. Memoirs 102, 5e38. Foster, H.E., Jackson, C.M., 2009. The composition of ‘naturally coloured’ late Roman vessel glass from Britain and the implications for models of glass production and supply. J. Archaeol. Sci. 36, 189e204. Foster, H.E., Jackson, C.M., 2010. The composition of late Romano-British colourless vessel glass: glass production and consumption. J. Archaeol. Sci. 37, 3068e3080. Foy, D., 2001. Secteur Nord de Tebtynis (Fayyoum), Le verre byzantin et islamique. Ann. Islamol. 35, 465e489. 183 Foy, D., Jezegou, M.P., 1997. Une epave chargee de !ingots et de vaisselle de verre, un temoignage exceptionnel du commerce et de la technologie du verre en Mediterranee antique. Verre 3.3, 65e70. Foy, D., Vichy, M., and Picon, M., 2000. Lingots de verre en Mediterranee occidentale (IIIe siècle av.J.-C. e VIIe siècle ap.J.-C.), Annales du 14e Congrès de l’Association Internationale pour l’Histoire du Verre: Milano, 1998, 51e57. Foy, D., Picon, M., Vichy, M., Thirion-Merle, V., 2003. Caractérisation des verres de la ﬁn de l’Antiquité en Méditerranée occidentale: l’émergence de nouveaux courants commerciaux. In: Foy, D., Nenna, M.-D. (Eds.), Échanges et commerce du verre dans le monde antique: actes du colloque de l’Association française pour l’archéologie du verre, Aix-en-Provence etMarseille, 7e9 juin 2001. Éditions Monique Mergoil, Montagnac, pp. 41e86. Freestone, I.C., 1994. Appendix: chemical analysis of ‘raw’ glass fragments. In: Hurst, H.R. (Ed.), Excavations at Carthage, 1. The Circular Harbour, North Side. The Site and Finds other than Pottery. British Academy Monographs in Archaeology, No. 4, vol. II. Oxford University Press, Oxford, p. 290. Freestone, I., 2006. Glass production in Late Antiquity and the Early Islamic period: a geochemical perspective. In: Maggetti, M., Messiga, B. (Eds.), Geomaterials in Cultural Heritage, Geological Society, London, Special Publications, vol. 257, pp. 201e216. Freestone, I., Gorin-Rosen, Y., Hughes, M.J., 2000. Primary glass from Israel and the production of glass in Late Antiquity and the Early Islamic period. In: Nenna, M.-D. (Ed.), La Route du Verre: ateliers primaires et secondaires de verriers du second millénaire av. J.-C. au Moyen Âge, Travaux de la Maison de l’Orient, 33, pp. 65e82. Freestone, I.C., Ponting, M., Hughes, M.J., 2002a. The origins of Byzantine glass from Maroni Petrera, Cyprus. Archaeometry 44, 257e272. Freestone, I.C., Greenwood, R., Gorin-Rosen, Y., 2002b. Byzantine and early Islamic glassmaking in the Eastern Mediterranean: production and distribution of primary glass. In: Kordas, G. (Ed.), Hyalos e Vitrum e Glass. History Technology and Conservation of glass and vitreous materials in the Hellenic World. 1st International Conference Rhodes e Greece 1e4 April 2001, Athens, pp. 167e174. Freestone, I.C., Wolf, S., Thirlwall, M., 2005. The Production of HIMT Glass: Elemental and Isotopic Evidence. Annales du 16e Congrès de l’Association Internationale pour l’Histoire du Verre, London, pp. 153e157, 2003. Freestone, I.C., Hughes, M.J., Stapleton, C.P., 2008. Composition and production of Anglo Saxon glass. In: Evison, V.I. (Ed.), Catalogue of Anglo-Saxon Glass Vessels in the British Museum. The British Museum, London, pp. 29e46. Freestone, I.C., Wolf, S., Thirlwall, S., 2009. Isotopic composition of glass from the Levant and the south-eastern Mediterranean Region. In: Degryse, P., Henderson, J., Hodgins, G. (Eds.), Isotopes in Vitreous Materials (Studies in Archaeological Sciences), pp. 31e52. Gallo, F., Marcante, A., Silvestri, A., Molin, G., 2014. The glass of the “Casa delle Bestie Ferite”: a ﬁrst systematic archaeometric study on Late Roman vessels from Aquileia. J. Archaeol. Sci. 41, 7e20. Gallo, F., Silvestri, A., Molin, G., 2013. Glass from the Archaeological Museum of Adria (North-East Italy): new insights into Early Roman production technologies. J. Archaeol. Sci. 40, 2589e2605. Gliozzo, E., Santagostino Barbone, A., D’ Acapito, F., 2013. Waste glass, vessels and window panes from Thamusida (Morocco), Grouping natron-based blue-green and colourless Roman glasses. Archaeometry 55, 609e639. Gorin-Rosen, Y., 2000. The ancient glass Industry in Israel: summary of the ﬁnds and new discoveries. In: Nenna, M.-D. (Ed.), La Route du Verre: ateliers primaires et secondaires de verriers du second millénaire av. J.-C. au Moyen Âge, Travaux de la Maison de l’Orient, vol. 33, pp. 49e63. Gratuze, B., Barrandon, J.-N., 1990. Islamic glass weights and stamps: analysis using nuclear techniques. Archaeometry 32, 155e162. Habachi, L., 1957. Tell Basta: Tell Basta, Supplément aux Annales du Service des Antiquités de l’Égypte: Cahier N 22, Cairo. Harden, D.B., 1936. Roman glass from Karanis found by the University of Michigan archaeological expedition in Egypt, 1924e1929, Michigan. Hill, M., Nenna, M.-D., 2003. Glass from Ain et-Turba and Bagawat Necropolis in the Kharga oasis (Egypt), Annales du 15e Congrès de l’Association Internationale pour l’Histoire du Verre: New York-Corning 2001, 88e92. Isings, C., 1957. Roman Glass from Dated Finds. Archaeologica Traiectina, Groningen. Jackson, C.M., 2005. Making colourless glass in the Roman period. Archaeometry 47, 763e780. Jackson, C., Price, J., Lemke, C., 2009. Glass Production in the 1st Century AD: Insights into Glass Technology, Annales du 17e Congrès de l’Association Internationale pour l’Histoire du Verre: Antwerp 2006, 150e155. Jennings, S., 2006. Vessel glass from Beirut BEY 006, 007 and 045. Berytus, 48e49. Kato, N., Nakai, I., Shindo, Y., 2009. Change in chemical composition of early Islamic glass excavated in Raya, Sinai Peninsula, Egypt: on-site analysis using a portable X-ray ﬂuorescence spectrometer. J. Archaeol. Sci. 36, 1698e1707. Kucharczyk, R., 2004. Two Roman toilet bottles from Kom el-Dikka. Pol. Archaeol. Mediterr. 15, 43e45 (Reports 2003). Kucharczyk, R., 2006. Late Roman/Early Byzantine glass from the Auditoria on Kom el-Dikka in Alexandria. Pol. Archaeol. Mediterr. 17, 45e50 (Reports 2005). Kucharczyk, R., 2010. Glass from area F on Kom el-Dikka (Alexandria). Pol. Archaeol. Mediterr. 19, 56e70 (Reports 2007). Marchini, C., 1999. Glass from the 1993 Excavations at Ismant El-Kharab. In: Hope, C., Mills, A. (Eds.), Dakhleh Oasis Project: Preliminary Reports on the 1992e93 and 1993e94 Field Seasons, Oxford, pp. 75e82. Meyer, C., 1992. Glass from Quseir al Qadim and the Indian Ocean Trade, Chicago. 184 D. Rosenow, Th. Rehren / Journal of Archaeological Science 49 (2014) 170e184 Mirti, P., Casoli, A., Appolonia, L., 1993. Scientiﬁc analysis of Roman glass from Augusta Praetoria. Archaeometry 35, 225e240. Nenna, M.-D., Vichy, M., Picon, M., 1997. L’ Atelier de verrier de Lyon, du Ier siècle après J.-C., et l’origine des verres « Romains». Rev. d’Archéomét. 21, 81e87. Nenna, M.-D., 2000. Ateliers de production et sites de consommation en Égypte, Ve s. av. J.-C. VIIe s. apr. J.-C. In: Annales du 14e congrès de l’AIHV: Venise-Milan 1998, 20e24. Nenna, M.-D., 2007. Production et commerce du verre à l’époque impériale: nouvelles découvertes et problématiques. Facta 1, 125e147. Nenna, M.-D., Picon, M., Thirion-Merle, V., Vichy, M., 2005. Ateliers primaires du Wadi Natrun: nouvelles decouvertes. Annales du 16e Congrès de l’Association Internationale pour l’Histoire du Verre, London, pp. 59e63, 2003. Nenna, M.-D., 2010. The glass vessel Decor of a Funerary Chapel in the Bagawat Necropolis, Kharga Oasis, Egypt. In: Drauschke, J., Keller, D. (Eds.), Glass in Byzantium e Production, Usage, Analyses. RGZM, Mainz, pp. 199e214. Paynter, S., 2006. Analyses of colourless Roman glass from Binchester, County Durham. J. Archaeol. Sci. 33, 1037e1057. Paynter, S., 2008. Experiments in the reconstruction of Roman Wood-Fired Glassworking furnaces: waste products and their formation processes. J. Glass Stud. 50, 271e290. Peacock, D., 2011. Glass. In: Peacock, D., Blue, L. (Eds.), Myos Hormos e Quseir alQadim: Roman and Islamic Ports on the Red Sea, Finds From excavations 1999e2003, Oxford, vol. 2, pp. 57e78. Picon, M., Vichy, M., 2003. D’orient en occident: l’origine du verre a l’epoque romaine et durant le haut Moyen Age. In: Foy, D., Nenna, M.-D. (Eds.), Échanges et commerce du verre dans le monde antique: actes du colloque de l’Association française pour l’archéologie du verre, Aix-en-Provence et Marseille, 7e9 juin 2001. Éditions Monique Mergoil, Montagnac, pp. 17e32. Picon, M., Thirion-Merle, V., Vichy, M., 2008. Les verres au natron et les verres aux cendres du Wadi Natrun (Egypte). Bull. l’Assoc. Francaise l’Archeol. Verre, 36e41. Rehren, Th, 2000. Rationales in Old World base glass compositions. J. Archaeol. Sci. 27, 1225e1234. Rehren, Th, Cholakova, A., 2010. The early Byzantine HIMT glass from Dichin, Northern Bulgaria. Interdiscip. Stud. 22/23, 81e96. Rehren, Th, Marii, F., Schibille, N., Swann, C., Stanford, L., 2010. Glass supply and circulation in early Byzantine southern Jordan. In: Drauschke, J., Keller, D. (Eds.), Glass in Byzantium e Production, Usage, Analyses. RGZM, Mainz, pp. 199e214. Rehren, Th., Connolly, P., Milevska, B., Schibille, N., Schwarzer, H., 2014. Changes in glass composition in Pergamon (Turkey) from Hellenistic to late Byzantine times, in press. Sayre, E.V., Smith, R.W., 1961. Compositional categories of ancient glass. Science 133, 1824e1826. Shaw, I., 2003. The Oxford History of Ancient Egypt, Oxford. Schibille, N., Degryse, P., O’Hea, M., Izmer, A., Vanhaecke, F., McKenzie, J., 2012. Late Roman glass from the ‘Great Temple’ at Petra and Khirbet et-Tannur, Jordan e Technology and provenance. Archaeometry 54 (6), 997e1022. Schibille, N., 2011. Supply routes and consumption of glass in ﬁrst millennium CE Butrint (Albania). J. Archaeol. Sci. 38, 2939e2948. Schlick-Nolte, B., Werthmann, R., 2003. Glass vessels from the burial of Nesikhons. J. Glass Stud. 45, 11e34. Shortland, A.J., Schachner, L., Freestone, I.C., Tite, M.S., 2006. Natron as a ﬂux in the early vitreous materials industry: sources, beginnings and reasons for decline. J. Archaeol. Sci. 33, 521e530. Silvestri, A., 2008. The coloured glass of Iulia Felix. J. Archaeol. Sci. 35, 1489e1501. Silvestri, A., Molin, G., Salviulo, G., 2008. The colourless glass of Iulia Felix. J. Archaeol. Sci. 35, 331e341. Smirniou, M., Rehren, Th, 2011. Direct evidence of primary glass production in Late Bronze Age Amarna, Egypt. Archaeometry 53, 58e80. Smirniou, M., Rehren, Th, 2013. Shades of blue e cobalt-copper coloured blue glass from New Kingdom Egypt and the Mycenaean World: a matter of production or colorant source? J. Archaeol. Sci. 40, 4785e4792. Tanimoto, S., Rehren, Th, 2008. Interactions between silicate and salt melts in LBA glassmaking. J. Archaeol. Sci. 35, 2566e2573. Velde, B., 1990. Alumina and calcium oxide contents of glass found in western and northern Europe, ﬁrst to ninth centuries. Oxf. J. Archaeol. 9, 105e117. Wedepohl, K.H., Baumann, A., 2000. The use of marine molluskan shells for Roman glass and local raw glass production in the Eifel area (Western Germany). Naturwissenschaften 87, 129e132. Wedepohl, K.H., Pirling, R., Hartmann, G., 1997. Römische und fränkische Gläser aus dem Gräberfeld von Krefeld-Gellep. Bonn. Jahrb., 177e189. Whitehouse, D., 2002. The transition from natron to plant ash in the Levant. J. Glass Stud. 44, 193e196.

RELATED PAPERS

RELATED TOPICS

Log In

Herding cats – Roman to Late Antique glass groups from Bubastis, northern Egypt (Rosenow & Rehren 2014, JAS 49)

Herding cats – Roman to Late Antique glass groups from Bubastis, northern Egypt (Rosenow & Rehren 2014, JAS 49)

Related Papers

RELATED PAPERS

RELATED TOPICS