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.