Analysis of the Composition of Glass Objects from Qumrân, Israel, and Comparison with Other Roman Glass from Western Europe - article ; n°1 ; vol.33, pg 113-121

-

English
10 Pages
Read an excerpt
Gain access to the library to view online
Learn more

Description

Travaux de la Maison de l'Orient méditerranéen - Année 2000 - Volume 33 - Numéro 1 - Pages 113-121
Dans cet article, on étudie une série de verres romains découverts sur le site archéologique de Khirbet Qumrân en Israël. L'analyse des éléments majeurs et mineurs ainsi que des traces fournit des informations sur les verres eux-mêmes et sur le site dans lequel ils ont été mis au jour. La composition du verre est pratiquement la même pour tous les objets de ce groupe important et confirme l'hypothèse que Qumrân était un centre de production de parfum. On a obtenu un meilleur aperçu des variations et/ou des similitudes de composition ainsi que des informations sur les matières premières en comparant ces verres avec d'autres verres romains provenant de sites de tout l'Empire.
In this work a Roman glass collection is studied, which was excavated at the archaeological site of Khirbet Qumrân in Israel. These glasses are well-dated (1st century AD) and were used by a particular community. The investigation of the major, minor as well as trace composition of the different objects presented an opportunity to deduce information about the glass objects themselves and the site they were found in. The composition of this large collection is nearly the same for almost all the objects and supports the view that Qumrân was a centre of the perfume industry in the Middle East at the time of production. A better insight into the variations and/or similarities in composition and information on the raw materials was obtained by comparing the glass vessels to other Roman glasses from locations throughout the Roman Empire.
9 pages
Source : Persée ; Ministère de la jeunesse, de l’éducation nationale et de la recherche, Direction de l’enseignement supérieur, Sous-direction des bibliothèques et de la documentation.

Subjects

Informations

Published by
Published 01 January 2000
Reads 63
Language English
Document size 1 MB
Report a problem

Ann F. B. Aerts
Monsieur Koen Janssens
Bruce Velde
Freddy C. V. Adams
Helena Wouters
Analysis of the Composition of Glass Objects from Qumrân,
Israel, and Comparison with Other Roman Glass from Western
Europe
In: La Route du verre. Ateliers primaires et secondaires du second millénaire av. J.-C. au Moyen Âge. Colloque
organisé en 1989 par l'Association française pour l'Archéologie du Verre (AFAV). Lyon : Maison de l'Orient et de la
Méditerranée Jean Pouilloux, 2000. pp. 113-121. (Travaux de la Maison de l'Orient méditerranéen)
Résumé
Dans cet article, on étudie une série de verres romains découverts sur le site archéologique de Khirbet Qumrân en Israël.
L'analyse des éléments majeurs et mineurs ainsi que des traces fournit des informations sur les verres eux-mêmes et sur le site
dans lequel ils ont été mis au jour. La composition du verre est pratiquement la même pour tous les objets de ce groupe
important et confirme l'hypothèse que Qumrân était un centre de production de parfum. On a obtenu un meilleur aperçu des
variations et/ou des similitudes de composition ainsi que des informations sur les matières premières en comparant ces verres
avec d'autres verres romains provenant de sites de tout l'Empire.
Abstract
In this work a Roman glass collection is studied, which was excavated at the archaeological site of Khirbet Qumrân in Israel.
These glasses are well-dated (1st century AD) and were used by a particular community. The investigation of the major, minor as
well as trace composition of the different objects presented an opportunity to deduce information about the glass objects
themselves and the site they were found in. The composition of this large collection is nearly the same for almost all the
and supports the view that Qumrân was a centre of the perfume industry in the Middle East at the time of production. A better
insight into the variations and/or similarities in composition and information on the raw materials was obtained by comparing the
glass vessels to other Roman glasses from locations throughout the Roman Empire.
Citer ce document / Cite this document :
Aerts Ann F. B., Janssens Koen, Velde Bruce, Adams Freddy C. V., Wouters Helena. Analysis of the Composition of Glass
Objects from Qumrân, Israel, and Comparison with Other Roman Glass from Western Europe. In: La Route du verre. Ateliers
primaires et secondaires du second millénaire av. J.-C. au Moyen Âge. Colloque organisé en 1989 par l'Association française
pour l'Archéologie du Verre (AFAV). Lyon : Maison de l'Orient et de la Méditerranée Jean Pouilloux, 2000. pp. 113-121.
(Travaux de la Maison de l'Orient méditerranéen)
http://www.persee.fr/web/ouvrages/home/prescript/article/mom_1274-6525_2000_act_33_1_1878route du verre La
TMO 33, Maison de l'Orient, Lyon, 2000
ANALYSIS OF THE COMPOSITION OF GLASS OBJECTS FROM QUMRAN,
ISRAEL, AND COMPARISON WITH OTHER ROMAN GLASS
FROM WESTERN EUROPE
A. Aerts *, K. Janssens *, B. Velde **, F. Adams * and H. Wouters ***
Abstract
In this work a Roman glass collection is studied, which was excavated at the archaeological site of
Khirbet Qumrân in Israel. These glasses are well-dated (1st century AD) and were used by a particular
community. The investigation of the major, minor as well as trace composition of the different objects
presented an opportunity to deduce information about the glass objects themselves and the site they were
found in. The composition of this large collection is nearly the same for almost all the objects and supports
the view that Qumrân was a centre of the perfume industry in the Middle East at the time of production. A
better insight into the variations and/or similarities in composition and information on the raw materials was
obtained by comparing the glass vessels to other Roman glasses from locations throughout the Roman
Empire.
RÉSUMÉ
Dans cet article, on étudie une série de verres romains découverts sur le site archéologique de Khirbet
Qumrân en Israël. L'analyse des éléments majeurs et mineurs ainsi que des traces fournit des informations
sur les verres eux-mêmes et sur le site dans lequel ils ont été mis au jour. La composition du verre est
pratiquement la même pour tous les objets de ce groupe important et confirme l'hypothèse que Qumrân
était un centre de production de parfum. On a obtenu un meilleur aperçu des variations et/ou des
similitudes de composition ainsi que des informations sur les matières premières en comparant ces verres
avec d'autres verres romains provenant de sites de tout l'Empire.
DESCRIPTION OF THE ARCHAEOLOGICAL SITE AND
THE GLASS COLLECTION OF QUMRÂN
In the 1950s a Roman settlement was discovered on the north-western shore of the Dead Sea, at the
eastern edge of the Judean desert in Israel (Fig. 1). This site, Khirbet Qumrân, was excavated for the first
time in 1951 by a team led by Father R. de Vaux as an extension of the excavations of the nearby caves
where numerous manuscripts, the "Dead Sea Scrolls", were discovered; five other archaeological
campaigns followed regularly in the period 1953-1958. Few archaeological sites around the Mediterranean
have stirred up so many questions as Qumrân. A lot of controversies exist about the role of this site in early
history. The established interpretation is that Qumrân was the monastic residence of a dissident Jewish
religious order, the Essenes, who had hidden their "library" in the caves. l On the other hand, it was
suggested by Donceel et al. 2 that, instead of a religious centre, the site was characterized by industrial and
* University of Antwerp, Department of Chemistry, Universiteitsplein 1, 2610 Wilrijk, Belgium.
** Laboratoire de Géologie de l'École Normale Supérieure, 24 rue Lhomond, 75231 Paris Cedex 05, France.
*** Royal Institute for Cultural Heritage, Jubelpark 1, 1040 Brussels, Belgium
1. Baillet, Milik and Vaux 1962; Fritsch 1956; Laperrousaz 1961; Milik 1972; Vaux 1973.
2. Donceel and Doucel-Voûte 1994. 114 A. Aerts, Κ. Janssens, Β. Velde, F. Adams & H. Wouters
commercial activities in the field of the perfume industry, centred around a well-appointed villa. The
presence of a large collection of coins and luxury goods strengthens this counter-hypothesis. 3
Next to an extensive series of terracotta oil lamps and a collection of lithic objects, a group of 90
(fragments of) glass objects, such as bottles, pearls, goblets and cups was recovered during the excavations.
Almost all the glass fragments were found in the central quadrangle of the site. 4 The objects are assumed
to date from the period 4 BC-68 AD and therefore remained buried for about 1900 years. All the objects,
except for one purple coloured, were made from naturally coloured, blue-green glass. Through contact with
water in the soil the objects showed extensive surface corrosion and were in a state of advanced
decomposition.
The glass fragments from the excavations of Qumrân were confided to the Royal Institute for Cultural
Heritage (IRPA-KIK), Brussels, Belgium, where examination and restoration was carried out. First of all,
the fragments were sorted, the still adhering soil was removed and the glass pieces were cleaned. The
second intervention was consolidation and fixation of the alteration products. Different solutions based on
the acrylic resin Paraloid B72 were used. Some fragments could be regrouped and glued together. 5 In
Fig. 2, a photograph of some of the objects is shown after restoration. Parallel to the restoration, a
descriptive index was worked out in which each specimen was described individually by eighteen
properties covering the form, the dimensions, the fabrication technique, the condition of conservation and
the restoration treatment. 6 It is important to remark that in most cases the determination of the form
proposed in the index follows an extrapolation beginning from a single preserved fragment (a base, a
fragment of the body, a neck, a handle or a lip). Of the eighty nine identified objects, only four recipients
presented a complete profile after restoration. Table 1 shows the typological classification of the fragments
which were analysed.
TYPE SUBTYPE NUMBERS
Ointment vessels 4,9, 11, 13, 15, 16,26,28,31,37,
36,38,41,42,45,47,51,56
Goblets with flat base 6, 20, 29, 40
3,49 ribbed
with widened lip 10, 21b, 23b, 44, 54
moulded 12, 60, 65, 66, 73, 74
with fluted body 14
decorated with glass strings 39
incised 43,68
with vertical flattening 75
with band at the top 21
Cups ribbed with flat base 2, 18, 30, 49b, 55
moulded ribbed 35
with decoration (?) 8, 48, 63, 76
ribbed with rounded base 7
17, 53, 69 (?)
Flasks 27, 32, 50, 62
Biconical recipients 5, 23, 34
Bottles 19, 22, 25, 27b, 33, 59, 61
Chalice on foot 24
Indeterminable shapes 52,70,71,72,77
Table 1 - Typological classification of the glasses presented in the card-index.
3. Ibid.
4. Donceel and Donceel- Voûte 1994.
5. Fontaine-Hodiamont 1993.
6. FileIRPA2L/123. Analysis of the Composition of Glass Objects from Qumrân, Israel 1 15
DISCUSSION OF THE COMPOSITION
In the literature, there are a number of reports on the chemical analysis of glass from the Roman
period, excavated at various locations in Europe such as 5 samples from window glass excavated at
Wroxeter (Great-Britain), 7 103 fragments of waste glass from a secondary workshop in Mancetter and
78 from one in Leicester (Great-Britain), 8 53 glasses from the in Jalame (Israel), 9
340 fragments from different places in and around Cologne (Germany), 10 59 samples of which
25 originate from a secondary workshop in Saintes (France), the others from 5 different sites in western
France, n 58 glass fragments from Aosta (Italy) 12 and 10 glass objects from Aquileia (Italy). 13 In many
cases, however, only a few objects were found or analysed in each site and in almost all cases only major
element analysis was performed.
In the present study, where an extensive series of objects is available, a unique opportunity presented
itself to evaluate to what extent detailed investigations on the composition of the various glass objects is
able to reveal information on the provenance and history of the objects themselves and/or of the site they
were found in. Therefore, next to the use of the conventional method of electron probe X-ray microanalysis
(EPXMA), which is suitable for determining the major composition of the glass, microscopic synchrotron
radiation induced X-ray fluorescence analysis (μ-SRXRF) was employed to determine, in a completely
non-destructive manner, the trace composition of a selected set of the Qumrân glass samples. This method
allows the determination of the trace of the glass down to the ppm level. The accuracy of both
methods is 10 % or better, except for Mg since an overestimation of the spectrum background in this region
results in an underestimation of the concentration of this element. Further details about the experimental
procedures can be found elsewhere. 14
The major (75 samples) and the minor/trace composition (64 samples) of a set of glass objects was
determined and the results were compared. Taking into account the accuracy of EPXMA and μ-SRXRF,
on first sight, the compositions all appeared to be (nearly) identical and conveyed the information that all
glass samples were of the typical low magnesium/low potassium soda-lime-silica type, like most glass
from the Roman period.
A closer inspection of the major composition together with the trace element patterns revealed that
the large group of objects could be subdivided into three classes containing 45, 9 and 5 objects respectively.
As can be seen from Table 2, where the average major, minor and trace composition of these three groups
is listed, the difference between group III and group I is due to the copper, tin, antimony and lead
concentrations which are all lower in group III. Within the series of 45 objects that constitute group I, in
all cases antimony is present while in the objects of group III no antimony is detected. Next to the
previously mentioned elements, also the manganese concentration in the objects of group III is
significantly lower than in group I and II. Group II has a lower CaO concentration and also lower copper,
tin, antimony and lead concentrations than group I. Groups II and III appear to be similar with respect to
these trace metals. Because the trace fingerprints of the three groups differ only significantly in a relatively
limited number of elements, one can reasonably assume that all the objects were either made from batches
of bulk glass which were closely related or else they were manufactured using two different procedures,
7. Cox and Ford 1993.
8. Jackson et al. 1991
9. Brill 1988.
1 0. Van Rolf and Rottländer 1 990.
1 1 . Velde and Gendron 1 980.
12. Mirti, Casoli and Appolonia 1993.
13. Calvi, Tornati and Scandellari 1968.
14. Aerts 1998 and Aerts et al. submitted. 116 A. Aerts, Κ. Janssens, Β. Velde, F. Adams & H. Wouters
each introducing slightly different concentrations of some trace elements into the glass objects being made,
next to a different amount of CaO.
Oxides Group I (n=45) Group II (n=9) Group III (n=5)
Na2O 16.48 ±0.40 17.20 ±0.35 16.28 ± 0.60
MgO 0.23 ±0.13 0.01 ± 0.01 0.07 ±0.13
2.51 ±0.07 2.35 ± 0.34 2.42 ±0.10 A12O3
69.46 ± 0.62 71.69 ±0.42 70.92 ± 1.62 SiO2
0.08 ± 0.04 0.02 ± 0.04 0.00 ± 0.00 p2o5
0.16 ±0.11 0.17 ±0.07 0.20 ± 0.07 SO3
Cl 0.82 ± 0.07 1.16 ±0.05 1.06 ±0.05
K2O 0.84 ± 0.06 0.58 ±0.12 0.61 ±0.11
CaO 8.41 ±0.55 5.52 ±0.61 7.54 ± 0.42
0.04 ± 0.04 0.05 ± 0.04 0.02 ± 0.02 TiO2
MnO 0.43 ± 0.06 0.84 ±0.13 0.09 ± 0.09
0.52 ± 0.06 0.33 ± 0.04 0.39 ±0.13 Fe2O3
12 ± 12 23 ±27 9±8 Cr2O3
NiO 7±5 15 ±7 8±8
CuO 209 ± 95 50 ±43 83 ± 137
ZnO 35 ±19 26 ±7 18 ±13
Br 7±11 8±7 7±4
Rb2O 14 ±4 12 ±3 13 ±3
SrO 637 ± 145 534 ± 130 570 ± 137
9±2 6±1 8±3 Y2O3
79 ±17 63 ±16 59 ±15 ZrO2
3±2 3±1 0±0 Mo2O3
56 ±34 42 ±21 113 ±49 SnO2
354 ± 190 29 ±31 1±1 Sb2O5
BaO 234 ±127 151 ±59 151 ±92
PbO 156 ± 64 17 ± 14 13 ± 9
Table 2 - Average concentration and standard deviation for oxide concentration of major, minor and
trace elements in the glass fragments belonging to the three groups ofQumrân objects.
Concentrations above the horizontal line are given in w %, below it in ppm.
It was assumed that during this period a stable glass was made from calcareous sand and as a flux,
natron. The two types of sand suitable for glass-making which are specifically mentioned by location in
classical writings are those at the mouth of the river Belus on the Syrian coast and the seashore deposit
[Pliny (AD 23-79)] near the mouth of the river Volturnus, north-west of Naples and the ancient harbour of
Pozzuoli. 15 Turner found that this sand contained about 9 % of lime 16 and other analyses have given
similar lime contents. 17 Therefore, an explanation of the difference in CaO content in the subgroups in the
Qumrân samples could be the use of different sands, as was also suggested by Nenna et al. 18 Next to this
lime also 3.6-5.3 % of alumina and about 1.5 % of magnesium carbonate is present. 19 When mixed with
alkali it is possible to fabricate a durable glass.
15. H.N. XXXVI, 65-66.
16. Turner 1956.
17. Engle 1973, p. 1-26 and ead. 1978, p. 10.
18. Nenna, Vichy and Picon, 1997.
19. Turner 1956. Analysis of the Composition of Glass Objects from Qumrân, Israel 1 1 7
Considering the low concentrations of P2O5, K2O and MgO and the presence of Cl in the glass, it was
assumed that as alkali source material, natron, a natural sodium sesquicarbonate (Na2CO3.NaHCO3.2H2O)
from the Wadi Natrûn, north-west of Cairo, was used in the glass-making activities. The composition of
natron is complex and variable: the sodium carbonate content varies from 22.4 to 75.0 per cent, sodium
bicarbonate from 5.0 to 32.4 per cent, sodium chloride from 2.2 to 26.8 per cent, sodium sulphate from
2.3 to 29.9 per cent plus water and insoluble material. It had already been employed from very early times
as detergent, in medicine and for embalming. 20
The colour of the glass produced would not have been up to the standards of modern colourless glass
because, although the iron oxide content was low in comparison with other sands that Turner examined, it
was high in comparison with that of present day glass. 2l The concentration of iron and manganese in the
glass is approximately the same. Iron in the Fe(II) state results in a blue-green colour in glass which can
be partially decolourised by adding manganese due to the following equilibrium reaction.
Fe2+ + Mn3+ <-> Fe3+ + Mn2+
When conditions during melting are such that the purple from the manganese just compensates for
the colour of the iron, a colourless glass results. 22 Object #21b on the other hand shows the typical purple
colour due to a high manganese concentration (3.9 % MnO). When correlations between elements were
considered, copper was found to be positively correlated with both zinc (Fig. 3) and lead, meaning that
copper was probably added to the glass melt as a Cu-based alloy which was softened by lead. Antimony
also showed a positive correlation with lead while manganese and iron were positively correlated up to a
manganese concentration of 0.6 %, suggesting that above this concentration manganese was added
intentionally as a decolouriser or colouring agent with high concentration.
Regarding the question of the provenance of the objects, the above-discussed compositional
information suggests that almost all objects found at Qumrân were made from the same batch of glass.
Either they were made locally in a glass workshop (e.g., by remelting and working of a large stock of bulk
glass) or they were imported ready-made in large quantity from elsewhere. In either case, at Qumrân, there
seems to have been a relatively large demand for glass vessels of various types such as goblets and
ointment pots. This is unusual as glass objects in ancient times were hard to make and thus fairly rare. This
tentative conclusion appears to be consistent with the hypothesis that at the Qumrân site some industrial
activity in the field of ointment/perfume manufacture was located during ancient times, the glass vials and
bottles being used as receptacles for perfume, ointment, etc., as proposed by P. Donceel- Voûte. 23
COMPARISON WITH OTHER ROMAN GLASS OBJECTS
In general, glass produced in early times and up to the Middle Ages was obtained by melting together
a siliceous constituent (e.g., quartz sand), a fluxing agent (e.g., natural soda or sodium-rich plant ashes)
and one or more colourant, decolourant and/or opacifying substance(s). It is logical to assume that the use
and type of these constituents varied, depending on the manufacturing centres and the historical period. 24
Roman vessel glass from the period lst-6th century AD, however, shows the remarkable property that its
major composition is very similar irrespective of the geographical location and time period the material
originates from; even post-Roman glass (9th century) maintains this composition for the most part. The
20. Frank 1982; Newton and Davison 1991.
21. Turner 1956.
22. Newton and Davison 1991.
23. Donceel- Voûte 1994.
24. Velde 1990. 118 A. Aerts, Κ. Janssens, Β. Velde, F. Adams & H. Wouters
reasons for this high degree of compositional similarity are unclear and are at present subject to speculation
in the literature. The consistency in composition suggests a certain discipline in the manufacturing
technique drawing on a constant supply of raw materials of high purity for sodium, silica and aluminium
which were used in quite proportions for a long period of time over a large geographical area. 25
Either all the bulk glass was manufactured (in the form of ingots) in one or a few locations in the
Middle East or all the glass was made with raw materials originating from the same source. A Middle-
Eastern source seems to be the most logical explanation of the pure soda component. 26 It was assumed
that Roman galleys loaded the natron in Alexandria and brought it to the head of the Adriatic from where
it was further transported up the Rhine. Since natron-type glass continued to be made long after the
breakdown of the Roman imperial power in the west, it seems that traders and merchants continued to
import natron into the north-west of Europe using these or other trade routes (via the Black Sea, up the
Danube, the Vistula and via the Baltic sea), even after the fall of the Roman Empire. 27 The hypothesis that
this material was already fused into glass near the site of extraction and that this raw glass was then
transported in the form of ingots, frit or cullet to the site where production of objects took place, is also
very plausible since it eliminates the necessity of finding the pure sources of CaO, SiO2 etc. at various
places, situated in various climatic zones throughout Europe. Furthermore, an insoluble, highly stable
material such as glass ingots is much easier to handle during transportation than to treat natron which is
corrosive when wet. A third hypothesis is that for the manufacture of glass during the later period of the
Roman Empire, glass from earlier periods was extensively recycled. 28
In order to obtain more detailed information of the differences and similarities of Roman glass from
different periods and geographical locations, an extensive set of Roman glass samples (about 250 samples)
was analysed in addition to the Qumrân series for the major, minor and trace composition. The glass
objects originated from Cologne and Trier (Germany), Rouen (France), Oudenburg, Tongeren,
Grobbendonk and several sites in Wallonia (Belgium) and Maastricht (The Netherlands); they cover a time
period from the 1st to 6th century AD. The aim was to deduce whether the compositionally distinct groups
which were found in the Qumrân data set, could also be found in glass collections originating from other
locations and/or other time periods in the Roman Empire or whether they are typical only for glasses from
the Middle-East. The concentration found for the different elements in the various series were compared
in order to obtain more information about the raw materials employed which could lead to a better
interpretation of the subdivisions in the set of objects. Only general information will be given here, details
about the composition of the separate glass collections can be found in the study of A. Aerts. 29 The results
were also compared with the results from Jackson obtained by inductively coupled plasma atomic emission
spectrometry (ICP-AES) on glass waste from two Roman sites at Mancetter and Leicester in Great-
Britain. 30
When the chemical compositions of all the glasses of the fourth century or older were compared, it
became clear that they are indeed very similar, especially in major elements, containing approximately 66-
72 % SiO2, 16-18 % Na2O and 6-8 % CaO; sometimes only slight differences could be seen which were
not sufficient to differentiate between various locations or age. All the glasses showed the typical low
magnesium/low potassium soda-lime-silica composition. When minor and/or trace elements were
considered, differences, mainly related to the colour of the glass, could be seen between the various
objects. Comparable concentrations of trace elements such as copper, cobalt, tin, antimony and lead, at
different levels, could be found in all the groups of glass objects, including the glasses from Qumrân. Yet
some subtle differences were present such as a high concentration of zirconium in the glasses from
Mancetter and Leicester.
25. Velde 1990; Velde and Sennequier 1985.
26. Henderson 1985.
27. Newton 1980.
28. Velde and Gendron 1980; Velde 1990.
29. Aerts 1998.
30. C. Jackson, personnal communication. Analysis of the Composition of Glass Objects from Qumrân, Israel 119
One of the elements covering the largest range of concentration levels was definitely antimony, which
in ancient times was not only used as decolouriser but also as opacifying agent and as means to remove
air-bubbles from the glass melt. Its use as decolourant seemed to have been largely taken over by
manganese during later centuries except for the most western part of the Roman territory (Mancetter and
Leicester in Great-Britain and Oudenburg in Belgium) where it was present in relatively high
concentrations in almost all glasses, even the ones dated to the third century. The raw materials used for
antimony were on the one hand stibina, a Sb-S mineral, and on the other hand a Sb-Pb mineral or alloy, of
which the latter seemed to have been used more often. Evidence for the use of stibina was only found in
some of the glasses from Maastricht and in the ones from Trier, Rouen and Oudenburg.
The most common colouring elements were iron, manganese, copper and cobalt. Iron was present in
all the glasses and had probably entered the glass batch as an impurity present in the raw materials used.
Next to being used as a decolouriser, an excess of manganese results in a purple colour. The use of copper
can lead to a red, blue or green glass, depending on the atmosphere in the glass furnace. Copper can be
added to the glass melt as either bronze or brass and it seemed that both were used during the different
centuries and all across the Roman Empire since positive correlations were found between CuO and ZnO
for some glasses and between CuO and SnO2 for others. Colouring with cobalt (even from 50 ppm on)
results in a brightly blue glass. This element could have been added to the glass batch as either a cobalt ore
or a blue glass cullet rich in cobalt.
Somewhere during the second half of the fourth century a drastic change in glass production appears
to have occurred; this is reflected in stylistic characterisations, colour as well as composition of the glass
vessels. The number of forms is strongly reduced to only common tableware, most of the glass shows a
dark green or even brown colour and the concentration of iron, manganese, titanium and some trace
elements such as copper, zinc and zirconium becomes significantly higher. The reason for this change in
composition could be explained by recycling of older glass introducing impurities to the glass during this
process and/or the use of less pure or unpurified raw materials.
CONCLUSIONS
Since the major composition of all these glass objects is so similar, it is likely that probably only a
few workshops produced glass from raw materials, which was then transported as ingots throughout the
Roman Empire. At different sites these were remelted, colourants, decolourants and/or opacifiers were
added, introducing amounts of the various trace elements, and objects were formed. During later
centuries older glass seems to have been recycled. Probably, further information can be gained if more
glass samples are measured, compared and especially when correlations between trace elements are studied
in detail providing indications about raw materials used. This might lead to a better understanding of trade
routes and/or techniques in glass production during the Roman period. According to our results, it can also
be stated that the glass-makers during that time appear to have had a quite good about how
to produce a durable glass and which elements or compounds to use as colourants, decolourants and/or
opacifiers.
Acknowlegdements
The authors would like to thank Mrs C. Fontaine- Hodiamont from the Royal Institute for Culture
Heritage, Brussels, Belgium for the typological classification of the Qumrân objects, Prof. R. Rottländer from
the University of Tübingen, Germany, Mr A. Verhoeven from the Institute for the Archaeological
Patrimonium, Flanders, Belgium and Mr W. Dijkman from the Department of Town Development and
Ground Matters, Maastricht, The Netherlands for providing us samples and Dr C. Jackson from the
University of Sheffield, Great-Britain for information on ICP-AES results. 120 A. Aerts, Κ. Janssens, Β. Velde, F. Adams & H. Wouters
{ Callirhoé
Illustration non autorisée à la diffusion
Fig. 2 - A few Roman glass objects after Fig. i - Location of Qumrân at the north-western shore restoration found at Qumrân, Israel. of the Dead Sea. Copyright IRPA/KIK.
r
140 ■■
120 ■■
100 ·■
80 ■■
60 ■■
20
t I » I
0 100 200 300 400 500 600 700 800 900 1000
CuO (ppm)
Fig. 3 - Positive correlation between the concentration of CuO and ZnO in the glass objects from Qumrân. Analysis of the Composition of Glass Objects from Qumrân, Israel 121
REFERENCES
Aerts Α., 1998, Microscopic Analysis of Roman Vessel Jackson CM., Hunter J.R., Warren S.E. and
Glass, PhD dissertation, University of Antwerp, Cool H.E.M., 1991, "The Analysis of Blue-green
Antwerp, Belgium. Glass and Glassy Waste from Two Romano-British
Working Sites", Archaeometry' 90, Birkhauser Aerts Α., Janssens K., Adams F. and Wolters Η., Verlag, Basel, p. 295-304. submitted, "Microanalysis of Roman Glass from
Khirbet Qumrân", Journal of Archaeological Laperrousaz E.M., 1961, Les Manuscrits de la Mer
Science, Proceedings of the Illinois Urbana Morte, PUF, Paris.
Archaeometry Conference.
Milik J.T., 1972, "Qumrân", Encyclopaedia Uni- Baillet M, Milik J.T. and Vaux R. de, 1962, versalis, 13, p. 896-898. Discoveries in the Judean Desert, volume 3,
Clarendon Press, Oxford. Mirti P., Casoli A. and Appolonia L., 1993, "Scientific
Analysis of Roman Glass from Augusta Praetoria", Brill R.H., 1988, "Scientific Investigations of the
Archaeometry 25, 1993, p. 225-240. Jalame Glass and Related Finds", in Weinberg G.
(ed.), Excavations at Jalame: Site of a Glass Nenna M.-D., Vichy M. and Picon M., 1997, "L'atelier Factory in Late Roman Palestine, University of de verrier de Lyon, du Ier siècle ap. J.-C, et l'origine Missouri, Columbia, p. 257-294. des verres 'romains'", Revue d'Archéometrie 21,
Calvi M.C, Tornati M. and Scandellari M.L., 1968, 1997, p. 81-87.
"Ricerche Technologiche", in Calvi M.C., / Vetri
Newton R.G., 1980, "Recent Views on Ancient Romani del Museo de Aquileia, Pubblicazioni
Glasses", Glass Technology 21(4), 1980, p. 173-183. dell' Associazione nationale per Aquileia 7.
Cox G.A. and Ford B.A., 1993, "The Long-term Newton R. and Davison S., 1991, Conservation of
Corrosion of Glass by Ground-water", J. Mat. Glass, Butterworths, London, Chapter 3.
Sei. 28, 1993, p. 5637-5647. Turner W.E.S., 1956, "Studies in Ancient Glasses and
Donceel- Voûte P., 1994, "Les mines de Qumrân Glass-making Processes. Part V. Raw Materials and
réinterprétées", Archéologia 298, 1994, p. 24-35. Melting Processes", Transactions of the Society of
Donceel R. and Donceel- Voûte P., 1994, "Methods of Glass Technology 40, p. 277-300.
Investigation of the Dead Sea Scrolls and the Van Rolf C and Rottländer Α., 1990, "NaturKhirbet Qumrân Site. Present Realities and Future wissenschaftliche Untersuchungen zum Römischen Prospects. The Archaeology of Khirbet Qumrân", Glas in Köln", Kölner Jahrbuch für Vor- und Annals of the New York Academy of Sciences 722, Frühgeschichte 23, 1990, p. 563-582. 1994, p. 1-38.
Vaux R. de, 1973, "Archaeology and the Dead Sea Engle Α., 1973, Readings in Glass History 1, Phoenix
Scrolls", The Schweich Lectures of the British Publications, Jerusalem.
Academy 41b. Engle Α., 1978, Readings in Glass History 10, Phoenix
Publications, Jerusalem. Velde B., 1990, "Alumina and Calcium Oxide
Content of Glass Found in Western and Northern Fontaine-Hodiamont C, 1993, "Quatre-ving-neuf
Europe", Oxford Journal of Archaeology 9, 1990, verres très fragmentaires provenant du site de
p. 105-117. Khirbet Qumrân (Cisjordanie, Ier s. ap. J.-C.)",
Bulletin de l'IRPA/van het KIK (Selectie/Sélection) Velde B. and Gendron C, 1980, "Chemical 25, 1993, p. 277-280. Composition of Some Gallo-Roman Glass
Frank S., 1982, Glass and Archaeology, Academic Fragments from Central Western France",
Press, London, Chapter 4. Archaeometry 22, 1980, p. 183-197.
Fritsch CT., 1956, The Qumrân Community: Its Velde B. and Sennequier G., 1985, "Observations on History and Scrolls, MacMillan, New York. the Chemical Compositions of Several Types of
Henderson J., 1985, "The Raw Materials of Gallo-Roman and Frankish Glass Production",
Early Glass Production", Oxford Journal of Annales AIHV 9 (Nancy 1983), AIHV, Liège,
Archaeology 4, 1985, p. 267-291. p. 127-147.