Nederlandse
Archeologische
Rapporten
081
Early medieval
glass production
in the Netherlands
A chemical and isotopic investigation
Hongjiao Ma, Julian Henderson,
Yvette Sablerolles, Simon Chenery
and Jane Evans
Early medieval
glass production
in the Netherlands
a chemical and isotopic investigation
Hongjiao Ma, Julian Henderson,
Yvette Sablerolles, Simon Chenery
and Jane Evans
Colophon
Nederlandse Archeologische Rapporten 81
Early medieval glass production in the Netherlands: a chemical and isotopic investigation
Authors: Hongjiao Ma, Julian Henderson, Yvette Sablerolles (University of Nottingham), Simon Chenery and Jane Evans (British
Geological Survey)
With a contribution from Menno Dijkstra (University of Amsterdam)
Authorisation: Hans Huisman and Rik Feiken (Cultural Heritage Agency of the Netherlands)
Illustrations (unless otherwise stated): Julian Henderson, Yvette Sablerolles and Marjolein Haars (BCL-Archaeological Support)
Cover design: J. Ranzijn (Happyfolio)
Cover: A selection of tesserae from the Wierum terp (province of Groningen) (Photograph: Jelle Schokker of the Noordelijk
Archeologisch Depot).
Design and layout: Xerox/Osage
ISBN/EAN: 978-90-76046-85-3
© Cultural Heritage Agency of the Netherlands, Amersfoort, 2023
Cultural Heritage Agency of the Netherlands
P.O. Box 1600
3800 BP Amersfoort
the Netherlands
www.cultureelerfgoed.nl
3
—
Contents
Summary
5
Samenvatting
7
1
1.1
9
1.2
1.3
1.4
1.5
2
2.1
2.2
2.3
2.4
2.4.1
2.4.2
3
3.1
3.2
3.2.1
3.2.2
3.2.3
3.3
3.4
3.4.1
3.4.2
3.5
3.5.1
3.5.2
3.6
3.6.1
3.6.2
3.7
3.8
3.9
3.10
3.11
3.12
Introduction
The research project ‘Early medieval glass
production’
Research questions
Research approach
Structure of the monograph
Acknowledgements
Technology and raw materials used for
Dutch early medieval glass
Introduction
Furnaces, raw materials and glass sources
Evidence for the glass industry in early medieval
Europe outside the Netherlands
Scientific analysis of early medieval glass
in Europe
The principal glass types
Summary of existing scientific analyses of
early medieval Dutch glass before the start
of this project
Evidence for early medieval glass-working
in the Netherlands
Introduction
Maastricht, Limburg Province
Maastricht-Jodenstraat (MAJO)
Maastricht-Mabro
Maastricht-Rijksarchief
Susteren-Salvatorplein, Limburg Province
Wijk bij Duurstede (Dorestad), Utrecht Province
Wijk bij Duurstede – Parkeerplaats Albert Heijn
(PPAH)
Wijk bij Duurstede – Veilingterrein and
Frankenweg/Zandweg
Utrecht, Utrecht Province
Utrecht-Domplein
Utrecht–Oudwijkerdwarsstraat
Leidsche Rijn, Utrecht Province
Leidsche Rijn-LR 51/54
Leidsche Rijn-Leeuwesteyn Noord
Oegstgeest–Nieuw Rhijngeest Zuid (Rijnfront),
Zuid-Holland Province
Rijnsburg-Abdijterrein,
Zuid-Holland Province
Valkenburg-De Woerd, Zuid-Holland Province
Den Haag-Frankenslag, Zuid-Holland Province
Bloemendaal-Groot-Olmen, Noord-Holland
Province
Wijnaldum-Tjitsma, Friesland Province
9
9
10
10
10
3.13
3.14
Wierum, Groningen Province
Deventer-Stadhuiskwartier, Overijssel Province
4
The materials, analytical techniques and
methodology
Introduction
An overview of the sites and glass samples
Electron probe microanalysis (EPMA) for major
and minor chemical composition
Laser ablation inductively coupled plasma mass
spectrometry (LA-ICP-MS) analysis for
trace element compositions
Thermal ionization mass spectrometry (TIMS)
analysis to determine Nd and Sr isotopic
compositions
How analytical data is used in this study
4.1
4.2
4.3
4.4
4.5
13
13
13
16
17
17
19
29
29
30
31
35
36
37
39
40
41
43
43
46
47
47
48
48
49
51
52
53
53
4.6
5
5.1
5.2
The analytical results and discussion
Introduction
Glass samples from Maastricht (Jodenstraat
and Mabro sites)
5.2.1 Naturally coloured and cobalt blue
bead-making glass waste
5.2.2 Highly coloured opaque glass
5.2.3 Vitreous and semi-vitreous materials attached
to the crucibles
5.2.4 Glass artefacts
5.2.5 Crucibles from the Mabro site, Maastricht
5.3
Glass samples from Gennep
5.4
Glass samples from Wijnaldum
5.4.1 Highly coloured opaque glass beads
5.4.2 Colourless glass beads
5.4.3 Vessel glass
5.4.4 Bead production materials
5.5
Glass samples from Utrecht
5.6
Glass samples from Wijk bij Duurstede
(Dorestad)
5.6.1 Vessel glass
5.6.2 Other glass
5.7
Glass samples from Susteren
5.7.1 Trail decorated glass beads
5.7.2 Window glass
5.7.3 Glass attached to crucibles
5.7.4 Vessel glass
5.8
Glass samples from Deventer
5.8.1 Wood ash glass
5.8.2 Natron glass
5.8.3 Mixed alkali glass
5.8.4 Plant ash glass
5.9
Nd-Sr isotope analysis
5.10
Discussion
5.10.1 The base glass used for bead making
at Jodenstraat
55
56
59
59
59
61
61
62
64
67
67
67
67
68
70
72
72
74
74
74
75
77
78
78
79
79
80
81
81
82
82
82
83
83
84
85
85
86
89
89
4
—
5.10.2 The use of crucibles in on-site lead tin yellow
colourant production in early medieval
northwestern Europe
5.10.3 The separate production of a tin-based white
opacifier at Maastricht, Jodenstraat
5.10.4 The other chemical characteristics of the glass
and its archaeological implications
5.10.5 A comparison of 7th–11th century vessel glass
from Comacchio with early medieval Dutch
glass and the suggested supply of raw glass in
the two areas
5.10.6 Wood ash glass and mixed alkali glass
5.10.7 Plant ash glass
5.11
Summary
6
6.1
6.2
Synthesis and conclusions
The Early Merovingian period
(450-550 AD)
The Middle Merovingian period (550– 650 AD)
89
6.3
6.4
6.5
The Late Merovingian period (650 – 750 AD)
The Carolingian period (750 – c. 850 AD)
The late phase, including the Ottonian period
(c. 850 – c. 1000 AD)
101
102
104
Answering the research questions
107
91
7
92
95
96
97
98
99
99
100
Bibliography
115
Appendices
Appendix I sample list
Appendix II major and minor chemical compositions
of samples analysed by electron probe
microanalysis
Appendix III trace element chemical compositions
of samples analysed by LA-ICP-MS
Appendix IV photos of the samples from Maastricht
and Utrecht
125
126
154
170
183
5
—
Summary
This monograph brings together for the first
time comprehensive combined archaeological,
technological and scientific investigations, using
chemical (major, minor and trace element
concentrations) and isotopic (87Sr/86Sr and
143Nd/144Nd) analysis, of early medieval glass
production in the Netherlands. We selected 276
samples of glass from Gennep, Maastricht
(the Jodenstraat and Mabro sites), WijnaldumTjitsma, Utecht (the Domplein and Oudwijkerdwarsstraat sites), Susteren-Salvatorplein,
Wijk bij Duurstede (the Hoogstraat and vicus
sites) and Deventer-Stadhuiskwartier, dating to
between the late 4th and 11th centuries covering
the Merovingian and Carolingian periods for
compositional analysis. In addition, 20 samples
were subjected to isotope analysis. The results
of our trace element and isotopic analyses have
provided new and highly significant insights into
early medieval glass production in the Netherlands.
Several different compositional types of
glass have been identified. Both high and low
lead glasses have been found along with pristine
and recycled (sub-types of) natron glass (HIMT,
Foy 2, Egypt II, Levantine II), plant ash glass,
wood ash glass and mixed-alkali glass. The best
evidence in early medieval Europe for the onsite production of the lead-tin yellow and tin
oxide colorant/opacifier in crucibles excavated
from the 6th-7th century AD Jodenstraat site in
Maastricht is discussed in detail. It is associated
with comprehensive evidence for the manufacture
of brightly coloured monochrome glass beads
(also found at Wijnaldum), a craft specialisation.
The base glass for the beads was imported
‘pristine’ Foy 2 glass. Our results show that a
higher proportion of Merovingian glasses were
imported ‘pristine’ (Egyptian) glasses than the
glasses used in the Carolingian period, when the
majority of glasses were recycled, potentially
multiple times. While it is often suggested that
elevated levels of antimony and lead started to
occur in the Carolingian period we have found
elevated levels already in Merovingian glasses,
an indication that a small proportion of Roman
tesserae and/or coloured Roman vessel glass
was being added to the glass melt then.
By combining trace element and isotope analysis
we have been able to demonstrate that
Carolingian recycled glass contained a small
proportion of wood ash glass. Wood ash glass
with elevated concentrations of Cs, Rb, Ba and Sr
started to be manufactured in Europe from
around 800 AD and added to natron glass to
produce mixed alkali glass. Even though mostly
recycled glass was in use by this time much of
the Dutch natron glass can still be attributed an
ultimate source in Egypt (recycled Foy 2 glass),
with limited evidence for the use of ‘pristine’
glass. Our analyses provides evidence that
recycled Foy 2 was still in use as late as the mid
10th century AD.
When the results of our analyses for
Carolingian natron glasses are compared with
contemporary (7th-11th century AD) northern
Italian glasses from the site of Comacchio and
Spanish glasses from Tolmo de Minateda an
interesting contrast is revealed. Both sets of
glasses are largely recycled: whereas Dutch
recycled natron glass has an ultimate source in
Egypt, the recycled natron glasses from
Comacchio and Tolmo de Minateda show far
more evidence for the use of imported Levantine
glasses instead. Only a single example of
Levantine (II) glass has been found amongst
Dutch early medieval glass with no evidence that
such glass formed part of the recycling process.
This is a clear reflection of differing trade
contacts and glass supplies between northern
and southern Europe. By the 9th-11th centuries AD
the widest range of glass types was in circulation
and, perhaps surprisingly, glass from Deventer
includes pristine glass. The glass from Deventer
consists of thirteen natron glasses (three Roman,
three pristine Egyptian II, six recycled Foy 2,
one pristine Foy 2), one plant ash glass, four
mixed alkali glasses and nineteen wood ash
glasses. Like plant ash glass production in
western Asia, the primary production of wood
ash glass would have formed a decentralised
network because wood ash was also widely
available.
Although glass beads were certainly made
in the Netherlands along the Meuse valley in the
Merovingian period, probably with a ‘permanent’
workshop in Maastricht (even if the bead makers
took part in other industries) we suggest that
bead workers were mobile further north, visiting
Rijnsburg, Wijnaldum and perhaps ValkenburgDe Woerd. A very likely source of vessel glass
probably existed in Cologne. Scientific analysis
of Merovingian bowls and beakers (450-550 AD)
shows a correlation between vessel type,
chemical composition and colour, suggesting
that glass of specific colours were selected
during vessel manufacture. This may simply
6
—
have been part of batch production but we
suggest that colour selection was related to a
memory of the exotic origin of the glass and
drinking rituals, including the colour of liquid the
vessels contained.
We have shown that the key types of raw
glass were from Egypt- Foy 2 and HIMT in the
Merovingian period and Egypt II in the Carolingian
period; a single Levantine II (punty) glass is the
only example of Levantine glass we have found;
mixed-alkali glass probably derived from
northern France; wood ash glass perhaps from
Belgium, northern France or more likely from
Germany (perhaps using the Viking trade
network); plant ash glass was imported from
western Asia as beads and raw glass (also perhaps
by the Vikings); the raw glass was made into
characteristic early medieval vessel types and
incorporated into glass beads.
7
—
Samenvatting
Deze monografie combineert voor het eerst
uitgebreid archeologisch, technisch en natuurwetenschappelijk onderzoek, met chemische
(hoofd- en sporenelementen) en isotopenanalyses (87Sr/86Sr and 143Nd/144Nd), van
vroegmiddeleeuwse glasproductie in Nederland.
We selecteerden 290 glasmonsters uit Gennep,
Maastricht (Jodenstraat en Mabro), WijnaldumTjitsma, Utrecht (Domplein en Oudwijkerdwarsstraat), Susteren-Salvatorplein, Wijk bij
Duurstede (Hoogstraat en vicus) en DeventerStadhuiskwartier, die met dateringen tussen de
late vierde en de elfde eeuw de Merovingische
en Karolingische perioden beslaan. Daarnaast
werden twintig van deze monsters geselecteerd
voor isotopenanalyse. De resultaten van de
sporenelement- en isotopenanalyses geven
belangrijke nieuwe inzichten in de vroegmiddeleeuwse glasproductie in Nederland.
Er zijn verschillende typen glas met verschillende
samenstelling geïdentificeerd. Zowel glas met
hoge als met lage gehaltes aan lood zijn
aangetroffen, naast vers en gerecycled natron
glas (sub-typen HIMT, Foy 2, Egypte II, Levantine
II), glas op basis van de as van planten of hout
en gemengd alkali glas. Het meest overtuigende
bewijs voor de lokale productie van lood-tin geel
en tin oxide als kleurstof en opacifier in vroegmiddeleeuws Europa - smeltkroesjes uit de
zesde-zevende eeuw n.Chr. opgegraven in
Maastricht Jodenstraat - wordt in detail besproken.
Tijdens de opgraving is er uitputtend bewijs
gevonden voor het maken van helder gekleurde
monochrome kralen, een gespecialiseerd
ambacht. Het basisglas voor de kralen was
geïmporteerd “vers” glas van type Foy 2.
Onze resultaten laten zien dat een hoger
percentage van Merovingisch glas bestond uit
geïmporteerd “vers” (Egyptisch) glas, vergeleken
met glas uit de Karolingisch periode, toen het
meeste glas werd gerecycled – één of meerdere
keren. Hoewel vaak wordt gesuggereerd dat
verhoogde gehaltes aan antimoon en lood pas
beginnen in de Karolingische tijd, doordat kleine
hoeveelheden Romeinse tesserae of glas van
Romeins glazen vaatwerk werden toegevoegd
aan gesmolten glas, vonden we al verhoogde
gehaltes van antimoon en lood in Merovingisch
glas. Met een combinatie van sporenelement- en
isotopenanalyses hebben we kunnen aantonen
dat Karolingische gerecycled glas een klein
aandeel hout-as glas bevat. Hout-as glas met
verhoogde concentraties van cesium (Cs),
rubidium (Rb), barium (Ba) en strontium (Sr)
werd voor het eerst geproduceerd in Europe
vanaf ongeveer 800 n.Chr., en het werd
toegevoegd aan natronglas om gemengd
alkaliglas te maken. Hoewel in deze periode
vooral gerecycled glas in gebruik was, kan veel
van het Nederlandse natronglas nog steeds
worden gelinkt aan een oorspronkelijk herkomst
in Egypte (gerecycled Foy 2 glas), met beperkte
aanwijzingen voor het gebruik van “vers” glas.
Onze analyses laten zien dat gerecycled Foy 2
glas zelfs nog in gebruik was in het midden van
de tiende eeuw n.Chr.
Als de resultaten van onze analyses van
Karolingisch natron glas worden vergeleken met
contemporain (zevende-elfde eeuws) NoordItaliaans glas uit Comacchio en Spaans glas uit
Tolmo de Minateda komt een interessant
contrast aan het licht. De assemblages bestaan
vooral uit gerecycled glas, maar terwijl het
Nederlandse gerecycled glas in oorsprong uit
Egypte komt, bevat het glas uit Comacchio en
Tolmo de Minateda aanwijzingen voor een
Levantijnse oorsprong. Slechts één voorbeeld
van Levantijns (II) glas is aangetroffen onder
Nederlands vroegmiddeleeuws glas, en er zijn
geen aanwijzingen dat dit soort glas een rol
speelde bij recycling. Dit is een duidelijke
weerslag van verschillen in handelscontacten en
glasleveringen tussen Noord- en Zuid-Europa.
In de negende-elfde eeuw was de variatie in
glastypes die werden gebruikt het grootst,
inclusief – wellicht verrassend – “vers” glas uit
Deventer. Het glas uit Deventer bestaat uit
dertien stuks natron glas (drie keer Romeins,
drie keer “vers” Egyptisch II, zes keer gerecycled
Foy 2, een keer “vers” Foy 2), een keer plant-as,
vier keer gemengd alkali en negentien keer
hout-as glas. Net als plant-as glasproductie in
West-Azië vormde de productie van hout-as glas
een decentraal netwerk omdat hout as
algemeen beschikbaar was.
Hoewel Merovingische glazen kralen zeker
werden gemaakt in Nederland in de Maasvallei,
waarschijnlijk met een permanente werkplaats
in Maastricht (zelfs als de kralenmakers ook
andere ambachten uitoefenden), waren de
kralenmakers verder naar het noorden
waarschijnlijk mobiel. Daar bezochten ze Rijnsburg,
Wijnaldum en mogelijk ook Valkenburg – de
Woerd. Keulen was hoogstwaarschijnlijk een
belangrijke bron voor glazen kommen en bekers.
Natuurwetenschappelijke analyse van
8
—
Merovingische kommen en bekers (450-550 n.
Chr.) tonen een correlatie tussen typologie,
chemische samenstelling en kleur, wat suggereert
dat glas met specifieke kleuren werd geselecteerd
voor deze toepassing. Dit zou simpelweg
onderdeel kunnen zijn van grootschalige
productie, maar we suggereren dat de selectie
van de glaskleur te maken had met herinneringen
aan de exotische herkomst van het glas en
drankrituelen, inclusief de kleur van de vloeistof
die in de glazen had gezeten.
We hebben laten zien dat de belangrijkste types
ruw glas tijdens de Merovingische periode
Egypte – Foy en HIMT - waren, en tijdens de
Karolingische periode Egypte II. Het enige
voorbeeld van Levantijns glas is een Levantijns II
puntige glastype. Gemengd alkaliglas kwam
waarschijnlijk uit Noord-Frankrijk, hout-as glas
mogelijk uit België, Noord-Frankrijk of Duitsland
(wellicht via het Viking handelsnetwerk verkregen).
Plant-as glas kwam uit West-Azië als kralen en
ruw glas (ook wellicht via de Vikingen); het ruwe
glas werd verwerkt tot typische vroegmiddeleeuwse kommen en bekers, en tot
glazen kralen.
9
—
1 Introduction
1.1
The research project ‘Early medieval
glass production’
The research project ‘Early medieval glass
production’ was one in a series of studies
referred to as Pre-Malta research (‘pre-Malta
onderzoek’), and as such falls under the
programme Knowledge for Archaeology.
The programme aims to obtain datasets from
(not yet fully elaborated) excavation data
before the introduction of the Valetta
convention in 2007. With the use of currently
developed research methods and techniques,
the Pre-Malta research programme enables
the study of archaeological remains from old
excavations, which means that substantial new
knowledge can be obtained about the past.
This research project involved a series of
important goals related to the production
technology, provenance, glass supply, recycling,
trade and use of early medieval glass in the
Netherlands, building on, and expanding
significantly on, existing published research.
The main goals were:
• To carry out a full chemical and isotopic analysis
of all available glass samples using cutting edge
techniques.
• To establish the raw materials used to make the
transparent, translucent and opaque glasses
samples.
• To consider whether the glass has been
recycled.
• To attempt to suggest a source for the glass
(i.e. provenance).
• To investigate the change in glass raw materials
over time.
• To establish if there are any sub- and supraregional supply patterns for unrecycled early
medieval glass found on Dutch early medieval
sites from within Europe, the Levant, Iraq
and Iran.
• To investigate in more detail whether there are
chronological changes in the use of pure
imported as opposed to recycled glass
moving from the Merovingian into the
Carolingian period.
• To compare the glass compositions and technologies used in the manufacture of glass beads
and vessels and investigate if there is evidence
for the use of raw material specialization.
• To establish if there is any evidence for local
specialisation of glass bead production as
reflected in their chemical compositions.
• To investigate the evidence for the production
of lead-tin oxide opacified glass found in the
Netherlands given the large number of
crucibles containing a yellow substance that
have been discovered in early medieval Dutch
contexts especially in Maastricht and whether
there is evidence for primary glass making.
Scope of the research project
The project focuses on the simple, monochrome
beads, raw materials and production waste from
Merovingian and Carolingian contexts along
with contemporary vessel glasses.1 A small
number of glasses dating to c. 900-1000 AD are
included for comparison. The examination of
polychrome beads and glass vessels have been
included because they provide information to
(better) answer the research questions. The
starting point for this project’s research is
Henderson and Sablerolles’ research plan from
2020. This plan is included as an appendix in
the report ‘An Overview of Dutch Early Medieval
glassworking, published chemical and isotopic
analyses of glass beads and vessels, raw material
provenance of beads and vessels, changes in
raw material use over time and a plan for future
scientific analysis’, which served as preparation for
this project.2 The substantive information from
that report has been largely incorporated into
this report.
1.2
Research questions
The central questions of the study are:
1. What raw materials were used in the local
production of simple, monochrome Early
Medieval beads?
2. Where were these raw materials obtained
from?
Sub-questions here are:
i. What substances were used to make the
different colours of glass in the artefacts
tested?
ii. What compositional groups can be
distinguished in the glasses based on
chemical analyses?
1
2
Between AD 480 and 987.
Henderson & Sablerolles 2020.
10
—
iii. What does this tell us about dating of
primary glass production of these groups?
iv. What do the isotope ratios (Sr, Nd) obtained
from the glasses of selected compositional
types tell us about the their origin and dating?
v. What networks inside and outside the
Netherlands were used in obtaining glass,
including the colourants used?
vi. To what extent were the raw materials or
semi-finished products derived from primary
production, or to what extent from systematic
recycling of glass, including Roman?
The research also provides building blocks for
two NOaA questions3:
• What are the nature, manifestations, extent
and context of craft specialization? (NOaA 2.0
question 67)
• Where do non-local raw materials of utilitarian
objects come from? (NOaA 2.0 question 139)
1.3
3
https://noaa.cultureelerfgoed.nl.
Research approach
We have been able to carry out a comprehensive
scientific analysis of a wide range of early
medieval glass samples and production waste.
The scientific techniques we used were Scanning
Electron Microscopy, Electron Probe Microanalysis,
Laser Ablation Inductively Coupled Plasma Mass
Spectrometry and Thermal Ion Mass Spectrometry,
resulting in the largest database of chemical and
isotopic analyses for early medieval Dutch glass.
The Covid-19 virus prevented us from
travelling and taking new samples which made it
difficult to plan and led to delays in scientifically
analysing some samples. We had hoped to work
on new glass samples excavated from secure
archaeological contexts. Although possible for
glass from Maastricht, Utrecht, Gennep,
Wijnaldum, Susteren and Deventer, the context
information for Dorestad glasses studied here
were unavailable at the time of sampling in the
early 1990s. It was nevertheless possible to
provide dates for the Dorestad glass according
to vessel form.
This project has involved a collaborative
team of archaeologists, archaeological scientists
and a geologist: Hongjiao Ma, Julian Henderson,
Yvette Sablerolles, Simon Chenery, Jane Evans
and Menno Dijkstra linking archaeologists,
archaeological scientists and geologists.
1.4
Structure of the monograph
The remaining chapters of the monograph
develop in a logical sequence. Chapter 2
provides essential information about early
medieval European glass technologies, including
raw materials, evidence of European early
medieval glass production outside the Netherlands,
the types of glass found in early medieval Europe
followed by a review of published results for
early medieval glass from the Netherlands.
Chapter 3 discusses the existing archaeological
evidence for early medieval glass production on
the Netherlands.
In Chapter 4 the sites from which glass
samples used in this study are introduced briefly
followed by a description of the three main
analytical techniques used to investigate the
samples chemically and isotopically: electron
probe microanalysis (EPMA), laser ablationinductively coupled plasma- mass spectrometry
(LA-ICP-MS) and thermal ion mass spectrometry
(TIMS). Chapter 5 presents and discusses the
results of the chemical and isotopic analyses of
the glass and crucible samples.
Chapter 6 is a synthesis of archaeological
and scientific results according to chronological
periods (450-550, 550-650, 650-750, 750-850
and 850-1000 AD) for the work together with
conclusions. Chapter 7 provides succinct answers
to the research questions and sub-questions
listed in Section 1.2 above.
1.5
Acknowledgements
We are grateful to W. Dijkman, Senior
Conservator Archeologie en Erfgoed, Team
Programma en Innovative, Centre Céramique–
Kumulus – Natuurhistorisch Museum,
Maastricht for allowing us to sample glass and
glass working material from Maastricht, to H.
Wynia, the municipal archaeologist of Utrecht
for allowing us to sample glass from Utrecht, to
J. Schokker of the Noordelijk Archeologisch
Depot for allowing us to sample glass from
Wijnaldum, A. Peddemors then curator at the
Rijksmuseum van Oudheden, Leiden for
permission to sample glass from Dorestad, W.
van Es of the Rijksdienst voor het
11
—
Oudheidkundig Bodemonderzoek also for
permission to sample Dorestad glass and to E.
Mittendorff, project leader archaeology,
Deventer for sending us glass from Deventer so
that we could sample it and both the former
director of the Provincial Limburgs Museum in
Venlo, G. Jansen and H. Stoepker, for permission
to sample glass from Susteren. We are also
grateful to Dr. A. Kronz of Göttingen University
for sharing some unpublished data. We are very
grateful for the support of H. Huisman and
R. Feiken of the Cultural Heritage Agency of the
Netherlands for being extremely supportive and
for being very flexible throughout the project,
at the very difficult time of Covid. Finally we are
grateful to Dr Matthew Delvaux for permission
to reproduce Figure 3.7 and Bill Bolton for
producing publishable versions of all the Figures.
2 Technology and raw materials used
for Dutch early medieval glass
2.1
Introduction
This chapter provides the basic background
information about Dutch early medieval glass
technology including a discussion of the raw
materials used which can be suggested from
glass chemical compositions and the furnaces
used for making the likely sources of glass in
Europe and western Asia. The main primary raw
materials are an alkaline flux, a silica source and
a calcium source. Fluxes are provided by natron,
plant ash or wood ash; silica is normally provided
by sand; calcium is provided by shell fragments in
sand or calcium compounds in plant ash or wood
ash . Mineral rich colorants were added separately
to these glasses or opacifiers were developed
from them by heat treating the glasses.
The chapter covers glass production dating to
between the late Hellenistic period and the early
Islamic period as well as evidence for centralized
and decentralized production organisations.
It also discusses the compositional evidence for
the recycling of glass in the second half of the
1st millennium AD.
A separate sub-section (Section 2.3) is
devoted to a discussion of the evidence for
the early medieval glass industry outside the
Netherlands, especially in Belgium, Denmark
France, Germany, Hungary, Ireland, Italy and
the United Kingdom. The next sub-section
(Section 2.4.1) discusses the main compositional
types of glass in some cases associated with
primary glass making sites such as in SyroPalestine, Egypt, Syria, Iraq and northern
Europe. The main compositional types are
pristine Egyptian and Levantine natron glasses,
Roman glass, HIMT and its variations, plant ash
glass, mixed-alkali glass and wood ash glass.
Section 2.4.2 is a summary of the existing
published chemical analyses of early medieval
Dutch glass from Maastricht, Susteren,
Dorestad, Rijnsburg, Wijnaldum, Lent,
Borgharen and Sittard with interim
interpretations.
2.2
13
—
Furnaces, raw materials and
glass sources
Most early medieval glass found in the
Netherlands is what is known as soda-lime
(natron) glass. From about the 2nd century BC,
during the Late Hellenistic period, this kind of
glass was fused from raw materials in massive
rectangular tank furnaces. The earliest example
of a tank furnace yet discovered is in Beirut,
possibly dating to the 2nd century BC.4 Roman
glass tank furnaces have also been found,
in Egypt5 and Syro-Palestine, such as Jalame.6
The Levant continued to be the primary centre
for the production of raw furnace natron glass
on a massive scale into the Byzantine period,
especially in the 6th–8th centuries AD7 with a
probable dip in the scale of production in the
early Byzantine period.
The glass fused in these tank furnaces from
raw materials attached itself to the floor of the
furnace and, once it had cooled down, would
have been removed, perhaps with a pickaxe,
to produce chunks of raw furnace glass.8 These
chunks would then have been reheated in a
crucible within a second furnace type, perhaps of
a beehive shape.9 This would then have enabled
the glass-workers to work the glass into a range
of glass artefacts using metal implements such
as gathering rods to make beads, and hollow
tubes, known as blowing irons, for blowing glass
into vessels. This second stage of glass production
is known as secondary production. It could have
occurred on the same site as where the glass
was fused (primary glass production) or on other
sites at a distance from where the glass was
fused.10 Raw furnace glass manufactured at
primary glass making centres was sometimes
traded by boat. Excavations of shipwrecks have
revealed the extent of the trade, such as the
3rd century BC Sanguinaire found off the coast of
Corsica which had at least 550 kg of glass including
raw glass11, the 2nd-3rd century AD Mljet wreck off
the Croatian coast produced about 100 kg of raw
glass12 and the 2nd-3rd century AD Ouest Embiez 1
(Var) which produced between 350 and 700 kg of
raw glass chunks, each weighing up to 25 kg.13
Furthermore raw glass has been excavated from
the Golfe de Fos near the mouth of the Rhône14
and two metric tons of raw glass came from the
early 11th century wreck at Serçe Limani, Turkey.15
4
5
6
7
8
9
10
11
12
13
14
15
Kouwatli et al. 2008.
Nenna 2015, 19.
Phelps et al. 2016.
Gorin-Rosen 2000; Tal, Jackson-Tal &
Freestone 2004; Nenna 2015; Freestone
et al. 2000; Phelps et al. 2016.
Gorin-Rosen 2000.
Henderson 2000, 38–42.
Henderson 1989; Freestone et al. 2000;
Phelps et al. 2016, Henderson et al. 2021.
Alfonsi & Gandolfi 1997.
Rossi 2009.
Fontaine & Foy 2007.
Foy & Nenna 2001.
Bass 1984; Bass et al. 2009.
14
—
16
17
18
19
20
21
22
23
24
25
26
Henderson 1989; Freestone et al. 2000.
Shortland 2004; Henderson 2013, 51–53.
Henderson 2013, 51–52.
Henderson 2013, 52.
Nenna 2015.
Brems et al 2013a; 2013b.
Boschetti et al. 2016; Henderson, Sode &
Sablerolles 2019; Crocco et al. 2021.
Lahlil et al. 2010; Boschetti et al. 2016;
Boschetti et al. 2020.
Henderson 1991a; Barber, Freestone &
Moulding 2009.
Schibille & Freestone 2013; Boschetti et
al. 2016; Henderson, Sode & Sablerolles
2019.
Henderson, Sode & Sablerolles 2019;
Crocco et al. 2021.
The presence of geographically separated
primary and secondary glass-making sites has
led to a suggested decentralized model for
classical glass production in western Asia.16
The discovery of raw furnace glass on sites
where there is no evidence for primary glass
production either suggests that it was being
traded through the site or that it was worked
on the site. The existence of crucibles with a
layer of glass on the inside supports the latter
suggestion and there are examples of this from
early medieval contexts in the Netherlands
(see Chapter 3 for information about the
industrial evidence for glass production in the
Netherlands). There is no archaeological
evidence for the primary manufacture of
translucent or transparent natron glass in the
early medieval Netherlands.
Natron glass was manufactured from a
combination of sand and a mineral flux called
natron or natrun.17 The sand that occurs on the
coastal beaches of the Levant is ideal for glass
production and is referred to as such by both
Strabo and Pliny.18 The second primary raw
material was natron. The main source of this
evaporite mineral flux was in the Egyptian
western desert at Wadi el Natrun,19 close to
some primary production sites for Roman
glass.20 This mineral is an evaporite which is
formed seasonally and would have been
shipped or traded to glass makers on the
Levantine coast. A third crucial component of
natron glass, which gives it durability, is lime.
Lime was provided by the marine shells in
the sand. It appears that the proportion in the
Levantine coastal sand was just right for the
production of durable natron glass. It is possible
that the shell fraction was separated by glass
makers and mixed with sand in the correct
proportion prior to glass production, though
no archaeological evidence for this has been
found. The availability of sand with these
characteristics would have been one reason
why primary glass-making furnaces were located
on the Levantine coast. Both archaeological and
scientific evidence confirms that this is the case.
Strontium and neodymium isotope and
mineralogical analysis of multiple beach
deposits around the Mediterranean has
suggested which sands would have been
suitable for glass making.21
In spite of the existence of important
Byzantine glass-making sites in the Levant in the
6th–8th centuries a range of political, social and
economic factors would not necessarily provide
a guarantee that fresh natron glass would have
found its way to early medieval glass-working
sites in the Netherlands.
While it is widely accepted that some form
of natron glass (whether pristine or recycled)
was used for the manufacture of early medieval
objects in the Netherlands (to be discussed in
much more detail below and in Chapter 5) there
is one source of fully fused coloured glass that
was also used: glass tesserae.22 Evidence for the
reuse of these cubes of generally opaque glass
has been discussed in many archaeological and
scientific studies and this study is no exception.
Their discussion is relevant in this section
because they were coloured and opacified: such
colourants and opacifiers were also sometimes
used in Dutch early medieval glass.
Roman glass tesserae are invariably made
from natron glass. Most are opacified with small
crystals, especially of calcium antimonate
(Ca2Sb2O7 or Ca2Sb2O6). Without additional
colourants this produces an opaque white
colour.23 Opaque yellow tesserae are coloured by
lead antimonate crystals (Pb2Sb2O7) with a
smaller number of opaque yellow tesserae
coloured with lead stannate crystals (Pb2Sn2O7).
Dull red tesserae are opacified and coloured with
copper droplets also found in Roman enamels.24
Opaque turquoise blue tesserae are coloured
with copper and calcium antimonate, opaque
yellow-green tesserae with copper and lead
antimonate; opaque blue tesserae are coloured
by a combination of cobalt and calcium
antimonate. Therefore if elevated levels of
antimony, lead, copper and sometimes tin are
found in translucent early medieval glass a likely
source is recycled Roman glass tesserae.25
Alternatively, such elevated levels of colourants
can be explained by the use of fragments of
highly coloured vessel glass. The most significant
collection of glass tesserae in the Netherlands
has been discovered at Wierum.26
Dutch early medieval glass was also
coloured deliberately with low levels of
transition metal ions: cobalt to produce a deep
translucent blue colour, copper for a turquoise
colour and manganese for a purple colour.
Various shades of green, amber and pale blue
could be produced by modifying the furnace
atmosphere in which the glass was melted if the
glass contained iron and manganese. Amber and
15
—
pale blue colours are produced in an oxygendeficient furnace atmosphere, green in a more
oxidising atmosphere.
There is evidence for the production of one
particular colour of glass in northwestern Europe
before c. 800 AD and as early as the 6th century:
opaque yellow. Opaque yellow vitreous materials
have been found in crucibles from several early
medieval sites in Ireland, Denmark and the
Netherlands. The evidence for its production and
its scientific analysis will be discussed in more
detail in Chapter 5.
From around 800 AD, glass technology in
western Asia underwent a technological
transition, especially during the Abbasid
caliphate. The Abbasid glassmakers had a
marked effect on glass technology in the western
Asia and the Mediterranean: instead of natron
they made glass using ashes of salt-tolerant
shrubby plants. These plants grew in semidesert, evaporitic and maritime environments in
western Asia and parts of the Mediterranean
basin. Because the plants used for the flux could
grow in inland locations, one result was that
primary glass production became more
widespread across western Asia and into central
Asia. This led to a fully decentralized production
system with multiple primary production centres,
many located in cosmopolitan hubs on the silk
road.27 In inland locations suitable plants were far
more accessible as a source of flux than the far
more limited sources of the evaporitic mineral
used by the Romans to make natron glass.
Like the Romans, the Abbasids fused glass
raw materials (plant ashes and sand) in large tank
furnaces28 and added colourants to the glass to
produce deeply coloured glasses as part of the
secondary phase of production. It is worth noting
that, unlike natron, these plants had a highly
variable composition depending on a range of
environmental factors. One of these is the
geological nature of the soil in which the plants
grew. Although this might be viewed as
potentially confusing, using scientific analysis has
enabled plant ash glasses to be provenanced in
increasingly more geographically defined ways
(see below). Although the Muslims were partly
responsible for this transition in glass production
the other possible influence on this technological
change was the pre-Islamic manufacture of plant
ash glass by the Sasanians between the 3rd and 7th
centuries29 in modern Iran and Iraq for which
there is no published direct evidence for primary
glass production from raw materials. It is
nevertheless likely that the glass was made at
sites like Veh Ardašīr and Ctesiphon30 and Brill31
has suggested -based on the presence of tank
furnace fragments found on rural sites, some of
probable Sasanian date - that this is evidence for
primary glass production. Most of the glass
found in early medieval northwestern Europe,
including Dutch contexts, dating to after c.
800 AD therefore shows a dependence on the
import of ready-made glass combined with a
transition that occurred in western Asian glass
technology, with the production and export of
plant ash glass, especially in the Carolingian
period.
The exception to this dependence on
imported glass made in the Mediterranean basin
and western Asia (with associated recycling
during in the Carolingian period) was the use of
some of the earliest glass fused from raw
materials in northwestern Europe, from tree
ashes. Some of the earliest examples date to the
8th century, for example from the Loire valley in
France.32 However, its production became
widespread in the 11th century and later, especially
in response to the massive demand for cathedral
and church windows, such as in the Weald of
Kent in southern England, and France,33 but the
period between c. 800 and 1000 AD was one of
transition too.34
One of the hallmarks of the technological
transition in early medieval northwestern Europe
is the occurrence of mixed-alkali glass, a likely
combination of different proportions of wood
ash glass and natron glass. It is more likely that
fully fused glasses were mixed than that wood
ash was added to natron glass. This would have
formed part of a period of experimentation with
the new alkali raw material – wood ash. Like the
plant ash used to make glass in western Asia
discussed above, wood ash has a highly variable
chemical composition depending on the
geological characteristics of the soil in which the
tree grew, the tree species, the season in which
the ash is burnt and the part of the tree.35 As with
plant ash glass it is becoming increasing clear
that in some cases scientific analysis can help to
provide a geographical provenance (often
regional) for such glasses.36 From around 800 AD
glass linen smoothers make an appearance in
early medieval Europe. Scientific analysis has
revealed that these were made in Europe using
glassy slags derived from lead-silver cupellation.37
27 Henderson et al. 2021; Henderson 2022.
28 Aldsworth et al. 2002; Henderson et al.
29
30
31
32
33
34
35
36
37
2005a; Henderson et al. 2021.
Mirti et al. 2008; Mirti et al. 2009.
Simpson 2014, 204.
Brill 2005, 66.
Aunay et al. 2020.
Wedepohl 2008; Meek, Henderson &
Evans 2012; Henderson 2013, 104–108.
Henderson 2013, 97–108; Aunay et al.
2020.
Jackson, Booth & Smedley 2005.
Meek, Henderson & Evans 2012;
Adlington et al. 2019.
Gratuze et al. 2003.
16
—
Summarizing, the main types of glass in use in
Early Medieval North-Western Europe,
and therefore potentially available in the
Netherlands, were:
• imported pristine raw furnace natron glass;
• recycled and mixed natron glass (including
tesserae);
• imported pristine plant ash glass (made
from shrubs);
• wood ash glass made in northwestern Europe;
• mixed-alkali glass (mixed wood ash and
natron glass)
2.3
38 Willmott & Welham 2013; 2015.
39 Wedepohl, Winkelmann & Hartmann
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
1997; Gai 2005.
Willmott & Welham 2015.
de Sigoyer et al. 2005.
Andersen & Sode 2010.
Henderson & Ivens 1992.
Van Wersch et al. 2014.
Dell’Acqua 1997.
Sanke, Wedepohl & Kronz 2002.
Szöke, Wedepohl & Kronz 2004.
Dodt, Kronz & Simon 2021.
Dodt, Kronz & Simon 2021, Abb. 10.
Koch 1987.
Stepphun 1998, 94–96, Abb. 24.
Kronz et al. 2016.
Haevernick 1979.
Dodt, Kronz & Simon 2021, Abb. 11.
Evidence for the glass industry
in early medieval Europe outside
the Netherlands
Evidence for the glass industry in early medieval
Europe is quite sparse. Glass furnaces have been
found at Glastonbury (UK), associated with the
evidence for glass working,38 and Paderborn
(Germany).39 The evidence from Paderborn and
Glastonbury is only for glass working, not glass
making (the fusion of raw materials) and
Paderborn is probably later. The evidence at
Glastonbury consists of a furnace, a crucible
fragment with green glass adhering, lumps of
glass, spills of glass, a moil (the glass encircling
the tip of a blowpipe constituting production
waste which does not get recycled), pulls of
glass, a cast slab and a bichrome cable as well
as vessel and window glass fragments.
A possible glass furnace associated with glass
fragments and bichrome cables has also been
found at Barking in London.40 Glass furnaces
have also been found at Huy (Belgium).41
Evidence for glass working in the form of
crucibles containing glass, dribbles and drops
of glass, melted glass rods and cables, half made
beads, vessels and window glass has been found
(either separately or together) at a variety of
places. Evidence for beadmaking has been found
at Ribe (Denmark)42 and at Dunmisk (Ireland),43
where glass studs were also made. Some other
places have yielded evidence for glass blowing,
including the abbeys of Stavelot (Belgium)44 and
San Vincenzo al Volturno (Italy).45 Other evidence
has been found at the Carolingian monasteries
of Lorsch and Corvey (Germany)46 and Zalavar
(Hungary).47
There is also clear evidence that glass was
worked at Cologne during both the Merovingian
and Carolingian empires.48 Evidence of glass
production consists of fragments of glass
furnace floors, which were presumably beehiveshaped furnaces,though no plans of the
excavations are published so it is difficult to
judge. Multiple crucible fragments with glass
adhering, dribbles and drops of glass, reticella
rods, scraps of glass, tesserae and vitrified bricks
have been found. A distribution of Merovingian
loop decorated bowls down the Rhine has been
recorded suggesting that they were made in
Cologne mainly from HIMT-2 glass (see below). 49
Koch has shown that there is also a distribution
of early Merovingian cone beakers down the
Rhine, further supporting Rhenish production,
probably in Cologne. 50
At Hedeby (Germany) a possible glassworking area was found, including a possible
furnace.51 Two crucible fragments containing
wood ash glass, a single one with high lead glass
combined with wood ash and soda-lime glass
and raw glass have been reported.52 Evidence for
glass production has also been found at Cordel
(Germany) although an early medieval date has
been called into question.53
Scientific analysis of the glass from Cologne
using electron probe microanalysis alone shows
that secondary glass making involved HIMT-2,
which was originally probably fused in the early
to mid 4th century AD, and was mainly used to
make funnel and bell beakers – as well as funnel
beakers from Hedeby (Germany).54 A plot of
weight % Fe2O3/TiO2 versus Fe2O3/Al2O3 provides
evidence of a single funnel beaker from Cologne
made with Egypt-2 glass (originally made
between c. 720 and 780 AD), and funnel beakers
from Hedeby made from Egypt-1 glass
(originally made between c. 760/780 and 870 AD).
However, unless failed examples of funnel
beakers and moils, both of the appropriate
chemical composition, are discovered it is
difficult to be absolutely certain that the vessels
were blown in Cologne from weak HIMT (HIMT-2),
although it remains likely. Being the commonest
glass compositional type at the time, weak HIMT
is clearly not diagnostic to a specific production
centre. Therefore, Cologne has provided
evidence for glass working but not definite proof
for the manufacture of funnel beakers there.
17
—
2.4
Scientific analysis of early medieval
glass in Europe
2.4.1 The principal glass types
Syro-Palestinian and Egyptian glass
The primary characterization of pristine Levantine
and Egyptian natron glass was carried out by
Nenna et al.,55 Foy et al.,56 Freestone et al.,57 Phelps
et al.,58 Freestone et al.59 and Schibille et al.60
These studies focused on glass which derived
from primary glass-making sites in the SyroPalestinian area and Egypt and shows that
different proportions of minerals such as zircons,
chromite and feldspars can characterize the sands
used to make glasses at different production sites
and at different times. Two compositional groups
of natron glass produced in Israel in the mid-late
first millennium AD have been widely recognized
so far. Levantine I is defined according to the
chemical compositions of sixth to seventh
century glass from Dor and Apollonia. The sand
used for making Levantine I glass was probably
derived from the Bay of Haifa, close to the
mouth of the river Belus of antiquity. As Phelps
et al.61 have noted the use of the term Levantine I
has masked other compositionally related but
distinct glass, such as that made at 4th century
Jalame. Levantine II was defined using the
chemical composition of furnace glass produced
at Bet Eli’ezer and is dated to the 8th century Late
Byzantine – Umayyad period.62
Two compositional groups were recognized
by Gratuze and Barrandon63 in their study of
early Islamic glass coin weights from Egypt.
Since then the two groups have been referred to
as Egyptian I and Egyptian II.64 Egyptian I glass
typically has high alumina (3–4.5 wt%) and low
lime (3–4 wt%).65 It has been suggested that this
glass was produced in factories near the famous
natron source at Wadi el Natrun. Egyptian II
glass has relatively high lime (c. 9%) and low
alumina (typically 1.5–2.5%). Recent detailed
analysis of Egyptian natron glasses has revealed
the existence of Egypt 1A dating to before
725 AD, Egypt 1B dating to between 720 and
780 AD and Egypt 2 dating to between 760/780
and 870 AD.66
Because these glasses from different
primary production sites have clear compositional
characteristics, they should be identifiable
amongst early medieval Dutch glass. It has been
noted that the levels of sodium oxide decreased67
and aluminium oxide increased over time68 in
these pristine glasses, due to a shortage of
natron, and the use of different sand deposits,
respectively.
Other natron glass
Tesserae
When glass tesserae made out of natron glass
were mixed with other natron glass to extend its
volume, certain compositional characteristics in
the tesserae are passed on to the bulk glass. As
discussed above the occurrence of elevated
levels of antimony, copper, lead and sometimes
tin in translucent vessel glass suggests that a
stock of glass tesserae has been mixed into the
bulk glass.69 Elevated levels of antimony indicate
this especially because calcium antimonate was
used almost universally as the opacifier in a high
proportion of Roman glass tesserae.
HIMT and its variations
High iron, manganese and titanium (HIMT)
oxide levels that are found in 4th–5th century
HIMT natron glass indicate a probable Egyptian
source.70 High levels of these oxides as well as
zirconium show that sands with high proportions
of minerals bearing these elements were used to
make the glass. A higher iron variation has also
been identified.71
Much work has been carried out to investigate
variations of HIMT glasses which have been
found in the Mediterranean and in northern
Europe. The most important variation of HIMT
sensu stricto is the Foy 2 compositional group.
Foy 2 was originally identified in glass from
Carthage72 and includes 6th century series 2.1,
which contains elevated V, Ti and Zr, and series 3.2
as originally described by Foy and colleagues.73
These types of glass are regarded as ‘weaker’
types of HIMT with lower concentrations of iron,
manganese and titanium and variously labelled
HLIMT (high lime, iron, manganese and titanium),74
weak HIMT75 and HIMT 2. Foy 2.2 was originally
defined by Foy and colleagues as a recycled
version of Foy 2.1.76 This type of glass may have
been recycled multiple times and is found across
the Mediterranean and Europe in contexts
dating to as late as the 9th century.77 The ultimate
origin of the original base glass used in these
glasses – which would then have been recycled
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
Nenna, Vichy & Picon 1997.
Foy et al. 2003.
Freestone, Gorin-Rosen & Hughes 2000.
Phelps et al. 2016.
Freestone et al. 2018.
Schibille et al. 2019.
Phelps et al. 2016.
Freestone, Gorin-Rosen & Hughes 2000.
Gratuze & Barrandon 1990.
Freestone, Gorin-Rosen & Hughes 2000.
Wt% = percentage of each oxide by
weight.
Schibille et al. 2019.
Henderson 2002.
Phelps et al. 2016.
Henderson 1991a; Schibille & Freestone
2013; Boschetti et al. 2016; Henderson,
Sode & Sablerolles 2019; Crocco et al.
2021.
Foy et al. 2003.
Ceglia et al. 2015.
Schibille, Sterrett-Krause & Freestone
2016.
Foy et al. 2003.
Ceglia et al. 2019.
Conte et al. 2014.
Foy et al. 2003.
Bertini, Henderson & Chenery 2020.
18
—
and mixed with other glasses – was probably
Egypt, with elevated proportions of heavy
minerals such as zirconium characterizing
Egyptian glass.
Foy 2.1 high iron found in Byzantine glass
weights78 has also been recognized in AngloSaxon Britain, Serbia, Merovingian France and
Spain.79 Furthermore, another compositional
variation, HIT (high iron and titanium), has been
recognized from 5th–6th century Bulgaria80 and
possibly from 5th–6th century Albania.81 HIT is
unlikely to have been made in Egypt or the
Levant; where it was made precisely is unknown.
A plant ash variant of the Foy 2 family has been
recognized by Schibille et al.82 in Byzantine glass
weights.
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
Schibille et al. 2016.
Ares et al. 2019.
Smith, Henderson & Faber 2016.
Conte et al. 2014.
Schibille et al. 2016.
Henderson et al. 2021.
Henderson et al. 2016; Siu et al. 2020;
Schibille et al. 2020.
Henderson 2002, Table 1, analyses 13,
22–25, 31, 41, 48 and 49.
Henderson 1999, 288, Table 2, analysis
32.
Sablerolles & Henderson 2012, analysis
129.
Henderson 2023.
Dekówna 1980, Table 44.
Kronz et al. 2016.
Langbroek 2021b.
Jankowiak 2021.
Aunay et al. 2020; Pactat et al. 2017.
Kronz et al. 2016.
Dekówna 1978, Tables 2 and 4; Dekówna
1980, Table 48, nos. 3, 4.
Kronz et al. 2016.
Wedepohl, Winkelmann & Hartmann
1997.
Kind, Kronz & Wedepohl 2002, Table 2.
Henderson 1991b, 129, Catalogue
numbers 240 and 250, Appendix;
Henderson 1993.
Henderson 2012.
Henderson 2012.
Henderson 2023.
Pactat et al. 2017.
Kronz et al. 2016.
Van Wersch et al. 2014.
Plant ash glass
From c. 9th century AD plant ash glasses
manufactured in the Islamic domain in western
Asia and the southern Mediterranean were
made in cosmopolitan centres along the silk
roads stretching from Spain to central Asia.83
Because local sources of plant ashes were used
to make the glass their chemical and isotopic
characteristics relate to the geological or
geographical location in which they were made.
A range of minor and trace elements, including
their ratios, can be used to characterize the raw
materials used to make the glasses, such as Ca,
Mg, Ba, Zr, Ti, Cs/K, Li/K, Li/Na, B/Na, 1000Zr/Ti,
La/Ti, Cr/La, Ce/Zr, Y/Zr, Mg/Ca.84
Examples of plant ash glasses that have
been found amongst early medieval European
glass are millefiori, blob decorated, melon and
chopped beads from 10th century Viking age
burials at Peel on the Isle of Man,85 a colourless
silver foil glass bead from Wijnaldum,86 the blue
cable applied to the rim of an 8th–9th century
pale green funnel beaker from Dorestad,87 in
the matrices of two 9th century trail-decorated
funnels and in the body of a funnel beaker as
well as beads found at Susteren.88 For a closer
consideration of Susteren examples see below,
and the use of trace element analysis in
Chapter 5. Dekówna89 reports the analysis of
eight artefacts, a glass drop, a rod, five segmented
beads and an annular bead from Hedeby
(Germany) that are plant ash glasses opacified
with lead stannate. Kronz et al.90 note that some
plant ash glasses from Hedeby were mixed with
lead. Moreover, typical Islamic millefiori glass
beads have been found at Dorestad.91
Like the distribution of Islamic coins
(dirhams),92 it appears that Islamic plant ash
glasses are mainly restricted to the northern
Netherlands, Scandinavia and the Baltic,
coinciding with the presence of Vikings, and
far fewer examples have been found in France,93
with negligible numbers in Germany until the
9th century.94
Mixed-alkali glass
Most glasses with mixed-alkali compositions
that have been published mainly date to the
9th–10th centuries, reflecting the transition from
natron to wood glass technology in northern
Europe. There is no doubt that wood ash glass
was made in northern Europe. Mixed-alkali
glasses have been found at Hedeby (Germany)95
including a single funnel beaker,96 and at
Paderborn (Germany)97 with two examples from
the abbey at Fulda (Germany) with relative
potassium oxide/soda levels of 7.7/6.6 and
6.0/9.1.98
Two pre-851 and an 11th–12th century AD
(possibly redeposited) mixed-alkali green
window glass have been found at Lurk Lane,
Beverley (UK).99 Two examples of mixed-alkali
glass have also been found at Dorestad: a late
8th century yellow-green palm funnel100 and a
9th century AD yellow-green funnel beaker.101
The analysis of a yellow-green tubular base of
a classic funnel beaker from Susteren was also
of a mixed-alkali composition.102
Comprehensive evidence for mixed-alkali
glass working has been found at the c. 800–
866 AD Carolingian site of Méru, Oise (France).
The archaeological evidence is in the form of
furnace walls and crucibles containing glass.
Both natron glass and glasses with potassium
oxide levels of between c. 2% and 12% were
found. This potentially suggests that Méru
was one possible location where the mixture
of wood ash and natron glass or, less likely,
the addition of wood ash to natron glass,
actually occurred103 although the case is not
completely proven. It is also suggested that
mixing of natron and wood ash glass occurred
at Hedeby in Germany.104
Wood ash glass
High potassium levels were introduced using
wood ash as a flux. High potassium glasses have
been published from the abbey of Stavelot
(Belgium).105 These glasses are quite early, from
19
—
contexts dating to between the second half of
the 7th century to the early 9th century AD.
The twelve examples are characterized by soda
levels of below 3%, potassium oxide levels of
between 10.14% and 20.21% and calcium oxide
levels of between 13.84% and 20.23%. All
contain high magnesium oxide and phosphorus
pentoxide levels; the rest of the glass from the
site is natron glass – there is no plant ash glass.
In addition, full wood ash window glasses
characterized by very high potassium and
calcium oxide levels like those from Stavelot
have been found in probable late 8th to early
9th century AD contexts at Baume-les-Messieurs,
Jura (France).106 Wood ash glasses have also been
found at the church of St Hermès-et-Alexandre
(Belgium).107
In Germany wood ash glass has been found
at ecclesiastical sites, at the abbeys of Lorsch,
Corvey, Brunshausen–Gendersheim and Fulda.108
Excavations of the 10th century site of La Milesse,
Sarthe (France) has produced large-scale evidence
for working high potash glasses. The glasses that
were scientifically analysed had a relatively
restricted compositional range which led Pactat
et al.109 to suggest that the glass was fused there.
Kronz et al.110 have reported the presence of nine
funnel beakers made with wood ash (out of a
total of 61 wood ash glasses) and two crucibles
containing wood ash glass from Hedeby
(Germany). It was noted that wood ash glass
contains higher calcium oxide levels in glass that
dates to post 1200; the range detected was
between c. 10% and 22% with the majority being
between c. 10% and 16%.111 Some of the earliest
wood ash glasses contain up to about 11%
calcium oxide, such as from Hedeby. Six funnel
beakers from Borg in Norway112 dating to
between 800 and 1000 AD contain between 11.2
and 12.7% calcium oxide.
High lead glass has been reported from a
variety of northern European sites. The first
high lead glass appears c. 10th century AD.
Four examples have been reported from
Hedeby with 22 other glasses being a mixture
of plant ash glass and lead.113
Glass making in early medieval Europe
therefore only involved wood ash to produce
high potassium glasses. Before around 800 AD
glass-workers relied on the import of scrap and
raw glass which was either formed directly into
objects or mixed with other glass.
2.4.2 Summary of existing scientific
analyses of early medieval Dutch
glass before the start of this project
Scientific analysis of early medieval glasses from
Maastricht, Susteren, Wijk bij Duurstede
(Dorestad), Rijnsburg, Wijnaldum, Lent,
Borgharen and Sittard had already been carried
out prior to this study. What follows is a brief
summary of the results from these investigations.
Maastricht
The early medieval glass and glass-bearing
objects from Maastricht114 include a good
number of glass beads (including failed beads),
glass rods and crucibles with vitreous materials
attached. Seventeen glass crucibles with
colourless, pale green and opaque yellow
residues attached were discovered at
Jodenstraat and are discussed in much more
detail in Chapter 5. Apparently sintered material
was attached to one of the crucibles found at the
Mabro site and it was suggested that this might
possibly be frit (though see Chapter 5).
The results of the analyses are given as
means and standard deviations for each glass
colour, for window glass samples and for
two separate samples of crucible glass.115
Using these results, a translucent blue splinter of
glass resulting from glass working could possibly
be of a pristine Levantine glass (but see
Chapter 5). The glass is coloured with 0.1%
cobalt oxide in the presence of 0.2% cupric
oxide. The three window fragments contain
elevated MgO, K2O, TiO2 and MnO, with low
antimony oxide concentrations, so are likely to
be recycled/mixed glass varieties of weak HIMT.
Yellow-green glass from inside a crucible is
contaminated with 8.7% Al2O3; it contains low
CaO (3%) as well as elevated TiO2 and Fe2O3. It is
therefore difficult to estimate what its original
composition was, especially if the same elements
(e.g. Al and Fe) were present both in the original
glass and in the crucible wall from which they
migrated into the glass. Colourless glass
attached to another crucible wall has very similar
elemental contamination.
The opaque yellow residues (n=4) contain
the highest PbO levels (30.6±10.02) associated
with tin oxide because they relate to the
production of lead-tin oxide (Pb2Sn2O7/Pb2SnO4).
106 Van Wersch et al. 2015.
107 Van Wersch, Mathis & Hoffsummer
2009.
108 Sanke, Wedepohl & Kronz 2002; Kind,
109
110
111
112
113
114
115
Kronz & Wedepohl 2002; Wedepohl
2003.
Pactat et al. 2017.
Kronz et al. 2016.
Kronz et al. 2016, Abb. 9.
Henderson & Holand 1992.
Kronz et al. 2016.
Sablerolles, Henderson & Dijkman 1997.
Sablerolles, Henderson & Dijkman 1997,
Table 2.
20
—
They contain slightly elevated MgO and
relatively high manganese and iron oxides of
above 1%. The opaque white residues (n=2)
contain high tin oxide and lower lead oxide
because they are opacified with tin oxide
crystals. These opaque white residues also
contain higher magnesia levels (1.3±0.2%) than
the opaque yellow glasses, something that has
been found in Roman opaque white enamels
and tesserae.116
The four opaque red glasses analysed
contain only slightly elevated levels of magnesia
and potassium oxide (means of 1.15% and 0.83%
respectively). High levels of ferrous oxide
(3.8±0.63%) may indicate that iron-rich crystals
contribute to the red colour (see Chapter 5 for a
more detailed interpretation); 0.37% ZnO
possibly indicates that brass filings were used as
a source of copper colourant. Two opaque green
glasses which contain elevated MgO (with a
mean of 1.2%) are coloured with cupric oxide
(mean of 4%). They are opacified with lead
stannate. Zinc has also been detected,
suggesting that brass was added as the copper
source. A single opaque turquoise glass was
probably opacified with lead-tin oxide. It is
coloured with 2.3% CuO and contains elevated
MgO at 1.2%.
116 Henderson 1991a.
117 Henderson 2023.
118 Henderson 2012.
Susteren
Six beads, eleven vessels, two crucibles and ten
windows were analysed using electron probe
microanalysis and Scanning Electron
Microscopy.117 The glass beads and vessels are
primarily of natron glass having similar
compositional characteristics to those from
Wijnaldum and Maastricht. All contain between
0.74% and 0.83% MgO. In beads 4 and 6 the K2O
levels are above 1%. In most cases the TiO2 levels
are between 0.1% and 0.2%; bead 5 contains
0.22% TiO2. These natron glasses contain
between 2.49% and 2.76% Al2O3 with the
exception of bead 6 which contains 3.05% Al2O3.
Being beads the glasses are coloured in various
ways: the ‘black’ (deep translucent brown) body
of bead 1 is coloured with 5.09% Fe2O3; a
combination of MnO and Fe2O3 has produced the
green colour in the bodies of beads 2, 5 and 6.
Most of the translucent bead bodies (and
opaque decoration) contain low levels of Sb, Pb
and Cu oxides – an indication of a level of
recycling involving the addition of coloured
Roman glass, including tesserae (which are
invariably coloured and opacified with Ca2Sb2O7).
The bodies of beads 3 and 4 contained high
levels of magnesia, potassium oxide and
phosphorus pentoxide and are therefore plant
ash glasses. Bead 4 contains the lowest
concentration of Al2O3 at 2.06%.
The eleven glass vessels from Susteren that
were analysed consist of nine soda-lime natron
glasses, one mixed-alkali glass (no. 19) and one
plant ash glass (no. 28). The natron glasses
contain between 0.76% and 1.2% MgO, up to
1.24% K2O and some have elevated P2O5 (e.g. no.
23 with 0.37%). TiO2 concentrations range from
0.09% (no. 25) to 0.4 (no. 26) so are variations
of HIMT.
The Susteren vessel glasses are colourless,
cobalt blue, blue-green, pale green and pale
yellow. The pale yellow glass is probably
coloured with ferrous oxide. All glasses contain
elevated concentrations of CuO, Sb2O3 and PbO.
No copper was detected in the colourless sample
(no. 26). These oxides are indications of mixing
and recycling.
The mixed-alkali glass (no. 19), a pale green
funnel beaker with a tubular base, contains
8.76% Na2O and 8.5% K2O. It also contains
3.69% MgO, 1.5% P2O5 and 9.73% CaO all of
which are quite distinctive characteristics
associated with the inclusion of an organic flux.
The blue-green plant ash glass trechterbeker
fragment (no. 28) contains the lowest Al2O3
(2.03%) of the vessel glasses showing that a
purer silica source was used. It is also characterized
by 4.2% MgO and 2.48% K2O.
The opaque red, yellow and white decorative
elements used on beads 2, 3, 4, 5 and 6 mainly
follow the same pattern of colourant use as
discussed for the Wijnaldum and Maastricht
opaque glasses: elevated Fe (4.59%) and CuO2
(1.74%) in red glass, high PbO and SnO2 in opaque
yellow (probably in the form of Pb2Sn2O7
crystals) as well as a combination of Pb and Sb
which are probably responsible for opaque
yellow (in the form of Pb2Sb2O7 crystals) in the
decoration of bead 5.
Wijk bij Duurstede (Dorestad)
Forty two vessels, one glass chip, five tesserae,
one rod and two linen smoothers were analysed
using electron probe microanalysis.118 Of these,
39 of the vessels are of a soda-lime natron glass
composition. The remaining three consist of one
plant ash blue trail from the rim of a trechterbeker
21
—
(no. 129b), and two potassium (wood ash) palm
funnel (late 8th century) and funnel beaker
(9th century) glasses (nos 103 and 136).
These contain relatively low potassium oxide
levels at 8.6% and 7.9% respectively, associated
with 1.6% and 1.2% soda levels and 14.1%
and 13.8% calcium oxide levels. These
represent early examples of wood ash glasses.
All the natron glass has very similar
characteristics to those already discussed for
Wijnaldum, Maastricht and Susteren.
The glasses contain between 2.1% and 3.34%
Al2O3 but mainly fall between 2.5% and 3.0%,
CaO concentrations of between 5.85% and
8.59% and P2O5 at 0.37% and 1.8%. Some
contain slightly elevated levels of MgO and K2O
and some elevated levels of P2O5 and MnO.
Unsurprisingly, such characteristics suggest that
the glasses have been recycled or mixed and are
weak HIMT.
The 32 Dorestad blue-green and pale green
vessel glasses are coloured by a combination of
MnO and Fe2O3, some with Fe2O3 levels up to
2.7%. The greenish translucent glasses are also
coloured with CuO levels up to 2%. There is a
statistically coherent number of pale green/
nearly colourless as opposed to blue-green
vessel fragments from Dorestad. There is some
evidence that the pale green and especially the
nearly colourless glasses contain lower Al2O3
than blue-green glasses, though there are
exceptions. MnO levels in all shades of green are
at similar levels, mainly at between 0.5% and
0.7%. The combination of MnO and Fe2O3 will
impart colour to the pale green glass given a
specific oxidizing/reducing atmosphere in the
glass furnace. The elevated CuO levels in the
blue-green glass therefore appear to provide the
deeper green colour. The lower Al2O3 in the pale
green and colourless glass suggests that the
glass was made with a slightly different sand
source from the blue-green glasses.
The blue plant ash glass trail used to
decorate a trechterbeker (no. 129a) is coloured
with cobalt oxide (0.06%). The body of the nearly
colourless beaker decorated with gold foil might
be expected to be a purer glass but it has a very
similar ‘intermediate’/Foy 2 composition to
other green Dorestad glasses with no obvious
use of a decolourizer or evidence that a ‘special’
glass was used. However, the furnace atmosphere
must have been controlled carefully to produce
the colourless glass. The red beaker base
contains 2.73% Fe2O3 and 1.51% Cu2O associated
with 0.66% ZnO and 1.4% PbO. The copper-rich
colourant (probably in the form of cuprite droplets)
used may therefore well have included scrap brass.
Rijnsburg
Thirteen Merovingian glass samples from the
glass-working site of Rijnsburg were analysed
using electron probe microanalysis. Six were
beads or unfinished beads, six were rods and
one was a sample of crucible glass.119 The only
translucent glass is a turquoise rod of a natron
composition.
All the yellow glasses were opacified with
lead-tin oxide and both white glasses opacified
with SnO2 but unusually they contained low
levels of MgO, whereas seven other opaque
white early medieval glasses from Maastricht
and Wijnaldum contain elevated MgO levels.120
It can be suggested that the Rijnsburg white
glasses were made at a separate source from
other white glasses. Although only two red
glasses were analysed the same thing is true for
them: neither of them contain elevated MgO
and K2O, something that is found almost
universally in other early medieval opaque
red glasses.
The chemical compositions of the two red
and two orange glasses from Rijnsburg are quite
different from each other. The red glasses
contain much higher Fe2O3 than detected in the
orange glasses (5.7% and 3.2% versus 0.8% and
0.5% respectively). On the other hand, the
orange glasses contain 20% and 18% Cu2O as
opposed to 5.7% and 3.2% Cu2O in the red
glasses. The orange glasses are therefore likely
to contain denser and larger Cu2O crystals than
the opaque red glasses. The latter are liable to
be in the form of micron sized copper droplets
or cuprite. The opaque yellow material on the
inside of the crucible fragment consists almost
entirely of Pb(O) and Sn(O2): 79.8% and 11%
respectively, with 4% SiO2 and may be evidence
for production of lead-tin pigment on site.
Wijnaldum
The scientific analysis of the glass and glassbearing artefacts from Wijnaldum reported on
38 electron microprobe analyses of twelve
vessels dating to between 450 and 900 AD,
24 beads dating between 550 and 900 AD,
a vitreous blob attached to a crucible dated to
250–350 AD and glass attached to a crucible
119 Dijkstra, Sablerolles & Henderson 2011.
120 Dijkstra, Sablerolles & Henderson 2011,
fig. 12.
22
—
121
122
123
124
125
126
Henderson 1999.
Henderson 1985; Matin 2019.
Steppuhn 1998, 100–101.
Henderson 1991a; Silvestri et al. 2012.
Corbella 2017.
Van Os et al. 2014.
dating to between 575 and 620 AD.121 The
translucent glasses are mainly glass vessels;
three are glass beads. The beads are mainly
deliberately coloured with the oxides of
transition metals, manganese, iron and copper.
Opacification is due to the formation of tin
oxide, SnO2 (white), lead-tin oxide Pb2Sn2O7
(yellow) and cuprous oxide Cu2O (red).122 High
iron associated in the opaque red glasses would
have acted as an internal reducing agent.
Apart from one, the translucent glass
samples are natron glass containing elevated
MnO and Fe2O3. The exception is a single drawn
segmented silver foil bead which has a full plant
ash composition with much higher levels of MgO
(5.5%) and K2O (2.2%) which are comparable
with Islamic glasses found at Hedeby.123 Elevated
MgO levels of c. 1% are present in two vessel
glasses; the earliest glass vessel dating to
c. 450 AD contains 0.3% TiO2 which is also
unusually high – its significance will be discussed
below. Three colourless glasses were analysed,
one of which is the silver foil bead. One glass is
probably decolourized with 0.4% antimony
trioxide. Some translucent glasses also have
trace levels of TiO2, CuO, Sb2O3, SnO2 and PbO,
all indicators of glass recycling or mixing,
including the use of Roman glass tesserae.
The dull opaque red glass bead contains
high potassium and magnesium oxides,
a characteristic of Roman enamels and
tesserae124 so this is evidence that this kind of
glass continued to be used in the early medieval
period. An opaque red globular bead contains
unusually high CaO (11.7%) and relatively low
Al2O3 (1.9%). The presence of tin suggests that
scrap bronze was used as a copper-rich
colourant in the bead; tin is absent from the
other opaque red beads. All red glass beads
contain lead ranging from 1.2% to 14.2%.
An orange bead is opacified with copper in the
presence of iron, the orange rather than red
colour was possibly attributable to differences
in the sizes of copper crystals in the glasses.
Lead stannate in crystalline form is the opacifier
in opaque yellow-green glass beads (numbers 18
and 23) which are otherwise coloured with
copper and iron respectively. The opaque yellow
glasses contain between 21.5% and 54.5% PbO
and are opacified with Pb2Sn2O7/Pb2Sn2O6
crystals. The opaque yellow material on the
inside of the flat tray or less likely a furnace
fragment (number 38) contains 63.5% PbO.
It may have been a flat plate on which the
opaque yellow residue was heated probably at
c. 650°C. It is clear from the composition that the
glass-like material is contaminated by interaction
with the ceramic substrate. A single vitreous blob
from a crucible that was analysed (number 37)
contains high Al2O3 (9.7%), 63.5% PbO and 2%
total alkali – it is probably a fuel ash slag.
Both an opaque yellow tessera and the yellow
spiral trail decorating the rim of a beaker from
Wijnaldum are opacified with lead antimonate.
Though opaque glasses contain elements
associated with their colour and opacity they
also can contain elevated MgO, K2O, CoO, CuO.
The presence of these oxides could suggest that
the base soda glass used to make the opaque
glasses was recycled, though more detailed
analysis of opaque yellow glasses in Chapter 5
provides interesting new information.
Lent
Corbella125 used p-XRF to analyse 30 glass beads
that were found in four graves in the Merovingian
cemetery of Lent. Most glasses are probably of a
natron composition with additives to modify the
glass colours. It was not possible to analyse
sodium. In this case the data should be treated
as indicative because it’s possible that different
colour compositions were combined in one
analysis. The red glasses analysed contained
elevated potassium, iron and copper as found in
other such glasses. The detection of MgO was
unreliable. Two of the glasses analysed apparently
contained alumina levels above 4% with relatively
low CaO so potentially might have originated in
south Asia. The white glass analysed contained
low tin and elevated antimony so may be
opacified with calcium antimonate. Two green
glasses may be coloured with a combination of
iron and manganese oxides.
Borgharen
The results of semi-quantitative analyses of the
glasses from Borgharen produced using p-XRF
are listed according to their colour.126 The opaque
white glasses mainly contain tin with low levels
of antimony and are therefore likely to be
opacified with SnO2 crystals. However, some
very high levels of P2O5 have been listed. If these
are so high, which does not seem likely, then
bone ash needs to be considered as a possible
white opacifier. Opaque blue glasses are also
apparently opacified with tin oxide. The glasses
23
—
contain relatively low levels of lead oxide.
According to these results, the same opacifier
was used in opaque brown glasses. The
colourless glass may have been clarified with
manganese oxide. The translucent green glass
seems to contain tin and lead so may be semiopaque. The opaque orange glass contains c. 6%
PbO and unusually high CaO. A compositionally
quite consistent series of eight opaque red
glasses indicate that they contain high tin and
iron. It is not completely clear from these analyses
that copper is responsible for the colour and/or
the opacification. However, a comparison with
analyses using ICP techniques and SEM EDX
show that these analyses are fully quantitative. 127
Sittard
A plot of relative levels of Pb and Sn in all
analysed glass from Borgharen and Sittard using
p-XRF shows very similar patterns for both sets
of beads.128 These data for Sittard, in particular,
have a strong positive correlation. The beads
from Sittard contain Sn levels of between 3%
and 12.5%. Such high levels must be due to a
high density of Pb2Sn2O7 crystals in the glasses –
analysis of the matrix glass between the crystals
using ED-XRF would likely produce a lower level
of Sn in the matrix. The Sittard beads contain
PbO levels of between 12% and 35%. It is
notable that the Borgharen data are more
scattered in terms of relative Pb and Sn levels,
especially at levels below 12% PbO and c. 3%
SnO2. The overall positive correlation between
the Pb and Sn suggests that these colouring/
opacifying elements were added together to
the glass melt.
There is a contrast in the levels of CaO
between the yellow glasses from Sittard and
Borgharen,129 with Sittard containing significantly
lower levels. A majority of glasses from Sittard
contain c. 2% CaO whereas Borgharen glasses
have a peak of CaO levels at 4% so this suggests
that different recipes may have been used to
make the glasses at the two sites.
Discussion
‘Roman’ natron glass was still in circulation
c. 800 AD, and later130 so by examining
scientifically European early medieval glass
assemblages of vessels and beads it becomes
possible to assess the extent to which glasses
have been imported from the Middle East
(whether natron or plant ash glass). Thus the
scientific investigation of this technological
transition also provides a way of demonstrating
how local glass production developed in Europe,
especially with the emergence of glass made
using wood ash. Characteristic Merovingian and
Carolingian vessel types were blown and bead
types made in Europe from imported raw
furnace glass or glass cullet in a secondary
production process.
For the high lead glasses, rather than
removing the PbO and recalculating the totals to
100% the use of ratios in Figs. 2.1 and 2.2 is
considered to be an alternative and equally
acceptable means of presenting the data so as to
investigate its provenance.
All natron glasses (the majority of the
analyses considered) including the base glass for
opaque glasses were possibly fused in Egypt or
the Levant. The natron source would have been
Wadi el Natrun in Egypt.
Therefore, what follows is a consideration
of the provenance of the glasses and by inference
also the provenance of the raw materials since
local or easily accessible raw materials would
have been used.
The Wijnaldum (Wij) glasses have a wide
compositional range according to major (Na2O,
CaO and SiO2) and minor components (Al2O3).
There are three outliers (Wij 16, 17 and 32) which
obscure some of the finer detail of the remaining
data in Fig. 2.1. They plot as outliers because Wij
16 contains a very high CaO level (11.7%), Wij 17
is probably not a glass and Wij 32 is a plant ash
glass (5.5% MgO and 2.2% K2O) rather than a
natron glass. Wij 37 has a very unusual
composition. It is possibly a plant ash glass,
but contains 9.7% Al2O3 and 3.7% Na2O.
Therefore in Fig. 2.1 the data for Wij 16, 17,
32 and 37 have not been included. This allows for
a more sensible comparison between the results
for glass from other early medieval Dutch sites
as well as other selected contemporary sites.
This plot is an approximate way of defining
compositional types and potential for raw material
provenance used by other researchers.131 The plot
suggests that Wij 1, 2, 3, 9, 13, 18, 20, 23, 25, 30,
and 31 are potentially Levantine glasses (though
see Chapter 5) and that Wij 24 is of the HIMT type
and therefore potentially made in Egypt.
The main bulk of early medieval Dutch natron
glass fall into the Foy 2 compositional types (and
its variants).132 These glasses have undergone
127
128
129
130
131
132
Personal comment Hans Huisman.
Huisman et al. 2019.
Van Os et al. 2014.
Phelps et al. 2016.
Bertini, Henderson & Chenery 2020.
Foy et al. 2003; Ceglia et al. 2015; 2019;
Schibille, Sterrett-Krause & Freestone
2016; Bertini, Henderson & Chenery
2020.
24
—
Fig 2.1 A plot of Na2O/SiO2 versus CaO/Al2O3 in early medieval glass beads and vessels compared to glasses from Jarrow and Monkwearmouth, UK (Brill 2006;
Freestone & Hughes 2006).
133 Freestone, Gorin-Rosen & Hughes 2000.
134 Phelps et al. 2016.
135 Gallo et al. 2015.
multiple recycling/mixing episodes. Even if they
were originally pristine Egyptian glasses before
they were mixed their provenance cannot now
be identified exactly. The glasses typically
contain elevated Fe2O3, TiO2, MnO and also some
elevated MgO, P2O5 and K2O.
The plotted points for the mean results of
both translucent and opaque glasses from
Maastricht all fall within the area that coincides
with the intermediate/‘Foy 2’ glasses. This is to
be expected given some of their other
compositional characteristics discussed above,
including quite high MgO, MnO, and Fe2O3 and
therefore come into consideration for HIMT or
related glasses. None of the glasses contain high
TiO2 levels. A single translucent blue chip of glass
appears to have a pristine Levantine
composition: with a Na2O/SiO2 ratio of 0.14 it
falls well below other glasses considered here
– due to its 74% SiO2 – and can therefore be
tentatively suggested as a product of Bet Eli’ezer
characterized by such high SiO2 levels133 although
its Al2O3 level of 2.4% is c. 0.5% lower than would
be expected.
Fig. 2.2 should be a more precise way of
defining different compositional types of natron
glasses. There are five outliers: Wij 11, 13, 14, 15
and 33. The reason why these are outliers is that
they contain high levels of iron: blue-green no.
11 (2.2%), opaque red no. 13 (4.6%), opaque red
no. 14 (4.1%), opaque red no. 15 (4.4%) and dark
olive green no. 33 (3.2%).
Without plotting the outliers in Fig. 2.2 a
better classification of Wijnaldum glasses than
seen in Fig. 2.1 becomes evident. By far the
largest proportion of samples fall into the same
‘intermediate’/‘Foy 2’ composition as seen in
Fig. 2.1. None of the Wijnaldum glasses fall
below the Fe/Ti value of 0.5, a value that may
be technologically significant for the few
Susteren and Dorestad intermediate/‘Foy 2’
glasses that do. The low values of Fe/Al of <0.2
for four Wijnaldum glasses (numbers 1, 9, 30
and 31) could suggest that they are unrecycled
Levantine glasses, but see Chapter 5.
Wijnaldum 1 could potentially be of a
Levantine–Apollonia type134 and Wijnaldum
30 perhaps of the Levantine–Jalame type.135
25
—
Fig 2.2 A plot of Fe2O3/Al2O3 versus Fe2O3/TiO2 in Dutch early medieval glass beads and vessels compared with results from Jarrow and Monkwearmouth UK
(Brill 2006; Freestone & Hughes 2006).
These glasses are characterized by low MgO and
K2O at c. 0.5%–0.7% weight, Al2O3 c. 3%, and low
iron and titanium. Overall, they lack the elevated
MgO, K2O, TiO2, Fe2O3 and MnO (and sometimes
P2O5) that characterize ‘Foy’ glasses, found in the
bulk of the remaining data. However a more
detailed interpretation is given in Chapter 5.
Three Wijnaldum glasses, 5, 7 and 12,
appear to plot with HIMT (high iron, manganese
and titanium) glasses in Fig. 2.2. Careful evaluation
of the results indicates that glasses 5 and 7 are
unlikely to be HIMT, glass 12 is more typical of
the composition (i.e. it contains 0.3% TiO2).
The difficulties in allocating a glass type to
these natron glasses are overcome significantly
by analysing them with the more sensitive
technique, laser ablation inductively coupled
plasma-mass spectrometry (LAICP-MS):
see Chapter 4.
Maastricht mean results are plotted in
Fig. 2.2 (triangles). All of the glasses, irrespective
of whether they are translucent or opaque,
fall into the area of the plot occupied by
‘transitional’/‘Foy 2’ glasses. This suggests
provisionally that all the glasses analysed from
the site are recycled/mixed and therefore
conform to the bulk of other analyses of early
medieval glasses from the Netherlands using the
relative values of Na2O/SiO2 and CaO/Al2O3. None
fall into the area of the plot occupied by pristine
Levantine natron glasses or HIMT glasses. See
Chapter 5 for a more detailed consideration.
Although it was suggested above that the
translucent blue chip of glass from Maastricht
might be a pristine Levantine glass, when
plotted in Fig. 2.2 this appears not to be the
case. As noted above most have elevated MgO,
MnO and Fe2O3 concentrations, above the values
found in pristine Levantine glasses (and pristine
Egyptian HIMT glasses); trace element analysis
(Chapter 5) reveals a more definitive means of
classifying the glass. Early medieval opaque and
translucent glasses have not been considered
together before in this way. This initial evidence
therefore suggests that recycled/mixed early
medieval glass was used as a base glass for
26
—
136
137
138
139
140
Bertini, Henderson & Chenery 2020.
Freestone, Gorin-Rosen & Hughes 2000.
Kato, Nakai & Shindo 2009.
Ceglia et al. 2015.
Schibille, Sterrett-Krause & Freestone
2016.
141 Bertini, Henderson & Chenery 2020.
142 Henderson et al. 2016.
143 Henderson, Ma & Evans 2020.
adding opacifying compounds – and this is also
apparently true for the single translucent blue
glass sample from Maastricht. Due to the high
level of contamination of both translucent
glasses in the crucible fragments from Maastricht
the results have not been included in Fig. 2.2.
The Susteren glasses plot with 8th and
9th century Dorestad glass having Na/Si oxide
ratios of above 0.24. Nevertheless, Levantine
glasses can also plot within the area occupied by
glasses with ‘Foy 2’ and ‘intermediate’
compositions136 so as noted above a better
discrimination is needed; an attempt at this is
given in Fig. 2.2. It appears that none of the
Susteren glasses are pristine (unrecycled)
Levantine I, Levantine II137 or N1 Levantine glasses
as defined by Kato et al.138 which are
manufactured at primary production sites. The
compositions for glass beads 1 and 6 as well as
vessels 21, 23, 24 and 27 from Susteren are
possible contenders for Levantine glasses but all
contain higher levels of MgO, P2O5, K2O, TiO2,
MnO and Fe2O3 than pristine natron glasses. This
is also true of all other natron glasses from the
site. These elevated levels of impurities show
that almost all of the Susteren natron glasses
have been recycled, perhaps several times.
A nearly colourless funnel beaker from
Susteren with a dark blue incalmo rim (no. 26) is
the only example of an HIMT composition. It has
the highest Fe2O3 (1.34%), TiO2 (0.4%) and MnO
(1.79%) levels of all the Susteren samples. It also
contains a low calcium oxide level of 6.2%,
another characteristic of HIMT glass. All other
Susteren natron glass is of the
‘intermediate’/‘Foy 2’ type with higher calcium
oxide and elevated levels of MgO, P2O5, K2O,
TiO2, MnO and Fe2O3 compared to pristine
Levantine I and other natron glasses.
The Susteren glass plots close to Jalame
(Levantine I) on a major/minor component plot
of Na/Si oxides versus Ca/Al oxides with
relatively high Na/Si oxide values of above 0.24.
However it is clear that when sand impurities are
plotted in Fig. 2.2 apart from the HIMT sample
they fall into the same plotted area as ‘Foy 2’
compositions as defined by Ceglia et al.139 and
Schibille et al.140 and ‘intermediate’ glasses from
Comacchio as defined by Bertini and
colleagues.141
In Fig. 2.1 Dorestad ‘8th’ century (pre-750)
glasses are plotted separately from Dorestad
‘9th’ century (750–850 AD) glasses. Almost all
‘8th’ century glasses have CaO/Al2O3 values of
between 2.5 and 3.09 or close to this range,
whereas ‘9th’ century (750–850 AD) glasses have
a far wider range of such values, of between 1.86
and 3.5. This suggests that raw materials with a
wider compositional range were used and/or
that mixing/recycling introduced a wider
compositional range in the glasses that date to
before 750. On this basis all Dorestad glasses can
be classified as ‘intermediate’/‘Foy 2’.
Figure 2.2 (from which Dorestad outliers
have been removed) on the other hand provides
further interesting insights into the Dorestad
glasses. Almost all natron glasses are of the
‘intermediate’/‘Foy 2’ composition according to
this plot. However, four Dorestad glasses have
low Fe/Ti oxide values of between 0.48 and 4.63
and this distinguishes them from the vast
majority of early medieval European
‘intermediate’/‘Foy 2’ glass and extends the
values of such glass. All ‘8th’ century Dorestad
glasses have Fe/Ti ratios of above 5. This
increasingly wide range of Fe/Ti oxide values in
early medieval European glass deserves to be
revisited in more detail using trace element
analysis. The single Dorestad glass that plots in
the ‘HIMT’ area is in fact a plant ash glass with
low Al2O3 (1.9%). This however raises the issue of
whether the silica sources used to make plant
ash glasses would benefit from such an
approach. Other trace element ratios have been
used to investigate Islamic plant ash glasses,142 as
have radiogenic isotopes.143
Rijnsburg glass tends to have lower Na/Si
oxide values than most translucent early
medieval glass (Fig. 2.1) which is intriguing
because it suggests that the base glass could be
pristine (unrecycled) Levantine glass, though this
would still need to be confirmed with trace
element analysis. If this is the case it would show
that relatively pure glass was imported and used
as the base glass for making the opaque
Rijnsburg material. Three glasses with Na/Si
values above 2.2 are (based on Fig. 2.1) probably
of an ‘intermediate’/Foy composition.
As can be seen in Fig. 2.2 six Rijnsburg
glasses plot in the area of Fe/Al versus Fe/Ti
oxide values that is consistent with being
pristine Levantine glasses too, so this
substantiates the above suggestion. As noted
above the use of ratios allows a direct
comparison with translucent glasses even though
the opaque glasses contain high levels of PbO
27
—
and Sn. This is an unexpected and intriguing
result which indicates how important further
research using trace element and isotopic analysis
would be.
The two Rijnsburg glasses that have been
badly contaminated when being worked, as well
as the material attached to the Rijnsburg
crucible, and glasses with very high Fe levels
have not been included because, for different
reasons, they cannot be compared with the
other plotted glasses with any validity.
It is impossible to plot Lent data on Fig. 2.1
because no soda was quoted. An attempt to
plot the data for Lent in Fig. 2.2 was made and,
where data was available, all plotted well away
from the other data. This suggests that the
quality of the data is too low to be considered
further here.
A small number of early Islamic plant ash
glasses were used to make early medieval beads
and vessels. The glasses have elevated MgO and
K2O concentrations and are quite distinct from
natron glasses as discussed above. A plot of
MgO versus CaO in relation to Levantine,
northern Syrian and Iraqi/Iranian early Islamic
glasses (not given here)144 suggests that glasses
from Wijnaldum and Susteren ultimately derive
from Iraq or Iran, that a second glass from
Susteren probably derives from Syria and a
single glass from Dorestad probably derives
from the Levant. Their provenances using trace
element analysis are discussed in more detail in
Chapter 5.
As already discussed, the bulk of Dutch early
medieval glasses are of the ‘intermediate’/‘Foy 2’
composition. Given the range of impurities in
these glasses they have undergone potentially
multiple episodes of recycling and mixing.
Most glasses of these highly variable compositions
date to, at the earliest, the 8th century and, using
major and minor chemical analysis, there is no
clear development or change in their chemical
compositions over time.
However, a number of observations can be
made when the compositions of glass from
different early medieval sites are compared.
Both Rijnsburg and Wijnaldum glasses appear
to include more examples of imported pristine
glass than the later sites of Susteren and
Dorestad. At Rijnsburg such glass would have
been used as the base glass for making beads.
However, currently there is no actual evidence
for mixing colourants with pristine glass there.
Based on these results no pristine glass was
detected at Susteren or Dorestad and only a
single probable example has been detected
amongst the Maastricht glass. This suggests that
there may have been a more direct supply of
pristine glass c. 600 AD to some sites where the
glasses were coloured/opacified (and possibly
made into beads or vessels on the same sites)
than for the Foy 2 glasses where a number of
recycling events would probably indicate that more
intermediate sites were involved.
Moreover, the variation in composition of
natron glasses as defined by major oxides (Na2O,
SiO2, CaO) as well as Al2O3 appears to decrease
with time when glasses dating to before
c. 750 AD (‘8th century’ in Fig. 2.1) are compared
with those that date to after c. 750 AD (‘9th century’
in Fig. 2.1). All of these observations would
benefit from further scientific analyses using
more sensitive techniques of chemical analysis
and also isotopic analysis (see Chapter 5).
Matthes145 has suggested that there was
an increase in the levels of As and Sb in opaque
yellow glass over time from c. 600 AD.
The implication could be that the lead sources
changed over time. However significant levels
of As have not been detected in Dutch early
medieval opaque yellow glass using electron
microprobe analysis and the increase in Sb levels
might relate to increasing levels over time in the
base glass used (see Chapter 5).
Early Islamic plant ash glass was introduced
in the early 9th century so we can be sure that
these few early medieval examples date to this
time, or later. Further analyses of early medieval
Dutch glasses will certainly reveal more examples
and it may eventually be possible to suggest the
key centres that such glasses and glass artefacts
were imported from.
Based on this provisional review we can
suggest that glass used to make early medieval
vessel, bead and window glass found in northwestern Europe was mainly imported and
reworked there. Primary production for natron
glass occurred in the Levant and Egypt where
coastal sand and natron from Wadi el Natrun
were fused in large tank furnaces. Glass furnaces
have been found on a number of early medieval
northwestern European sites but all of these
show evidence of secondary production (glass
working). It is conceivable that glass was reheated
and blown into characteristic early medieval
vessel types such as bowls and beakers or made
144 Henderson et al. 2016.
145 Matthes 1998, fig. 10.
28
—
146 Foy 2003; Ceglia et al. 2015; 2019;
Schibille, Sterrett-Krause & Freestone
2016; Bertini, Henderson & Chenery
2020.
147 Henderson 2012; Henderson 2023.
into beads and windows; the weight of evidence
suggests vessels were made in Germany.
The exceptions are the production of an opaque
yellow material (originally suggested to be
glass), especially at Maastricht and, later, when
the first wood ash glass was made from about
800 AD. It is possible that mixed-alkali glass was
made at Méru by adding wood ash to natron
glass but a more likely technique was to mix the
two types of glasses.
Although most early medieval northwestern
European glass is natron glass only small
numbers of pristine (unrecycled) Levantine and
Egyptian glass occur. Most glass that has been
found dating to between the 4th and 10th centuries
is HIMT (characterized by high levels of iron,
manganese, titanium and zirconium) and its
recycled variants. Based solely on the major and
minor elemental analysis of early medieval
Dutch natron glass, most appear to fall into the
recycled Foy 2 compositional types.146 These
Dutch early medieval glasses typically contain
elevated Fe2O3, TiO2, MnO and also some elevated
MgO, P2O5 and K2O.147 A much more detailed
consideration of the probable origins of early
medieval Dutch glass using trace element
analysis is given in Chapter 5 of this book.
Concluding remark
Glass from the earlier periods, between 450–550
and 550–650 AD contain somewhat more
examples of unrecycled (‘pristine’) glasses and a
smaller number of recycled glasses. By the time
of the Carolingian empire from about 750 AD
almost all the glass is recycled as exemplified by
glass found at Dorestad. Exceptions are Egyptian
II glass from Hedeby and Cologne. By 800 AD
primary glass production of wood ash glass
occurs in northwestern Europe and, as noted
above, at a time of technological transition,
the production of mixed-alkali glasses occurred.
Plant ash glass was imported from western Asia
in the form of beads and raw glass; it was used
in the manufacture of early medieval European
beads and occasionally glass vessels.
29
—
3 Evidence for early medieval glassworking in the Netherlands
Introduction
ISIA
N FR
TER
S
EA
R
T
OO
W
E
N
O
E
ST
R
ers
STER
GO
Lauw
S
L FRISIA
TRA
N
CE
H
Wierum
GO
Wijnaldum
BloemendaalGroot Olmen
Almere
Deventer
ValkenburgDe Woerd
Leidsche
Rijn
Utrecht
Wijk bij Duurstede
(Dorestad)
AA
SL
AN
D
Den Haag - Frankenslag
Rijnsburg
Oegstgeest
IJss
el
RIJ
NL
AN
D
WE
STE
RN FR
ISIA
K E NNEMERLAND
TE
XE
ie
L
Vl
M
Rh
ine
Ma
as
ine
Rh
Susteren
t
eld
Sch
Maas
Glass production can be divided into two
different stages. The first is primary production
involving the fusion of raw materials (see Chapter
2). The second is remelting the glass made in the
first stage to make glass objects. There is no
recorded evidence for primary glass production
in the early medieval Netherlands. There is,
however, evidence for secondary production,
not just the remelting of glass but also the
manufacture of pigments for glass coloration.
In the Merovingian period, comprehensive
evidence for glass bead making has been found
at Maastricht-Jodenstraat, Rijnsburg-Abdijterrein
and Wijnaldum (fig. 3.1). The evidence falls
between the last quarter of the 6th and first half
of the 7th century. As Pion has shown, this period
coincides with a steep reduction of imported
oriental beads in the west, giving birth to a new
production of highly coloured Merovingian bead
types.148 Maastricht, developed from an old
Roman town was one of several production
centres in the middle Meuse valley149, well within
the borders of the Merovingian empire. On the
other hand, the production of popular Merovingian
bead types in the central places150 of Rijnsburg in
the Rhine delta and the northern terp site of
Wijnaldum shows that the demand for
fashionable beads reached well beyond the
borders of the Merovingian empire (fig. 3.1).
There is scantier evidence for glass production
in the Carolingian period in the form of waste,
perhaps in part because the custom of wearing
colourful beads became less popular in the
Frankish heartlands151, though they were still
being worn in the northern periphery of the
Carolingian empire. Imports of Islamic period
beads from the end of the 8th century onwards152
in settlements along the Rhine and in cemeteries
north of the Rhine, traded through Viking trade
networks to the west, may also have contributed
to a reduction in Frankish glass bead production.
Another notable difference with the preceding
period is that that the evidence for glass working
derives from a more varied range of site types
and now includes an early emporium (Dorestad)
and ecclesiastical centres (Utrecht-Domplein,
Susteren) (fig. 3.1). From the Carolingian period,
not all phases of the production processes are
E
A
3.1
Maastricht
Frisian inhabited area
Holocene coastal area
0
50km
Pleistocene hinterland
Fig. 3.1 Sites in the Netherlands with evidence for glass working in the early medieval period
(adapted from Dijkstra 2011, 12).
represented by the glass working waste, so it is
more difficult to attribute what evidence there is
to specific products. Glass beads were probably
made at Dorestad, while possible evidence for
the production of the earliest highly coloured
stained window glass was found at the early
medieval monastery of Susteren.
In the following the sites are considered starting
with southern sites and moving northwards.
148
149
150
151
152
Pion 2014.
Van Wersch 2012.
Nicolay 2017.
Delvaux 2017.
Sablerolles & Van der Linde-Louvenberg
2019 (Leiderdorp-Plantage); Langbroek
2021b (Dorestad); Van Es & Schoen
2007/2008 (Zweeloo cemetery).
30
—
3.2
Maastricht, Limburg Province
Excavations carried out in Maastricht in the
1980s and early 1990s yielded evidence of an
impressive range of craft activities carried out
during the 6th and 7th centuries. The location of
Maastricht is shown in fig. 3.1. In total, nine
different sites were identified in quite different
parts of the settlement.153 Apart from one, these
were all located on the west bank of the Meuse.
Here a Roman castellum had been constructed
in the 4th century which developed into an
important political, religious and economic
centre in the Middle Meuse region during the
Merovingian period. Maastricht had its own
bishop, see and mint. Gold coins minted in
Maastricht have been found all over northwestern Europe, with a significant concentration
in the northern Netherlands, especially in
Friesland Province.154
Evidence of glass-working was found at six
locations, the Mabro site (Onze Lieve Vrouweplein),
the Derlon site, the Jodenstraat site, the
Rijksarchief site, the Boschstraat area and the
Lage Kanaaldijk site.155 On all of these sites glassworking was combined with one or two other
high-temperature crafts (table 3.1), namely
pottery production (Lage Kanaaldijk), ironworking (Boschstraat, Rijksarchief), copper-alloy
working (Derlon, Boschstraat, Jodenstraat) and
possible gold-working (Mabro). Antler-working
took place at the Mabro site, Derlon site,
Boschstraat area and Rijksarchief site, while
waste from amber-working was mixed in
with glass bead production waste from the
Jodenstraat site.
It is likely that glassworkers worked closely with
other craftsmen and shared knowledge about
furnace technology, fuel and other materials. For
instance, the glass crucibles used by the
glassworkers on the Jodenstraat, Mabro and
Rijksarchief sites were all repurposed ovoid
coarse-ware cooking pots (Wölbwandtöpfe) which
were most probably made in Maastricht. At the
Céramique site on the east bank of the river
Meuse, four cross-draught kilns were found
filled with wasters, dominated by coarse-ware
Wölbwandtöpfe (fig. 3.2).156 Remarkably, these
pots are also used as glass crucibles in Merovingian
Rijnsburg (see Section 3.8) and
Valkenburg-De Woerd (see Section 3.9) and
there is continuity in the reuse of Wölbwandtöpfe
as glass crucibles into the Carolingian period
(globular pots type Dorestad W III) (see
discussion by Menno Dijkstra, Section 3.5.1)
1:4
0
20cm
Fig. 3.2 Maastricht-Céramique site: A restored cooking
pot (Wölbwandtopf ) from the fill of a furnace
(Photograph: Wim Dijkman).
Table 3.1 Maastricht: Early medieval sites with evidence for glass-working in
combination with other crafts.
Maastricht sites
High-temperature crafts
glass
Derlon
X
pottery
Other crafts
iron
copper-alloy
gold?
X
Mabro
X
Jodenstraat
X
153 Dijkman 2013.
154 Pol 1999, fig. 7.
155 Sablerolles, Henderson & Dijkman 1997,
Boschstraat area
X
X
Rijksarchief
X
X
295, fig. 1; Dijkman 2013, fig. 1.
156 Dijkman 2013, fig. 4.
Lage Kanaaldijk
X
antler
X
X
X
X
X
amber
X
X
X
X
31
—
3.2.1 Maastricht-Jodenstraat (MAJO)
Site:
1
Site type:
specialist craft zone
Province:
Limburg
Municipality: Maastricht
Place:
Maastricht
Toponym: Jodenstraat 30
Start date 580/590 (late 6th century)
End date:
610/620 (early 7th century)
Description: In 1988 excavations by the
Gemeentelijk Oudheidkundig Bodemonderzoek
Maastricht (GOBM), the Municipal
Archaeological Service of Maastricht, took place
at Jodenstraat 30.157 The site was situated north
of the late Roman castellum, near the Via Belgica
which connected Tongeren and Maastricht.
A rubbish pit filled with the debris from glass
bead making was found.158 The pit also contained
waste from copper-alloy-working and amberworking.159 Based on the pottery finds, the pit
was filled sometime in the late 6th to early
7th century. To date, the glass assemblage
represents the most comprehensive evidence for
6th–7th century glass bead making in Europe.
Glass production waste: The debris consisted of
750 glass objects which represent the full range
of waste from glass bead production.
The production waste was divided into eight
main groups: glass rods (n=369), ‘punty’ glass
from a beadmaker’s tool (n=36), glass threads
with and without tweezer marks (n=17), glass
drops (n=39), finished and failed beads (n=123),
crucibles (n=38, EMN=17), cullet or scrap glass
(n=20), glassy slags/fuel ash slags (n=53) and
non-diagnostic fragments (n=55) which include
(small lumps of) melted glass and fragments
that are too small to classify. All waste categories
are dominated by opaque yellow glass (apart
from scrap glass and glassy slags). Opaque
yellow, green, red, white, red and blue make up
57%, 17%, 16%, 7% and 0.9% of colours
respectively. The balance are translucent light
green (2%) and blue-green (0.6%), and
transparent colourless (0.2%) of glass.
Almost all beads are wound and have
tapering perforations showing they were made
by winding melted glass around a mandrel,
a bead-making tool with a conical point. Such a
tool may have been found at the Rijksarchief site
(see Section 3.2.3). A few beads were made by
perforating a section of a ‘composite’ glass rod
which was made by fusing strands or slender
157 The site remains unpublished, apart
Fig. 3.3 Maastricht-Jodenstraat site: From left to right: opaque yellow melted rods ends and threads with tweezer
marks, finished pentagonal beads, drawn and composite rods, failed, cracked beads (Photograph: Gemeentelijk
Oudheidkundig Bodemonderzoek Maastricht (GOBM)).
from the rubbish pit with production
waste.
158 Sablerolles, Henderson & Dijkman 1997.
159 For waste from amber-working, see
Dijkman 2013.
32
—
rods together in order to make a glass rod with
a large enough diameter for making a bead.
The vast majority of the beads (79%) are
cylindrical in shape, predominantly with a
pentagonal section (76%), while some have a
square (2%) or round (1%) section. The polygonal
beads were believed to have been shaped with
tongs but are much more likely to have been
shaped by hand with a small wooden tool known
as a paddle. The remainder of the beads are
made up of small globular beads of opaque
yellow glass and medium-sized bi-conical beads.
The majority of the beads are split in half along
the length of the perforation (72%) (fig. 3.3).
This is a commonly observed phenomenon on
bead-making sites (see Section 3.8 and Section
3.12) and occurs when the glass is overheated
during the making of the beads, or when the
beads have not been annealed or cooled down
properly, causing the beads to crack.160
Crucibles are represented by 38 fragments from
at least 17 coarse-ware cooking pots
(Wölbwandtöpfe). In 15 cases, only the lower
halves of the pots were used to melt what was
originally believed to be highly coloured opaque
yellow glass (see Section 5.2.3). More recently, a
crucible base with an opaque white deposit was
identified by one of the authors of the 1997
160 Gam 1990; Heaser 2018.
161 Dijkman 2013.
162 Cf. Sode 2004; Risom 2013, 56–57, 60.
publication (fig. 3.4).161 Table 4.2 lists further
selected crucibles and associated glassworking
debris from Jodenstraat (MAJO 1-27); images
are in Appendix IV (figures appendix IV.11-38).
No crucibles with opaque red, green and blue
glass have been found.
Once made, opaque glass would have been
worked into beads, for instance by gathering
some melted glass onto a solid metal rod or
punty and winding it around a mandrel, a beadmaking tool with a conical point. However, the
presence of opaque yellow and white glass rods
suggests that these were an essential phase in
the bead-making process.
The rods were made by attaching two metal
rods (punties) to a gather of melted glass, then
pulling the glass in opposite directions creating
a drawn glass rod of several metres’ length with
a more or less round cross-section The long
glass rods could have been fragmented into
shorter sections, as suggested by the presence of
a particular type of glass rod breaking splinter.
Short rod sections could have been pre-heated,
picked up with a punty, as indicated by the
presence of punty glass among the waste
products, and wound around a mandrel.162
Alternatively, a production technique in use by
modern beadmakers may have been used.
Fig. 3.4 Maastricht-Jodenstraat site: Crucible base with opaque white glass, drawn and twisted opaque white rods
and cracked, failed beads (Photograph: Gemeentelijk Oudheidkundig Bodemonderzoek Maastricht (GOBM)).
33
—
Table 3.2 Maastricht, Jodenstraat (MAJO): selected artefacts and photo numbers.
Find number
Fragment
Glass colour(s)
Sample
1-1-7-4
crucible base
opaque yellow
MAJO 1
1-1-5-17
crucible body
deep translucent and yellow
MAJO 2
21
figure appendix IV.12
1-1-7-11
crucible base
opaque yellow
MAJO 3
22
figure appendix IV.13
1-1-7-3
crucible base
opaque yellow
MAJO 4
23
figure appendix IV.14
1-1-7-2
crucible base
opaque yellow
MAJO 5 (inside) &
MAJO 5 (outside)
24
figure appendix IV.15 (inside) &
figure appendix IV.16 (outside)
1-1-7-8a
crucible base
opaque yellow
MAJO 6
25
figure appendix IV.17
1-1-7-8b
crucible base
opaque yellow
MAJO 7
26
figure appendix IV.18
1-1-7-8c
crucible base
dark translucent
MAJO 8
29
figure appendix IV.19
1-2-3
possible brick fragment
white and opaque yellow
MAJO 9
30
figure appendix IV.20
1-1-7-451 a and b
fragments
blue
MAJO 10
39-40
figure appendix IV.21
1-1-7-21-22a
scrap
red
MAJO 11
41
figure appendix IV.22
1-1-7-21-22b
scrap
red
MAJO 12
42
figure appendix IV.23
1-17-28
scrap
green
MAJO 13
43
figure appendix IV.24
1-2-3-VG3-2
window
yellow-green
MAJO 14
44
figure appendix IV.25
1-2-3-VG3-1
window
yellow-green
MAJO 15
45
figure appendix IV.26
1-2-3-VG3-3
window
pale yellow-green
MAJO 16
46
figure appendix IV.27
1-1-7-500
thin rod
green
MAJO 17
47
figure appendix IV.28
1-1-7-431
drop
yellow weathered
MAJO 18
50
figure appendix IV.29
1-1-5-388
drop
red
MAJO 19
51
figure appendix IV.30
1-1-7-463
drop
dark green
MAJO 20
52
figure appendix IV.31
1-1-7-462
pulled rod
milky blue
MAJO 21
53
figure appendix IV.32
1-1-7-583-594
thin rod
red
MAJO 22
54
figure appendix IV.33
296
twisted rod
opaque white
MAJO 23
58
figure appendix IV.34
1-2-5a
beaker base
green
MAJO 24
60
figure appendix IV.35
1-2-5-b
punty glass
blue
MAJO 25
61
figure appendix IV.36
1-1-7-503
ribbed flat
blue-green
MAJO 26
68
figure appendix IV.37
309-349
rod fragments
green
MAJO 27
73-74
figure appendix IV.38
This involves heating up the end of a longer
section of a glass rod, attaching the hot glass
to a mandrel and winding it directly around
a mandrel.
According to experimental glass beadmaker
Sue Heaser, finds of glass rods of about 5–10 mm
thick on many early medieval bead-making sites
suggest that that beadmakers in the first
millennium AD knew the technique.163 She states
that this method gives more control than using
a gather of molten glass attached to a punty. She
goes on to explain that only the end of the glass
rod is heated to liquid point so that the rest of
the rod remains cool and rigid. The cool end is
used as a handle by the beadmaker to control
the hot glass at the other end as it is applied to
Sample number
20
the mandrel. This could be another explanation
for the presence of ‘composite’ glass rods at the
Jodenstraat site made up of lots of strands or
slender rods which are fused or twisted together
(fig. 3.3 and fig. 3.4). Modern beadmakers use
these leftover bits to fuse onto the end of glass
rods when they have become too short to
handle.164 Several examples of rods with melted
ends (fig. 3.3) among the waste material from
the pit could have been destined for this kind
of recycling by melting the ends and pressing
two of the same colours together.165
Chemical analyses published in 1997
suggested that the lead-tin-yellow crucible
residues has the same composition as the
opaque yellow glass rods and other waste
Photo number
figure appendix IV.11
163 Heaser 2018.
164 Personal comment Ingrid Pears, Hot
Glass Studio, Thoresby, Notts., UK.
165 Cf. Heaser 2018.
34
—
products, including (failed) beads. Moreover, the
chemical compositions of opaque white, red and
blue-green glass rods was also linked to waste
products of corresponding colours. However,
only trace analysis can give incontrovertible
evidence that this is indeed the case (see Section
4.4). No crucibles fragments bearing opaque red,
blue-green or blue glass were found and it was
believed that these must have been imported in
the form of glass rods that were directly worked
into beads. In that case, the base glass used for
the presumedly imported glass colours is likely
to be different from the base glass used for the
locally made opaque yellow and white glasses.
A more detailed scientific investigation looking
into this matter is given in Chapter 5.
Fragments of the rim, (upper) bodies and
base of two crucibles show that complete pots
were used to melt colourless glass and translucent
blue-green glass. Drops of translucent greenish
glass amongst the waste products suggest this
glass was worked on or near the site (fig. 3.5).
Fragments of translucent greenish scrap glass
– old vessel glass and the earliest early medieval
grozed window glass from the Netherlands –
were interpreted as cullet destined to be
remelted to create this type of base glass.
Further detailed investigations of the base and
166 Callmer & Henderson 1991, fig. 2.
167 Andersen & Sode 2010, 32, fig. 11, 34,
table 11.
168 Preiß 2010, nr 38, fig. 10; Sablerolles &
Henderson 2012, afb. 6.19, NO 5041.
scrap glass using trace element analysis are
given in Chapter 5.
Not included in the 1997 publication are a couple
of fragments of Roman faience melon beads and
a handful of very small fragments of crushed
translucent dark blue glass (fig. 3.5) as well as an
almond-shaped bead of translucent dark blue
glass. Although it is impossible to establish from
what kind of object(s) the translucent blue
crushed fragments come from, it is tempting to
suggest they could derive from early planoconvex glass ‘cakes’ of translucent dark blue
glass which are found on later 8th–9th century
bead-making sites in southern Scandinavia,
such as Åhus (Sweden)166 and Ribe (Denmark)167
and at Dorestad168 (see Section 3.4.2). With this in
mind, it will be interesting to see whether the
translucent dark blue crushed fragments can be
chemically linked to a translucent dark blue bead
found among the waste. Moreover, could this
translucent dark blue glass have been used as a
base glass to make opaque blue glass on site
given there are three beads and a drop of
opaque blue glass as well as an opaque greyish
blue glass rod. This will be further discussed
in Chapter 5.
Fig. 3.5 Maastricht-Jodenstraat site: From left to right: fragments of Roman faience melon beads, crushed
translucent dark blue glass and yellowish green window and vessel glass of translucent greenish glass with glass
drops in matching colours (Photograph: Gemeentelijk Oudheidkundig Bodemonderzoek Maastricht (GOBM)).
35
—
3.2.2 Maastricht-Mabro
Site:
2
Site type:
dump zone
Province:
Limburg
Municipality: Maastricht
Place:
Maastricht
Toponym: Mabro
Start date: 500 (6th century)
End date:
700 (7th century)
Description: Excavations carried out in 1981 by
the GOBM or Municipal Archaeological Service
of Maastricht at the site of the Maastrichtse
Broodfabriek (Mabro) or the Maastricht Bread
Factory, located at the Onze Lieve Vrouweplein
(MAVP) 16–18 remain unpublished.
Glass production waste: The site produced
twelve fragments of glass crucibles, a substantial
number of beads, glassy slags, one of which
might be a fragment of melted furnace wall,
and at least one red-burnt fragment of clay
covered with translucent greenish glass.169
Eleven crucible fragments can be dated to
the 6th or 7th centuries (Wölbwandtöpfe), of which
ten were made available for sampling (table 3.3).
Additionally, one rim fragment (find no. 1-5-OA)
is of a late 4th–early 5th century bowl (type Alzey),
perhaps recovered from a late Roman grave and
reused in the Merovingian period. Chemical
analysis of the latter was included in the 1997
Maastricht-Jodenstraat publication.170
The fragment is covered with translucent
greenish glass below the rim and a vitrified,
off-white granular material on the inside and
outside of the rim. At the time, this was
tentatively interpreted as overheated frit. Frit
is a half-product of glass-making, so this could
constitute the earliest evidence of glass-making
in the early medieval west. This material has
now been re-analysed using scanning electron
microscopy and the results are presented in
Section 5.2.
Two rim fragments of Merovingian crucibles
show similar deposits (table 3.3). As is the case
on the Jodenstraat site, the crucibles either
contain colourless or translucent natural green
glass or opaque yellow glass; no crucible fragments
with other colours were found.
Table 3.3 lists ten crucible fragments
together with selected images (MABRO 1-10)
from Maasricht-Mabro and the sample numbers
used in scientific analysis. The images are at
Appendix IV (figures appendix IV.1-10).
Judging from photographs of the beads, they are
all made by winding and are either monochrome
or decorated with trails in contrasting colours.171
Using the bead typology developed by Pion in
2014 for beads from six Merovingian cemeteries
in Belgium, later adapted by Vrielynck, Mathis
and Pion,172 the Mabro beads cover a long period
between 480–530 (P1) and 620–670 (P5),
but they mostly date in period 560–610 (P3)
and period 610–640 (P4) (table 3.4).
This site has not been published, so it is
impossible to state which beads are likely to be
local products, but given the dates for the
crucibles, those beads dating to roughly the
Table 3.3 Maastricht-Mabro: Selected crucibles and their glassy contents, together with
photo numbers and sample numbers.
Find number
Fragment
Glass colour(s)
Sample
Photo number
03-04-2000
rim
white: frit-like
MABRO 1
figure appendix IV.1
3-OA-55
body
colourless/white: frit-like
MABRO 2
figure appendix IV.2
01-03-1951
body?
translucent green
MABRO 3
figure appendix IV.3
3-OA-1
rim
colourless
MABRO 4
figure appendix IV.4
1-5-OA
rim
colourless/white: frit?
MABRO 5
figure appendix IV.5
3-OA-40 (= 3-AA-40)
base
opaque yellow
MABRO 6
figure appendix IV.6
02-02-2018
base (burnt clay?)
colourless/pale green
MABRO 7
figure appendix IV.7
03-05-2024
base
opaque yellow/green
MABRO 8
figure appendix IV.8
03-05-2024
body?
deep translucent
MABRO 9
figure appendix IV.9
03-04-2012
base
green
MABRO 10
figure appendix IV.10
169 Information and photographs of the
finds were kindly provided by Wim
Dijkman, Senior Conservator
Archeologie en Erfgoed, Team
Programma en Innovatie, Centre
Céramique – Kumulus –
Natuurhistorisch Museum.
170 Sablerolles, Henderson & Dijkman 1997,
307–308, pl. 25, 1.
171 With many thanks to Wim Dijkman for
providing the photographs.
172 Pion 2014; Vrielynck, Mathis & Pion
2018.
36
—
Table 3.4 Maastricht-Mabro: Typology and bead periods (Pion 2014, Vrielynck, Mathis & Pion 2018).
Find number
Form
Colour
Decoration
Typology
Period
01-02-1963
globular
opaque white
translucent light blue crossing trails
B3.3-3a
P4
02-01-2000
cylindrical, round section?
opaque yellow
-
B1.4-1a?
P3?
03-04-2025
short-cylindrical
opaque red
-
B1.4-2a
P3
03-05-1932
bi-globular
opaque yellow
-
B1.2-1b
P4
03-05-1933
globular, medium
opaque white
-
B1.1-4a
P1-5
03-06-2004
globular, small
opaque yellow
-
B1.1-2a
P2-5
3-OA-0
disc
opaque white
translucent dark blue crossing trails
B3.3-2a
P3
3-OA-40.1
cube
opaque red
opaque yellow borders & crosses
B11.1-5
P2
3-OA-52
cylinder, square section
opaque red
opaque yellow dots
B6-2-1b
P5
Period: P1 = 480-530 AD; P2 = 530-560 AD; P3=560-610 AD; P4=610-640 AD and P5=620-670 AD.
2nd half of the 6th and first half of the 7th century
are the most likely candidates: a small globular
bead, a bi-globular bead and a possible
cylindrical bead of opaque yellow glass,
a medium-sized globular bead of opaque white
glass and a short cylindrical bead of opaque red
glass. There are two polychrome, trailed beads:
an opaque white globular bead with translucent
light blue narrow crossing trails and an opaque
white disc-shaped bead with translucent dark
blue crossing trails.
3.2.3 Maastricht-Rijksarchief
Site:
3
Site type:
mixed crafts zone
Province:
Limburg
Municipality: Maastricht
Place:
Maastricht
Toponym: Rijksarchief
Start date: 480–490 (late 5th century)
End date:
600?
Description: Excavations carried out in 1990–1991
by the Municipal Archaeological Service of
Maastricht (GOBM) are briefly discussed by
Hulst.173 The site was situated in the middle of
the delta of the river Jeker, south of the
4th century Roman castellum and not far from
the old Roman road which connected Tongeren,
the Roman capital of the civitas Tungrorum,
to the new centre in the region, Maastricht.
Dozens of rubbish pits were found in this area
containing (late) 5th and 6th century pottery, Roman
tiles, chunks of local sandstone (‘kolenzandsteen’)
and flint, waste from glass-working, antlerworking and iron-working. No traces of buildings
were found, apart from three scattered postholes
and remains of a small ditch.
Glass production waste: Hulst174 lists 55 fragments of
early Merovingian vessel glass, two glass rods,
(fragments of) 25 beads, drops of glass, melted
glass and small, dark vitreous spheres
(‘glasbolletjes’) which may also be linked to ironworking on the site. Furthermore, one fragment
of a glass crucible, glassy slags and vitrified
fragments of (a) furnace floor with glassy slags
and iron slags adhering to them were found.
One pit contained traces of firing and may have
been the firing pit of a dismantled furnace. A
truly remarkable find is that of a forged iron rod
which is round in section at one end and square
in section at the other (fig. 3.6). Hulst remarks
that the perforations of the beads match the
circumferences of the iron tool. The rod is
interpreted by Hulst as a bead-making tool or
‘mandrel’, a forged iron rod with a conical point
on which beads are formed by winding a glass
thread around it.
1:1
0
173 Hulst 1992.
174 Hulst 1992.
5cm
Fig. 3.6 Maastricht-Rijksarchief site: Forged iron mandrel, square-sectioned, with a round-sectioned conical point
(Photograph: Wim Dijkman).
37
—
The opposite, square-sectioned end of the rod
was probably originally inserted into a wooden
handle.175 A similar tool from a bead-making site
at Paviken on the Baltic island of Gotland was
deemed too slender for an awl and has been
interpreted as a mandrel (fig. 3.7).176 The squaresectioned end is hollow and could have been
wrapped around a wooden handle. A mandrel
with the remains of a handle was found at RibeDommerhaven, where beads were made on a
very large scale in the 8th–9th centuries.177
According to Hulst it is obvious that beads were
made on or near this site sometime in the 6th
century. He mentions that the beads are mostly
made of monochrome glass, including blue,
brownish-red and yellow. Some beads are
decorated with glass trails of blue, white or
yellow glass. The beads include finished
specimens and failed beads. Among the failed
beads are examples where the winding of the
spiral glass trail around the mandrel had gone
wrong. Cracked specimens have been found.
This can occur after the glass has been
overheated or when the finished beads have not
been annealed properly. Both rods are
monochrome opaque brownish-red.
Recent photographs of some of the finds178
show that the scrap glass includes a thick-walled
fragment of early/mid-Roman blue-green glass
(probably window glass or a square bottle),
a yellow-green knocked-off rim of a late Roman
cup of Isings type 96a,179 a rim fragment of
a Merovingian bell beaker of yellowish-green
glass with dark inclusions and lots of bubbles,
decorated with vertical optic blown ribs and an
opaque white trail below the straight, firerounded rim. The latter is probably contemporary
with the bead-making. Among the beads, there
is a sub-biconical opaque red bead with an
opaque yellow spiral. This is the only specimen
which can be securely dated to Pion’s Period 3
(560–610 AD), while a seemingly dark/dirty
bi-globular bead with whitish deposits is probably
of Pion’s Period 4 (610–640 AD).180 A spiral bead of
translucent light greenish glass is very similar to
one found at Leidsche-Rijn Leeuwesteyn Noord
(Rijnfront) (see Section 3.6.2).
1:1
0
5cm
Fig. 3.7 Paviken, Gotland, Sweden: possible mandrel with a bead added for museum display
(Photograph: Matthew Delvaux).
3.3
Susteren-Salvatorplein, Limburg
Province
Site:
4
Site type:
Monastery
Province:
Limburg
Municipality: Echt-Susteren
Place:
Susteren
Toponym: Salvatorplein
Start date: 714
End date:
1802
Description: In the 1990s, excavations by the
former Archaeological State Service (ROB) took
place immediately north of the basilica of
St Amelberga, the Romanesque church which is
still standing today. These revealed remains of
a monastery which was inhabited from the
8th century until it was dissolved in 1802 during
the French occupation. The monastery was
almost completely demolished in the early
19th century. The results have recently been
published by Henk Stoepker.181
The foundation of the monastery and a
small church is recorded in a charter from 714
when Pepin II and his wife Plectrude donated
a small estate on the river Suestra to the AngloSaxon missionary Willibrord. The stream was
part of a larger drainage system in the Limburg
Meuse Valley. The excavations yielded only
scant evidence for late Merovingian habitation,
consisting of a timbered building with several
associated waste pits and an oven, a cistern and
a few graves. Very few portable finds were
recovered from this period.
The period of the late 8th century and
9th century sees an increase in habitation and
there were now two stone buildings, one of
which was circular and may have been a funeral
chapel, a timbered building, as well as the
above-mentioned cistern; to the east of the
buildings was a craft zone with a bell-casting pit
175 Heaser 2018.
176 Lundström 1981, 99–100, fig. 10:4. With
177
178
179
180
181
many thanks to Matthew Delvaux,
Princeton University, for providing the
reference and translating the Swedish
text into English.
Sode 2004, 86, fig. 3.
Information and photographs are
kindly provided by Wim Dijkman,
Senior Conservator Archeologie en
Erfgoed, Team Programma en
Innovatie, Centre Céramique – Kumulus
– Natuurhistorisch Museum.
Isings 1957, 113–114.
Bead type B5.2-2a, Period 3; Bead type
B1.2, Period 4 (Vrielynck, Mathis & Pion
2018).
Stoepker 2021.
38
—
182 Sablerolles 2023 (basispublicatie
chapter 29).
183 Pottery identification by Jan de Koning.
184 Stoepker 2021, 229, afb. 11.11, V12-053185
186
187
188
GL-10. Pottery identification by Jan de
Koning.
Henderson 2023 (basispublicatie
chapter 31).
Stoepker 2021, 231, table 11.1, V09-129GL-01, V04-194-GL-01.
Stoepker 2021., 230, afb. 11.12.
Freestone 2015,.
and five ovens. This development is mirrored by
an increase in the number of graves; these are
presumably associated with the early medieval
abbey church which is likely to be found beneath
the present-day Romanesque church. Finds of
highly coloured quarries are typical for ecclesiastical
contexts and are a testament that stained glass
windows were in use, most likely in the
abbey church.
A watercourse (complex 4300/4400) north
of the habitation zone was used as a refuse
dump from which many Carolingian period finds
were retrieved. As much as 85% of Carolingian
pottery was imported from the German Rhineland,
showing the monastery was firmly embedded
in the Rhenish trade system, despite the location
of the monastery on the Meuse. Wine glasses
such as (palm-)funnels probably came from
the same Rhenish production centres as
the ceramics.
No indications were found for animal
husbandry, although it can be assumed that
nearby farms on (a) monastic estate(s) would
have provided the monastery with animal
products. The over-representation of certain
skeletal elements of pigs shows these were
specifically imported for the consumption of
meat, one of the few signs of luxury enjoyed
in the monastery.
During the first half of the 10th century,
the habitation zone was cleared and a large
amount of settlement refuse, including remains
of buildings, was dumped in watercourse 4300,
perhaps as a result of a Viking raid in 881–882,
although no evidence was found to support this.
For the period of c. 900 to c. 1050 only one
wooden building, a well and graves are
discernible, and possibly some ovens. There are
significantly fewer portable finds than in the
previous century. The building of a new,
Romanesque church in the second half of the
11th century ushers in a new phase of the monastery,
characterized by the construction of stone-built
cloisters. The digging of watercourse 4200 in the
11th century, intersecting the early medieval
watercourse 4300/4400 caused a lot of early
medieval material to be redeposited in later
contexts.
Glass production waste: Five fragments of
glass production waste make it likely that glass
was worked in the monastery, probably during
the early medieval period.182 The finds consist of
two fragments of glass crucibles, a partially
melted Roman tessera, a glass fragment from
glassblowing tool and a possible fragment of
opaque yellow raw glass. The crucible fragments
derive from the same context in watercourse
4310 (800–1300 AD) which mostly contains
redeposited Carolingian material (60%) and
some 10th century (17%), 11th–12th century (22%)
and Iron Age/Roman period (1%) material.
One crucible fragment is probably made of
Carolingian Badorf ware183 and presumably
derives from the same type of cooking pot that
was used at Utrecht-Domplein (Dorestad type W
III) (see Section 3.5.1) and probably also at
Leidsche Rijn (see Section 3.6.2). Unfortunately,
the translucent light (bluish-)green natron glass
is too contaminated to be linked to either the
window or the vessel glass found at Susteren.
The other crucible fragment of possible grey
Meuse Valley ware is covered on the inside with
a thin layer of translucent dark blue natron glass
with a small area of colourless glass (fig. 3.8a).184
The chemical composition of the dark blue glass
in the crucible can be chemically linked to the
dark blue glass of two translucent dark blue
window quarries:185 an irregularly shaped quarry
from a grave dating to the late 10th/11th century
and a small triangular quarry from a context in
watercourse 4400 with predominantly high
medieval material with some early medieval
finds (8%).186 High concentrations of antimony
in all three glasses show that antimony-rich
Roman tesserae were mixed in to colour
the glass.
The find of a partially melted, opaque dark blue
Roman tessera (fig. 3.8b) could suggest that dark
blue glass for the production of window glass
was made in the monastery by adding blue
tesserae to a colourless base glass.187 The tessera
was found together with the above-mentioned
triangular dark blue quarry and can be either
high or early medieval. The number of artefacts
involved is small but the presence of the crucible
clearly shows that dark blue glass was being
worked. The practice of recycling Roman tesserae,
especially blue tesserae for colouring window
glass, was carried on into the 12th century.188
A thick-walled fragment of translucent
dark bluish-green glass is covered on the
concave inside with dark grey iron scale from a
glassblowing tool (fig. 3.8c). The fragment
comes from the intersection of high medieval
watercourse 4200 (11th–13th century) and early
39
—
such as a spiral trail. In the Carolingian period,
self-coloured spiral trails were especially popular
on the necks of globular jars. One bluish-green
vessel fragment from Susteren may derive from
such jars.
A fragment of opaque yellow glass from a
17th century cesspit, which also contains some
early medieval material, may be a raw glass
fragment struck from an early medieval glass
ingot. In northern France yellow glass was
worked in several monasteries during the 8th
and early 9th centuries to produce reticella wares,
especially bowls decorated with yellow spirals
and reticella rods.191 A body fragment of such a
bowl was also found at Susteren.192
a
b
c
3.4
0
1:1
Wijk bij Duurstede (Dorestad),
Utrecht Province
5cm
Fig. 3.8 Susteren-Abdijterrein: Glass-working production
waste: a) fragment of a crucible with translucent dark
blue glass on the inside (V12-053-GL-10); b) a partially
melted, opaque dark blue tessera (V04-194-GL-02); c)
a glass fragment from a gathering or bit iron (V12-078GL-01) (Photographs: Limburgs Museum, Venlo.
Drawing: SAGA Archeologie).
medieval watercourse 4300; the find context
contains 25% early medieval pottery. It is
possible that both crucible fragments and the
glass fragment from the gathering iron were
deposited closely together in watercourse 4310
and that the latter was redeposited when
watercourse 4200 was dug.189 The glass is well
preserved and is likely to be natron glass rather
than wood ash glass which makes an early
medieval date more likely.
Whether early or high medieval in date,
this fragment represents the only direct evidence
for glassblowing in the Netherlands since the
Roman period. The fragment is interpreted by
the archaeologist and experimental glassblower
Mark Taylor190 as glass that was broken off a
gathering iron or bit iron. A bit iron is a long,
thin iron rod with a flat end which is used by
glassblowers to add handles, feet or decorations
to a glass vessel. Since no early or high medieval
vessels with handles or added feet are known
from the Netherlands, the fragment therefore
most likely results from decorating a translucent
bluish-green vessel with a self-coloured decoration
An overview of glass-working waste from Dorestad
was published by Preiß in 2010.193 His inventory
counts 84 finds which are made up of a few old
finds without contexts as well as in situ finds
from excavations in the late 1960s and ’70s in
the harbour area (Hoogstraat excavations) and
the settlement on the river bank (vicus).
However, the majority of the finds (60%) were
retrieved during more recent excavations in the
agrarian settlement, the Parkeerplaats Albert
Heijn (PPAH) excavations in 1992–1993 and the
Veilingterrein excavations in 2007–2008,
probably due to wet-sieving.
The largest category of glass waste is
formed by tesserae (43%) which were recycled
on a large scale in the early medieval period.
They were melted down to make beads, especially
in Scandinavia, or to increase the volume of
glass batches intended for the production of
window and vessel glass.194 Preiß’s group of
deformed glass (29%) may also include accidentally
melted vessel fragments. The remainder is made
up of glass drops and threads (15%), glass lumps
(7%) and miscellaneous (6%).
A find worth mentioning in the last category
is that of a crucible which was found before
1978; it comes from the vicus on the river bank
and is now lost.195 Isings described the fragment
as follows: ‘Fragment of a crucible, pinkish grey
ceramic. Covered by a layer of greyish green
glass on the outside and a thick layer of green
to bluish green glass on the inner surface.’196
189 Stoepker 2021, 229.
190 Many thanks to Mark Taylor of ‘Heart of
191
192
193
194
195
196
England Glass’ (https://
heartofenglandglass.co.uk) and
‘The Glassmakers’ (http://www.
theglassmakers.co.uk).
Louis 2015, fig. 4b, c, d; Cabart, Pactat &
Gratuze 2017; Henton 2020.
Stoepker 2021, 201, table 10.03, V08-190GL-07.
Preiß 2010.
Henderson, Sode & Sablerolles 2019;
Schibille & Freestone 2013, 2–3, fig.
1C, D.
Isings 1978; Preiß 2010.
Isings 2015, 444, No. 16035.
40
—
Perhaps the glass on the outside seems greyish
green because the outside of the crucible had
discoloured to a dark grey colour as the result
of reheating, as can be seen on the crucible
fragments from Utrecht-Domplein (see
Section 3.5.1).
The glass production waste from the PPAH
excavations and the Veilingterrein excavations is
discussed in more detail below.
3.4.1 Wijk bij Duurstede – Parkeerplaats
Albert Heijn (PPAH)
197 Van Dockum 1997.
198 Nyst 2003, 13.
199 Nyst 2003, 32, catalogue 4.2; Preiß 2010
passim.
200 Van Es & Verwers 1980.
201 De Koning 2012, 186.
Site:
5
Site type:
Settlement
Province:
Utrecht
Municipality: Wijk bij Duurstede/Dorestad
Place:
Wijk bij Duurstede
Toponym: Parkeerplaats Albert Heijn (PPAH)
(Steenstraat/Zandweg)
Start date: 750–775
End date:
c. 1250
Description: In 1992 excavations by the former
Dutch National Service for Archaeological
Heritage (ROB) were carried out at the intersection
of Steenstraat and Zandweg before the
construction of a car park for a planned new
supermarket (Albert Heijn), hence the toponym
Parkeerplaats Albert Heijn (PPAH).197
Fig. 3.9 Dorestad-Parkeerplaats Albert Heijn (PPAH) site: Glass working crucible with melted
opaque white glass, probably from a Roman tessera. Defects in the transparent base may
indicate where crushed tesserae were attached but have fallen out. Rim diameter 17.5 cm
(Preiß 2010, 125, fig. 107). Not to scale.
Both Carolingian and later settlement traces
were found. In the Carolingian period this area
belonged to the settlement west of the vicus on
the river bank. This settlement has a more
agrarian nature as evidenced by farm buildings
discovered later during the Veilingterrein
excavations (2007–2008) further north along
Zandweg (see Section 3.4.2). Carolingian features
include ditches which seem to enclose (farm)
yards on which posthole clusters, pits and wells
were found. Most of the contents of the pits
were sieved which yielded an enormous amount
of glass fragments, including a rare gold-foil
beaker, glass beads, bone artefacts and birds and
fish bones. The bone artefacts included
production waste. The adjoining Albert Heijn
supermarket site also produced Carolingian
period loom weights and traces of metalworking.198
Pottery finds were made up of the usual
range of Carolingian wares, for instance Rhenish
and Eifel ceramics, as well as younger wares
from Pingsdorf, Andenne and Paffrath. The finds
prove this part of the settlement remained
inhabited after the Viking attacks and only
shifted in a south-easterly direction in the
mid-13th century, where the town of ‘Wiic bi
Duerstede’ would develop.
Glass production waste: Some thirty-six fragments
of glass-working waste, including eight tesserae,
were found distributed throughout this part of
the settlement (trenches 810–815).199 Among
these was a large fragment of a glass crucible
(fig. 3.9). It was found in a pit together with
other glass-working waste made up of two
blue tesserae, a regular and an irregular drop of
translucent pale green glass, a small, dark
sphere, six melted lumps of translucent pale
green vessel glass and a melted fragment of
‘black’, deep olive-green glass. The crucible
belongs to a pluriform group of bowls (type WX
in the Dorestad typology)200 which are late
Merovingian in origin and are made in different
production centres in the Rhineland. On the
Dorestad-Veilingterrein site they are mostly
8th century in date and are also found in
Carolingian yards.201
The inside of the crucible is covered with a thin
layer of almost colourless glass of c. 1 mm
thickness with a thicker patch of opaque white
glass which probably represents (a fragment of)
a melted tessera. Preiß points out that defects in
41
—
the translucent glass probably indicate locations
where other (crushed?) tesserae had been
attached.202 The crucible can be linked to beadmaking, but also to vessel or window glass
production.
3.4.2 Wijk bij Duurstede – Veilingterrein
and Frankenweg/Zandweg
Site:
6
Site type:
Settlement
Province:
Utrecht
Municipality: Wijk bij Duurstede
Place:
Wijk bij Duurstede
Toponym: Veilingterrein (Zandweg) &
Frankenweg/Zandweg
Start date: c. 600–650
End date:
900
Description: Due to their proximity these two
sites have been treated as one. The archaeological remains belong to the same agrarian
settlement west of the vicus and harbour works
as the PPAH-excavations (see Section 3.4.1).
The excavation Frankenweg/Zandweg
was carried out by the Archeologisch Diensten
Centrum (ADC) in 2001 before the planned
construction of an apartment block on this
site.203 During the Merovingian phase (c. 600–
725) no buildings were found, but four wells
indicate there must have been some. It was
only possible to reconstruct one Carolingianperiod building with a boat-form, possibly a
farmhouse. Other Carolingian features are
ditches and eight wells. Metal slags were found
as well as waste products from antler-working
and possibly glass-working. Habitation
decreased dramatically during the third phase
(late 9th–10th century) and was discontinued
during the high middle ages when the area was
in use for arable farming.
Immediately south of the 2001 excavation,
the ADC carried out another, much larger
excavation (1.7 ha) on the site of a former fruit
auction (Veilingterrein) along Zandweg.204
The archaeological remains belong to the same
agrarian settlement west of the vicus and
harbour works as the PPAH excavations
(see Section 3.4.1). Habitation started in the
Merovingian period around 650. Three large
farm yards were identified in this area. On these,
the remains of two farmhouses were identified
as well as twelve wells, many waste pits and
inhumations. In the third quarter of the
8th century a new partition of the land took
place into rather narrow strips of land, clearly
delineated by ditches. These boasted eight
buildings, 112 wells, many pits, latrines and
oven-pits. Seven boat-shaped farmhouses
were identified and one building with straight
sides, its function is uncertain. The farm
houses do not show obvious differences in
size or layout.
Several yards yielded evidence for ironsmithing and weaving wool, one (yard K4)
provided clear evidence for specialized crafts,
namely amber-working and the production of
brass (terminus post quem 800), while finds of
two touchstones with traces of gold and the
largest concentration of coins point to trade
activities in this yard.
A large amount of pottery from the
German Rhineland and the Eifel, mill stones
from the Eifel, wine glasses most probably from
the Rhineland, wine (in barrels) from the middle
Rhine region, combs from Scandinavia, and
Roman tesserae, possibly from the Mediterranean,
underline the international character of the
settlement and the importance of trade.
The period between 875 and c. 1050 saw a steep
decline in habitation and only one farmhouse
can be identified, while three were found dating
between c. 1050 and 1300. From c. 1300 onwards,
the area was used for arable farming.
Glass production waste: The glass production
waste from the Veilingterrein was published in
2012 in a monograph on the excavations, and
recently in an overview article on the beads from
Dorestad.205 Tesserae make up the largest
category (n=13), almost all in the blue/green
colour spectrum. Fragments of translucent
bluish-green and dark blue glass point to
imports of (chunks or ingots of) raw glass.
One bluish-green flake was struck off from a
larger lump of raw glass and has an imprint of a
glassworker’s tool (fig. 3.10). The small diameter
of the tool makes it more likely it was a beadworking tool (punty).
Two convex fragments of dark blue glass
(fig. 3.10) clearly belong to plano-convex ‘cakes’
also found on Scandinavian bead-making sites
such as Åhus, Sweden.206 There are three glass
drops in corresponding colours (fig. 3.10).
Two glass rods of opaque yellow glass (square-
202
203
204
205
Preiß 2010.
Sier, Van Doesburg & Verwers, 2004.
Dijkstra, 2012.
Sablerolles & Henderson 2012, 326–333;
Langbroek 2021b.
206 Callmer & Henderson 1991, fig. 2.
42
—
6071
5041
0
1:1
4834
5cm
6500
Fig. 3.10 Dorestad-Veilingterrein site: A translucent bluish green flake of raw glass with imprint from a probable
punty (6071), a convex fragment from a translucent blue glass ‘cake’ (5041), a plano-convex drop of bluish green
glass (4834), an opaque yellow glass lump with a section of an opaque yellow composite rod and a fragment of
punty glass (6500) (Sablerolles & Henderson 2012, fig. 6.19).
207 Sablerolles & Henderson 2012, 329–330,
208
209
210
211
212
213
afb. 6.21, findnr. 5791; Langbroek 2021b,
table 7, findnr. Veilingterrein 5791.
Langbroek 2021b, fig. 12.5, 64, table 7,
findnr. Veilingterrein 5791.
Lassaunière et al. 2016; Henon 2020.
Langbroek 2021b, table 7, findnr.
Veilingterrein 1195.
Sablerolles & Henderson 2012, 326–333,
afb. 6.21, 5791; Langbroek 2021b.
Van Doesburg 2004.
Langbroek 2021b, table 7, findnrs
Zandweg WD 754.2.63b, WD 754.2.63b.
sectioned) and white glass (round-sectioned) are
likely to be associated with bead-making.
A lump of opaque yellow glass has part of a
composite opaque yellow glass rod and yellow
punty glass from a glassworker’s tool melted
onto it (fig. 3.10). Composite glass rods were also
found on the Merovingian bead-making site of
Maastricht-Jodenstraat and can be linked to
bead-making (see Section 3.2.1).
An opaque green object, first believed to
perhaps represent the pinched end of a glass
rod,207 is actually a failed, wound bead and
represents waste from glass bead production.208
A fragment of a twisted bichrome rod has a
translucent bluish-green core with an opaque
white trail twisted round it. In the Carolingian
period reticella rods with translucent greenish
cores and opaque twisted trails were used to
decorate vessel glass. Twisted bichrome cables
have been found at the French monasteries of
Hamage and St Amand-les-Eaux on the river
Scarpe, where they were used to decorate
globular jars and bowls.209
A thick, flat piece of opaque yellow glass
shows black traces of iron oxide scale in the
fractures on the sides as well as what seem to
be small amounts of ceramic (from a crucible?)
or red-baked clay.210 This yellow glass could
have been intended for bead-making or for
decorating 8th century vessel glass, funnels, jars
and bowls, with yellow trails. Finally, a quantity
of melted bluish-green glass adheres to a
fragment of red-baked clay.211
In several contexts on the abovementioned yard K4, a small concentration of
glass bead production waste, including four
tesserae, was found together with waste from
amber-working, raising the possibility that either
a beadmaker was working here side by side
with an amber-worker or that both crafts were
carried out by the same craftsman. A combination
of waste from amber and a small concentration
of bead-working waste, including five tesserae,
is also found on the northernmost yard (yard
K1), which saw most activity around 800.
Finally, from the 2004 excavation
(Frankenweg/Zandweg), a tessera of ‘bright blue
glass’ and a lump of clay covered with a thick
layer of ‘blobby greenish glass’ were found
immediately north of yard K1 on the Veilingterrein
site.212 These are not included in Preiß’s 2010
overview of glass production waste from Dorestad.
Nor are two small spheres of whitish glass,
possibly weathered translucent glass.213
These finds point to the local production of
beads, but the production of vessel glass cannot
be excluded.
There is a relatively small amount of glass
production debris, but it is worth considering
that, in contrast to the situation in the famous
Danish bead-making site of Ribe, no original
floor surfaces were preserved, so only finds from
pits, wells and ditches were recovered, and that
wet sieving was only carried out in specific
43
—
instances.214 It is therefore hard to gauge whether
this constitutes small-scale production at
household level, production for local demand or
(supra-)regional production.
3.5
Utrecht, Utrecht Province
Utrecht may be one of those rare places in the
Netherlands where there are indications for
Merovingian glass-working – though we need to
await the final results of the post-excavation
work to establish if there is more comprehensive
evidence – as well as evidence for glass-working
in the Carolingian period.
3.5.1 Utrecht-Domplein
Site:
9
Site type:
Proto-urban settlement
Province:
Utrecht
Municipality: Utrecht
Place:
Utrecht
Toponym: Domplein
Start date: c. 40
End date:
present day
Description: The Domplein (Dom Square) in
Utrecht city centre is named after the Domkerk
(Dom church or St Martin’s Cathedral) which
dominates the central square. Excavations
carried out in the 20th and 21st centuries have
made clear that the square was continually
inhabited from the Roman period to the
present-day. Small-scale excavations took place
between 1927 and 1949.215 These revealed parts
of an auxiliary fort – Traiectum – which was first
constructed in the 40s of the 1st century. After
four wooden phases, it was rebuilt in stone
around 200 and abandoned during the course of
the 3rd century. Furthermore, evidence for
Carolingian and high medieval habitation was
found, as well as remains of the 11th century
imperial palace – Lofen.
Excavations in 1993 of part of the Heilig
Kruiskapel (church of the Holy Cross), which was
demolished in 1829, made clear that the church
was probably founded around 700, rather than
in the 10th century as was suggested in 1929. This
has led to speculation that this simple hall
church may be identified as one of two churches
reputedly built here by the Anglo-Saxon
missionary Willibrord around 695: either the
now-lost church of St. Salvator (SintSalvatorkerk) or the church of St. Martin (SintMaartenskerk), the predecessor of the presentday Dom church.
Glass production waste: In total, 17 fragments of
glass crucibles were found during excavations in
1933 at the site of the Roman castellum on or
near the Domplein (table 3.5).216 Of these, five
were published in 1934 and illustrated by the
famous Dutch artist and illustrator Anton Pieck
(fig. 3.11).217 Three of the published fragments
(body fragments 203, 204 and B8) were not
among the crucibles made available for
sampling carried out at the Cultural Heritage
Agency of the Netherlands (RCE).218 Two of the
latter were found just west of the Domplein in
‘Flora’s hof’ (courtyard). Judging from the
illustration (fig. 3.11), the glass colours in these
214 Dijkstra 2012, 25–27, 591.
215 Wynia 2013.
216 Vollgraff & Van Hoorn 1934, 63–64. The
objects are part of the collection of the
Provinciaal Utrechts Genootschap van
Kunsten en Wetenschappen (PUG) or the
Provincial Society for the Arts and
Sciences (PUG findnumber 234).
217 Vollgraff & Van Hoorn 1934, pl. XXII.
218 The crucibles were sampled by Hans
Huisman, Dutch National Heritage
Agency (RCE), Amersfoort.
Fig. 3.11 Utrecht-Domplein site: Illustration of glass crucible fragments by Anton Pieck (Vollgraff
& Van Hoorn 1934, 63-64, pl. XII); top row: findnrs 234 (rim), 203 (body); middle row: findnrs 204
(body), B8 (base); bottom row: findnr B8 (body) (Vollgraff & Van Hoorn 1934, pl. XXII).
44
—
crucible fragments are a deep blue-green rather
than ‘dark green’ and pale bluish green rather
than ‘pale green’. Six photos of the glass
production waste are given at Appendix IV
(figures appendix IV.39-45).
219 Isings, Rauws, Lägers & De Kam, 2009,
48–49.
220 Isings, Rauws, Lägers & De Kam, 2009,
221
222
223
224
225
48–49.
Barfod, Feveile & Sindbæk 2022.
Sablerolles and Henderson 2012;
Langbroek 2022.
See for instance Baumgartner & Krüger
1988, 71, No. 14, 72, No. 16.
Stoepker 2021, 230–232, Table 11.1;
Sablerolles 2023.
Van Es & Verwers 1980, 81–87.
At the time of the excavation, the fragments
were erroneously believed to be Roman in date,
but in a publication of 2009 a rim fragment
(DPL-234) and a base fragment (DPL 1933-zn3)
were published as fragments of (a) Carolingian
globular pot(s).219 Judging from photographs of
the objects, all the other fragments probably
also belong to this type of pot (see Dijkstra’s
contribution below). Based on the find locations,
the glasses inside the pots and their fabrics, at
least eight pots are represented, but this number
may have to be adjusted in future when the
fragments themselves are studied.
Two rim fragments have off-white deposits
on the inside below the rim, on top of the rim
and on the outside, just below the rim (DPL 1933
234, DPL 1933 zn2. The composition of these
white deposits will be discussed in more detail in
Chapter 5. A rim fragment (DPL 1933 234) has
two very thick deposits of translucent glass
which start just below the rim and are sticking
out above it. This is likely to be the result of an
attempt at removing the last remains of viscous
glass from the crucible which solidified just
before it could be poured out.
It was suggested by Isings et al. that glass
production on the Domplein is linked to making
red glass inlays for jewellery or beads.220 A recent
discovery of a glass bead workshop at Ribe
dated to between 760 and 790 yielded fragments
of a crucible with the remains of recycled Roman
green glass with streaks of red and brownish
coloration caused by iron and copper. It is
suggested that this glass was used for the
production of opaque red cylindrical beads and
black wasp beads221, bead types that were also
found at Dorestad.222
The glass inside the crucibles at Utrecht
Domplein may well represent an attempt at
making opaque red glass. However, red marbled
translucent blue-green/bluish-green glass was
popular in the late Merovingian and Carolingian
periods for the production of vessel glass,
especially bowls and bulbous jars,223 and for
flat glass intended for stained glass windows.
A fragment of a Carolingian red marbled bluegreen glass quarry was found at the early
medieval monastery at Susteren (Limburg
Province) (fig. 3.12), where there are indications
for the production of window glass, most likely
in the Carolingian period (see Section 3.3).224 In
the context of the Domplein, production of
window glass is certainly a possibility since there
was at least one church here in the 8th century
(see above). Interestingly, the monastery at
Susteren had special links with the church of
Utrecht as it had been founded by the AngloSaxon missionary Willibrord (see Section 3.3).
1:1
0
5cm
Fig. 3.12 Susteren-Abdijterrein: A fragment of quarry of
translucent blue-green glass with red marbling
(Stoepker 2012, table 11.1; Sablerolles 2023).
Glass crucibles: Utrecht-Domplein
Contribution by Menno Dijkstra (University of
Amsterdam)
Based on the photographs, the crucible
fragments probably all derive from Carolingian
globular pots (Dorestad type W III)225 dating
between 750 and 875/900 (fig. 2.13). This is
indicated by the three rim fragments as well as
the relatively thin-walled body fragments and
the lenticular base.
1:4
0
20cm
Fig. 3.13 Dorestad-Veilingterrein: Example of a complete
Carolingian globular pot of Badorf ware (Dorestad type
W IIIA) (De Koning 2012, afb. 4.28, findnr 503)
(Photograph: Archeologisch Diensten Centrum (ADC).
Height c. 17 cm.
45
—
Table 3.5 Utrecht-Domplein: fragments of glass crucibles together with photo numbers and those scientifically
analysed (Description of missing fragments from Vollgraff & Van Hoorn 1934, 63–64).
Find number (modern)
Find number (old)
Crucible
Glass
Sample
DPL 1933 zn2
?
1
rim
translucent bluish green & off-white
deposit
DOM 5
35
figure appendix IV.43
DPL – zn3 (Flora’s hof)
B8
1
base
translucent dark green glass (1 mm
thick) marbled with opaque (purplish)
red glass (0.5 mm thick)
DOM 3
33
figure appendix IV.41
Missing (Flora’s hof)
B8
1
body
translucent pale green to dark green,
red marbled
-
-
-
Missing
203
1
body
translucent dark green layer (1-6 mm
thick)
-
-
-
Missing
204
1
body
translucent pale green glass, cracked
(0.5–3 mm thick)
-
-
-
DPL 1933 234
234
1
rim
translucent bluish green marbled with
opaque red
DOM 4
34
figure appendix IV.42
DPL 1933 73-84
?
4
rim
remains of translucent ‘garnet’ red glass
in translucent bluish green glass
DOM 6
36
figure appendix IV.44
DPL 1933 77-53
B7
1
body
colourless glass (crucible: thin-walled
beige fabric)
DOM 2
32
figure appendix IV.40
DPL 1933 77-56
B26
1
body
remains of opaque red glass
-
DPL 1933 77-36
B42
2
body
colourless glass
DOM 1
DPL 1933 77-57
B42
1
body
colourless glass
-
-
-
DPL 1933 77-31
B45
1
body
remains of opaque bright red glass
-
-
-
DPL 1933 77-58
B45
1
body
opaque bright red glass
-
-
-
Total
N
Sample number
Photo number
31
figure appendix IV.39
17
Perhaps this type of globular pot was preferred
because of its closed form. Crucible fragments of
similar pots were probably found at the monastery
of Susteren (see Section 3.3) and in Leidsche Rijn
– Leeuwesteyn Noord (See Section 3.6.2).
The photographs show that the fragments
are of a reddish colour on the inside and grey on
the outside. One thin-walled fragment is beige
on the inside (DPL 1933 77-53). The photographs
show that the outside surfaces of the sherds are
marked by small blisters which are located in
places where the temper pierces the surface.
These blisters are probably caused by reheating
of the pots when reused as glass crucibles and
may represent the early phases of vitrification.
The presence of rims and the fact that the
base fragment and the body fragments are not
covered by spilt glass on the fractures, indicate
only complete pots were used. This is in contrast
to the late 6th-century Maastricht-Jodenstraat
(see Section 3.2.1) and early 7th-century
Rijnsburg-Abdijterrein sites (see Section 3.8)
where complete pots contained colourless or
translucent blue-green glass, while only the
lower parts of the pots were used for working
relatively small amounts of highly coloured
opaque yellow glass: by removing the upper parts
of the pots, the beadmakers created wide-open
‘bowls’ that allowed for easy access to the glass.226
It is difficult to make a statement about the
types of fabrics represented on the basis just of
photographs (table 3.5). Moreover, it can be hard
to distinguish the different types of tempers in
rather thin-walled Rhenish fabrics from Badorf
and Walberberg in the Vorgebirge and Mayen in
the Eifel. Parts of the photographs of the crosssections show the presence of quartz sand, but
this can be present in both Mayen and Vorgebirge
wares. Seven sherds show black specks, which
are mostly tiny cavities, but some of them could
be fragments of augite or hornblende, which are
typical for Mayen fabrics (see table 3.6).
Other typical Mayen volcanic temper-like
pumice grains and off-white specks (clay pellets)
are almost absent, but this is not uncommon in
thin-walled fragments. This thinness could also
explain the absence of small red particles, probably
ferronic nodules, which are sometimes noted in
Vorgebirge fabrics.227 Only the fabric of the
sectioned rim fragment DPL 77-234 can with
226 Sablerolles, Henderson & Dijkman 1997,
304; Dijkstra, Sablerolles & Henderson
2011, 185–186.
227 Fabric details are based on Redknap
1988, 5 and 11; Bardet 1995, 221–230;
Keller 2012, 213.
46
—
Table 3.6 Utrecht-Domplein: Rim type and fabric of the Carolingian globular pots used
as glass crucibles.
Find number
Crucible
Rim type
Fabric (based on photographs)
DPL 1933 zn2
rim
W IIIA
coarse quartz sand, Walberberg fabric (Dorestad w-4?)
DPL 1933 234
rim
W IIIA
black specks etc., Mayen fabric (Dorestad w-12)
DPL 1933 73-84
rim
W IIIA
coarse quartz sand, Walberberg fabric (Dorestad w-8)
DPL 1933 zn3
base
-
black specks, Mayen ware?
DPL 1933 77-31
body
-
black specks, Mayen ware?
DPL 1933 77-36
body
-
black specks, probably Mayen ware
DPL 1933 77-53
body
-
not determinable
DPL 1933 77-56
body
-
Mayen or Walberberg fabric?
DPL 1933 77-57
body
-
black specks, probably Mayen ware
DPL 1933 77-58
body
-
black specks, Mayen ware?
W IIIA = globular pot of Badorf ware, dating between 750 and 875/900 (Dorestad type W IIIA).
confidence be identified as typical for the
production centre at Mayen (Dorestad fabric
w-12). Rim fragment DPL 73-84 can be identified
as a Walberberg fabric (Dorestad fabric w-8) due
to the temper with coarse quartz sand.
3.5.2 Utrecht–Oudwijkerdwarsstraat
228 Post-excavation is ongoing at the time
of writing. Many thanks to Marieke
Arkema, external archaeologist,
Municipality of Utrecht,
Ontwikkelingorganisatie Ruimte,
Duurzame Stad, Erfgoed, for all initial
information on the site.
Site:
10
Site type:
Proto-urban settlement
Province:
Utrecht
Municipality: Utrecht
Place:
Utrecht
Toponym: Oudwijkerdwarsstraat
Start date: c. 600 (?)
End date:
c. 750
Description: During building activities in
Oudwijkerdwarsstraat just east of Utrecht old
city centre, archaeologists of the municipality of
Utrecht recovered remains of an early medieval
settlement.228 It comprises part of a large
settlement that stretched along the former bank
of the Kromme Rijn for at least 150 metres. An
initial assessment of the pottery points to
habitation in the 7th and first half of the 8th
centuries. This find is very important for two
reasons: not much is known about early
medieval life in this part of Utrecht, and the
finds are extraordinarily well preserved.
No house plans were found, since only the
rear parts of farm yards were found. However,
there are remains (postholes, clay) of probable
subsidiary buildings, while ditches may
represent yard boundaries. Waste pits yielded
pottery, animal bones, millstones and loom
weights, offering an insight into the economy of
the settlement. About 1400 amber fragments,
ranging in size from splinters to fairly large
lumps, and including half-products, show amber
was worked in the settlement. Never before has
so much amber been found in Utrecht. The
amber finds were concentrated in two shallow
pits which yielded many other finds including 13
glass fragments. Furthermore, silver and copper
coins and glass beads were also found.
Glass production waste: Glass production waste is
scanty. An irregular drop/melted lump of
translucent bluish-green glass (with a yellowish
tinge) (WP 5-1-135) could be the result of glassworking, but could also be accidentally melted
vessel glass. However, when vessel glass is
accidentally melted, for instance in a hearth, it is
usually possible to tell it is a deformed vessel
fragment. The glass drop comes from a pit (S135)
which yielded animal bones, pottery and burnt
clay. This pit was only discovered when
sectioning an adjacent pit (S62) which yielded
pottery, metal, burnt clay, animal bones and
three amber fragments. The pits are at a
distance of 30 metres from the two shallow pits
mentioned above with concentrations of amber.
A small amount of glass (WP 6-1-170) seems to
be crushed, perhaps in preparation for being
melted down. Two photos of the glass
production waste -OUDWIJ 1 (sample 78) and
OUDWIJ 2 (sample 79)- are given at Appendix IV
(figures appendix IV.45 and 46).
47
—
3.6
Leidsche Rijn, Utrecht Province
The large early medieval settlement excavated
at Leidsche Rijn has yielded sparse evidence for
glass-working in the late Merovingian and
Carolingian periods.
3.6.1 Leidsche Rijn-LR 51/54
Site:
11
Site type:
Riverine settlement
Province:
Utrecht
Municipality: Utrecht
Place:
Utrecht
Toponym: Leidsche Rijn 51/54
Start date: 575
End date:
775
Description: In the period 2000–2001 a field
survey, coring and a trial excavation were
conducted as a result of plans to expand the
A2 highway in Leidsche Rijn located in the
western part of the city of Utrecht.229
The archaeological research demonstrated that
an early medieval settlement was situated here.
In 2005 a part of this settlement was
excavated by the municipality of Utrecht
(project LR51 and LR54). The excavations
revealed that the early medieval settlement was
located on relatively high land along the
northern bank of the Oude Rijn. The second half
of the 7th and the early 8th century formed the
heyday of the settlement. In the course of the
second half of the 8th century habitation
declined. However, a few stray finds from the
9th century suggest that the area was also
inhabited later, in the Carolingian period.
In total 88 early medieval buildings were found,
which comprised 14 farmhouses and 74
outbuildings. Noteworthy are 34 large
outbuildings with very long, pointed wooden
posts, maybe granaries.
As only pollen (no grains) of oats, rye and
wheat were found, the arable fields were probably
not very close to the settlement. Farm animals
were reared, especially cows and sheep. The lack
of bones of 2-to-4-year-old cattle in the oldest
and youngest habitation phases may point to
the export of cattle during these periods. The
diet was supplemented with riverine and
a
b
0
1:1
5cm
Fig. 3.14 Leidsche Rijn - Leeuwesteyn Noord: A fragment of a glass crucible, probably from a
Carolingian globular pot (Dorestad type W III), a) covered on the inside with red marbled
translucent greenish glass (left) and b) on the outside with a blob of translucent greenish glass
(right) (Sablerolles 2019, fig. 7.15).
imported marine fish as well as game. The
inhabitants were engaged in (occasional)
shipping activities and artisan production: iron
was produced from imported iron ores and was
also worked in the settlement; bronze and lead
were worked.
Combs were produced from antler and
seven fragments of raw amber point to amberworking. As much as 80% of the pottery was
wheel-thrown, imported from the Rhineland
together with the remains of at least seven late
Merovingian drinking glasses
(palm-funnels). Trade activities may have taken
place, judging by the discovery of 37 early
medieval coins. It is suggested that imported
wares such as pottery, glass, millstones, iron ore,
coal and amber could have been obtained with
agricultural surplus stored in the granaries,
as well as animal products.
Glass production waste: An opaque blue tessera
was found in a waste pit on yard 3 in the eastern
part of the settlement.230 This yard also yielded
most evidence for craft production: metalworking (iron and bronze), bone- and antlerworking and amber-working. Moreover, almost
all of the twelve glass beads were found on this
yard, one of them in the same pit as the blue
tessera, as well as most fragments of old,
Roman and Migration period glass, which might
represent cullet intended for recycling.231 The pit
is dated to the late Merovingian period which
would make the tessera the earliest specimen in
the Netherlands together with a tessera from
Oegstgeest – Nieuw Rhijngeest Zuid (Rijnfront)
(see Section 3.7).
229 Nokkert, Aarts & Wynia, 2009.
230 Isings 2009, 247, afb. 11.2, table 11.1.
231 Isings 2009, 249, afb. 11, table 11.2, 250–
251.
48
—
Although there is no glass bead production
waste to accompany the blue tessera, the fact
that it was found together with raw amber may
point to a glassworker and amber-worker being
active in this part of the settlement.
3.6.2 Leidsche Rijn-Leeuwesteyn Noord
232 Norde 2019.
233 Sablerolles 2019, 136–138.
234 Norde 2019, 274–275, fig. 14.1, 7.
Site:
12
Site type:
Riverine settlement
Province:
Utrecht
Municipality: Utrecht
Place:
Utrecht
Toponym: Leeuwesteyn Noord
Start date: 575
End date:
850
Description: In 2015 and 2016 excavations,
commissioned by the municipality of Utrecht and
carried out by RAAP, took place immediately east
of the early medieval settlement Utrecht-Leidsche
Rijn LR 51/54 which was excavated a decade earlier
(see Section 3.6.1).232 The settlement, on the
northern bank of the Old Rhine, was probably
located in an outer, not an inner bend of the
river as had previously been assumed.
The combined excavations have unearthed
275 metres of a large settlement which continues
in easterly and westerly directions. The excavations
confirm the idea proposed by the 2005 excavators
that the settlement continued into the 9th century,
probably till around 860.
Just as in Dorestad-Veilingterrein, the area
was reorganized in the Carolingian period and
(parts of) yards are delineated by north/south
oriented ditches and picket fences.
Remains of two farms and several
subsidiary buildings were found. This brings the
total of houses in the Leidsche Rijn settlement
to 14, while a total of 57 large subsidiary
buildings were found. Of the latter, 37 are large,
two-aisled buildings constructed with very long
posts, deeply driven into the ground. They are
interpreted as warehouses. It has been pointed
out that similar structures are also found in
some other 7th–8th century settlements along the
Old Rhine (see Section 3.7) and it has been
suggested that these could point to the
existence of specialized settlements.
It was argued that the paucity of cereal
grains in botanical samples collected during the
2005 excavations may indicate that (this part of)
the settlement was geared towards craft and
trade rather than agriculture. There is evidence
for antler-working; metal-working evidence
consists of iron-smithing and bronze-working.
There is possible evidence for lime-burning.
A glass crucible indicates glass was worked in
the settlement.
Glass production waste: A body fragment of a
crucible was recovered from one of the
uppermost layers of the fill of the Old Rhine
which contained both Merovingian and
Carolingian pottery (fig. 3.14).233 The fragment is
reddish on the fracture, and grey on the outside,
and is probably of a Carolingian globular pot
(Dorestad type W III, fabric 12), similar to the
crucibles from Utrecht-Domplein (see Section
3.5.1). The crucible is covered on the inside with
a thin layer (1–1.5 mm) of translucent (bluish-)
green glass marbled with opaque red glass,
comparable to some of the Utrecht-Domplein
specimens. A drop of translucent greenish glass
was spilt on the outside.
In this context, it is less likely that this glass
was being worked for the production of window
glass as was suggested for Utrecht-Domplein.
Perhaps, therefore, this glass was used for the
production of beads. The only bead recovered
from the site is a spiral bead from a Carolingian
pit, made of faintly translucent greenish glass
full of small bubbles, dark inclusions (iron
fragments from a beadmaker’s tool) and what
looks like black/dark red streaks.234 This bead
may be one of the local products. It is similar to
a bead found on the Maastricht-Rijksarchief site
where beads were made (see Section 3.2.3).
3.7
Oegstgeest–Nieuw Rhijngeest Zuid
(Rijnfront), Zuid-Holland Province
Site:
13
Type:
Riverine settlement
Province:
Zuid-Holland
Municipality: Oegstgeest
Place:
Oegstgeest
Toponym: Nieuw Rhijngeest Zuid (Rijnfront)
Start date: c. 550
End date:
725
Description: From 2009 until 2014 the University
of Leiden excavated a settlement dating to
Merovingian times. Smaller parts of the
settlement had already been excavated by
49
—
ARCHOL and ADC.235 The Oegstgeest settlement
has been excavated almost completely. It was of
a modest size, with c. five or six contemporary
farmsteads during its existence. Its population
will not have been larger than c. 60 persons.
It was located along the northern bank of the
Old Rhine close to its estuary, some five km from
the coast. The settlement was set in a landscape
that was subjected to both riverine and maritime
influences and was intersected by river arms,
gullies and creeks, creating four quarters or
‘islands’ interconnected by one bridge and
several small dams. Structures connected to
shipping are jetties, quay works, and land
abutments. As is the case at Rijnsburg-Abdijterrein
(see Section 3.8) and Valkenburg-De Woerd
(see Section 3.9), the farmhouses were laid out
on a grid at right angles to a main gulley of the
Oude Rijn. Each yard, enclosed by fences made
of wattle, consisted of a farmhouse with several
associated outhouses, multiple pits and wells.
The postholes of the outhouses were remarkably
deep and it is suggested that the buildings may
have had raised floors enabling safer storage of
agricultural produce.
Animal husbandry in the floodplains (cattle),
some agriculture on the river levees, and fishing
were the backbone of the economy. They also
practised crafts such as smithing, and above all
casting of copper alloy objects. Small and
unexpectedly large crucibles show that this must
have taken place to satisfy the needs of themselves
and those in other settlements. Antler combs
were made and amber-working was a widespread
activity in the settlement. There is some tentative
evidence for glass-working. Imported pottery
from the German Rhineland, grain from löss
areas, probably the Main area, wine (barrels)
from the middle Rhine area and some exotic
imports testify that this settlement made good
use of its advantageous location along one of
the most important early medieval supraregional waterways, with the possibility to
engage in exchange with the wider, early
medieval world.
Glass production waste: Five fragments of
possible glass production waste were found in
five different locations: two fragments of
tesserae, a possible fragment of a glass rod,
a black drop and a trapezium-shaped fragment
of translucent blue-glass with rounded edges.236
Two weakly transparent green-blue fragments
from a pit and a well from probable Roman
tesserae could be the earliest evidence from the
Netherlands for the reuse of tesserae by early
medieval beadmakers, together with a specimen
from the Merovingian settlement of Leidsche
Rijn L51/54 (see Section 3.6.1). A colourless
fragment is interpreted as part of glass rod used
for winding beads. Since wound colourless glass
beads were not common in the 7th century,
it could perhaps also be a fragment of a reused
Roman glass stirring rod. A drop of brown,
almost black glass was split when it was heated.
Small, spherical black drops have been found in
association with glass bead production waste
in Åhus, Dorestad, Maastricht-Rijksarchief and
Wierum.237
The wound, monochrome and trailed beads
found in the settlement (n=28) mostly date to
the 7th century and show many similarities with
half-products and wasters of beads found in the
contemporary nearby bead production site at
Rijnsburg-Abdijterrein (see Section 3.8). 238
The shapes, colours and decorations used are
strikingly similar, including small flattened
globular beads of opaque yellow, red and
white glass, white beads with translucent bright
blue crossing trails and a red bead with white
crossing trails, and it is suggested that these
beads could have been made in RijnsburgAbdijterrein.239 Alternatively, they could have
been made by travelling beadmakers who
visited riverine settlements along the Old Rhine,
including Rijnsburg-Abdijterrein and possibly
Valkenburg-De Woerd.
3.8
Rijnsburg-Abdijterrein,
Zuid-Holland Province
Site:
14
Site type:
Riverine settlement
Province:
Zuid Holland
Municipality: Katwijk
Place:
Rijnsburg
Toponym: Abdij
Start date: 600
End date:
12th century
Description: Between 1944 and 1966 a series of
excavations by the archaeological institutes of
the universities of Groningen and Amsterdam
and the former National Service for
Archaeological Heritage (ROB, now RCE) took
place on the site of a former Benedictine abbey
235 Hemminga & Hamburg 2006;
236
237
238
239
Hemminga et al. 2008; Dijkstra 2011, 134;
De Bruin 2018, 20–25; De Bruin, Bakels
& Theuws 2021.
Langbroek 2021a, table 12.2, fig. 12.6.
Callmer & Henderson 1991, Table 1C, 1;
Preiß 2010, number 47; Henderson,
Sode & Sablerolles 2020, 78.
Langbroek 2021a, fig. 12.2.
Langbroek 2021a.
50
—
240 Dijkstra, Sablerolles & Henderson 2011;
Dijkstra 2011, 114-133.
241 Nicolay 2017.
242 Koch 1977, 207, Farbtaf. 3, Gruppe 34.
243 Pion 2014; Vrielynck, Mathis & Pion
2018.
(c. 1130–1574), immediately east and north of
the present-day church.240 The excavations of
this convent led to the discovery of various older
settlements lying underneath. The oldest phase
of the settlement dates to around 600 AD,
when the southern bank of the Vliet was divided
with wattle fences into small north-south
oriented plots where several rectangular
longhouses and secondary buildings were built,
that can be associated with three or four
generations of occupation lasting until around
720. It was situated in a tidal saltmarsh at the
southern bank of a creek (the Vliet) of the
Oude Rijn, close to the former mouth of the
river. The economy was primarily based on
farming, although traces of oven-/furnace-like
structures, possible outdoor hearths, smithing
slags and a few tuyères point to iron-working,
while two crucibles, bronze fragments and a
probable casting mould point to bronze casting.
One of the smaller buildings close to the creek is
believed to be a possible home of a smith and
his family. A few lumps of amber may point to
amber-working. The evidence for the production
of Merovingian type beads also comes from this
phase of the settlement. During the late
6th century and the early decades of the 7th,
the settlement can be considered part of a
probable central place complex located in the
mouth of the Rhine.241 Therefore, it could have
been at the invitation of a local or regional ruler
that a Merovingian beadmaker travelled to the
settlement to produce fashionable glass beads.
In the second phase (720-890) a new type
of boat-shaped house appears; apart from its
shape, there was a difference in orientation
pointing to a different organisation of the plots.
It is not clear whether there was a short hiatus in
habitation. In the Carolingian period a chapel
and a cemetery were added.
Settlement traces which can be identified
with the fortress of Rinasburg date between
890-1050 (phase 3), followed by the building of a
new church accompanied by a farm or possibly a
rectory (1050-1130, phase 4) and a Benedictine
nunnery in the 12th century (phase 5).
Glass production waste: Virtually all waste from
glass bead production debris was found in a
feature which mainly consisted of fired clay,
the remains of a hearth or possibly a
rudimentary glass furnace. There may be a
connection with one of the small buildings in the
immediate surroundings, a possible home of a
smith and his family. Glass bead production took
place during the second generation (phase 1b,
c. 610–640) or third generation of building
(phase 1c, c. 640–680).
The glass bead production waste (objects)
consists of finished, unfinished and failed beads
(n=68), glass rods (n=45), punty glass from a
beadmaker’s tool (n=3), crucibles (n=8) and one
undiagnosed fragment. Two lumps of fired clay
covered with translucent greenish glass may be
from the furnace floor.
Several categories represented among the
bead production waste from MaastrichtJodenstraat (see Section 3.2.1), such as glass
drops and pulled threads, are missing at
Rijnsburg. Scrap glass is also lacking. The latter is
perhaps a coincidence, but it was noted that not
a single fragment of Merovingian glass vessel
was found in the entire settlement or in the
nearby cemetery.
Excluding the crucibles, the waste categories
are dominated by opaque yellow glass (an average
of 49.6%), followed by opaque white (24.8%),
opaque red (18.8%), opaque turquoise (5.1%)
and opaque orange (1.7%).
Beads that were produced in the settlement
include monochrome globular, bi- and tri-globular
beads of opaque yellow glass, and bi-globular
beads of red glass. Trailed beads include biglobular beads of red glass with both white
crossing trails and a white spiral, tri-globular
beads of opaque red glass with opaque
white crossing trails and white beads with
translucent blue crossing trails.
In the context of the beads from the
cemetery of Schretzheim, Koch was quoted as
stating that beads with narrow crossing trails
represent billige Massenware and are ubiquitous
in necklaces of the later 6th and 7th centuries.242
This date can now be refined by a more recent
bead typology developed for beads from
cemeteries in Belgium by Pion and Vrielynck,
Mathis and Pion. 243 The above-mentioned bead
types are all typical for Pion’s Bead period 4
(600–640) (Table 3.7). Given the types of beads
that were produced here, it is therefore most
likely that the rudimentary furnace was in use
during settlement phase 1b (c. 610-640).
Most of the beads were split during
manufacture, either due to overheating of the
glass or not annealing the beads properly after
manufacture, a common occurrence on beadmaking sites (see Section 3.2.1 and Section 3.12).
51
—
Table 3.7 Rijnsburg-Abdijterrein: A selection of locally produced beads, their typology
and bead periods (Pion 2014; Vrielynck, Mathis & Pion 2018).
Form
Colour
Decoration
Type
Period
Bi-globular
opaque yellow
-
B1.2-1b
P4
Tri-globular
opaque yellow
-
B1.2-1c
P4
Bi-globular
opaque red
-
B1.2-2b
P4
Bi-globular
opaque red
white crossing trails & 1 white spiral
B5.2-1h
P4
Tri-globular
opaque red
white crossing trails
B3.2-1c
P4
Globular
opaque white
translucent blue crossing trails
B3.3-3a
P4
Period: P4=610-640 AD.
Chemical analysis of the glass from
Rijnsburg-Abdijterrein indicates that the rods
are of a very similar composition to the beads
and therefore that the beads are very likely to
have been made from the rods on site.
Furthermore, the opaque yellow glass from the
crucible fragments – although not associated
with the production waste from the furnace –
is proven to be of the same general chemical
lead-stannate composition as the opaque
yellow glass production waste.
Five fragments of glass-bearing crucibles
were retrieved from different contexts within
the Merovingian settlement, although one may
derive from a section through the furnace.
During the campaign of 1963 three fragments of
crucibles were found, of which two smaller
fragments derive from contexts which had
intrusions from later periods. There are six
base fragments and two body fragments.
These belong to coarse-ware cooking pots
(Wölbwandtöpfe) which were deliberately broken
to obtain their bases for use as shallow, dish-like
crucibles, comparable to those from the
Maastricht-Jodenstraat site (see Section 3.2.1).
When initially investigated, the crucibles
were thought to have glass attached to them.
However, no detailed scientific analysis was
carried out as part of this project to ascertain
whether this material is glass or not. A base
covered with what appears to be yellow glass on
the inside and on the fracture may come from a
section through the ‘furnace’. A body fragment
has what appears to be opaque yellow glass
over a white layer, together with a greenish spot
with streaked colourless and yellowish glass-like
material lying over it; the streaked material
covers the fracture while opaque yellow and
yellowish/white spots can be seen on the
outside. A small base fragment has opaque
yellow material sticking to the outside of the
base and colourless glass with small opaque
yellow spots on the inside; a second small base
has the same characteristics. A body fragment
has a thin layer of colourless glass on the inside
and an irregular, bubbly glass layer on the outside,
indicating this was probably overheated and
bubbled over. None of the fragments show any
traces of vitrification. The remaining three bases
show traces of vitrification on the outside and
may have been used for glass- or metal-working.
It was suggested that a colourless base
glass was modified on site using lead-tin-yellow
pigment. Further scientific research needs to be
carried out in order to investigate/ confirm
whether fully formed yellow glass is present.
3.9
Valkenburg-De Woerd, Zuid-Holland
Province
Site:
15
Site type:
Riverine settlement
Province:
Zuid-Holland
Municipality: Katwijk
Place:
Valkenburg
Toponym: De Woerd
Start date: 525
End date:
950
Description: Excavations between 1986 and 1988
by the former National Service for Archaeological
Heritage (ROB, now RCE) revealed the remains
of an early medieval settlement at Valkenburg-De
Woerd. The provisional findings were published
in 1987 and 1990.244 The settlement was laid out
244 Bult & Hallewas 1987; Bult, Van
Doesburg & Hallewas 1990.
52
—
245
246
247
248
Jezeer & Jongma 2002 (in Dijkstra 2011).
Dijkstra 2011, 172.
Magendans & Waasdorp 1989.
Magendans & Waasdorp 1989, 33; see
also Dijkstra, Sablerolles & Henderson
2011, 192.
on a natural levee along the inner curve of a
meander of the Oude Rijn, between gulleys on
either side. The settlement was just a few
kilometers away from the early medieval
settlement at Oegstgeest-Nieuw Rhijngeest Zuid
(Rijnfront) on the opposite side of the river.
The history of Valkenburg-De Woerd begins
at the establishment of the Roman limes. In the
mid-first century CE a military entrepôt harbour
was laid out here, which must have been part of
the vicus of castellum Valkenburg. Roman occupation
ceased around 230 AD. Ceramic finds from a
transect cut across a gulley of the Oude Rijn date
between 525 and 950, with most finds dating to
the 8th–9th centuries.245
The shore was divided into plots laid out on
a grid at right angles to the river, similar to the
situation at Oegstgeest-Nieuw Rhijngeest Zuid
(Rijnfront) (see Section 3.7) and RijnsburgAbdijterrein (see Section 3.8). There were
probably six to eight yards (width c. 50 m)
divided by ditches, simultaneously at any given
time during the Merovingian and Carolingian
periods. The plans of the buildings are very
unusual, mostly two-aisled, and are difficult to
interpret. Farms like those found at RijnsburgAbdijterrein and Oegstgeest Nieuw Rhijngeest
Zuid (Rijnfront) are lacking. Dijkstra points out
that in the Merovingian period, two-aisled
buildings functioned as barns and he hypothesises
that the buildings on De Woerd may have
combined two functions: traders may have lived
and worked in them, while the buildings were
used to store products or agricultural produce
during the trading high-season, drawing a
comparison with two- and three-aisled buildings
on the dams in the Dorestad harbour.246
In the south-eastern part of the settlement
the remains of revetments and a jetty were
found, underlining the importance of the river as
a mode of transport. There is evidence of boneand antler-working and of livestock rearing.
Glass production waste: A glass crucible
fragment (Find No. 510-4-307) was found in one
of the trenches (trench 510) cut across the river.
The crucible is covered on the inside with a thin,
even layer of translucent pale greenish glass.
A recent examination of the crucible by Epko
Bult, University of Leiden, revealed it is a lower
body fragment of a Merovingian Wölbwandtopf
dating to the 7th rather than the 6th century.
The crucible may, therefore, be contemporary
with the crucibles from the nearby settlement at
Rijnsburg-Abdijterrein and could be linked to
bead-making (see Section 3.8).
3.10 Den Haag-Frankenslag, ZuidHolland Province
Site:
16
Site type:
Coastal settlement
Province:
Zuid-Holland
Municipality: Den Haag
Place:
Den Haag
Toponym: Frankenslag (Johan van
Oldenbarneveltlaan 91–95)
Start date: 500–550
End date:
around 700
Description: Small-scale excavations (385 m2)
carried out by the municipality of Den Haag in
1984 yielded part of a Merovingian settlement
located on the eastern side of a coastal barrier.247
The settlement started in the first half of the
6th century and ended in the late 7th or early
8th century. Shortly afterwards, there is evidence
for arable farming until the settlement was
covered by drift sands (the Younger Dune
formation phase-0). The remains of pits,
hearths, three houses, and two successive
sunken huts were found which were probably
used for weaving.
The inhabitants grew rye and barley on the
nutrient-poor sandy soils, and reared cattle and
sheep. They supplemented their diet with locally
caught marine and riverine fish, game and wild
fruit. Locally sourced bog iron was processed for
the production of iron. Finds of Rhenish pottery
and millstones, bronze and lead are believed to
have been obtained by generating agricultural
surplus. The Meuse and Rhine river systems
could have been accessed over land (by way of
the coastal barriers or the beach) or by sea.
Glass production waste(?): A few sherds of brittle
hand-made pottery were found in a sunken hut.
They are covered on the inside and outside with
dark, deep blue-green ‘glass’, perhaps due to
vitrification of the fabric of the crucible.248 It is
not clear if these fragments represent waste
from glass- or metal-working.
53
—
3.11
Bloemendaal-Groot-Olmen, NoordHolland Province
Site:
17
Site type:
Coastal settlement
Province:
Noord-Holland
Municipality: Bloemendaal
Place:
Bloemendaal
Toponym: Groot Olmen
Start date: 675
End date:
850
Description: In the dunes of the National Park
Zuid-Kennemerland near Bloemendaal early
medieval remains were found at 14 different
locations.249 The remains, which had been buried
under the Younger Dunes (formed between 1200
and 1600), appeared when the area was restored
to its former ‘driftsand’ state by de-turfing.
In 2006 and 2007 Hollandia excavated a
settlement (location 1-3) dating between the
5th and 7th centuries. A survey combined with
some small trial trenches carried out by the
ROB prior to the Hollandia excavations showed
that locations 4, 5, 8 and 14 were in use during
the 8th and early 9th century. Hollandia excavated
locations 8 and 14 which were part of the
same settlement.
In total, seven buildings were discovered,
one barrel-lined well and remnants of fences.
Site 8, where habitation layers were partially
preserved, represented a single building dated
to the 9th century. Site 14 yielded the remains of
six buildings, including three house plans with
a distinct boat-shaped form comparable to
‘urban farms’ found in Dorestad-De Heul.
This imported building tradition most probably
originated in the central riverine area and
the Veluwe.
Evidence points to the agrarian nature of
the settlements, while marine fish and molluscs
played a more important role in the diet than
in the older settlement (location 1-3). Pottery,
glass, millstones and whetstones were imported
from the Rhineland and the Eifel. In the 8th century
the North-Holland coastal region was incorporated
into the Carolingian empire and it has been
suggested by de Koning that the settlement at
Bloemendaal-Groot Olmen may have been
connected to a royal domain which, according to
historical sources, was located in the area
around nearby Velsen.250
Glass production waste: A surface find of an
opaque dark blue tessera was found near
location 14 (8th–9th century).251 This location
also yielded a few fragments of thick-walled,
blue-green Roman glass which could be cullet
intended for recycling, perhaps to make the kind
of globular ‘bottle’ blue-green bead that was
also recovered from this location.252
3.12 Wijnaldum-Tjitsma, Friesland
Province
Site:
18
Site type:
Terp settlement
Province:
Friesland
Municipality: Harlingen
Place:
Wijnaldum
Toponym: Tjitsma
Start date: c. 50 AD
End date:
950–12th century?
Description: In 1990 fragments of a 7th century
gold cloisonné royal brooch were found in a
field on the Tjitsma terp near present-day
Wijnaldum by metal detecting. Its footplate had
already been found by chance in the 1950s.
These finds were the catalyst for the excavations
that were carried out by the Universities of
Groningen and Amsterdam on the eastern crest
of the terp settlement between 1991 and 1993.253
Although they yielded a wealth of information,
no tangible remains of the king or a royal residence
were found. A second volume on the ceramic
assemblage was published in 2014.254
The early medieval artificial mound or terp
settlement at Wijnaldum was located on a salt
marsh ridge which was oriented east-west.
There is evidence it was settled since the
1st century AD. It was one of a number of closely
spaced terps by the salt marsh which was open
to the sea. It is assumed that it was quite densely
populated since the Roman period, including
during the early middle ages.
A recent field survey has shown that the
beginning of habitation probably started as early
as the 1st century.255 The end of habitation on the
terp may have come in the 12th century when the
last farmstead may have moved to a separate
house terp, just like other farms in the
terp region.
The heyday of the terp settlement was the
period between 550 and 650 when the area
249
250
251
252
De Koning 2015.
De Koning 2015.
De Koning 2015, 317–318, afb. 11.6.
Sablerolles & De Koning 2015, 311–316,
afb. 11.1, 3, 4, 10.
253 Besteman et al.1999.
254 Nieuwhof 2020.
255 Kaspers 2020.
54
—
surrounding Wijnaldum, northern Westergo,
had developed into the centre of a kingdom
that covered the entire terp region of the
northern Netherlands. According to Nicolay,
the distribution of gold jewellery in a distinctive
style suggests that the king residing at or near
Wijnaldum had retainers across this entire area
(see Section 3.13).256 The evidence for glass
working on the terp dates to this period and it
seems likely that the elite status of the
settlement played a role in attracting a travelling
Merovingian beadmaker to visit the settlement.
Traces of habitation in this period are
modest though and include the remains of
six buildings divided over four households:
four (possible) sod houses, two granaries and a
sunken hut. Each house was built on a house
platform built from sods. The houses were N-S
orientated towards two large boundary ditches
running east-west immediately south of the
platforms. One (possible) house yielded
evidence for two hearths and evidence for
metal-working.
Wheel-thrown pottery imported from the
Rhineland makes up 63.7% of the total ceramic
assemblage on the terp during this period and it
is thought that Wijnaldum or northern Westergo
was a distribution centre for Merovingian
pottery; traders of Frankish goods such as
pottery (or its contents) may have depended on
the Wijnaldum elite for access to markets in the
northern coastal area.
The northern Netherlands became
incorporated into the Frankish empire during the
8th century, an area equivalent to present-day
Friesland in 734, and Groningen in 784 AD;
northern Westergo was no longer the political
centre controlling the area. An increasingly more
256 Nicolay 2014, 20–23.
257 Sablerolles 1999, 263–266.
even distribution of imported Carolingian
pottery across the northern coastal area probably
shows that traders were able to access the area
and were no longer obstructed or controlled by
the political centre. A reflection of this was the
percentage of imported pottery at Wijnaldum
during the Carolingian period, which increases to
c. 13.3%. Habitation was concentrated in the
south-eastern part of the excavated area of the
site during the Carolingian period, on the
southern flank of the terp. The highest parts of
the terp were used as arable fields, also found
on other terps during the 1st millennium.
Glass production waste: Glass-working evidence is
sparse.257 The most important object is a very
thick fragment of baked clay, possibly part of a
tray or a glass furnace, covered with a thick layer
(1.0–1.3 cm) of weathered opaque yellow glass
which has permeated through the porous,
pinkish-orange fabric (fig. 3.15). It was found
amongst waste from metal-working by a
blacksmith/bronze-caster. The dump is very
closely dated to the last quarter of the 6th and
the first quarter of the 7th century and is
contemporary with the glass-working evidence
from the Jodenstraat site in Maastricht. Two
small (flattened) globular beads of opaque
yellow and white glass accompanied this find
and are among the likely local products. This
type of bead was also found among the bead
production waste from the Jodenstraat site in
Maastricht, where only yellow examples are
represented.
Many glass beads found on the terp were in one
of two large boundary ditches (550–600) and
many of these simple, wound beads – including
small flattened globular beads of opaque yellow,
Fig. 3.15 Wijnaldum-Tjitsma: Detail of possible furnace floor or tray with opaque yellow glass permeating through
the fabric (Photograph: Henk Faber Bulthuis, Noordelijk Archeologisch Depot, Nuis).
55
—
red and white glass – could have been local
products.258 Several halves of short cylindrical
beads were retrieved from the above-mentioned
large boundary ditch. They are split lengthwise,
along the perforation, and are probably failed
beads due either to overheating of the glass or
as a result of not annealing the beads properly,
causing them to crack (see Section 3.2.1).
The same ditch also yielded three fragments of
unworked amber, suggesting glass and amber
bead-making could have been carried out at the
same time.259 An inhumation burial on the terp
from 550–600 AD contained a necklace with at
least 22 small, rather roughly shaped amber
beads which were perhaps made locally.260
A transverse breaking splinter of an opaque
greenish-white rod (see Section 3.2.1) comes
from a 5th century context. It is, however, not
securely dated, so perhaps this fragment is
contemporary with the above-mentioned
furnace or tray fragment. Moreover, almost all
context-dated opaque white beads from the
terp date to the second half of the 6th century or
between 575 and 625.
In view of the paucity of the material, this
production waste was interpreted as relating to
just one production event. Because the glass
waste production was found among that of a
bronze-caster, it was suggested that beadmaking could have been a secondary activity
carried out by, for instance, a bronze-caster or
a gold- or silversmith who visited the terp
occasionally. The possibility of a travelling
beadmaker, however, cannot be excluded as it
would be logical for such a craftsman to seek out
(more) permanent high-temperature craftsmen
on the terp.
There is also some very limited evidence for
bead-making on the terp in the Carolingian
period. It consists of a fragment of translucent
deeply coloured blue-green (turquoise) ‘punty’
glass from around a beadmaker’s tool. It is from
a context with a reliable date between 750–800.
An opaque yellow tessera from a ditch is
probably dated between 750–770.261 The tessera
clearly shows thin swirling layers of colourless
glass within the yellow matrix, indicating that
the yellow opacifier is not fully homogenized
with the translucent base glass (fig. 3.16).
2:1
0
2,5cm
Fig. 3.16 Wijnaldum: Opaque yellow tessera of
opaque yellow glass streaked with colourless glass.
The dimensions of the yellow tessera are: length
1.25 cm; height 0.81 cm and width 0.97 cm.
(Photograph: Henk Faber Bulthuis, Noordelijk
Archeologisch Depot, Nuis).
These two fragments do, of course, not constitute
solid evidence for bead-making on the terp in
the second half of the 8th century, but they at
least raise the possibility, especially in view of
the recently published tesserae finds from the
terp of Wierum (see Section 3.13).262
3.13 Wierum, Groningen Province
Site:
19
Site type:
Terp settlement
Province:
Groningen
Municipality: Winsum
Place:
Wierum
Toponym: Wierum
Start date: c. 400 BC
End date:
late middle ages.
Description: The largest find of Roman tesserae
in the Netherlands originates from the terp of
Wierum near Wierumerschouw (Groningen
Province) in the northern coastal region,
which was a frequently flooded salt-marsh
area.263 The terp was located on the wide river
Hunze, later renamed Reitdiep, which connected
the Wadden Sea and North Sea with inland
locations . The find is regrettably without a
context and is likely to have been discovered
between 1912 and 1916 when an estimated 3.5 ha
of the original 5 ha of the site was dug
commercially for its fertile soil. Only c. 1.5 ha of
the original terp remained. In addition to the
results of a coring programme, that provided
information on the original circumference and
the subsoil of the terp, in 1983 an overview of
the finds was published.264
258 These finds come from boundary ditch
1233 (Sablerolles 1999, 270–273 passim).
Sablerolles 1999, 277, cat. nr 226–228.
Sablerolles 1999, 276, cat nr 191–213.
Sablerolles 1999, cat nr 216.
Crocco et al. 2021.
Nieuwhof 2006; Crocco et al. 2021 and
references therein .
264 Miedema 1983.
259
260
261
262
263
56
—
265 Nicolay 2014.
266 Preiß 2010, 124, 130 number 47; Callmer
& Henderson 1991, Table 1C, 1.
In 2004 Groningen Province decided to
restore the terp to its original size and shape
using soil dredged from the river Reitdiep.
Archaeological excavations carried out by the
Groninger Instituut voor Archeologie (GIA) of
the University of Groningen (Rijksuniversiteit
Groningen) revealed that the site was inhabited
from the 4th century BC or slightly earlier, until
at least the late middle ages, probably with an
interruption in the 4th century AD. Unfortunately,
no farmhouses and outbuildings or artisanal
areas were excavated. It is unlikely that there
was much labour specialisation because Wierum
was mainly a self-sufficient agricultural
settlement, like all terps.
The find of a crescent-shaped gold pendant
suggests that one of the retainers of the king
who resided in or near Wijnaldum lived at
Wierum in the Merovingian period (see Section
3.12).265 During the 8th or possible the 9th century
the region became incorporated into the Frankish
empire. In the course of the Merovingian period a
local leader may have made the settlement of
Wierum his home. Because of its advantageous
position on the river Hunze, it is surmised that
by the Carolingian period it may still have had
regional political significance. The combination
of its convenient location and political status
may have attracted itinerant craftsmen,
including beadmakers.
Glass production waste: The assemblage has been
interpreted as a supply of ‘raw’ glass of an early
medieval glass beadmaker, most likely active on
the terp in the 8th/9th century. This may have
been a travelling beadmaker visiting terp sites
such as Wijnaldum in the northern coastal
region, which was most easily accessible by boat
from the central riverine area with Dorestad at
its centre.
The glass finds are dominated by
(fragments of) 201 tesserae. Most tesserae are
affected by heat: something which can be the
result of having been in a high-temperature
workshop environment. Other glass finds are
made up of five fragments of highly coloured
early Roman vessel glass, one fragment of
possible naturally tinted Roman or early medieval
vessel glass, three plano-convex drops of
opaque green glass, almost colourless glass and
translucent dark blue glass, and 13 irregular
drops/melted lumps of (almost) colourless, pale
green and pale blue-green glass. The latter may
be recycled Roman vessel glass or Roman gold-
foil tesserae stripped of their gold-foil. A small,
matt grey sphere may be a globular glass drop
of a type also found at bead-making sites of
Maastricht-Rijksarchief (see Section 3.2.3),
Wijk bij Duurstede (Dorestad) (see Section 3.4),
and Åhus in Sweden.266 Chemical analyses
confirm the Roman date of the analysed
glass finds.
Apart from the glass finds, the assemblage
includes four stone tesserae: two of green
porphyry, one of purple porphyry and a white
tessera, probably white marble, which is still
embedded in mortar showing it was robbed
from an ancient mosaic. Three more tesserae
show the remains of mortar adhering to one
side. Furthermore, there is a fragment of
Egyptian blue and two fragments of amber.
The latter may indicate that the production of
glass and amber beads was closely linked.
Two fragments of basalt may derive from
millstones imported from the Eifel. It is argued
that the stone tesserae and the Egyptian blue
pebble could have been collected accidentally
with glass tesserae during the frequent
spoliation of lavishly decorated Roman
buildings. The Egyptian blue and highly coloured
vessel fragments may even indicate that the
collection originates from one or more buildings
that contained a combination of first century AD
shell mosaics, early glass mosaics and glass
tesserae mosaics, or transitional forms thereof.
3.14 Deventer-Stadhuiskwartier,
Overijssel Province
Site:
20
Site type:
(Proto) urban settlement
Province:
Overijssel
Municipality: Deventer
Place:
Deventer
Toponym: Stadhuiskwartier
Start date: c. 850
End date:
c. 1200 (thereafter medieval city)
Description: Deventer is situated in the east of the
Netherlands, on the river IJssel, a tributary of the
Rhine, which flowed into Lake Almere, now the
IJsselmeer, which gave access to the Wadden Sea
and the North Sea.
The excavations in the inner city of Deventer,
project 312 (2007–2009) and project 434 (2012–
2013) revealed multi-period occupation,
57
—
including a late-mesolithic camp, lateprehistoric settlement traces and especially
many remains of the medieval city and its early
medieval predecessor.267
The earliest phase of the medieval
settlement consists of several scattered buildings
and a layer of arable land dating to the 8th and
early 9th centuries. In the third quarter of the
9th century the land was reorganized on a large
scale. The area was levelled and divided into
new, regular plots. This development can be
seen as the start of the process of urbanisation
in Deventer. A new type of urban house was
introduced which is clearly different from
farmhouses in the surrounding agrarian
settlements, e.g. they lack a stable. Many floor
remains of this type of house were found, as
well as a large number of cesspits, waste pits
and wells. The finds indicate a large increase in
craft activities in the late 9th century. At the end
of the 9th century a defensive rampart was
constructed around the settlement.
From the 10th century onwards several
timber houses with cellars were present as well
as secondary buildings with cellars which
probably had an artisanal function. From the late
9th and 10th centuries there is evidence for boneworking, iron-working (smithing slags) and
textile production from different locations.
Production waste from different crafts is found
together in the same waste pits on the same plots.
During the 10th and especially the 11th century
large tuff (stone) houses appear. Initially, the tuff
is sourced from old Roman building material,
transported along the Rhine from the Roman
fort at Xanten, Germany. In the late medieval
period the area developed as the centre of the
medieval city with a town hall and houses
belonging to members of the urban elite.
The glass finds include vessel glass,
window glass and some glass beads. Among
the vessels are fragments of very thinly blown
funnel beakers which are mostly made of a wellpreserved, clear bluish-green glass. There are
also fragments of thick-walled, curved vessels
of heavily weathered glass. The window glass is
mostly made of heavily weathered light glass
that is greenish where it is possible to see the
colour. Several quarries have preserved sides
that were nibbled with a grozing iron in order to
give them a distinct shape. The quarries would
have been mounted in lead strips.
Glass production waste: Glass production waste is
scanty and dates between c. 850 and c. 1050
(unpublished data). All glass production waste
products were found in waste pits or cesspits.
Two pits dating between 900 and 925 (project
312, K60 and K74) and two pits dating to the first
half of the 9th century (project 343, K116 and K174)
also yielded production waste of smithing,
bone-working and textile production.
A hollow, glassy slag dates to 850–900
(434/16203) and is very similar to glassy slags
found in a 10th century glass workshop in La
Milesse (Sarthe, France) where wood ash glass
was made from raw materials and blown into
glass vessels.268 Three fragments date between
900 and 925. A heavily weathered chip of glass
with a conchoidal fracture and with
characteristic concentric ribs (312/29057) was
struck off a larger chunk of raw glass. A heavily
weathered fragment has one convex surface and
is more or less triangular in section (312/29028).
It may be a transverse breaking splinter struck
off a glass ingot with at least one curved side. A
small, heat-affected fragment (312/29028) may
be part of a pulled thread, but this is not certain.
Three glass production waste fragments
date to the period 900–950. A small lump of
translucent clear bluish-green raw glass
(project 434/99144) has a conchoidal fracture
and was struck off a larger chunk of raw glass.
It is of a similar quality and colour to a funnel
beaker fragment with optic blown oblique ribs
(project 434/10380). A heavily weathered
fragment with a triangular section (project
434/99154) is similar to fragment 312/29028 and
may be a transverse breaking splinter. A heavily
weathered fragment with two irregular, heataffected surfaces (project 434/99154) could be a
partially melted chip of raw glass.
An intriguing fragment (project 312/29048)
dating to 950–1050 is difficult to interpret.
It consists of two layers of translucent bright
bluish-green and deep turquoise glass covered
by a very thin film of opaque red glass.
The fragment has two irregular surfaces which
are heat-affected, probably as a result of being
in a high-temperature glass workshop
environment. The turquoise colour is very
similar to that of a contemporary fragment of
very thin flat glass, either window glass or a
glass inlay with very fine grozing, from the same
area (project 312/29063). A fragment of a deep
turquoise quarry dating between 900–950
267 All information about the excavations
has been kindly provided by Emile
Mittendorff, Project leader Archaeology,
Deventer.
268 Cf. Raux et al. 2015, Fig. 3F.
58
—
269 Isings 1957, 113–114.
comes from another location (project 434
434/99139).
A knocked-off rim of a late Roman yellowgreen cup of Isings type 96a269 (project 434/
99116) dated to 900–950 may be cullet intended
to be melted down for the production of vessel
or possibly window glass. It is not unusual to
find old glass among early medieval glass
production waste (see for instance MaastrichtJodenstraat Section 3.2.1). An alternative
explanation is that Roman glass was accidentally
mixed in with Roman pottery (Samian ware) and
tegulae fragments which are regularly found in
Deventer in 10th–12th century contexts. It is
thought that the tegulae and possibly (part of)
the pottery had been transported to Deventer
together with the Roman tuff that was reused to
build stone houses. Part of the tegulae have
stamps proving they were made in Xanten.
There are as yet no indications for Roman
habitation in Deventer.
59
—
4 The materials, analytical
techniques and methodology
4.1
Introduction
This chapter provides a brief introduction to
the research samples studied and describes the
analytical techniques used and methodology
applied in the scientific analysis of the samples.
The rationale of how the analytical data collected
were used for the interpretation and comparison
with previously published data is also explained.
4.2
An overview of the sites and
glass samples
A detailed description of the evidence for the
Dutch early medieval glass-working is given in
Chapter 3.
One of the earliest sites providing glass for
this project is Gennep in the province of Limburg.
It is a 5th century AD Frankish settlement probably
founded around 400 AD located on a high river
terrace overlooking the confluence of the rivers
Meuse and Niers. It is located between the late
Roman fortress of Cuijk and the burgus of Asperden
on the Niers to the east.270 It did not yield any
evidence for glass working. Some 200 glass
vessels were found at the site which were mainly
table ware, mostly drinking vessels. The samples
analysed were all typical Frankish glass vessels
consisting of bowls and cones.271 They have
provided critical compositional data for an early
phase of the Dutch middle ages with which to
compare other early medieval glasses.
Excavations on nine Merovingian sites
mostly on the west bank of the river Meuse in
Maastricht has produced some of the most
comprehensive evidence for Early Medieval
glassworking yet found in Europe. The best
evidence for a glass industry was found during
excavations at the Jodenstraat (MAJO) site in
Maastricht (see Section 3.2.1). Evidence of glass
bead making, including 38 fragments of crucibles
containing opaque yellow and white glass were
found with more good evidence from the
Mabro site in Maastricht (see Section 3.2.2).
Crucibles containing glass were sampled along
Table 4.1 Photographs of crucibles from
Maastricht, Jodenstraat (MAJO) together
with their sample numbers.
Sample
MAJO 1
Sample number
Photo number
20
figure appendix IV.11
MAJO 2
21
figure appendix IV.12
MAJO 3
22
figure appendix IV.13
MAJO 4
23
figure appendix IV.14
MAJO 5
(inside)
24
figure appendix IV.15
MAJO 5
(outside)
24
figure appendix IV.16
MAJO 6
25
figure appendix IV.17
MAJO 7
26
figure appendix IV.18
MAJO 8
29
figure appendix IV.19
MAJO 9
30
figure appendix IV.20
MAJO 10
39-40
figure appendix IV.21
MAJO 11
41
figure appendix IV.22
MAJO 12
42
figure appendix IV.23
MAJO 13
43
figure appendix IV.24
MAJO 14
44
figure appendix IV.25
MAJO 15
45
figure appendix IV.26
MAJO 16
46
figure appendix IV.27
MAJO 17
47
figure appendix IV.28
MAJO 18
50
figure appendix IV.29
MAJO 19
51
figure appendix IV.30
MAJO 20
52
figure appendix IV.31
MAJO 21
53
figure appendix IV.32
MAJO 22
54
figure appendix IV.33
MAJO 23
58
figure appendix IV.34
MAJO 24
60
figure appendix IV.35
MAJO 25
61
figure appendix IV.36
MAJO 26
68
figure appendix IV.37
MAJO 27
73-74
figure appendix IV.38
with material from rods, bead fragments,
splinters, drops and punty glass. Samples from
both Jodenstraat and Mabro sites were analysed.
Tables 4.1 and 4.2 are list of sampled crucibles
from Jodenstraat and Mabro respectively,
together with their photograph numbers
provided here.
270 Brüggler 1994.
271 Sablerolles 1992; 1993.
60
—
Table 4.2 Photographs of crucibles from
Maastricht, Mabro together with their
sample numbers.
Sample
Sample number
Table 4.3 Photographs of crucibles from
Utrecht, Domplein together with their
sample numbers.
Photo number
Sample
Sample number
Photo number
MABRO 1
7
figure appendix IV.1
DOM 1
31
figure appendix IV.39
MABRO 2
8
figure appendix IV.2
DOM 2
32
figure appendix IV.40
MABRO 3
9
figure appendix IV.3
DOM 3
33
figure appendix IV.41
MABRO 4
10
figure appendix IV.4
DOM 4
34
figure appendix IV.42
MABRO 5
11
figure appendix IV.5
DOM 5
35
figure appendix IV.43
MABRO 6
12
figure appendix IV.6
DOM 6
36
figure appendix IV.44
MABRO 7
13
figure appendix IV.7
MABRO 8
14
figure appendix IV.8
MABRO 9
15
figure appendix IV.9
MABRO 10
16
figure appendix IV.10
The Wijnaldum-Tjitsma (henceforth Wijnaldum)
terp which has its heyday c. 550-650 AD when
the area surrounding Wijnaldum, northern
Westergo, had developed into the centre of a
kingdom that covered the entire terp region
of the northern Netherlands (see Section 3.12).
Many beads, a tessera, vessel fragments, a rod
and a possible furnace or thick tray fragment with
opaque yellow glass adhering, were analysed.
Archaeological investigations of the protourban site at Utrecht dating to c. 700-10th century
AD also produced 17 crucible fragments with
glass adhering (probably 8th-9th century Carolingian
pots) (see Section 3.5.1). These were found at
the Domplein site and sampled for this project
(a list if the crucible samples together with photo
numbers are given in Table 4.3). In addition,
scanty evidence for glass working was found at
the Utrecht Oudwijkerdwarsstraat site dating to
the 7th- first half of the 8th century AD. This glass
was also sampled (Table 4.4). Excavations at
the Carolingian site of Susteren-Salvatorplein
(henceforth Susteren), a monastic site, produced
two crucible fragments along with polychrome
beads, windows and vessels (see Section 3.3).
Excavations of the famous emporium of
Wijk bij Duurstede at the Hoogstraat and vicus
sites (henceforth Dorestad) dating to between
c. 600 and 900 AD produced a wide range of
glass artefacts. Those selected for scientific
analysis were mainly fragments of funnel
beakers, bowls and bell beakers but also
tesserae, linen smoothers.
The latest site which produced glass included in
this project was Deventer-Stadhuiskwartier
(henceforth Deventer, see Section 3.14). The site
dates to between c. 850 at the earliest and the
10th-11th centuries AD. Site excavations produced
raw chunks of glass as well as vessel glass, glass
beads and window glass.
The quantitative major and minor chemical
composition and trace element chemical
composition have been determined for each
glass sample studied in this work. Altogether
279 glass objects have been sampled in this
project. Neodymium and strontium isotopic
compositions were determined for 20 glass
samples which were selected based on their
chemical characteristics. The major, minor and
trace element compositions of our samples
constitute the primary data for this project.
The compositional group and/or formula group
for each sample have been identified using
certain compositional and isotopic
characteristics. The technical details of the
analytical methods used to produce chemical
and isotopic data are elaborated below.
Table 4.4 Photographs of glassworking
evidence from Utrecht,
Oudwijkerdwarsstraat.
Sample
Sample number
Photo number
OUDWIJ 1
78
figure appendix IV.45
OUDWIJ 2
79
figure appendix IV.46
61
—
4.3
Electron probe microanalysis
(EPMA) for major and minor
chemical composition
4.4
Quantitative major and minor chemical
compositions of our samples were determined
on the JEOL JXA-8200 electron microprobe
housed in Nanoscale and Microscale Research
Centre, University of Nottingham. Fragments
of each ceramic shard were mounted in crosssection in epoxy resin blocks and polished to a
0.25 mm diamond paste finish so as to reveal
a fresh flat analytical surface. The blocks were
carbon coated to prevent surface charging and
distortion of the electron beam during analysis.
The EPMA system is equipped with four
wavelength-dispersive X-ray spectrometers
with LIF, TAP, PETJ and LIFH crystals, a single
energy dispersive X-ray spectrometer and both
secondary and backscattered detectors.
A defocused electron beam with a diameter
of 40 µm was used so as to prevent volatilization
of light elements such as sodium. The probe
was run at an accelerating voltage of 15 kV and
a beam current of 20 nA.
The system was calibrated with a mixture
of mineral and oxide standards. A ‘Phi-rho-z’
correction program was used to quantify the
results. The Corning B glass standard was
routinely used as a secondary standard to check
for accuracy and precision and to monitor any
drift in the instrument. The analytical precision
and accuracy achieved by using the Corning B
standard are listed in Table 4.5
Laser ablation inductively coupled
plasma mass spectrometry (LAICP-MS) analysis for trace element
compositions
The trace elemental compositions of our glass
samples were determined using the LA-ICP-MS
instrument consisting of either a NewWave
UP193FX excimer (193 nm) or Elemental Scientific
Lasers imageGEO (193nm) laser system and an
Agilent 7500cs series instrument housed in the
Analytical Geochemical Laboratories of British
Geological Survey. The same samples used by
EPMA major and minor chemical analysis are
analysed for their trace element compositions.
Prior to analysis the carbon coating of the samples
was removed and the samples were cleaned by
rubbing a tissue soaked in dilute acid over the
surface for a few seconds. The sample was
placed in a two volume ablation cell with a 0.8 L
min−1 He flow. In addition to the sample block,
NIST glass standards SRM610 and 612 as well as
USGS glasses standards GSD-1G and BCR-2G
were placed in the chamber. The UP193FX laser
was fired for 40s at 10 Hz using a beam diameter
of 70 μm; whereas the imageGEO was fired at
20Hz or 10s using a square 50 X 50 μm beam.
Fluence and irradiance as measured by the
internal monitor were typically 3 J/cm2 and 0.85
GW/cm2 respectively for both laser systems.
With the UP193FX laser prior to introduction into
the ICP-MS the He flow was mixed, via a
Y-junction, with 0.85 L min−1 Ar and 0.04 L min−1
N2 gas flows supplied by a Cetac Aridus
desolvating nebulizer. The Aridus allowed
introduction of ICP-MS tuning solutions and
optimization of the Aridus sweep gas (nominal 4
L min−1 Ar). During solid analysis by the laser, the
Aridus only aspirated air. The imageGEO system
mixed the argon gas as above but added the N2
Table 4.5 The recommended composition for the Corning B standard compared to average analytical results (n = 44)
and associated standard deviations and errors using the electron microprobe.
SiO2
Al2O3
Na2O
K2O
CaO
4.36
17
Measured
62.43
4.65
16.83
1.04
8.75
1
0.17
0.6
0.09
0.26
1.4
6.7
1
4
2.2
12.4
Error %
8.56
FeO
61.55
Standard deviation (n=44)
1
TiO2
Quoted
0.089
MnO
MgO
0.31
0.25
0.1
0.3
0.02
0.05
3.2
CoO
CuO
P2O5
Sb2O5
1.03
0.05
2.66
0.82
0.46
0.26
1.05
0.06
2.39
0.85
0.51
0.02
0.05
0.04
0.12
0.07
0.04
4
1.9
20
10.2
3.7
10.9
62
—
gas flow internally. Tuning was by rastering the
laser beam over the glass standards.
The ICP-MS instrument was set for a dwell
time of 7 ms for each of the 47 isotopes of
interest to give one time-slice Data were
collected in a continuous time resolved analysis
(TRA) fashion as a repetitive series of time-slices.
Prior to laser firing a period of at least 120 s of
‘gas blank’ was collected, then three ablations
being made on the SRM610; three ablations on
GSD-1G; 3 ablations on the SRM610; three ablations
on the BCR-2G, 3three ablations on up to
eight samples and finally three ablations on
the SRM610; three ablations on GSD-1G.
The SRM610 and GSD-1G were used to calibrate
the system whilst the SRM612 and BCR-2G were
used as a quality control (QC) materials. The full
quality control report of our trace element
analysis is listed in Table 4.6. Calibrations and
data reduction were performed using Elemental
Scientific Lasers Iolite4 software, with data
compilation in Microscoft Excel 2016.
The nature of laser ablation means that
there is some variability in ablation volume and
transport efficiency with different materials
(matrix effects). Therefore, accepted practice
is to normalize results to an internal standard
element; in the current study Si was chosen for
this purpose with its concentration being known
in the NIST glasses and provided by the EPMA
data for the study glasses.
The 26 trace element pattern has been used
in many recent publications to identify pristine
natron glass of different compositional types.
We have adopted this approach here in the
discussion of our results below.
4.5
272 Birck 1986.
Thermal ionization mass
spectrometry (TIMS) analysis to
determine Nd and Sr isotopic
compositions
For isotopic analysis, a small glass fragment was
first sampled and transferred to a clean (class
100, laminar flow) working area for further
preparation. In the clean laboratory, the samples
were cleaned ultrasonically in Milli-Q water,
dried on a hotplate and then weighed into
pre-cleaned Teflon beakers.
For Sr isotopic analysis, the samples were
spiked with 84Sr tracer solution and dissolved in
Teflon distilled 8M HNO3 and Ultrapure 29M HF.
Samples were converted to chloride form using
Teflon distilled 6M HCl. The samples were then
taken up in calibrated 2.5M HCl and centrifuged.
Strontium was collected using Eichrom AG50 X8
resin columns. Each sample was then loaded on
to a single Re filament with TaF, following the
method of Birck.272 The 87Sr/86Sr and Strontium
elemental concentrations were determined by
Thermal Ionization Mass spectroscopy (TIMS)
using a Thermo Triton multi-collector mass
spectrometer at the National Environmental
Isotope Facility of the British Geological Survey.
The international standard for 87Sr/86Sr, NBS987,
loaded in the same way, gave a value of 0.710259
± 0.000018 (n = 21, 2σ) during the analysis of
these samples, and sample data was normalized
to the accepted value for this standard of
0.710250. Procedural blank values were in the
region of 100 pg.
For Nd isotopic analysis, fractions were
dissolved in 1 ml of 2% HNO3 prior to analysis on
a Thermo-Electron Neptune mass spectrometer,
using a Cetac Aridus II desolvating nebulizer.
0.010 L min-1 of nitrogen were introduced via the
nebulizer in addition to argon in order to
minimize oxide formation. The instrument was
operated in static multi-collection mode, with
cups set to monitor 142Ce, 143Nd, 144Nd, 145Nd,
146Nd, 147Sm, 149Sm, and 150Nd. 1% dilutions of
each sample were tested prior to analysis, and
samples diluted to c. 20 ppb. Jet sample cones
and X-skimmer cones were used, giving a typical
signal of c. 800–1000 V/ppm Nd. Correction for
144Sm on the 144Nd peak was made using a ratio
for 147Sm/144Nd derived from multiple analyses of
SpecPur© samarium solution. This correction
was insignificant due to the efficiency of the
column separation. Data are reported relative to
146Nd/144Nd = 0.7219. The Nd standard solution
JND-i was analysed during each analytical
session and sample 143Nd/144Nd ratios are
reported relative to a value of 0.512115 for this
standard.
63
—
Table 4.6 Summary of quality control (QC) data for analysis of glass samples.
Reference material: SRM612
Element
Li
Number of analyses=101
Measured isotope
Nuber of analytical sessions=3
Expected concentration (mg/kg)
7
39.5
Mean concentration (mg/kg)
Standard deviation
40.2
RSD%
Error%
1.5
3.8
-2
B
11
37.3
34.3
2.7
7.2
9
Na
23
99780
103858
1867
2
-4
Mg
24
58
68
4
7
-15
Al
27
11102
11167
295
3
-1
P
28
73
46.6
152
208
56
K
31
53
62.3
4
8
-15
Ca
39
84382
85002
2298
3
-1
Ti
42
39.8
44
4
9.9
-9
V
47
37.8
38.8
1.5
3.9
-3
Cr
51
34.8
36.4
1.7
4.8
-4
Mn
52
38
38.7
1
4
-3
Fe
55
46
51
3
7
-11
Co
56
34.3
35.5
0.9
2.7
-3
Ni
59
37.7
38.8
1.4
3.7
-3
Cu
60
36.7
37.8
1.2
3.3
-3
Zn
63
37.6
39.1
2.2
6
-4
As
66
33.2
35.7
2.2
6.5
-7
Rb
75
31.3
31.4
0.8
2.5
0
Sr
85
76.9
78.4
2.9
3.7
-2
Y
88
38.4
38.3
1.1
3
0
Zr
89
38.5
37.9
1.1
2.9
2
Nb
90
38.3
38.9
1.2
3.2
-2
Mo
93
35.6
37.4
1.6
4.6
-5
Sn
95
37.6
38.6
1.6
4.1
-3
Sb
120
33.6
34.7
1
3.1
-3
Cs
121
41.2
42.7
1.1
2.6
-3
Ba
133
38.7
39.3
1
2.5
-1
La
138
35.7
36
0.9
2.6
-1
Ce
139
37.7
38.4
1.2
3.3
-2
Pr
140
37.3
37.9
1.2
3.3
-2
Nd
141
34.9
35.5
1.4
4
-2
Sm
146
37.3
37.7
1.5
4
-1
Eu
147
35
35.6
1.3
3.6
-2
Gd
153
37.7
37.3
1.4
3.7
1
Tb
157
36.6
37.6
0.9
2.6
-3
Dy
159
35.6
35.5
1.1
3.1
0
Ho
163
37.8
38.3
1
2.6
-1
Er
165
38.4
38
1
2.7
1
Tm
166
36.6
36.8
1.1
2.9
-1
Yb
169
38
39.2
1.2
3.2
-3
Lu
172
36.6
37
0.9
2.5
-1
Hf
175
37
36.7
1.2
3.4
1
Ta
178
36.9
37.6
1
2.6
-2
Pb
208
38.2
38.6
1.1
2.9
-1
Th
232
37.5
37.8
1.2
3.1
-1
U
238
36
37.4
1.2
3.4
-4
64
—
4.6
How analytical data is used in this
study
The type of glass whether natron, plant ash or
wood ash, can be easily identified by major and
minor oxide contents such as Na2O, K2O, CaO
and MgO. The majority of the glass studied here
is natron glass; wood ash glass and plant ash
glass only account for a small fraction of the
samples. One of the main aims of this study is to
categorize the majority which are natron glasses
according to different compositional types
(see Section 2.4.1). These are related to their
provenance, so we can gain an insight into raw
glass supply in the early medieval Netherlands.
Because a large number of samples are involved
in this study, the job of categorizing natron glass
samples into compositional groups has mainly
been achieved by using three plots, Al2O3/SiO2
against TiO2/Al2O3, Pb against Sb, and a 26 trace
element pattern.
Firstly the Al2O3/SiO2 against TiO2/Al2O3 plot
is used to provide a preliminarily classification of
natron glasses into compositional groups: TiO2,
Al2O3 and SiO2 essentially represent the heavy
mineral, feldspar and quartz contents of the
sands used for making the glass,273 which can
reflect their provenance very well (Figure 4.1).274
The Pb against Sb plot is used to show the levels
of impurities brought in by recycling of the glass
samples. In ‘pristine’ (non-recycled) natron glass
the levels of a few correlated elements such as
Pb, Sb and Cu are very low, but for recycled glass
the levels of these elements are much higher.
Pb and Sb (both in ppm) are the most consistent
demonstrators among these elements, so they
have been chosen to distinguish ‘pristine’ glass
samples from recycled glass samples.
The criterion for the identification of a ‘pristine’
glass is that the Pb and Sb contents are both
under 1000 ppm, following previous conventions.275
The compositional groups of ‘pristine’ glass
samples identified using the two previously
mentioned plots are then confirmed by using
the 26 trace element patterns of the samples.
Although it has been found that the rare earth
element patterns of all natron glasses tend to be
very similar, when lighter trace elements (excluding
some elements which may have been introduced
with the colourant, such as transition metals) are
included, the patterns of the four main different
compositional groups, HIMT sensu stricto, Foy 2,
Egyptian II and Levantine,276 can be distinguished
very well (Figure 4.2).277 The 26 trace elements
used in this study are V, Cr, Rb, Sr, Y, Zr, Nb, Cs,
Ba, La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm,
Yb, Lu, Hf, Th, U, and their normalized
concentrations compared to that for the upper
continental crust.278.
273 Freestone et al. 2018.
274 Data sources: Foy et al. 2003 (HIMT
275
276
277
278
sensu stricto, Levantine I), Schibille,
Sterrett-Krause & Freestone 2016 (Foy 2),
Schibille et al. 2019 (Egyptian I and
Egyptian II), Freestone et al. 2015
(Levantine II), Silvestri, Molin & Salviulo
2008 (Roman Mn and Roman Sb).
Foster & Jackson 2009.
The trace element patterns of Levantine
I glass and Levantine II glass are very
similar. Thus only the pattern of
Levantine I glass is shown here to
demonstrate its difference with that of
other compositional groups. It is
difficult to distinguish Egyptian I glass
and HIMT sensu stricto glass by their trace
element patterns. This it is not too
much of a problem here since Egyptian I
glass is not a significant compositional
group in northwestern Europe.
Schibille, Sterrett-Krause & Freestone
2016; Bertini, Henderson & Chenery
2020.
Kamber et al. 2005.
Figure 4.1 Al2O3/SiO2 against TiO2/Al2O3 plot showing the compositional differences between major groups of natron
glasses.
65
—
Figure 4.2 The 26 trace element patterns of four major compositional groups of natron glass. The trace elements are V, Cr, Rb, Sr, Y,
Zr, Nb, Cs, Ba, La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu, Hf, Th and U
5 The analytical results and discussion
5.1
Introduction
Glass samples from nine sites were studied,
including two sites in both Maastricht and
Utrecht. In this chapter the chemical compositional
features of glass samples from each site will be
elaborated separately. The major and minor
chemical compositions of analysed samples are
given in Appendix II. The trace element data of
analysed samples are given in Appendix III.
5.2
Glass samples from Maastricht
(Jodenstraat and Mabro sites)
Apart from two glass vessel fragments and three
window glass fragments, which may be glass
cullet, all the rest of the 51 glass samples from
Jodenstraat are the remains of on-site bead
making, and they all came from one pit filled
between late 6th century and early 7th century
AD.279 The glass samples from the bead-making
context can be roughly divided into three groups:
translucent naturally coloured and cobalt blue
glass waste, highly coloured opaque glass, and
vitreous materials attached to crucibles
(Appendix I).
67
—
5.2.1 Naturally coloured and cobalt blue
bead-making glass waste
The bead-making glass waste from Jodenstraat
includes glass drops, glass rods, punty glass,
glass fragments and glass attached to crucibles.
Because of the shape and state of the glass
samples found in the bead-making context,
and the similarity of their chemical compositions
with that of the glass beads, it has been suggested
that bead making involved firstly colouration of
the naturally coloured base glass followed by
further procedures to make the glass into beads.280
If this suggestion stands, the bead-making glass
waste should reflect the chemical features of the
base glass used by the bead makers. Their low
K2O and MgO and elevated Fe2O3, TiO2 and MnO
contents suggest that they can all be categorized
as the so-called HIMT natron glass dominating
northwest Europe during the 4th–7th century.
HIMT glass is a general name for early medieval
natron glass with elevated Fe2O3, TiO2 and MnO
contents (see Section 2.4.1).
There are some further compositional
groupings that can be demonstrated by the
ratios of TiO2, Al2O3 and SiO2, which represent
essentially the heavy mineral, feldspar and quartz
contents of the sands used for making the glass.
In Figure 5.1 we can see that the naturally
coloured and blue glass waste from Jodenstraat
forms one tight group, and it agrees very well
with the well understood Foy 2 compositional
group of natron glass (compare with Figure 4.1),
including one of the crucible fragments with pale
green glass attached (Joden 29) which is least
Figure 5.1 A plot of Al2O3/SiO2 versus TiO2/Al2O3 for Jodenstraat naturally coloured bead-making waste and highly
coloured opaque glass and beads.
279 Sablerolles, Henderson & Dijkman 1997.
280 Sablerolles, Henderson & Dijkman 1997.
68
—
Figure 5.2 The 26 trace element patterns for Jodenstraat naturally coloured and blue bead-making waste and
Jodenstraat highly coloured opaque glass and beads compared to that of Foy 2 glass.
contaminated. The latter shows no evidence that
lead or tin had been added (yet). The slightly
elevated PbO contents of some naturally coloured
glass waste may have been caused by
contamination during the production process since
lead was an important raw material ingredient in
bead making of this period in northwestern
Europe. The elevated PbO content of early
medieval glass could also indicate glass recycling.281
However, the Sb contents, which are correlated
with PbO in recycled early medieval glass, and
often used together with PbO as an indicator of
glass recycling, are all very low at ≤1000 ppm level.
281 Foster & Jackson 2009.
Therefore we are more inclined to believe that this
glass waste is ‘pristine’ glass rather than recycled,
and that the elevated PbO was introduced during
the production procedure. The 26 trace element
pattern in the naturally coloured and blue glass
waste from Jodenstraat agrees very well with that
for Foy 2 glass published previously (Figure 5.2)
confirming that they are Foy 2 glass. Amongst the
bead making waste there is a group of five blue
samples: four blue fragments (Joden 37-40) and
one translucent blue coloured bead fragment
(Joden 61). In the Al2O3/SiO2 against TiO2/Al2O3 plot,
these five samples locate in the area of Foy 3.2,
one of the two subgroups of Foy 2, the other
subgroup being Foy 2.1 (see Section 2.4.1). The flat
tapered bead fragment (Joden 61) does not contain
elevated PbO, nor is it opacified with SnO2 like the
many other highly coloured beads found at
Jodenstraat. The four cobalt blue fragments (Joden
37-40) and the cobalt blue bead (Joden 61) have
the same composition which suggests this type of
non-tin opacified blue beads was made directly
from ‘pristine’ cobalt blue coloured raw glass.
5.2.2 Highly coloured opaque glass
The highly coloured glasses found at Jodenstraat
are beads, rods and drops, potentially all linked to
the process of bead making. These highly coloured
glass samples are in four basic colours: yellow,
white, red and greenish-blue. They all have
elevated PbO contents ranging from a little over
2.0 wt% to over 40 wt% PbO, which demonstrates
that lead was used as one of the key raw materials
in bead making in addition to natron glass. As
shown in figure 5.3, the average PbO contents in
these yellow, white, red and greenish-blue glasses
are not quite the same especially for yellow glass:
its PbO contents are much higher than that of the
other three colours. The reason for the different
PbO concentrations in yellow glass and in other
colours is addressed in detail in Section 5.10.3
below.
Figure 5.3 The PbO contents of highly coloured opaque
glasses from Jodenstraat (G blue = greenish-blue).
69
—
a
b
c
d
e
f
Figure 5.4 Backscattered SEM images of yellow (top left: 5.4a), green (top right: 5.4b), white (middle left: 5.4c), red
(middle right: 5.4d) coloured glass from Jodenstraat and crystalline inclusions: SnO2 opacifiers found in white, green
and red colour glass (bottom left: 5.4.e), ground fayalitic slag surrounded by 0 valence micron sized particles in red
glass (bottom right: 5.4f ).
From the backscattered SEM images of the
highly coloured opaque glass samples from
Jodenstraat (Figure 5.4), we can see that they all
contain some highlighted (pale grey) inclusions
of variable sizes. The quantitative EPMA
analyses of these phases show that they differ
compositionally. The typical chemical
compositions of the three crystalline phases of
different highly coloured glasses are listed in
Table 5.1. According to their quantitative
chemical compositions and backscattered SEM
images, the crystalline phases in yellow glass can
be identified as lead tin yellow II (PbSn(Si)O3)
with a PbO/(SnO+SiO2) ratio of about 2:1
(Figure 5.4). Lead tin yellow II was widely used as
the colourant in yellow coloured beads in early
medieval Europe.282 Three types of crystalline
inclusions are found in the glass matrix of the
red coloured glass (Figure 5.4): firstly a phase
containing SnO2 of 50–70 wt% and variable SiO2
and PbO contents; secondly, a high iron phase,
which was also observed in similar medieval red
colour beads from England and identified as
ground fayalitic slag;283 finally, micron sized
particles of 0 valence metallic copper, which
are scattered evenly in the glass matrix.
282 Heck, Rehren & Hoffmann 2003.
283 Peake & Freestone 2012.
70
—
Table 5.1 The typical chemical compositions of the three crystalline phases found in
Jodenstraat highly coloured glasses.
SiO2 (wt%)
284 Bandiera et al. 2020.
Lead tin yellow II
SnO2 opacifiers
Fayalitic slag
8–18
15–30
10–25
Al2O3 (wt%)
1–3
2–3
<1
Na2O (wt%)
3–5
5–8
1–5
K2O (wt%)
-
-
CaO (wt%)
-
1
-
SnO2 (wt%)
15–24
50–70
-
PbO (wt%)
60–65
2–15
<1
FeO (wt%)
-
-
65–90
According to previous studies we can
conclude that the phase containing SnO2, the
dominant composition, acts as the opacifier in
the glass matrix. The red glass is mainly coloured
by micron sized particles of 0 valence copper,
like many other types of red glass and ceramic
glazes,284 and the ground fayalitic slag would
have acted as an internal reducing agent during
the formation of the copper particles.
The crystalline inclusions found in the white
glass and the greenish-blue glass beads are the
same (Figure 5.4), and they also contain the first
type of inclusion found in the red glass, mainly
SnO2 at 50–70%, with variable SiO2 and PbO %
levels, acting as the opacifier in the glass matrix.
The amount of SnO2 found in the white glass is
much higher than that found in the red and
greenish-blue glass, producing its opacified
white colour. The colour of the greenish-blue
glass is caused by a copper based colourant
(CuO), as indicated by high copper contents,
which would have dissolved in the glass matrix
so it cannot be observed as a separate phase in
the SEM images.
Jodenstraat are also very low at ≤1000 ppm, the
same as found in naturally coloured and blue
bead-making waste found on the site. These
results indicate that the base glass of
Jodenstraat highly coloured opaque glass is also
‘pristine’ Foy 2 glass and that the bead-making
debris formed during the production process
was produced using this same base glass.
The observation that Jodenstraat highly
coloured opaque glasses generally have higher
Al2O3 contents than the bead-making waste,
demonstrated by their clustering to the right of
colourless bead-making waste in Al2O3/SiO2
against TiO2/Al2O3 plot (Figure 5.1), can be
attributed to the addition of the lead tin yellow
II colourant and tin opacifiers during the
colouring process: the lead tin yellow II
colourant and tin opacifiers have higher Al2O3/
SiO2 than the base glass (Table 5.1).
In the plot of Al2O3/SiO2 against TiO2/Al2O3
(Figure 5.1), Jodenstraat highly coloured opaque
glass samples form a tight cluster in the area of
the Foy 2 compositional group (with Figure 4.1 as
reference), located slightly to the right of the
cluster of Jodenstraat colourless bead-making
waste. In Fig. 5.2 it can be seen that the 26 trace
element pattern of the averaged composition of
these highly coloured glass samples is identical
with that of Jodenstraat colourless bead-making
waste and the Foy 2 pattern published
previously. Additionally, the Sb contents of all
highly coloured opaque glass samples from
From twelve crucibles retrieved from the beadmaking context of Jodenstraat, three types of
vitreous materials attached to them have been
examined scientifically. They are naturally
coloured natron glass (two samples), naturally
coloured glass mixed with bright yellow residues
(eight samples), and a white melt with a light
yellow tinge (two samples). The description of
the vitreous residues attached to each crucible is
listed in Table 5.2.
Naturally coloured natron glass has been
found attached to crucibles Joden 21 and Joden
29 from the bead-making context of
5.2.3 Vitreous and semi-vitreous
materials attached to the crucibles
71
—
Jodenstraat, and their chemical compositions
have been discussed along with other naturally
coloured glass in Section 5.2.1: they conform to a
Foy 2 composition.
The naturally coloured glasses mixed with
bright yellow residues in 8 crucibles turned out to
be very pure lead oxide-silica glass. The Na2O
concentrations are very low in these lead glasses
(Appendix II), which shows that natron glass was
not involved in the procedure that produced these
lead glasses and the yellow/white residue mixture.
Yellow residues attached to crucibles from early
medieval northwest European sites have been
studied before285 and the analytical results for
yellow residues attached to crucibles from
Jodenstraat are the same, namely lead tin oxide.
Chemical compositions and SEM images
show that the main phase of the bright yellow
lead tin residue is lead tin yellow II (PbSn(Si)O3)
(Figure 5.5a), where the (SnO+SiO2) to PbO
weight ratio is close to 1:2. It has been suggested
that these crucibles containing lead tin yellow
residues are evidence of on-site production of
the yellow colourant which was then added to
base glass during the manufacture of yellow
beads.286 The chemical compositions and SEM
morphologies of the lead tin yellow II crystallites
attached to the Jodenstraat crucibles are very
similar to the lead tin yellow II found in the
yellow glass beads and yellow bead-making
debris from Jodenstraat. Therefore we also
suggest that these crucibles are remains of onsite lead tin yellow colourant production, and
that the lead tin yellow II produced in the
crucibles would have been used directly to
colour the base glass to create a yellow colour.
More details regarding the procedures of on-site
lead tin yellow II production and how the
crucibles were used during the process are
addressed in Section 5.10.2 below.
A white melt with a light yellow tinge has
been found attached to two crucibles (Joden 23
and Joden 30). Their chemical compositions show
that this white melt also contains SnO2, PbO and
SiO2 as the main components, but that their
weight ratios are quite different from that of lead
tin yellow II (Appendix II). SnO2 is the dominant
component, ranging from 50 wt% to 70 wt% in
different areas of the white melt; the PbO and
SiO2 contents are variable. The SEM morphology
of the tin white crystallites is also quite different
from that of lead tin yellow in that no lead silica
glass surrounds the tin oxide in the former
whereas it does in the latter (Figure 5.5b).
a
b
Table 5.2 Description of the vitreous residues attached to Jodenstraat
crucibles
Sample number
Vitreous residue description
Maastricht-Jodenstraat 19
lead yellow ii surrounded by pure lead glass
Maastricht-Jodenstraat 20
lead yellow ii surrounded by pure lead glass
Maastricht-Jodenstraat 21
natron glass
Maastricht-Jodenstraat 22
pure lead glass with small yellow spots
Maastricht-Jodenstraat 23
white melt
Maastricht-Jodenstraat 24
pure lead glass with small yellow spots
Maastricht-Jodenstraat 25
pure lead glass with small yellow spots
Maastricht-Jodenstraat 26
pure lead glass with small yellow spots
Maastricht-Jodenstraat 27
lead yellow ii surrounded by pure lead glass
Maastricht-Jodenstraat 28
lead yellow ii surrounded by pure lead glass
Maastricht-Jodenstraat 29
natron glass
Maastricht-Jodenstraat 30
white melt
Figure 5.5 Backscattered images of yellow residue (left: 5.5a) and white melt (right: 5.5b) found attached to crucibles
from Jodenstraat.
285 Henderson & Ivans 1992; Heck, Rehren
& Hoffmann 2003; Peake & Freestone
2014.
286 Heck, Rehren & Hoffmann 2003; Peake
and Freestone 2014.
72
—
Moreover, the chemical compositions and SEM
morphologies of this tin white melt are very
similar to that of the tin opacifiers in white, red
and greenish-blue beads. Therefore, we think
these white melts attached to crucibles are
evidence for on-site production of tin opacifiers
in early medieval northwestern Europe, and they
constitute the first such evidence to be reported.
More details of the separate production of the
tin white opacifier at Jodenstraat is addressed in
Section 5.10.3 below.
un-recycled (‘pristine’) glass. The 26 trace
element patterns of Joden 45 and Joden 46
confirm that they have the same patterns as
Foy 2 and the HIMT sensu stricto respectively
published previously, but the trace element
pattern of Joden 44 shows some clear
differences from that of the Foy 2 glass pattern
which may be related to the unusually high MnO
content (1.9%) in the sample (Figure 5.7a).
5.2.5 Crucibles from the Mabro site,
Maastricht
5.2.4 Glass artefacts
287 Silvestri, Molin & Salviulo 2008.
288 Pactat et al. 2017.
Five glass samples from Jodenstraat, three
window glass fragments (Joden 44–46) and
two glass vessel fragments (Joden 60, 68), may
be cullet. The elevated MnO, Fe2O3 and TiO2
contents suggest that they are HIMT natron
glass. Their Al2O3/SiO2 and TiO2/Al2O3 ratios show
that the window glass (Joden 46) belongs to the
HIMT sensu stricto compositional group, samples
44, 45 and 60 distribute in the area of the Foy 2
compositional group and sample 68 could be a
piece of Roman glass (Figure 5.6a with Figure 4.1
as reference). The form of Joden 68 also
suggests that it is a Roman vessel fragment
(thick naturally coloured ribbed green glass),
but it does not have high Mn or high Sb contents
normally found in decolourized and green
Roman glasses.287
The Sb and Pb contents of the five glass
samples show that apart from Joden 60,
which has Sb and Pb contents over 1000 ppm,
the balance have low Sb and Pb contents <1000
ppm (Figure 5.6b). Therefore, this result shows
that Joden 60 was made using recycled glass
while the other four samples were made from
Ten crucibles with vitreous residues attached
from the Mabro site in Maastricht, which is
located very close to Jodenstraat, were also
analysed here. Lead tin yellow II surrounded by
pure lead silica glass has been found in two
crucibles, Mabro 12 and Mabro 14. ‘Mixed alkali’
glass was found attached to crucible Mabro 7,
a high potassium oxide glass containing grey
mainly angular unmelted silica grains was found
attached to crucible Mabro 9 (Figure 5.8),
though both are contaminated with 7.2 and
7.9% aluminium oxide, and high iron and
titanium oxide. In Figure 5.8 the body of the
glass is the homogenous pale grey layer on
the right-hand side.
The glass in crucible 9 contains 13% K2O but
low levels of MgO and P2O5 so it is unlikely to be
evidence for working wood ash glass. Moreover,
it contains 7.54% Al2O3 and 3.49% Na2O.
The presence of angular unmelted SiO2 grains
and high Al2O3 suggests that what remains is the
interaction layer with the crucible fabric.
The mixed alkali glass in crucible 5 contains
levels of MgO and phosphorus pentoxide that
are significantly lower than in typical mixed
Figure 5.6 Plots of Al2O3/SiO2 against TiO2/Al2O3 (a) and Sb against Pb (b) for Jodenstraat glass artefacts.
73
—
a
b
Figure 5.7 The 26 trace element patterns for Joden 44–45 compared to that of relevant natron glass types published
previously (top: 5.7a) and the same for Joden 46 (bottom: 5.7b) (Joden= Jodenstraat, Maastricht).
Figure 5.8 Backscattered SEM image of glassy residue
attached to crucible Mabro 9.
alkali glasses, such as those from Méru in
France.288 It is therefore more likely to be a
contaminated natron glass.
Soda glass with an unusual chemical
composition was found attached to five crucibles:
Mabro 8 and Mabro 11 contain high Al2O3 contents
at 9.87% and 12.48% respectively as well as high
iron and titanium interpreted as contamination
by interaction with the crucible fabric. They may
therefore be contaminated natron
glasses. Mabro 13, Mabro 15 and Mabro 16 have
higher CaO contents at 6.74%, 7.52% and 5.82%,
approaching normal levels for natron glass.
Although Mabro 13 and 15 contain relatively high
Cl levels at 0.52% and 0.38% they contain
normal levels of alumina, iron and titanium. All
three contain elevated K2O levels of up to 2.6%
but these are not paired with elevated magnesia.
The elevated K2O levels may be due to
contamination.
No vitreous phase was found in Mabro 10.
Analysis of the ‘frit-like’ material on the rim of
crucible 11 dating the late 4th-early 5th century
74
—
shows that it is a fuel ash slag with no detected
CaO, high Cl (0.32%), high K2O (4.56%) but
relatively low MgO (1.15%), high Fe2O3 (2.9%)
and high MnO (0.98%). Paynter has shown that
melting a natron glass in wood fired furnace can
lead to fuel ash slags and contaminated glass of
highly variable compositions.289
5.3
289 Paynter 2008.
290 The trace element patterns of HIMT
glass and Egyptian I glass are very
similar, so we are unable to further
categorize the two samples which plot
in the overlapping area of HIMT glass
and Egyptian I glass, Ge 44 and Ge 45,
into a more specific compositional
group using trace elements.
291 Freestone 2015.
Glass samples from Gennep
The 28 vessel glass fragments from Gennep are
the earliest glass studied here: they are tightly
dated to between the late 4th century
and mid-6th century AD. Their low K2O and MgO
contents and elevated MnO, TiO2 and Fe2O3
contents suggest that they are all a type of HIMT
natron glass. The TiO2/Al2O3 and Al2O3/SiO2 ratios
of Gennep glass show that seven of them locate
in the area of the HIMT sensu stricto compositional
group (Ge 44 and Ge 45 fall in the overlapping
area of HIMT sensu stricto and Egyptian I
compositional groups) and 21 of them cluster
closely in the area of the Foy 2 compositional
group (Figure 5.9a with Figure 4.1 as reference).
Pb and Sb concentrations are both <1000 ppm
and are used here as the criterion to provide a
preliminary distinction between ‘pristine’ glass
from recycled glass. According to this criterion,
12 out of the total 28 samples can be regarded as
‘pristine’ glass: six of them belong to the HIMT
sensu stricto group and the other six belong to the
Foy 2 group (Figure 5.9b).
The 26 trace element patterns of the
six ‘pristine’ HIMT sensu stricto glass samples are
essentially identical and their average pattern is
very similar to that of HIMT glass reported
previously thus confirming their HIMT
Figure 5.9 Plots of Al2O3/SiO2 against TiO2/Al2O3 (a) and Sb against Pb (b) for Gennep glass vessels.
compositional type (Figure 5.10a).290 The 26 trace
element pattern of the average composition of
the six ‘pristine’ Foy 2 glass is also identical with
the Foy 2 glass pattern reported previously
(Figure 5.10b).
It has been suggested that the recycling of
Roman tesserae and coloured vessel glass to
supplement the dwindling supply of natron glass
in northwestern Europe started approximately in
the early 8th century, which is signified by the
elevated Sb and Pb contents of glass in this
period.291 However, our Gennep vessel glass
samples are securely dated to late 4th to
mid-6th century AD, and we already see elevated
Sb and Pb contents for the majority of them
(17 out of 29).
5.4
Glass samples from Wijnaldum
Apart from two pieces of evidence for bead
production, one greenish-white glass rod
splinter (WIJ 41) and one turquoise punty glass
(WIJ 42), the rest of the 40 glass samples from
Wijnaldum can be separated into three groups:
highly coloured opaque glass beads, colourless
glass beads (four out of five were metal foil and
the other one was trail-decorated) and glass
vessels.
5.4.1 Highly coloured opaque glass beads
The highly coloured glass beads are in three
basic colours: opaque yellow, opaque white and
opaque red. Wijnaldum beads share a similar
75
—
a
b
Figure 5.10 The average 26 trace element patterns of Gennep ‘pristine’ HIMT sensu stricto glass and ‘pristine’ Foy 2
glass compared to that of relevant natron glass types published previously (top: 5.10a) and the same for Gennep
‘pristine’ Foy 2 (bottom: 5.10b).
date to Maastricht Jodenstraat beads, and the
chemical compositions of glass beads are also
very similar. First of all, the highly coloured glass
beads from Wijnaldum all have elevated PbO
contents: 23–56% in yellow beads, 3–23% in red
beads and 1–5% in white beads. Secondly, they
share the same compositional feature of low Sb
contents as found in Jodenstraat opaque beads:
all highly coloured beads from Wijnaldum have a
Sb content of <1000 ppm. Moreover, the
colouring mechanisms of Wijnaldum beads are
the same as for Jodenstraat beads. The
Wijnaldum yellow beads are coloured by lead tin
yellow II; the Wijnaldum white beads and red
beads are opacified by tin oxide; and the
Wijnaldum red beads are coloured by micron
sized copper particles with iron-rich fayalitic slag
acting as an internal reducing agent (Figure 5.11).
The glass working tray covered with a
contaminated opaque yellow vitreous layer was
analysed previously (see section 2.4.2).
5.4.2 Colourless glass beads
The five colourless glass beads are a special
group (WIJ 35–39), not seen in Maastricht. Unlike
highly coloured opaque glass beads, they do not
have elevated PbO contents. Since their Sb and
Pb contents are all under 1000 ppm, it suggests
that ‘pristine’ rather than recycled glass was
used to make them. Among the five colourless
glass beads, WIJ 35 (a gold foil bead) and WIJ 37
76
—
a
b
c
d
Figure 5.11 The backscattered images of a Wijnaldum
yellow bead (top left: 5.11a), white bead (top right:
5.11b), red bead (middle left: 5.11c), tin opacifiers found
in white beads and red beads (middle right: 5.11d) and
fayalitic slag surrounded by 0 valence copper micron
particles found in red beads (bottom: 5.11e).
e
(a silver foil bead) have completely different
chemical compositions from glass vessels and
other glass beads from Wijnaldum. Their lower
Na2O and CaO contents and higher K2O and MgO
contents compared to natron glass indicate that
they are plant ash glasses, they date to between
775 and 900 AD according to their context dates
and this correlates with the introduction of plant
ash glasses in western Asia by Islamic glassmakers.
The low MgO and K2O contents and
elevated Fe2O3, TiO2 and MnO contents of the
other three colourless glass beads (a gold foil
bead and a silver foil bead dating to 450-550, and
a colourless bead with red streaks dated to 750850) suggest they were made from a type of
Figure 5.12 Plots of Al2O3/SiO2 against TiO2/Al2O3 (a) and Sb against Pb (b) for Wijnaldum glass samples.
77
—
a
b
c
Figure 5.13 The 26 trace element pattern for colourless Wij 36, 38 and 39 compared to that of relevant Foy 2 glass
published previously (top: 5.13a), for Wij 10 and 15 compared to that of relevant HIMT and Egyptian II glasses
published previously (middle: 5.13b) and Wij 42 compared to that of relevant Levantine II glasses published
previously (bottom: 5.13c) (Wij = Wijnaldum).
HIMT natron glass; their Al2O3/SiO2 and TiO2/
Al2O3 ratios show that they belong to the Foy 2
compositional group (Figure 5.12a with Figure 4.1
as reference). The Sb and Pb contents of the
three Foy 2 beads are all very low at <200 ppm,
so this suggests that they were made from
‘pristine’ Foy 2 glass rather than from recycled
glass (Figure 5.12b) and two are early examples
of this kind of glass. The 26 trace element
pattern of the average composition of the three
beads confirms this suggestion (Figure 5.13a).
5.4.3 Vessel glass
The nine vessel glass samples from Wijnaldum
are dated between the mid-5th century and late
9th century. The low MgO and K2O contents of
these samples show that they are all made from
natron glass. Their Al2O3/SiO2 and TiO2/Al2O3
ratios show that WIJ 10 and WIJ 15 plot at the
overlapping zone of HIMT and Egyptian II
compositional groups while the other seven
78
—
samples belong to the Foy 2 compositional
group (Figure 5.12a with Figure 4.1 as reference).
Their Pb and Sb levels are both <1000 ppm and
can be used as a criterion to distinguish pristine
glass from recycled natron glass in a preliminary
way. We can see that WIJ 10 and WIJ 15 can both
be regarded as ‘pristine’ glass, while all Foy 2
glass samples are recycled glass (Figure 5.12b).
The 26 trace element patterns for WIJ 10 and WIJ
15 resolve the inconclusive identification of the
two samples using major chemical compositions
which suggested that they are Egyptian II glass
or HIMT sensu stricto glass. The trace element
patterns of WIJ 10 and WIJ 15 show a close
resemblance to that of Egyptian II glass with
clear compositional differences from that of
HIMT sensu stricto glass (Figure 5.13 middle).
Therefore we can confirm that WIJ 10, a blue
green funnel beaker, and WIJ 15, a dark blue
funnel beaker with an incalmo rim, were made
from ‘pristine’ Egyptian II glass. The dates of WIJ
10 (800–850 AD) and WIJ 15 (770–900 AD) are
also consistent with the suggested dates for
when Egyptian II glass was in circulatation in
the 8th–9th centuries AD, a time when the supply
of pristine Levantine glass was drying up.292
5.4.4 Bead production materials
292 P helps et al. 2016.
The high Na2O content and low K2O and MgO
contents in the greenish-white rod splinter
(WIJ 41) and turquoise punty glass (WIJ 42)
suggest that they are both natron glass.
However, their different Fe2O3 and TiO2 contents
and Al2O3/SiO2, TiO2/Al2O3 ratios suggest that
they do not belong to the same compositional
group. WIJ 42 has quite low Fe2O3 and TiO2
contents of 0.45% and 0.08% respectively, a
feature that differs from the dominating HIMT
natron glass discussed here, and the Al2O3/SiO2
and TiO2/Al2O3 ratios suggest it can be
categorized as having a Levantine II (8th–9th
century) natron glass composition (Figure 5.12a
with Figure 4.1 as reference); the 26 trace
element pattern for WIJ 42 confirms this
identification (Figure 5.13c).
The established date for this compositional
group agrees with the date provided from the
archaeological context (750–800 AD). The high
CuO content in WIJ 42 shows that the turquoise
colour was caused by copper. The greenish-white
glass rod splinter WIJ 41 has elevated TiO2, Fe2O3
and MnO contents like most of the natron glass
studied in this work. The Al2O3/SiO2 and TiO2/
Al2O3 ratios of WIJ 41 plot at the overlapping area
of HIMT sensu stricto and the Egyptian II
compositional groups (Figure 5.12a with
Figure 4.1 as reference). The elevated Sb and Pb
contents of WIJ 41, with 1650 ppm and 4933 ppm
respectively, suggest that it was made from
recycled glass (Figure 5.12b).
5.5
Glass samples from Utrecht
Nine samples from two sites in Utrecht were
analysed. Three glass fragments derive from
Utrecht Oudwijkerdwarsstraat which dates to
the 7th to mid-8th century AD. Six crucible
samples (Utr 31–36) are from Utrecht Domplein
which dates to between the mid-8th to late
9th century AD. The three glass fragments from
Utrecht Oudwijkerdwarsstraat (Utr 77–79) are
soda lime glasses. Utr 77, a fragment with a
yellow tinge and Utr 79, a piece of green debris
from glass working are natron glasses. Utr 78
has a modern composition and therefore will
not be discussed further.
In the plot of Al2O3/SiO2 against TiO2/Al2O3
(Figure 5.14a with Figure 4.1 as reference), Utr 77
plots in the area of the Foy 2 compositional
group, while Utr 79 plots in the area of Roman
glass. Utr 79 contains high CaO and slightly high
Cr2O3. The Sb content of Utr 77 is low at 70 ppm,
but its Pb content is high at 4034 ppm,
suggesting it is a recycled glass. Both Sb and Pb
contents of Utr 79 are low at <1 ppm and 92 ppm
respectively, consistent with the Sb and Pb
contents of non-Sb-decolourized Roman glass
(Figure 5.14 right).
Glass attached to six crucible fragments from
Utrecht Domplein was also investigated. There is
a thin very pale green, appearing colourless,
glass layer attached to crucibles Utr 31 and Utr
32. There is evidence from the chemical
compositions that the glass had interacted with
the body of the crucible: they contain 12.6% and
4.65% Al2O3 respectively. Utr 31 especially has
levels of Fe2O3 (3.17%) and TiO2 (0.64%) very
likely to be the result of these elements
migrating into the glass from the crucible fabric
at high temperatures. Utr 31 was probably
79
—
Figure 5.14 Plots of Al2O3/SiO2 against TiO2/Al2O3 (a) and Sb against Pb (b) for Utr 77 and Utr 79 (Utr = Utrecht).
originally a soda glass; a high level of K2O
(4.73%) (yet only 0.57% MgO) is probably also
due to contamination. A high level of antimony
(0.45% Sb2O3) is difficult to explain but could be
due to glass recycling.
Utr 32 contains 1.66% MgO, 1.9% K2O,
15.54% Na2O and 5.75% CaO, all characteristics
of a plant ash glass. However, a very low level of
P2O5 (0.06%) is unusual for a plant ash glass.
The thick layer of green and red striped
glass in crucible Utr 34 from Utrecht Domplein
also warrants more detailed discussion.
It contains 1.45% MgO and 1.42% K2O, levels that
are probably consistent with a natron glass, as
well as a probable uncontaminated level of Al2O3
(2.82%). However it also contains 0.56% P2O5 as
well as 4.91% PbO, 0.9% SnO2, 0.615% CuO and
1.5% Fe2O3. The copper in the red glass would be
in a reduced form (Cu2O) and also the iron; the
lead, tin and iron were probably introduced as
part of the colouring process along with the
copper. This composition is similar to red
coloured glass from Maastricht Jodenstraat and
Wijnaldum. It also contains 0.728% MnO and
0.12% CoO. The levels of MgO and K2O are below
1.5% and fall within the values for a natron glass,
yet the phosphorus level would be more in line
with a plant ash glass and may indicate a degree
of contamination. Red streaks of decoration are
sometimes found in early medieval vessel glass.
It can be assumed that it was produced
deliberately by mixing in small amounts of red
glass or red glass colorant.
Number 33 is too contaminated to be able
to discern the original chemical composition of
the glass, with an Al2O3 level of 36.89%. The
greenish glasses in crucibles 35 and 36 also
contain a mismatch between the potassium and
magnesium oxide levels as well as elevated
antimony. Number 36 contains 5.4% K2O and
8.21% Na2O so it is tempting to suggest this
might be a mixed alkali glass. However, the
relatively low level of MgO (0.84%) suggests that
the high K is due to contamination of the glass.
Both 35 and 36 contain high levels of Al2O3 at
7.41% and 7.68%, so again this shows that
contamination has occurred. They contain 1.1%
and 0.39% CuO probably originally added as
colourants to a natron glasses.
5.6
Glass samples from Wijk bij
Duurstede (Dorestad)
Apart from one raw glass chip, five tesserae
fragments, one opaque yellow glass rod and two
linen smoother fragments, the other 55 glass
samples from Dorestad are all glass vessel
fragments dated to between the mid-8th and
mid-9th centuries, and do not derive from a
glass-working context.
5.6.1 Vessel glass
Apart from two wood ash glass samples (Dor 103
and Dor 136), a yellow-green palm funnel and an
iridescent yellow-green funnel beaker base the
low K2O and MgO contents and elevated Fe2O3,
MnO and TiO2 contents of the rest of the 52
vessel glass samples from Dorestad suggest they
are all made from natron glass. Their Al2O3/SiO2
and TiO2/Al2O3 ratios show that apart from Dor
122, a pale green funnel beaker, which can be
categorized as Egyptian II glass, the other 51
samples cluster closely in the area the Foy 2
80
—
Figure 5.15 Plots of Al2O3/SiO2 against TiO2/Al2O3 (a) and Sb against Pb (b) for Dorestad vessel glass samples.
compositional group (Figure 5.15a with Figure 4.1
as reference). Apart from Dor 122, all the rest of
the natron glass vessels have either Sb contents
>1000 ppm or Pb contents >1000 ppm and
mostly both (Figure 5.15b). This suggests these
glass vessels were generally made from recycled
glass rather than ‘pristine’ glass.
Dor 122 is the clear exception in this
assemblage. Not only does it belong to a
different glass compositional group (Egyptian II),
its low Sb and Pb contents, 285 ppm and 461
ppm respectively, also suggest it could have
been made from ‘pristine’ glass (Figure 5.15b):
the 26 trace element pattern for Dor 122
confirms that it is Egyptian II glass (Figure 5.16).
The date of Dor 122 (9th century AD) is also
consistent with the suggested date for when
Egyptian II glass was in circulation, in the
8th–9th centuries AD.
The two wood ash vessel glass samples (Dor 103,
a palm funnel and Dor 136, a funnel beaker base)
have quite similar chemical compositions.
293 Wedepohl and Simon 2010
294 Krüger & Wedepohl 2003.
Their CaO contents are very high at over 13%,
their K2O and MgO contents are significantly
higher than that found in natron glass at over
7% and their Na2O contents are low at below
2%. Such compositions are quite typical of
relatively early wood ash glass, even if the K2O
levels are quite low for such glass.293
5.6.2 Other glass
The two dark green linen smoothers (Dor 150,
151) have a very peculiar composition. They have
high lead contents at around 22% and high Al2O3
contents at around 7%. Their high CaO, P2O5,
MgO and K2O contents also suggest they could
have been made from wood ash as a main
ingredient. Over a hundred linen smoothers
were found at an important medieval Viking city,
Hedeby in northern Germany.294 Analytical work
shows that they are generally of two glass
compositions: wood ash glass and wood
Figure 5.16 The 26 trace element pattern for Dor 122 compared to that of relevant Egyptian II natron glasses
published previously (Dor=Dorestad).
81
—
ash-lead glass. Our linen smoother samples are
wood ash-lead glass, similar to those from
Hedeby. Although the presence of the two linen
smoothers could result from Viking trade a clear
compositional match with slag resulting from
the refinement of lead-silver ores at the Carolingian
mine of Melle demonstrates that, along with
many other linen smoothers found in France,
Ireland, Germany, Norway Denmark and Belgium,
the origin of the vitreous slag used to make
them was Melle.295
It has been suggested that the glass
working carried out in the important trading
entrepôt of Dorestad was based on remelting
tesserae, imported raw glass and glass rods
(see Section 3.4). A raw glass chip from the
site has a typical Foy 2 natron glass chemical
composition, the same as nearly all of the vessel
glass from the site. The elevated Sb and Pb
contents of this raw glass chip, 4311 ppm and
2844 ppm respectively show that it was
recycled glass.
The five tesserae samples are in two colours,
turquoise (three samples) and blue (two samples).
From their major and minor chemical
compositions, we can see that they are typical
natron glass chemical compositions with high
Na2O, low K2O and low MgO; the turquoise samples
are coloured by copper and the blue samples are
coloured by cobalt. Three of the five tesserae (Dor
145, 146, 147) have very high Sb contents of over
10000 ppm, a concentration significantly higher
than that found in recycled Foy 2 glass. This
compositional feature confirms that they are
Roman glass with high Sb due to the presence of
calcium antimonate opacifying crystals.
The opaque yellow glass rod (Dor 149) has a
similar chemical composition to opaque yellow
glass from Maastricht, Jodenstraat and Wijnaldum.
It has high Na2O and PbO contents and low K2O
and MgO contents: it is also coloured by lead tin
yellow II. Therefore it would have been made
using the same procedure as the opaque yellow
glass from Jodenstraat and Wijnaldum: natron
glass was used as the base glass, and it was
coloured by lead tin yellow II. The Al2O3/SiO2,
TiO2/Al2O3 ratios and low Sb content of Dor 149
suggest that the base glass used could also have
been ‘pristine’ Foy 2 glass.
5.7
Glass samples from Susteren
The glass samples from Susteren fall into
four categories: trail decorated glass beads
(six samples), window glass (ten samples), glass
attached to crucibles (two samples) and vessel
glass (eleven samples).
5.7.1 Trail decorated glass beads
The trail decorated glass beads (Sust 1–6) do not
occur on the other sites studied here. The bodies
of the trail-decorated glass beads are mainly a
green colour of different shades, and the
decorated glass trails are red, yellow and white.
The matrix glass of all six trail-decorated glass
beads was analysed (Sust 1–6 body); five coloured
glass trails were analysed too (Sust 2–6 trail).
The chemical compositions of the bodies shows
that four of them are made from natron glass
and two (Sust 3 and 4) were made from plant
ash glass. In the Al2O3/SiO2 against TiO2/Al2O3
plot, three natron glass bead bodies (Sust 1, Sust
2 and Sust 6) distribute in the area of the Foy 2
compositional group while the Sust 5 body
composition locates in the area of Egyptian II
glass compositional group (Figure 5.17a with
Figure 4.1 as reference). The elevated copper,
antimony, lead and tin contents of four body
glass samples suggest they were made from
recycled glass (Figure 5.17a).
There are two other glass samples from
Susteren with compositional features like Sust 5:
Sust 14 (window glass) and Sust 22 (glass vessel).
They are recycled glass with high Sb and Pb
contents over 1000 ppm, and they plot together
in the area of the Egyptian II compositional
group in the Al2O3/SiO2 against TiO2/Al2O3 plot.
The chemical compositions of bead Sust 3 and
Sust 4 bodies are clearly different from that of
the other four trail-decorated bead bodies as
they are soda-lime-silica glass with high K2O
and MgO contents and are therefore plant ash
glasses. More details regarding the possible
origins of these glass beads are given in
Section 5.10.7.
As for the coloured glass trails, the red
colour (Sust 2) contains high levels of CuO
(1.74%) and Fe2O3 (4.59%) in a natron glass
295 Gratuze et al. 2003; Pactat et al. 2017.
82
—
matrix. This suggests that the colour could be a
result of 0 valence copper micron sized metallic
particles and the iron inclusions, perhaps
introduced in a slag, would have acted as an
internal reducing agent.296 The opaque yellow
colour of Sust 3 trail decorated glass bead has a
similar chemical composition to the yellow glass
beads from Maastricht Jodenstraat and Wijnaldum:
it is coloured with lead tin yellow II.297
The colouring mechanism of opaque white Sust
4 and Sust 6 is different from any other opaque
white glass beads studied in this work as they
are not coloured by tin-based opacifiers but by
calcium antimonate crystals. This suggests the
opaque white glass trails are recycled Roman
opaque white tesserae. Opaque yellow Sust 5 is
also different from other opaque yellow glass
studied here: it is coloured by lead antimonite
rather than lead tin yellow II. This also suggests
that the yellow glass used for the trail may have
been recycled Roman tesserae.298
5.7.2 Window glass
296
297
298
299
Peake & Freestone 2012.
Henderson 2023.
Henderson 2023.
Henderson 2023.
As for the window glass samples, apart from
Sust 11, which is a wood ash glass, the remaining
nine window glass samples are all natron glass.
In the Al2O3/SiO2 against TiO2/Al2O3 plot of these
natron window glass samples (Figure 5.17a with
Figure 4.1 as reference), Sust 16, which contains a
high Al2O3 content at 3.6%, does not plot in the
area of any recognized natron glass compositional
group, and Sust 14 plots in the area of the
Egyptian II glass compositional group, while the
remaining seven samples all distribute in the
area of the Foy 2 compositional group. All natron
window glass samples contain elevated Sb and
Pb contents (Figure 5.17b), suggesting they are
recycled rather than pristine glass. We also
noticed that Sust 7 and Sust 8 contain extra high
Sb2O3 contents (1.58% and 1.70% respectively),
which are much higher than that found in common
recycled Foy 2 glass. The wood ash glass Sust 11
has very high P2O5 (2.5%), MgO (5.6%), CaO
(11.0%) and K2O (14.2%) levels, which is a typical
of northern European wood ash glass dating to
the late 8th century at the earliest.
5.7.3 Glass attached to crucibles
The two glass samples attached to crucibles
(Sust 17, Sust 18) are natron glass but their
composition is different from common natron
glass. Sust 17 contains high Sb2O3 (1.5%) and K2O
(5.2%) contents, but less than 1 wt% MgO.
This peculiar composition may suggest the
addition of recycled Roman tesserae and
contamination from furnace wood ash. On the
other hand, Sust 18 contains high Al2O3 (7.74%)
and K2O (3.3%), which suggests that the glass
has been contaminated by interaction with the
crucible fabric.299
5.7.4 Vessel glass
Eleven vessel glass samples from Susteren have
been analysed. Their chemical compositions
show that apart from two samples, Sust 19 (the
pale green tubular base of a funnel beaker) and
Sust 28 (a pale green funnel beaker fragment
decorated with a green and white reticella rod),
the others are natron glass. The Al2O3/SiO2 and
TiO2/Al2O3 ratios of the nine natron glass samples
indicate that Sust 26 (a nearly colourless funnel
Figure 5.17 Plots of Al2O3/SiO2 against TiO2/Al2O3 (a) and Sb against Pb (b) for Susteren natron glass samples.
83
—
Figure 5.18 The 26 trace element pattern of Sust 26 compared to that of the relevant HIMT natron glass type.
beaker with a dark blue incalmo rim) belongs to
the HIMT sensu stricto compositional group and
Sust 22 (a blue-green funnel base) plots in the
area of the Egyptian II compositional group,
while the other seven samples are of a Foy 2
composition (Figure 5.17a with Figure 4.1 as
reference). The Sb and Pb concentrations show
that Sust 26 was made from ‘pristine’ glass while
the other eight samples were made from
recycled glass (Figure 5.17 right). The 26 trace
element pattern of Sust 26 matches with that of
pristine HIMT sensu stricto reported previously
(Figure 5.18). Therefore this confirms that Sust 26
was made from ‘pristine’ HIMT sensu stricto glass.
Sust 19 is a highly weathered funnel base which
contains Na2O and K2O at 8.87% and 8.5%
respectively. It is therefore a mixed alkali glass.
Sust 28 is a plant ash glass sample containing
0.3% P2O5, 4.2% MgO, 2.5% K2O and 7.0% CaO.
The trace element composition of Sust 28 has
shed light on its provenance; this will be
addressed along with other plant ash samples
identified in this work in the discussion below.
5.8
Glass samples from Deventer
Forty one samples from Deventer-Stadhuiskwartier were analysed. Apart from two
monochrome glass beads (Dev 2, 8), the rest of
the samples are glass vessels, window glass and
raw glass fragments. Four compositional types
of glass were identified. The largest number are
wood ash glass (19 samples), thirteen samples
are natron glass, four samples are mixed alkali
glass and one sample is plant ash glass.
Four samples were too weathered to be worthy
of analysis (DEV 16, 17, 23, 34). Deventer samples
are elaborated according to their compositional
types in the following.
5.8.1 Wood ash glass
The 19 wood ash glass samples are all vessel and
window glass. The sample numbers are: DEV
1,6,7,8, 14,15, 21, 24, 26, 27, 28, 29, 30a, 32, 33, 35,
36, 38 and 39. The main fluxing agent in wood
ash glass is K2O, with CaO in variable amounts.
These wood ash glasses generally have high
MgO contents over 3.5%, high P2O5 contents
over 2.0% and low Na2O contents of below
3.0%. Apart from sample Dev 24, which has a
CaO/K2O ratio of 5.8 which is the only example
of a high lime (23.6% CaO), low alkali (2.9%
Na2O, 4.1% K2O) glass (HLLA), the rest of the
19 samples have CaO/K2O ratios from 0.7 to 2.0
with an average at 1.1. Deventer 24 is a pale
blue-green window fragment: its chemical
composition is much more typical of medieval
and post-mediveal glass300 so a later production
date for it can not be ruled out (however see
Section 6.5). The chemical compositions of early
medieval wood ash glass are not especially well
defined and as can be seen from previously
published results they are quite variable.301
Analytical research on later (13th century) wood
ash glasses provide some broad compositional
trends but the origins in northern France,
‘Rhenish’ (based on borders of c. 1300 AD which
includes the Low Countries) and central Europe
300 Van Wersch et al. 2018.
301 Wedepohl & Simon 2010.
84
—
Figure 5.19 The average 26 trace element patterns of wood ash glass samples from Deventer compared to that of
relevant natron glass studies published previously.
do not provide especially clear distinctions and
in many cases the ‘types’ overlap.302 Moreover a
comparison with earlier Deventer wood ash
glasses offers a somewhat confused picture and
such a comparison is, in any case, inappropriate
given the difference in dates associated with the
different socio-economic contexts of
production. The 26 trace element pattern of the
average composition of these wood ash glass is
plotted in Figure 5.19 along with that of natron
glasses of selected compositional groups (Foy 2
and Egyptian II) reported previously. We can see
that the trace element pattern of wood ash glass
is quite different from that of natron glass,
especially the clearly elevated Rb, Cs and Ba
contents.
5.8.2 Natron glass
302 Adlington et al. 2019.
Among the thirteen natron glass samples (DEV 2,
3, 9, 10, 11, 12, 13, 19, 20, 25, 31, 40, 41), three
samples, surprisingly for such a late site, have
features suggesting that they are of a Roman
provenance: Dev 9 (an amber coloured fragment),
Dev 13 (a funnel beaker with a very high Mn
content suggesting Mn decolourisation) and Dev
20 (a typical Roman vessel shape, Isings 96a). All
date to between 900 and 950 AD. At the same
time, the three samples also distribute in the
Roman glass area in the Al2O3/SiO2 against TiO2/
Al2O3 plot and have low Sb and Pb concentrations
<1000 ppm (Figure 5.20). Therefore they can
definitely be identified as glass produced in the
Roman tradition.
The Al2O3/SiO2 and TiO2/Al2O3 ratios of the
other ten natron glass samples show that Dev 11
(the rim of a possible funnel beaker), Dev 31
(possible window glass) and Dev 40 (a fragment
of dark and light green layered glass) all plot in
the area of Egyptian II glass; they date to 900–
925, 890–925 and 950–1050 AD respectively
according to the find contexts. The rest belong
to the Foy 2 compositional group (Figure 5.20a
with Figure 4.1 as reference) and date to
between 850 and 1050. The Sb and Pb
concentrations of the ten samples suggest that
Figure 5.20 Plots of Al2O3/SiO2 against TiO2/Al2O3 (a) and Sb against Pb (b) for Deventer natron glass samples.
85
—
Figure 5.21 The 26 trace element patterns for Deventer 11, 31 and 40 compared to that for relevant Egyptian II natron
glass published previously.
the Egyptian II subgroup (Dev 11, Dev 31 and Dev
40) are all ‘pristine’ glass while only one Foy 2
monochrome blue glass bead (Dev 2) is a
‘pristine’ glass (Figure 5.20b) dating to 850–900
AD. The 26 trace element pattern of Dev 2
confirms that it was made from ‘pristine’ Foy 2
glass, while the patterns for Dev 11, Dev 31 and
Dev 40 confirm that they were made from
‘pristine’ Egyptian II glass (Figure 5.21). The rest
of the Foy 2 glasses are recycled.
Two mixed alkali glasses have elevated PbO
contents of between 2.4 and 2.8% lead oxide.
This could be interpreted as an addition of lead
glass, introduced during the recycling process
and which should not be regarded as a surprise
if a ‘potluck glass working’ strategy was used
during glass recycling.
5.8.3 Mixed alkali glass
The one plant ash glass identified from the
Deventer assemblage (Dev 37) is a green possible
beaker fragment with a little weathering, from a
context dating to between 950 and 1050. It has a
typical plant ash chemical composition. It is a
soda-lime-silica glass with high MgO and K2O
contents at 1.8% and 4.77% respectively.
Discussion of its possible origin is given in
Section 5.10.7.
In contexts dating to 850-900 AD, six out
of eight samples are wood ash or mixed alkali
glasses, with the remaining two being recycled
Foy 2. In contexts dating to between 900 and
950 AD there are ten wood ash glasses, three
Roman (natron) glasses, two recycled Foy 2
glasses and two Egyptian II glasses. Two glasses
dating to between 950 and 1000 are wood ash;
between 950 and 1050 four glasses are of the
wood ash type, one is plant ash, one recycled
Foy 2 and one Egyptian II. Given that there is no
evidence for a wood ash glass industry in the
Netherlands possible ways that wood ash glass
was imported, would have been as part of the
Viking trade network or from the various sites
further north.
The four mixed alkali glass samples from
Deventer (DEV 4, 5, 18, 22) are an interesting
group, which are not found amongst glasses
reported elsewhere in this study and are a
reflection of their production date. The samples
tested derive from archaeological contexts dated
to between 850 and 950 AD. Their high MgO
contents (>1.8%) suggest the potassium content
was introduced in the form of plant ash/wood
ash rather than potassium-containing minerals.
Their Na2O/K2O values are variable, between 0.5
and 1.7. Glass with similar features has also been
found in 8th–10th century sites in Germany, France
and the Netherlands (see Section 2.4.1). It has
been suggested that this type of mixed alkali
glass was produced in order to extend stocks of
soda glass,303 and the procedure can be
interpreted either as the mixing together of
wood ash and natron glasses or, less likely, by
the addition of an increasing amount of wood
ash in a mixture of glass cullet.304
5.8.4 Plant ash glass
303 Krueger & Wedephol 2003.
304 Pactat et al. 2017.
86
—
5.9
Nd-Sr isotope analysis
Twenty samples selected from Maastricht
(Jodenstraat), Wijnaldum and Dorestad were
analysed for their strontium and neodymium
isotopic compositions, and eighteen valid results
were obtained. In this study 143Nd/144Nd ratios
are represented in parts per 104 deviation from
the present-day value of a model evolution of
Nd isotopes in a chondritic Earth (Chondritic
Uniform Reservoir, CHUR)305 according to the
following equation:
ε Nd=
(143Nd⁄144Nd ) Sample
(143Nd⁄144Nd ) CHUR
-1 ×104, with 143Nd/144NdCHUR =0.512638
The strontium and neodymium isotopic
compositions of our 18 samples and their
chemical compositional groups are listed in
Table 5.3 and plotted in Figure 5.22. The two
wood ash glass samples, Dor 136 and Dor 150,
both lead-rich linen smoothers, have very
different Nd–Sr isotopic signatures from the
other samples. Their Nd isotopic signatures
reflect the Nd isotopic signatures of the sands
used for making them, and their Sr isotopic
signatures reflect the bio-available Sr isotopic
signatures in the calcium-rich raw materials
used to make them. The Sr from wood ash glass
would have mostly been introduced in the wood
ash used as the flux. The Nd–Sr isotopic
signatures of Dor 136 and Dor 150 have a similar
range to that of 14th–15th century forest glass
produced in Staffordshire, England.306 However
there are no other Nd–Sr isotopic datasets of
wood ash glasses by geographic region currently
available so we cannot suggest provenances
based on Nd–Sr results for two wood ash glass
samples at this stage.
The only plant ash sample, WIJ 37, from the
assemblage clusters together with natron glass
samples in figure 5.22 but it has a low 87Sr/86Sr
signature. Comparing the Nd–Sr isotopic
signatures of WIJ 37, a silver foil colourless bead,
with that of other available plant ash glass data
(Figure 5.23), we found that it overlaps with 3rd–7th
century Sasanian glass found in Veh Ardašīr, an old
Sasanian administrative centre 40 km to the
southeast of modern Baghdad.307 Although the
3rd–7th century Sasanian glass samples from Veh
Table 5.3 The Nd–Sr isotopic compositions and compositional groups of samples
analysed (WIJ=Wijnaldum; DO=Dorestad).
Sample number
305 Depaolo & Wasserburg 1976.
306 Meek, Henderson & Evans 2012.
307 Ganio et al. 2013.
Compositional group
143Nd/144Nd
87Sr/86Sr
εNd
Maastricht-Jodenstraat 47
‘pristine’ Foy 2
0.512
-5.6
0.709
Maastricht-Jodenstraat 49
‘pristine’ Foy 2
0.512
-7.6
0.709
Maastricht-Jodenstraat 58
‘pristine’ Foy 2
0.512
-5.4
0.709
Maastricht-Jodenstraat 63
‘pristine’ Foy 2
0.512
-5.4
0.709
Maastricht-Jodenstraat 68
Roman glass
0.512
-5.2
0.709
Maastricht-Jodenstraat 73
‘pristine’ Foy 2
0.512
-6
0.709
Maastricht-Jodenstraat 76
‘pristine’ Foy 2
0.512
-5.6
0.709
WIJ 13
recycled Foy 2
0.512
-5
0.709
WIJ 27
not identified
0.512
-6.8
0.709
WIJ 36
‘pristine’ Foy 2
0.512
-5.4
0.709
WIJ 37
plant ash glass
0.512
-6.4
0.709
DOR 111
recycled Foy 2
0.512
-6
0.709
DOR 113
recycled Foy 2
0.512
-5.6
0.709
DOR 115
recycled Foy 2
0.512
-6
0.709
DOR 128
recycled Foy 2
0.512
-5
0.709
DOR 136
wood ash glass
0.512
-9.9
0.715
DOR 147
Roman tesserae
0.512
-6.2
0.709
DOR 150
wood ash glass
0.512
-10.7
0.711
87
—
Figure 5.22 A Nd–Sr isotopic plot of the glasses analysed; WIJ 37, Joden 49, DOR 136 and DOR 150 are labelled
separately (WIJ= Wijnaldum; DOR = Dorestad; Joden = Jodenstraat, Maastricht).
Figure 5.23 A Nd–Sr isotopic plot of WIJ 37 compared with that of plant ash glasses from the Middle East published
previously (WIJ = Wijnaldum). Data source: 3rd–7th century Sasanian glass Nd–Sr isotopic data is from Ganio et al.
(2013), Tyre raw furnace glass from Degryse et al. (2010), al-Raqqa 9th century and 11th–12th century vessel glass data
from Henderson, Evans & Barkoudah (2009), the rest from Henderson, Ma & Evans (2020).
Figure 5.24 A Nd–Sr isotopic plot of the natron glass samples analysed. Joden 49 is labelled separately (Joden =
Jodenstraat, Maastricht). The ellipses are drawn according to Degryse & Schneider (2008).
Ardašīr are too early for the late 9th century date of
WIJ 37, this observation still sheds some light on
the possible geographical origin of this plant ash
glass sample. This is addressed in detail along with
other plant ash samples identified in this study in
Section 5.10.7.
We have suggested that Joden 68 was a Roman
ribbed vessel fragment: this is confirmed as a
Roman natron glass from its major and minor
chemical elemental composition; Dor 147 is a
88
—
Figure 5.25 The 26 trace element patterns for Joden 68 and Dor 147 compared to that for Levantine I natron glass
published previously (Joden= Jodenstraat, Maastricht; DOR = Dorestad).
Roman glass tessera. We did not try to make
further provenance identification based on their
trace element data since the geographical
production centre of this Roman glass type is
not yet clear. Joden 68 has a 87Sr/86Sr of 0.7090
and εNd of -5.03, and Dor 147 has a 87Sr/86Sr of
0.70892 and εNd of -6.20, typical Nd–Sr
signatures of glass produced in Levant.308
Following this lead, we compared the trace
element patterns of Joden 68 and Dor 147 with
that of early medieval Levantine glass, and we
found they both have similar trace element
patterns to Levantine I glass (Figure 5.25).
This observation indicates that Joden 68 and
Dor 147 were made with a similar sand source as
that used for making Levantine I glass.
Therefore we suggest that Joden 68 and Dor 147
are Roman glasses produced in the Levant.
308 Degryse & Schneider 2008.
309 Brems et al. 2018.
The Nd–Sr isotopic compositions of the thirteen
natron glass samples mostly distribute in the
typical range for natron glass Nd–Sr signatures,
εNd at between -5 and -7, and 87Sr/86Sr at
between 0.7085 and 0.7093,309 except for one
outlier: Joden 49, a drop of green glass with soil
contamination (Figure 5.24). The εNd of Joden
49 is -7.3, lower than the normal range of natron
glass -7 ≤ εNd ≤ -3, and this may be related to
contamination. Twelve out of the thirteen natron
samples have been identified as belonging to the
Foy 2 compositional group: seven ‘pristine’ Foy 2
glass and five recycled Foy 2 glass (Table 5.3),
and the 87Sr/86Sr ratios of the twelve samples
stretch in a rather wide range between 0.7085
and 0.7093. Six out of the seven ‘pristine’ Foy 2
glasses (except for the outlier Joden 49) actually
cluster closely at 87Sr/86Sr of 0.7086, a typical
value for glass produced in Egypt, the commonly
suggested origin of Foy 2 glass. Five of these
samples are the working debris from bead
production in Maastricht (a glass strand, a drop
and three rods) and a gold foil bead from
Wijnaldum.
The 87Sr/86Sr values for five recycled Foy 2
glass samples, a pale green funnel beaker from
Wijnaldum as well as two funnel beakers, one
palm funnel and a gold leaf decorated funnel
from Dorestad stretch between 0.7088 and
0.7093. This is higher than the typical Egyptian
range. These recycled ‘Foy 2’ glasses are all
dated to the late 8th to mid-9th century.
We suspect the Sr isotopic signatures of recycled
Foy 2 glass samples would have changed from
the ‘pristine’ Foy 2 values during the recycling
process, which would have involved mixing
different types of glass. The mixing of glass
types with very high 87Sr/86Sr values (such as
wood ash glass), into a mainly natron glass
recycling batch, would have caused the elevation
of 87Sr/86Sr values in the recycled natron glass
compared to ‘pristine’ natron glass. Evidence of
recycling/mixing wood ash glass along with
natron glass has also been noted in the trace
element patterns of recycled Foy 2 glass dated to
the late 8th century and later. This is addressed in
Section 5.10.4.
89
—
5.10 Discussion
bead-making artefacts found at the Maastricht
Jodenstraat site were made from the same glass
as the base glass used for making the opaque
glass beads: it is all ‘pristine’ Foy 2 glass.
5.10.1 The base glass used for bead making
at Jodenstraat
It has been suggested that a two-step procedure
was used for the manufacture of highly coloured
opaque glass beads in early medieval
northwestern Europe: first, naturally coloured or
colourless base glass was coloured and made
into rods or strands. Then coloured glass rods or
strands were softened and formed into beads.310
Apart from the highly coloured opaque glass
objects, the site of Jodenstraat in Maastricht also
yielded translucent naturally coloured and cobalt
blue bead-making waste, which is ideal to test
this two-step production proposition and
understand what kind of glass was used as the
base glass.
In this work we found three types of
evidence that prove that the naturally
coloured translucent glass-working debris is of
the same compositional type as the base glass
used to make opaque glass and thus that the
two-step production proposition stands. Firstly,
the Al2O3/SiO2 and TiO2/Al2O3 ratios of the
coloured opaque glass-working remains and
translucent glass samples cluster together and
they can all be categorized as being of the Foy 2
type. From the Al2O3/SiO2 versus TiO2/Al2O3 plot
we can see that, on average, highly coloured
glasses have a slightly higher Al2O3/SiO2 ratio
than naturally coloured glass debris: the cluster
of highly coloured glasses plot slightly to the
right of the cluster of naturally coloured glassworking debris (Figure 5.1). This can be attributed
to the higher Al2O3/SiO2 ratios of the lead tin
yellow colourant and tin opacifiers that were added
to the highly coloured glass matrix (Table 5.1).
Secondly, the Sb contents of the naturally
coloured glass-working debris and highly
coloured glass samples are all very low,
suggesting they were both made from ‘pristine’
glass. Finally, the 26 trace element patterns of
the average composition of the naturally
coloured bead-making artefacts and the average
composition of the highly coloured glass are
identical and the same as that of ‘pristine’ Foy 2
glass previously published. Therefore we can
suggest strongly that the naturally coloured
5.10.2 The use of crucibles in on-site lead
tin yellow colourant production in
early medieval northwestern Europe
Early medieval lead tin yellow residues attached
to crucibles from northwest European sites have
been studied before.311 The shapes of the crucibles
containing lead tin yellow residues are mostly
shallow and with a wide opening that resembles
a tray. It has been suggested this type of crucible
was not used for metallurgical processes but
was specifically used for making the lead tin
colourant for yellow coloured beads,312 which
were very popular in early medieval Europe. We
agree with the suggestion that the crucibles with
yellow residues were used for making the yellow
colourant for bead making, since the main phase
in the yellow residue is lead tin yellow II
(PbSn(Si)O3), the common colouring agent of
yellow beads at the time. However, from the
study of the rich material remains related to
bead making found at Jodenstraat Maastricht,
we have managed to provide some new insights
into how lead tin yellow was made in these tray
shaped crucibles.
It has been suggested that this type of
crucible may have been used for calcining lead
and tin, a chemical reaction which would result
in lead tin yellow I, and that lead tin yellow II
identified in the crucible may have been formed
by the reaction between the siliceous crucible
body and lead tin yellow I.313 Heck and colleagues
also suggested that the colourant produced in
the crucible was to be mixed, in a ratio of one to
one by volume, with natron glass to make the
yellow glass for beads.314 Peake and Freestone315
studied yellow beads from Tarbat Ness and
Eriswell and reviewed the results of Henderson
and Ivens316 and Heck and colleagues317 for the
yellow residues attached to crucibles, and
concluded that the reaction which occurred in
the crucible, resulting in the lead tin yellow II
colourant, involved silica as a raw material.
The thorough replication experiments to
produce lead tin yellow colourant by Rooksby318
and Matin and colleagues319 demonstrated that the
310 Sablerolles, Henderson & Dijkman 1997.
311 Henderson & Ivens 1992; Heck, Rehren
312
313
314
315
316
317
318
319
& Hoffmann 2003; Peake & Freestone
2014.
Heck, Rehren & Hoffmann 2003.
Heck, Rehren & Hoffmann 2003.
Heck, Rehren & Hoffmann 2003.
Peake & Freestone 2014.
Henderson & Ivens 1992.
Heck, Rehren & Hoffmann 2003.
Rooksby 1964.
Matin, Tite & Watson 2018.
90
—
a
b
c
d
Figure 5.26 Backscattered SEM image (top left: 5.26a) for Joden 19 showing the unmelted silica grains. Elemental lead
maps for Joden 19 lead (Pb) (top right: 5.26b), silica (Si) (bottom left: 5.26c) and tin (Sn) (bottom right: 5.26d) (Joden
= Jodenstraat, Maastricht).
yellow colourant for glass- and glaze-making
would have been produced in a two-step
procedure, as recorded in ancient Islamic
literature. Firstly, a lead tin calx, which contains
lead tin yellow I (PbSn2O4) as the main component,
is produced by calcining lead and tin together in a
stoichiometric ratio of Pb:Sn over 3.5. Then lead tin
calx is mixed with silica and the mixture is heated
to over 800°C for the lead tin yellow colourant
(lead tin yellow II) used in glass- and glaze-making
to be produced. During the second step, variable
amounts of SiO2 substitute for SnO2 in PbSn2O4
(lead tin yellow I) which causes a crystalline
conversion to PbSn(Si)O3 (lead tin yellow II).
No lead tin yellow I phase has been
identified in the yellow tin residues in this study,
and the chemical composition of lead tin yellow
II attached to crucibles is in good agreement
with that found in the yellow glass and yellow
beads, especially with respect to the Sn to Si
ratio, which could be variable depending on how
much Sn was replaced by Si during the reaction
involving lead tin yellow I and silica. Therefore
we suggest that the lead tin yellow colourant
production procedure used by early medieval
northwestern European bead makers would
have been very similar to that recorded in later
Islamic literature. This was a two-step procedure:
the tray shape crucibles with lead tin yellow
residues attached would have been used during
the second step when heating silica with lead tin
yellow I for making lead tin yellow II occurred.
91
—
Silica grains are observed in one of our samples
(Joden 19) in support of this suggestion
(Figure 5.26), also observed in the Early Christian
Irish evidence from Dunmisk.320
5.10.3 The separate production of a tinbased white opacifier at Maastricht,
Jodenstraat
Apart from crucibles containing lead tin yellow
residues, two crucibles containing tin white
residues (Joden 23 and Joden 30) were found at
Jodenstraat. Our study shows that the main
phase of the white residues is SnO2 (50–70 wt%)
in the presence of variable amounts of PbO and
SiO2 (see Section 5.2.3). Because the Sn:Pb ratio
in the tin white residues is very different from
that of lead tin yellow residues and very similar
to the tin-based opacifiers found in the white,
red and greenish-blue glass and beads, we can
strongly suggest that the tin-based opacifiers
were intentionally produced as a separate
material on site at Jodenstraat and that they
were probably used directly in the production
of the white, red and greenish-blue glass for
bead making.
Highly coloured glass opacified by tin-based
opacifiers re-emerged around the same time
when lead tin yellow glass became mainstream
in northwest Europe in c. 3rd century AD.321
Although early medieval northwestern European
lead tin yellow colourant production has been
studied on a few occasions, the two crucibles
containing tin-based opacifiers used to make
white glass are the first reported evidence
confirmed by scientific analysis that tin opacified
white glass was produced separately in early
medieval northwestern Europe.
In a previous study of bead-making
materials from Jodenstraat322 it was suggested
that the white, red and greenish-blue glass
rods found at the site for producing beads of
the same colours may have been imported
from other sources since no crucibles containing
glassy residues of these three colours were
found at the site. Two extra pieces of evidence
identified here suggest that the white, red and
greenish-blue glass rods for producing beads of
the same colours were also produced on site,
rather than being imported from elsewhere.
Firstly the base glass of all four different
coloured glasses and beads as well as the
colourless glass from working the glass were
made from a very similar ‘pristine’ Foy 2 glass
(see Section 5.10.1), and secondly the opacifiers
used in the white, red and greenish-blue glass
were probably also produced on site.
The contrasting PbO contents (Figure 5.3)
between yellow glass and the white, red and
greenish-blue glasses is probably also related to
the use of lead tin yellow colourant in the yellow
glass and tin-based opacifiers in white, red and
greenish-blue glass. When the lead tin yellow
colourant was produced, a pure lead silica glass
formed around lead tin yellow II in the lead tin
yellow colourant as we can see in Figure 5.5.
When in the third stage the colourant is added
to the melted natron base glass, the pure lead
silica glass would have melted and mixed into it
and the lead content of the resulting glass would
have increased greatly as a result.
On the other hand, the production of the tinbased white opacifier would have involved far
less lead; no pure lead silica glass formed around
the tin opacifier.323 So when the tin opacifier was
mixed into melted natron base glass, lower levels
of lead would have been introduced. This is a
reasonable explanation for the source of the lead
content in highly coloured opaque glasses and
beads, and why there is a big difference in the PbO
contents between yellow glass and glass of the
other three colours (Figure 5.3). Moreover, the
methods used to make the lead tin yellow
colourant and tin oxide opacifier were different.
As we can see in the SEM backscattered
images of the four highly coloured opaque
glasses (Figure 5.4), the bright phase in yellow
glass (lead tin yellow II crystals) consists of very
small particles of a similar size and they are
distributed homogeneously in the glass matrix,
while the tin oxide crystals in white, red and
green glass have heterogeneous crystal sizes and
a lot of them are in rather big aggregates. This
observation suggests that the lead tin yellow
colourant (lead tin yellow II) may have been
mixed with the base glass while they were both
in melted or semi-melted state, and that is why
the lead tin yellow II consists of small particles
that are distributed homogeneously in the glass
matrix. On the other hand, the white tin oxide
opacifier may have been added in a powdered
state after it was produced on site in a large
volume.
320
321
322
323
Henderson & Ivens 1992.
Matin 2019.
Sablerolles, Henderson & Dijkman 1997.
Matin 2019.
92
—
5.10.4 The other chemical characteristics
of the glass and its archaeological
implications
324
325
326
327
328
329
330
Foy et al. 2003.
Freestone et al. 2018.
Foy et al. 2003.
Foy et al. 2003.
Ares et al. 2019.
Bertini, Henderson & Chenery 2020.
Freestone 2015.
Glass fragments from non-industrial contexts
account for the majority of the samples studied.
They include glass vessels, window glass and
some raw glass fragments. Altogether 160 of
such glass samples have been chemically
analysed – 28 from Gennep, five from Jodenstraat
Maastricht, nine from Wijnaldum, three from
Utrecht, 56 from Dorestad, 21 from Susteren and
38 from Deventer. The majority (95) of these
samples have been identified as recycled Foy 2
glass, corresponding to the Foy 2.2 group of Foy
et al.324 Nine HIMT sensu stricto glasses have been
identified from the collection, and seven of them
are from the site with the earliest date: Gennep
(late 4th to mid-6th century). Eight ‘pristine’ Foy 2
glasses have been identified: they are from
Gennep (six samples) and Maastricht Jodenstraat
(two samples). They date to between the late 4th
and early 7th centuries.
Six ‘pristine’ Egyptian II glasses have been
identified: two from Wijnaldum, one from
Dorestad and three from Deventer. Their dates
provided by the archaeological contexts in which
they were found are as follows: three date to
8th–9th centuries AD, two date to the early
10th century and one dates to between 950 and
1050 AD. They are consistent with or slightly
later than the suggested date when Egyptian II
glass was widely circulated, in the 8th and
9th centuries AD. Twenty seven wood ash/mixed
alkali glass samples have been identified.
They are from three sites: Dorestad (two samples),
Susteren (two samples) and Deventer (24 samples)
and their context dates are consistent with the
suggested date when wood ash glass started to
be made and was circulating in Europe from the
late 8th century onwards.
Five Roman glasses have been identified:
they are from Jodenstraat (Joden 68), Utrecht
(Utr 79) and Deventer (Dev 9, 13, 20). Four plant
ash samples have been identified: Sust 3, 4 and
28, and Dev 37. The remaining eight samples are
five compositional outliers (Utr 78, Dor 95,
Sust 14, 16, 22) and three highly weathered
samples (Dev 16, 17, 23). We can see that the
identified compositional types of these
mainly vessel and window glass fragments are
consistent with our current understanding of the
dates when the different types of glass emerged,
peaked and disappeared.
It has been concluded that HIMT glass in
the widest sense (including HIMT sensu stricto,
Foy 2 etc.) was in use from the middle of the
fourth century until the seventh century.325
However, in this study we have found that glass
made from recycled glass with Foy 2 glass
compositional features (referred to as recycled
Foy 2 glass in this study, corresponding to
Foy 2.2 group)326 was still the dominant group
between early 8th century and late 9th century
(in Dorestad and Wijnaldum vessel glasses).
It remained in circulation till possibly as late as
the first half of the 10th century: four out of
38 glasses analysed from Deventer dated to
850–950 AD have been identified as recycled
Foy 2 glass- with a single example, perhaps
redeposited, in a context dating to 950-1000.
This means that recycled glass with Foy 2
compositional features was still in use centuries
after the ‘pristine’ Foy 2 glass supply ended in
northwestern Europe. The chemical composition
of recycled Foy 2 glass samples, have some
differences from that of ‘pristine’ Foy 2 glass, one
of which is that they nearly all contain elevated
Sb and Pb contents. This recycled version of Foy 2
glass (especially for Foy 2.1 subgroup) is labelled
as Foy 2.2 subgroup in the work of Foy and
colleagues,327 and is known from a very limited
number of assemblages in France, Italy and Spain
that are typically dated to the end of the seventh
and the eighth centuries AD.328
The 7th–11th century glass assemblage from
Comacchio, northern Italy, is compositionally
very similar to the samples studied here,
as recycled glass with Foy 2 compositional
features account for the majority of the glasses:
the authors of that paper labelled such recycled
glass ‘intermediate’.329 This observation is in line
with the suggestion that around the early
8th century, Roman tesserae became a ‘new’
glass source in northwestern Europe; it was
recycled along with other glass, supplementing
the dwindling natron glass supply.330 From the
26 trace element patterns of the average
compositions of recycled Foy 2 glasses from
Gennep, Wijnaldum, Dorestad, Susteren and
Deventer we can see that they all retain the basic
signature of ‘pristine’ Foy 2 glass (Figure 5.27).
However, we have also noticed that the
26 trace element patterns of recycled Foy 2
93
—
a
b
c
Figure 5.27 The average 26 trace element patterns of recycled Foy 2 glasses from Gennep (5.27a), Wijnaldum (5.27b),
Wijk bij Duurstede (Dorestad) (5.27c).
94
—
d
e
f
Figure 5.27 (continued) The average 26 trace element patterns of recycled Foy 2 glasses from Susteren (5.27d),
Deventer (5.27e) and ‘intermediate’ glass from Bertini, Henderson & Chenery (2020) (5.27f ).
95
—
glasses from sites of the Carolingian period
(Dorestad, Susteren) and slightly later at
Deventer clearly have elevated Cs and Rb
contents. Recycled Foy 2 glass from Gennep
and Wijnaldum, which mostly predate the
8th century, do not show these features so
strongly, nor do the recycled ‘intermediate’
group studied by Bertini and colleagues.331
This could suggest that glass of a new
compositional type started to be involved in
glass recycling as a minor component in the
Netherlands from around the start of the
Carolingian period.
We suspect that this ‘new’ glass could have
been wood ash glass since elevated Ba, Rb and
Cs contents are the clear discriminating
characteristics between wood ash and natron
glass as defined by 26 trace element patterns
with higher concentrations in wood ash glasses
(Figure 5.19). This suggestion is also supported
by the elevated Sr isotopic compositions of
recycled Foy 2 glass samples dated to late 8th to
mid-9th century from Wijnaldum and Dorestad,
and in line with the understanding that wood
ash glass started to be manufactured in Europe
from the late 8th century in northwestern Europe.
A further compositional characteristic of both
Merovingian glass from Gennep and Carolingian
glass from Dorestad is that potassium is
correlated with both Rb and Li. An intriguing
characteristic of these correlations (not shown
here) is that Rb levels in the Gennep ‘pristine’
Foy 2 glass are mainly below 12ppm whereas the
concentrations in Dorestad recycled Foy 2
glasses are mainly between 12ppm and 20ppm
with a small number containing levels upto
28ppm. This is another clear marker of the
increased degree of recycling in the later
(Carolingian) glasses and the same thing is true
for Li concentrations. Some Dorestad glasses
contain between 23 and 40ppm Li. These same
correlations found in 7th century and later Foy 2.1
glasses have been attributed to evidence of sitespecific contamination from e.g. muscovite in
the crucibles used for working the glass at Tolmo
de Minateda, Spain. 332 Our results suggest that
such contaminants including very similar
concentrations of Rb and Li as in the Tolmo glasses
may not be characteristic of local production in our
case. A more likely interpretation is that the
increase in concentrations of potassium,
rubidium, lithium and cesium in Carolingian
glasses compared to earlier glasses may be
attributable to muscovites associated with
sands used to make some of the glasses that
were mixed as part of the Carolingian recycling
processes involving Foy 2 glass.
5.10.5 A comparison of 7th–11th century
vessel glass from Comacchio with
early medieval Dutch glass and the
suggested supply of raw glass in the
two areas
We have mentioned that the recycled Foy 2 glass
in this study bears similar compositional features
to the ‘intermediate glass’ identified by Bertini
and colleagues333; they are both recycled glasses
and their major and minor chemical compositions
are very similar to that of ‘pristine’ Foy 2 glass.
Just as the recycled Foy 2 glass was the dominating
compositional group of vessel glasses from early
medieval Dutch sites, the recycled intermediate
glass was also the dominating compositional
group for glass of 7th–11th century dates found at
Comacchio, northern Italy (53 out of 77). However,
in the two studies only a very limited number of
‘pristine’ Foy 2 glasses have been identified. This
phenomenon suggests that ‘pristine’ Foy 2 glass
would have been the main raw glass imported to
the two areas for glass working prior to or in the
early part of the period. However, after this
supply waned, local glass working in the two
areas had to rely more on recycling contemporary
cullet and old Roman tesserae.
The chemical compositions of glass from
Comacchio and from the early medieval
Netherlands also reflect some differences in the
supplies of raw glass. First of all, Levantine glass
was one of the important sources of raw glass in
Comacchio, possibly after the ‘pristine’ Foy 2
glass supply waned in the area, but not in the
Netherlands. Altogether 17 pristine Levantine
glass samples were identified from the total of
the 77 items analysed from Comacchio.334 This is
a proportion that is much higher than the
number (four) of ‘pristine’ Foy 2 samples found
at the site. Besides, the chemical compositional
features of mixing Foy 2 glass with Levantine
glass were also noted in the intermediate glass
group, which suggests that apart from Foy 2
glass, Levantine glass was another source of
glass contributing to the recycling process.
331
332
333
334
Bertini, Henderson & Chenery 2020.
Schibille et al 2022.
Bertini, Henderson & Chenery 2020.
Bertini, Henderson & Chenery 2020.
96
—
335
336
337
338
339
340
341
342
Schibille et al. 2022.
Pactat 2021, Fig 7.
Cagno et al. 2012.
Genga et al. 2008.
Casellato et al. 2003.
Pactat 2021.
Phelps et al. 2106.
Krueger & Wedepohl 2003.
In contrast, apart from one punty glass
fragment from Wijnaldum (WIJ 42), no pristine
early medieval Levantine glass was found
amongst the samples analysed here, and no
mixing with Levantine glass can be suggested
from the chemical compositions of the
predominantly recycled Foy 2 glass found here.
Analysis of glass from another contemporary
site in the southern Mediterranean, Tolmo de
Minateda (Spain), also revealed the presence of
a higher proportion of Levantine I glass (33 out
of 253 samples).335 Therefore, we can be sure that
pristine Levantine natron glass did not arrive in
the Netherlands in the relatively large quantities
that reached northern Italy and Spain. Secondly,
no wood ash glass was identified amongst the
7th–11th century Comacchio glasses analysed.
By the 9th–11th centuries, wood ash glass
accounts for more than half of the samples
tested from Deventer with the proportion of
wood ash glasses found in contexts dating to
after 900 increasing somewhat, perhaps as a
result of the Viking trade network. Wood ash
glass was almost entirely dominant in the
manufacture of a range of glass vessels by the
10th century in France with minimal recycled
natron glass.336 Therefore this shows that wood
ash glass found in northern, western and central
Europe may not have been available in Comacchio
and other sites in northern Italy337 during the
period, just as Levantine glass may only have
had very limited availability in the early medieval
Netherlands. No wood ash glass was found at
the 9th to 13th century glass from the site of
Siponto in southern Italy338 and only plant ash
glass was being manufactured by the 13th14th century at the northern Italian site of
Germagnana.339
Moreover, six pieces of Egyptian II glass
have been identified amongst Wijnaldum,
Dorestad and Deventer glasses. The three
Deventer examples all contain lower levels of Li,
R, Cs and Ba than detected in contemporary
recycled Foy 2 glasses discussed above reflecting
their ‘pristine’ nature. Although Egyptian II do
not account for a high proportion of the total
glasses analysed, they are the only ‘pristine’
natron glass found from the 8th century and later
with only an ‘anecdotal’ occurrence in France.340
This could mean that amongst the limited
amount of ‘pristine’ natron glass that arrived in
the Netherlands during this period, Egyptian II
glass was quite important. Moreover, no Egyptian
II glass was identified amongst 7th–11th century
glass vessels from Comacchio even though a
higher proportion of Egyptian II glass was found
in the Levant after the 9th century.341. This
therefore reflects an important contrast in the
availability of raw glass in northern Italy and the
Netherlands and it also partly reflects a collapse
in Mediterranean trade.
5.10.6 Wood ash glass and mixed alkali glass
Wood ash glass and mixed alkali glass are
treated as one group here because they were
both produced using wood ash as one important
raw material. All wood ash glass and mixed alkali
glass identified came from Carolingian sites: two
wood ash glasses from Dorestad, one wood ash
glass and one mixed alkali glass from Susteren,
and twenty four wood ash and mixed alkali
glasses from Deventer. Their dates are all later
than the end of the 8th century, which is in line
with the date that wood ash glass technology
emerged in northwestern Europe.342 We have
also noted above that wood ash glass may
already have been involved in glass recycling in
the Netherlands as a minor component during
the Carolingian period as suggested by the
clearly elevated Rb and Cs contents of the
recycled glass from Dorestad, Susteren and
Deventer (see Section 5.10.4).
However, the proportions of wood ash in all
the glasses studied from the three sites are quite
different. In Dorestad and Susteren, wood ash
glass and mixed alkali glass only account for a
very small fraction of early medieval glass, while
in Deventer wood ash glass and mixed alkali
glass make up two thirds of the total vessel glass
from the site. This difference can largely be
attributed to the dates of glass samples from the
three sites. The glass from Dorestad and
Susteren dates to between the 7th and 9th centuries,
while the glass from Deventer dates to between
the mid-9th and mid-11th centuries. This suggests
that the supply of wood ash and mixed alkali
raw glass was quite limited in the Netherlands
during the late 8th century and early 9th century
and that they are likely to have formed a significant
part of the raw glass supply from the mid9th century onwards.
97
—
Figure 5.28 Plots of Cr/La against 1000Zr/Ti ratios (top) and Li/K against Cs/K ratios (bottom) of plant ash glasses
from Deventer and Susteren (labelled) and Wijnaldum compared to data published previously. Data source:
Henderson et al. 2016.
5.10.7 Plant ash glass
Altogether only six plant ash glasses were
identified from all the glass analysed here.
They are four glass beads, WIJ 35 and WIJ 37
(a gold foil and a silver foil bead), Sust 3 and Sust
4 (both trail decorated conical beads) and two
glass vessels, Sust 28, a pale green funnel beaker
and Dev 37, a possible beaker fragment. It is
difficult to give clear provenance identification
for plant ash glass by using major and minor
chemical compositions alone. However, by plotting
certain key trace element values343 (Cr/La against
1000Zr/Ti and Li/K against Cs/K) found in plant
ash glass samples along with a large dataset of
plant ash glasses of different origins from the
Middle East,344 we can provide some clues about
the possible origins of our samples (Figure 5.28).
From the two plots using key trace element
ratios we can see that three samples (WIJ 35,
WIJ 37 and Sust 28) cluster together in both
plots. Sust 3 clusters together with WIJ 35, WIJ 37
and Sust 28 in the Cr/La versus 1000Zr/Ti plot
and with Sust 4 in the Li/K versus Cs/K plot. Dev
37 clearly has a rather different trace element
composition from the remaining five samples.
In terms of the origins of the plant ash samples,
we can see WIJ 35, WIJ 37 and Sust 28 basically
cluster closely with samples from sites located in
the eastern zone of Western Asia, in Iraq and
Iran: Nishapur, Ctesiphon and Samarra. The Nd–
Sr isotopic ratio of WIJ 37 also clusters with
Sasanian glass samples from Veh Ardašīr, 40 km
to the southeast of modern Baghdad, Iraq (see
Section 5.9) providing a geological provenance
so this agrees with the trace element results.
Therefore we can suggest that WIJ 35,
WIJ 37 and Sust 28 originated from the eastern
zone of Western Asia. Dev 37 clusters closely
with samples from Damascus in both plots and
343 The approach was taken by Henderson
et al. 2016.
344 Henderson et al. 2016.
98
—
therefore derives from the Levantine region.
Moreover, it can be suggested that Sust 3 was
made in the eastern zone and Sust 4 possibly
northern Syria, though this is not entirely clear.
5.11
Summary
The analytical results for the early medieval glass
samples studied have provided new insights,
especially into two aspects of the glass used and
manufactured in the Netherlands, namely the
highly coloured opaque monochrome glass beads
produced in 6th-7th century AD Netherlands and
changes in the glass supply in the period between
the late 4th century and mid-11th century AD.
The analysis of the materials used to make
coloured beads from the Jodenstraat site in
Maastricht has confirmed the previously
suggested two-step manufacture mode for
coloured bead making (see Section 5.10.1)
and on-site lead-tin yellow colorant production
at the site (see Section 5.10.2). Our new
understanding from analysing these materials is
that ‘pristine’ Foy 2 glass was used as the base
glass for making beads at Jodenstraat
(see Section 5.10.1) and that the tin opacifiers,
to make both yellow and white glass, used in
making coloured beads were also produced
on-site (see Section 5.10.3).
The majority of glass analysed was imported,
and our aim in analysing it was mainly to identify
its chemical compositional group. We have
managed to group the glass according to its
compositional features for the majority of the
glass analysed and we have also identified shifts
in glass supply in the Early Medieval Netherlands
according to changes in glass compositional
groups over time (see Section 5.10.4 and
Section 5.10.5). The political and economic
factors probably responsible for the shifts in
glass supply to the Early Medieval Netherlands
are addressed in Chapter 6, along with evidence
from other historical studies.
99
—
6 Synthesis and conclusions
6.1
The Early Merovingian period
(450-550 AD)
McCormick345 considers that the study of textiles,
relics and coins, but also glass and ceramics, are
a good way to disentangle the complexities of
the early medieval economy. Our scientific
results for glass help to unravel some aspects of
the early medieval economy in the Netherlands
and build on the established picture for the
production and supply of early European glass.
The early Merovingian period dating to between
450 and 550 AD saw important economic, social
and cultural developments under the Pippinids
in the Meuse valley. The glass in use was
essentially a continuation of the Roman tradition
but with a reduction in the range of production
techniques for vessels. It mostly involved
translucent blue-green nearly colourless hues
for vessels and there was a continued import of
glass beads from the Mediterranean. There is
some evidence that glass beads were made in
Cologne. Bowls, cups, bottles and cone beakers
sometimes with trail decoration, including the
famous Kempston type,346 date to this period.
There is evidence for a workshop at Huy in
the Meuse valley where glass was worked but no
evidence for the types of objects/vessels made
there has been found.347 There is also evidence
for Rhenish production348 with possible evidence
in Cologne349 and concentrations of early Frankish
vessels around Mayen (Eifel) might suggest a
production centre in the area.350 Furthermore a
possible production site for 5th century Helle
bowls has been found in western Germany at
Asperden351, quite close to Gennep, where some
were discovered.352 It is possible that glass vessels
were made in the proto-urban centre of Maastricht
at this time.353 So the glass vessels found at
Maastricht and certainly those found at Gennep
studied here would have been imported from
one or more of these places.
The glass that we have analysed dating to
this phase mainly derived from Gennep,
with two vessels from Wijnaldum. By determining
major, minor and trace levels of element oxides
in glass vessels from Gennep dating to between
the late 4th and mid 6th centuries we have found
the use of (pristine) HIMT sensu stricto, Foy 2, and
quite early examples of recycled Foy 2. HIMT glass
was probably made in Egypt354 from the mid
4th to the 5th centuries and Foy 2 also probably
had an Egyptian origin from the second half of
the 5th and the 6th centuries. Therefore, raw
furnace glass would have been imported to
centres on the Rhine and Meuse to be remelted
and worked/blown into vessels. Glasses would
also have been recycled on these secondary
production sites though no crucibles in which
evidence for such glass mixing have been found.
There are some interesting relationships
between compositional type, colour and vessel
type for Gennep vessel glasses. Four cone beakers
are made from pristine HIMT, three of which are
decorated with spiral coils below the rims.
Seven other cones were made from recycled
Foy 2 glass. Whereas the Foy 2 glass is very pale
green ‘colourless’ or pale green, the HIMT glass
cones are olive-green, yellow-green and amberbrown due especially to higher levels of iron and
manganese. Pristine Foy 2 was used to make
four bowls with vertical loops below an outfolded
rim, similar to the decoration on Kempston
cones; pristine Foy 2 was not used to make
cone beakers in our data set. It appears therefore,
that a specific supply of pristine Foy 2 glass was
used to make the bowls we have analysed from
Gennep; single examples were made from pristine
HIMT and recycled Foy 2 glass. Of the five bowls
from Gennep decorated with an opaque white
feather pattern that we analysed, four are of
recycled Foy 2 glass and the fifth made from
pristine Foy 2. Both cone beakers and bowls
were made with nearly colourless (very pale
green) recycled Foy 2 glass, including the two
cone beakers from Wijnaldum dating to this
phase, one a Kempston cone.
Therefore, the majority of cone beakers at
this site/time were made from pristine pale
green Foy 2 glass or more deeply coloured
yellow-green and brown pristine HIMT glass. In
contrast the majority of ‘colourless’ bowls
decorated with feather patterns and some
‘colourless’ cone beakers were made from
recycled Foy 2 glass. Perhaps unsurprisingly this
suggests that colour was an important
consideration when it came to pale green or
nearly colourless drinking vessels made from
Foy 2 glass, where the colour of the liquid could
be observed depending on whether beer or wine
was being consumed. Nevertheless, the use of
strongly coloured olive-green, yellow-green and
amber-brown Egyptian HIMT glass to make cone
beakers suggests something else. It is possible
345
346
347
348
349
350
351
352
McCormick 2001, 281.
Evison 1972.
Van Wersch 2013.
Koch 1987.
Dodt, Kronz & Simon 2021.
Sablerolles 1993.
Brüggler 1994.
Sablerolles 1992, Sablerolles 1993,
Henderson 2000, 68-70.
353 Van Lith & Sablerolles 1995.
354 Arles et al. 2019.
100
—
355 Biblioteca Apostolica Vaticana, Rome,
356
357
358
359
360
361
362
363
364
Ms. I. Cod. Reg. Lat. 438, 25r (https://
digi.vatlib.it/view/MSS_Reg.
lat.438/0053). Bibliothèque nationale
de France, Paris, Département des
Manuscrits, Lat. 8085 fo 61v. https://
portail.biblissima.fr/ark:/43093/
mdata69093fcda745577d3ff9d21597dc3fc4c51ca346. Special
Collections University Library Leiden,
Codex Burmanni Q 3, folio 120v.
(https://disc.leidenuniv.nl/view/
item/1935754?solr_nav%5Bid%5D=22d3ef11949c6ef5312f&solr_
nav%5Bpage%5D=0&solr_nav%5Boffset%5D=0#page/138/mode/1up).
See for instance Arbman 1937, 41-44,
fig. 5a-b; Gaut 2011, 194, 255-265.
Aunay, et al.2020, 298 (Paris, BnF. Ms.
Lat. 1, f327v).
Later 2010; Van Winkelhof 2021.
Van Winkelhoff 2021.
Henderson, Sode & Sablerolles 2019.
Pion 2014.
Boschetti, Gratuze & Schibille 2020.
The Rural Riches project funded by the
European Research Council (ERC) of the
European Union, and directed by
Professor Frans Theuws is investigating
the sources and distribution of such
beads.
Matin 2019.
that glassblowers used whatever came to hand
but we have nevertheless found evidence for a
relationship between vessel type and colour so
probably some vessels were made in batches
using raw glass of the same colour.
Unstable Merovingian vessels such as bell
beakers, palm cups and deep palm cups as well
as Carolingian funnel-shaped and pointed
conical beakers are usually interpreted as
drinking vessels which had to be emptied before
placing them upside down. Indeed, illuminations
in several Carolingian manuscripts show the use
of 9th century glass cones as wine glasses.355 In
Viking period graves and settlements the
association between glass funnels and cones
and fine ceramic tablewares such as Tating jugs
and Badorf pitchers also indicates a drinking
function.356 However, an illustration in the First
Bible of Charles the Bald, also known as the
Vivian Bible, after Count Vivian of Tours who
commissioned the bible in 845, shows the use
of individually suspended glass cones as lamps.357
It appears that these Carolingian glass vessels
were multi-functional. The same is probably true
for unstable Merovingian glasses with rounded
bases. Experiments with bell beakers, palm cups
and deep palm cups have led researchers to
believe that all these vessel types could have
been used as lamps.358 Van Winkelhoff found
that mould-blown and optic blown vertically
ribbed glass vessels were especially effective in
that they emitted bright and clear “sunlike”
patterns and that small mould-blown palm cups
work especially well as lamps when hung at a
low level above the ground.359 She concluded
that such usage is especially relevant in the
context of a grave lamp, or votive lamp, hung as
a visual reminder of the deceased at the grave.
Although raw pristine HIMT glass imported
from Egypt would undoubtedly have passed
through intermediaries before being blown into
vessels, perhaps its exotic origin was still known
and was socially and/or ritually significant
depending on whether it was used in a domestic
context or in a burial. Six beads and a greenishwhite glass breaking splinter from Wijnaldum
date to this phase. Two beads are colourless
with gold and silver-foil respectively.
They appear to have been made using pristine
Foy 2 glass, adding to the evidence for a much
higher proportion of pristine glass in circulation
in the early Merovingian period than in the
Carolingian period. Out of the six beads a single
opaque yellow one is opacified with lead
stannate so dates to before the evidence of
the production of such glass in Maastricht;
the breaking splinter is opacified with calcium
antimonate in the Roman tradition found in
tesserae.360
6.2
The Middle Merovingian period
(550– 650 AD)
In our second period dating to between 550 and
650 AD the production of glass vessels became
somewhat diminished, with a reduction of
vessel types. For some reason beads were no
longer imported from the Mediterranean and
this seems to have led to the birth of Merovingian
bead production,361 which saw massive numbers
of glass beads being manufactured with very
similar designs. These were distributed across
Europe between the Anglo-Saxon realms, the
Merovingian territories as well as the Frankish
kingdom of Italy362 as late as the end of the
7th century. So, much imported glass would
have been used to make beads.363
In the Netherlands, glass beads, especially
opaque yellow ones, were being manufactured
in Maastricht, with especially good evidence for
their production dating to between the late
6th and early 7th century at Jodenstraat, Maastricht,
located well within the Merovingian empire.
Comprehesive evidence for bead making has
also been found at the central places of Rijnsburg
Abdijterrein on the Rhine delta and WijnaldumTjitsma, a northern terp site. Both of these
two sites are located well outside the border
of the Merovingian empire so there was clearly
a demand for fashionable beads outside the
empire; a local ruler may well have invited a
bead maker to the sites. The bead making
evidence at all three sites dates to the last quarter
of the 6th and first half of the 7th centuries AD.
We have found evidence for the manufacture
of lead tin yellow (lead tin yellow II) for colouring
glass in a number of crucibles at Jodenstraat,
Maastricht. The occurrence of silica crystals
associated with this yellow pigment, as well as
pure lead-silica glass in eight crucibles, is clear
evidence for the second step in the production
of lead tin yellow II when it was added to the
lead tin calx.364 The first step in the production of
lead tin yellow II – the production of lead tin calx
101
—
– would have happened elsewhere (perhaps
even in Maastricht), but not in the crucibles
we have examined. The same evidence for
the manufacture of lead tin yellow glass in a tray,
dating to between 575 and 625, was found at
Wijnaldum.
The yellow pigment was made in what
was originally 6th–7th century Merovingian
wheel-thrown ovoid-shaped domestic
coarseware storage pots which would have
had a constricted opening and an everted rim.
The top half of the pot was taken off for the
production of the yellow pigment.
This production process for opaque yellow
glass is described in later Islamic literature,
the earliest being Abu’l Qasim Kashani dating
to 1301.365 No lead-silica glass is in use for the
manufacture of vessels or, by itself, for the
manufacture of beads at the time.
We have found direct scientific evidence
linking the crucible yellow glass to the yellow
beads and yellow rods at Maastricht so they
are clearly part of the production process there.
The yellow beads and rods are united
compositionally because a pristine Foy 2 base
glass was used to make them. Translucent glass
vessel fragments and wasters from Maastricht
Jodenstraat were also made with pristine Foy 2
natron glass. Had the Roman vessel glass
fragments found at Jodenstraat been used as
a base glass instead this would have been
detected, but this is not the case. Although
similar evidence for the production of opaque
yellow lead tin colourant has been found
elsewhere,366 including at Wijnaldum (discussed
here), the Maastricht evidence constitutes the
best evidence for its production in terms of its
scale and for its use for making beads in northwestern Europe. Quite why there was this
demand for yellow glass beads is an intriguing
question. No evidence for the production of
opaque yellow glass in crucibles has been found
in Carolingian contexts in the Netherlands.
Further new evidence for the production of
opaque glass is for the manufacture of a tin
white opacifier, found in two crucibles from
Maastricht, Jodenstraat. This is the first evidence
for this from northwestern Europe. We have also
demonstrated that it was used to make the
white, red and greenish glass beads at Jodenstraat,
including their characteristic microstructures and
the use of the same base natron glass, Foy 2.
No crucibles in which the white pigment was
mixed with base glass have been found.
Whole pots, for which there is production
evidence from Maastricht itself, were used
for working translucent glass from both the
Jodenstraat and Mabro sites in Maastricht,
and also in Utrecht.
Detailed investigation of the ‘frit-like’
material observed on the rim of a crucible dating
to the late 4th-early 5th century from the Mabro
site in which translucent greenish glass was
reheated, which would be tentative evidence
for primary glass production367, has instead been
shown to have a variable composition and to
be fuel-ash slag.
Seventeen of the beads from Wijnaldum
date to this phase. Like the Maastricht Jodenstraat
beads they were made from highly coloured
opaque yellow, white and red glass. Their leadrich chemical compositions, the colourants and
the opacifiers used, as well as the evidence for the
use of a pristine base glass, are all very similar to
Maastricht beads. Two later (8th–9th century) gold
and silver-foil decorated colourless plant-ash
glass beads from Wijnaldum are discussed below.
Bead makers were evidently located in
proto-urban or urban centres in the Meuse
valley. This is in contrast to the situation further
north where the relatively small scale of production
suggests that bead makers travelled to centres
like Wijnaldum, Rijnsburg and perhaps
Valkenburg-De Woerd. Evidence for the
manufacture of very popular Merovingian bead
types with crossed swag decoration have been
found at Rijnsburg, one of the types being found
as far south as Italy. Callmer368 has suggested
that glass bead production was regarded as
having a magical aspect at the time. If magic
was considered as important, perhaps this is
one reason why beads were no longer imported.
6.3
The Late Merovingian period
(650 – 750 AD)
By 650–750 AD fewer glass beads were made;
there was a restricted range of rather poorly
made vessel forms, such as palm cups and deep
palm cups usually made with poor quality bubbly
glass full of inclusions, reflecting the high level of
recycling. A further reduction in imported beads
was a catalyst for more beads to be manufactured,
especially in Scandinavia. The artefacts that we
365 Allan 1973; Matin 2019.
366 Henderson & Ivens 1992; Peake &
Freestone 2014.
367 Sablerolles, Henderson & Dijkman 1997.
368 Callmer 2003.
102
—
have studied from this period are four opaque
(yellow, red, orange and white) glass beads from
Wijnaldum though these beads were probably
made earlier than their context dates.
6.4
369
370
371
372
373
Lassaunière et al. 2016.
Van Versche et al. 2015.
Pactat et al. 2017.
Pactat et al. 2017.
Foy et al. 2003.
The Carolingian period
(750 – c. 850 AD)
The Carolingian dynasty (750–887 AD) saw a
renaissance in vessel production, especially with
the use of new decorative techniques particularly
in northern France, but also with clear evidence
for the manufacture of pale green beakers from
the Rhenish area with an expansion in the scale
of production. It is also possible that highly
coloured vessels were made in monasteries in
the Netherlands. The vessels produced in France
included globular jars and reticella decorated
beakers: there is evidence for the production of
lead-tin yellow pigment and its use in reticella
rods to decorate glass vessels from the early
8th century site of Hamage, northern France.369
This must have created different markets for
both simple and more highly decorated vessels.
Although the glass-working evidence for the
Carolingian period is scantier, a contrast with the
Merovingian period is that it occurs on a wider
range of site types, including the emporium of
Dorestad and the ecclesiastical centres of
Susteren-Salvatorplein and Utrecht-Domplein.
Wearing glass beads became less fashionable in
the Frankish heartlands though they were still
worn in the northern periphery of the empire.
From the end of the 8th century onwards Islamic
glass beads were imported via Viking trade
networks and occur in settlements along the
Rhine and in cemeteries north of the Rhine.
The availability of Islamic glass beads may well
have impacted on the manufacture of Frankish
glass beads.
By this time the evidence for glass recycling
had increased, with a much higher proportion
of weak HIMT/recycled Foy 2 glass in circulation,
with few examples of HIMT or other pristine
glass types. Our analyses of fifty-five samples
of palm cups, palm funnels (including a gold-foil
palm funnel), bell beakers, funnel beakers and
a bowl from Dorestad, as well as a Kempston
cone, a bowl and four funnel beakers from
Wijnaldum, show that, with a few exceptions,
those who made these vessels relied on a supply
of recycled (Foy 2) natron glass. The same is true
for the Carolingian vessel glass from Susteren.
Therefore we have assembled very strong
evidence for recycled Foy 2 glass being the
dominant glass type in the 8th and 9th century
Netherlands, with pristine glass, especially Foy 2,
having almost gone out of use.
The few exceptions are the use of Egyptian
II glass used to make a possible pale green
funnel beaker from Dorestad, and the latest
dated funnel beakers from Wijnaldum (770–900):
one is a dark blue incalmo rim, the other a bluegreen colour. A single pale green funnel from
Susteren was made with pristine HIMT. The only
pristine Levantine (II) glass found in this study is
a turquoise punty dating to 750–800 from
Wijnaldum. So even when recycling was such a
dominant practice at this time, a small amount
of pristine glass was in circulation and would
have been imported in raw form and made into
funnel beakers at production centres.
This was a period of technological transition
when the first European wood-ash glass was
being manufactured and used to make vessels
such as those from Baume-les-Messieurs,370
and the possible production of mixed-alkali
glass by extending natron glass with wood-ash
(glass) at Méru, France.371 Reflecting this period
of technological transition, two wood-ash-lead
linen smoothers (partly made using slag from
silver smelting)372 and two yellow-green woodash palm funnels derive from Dorestad, and a
single piece of wood-ash blue-green window
glass from Susteren. A single mixed-alkali pale
green funnel was also found at Susteren. Raw
wood-ash and mixed-alkali glass used to make
the vessels would have been made more locally.
By using trace element analysis we have
demonstrated that there was an increase in the
levels of, for example, potassium and phosphorous
oxides over time. Even though these are initially
at low levels in the 8th–9th centuries, we have
suggested that this indicates that a small
proportion of wood-ash was being mixed into
the (recycled) Foy 2 glass (referred to by Foy et al.
as Foy 2.2)373 from the late 8th century and into
the Carolingian empire. This is supported by the
occurrence of elevated trace levels of cesium,
barium, rubidium and strontium in Foy 2 glass
after the 8th century, all characteristics of woodash glass.
We have also demonstrated that this is the
case with the first neodymium and strontium
103
—
isotope analyses of recycled Foy 2 glass and
observed that a probable explanation for an
increase in the strontium isotope ratio (when
compared to pristine Foy 2 glass) can be attributed
to the mixing in of a small proportion of woodash glass. These compositional and isotopic
results reflect the emergence of the first
European-made (wood-ash) glass technology.
The main evidence for early wood-ash glass
production zones, based on the occurrence of
dated glass objects, is in France and Germany.
Previous research has pointed to potential
evidence for the addition of wood-ash glass to
natron glass in Anglo-Saxon vessels from Jarrow,374
and Ares et al.375 have noted that a ‘plant ash’
component must explain the elevated levels of
potassium, magnesium and phosphorus oxides
above those found in natron glasses.
Nevertheless, elevated levels of these elements
could also be introduced into glasses with the
addition of a small proportion of fuel-ash slag,
also ultimately with a wood-ash component,
as seen in the analysis of material attached to a
crucible rim from Maastricht Mabro (crucible 9).
It is however more than a coincidence that
increasing levels of potassium and phosphorus,
and especially elevated concentrations of
cesium, rubidium, barium and strontium are
correlated with a time when wood-ash glasses
were being introduced in Europe, so this is a far
better explanation.
Window glass with a full wood-ash
composition has been found at the monastery of
Baume-les-Messieurs, Jura in France376 dating to
the late 8th century. It is possible that the presence
of small proportions of wood-ash (glass) in
recycled Dutch early medieval natron glass resulted
partly from wood-ash glass production further
south in Belgium and France – and we suggest
that a source to the east in Germany as mentioned
in Chapter 2.4.1 is more likely. There is no early
medieval wood-ash glass making evidence from
the Netherlands.
We have compared our results with those
from Comacchio in northern Italy377 and observed
some intriguing differences. In Comacchio a far
higher proportion of pristine Levantine glasses
was identified (17/77). This compares with a single
example amongst our data: a punty glass from
Wijnaldum. The mixture of Levantine and Foy 2
glass from Comacchio led to an ‘intermediate’
group being recognized. No examples of woodash glass were found at Comacchio, nor evidence
for its mixture with natron glass. We are therefore
the first to define these contrasting production
spheres using the characteristics of recycled
glasses: a southern European one exemplified
by Comacchio glass and Spannish glass from
Tolmo de Minateda378, with far greater evidence
for the use of Levantine glass, and a northwestern
European Dutch one with a reliance on an
admixture of wood-ash glass and no apparent
evidence for mixing with Levantine glass.
Our approach could be used to define recycling
traditions, reflecting trade links, in other parts
of Europe.
The manufacture of the first European
(wood-ash) glass would have partly been driven
by the demand for windows in monasteries and
churches: by the 10th century it is therefore no
coincidence that wood-ash glass was being
worked at the ecclesiastical site of La Milesse,
Sarthe in France. In our study nine out of the
eleven window glasses analysed from the monastic
site of Susteren are of a natron composition –
seven are recycled Foy 2, and the single example
of Egyptian II noted above. This is further
evidence for very different glass supply in the
two areas.
From the early 9th century another type of
glass was made, from ashed halophytic plants
and sand in Islamic cosmopolitan centres in
western Asia. Although the Sassanids had made
plant-ash glass earlier there is no current
evidence that it was used to make glass objects
in western Europe. Trace element analysis has
shown that the manufacture of plant-ash glass
by the Muslims formed part of a decentralized
production system379 and that it is possible to
link quite securely the provenance of Islamic
plant-ash glasses to production centres or
zones. A gold-foil bead (dated 775–850 AD)
and a silver-foil bead (dated 875–900 AD) from
Wijnaldum are plant-ash glasses imported from
western Asia; the bodies of two rather unique
trail-decorated conical beads and a blue-green
funnel from Susteren are also made of plant-ash
glass.
The Susteren beads and beaker would have
been made in the west using raw plant-ash glass
imported from western Asia. A probable funnel
beaker fragment of a west Asian plant ash
composition from Deventer was found in a
context dated to between 950 and 1050 AD.
Trace element analysis shows that the metal foil
beads from Wijnaldum were probably made in
374
375
376
377
378
379
Freestone & Hughes 2006.
Ares et al. 2019.
Van Wersch et al. 2015.
Bertini, Henderson & Chenery 2020.
Schibille et al. 2022.
Henderson et al. 2016.
104
—
380
381
382
383
384
385
386
387
388
389
Kronz et al. 2016.
Thedéen 2009.
Philippsen et al. 2021.
Langbroek 2021b.
Henderson & Holand 1992.
Gaut & Henderson 2011.
Pactat 2021.
Schibille et al. 2020.
Neri, Gratuze & Schibille 2018.
Pers. Comm Professor Kronz.
the eastern zone of Western Asia, in Iraq/Iran.
The glass used to make the two Susteren beads
probably derived from Iran/Iraq, and possibly
northern Syria, respectively. One of the Susteren
beads (Sust 4) is decorated with opaque white
glass produced in the Roman tradition (calcium
antimonate crystals) used almost universally in
Roman glass tesserae. Therefore the bead
combines an intriguing combination of western
European and west Asian traditions. The raw
plant-ash used to make the Deventer vessel
(fragment) derived from the Levant. The import
of raw plant-ash glass probably formed part of
the Viking trade network via centres like Hedeby.
Though only plant ash glass beads have been
reported so far, the occurrence of mixed plant
ash and lead glass at Hedeby provides indirect
evidence for the import of raw plant ash glass.380
The occurrence of other Islamic artefacts
made from plant-ash glass is further evidence of
such a trade network. Examples are the import
of large numbers of early Islamic glass beads to
Scandinavia from the late 8th century and later
in Viking-age contexts such as in Gotland burials,381
from precision-dated excavations at Ribe,
Denmark382 and millefiori decorated glass beads
from Dorestad.383 No examples were found at
Borg in Norway384 but five plant ash glass beads
were found at Kaupang.385 Early Islamic glass
beads are found along and to the north of the
Dutch Rhine, along the German Wadden sea and
on the Baltic coast. However minimal numbers
have been found along the German Rhine or in
Belgium and France. An exception is the occurrence
of 9th century small glass bottles made from
plant-ash glass imported from Islamic lands,386
which were probably used for the import of
specific western Asian liquids. The relative rarity
of Islamic plant ash glass in France at this time
may be because the demand was lower, due to
the greater availability of wood-ash and mixedalkali glass. However, a better explanation is that
France and Belgium did not form part of the
Viking trade network that existed to the north.
Further south, in Umayyad Spain, plant-ash
glass was imported from western Asia in the
8th and 9th centuries; 387 Islamic glass beads
dominated amongst those found at Illyricum,
Albania.388
6.5
The late phase, including the
Ottonian period (c. 850 – c. 1000 AD)
The last phase (850–1000) was a time when very
few glass beads were produced in the Netherlands
apart from evidence of small-scale production at
Dorestad. It includes the Ottonian Dynasty
(919-1024 AD). Beads were imported from
Scandinavia and continued to be imported from
Islamic western Asia. This phase is represented
in our research by glass from Deventer. The site
produced a wide range of glass compositional
types: nineteen wood-ash glasses, one pristine
and five recycled Foy 2 glasses, four mixed-alkali
glasses, three Egyptian II (natron) glasses, three
Roman (natron) glasses and one plant-ash glass.
A single example of high lime -low alkali window
glass is potentially a very early example of what
is generally considered a much later technology
though it appears there are other early examples
from east of the Rhine.389 Funnel beakers and
conical beakers were the dominant vessel forms
in the second half of the 9th and the 10th centuries.
By this time the vessels from Deventer were
made out of wood-ash glass, recycled Foy 2
glass, plant-ash glass and Egyptian II glass.
This shows that they were blowing these vessels
from whatever was available; unless the
different sources of natron glass were known
(which is a possibility for pristine glass) the
different origins of pale green recycled natron
glass – potentially with different working
properties - would normally be unknown. The
wood ash and natron glasses would certainly
have had different working properties: thickwalled wood ash glass vessels started to replace
the thin walled Carolingian beakers after c. 900
AD. A single chunk of raw Egyptian II and three
chunks of raw wood-ash glass were found on
the site. This could constitute tentative evidence
for a glass industry there, or it could simply
mean the glass was being traded through the
site.
The use of fresh high quality Egyptian II
glass for the manufacture of funnel beakers is
in contrast to the preceding phase when a very
high proportion of recycled glass was in use, for
example to make the funnel beakers found at
Dorestad. Apart from the four definite examples
of wood-ash funnel beakers from Deventer
(two dating to the late 9th century), there is an
105
—
8th–9th century palm funnel from Dorestad,390
as noted above, with another nine examples
from Hedeby.391 As is to be expected, many were
very poorly preserved. The balance of the Hedeby
funnel beakers analysed were twenty natron
(Foy 2) and one mixed-alkali glass.
In contexts dating to the 9th and 10th centuries
at Deventer 60% of the glasses were wood-ash
or mixed alkali glass. These would have been
manufactured as part of a decentralized production
system.392 Some of the earliest full wood-ash
glasses have been found at Stavelot in Belgium
and at Baume-les-Messieurs, Jura in France.
It also seems to be the case that France was a
centre for the production of mixed-alkali glass
(and perhaps wood- ash glass), one probable
location being Méru, France. In contrast to
Deventer, by the 10th century almost all glass
found in France was of the wood-ash type.393
Currently we do not know where the Deventer
mixed-alkali or wood-ash glasses were made.
Although France is one possible source for
Dutch wood-ash glass a more likely one was the
Viking trade network including through Hedeby,
perhaps from northern Germany, for example,
where funnel beakers and a crucible containing
wood-ash glass have been found.394 The higher
proportion of non wood-ash glass found in
10th century Deventer, including pristine Egyptian
II glass – and possible working debris- also
contrasts with the situation in France,
suggesting that a separate supply route from
Egypt was involved, probably including several
intermediaries.
390 Henderson 2012.
391 Kronz et al. 2016.
392 Meek, Henderson & Evans 2012; Van
Wersch et al. 2015; Adlington et al. 2019.
393 Pactat 2021.
394 Kronz et al. 2016.
107
—
7 Answering the research questions
In the introduction to this book a series of
research questions were posed. In this chapter
we discuss the extent to which we have been
able to answer them.
site. Nevertheless it would seem that the cobaltrich colorant was added directly to the pristine
base glass at some point.
2.
1.
What raw materials were used in the local
production of simple, monochrome Early
Medieval glass beads?
We have been able to identify the base glass
used in the manufacture of monochrome yellow,
white, red, and blue beads as a type of natron
glass, Foy 2 (see Section 5.10.1). No base glass
was fused from raw materials in the early
medieval Netherlands, so it needed to be
imported (see below). The majority of the
material examined scientifically that is relevant
to this research question forms part of some of
the most comprehensive evidence for the
manufacture of such beads in early medieval
northwest Europe, from Merovingian Maastricht
at the Jodenstraat site dating to the late 6th-early
7th century. The evidence consists of beads,
broken beads, rods, drops - and crucibles with
evidence for the manufacture of the opaque
yellow and white pigments attached to them.
The 26 trace element signature of the glasses,
determined by laser ablation inductively coupled
plasma mass spectrometry, showed that they
are unified compositionally by being made with
the same base glass: ‘pristine’ Foy 2.
The other raw materials which both coloured
and opacified the glasses used to make the
monochrome beads are the colorants. Tin-based
pigments were made at Maastricht which were
then used to make bright opaque yellow and
white glass. The tin based opacifier was also
found in red and greenish-blue coloured glass
beads which, in all likelihood, were also made
in Maastricht. A thick possible furnace or tray
fragment from Wynaldum also has a thick layer
of opaque yellow material attached to: it provides
probable evidence for the manufacture of leadtin yellow II pigment. Monochrome beads would
have been made there too. A separate source of
lead would have been involved.
Given that pristine Foy 3.2 glass (a sub group
of Foy 2) was used to make the translucent blue
bead and the four blue glass fragments found at
Maastricht it is a possibility that the cobalt colorant
was added to the base glass in Maastricht though,
unlike the evidence for the manufacture of lead
tin oxide II, there is no archaeological evidence
for the manufacture of cobalt blue glass at the
Where were these raw materials obtained
from?
Glass that has a Foy 2 trace element signature is
very likely to have derived from Egypt. 395 It was
made with silica sources that are characterised
by the presence of minerals that introduced
elevated elements such as titanium, manganese
and iron. It is difficult to be certain where the
tin used to make the opaque yellow and white
glasses was derived from. Possible sources are in
Cornwall in the UK396 and Turkey.397 Lead is a far
more common mineral with possible sources in
northern Spain, central and southern England,
the Saxon-Bohemian metalliferous mountains
(including the Erzgebirge) and Harz mountains
in Germany398, northern Italy and the Taurus
mountains in Turkey. Lead isotope analysis has
the potential to determine in an increasingly
precise way the source(s) of the lead used when
used in an appropriate way such as used for
European iron age glass399 and in ancient metal
research.400 The lead source used in early medieval
Dutch glass would be introduced either as an
impurity401 or deliberately as part of a colorant,
such as lead-tin yellow II. Therefore, lead
isotope analysis could potentially source the
lead raw material, but not the glass,
as discussed below.
Sub-questions here are:
i.
What substances were used to make the
different colours of glass in the artefacts
tested?
As noted above the monochrome opaque beads
from Maastricht, Jodenstraat were mainly
coloured with lead-tin oxide II and tin oxide.
Tin oxide was also combined with copper and
cobalt to produce opaque red and blue glass
respectively. Examples of yellow glass opacified
with lead-tin oxide II have also been identified
in glasses found at Wijnaldum and Dorestad.
Detailed analysis of the opaque red glass from
Maastricht, Jodenstraat revealed the presence
of micron-sized copper droplets, iron-rich
fayalitic slag402 and a crystalline phase containing
high tin associated with lead and silica: the
colorant is copper. The presence of micron sized
copper droplets and fayalitic slag was also found
in red beads from Wijnaldum.
395
396
397
398
399
400
401
402
Foy et al. 2003.
Meharg et al 2012.
Yener et al. 2015.
Wedepohl & Baumann 1997.
Huisman et al. 2017.
Artioli et al. 2020; Standish et al. 2021.
Henderson et al. 2005b.
Peake & Freestone 2012.
108
—
The pale green vessel glasses analysed were
coloured mainly with a combination of
manganese and iron oxides: if HIMT or one of its
recycled variants was used as the base glass,
elevated manganese (and iron) would have
modified the final colour, partly depending on the
melting atmosphere in the furnace. For example
HIMT glass cones form Gennep are olive-green,
yellow-green and amber-brown due to relatively
high levels of iron and manganese. Other vessel
glasses were coloured with cobalt producing a
blue colour and ferrous iron has produced an
amber colour. Deliberately added colorants/
opacifiers for glasses in both Merovingian
(Wijnaldum and Maastricht) and Carolingian
glasses (e.g. Susteren glass beads) are very similar:
elevated Fe oxide (up to around 5 weight %) and
CuO2 (up to around 1.7%) in red glass, high PbO
and SnO2 in opaque yellow (probably in the form
of PbSnO3 crystals) as well as a combination of Pb
and Sb which responsible for opaque yellow glass
(in the form of Pb2Sb2O7 crystals).
Opaque turquoise and blue tesserae from
Dorestad are coloured by copper and cobalt
respectively. Three of the five are opacified with
calcium antimonate. An opaque yellow glass
rod from Dorestad (DOR149) has a similar
chemical composition to opaque yellow glass
from Maastricht, Jodenstraat and Wijnaldum. It
has relatively high Na2O (10.6%) and PbO (10.7%)
contents and low K2O and MgO contents: it is also
coloured by lead tin yellow II. Therefore, it would
have been made using the same procedure as the
opaque yellow glass from Jodenstraat and
Wijnaldum. The Al2O3/SiO2, TiO2/Al2O3 ratios and
low Sb content of suggest that the base glass used
could also have been ‘pristine’ Foy 2 glass.
ii.
403 Foy et al. 2003.
404 Nenna 2014.
What compositional groups can be
distinguished in the glasses based on
chemical analyses?
The compositional groups that we have identified
amongst the glass samples that we have
chemically analysed are all known from the
literature (see Section 2.4.1 for a full discussion of
the glass types and associated literature).
By chemically and isotopically analysing recycled
Foy 2 (natron) glasses which was the dominant
composition between the early 8th century and
late 9th century we have been able to provide a
new explanation for some of the impurity levels
detected in the glass in a new way (see Sections
5.9 and Section 5.10.4).
We have identified the following compositional
types:
Natron (soda-lime) glasses:
• ‘Roman’
• High iron, manganese and iron (HIMT)
• Foy 2.1
• Foy 2.2 (recycled glass)
• Foy 3.2
• Egyptian II
• Levantine II
Plant ash (soda-lime) glass
Mixed-alkali (sodium and potassium) glass
High potassium glass
iii. What does this tell us about dating of
primary glass production of these groups?
There is no evidence for primary glass production
in the early medieval Netherlands. It is known
that glass of a ‘Roman’ composition was made
between the 1st and 4rd centuries AD on the
Levantine coast and in Egypt. Glass of the HIMT
composition was made between the 4th and 5th
centuries and it is probable that weak HIMT
(HIMT-2 = Foy 2) was also made from around the
mid 4th century though the recycled variants of
HIMT/Foy 2 glasses have been found in much
later contexts (see below) so will probably have
been recycled multiple times. Glass of Foy 2.1
and 3.2 compositions were probably made from
around the 6th century, the recycled Foy 2 (Foy
2.2) has been found in later contexts. Pristine
Levantine II glass was made from the 8th century
AD; Egyptian II was made between 760/780 and
870 AD.
Five Roman (vessel) glasses have been
identified, from Jodenstraat Utrecht and
Deventer. Our analyses suggest a Levantine
source for this relic glass. A single Levantine II
sample has been identified amongst our
samples dating to between 750 and 850.
Nintety-five mainly vessel glasses are of a
recycled Foy 2 glass composition (corresponding
to the Foy 2.2 group of Foy et al.).403 Highly
coloured opaque beads and translucent beads
from Maastricht, Jodenstraat were made with
pristine Foy 2 glass, as were the highly coloured
opaque glass beads from Wijnaldum.
Nine HIMT sensu stricto glasses have been
identified404, six being from Gennep, the site
with the earliest date (late 4th to mid-6th century)
from which we have obtained samples.
Eight ‘pristine’ Foy 2 glasses have been
109
—
identified: six from Gennep and 2 from
Maastricht, Jodenstraat, dating to between the
late 4th and early 7th centuries AD. Six ‘pristine’
Egyptian II glasses have been found: two from
Wijnaldum, one from Dorestad and three from
Deventer. Using context dates, three from
Deventer date to the 8th–9th centuries AD,
two date to the early 10th century and one dates
to between 950 and 1050 AD. They are consistent
with or slightly later than the suggested date
when Egyptian II glass was widely circulated,
in the 8th and 9th centuries AD. Previously it had
been suggested that HIMT glass in the widest
sense (including HIMT sensu stricto and Foy 2) was
in use from the middle of the fourth century until
the seventh century.405 However, our results show
that recycled Foy 2 glass (referred to as ‘recycled
Foy 2’ glass in this study, corresponding to Foy 2.2
group)406 was still the dominant compositional
group between early 8th century and late 9th
century in Dorestad and Wijnaldum vessel
glasses. Moreover, it remained in circulation till
possibly as late as the first half of the 10th century
according to our results from Deventer.
This shows that recycled Foy 2 glass was in
use centuries after the ‘pristine’ Foy 2 glass supply
dried up in northwestern Europe. The recycled
version of Foy 2 glass is labelled as Foy 2.2
subgroup in the work of Foy and colleagues,407
and is known from a very limited number of
assemblages in France, Italy and Spain that are
typically dated to the end of the seventh and into
the eighth centuries AD.408
Twenty six wood ash/mixed alkali glasses
have been found in this study. Two samples
derived from Dorestad, two from Susteren and
twenty-three from Deventer. These dates are
consistent with the suggested date from when
wood ash glass started to be made and was
circulating in Europe from the late 8th century
onwards.
Six plant ash samples have been identified;
all date to post-9th century AD the time when
plant ash glass started to be the dominant
technology in western Asia. Islamic plant ash
glasses started to appear further west and east
after this time. Plant ash glass was used to make
single funnel beakers from Susteren and
Deventer, the bodies of two decorated traildecorated beads from Susteren and it was
formed into single examples of gold and silver foil
beads found at Wijnaldum, imported from
western Asia. Raw plant ash glass was therefore
imported from western Asia to the west where it
was remelted to form funnel beakers and some
(trail-decorated) beads.
iv. What do the isotope ratios (Sr, Nd)
obtained from the glasses of selected
compositional types tell us about the their
origin and dating?
Thirteen natron glass samples have a typical
range of Nd–Sr signatures for such glass with
εNd between -5 and -7, and 87Sr/86Sr between
0.7085 and 0.7093,409 (Figure 5.24); one sample
was contaminated (see below). The twelve clean
natron samples are of the Foy 2 type: seven are
‘pristine’ Foy 2 glass and five are recycled Foy 2.
Their 87Sr/86Sr ratios stretch in a rather wide
range between 0.7085 and 0.7093. Six pristine’
Foy 2 glasses form a cluster at 87Sr/86Sr of 0.7086,
a typical value for glass produced in Egypt, the
commonly suggested origin of Foy 2 glass. Five
‘pristine’ glasses are from bead making at
Jodenstraat, Maastricht (late 6th- early 7th century
AD), the 6th being a gold foil bead from Wijnaldum.
The 87Sr/86Sr of 0.7088 and 0.7093 for recycled
Foy 2 (vessel) glasses all dating between the late
8th to mid 9th centuries AD is higher than the
typical Egyptian range of values: it is likely that
the 87Sr/86Sr values have been modified by the
recycling process. We suggest that the mixing of
wood ash glass, with very high 87Sr/86Sr values,
with natron glass is the explanation; this agrees
with the results from trace element analysis.
We determined the Nd–Sr isotopic signatures
for two wood ash-lead linen smoothers from
Dorestad with very high 87Sr/86Sr. A single plant
ash glass, an Islamic silver foil glass bead has the
lowest 87Sr/86Sr signature out of the 18 samples
analysed and may have a provenance near
Baghdad in Iraq. A single ribbed glass vessel and
a glass tesserae are of ‘Roman’ chemical
compositions and both have the anticipated
typical Levantine natron isotopic signatures
with 87Sr/86Sr values of 0.7090 and 0.7092
respectively.
v.
What networks inside and outside the
Netherlands were used in obtaining glass,
including the colourants used?
In all cases it is difficult to ascertain how many
intermediaries were involved during the process
of obtaining glass, whether it was unworked
‘pristine’ raw glass, recycled raw glass or fully
formed objects. The colorants used deliberately
405
406
407
408
409
Freestone et al. 2018.
Foy et al. 2003.
Foy et al. 2003.
Ares et al. 2019.
Brems et al. 2018.
110
—
(and their sources) which formed part of
networks of interaction are discussed under
question 2.
The source of many of the Carolingian
beakers found in the Netherlands is considered
to be Germany/Cologne. There is clear evidence
that opaque yellow (and other brightly coloured)
glass beads were made at both Maastricht and
Wijnaldum, but there is still a possibility that
beads found at Wijnaldum originated from
Maastricht, for example. Lead may well have
been obtained from Germany too; the tin source
may potentially have been Cornwall in the UK or,
less likely, Turkey.
Egypt was clearly the ultimate source
for Egyptian II glass and pristine Foy 2 glass.
Our isotopic results support an Egyptian
provenance which, up to now, has mainly been
suggested using the results of chemical analyses.
There is little doubt that pristine HIMT glass
was originally made in Egypt too, as recently
confirmed using Nd and Sr isotope analysis,
helping to distinguish it from Levantine glass.410
There is a single example of an early medieval
pristine Levantine glass in our study, underlining
that the primary source of pristine glass for early
medieval glass working was ultimately Egypt.
Plant ash glass would have been obtained
from western Asia and using trace element analysis
we have determined that northern Syria, Iraq/
Iran and the Levant are likely sources for
the plant ash glasses identified in this study.
Plant ash glasses were mass produced in early
Islamic cosmopolitan centres such as Damascus,
Baghdad and Samarra411 and started to appear
on European sites in any number as a result of
Viking trade networks.
In the absence of glass making furnaces in
the Netherlands, Belgium and northern France
are possible sources of wood ash glass but a
more likely source is Germany which formed
part of the Viking trade network, including
Hedeby in northern Germany. The probable
evidence for mixed-alkali glass production in
France suggests that this is one possible source
for the type of glass in the early medieval
Netherlands though again Germany may also
have been a source.
410 Gliozzo et al. 2019.
411 Henderson 2022.
vi. To what extent were the raw materials
or semi-finished products derived from
primary production sites, and to what
extent did they derive from systematic
recycling of glass, including Roman?
In ‘pristine’ (non-recycled) natron glass the
levels of a few correlated elements such as Pb,
Sb and Cu, Ba, Rb, Cs are low or very low, but for
recycled glass the levels of these elements are
higher. Elevated levels of Pb and Sb most
consistently demonstrate recycling has occurred,
so they have been chosen here as one of the
means of distinguishing ‘pristine’ glass samples
from recycled glass samples. The criterion for
the identification of a ‘pristine’ glass is that the
Pb and Sb contents are both under 1000 ppm,
following previous conventions.
We have noted that there is a higher
proportion of ‘pristine’ (unrecycled) natron glass
imported as a raw material, however indirectly,
from primary production sites found on
Merovingian sites than on Carolingian sites.
For example ‘pristine’ Foy 2 glasses from the late
6th-early 7th century Maastricht Jodenstraat site
were used as the base glass to make highly
coloured yellow, white and red opaque glass
beads there and perhaps the translucent glass
beads too. The same was found for the few
semi-finished products from the site. The late
4th to mid 6th century glass vessels analysed
from Gennep are pristine HIMT sensu stricto or
‘pristine’ Foy 2 glass. Nevertheless, elevated
levels of Sb and Pb suggest there was an admixture
of highly coloured Roman vessel glass or glass
tesserae to some glass as early as this.
In contrast if we use the results of 55 vessel
glasses from Dorestad dating to the Carolingian
period we have found that, apart from two
wood ash glasses, and an Egyptian II glass,
the remaining glasses are of the Foy 2 composition.
These Foy 2 glasses contain at least 1000 ppm of
Pb or 1000 ppm of Sb, or both, as well as
elevated levels of Rb and Cs, characteristics of
recycled glasses.
By the c. 850 and into the 10th-11th centuries
the site of Deventer provides an intriguing
contrast to earlier periods, reflecting an
important period of technological transition.
From the glass that is sufficiently unweathered
to provide a valid analysis nineteen out of thirtyeight samples are of a wood ash composition,
twelve are natron glass, four are mixed-alkali
and one is a plant ash glass. Therefore, wood
111
—
ash and plant ash glasses are unrecycled; the
mixed-alkali glasses are the product of recycling.
The twelve natron glasses surprisingly include
three of a Roman composition, three of a
‘pristine’ Egyptian II composition, one ‘pristine
Foy II glass and five recycled Foy 2 glasses.
Therefore the majority of glasses analysed from
Deventer are pristine and have not been recycled.
It would seem therefore, that the lowest
proportion of recycled glass was imported and
used in the late 6th-early 7th century and between
c. 850 and 10th-11th centuries, partly in the latter
case because decentralized primary production
of wood ash glass had developed. Between
these dates there was clearly a dependence on
using recycled glass for making the majority of
vessels found at Dorestad.
The research also provides building blocks
for two NOaA questions:
What are the nature, manifestations, extent
and context of craft specialization? (NOaA 2.0
question 67)
The description of glass craft specialization is
discussed in detail in Chapter 3, some of which is
ephemeral. Therefore only the most significant
evidence is discussed here. Monochrome glass
bead production was a craft specialization in the
early medieval Netherlands. The most
comprehensive evidence in Europe to date for
the manufacture, especially of opaque yellow
glass beads, has been excavated from the
Jodenstraat 30 site in Maastricht where a
rubbish pit filled with the debris from glass bead
making was found.412 The pit also contained
waste from copper-alloy-working and amberworking.413 Based on the pottery finds, the pit
was filled sometime in the late 6th to early
7th centuries. The debris from bead making
consists of 750 glass objects which represent the
full range of waste from glass bead production.
The production waste was divided into eight
main groups: glass rods (n=369), ‘punty’ glass
from a beadmaker’s tool (n=36), glass threads
with and without tweezer marks (n=17), glass
drops (n=39), finished and failed beads (n=123),
crucibles (n=38, EMN=17), cullet or scrap glass
(n=20), glassy slags/fuel ash slags (n=53) and
non-diagnostic fragments (n=55) which include
(small lumps of) melted glass and fragments
that are too small to classify. All waste categories
are dominated by opaque yellow glass (apart
from scrap glass and glassy slags). Almost all
beads are wound and have tapering perforations
showing they were made by winding melted
glass around a mandrel, a bead-making tool
with a conical point. Such a tool may have been
found at the Rijksarchief site. The crucibles
associated with bead production consist of
38 fragments from at least 17 coarse-ware
cooking pots (Wölbwandtöpfe). In 15 cases,
only the lower halves of the pots were used to
melt highly coloured opaque yellow glass.
Two crucible bases with opaque white pigment
are also present. Drops of translucent greenish
glass among the waste products suggest this
glass colour was worked on or near the site.
Excavations at the Maastricht Mabro site
produced twelve fragments of crucibles with
glass deposits; eleven of these date to the 6th7th centuries. The colours of the glass in the
crucibles are colourless or pale green or opaque
yellow. A number of wound beads have been
found: they are either monochrome or decorated
with trails in contrasting colours. The site has
not been published, so it is impossible to state
which beads are likely to be local products, but
given the dates for the crucibles, those beads
dating to roughly the 2nd half of the 6th and first
half of the 7th century are the most likely
candidates: a small globular bead, a bi-globular
bead and a possible cylindrical bead of opaque
yellow glass, a medium-sized globular bead of
opaque white glass and a short cylindrical bead
of opaque red glass. There are also two
polychrome, trailed beads: an opaque
white globular bead with translucent light blue
narrow crossing trails and an opaque white
disc-shaped bead with translucent dark blue
crossing trails.
Excavations of the Maastricht-Rijksarchief
site are discussed by Hulst.414 Dozens of rubbish
pits were full of the evidence for glass working,
antler working and iron working and dated by
(late) 5th and 6th century pottery. The evidence
of glass working consists of 55 fragments of
early Merovingian vessel glass, two fragments
of glass rods, 25 beads, drops of glass, melted
glass, glassy slags and fragment of a glass
crucible.415 Furthermore, one fragment of a glass
crucible, glassy slags (included some attached
to a furnace floor or thick tray fragment) were
found. A forged iron rod which is round in
section at one end and square in section at the
other match the perforations of beads found on
the site; it is an example of a very unusual bead-
412 Sablerolles, Henderson & Dijkman 1997.
413 For waste from amber-working, see
Dijkman 2013.
414 Hulst 1992.
415 Hulst 1992.
112
—
416 Sablerolles 2023. (basispublicatie
chapter 29).
417 Pottery identification by Jan de Koning.
418 Nyst 2003, 13.
419 Preiß 2010, 125.
making mandrel around which glass filaments
would have been wound.
Evidence for the manufacture of opaque
yellow glass was also found at Wynaldum in the
form of a thick fragment covered with a layer of
opaque yellow material of between 1.0-1.3 cm
thick. The yellow substance would probably
have been used to make beads once formed into
glass. Two small flattened opaque yellow and
white beads were found and are probably local
products. This evidence was found amongst
waste produced by a blacksmith/bronze-caster.
The dump dates to the last quarter of the 6th
and the first quarter of the 7th century and is
contemporary with the glass-working evidence
from the Jodenstraat site in Maastricht. A tessera
and a piece of punty glass glass were found in
Carolingian contexts. There is a possibility that
they relate to bead production.
Brightly coloured Merovingian monochrome
beads would presumably have been used by local
populations but would also have been exported.
As mentioned in Chapter 6 there was evidently a
demand especially for bright yellow glass beads
from at least the middle Merovingian period;
such glass must have had social significance but it
is difficult to suggest what it might have been.
The five fragments of glass production
waste have been found at the monastic site of
Susteren-Abdijterrein suggest that glass was
worked there during the early medieval period.416
The finds are two fragments of glass crucibles
(one with a layer of cobalt blue glass and the
other with a layer of bluish green and colourless
glass attached), a partially melted Roman tessera,
a glass fragment from glassblowing tool and a
possible fragment of opaque yellow raw glass.
The crucible fragments derive from a context
dated to 800–1300 which contains 60%
redeposited Carolingian material. One crucible
fragment, which contains bluish-green glass,
is probably made of Carolingian Badorf ware.417
The second crucible fragment which contains
cobalt blue blue glass was made of possible grey
Meuse Valley ware: it may have been used to
make dark blue window quarries on the site,
the first such evidence from the Netherlands.
A partially melted, opaque dark blue Roman
tessera may have formed part of the manufacture.
A translucent dark bluish-green glass fragment
with thick walls is covered on its concave inside
with dark grey iron scale from a glassblowing
tool. This fragment is the only direct evidence
for glassblowing in the Netherlands since the
Roman period.
Excavations at Wijk bij DuurstedeVeilingterrein and Frankenweg/Zandweg
revealed evidence for Carolingian bone- working
(and those at the adjoining Parkeerplaats Albert
Heijn (PPAH) site produced traces of metal
working and loom weights).418 Some thirty-six
fragments of glass-working waste were found,
including eight tesserae, a large fragment of a
glass crucible of a late Merovingian form found
in a pit with two blue tesserae, a regular and an
irregular drop of translucent pale green glass, a
small dark sphere, six melted lumps of translucent
pale green vessel glass and a melted fragment of
‘black’, deep olive-green glass. The crucible
contains almost colourless glass of c. 1 mm
thickness with a thicker patch of opaque white
glass which is probably a melted tessera.
Preiß points out that defects in the translucent
glass probably indicate locations where other
(crushed?) tesserae had been attached.419
The crucible may be linked to bead-making,
but it also may be linked to vessel or window
glass production at the site.
Excavations at Wijk bij Duurstede –
Veilingterrein and Frankenweg/Zandweg
revealed evidence for glass production along
with iron smithing, brass production, weaving
(wool) and amber working; glass and amber
working evidence were sometimes found in
close proximity. The largest category of glass
working evidence is tesserae, almost all a blue/
green colour. Fragments of translucent bluishgreen and dark blue glass probably result from
breaking up glass ingots, cakes or raw glass
chunks. There are dark blue drops, two square
sectioned opaque yellow glass rods and opaque
white glass which were probably used for beadmaking. A lump of opaque yellow glass has part
of a composite opaque yellow glass rod and
yellow punty glass from a glassworker’s tool
melted onto it. An opaque green object is waste
from glass bead production; there is also a
twisted bi-chrome (opaque white and bluegreen) rod fragment. Such rods were used or
decorating Carolingian glass vessels. A thick
opaque yellow glass fragment has embedded
iron oxide scale in it as well as ceramic fragments.
It may have been used in bead production or
decorating 8th century vessels. A ‘bright blue’
tessera and clay covered with a thick layer of
‘blobby greenish glass’ were found on the
113
—
Veilingterrein site.420 Moreover two small
spheres of whitish translucent glass were
found.421
The finds from Dorestad therefore provide
evidence for local bead production; vessels may
also have been made there. The relatively small
amount of glass production debris derives from
pits, wells and ditches; wet sieving was not
carried out universally. It is therefore difficult
to decide whether the evidence indicates that
production was on a small-scale, at household
level, or on a larger scale for export.
Excavations at Utrecht Domplein produced
twelve crucible fragments with a layer of glass
attached to their insides. The glass is either
apparently colourless, green with red streaks
of glass running through it or pale green. Red
marbled translucent blue-green/bluish-green
glass was popular for making late Merovingian
and Carolingian glass vessels. Excavations at
Utrecht Oudwijkerdwarsstraat produced an
irregular drop of bluish-green glass as well
as some crushed fragments. The discovery of
fourteen hundred amber fragments shows
that some craft activity occurred on quite a
large scale.
The evidence from Merovingian Leidsche
Rijn-Leeuwesteyn Noord consists of a single
crucible fragment with a layer of blue-green
glass with marbled opaque red streaks perhaps
used to make windows. The Merovingian
Rijnsburg-Abdijterrein site produced some
useful evidence for glass bead production
consisting of finished, unfinished and failed
beads, glass rods and three pieces of punty
glass from a beadmaker’s tool; eight crucible
fragments and two lumps of fired clay covered
with translucent greenish glass could have been
part of a glass-working furnace floor. A single
crucible fragment with green glass attached may
be contemporary. Nearly half of the waste from
Rijnsburg is opaque yellow. The crucible fragments
have yellow and colourless glass attached to
them. This material was not available for this
research project. It is possible that colourless
glass was modified on site using lead-tin yellow.
The chemical composition of the yellow glass in
the crucible, the beads and the rods are similar
and therefore likely to have been made on site.
The bead types made at Rijnsburg were
monochrome opaque yellow globular, bi- and
tri-globular beads and bi-globular red beads.
Trailed beads include bi-globular beads of red
glass with both white crossing trails and a
white spiral, tri-globular beads of opaque red
glass with opaque white crossing trails and
white beads with translucent blue crossing trails
were also possibly made at the site. It is likely
that the production phase occurred in the
7th century partly based on dated bead
typologies (see Section 3.8).
Two hundred and one (unstratified) glass
tesserae from the terp at Wierum have been
interpreted as a supply of ‘raw’ glass for making
beads in the early medieval period.422 The terp
probably dates to the 8th/9th centuries.
Most tesserae have rounded profiles so appear
to have been heated. Five fragments of highly
coloured early Roman vessel glass, one
fragment of Roman or early medieval vessel
glass, three opaque green nearly colourless and
translucent dark blue plano-convex drops of
glass and thirteen irregular drops/melted lumps
of nearly colourless, pale green and pale bluegreen glass have also been found. This evidence
from Wierum may have resulted from a travelling
beadmaker visiting terp sites such as Wijnaldum
in the northern coastal region, which was most
easily accessible by boat from the central
riverine area, with Dorestad at its centre.
The latest site to be considered in our
research project is Deventer- Stadhuiskwartier.
The evidence for the glass industry is scanty,
dates to between c. 850 and c. 1050 and derives
from waste pits or cesspits. Two pits yielded
production waste of iron smithing, boneworking and textile production; glassy slag dates
to 850–900. Glass working evidence dating to
between 900 and 925 consists of a heavily
weathered chip of glass with a conchoidal
fracture, a heavily weathered triangular fragment
and a small, heat-affected fragment which may
be part of a pulled thread. Waste from the 900–
950 phase consist of a small lump of translucent
clear bluish-green raw glass, a heavily weathered
fragment with a triangular section and a heavily
weathered fragment with two irregular, heataffected surfaces. An unusual fragment dating to
950-1050 consists of a translucent bright bluishgreen and deep turquoise glass layer covered by
a very thin film of opaque red glass. The turquoise
colour is very similar to a contemporary fragment
of very thin flat glass, either window glass or a
glass inlay, found in the same area.
It is evident that early medieval glass
working in the Netherlands was frequently
420 Van Doesburg 2004.
421 Langbroek 2021b, 64, table 7, findnrs
Zandweg WD 754.2.63b, WD 754.2.63b.
422 Crocco et al. 2021.
114
—
423 Verhulst 2002, 72-84.
associated in industrial areas on the sites with a
range of other industries including iron smithing,
copper-alloy production and brass production,
as well as amber working and weaving. It is likely
that the same fuels (yet to be determined) would
have been used for glass and metal production,
depending on the maximum temperature required.
The artisans involved in glass working may have
taken part in other activities associated with
shared aspects of other high temperature
industries, such as obtaining fuel, making
crucibles and building kilns/furnaces or separate
groups were involved in such activities.
During the Carolingian period such industrial
organisations, that were involved in several
different production activities on particular sites,
have been found, for example, at San Vincenzo
al Voltorno, Augsburg and Corvey.423
Where do non-local raw materials of utilitarian
objects come from? (NOaA 2.0 question 139)
There is evidence that much of the base glass
used to make utilitarian beads and vessels
originated in Egypt; by the Carolingian period
most of it was recycled to the extent that its
source is indeterminate but most still has
evidence that the ultimate source of almost all
‘pristine’ unrecycled glass was Egypt. The small
number of plant ash glasses used to make
utilitarian objects in the early medieval
Netherlands probably derived from Iraq/Iran,
Syria and the Levant. The lead and perhaps the
cobalt may have originated in Germany, but this
is to be confirmed - and there are other possible
sources. The possible sources of tin are more
restricted: Cornwall in the UK is one such source;
Turkey is another one.
115
—
Bibliography
Adlington, L.W., I.C. Freestone,
J.J. Kunicki-Goldfinger, T. Ayers,
H. Gilderdale Scott & A. Eavis
2019: Regional patterns in
medieval European glass
composition as a provenancing
tool, Journal of Archaeological
Science, 110, 104991.
Aldsworth, F., G. Haggerty,
S. Jennings & D. Whitehouse
2002: Medieval Glassmaking at
Tyre, Lebanon, Journal of Glass
Studies 44: 49–66.
Alfonsi, H. & Gandolfo P. 1997:
L’épave Sanguinaire A, Cahiers
d’Archéologie subaquatique 13,
35-74.
Allan, J.W., 1973: Abū’l-Qāsim’s
treatise on ceramics, Iran, 11,
111–120.
Andersen, J.H. & T. Sode 2010:
The Glass Bead Material, in:
M. Bencard & H.B. Madsen
(eds), Ribe Excavations 1970–76,
vol. 6., Ribe Esbjerg:
Sydjysk Universitetsforlag
(Ribe Excavations, Jutland
Archaeological Society
Publications), 17–59.
Artioli, G., C. Canovaro, P. Nimis
& I. Angelini 2020: LIA of
Prehistoric Metals in the
Central Mediterranean Area:
A Review, Archaeometry, 62,
53-85.
Aunay, C., A.A. Berthon,
B. Gratuze, M. Guérit,
J. Motteau & I. Pactat 2020:
Le verre creux du VIIIe au Xe
siècle dans la vallée de la Loire
moyenne et de la Vienne. Essai
typo-chronologique et
archéométrique, in: Inès Pactat
& C. Munier (eds), Le verre du
VIIIe au XVIe siècle en Europe
occidentale, Les Cahiers de la
MSHE Ledoux, Presses
universitaires de FrancheComté, 293–314.
Bandiera, M., M. Verità,
P. Lehuédé & M. Vilarigues
2020: The technology of
copper-based red glass sectilia
from the 2nd century AD Lucius
Verus villa in Rome, Minerals,
10(10), 875.
Arbman, H., 1937: Schweden und
das Karolingische Reich. Studien
zu den Handelsverbindungen des
9. Jahrhunderts. Kungl, Stockholm
(Vitterhets Historie och Antikvitets
Akademiens Handlingar Del 43).
Barber, D.J, I. Freestone &
K.M. Moulding 2009:
Ancient Copper Red Glasses:
investigation and analysis by
microbeam techniques,
Chapter 11 in: A.J. Shortland,
I.C. Freestone & T. Rehren
(eds), Mine to Microscope,
Oxford: Oxbow Books & the
David Brown Book Company.
Ares, J.D.J., A.V.E. Guirado,
Y.C. Gutiérrez & N. Schibille
2019: Changes in the supply of
eastern Mediterranean glasses
to Visigothic Spain, Journal of
Archaeological Science, 107,
23–31.
Bardet, A.C., 1995: Pottery
traded to Dorestad: some
explanatory archaeometrical
analyses of early medieval
Rhenish wares, Berichten van de
Rijksdienst voor Oudheidkundig
Bodemonderzoek, 41, 187–251.
Bass, G.F., 1984: The nature of
the Serçe Limani glass, Journal
of Glass Studies 26, 64–9.
Bass, G.F., B. Lledo, S. Matthews
& R.H. Brill 2009: Serçe Limani.
Volume 2: The glass of an eleventhcentury shipwreck, Rachel
Foundation Nautical Archaeology
Series, College Station, Texas
A & M University Press.
Boschetti, C., C. Leonelli,
R. Rosa, M. Romagnoli,
M. Valero Tévar & N. Schibille
2020: Antimony or Opacified
White Glass Tesserae, Heritage,
3, 549–560. https://doi.
org/10.3390/heritage3020032
Baumgartner, E. & I. Krüger
1988: Phoenix aus Sand und
Asche. Glas des Mittelalters,
München.
Boschetti, C., B. Gratuze &
N. Schibille 2020: Commercial
and social significance of glass
beads in Migration-Period
Italy: the cemetery of Campo
Marchione, Oxford Journal of
Archaeology, 39, issue 3, 319–342.
Bertini, C., J. Henderson &
S. Chenery 2020: Seventh to
eleventh century CE glass from
Northern Italy: between
continuity and innovation,
Archaeological and Anthropological
Sciences, 12:120. https://doi.
org/10.1007/s12520-02001048-8
Besteman, J.C., J.M. Bos,
D.A. Gerrets, H.A. Heidinga &
J. de Koning (eds) 1999: The
Excavations at Wijnaldum. Reports
on Frisia in Roman and Medieval
times, Volume I, Rotterdam/
Brookfield.
Birck, J.L., 1986: Precision
of K-Rb-Sr Isotopic Analysis
- Application to Rb-Sr
Chronology, Chemical Geology,
56, 73-83.
Boschetti, C., J. Henderson,
J. Evans & C. Leonelli 2016:
Mosaic tesserae from Italy and
the production of Mediterranean
coloured glass (4th century
BC–4th century AD). Part I:
Chemical composition and
technology, Journal of
Archaeological Science, Reports,
7, 303–311.
Brems, D., M. Ganio,
K. Latruwe, L. Balcaen,
M. Carremans, D. Gimeno,
A. Silvestri, F. Vanhaecke,
P. Muchez & P. Degryse 2013a:
Isotopes on the beach, part 1:
strontium isotope ratios as a
provenance indicator for lime
raw materials used in Roman
glass-making, Archaeometry 55,
214-234.
Brems, D., M. Ganio,
K. Latruwe, L. Balcaen,
M. Carremans, D. Gimeno,
A. Silvestri, F. Vanhaecke,
P. Muchez, & P. Degryse 2013b:
Isotopes on the beach, part 2:
neodymium isotopic analysis
for the provenancing of Roman
glass-making, Archaeometry,
449-464.
Brems, D., I.C. Freestone,
Y. Gorin-Rosen, R. Scott,
V. Devulder, F. Vanhaecke &
P. Degryse 2018: Characterisation
of Byzantine and early Islamic
primary tank furnace glass,
Journal of Archaeological Science:
Reports, 20, 722–735.
116
—
Brill, R.H., 2005: Chemical
Analyses of Some Sasanian
Glasses from Iraq, in: David
Whitehouse, Sasanian and Post
Sasanian Glass in The Corning
Museum of Glass, Corning: the
museum, 2005, 65–88.
Brill, R.H., 2006: Scientific
Investigations of Some
Glasses from Jarrow and
Monkwearmouth, in: R. Cramp,
Wearmouth and Jarrow monastic
sites, volume 2, English
Heritage.
Brüggler, M., 1994: Burgus und
Glaswerkstatt der Spätantike
bei Goch-Asperden, Berichte.
Grabung–Forschung–Präsentation,
Köln.
Bult, E.J., J. van Doesburg &
D.P. Hallewas 1990: De
opgravingscampagne in de
Vroeg-Middeleeuwse
nederzetting op de Woerd bij
Valkenburg (Z.H.) in 1987 en
1988, in: E.J. Bult & D.P. Hallewas
(eds), Graven bij Valkenburg III,
het archeologisch onderzoek in
1987 en 1988, Delft, 147–166.
Cabart, H., I. Pactat &
B. Gratuze 2017: Les verres du
Haut Moyen âge issus des
fouilles du monasterium
Habendum (Vosges, France),
Annales du 20e Congrès de
l’Association Internationale pour
l’Histoire du verre, 346–353.
Cagno, S., M.B. Badano,
F. Mathis, D. Strivay &
K. Janssens 2012: Study of
medieval glass fragments from
Savona (Italy) and their relation
with the glass produced in
Altare, Journal of Archaeological
Science, 39 (7), 2191–2197.
Callmer J., 1977: Trade beads and
bead trade in Scandinavia ca. 800–
1000 AD, Acta Archaeologica
Lundensia, 4 , 11, Habelt, Bonn.
Callmer, J., 2003: Wayland. An
essay on craft production in the
Early and High Middle Ages in
Scandinavia, in: L. Larsson & B.
Hårdh (eds), Centrality-Regionality:
the social structure of southern
Sweden during the Iron Age,
Stockholm. Acta Arch. Ludensia
Ser. 8°, 40, 337-361.
Callmer, J. & J. Henderson
1991: Glassworking at Åhus,
S. Sweden (eighth century AD),
Laborativ Arkeologi, 5, 143–154.
Casellato, U., F. Fenzi,
P. Guerriero, S. Sitran,
P.A. Vigato, U. Russo,
M. Galgani, M. Mendera &
A Manasse 2003: Medieval and
renaissance glass technology
in Valdelsa (Florence). Part 1:
raw materials, sands and
non-vitreous finds, Journal of
Cultural Heritage, 4, (4), 337-353.
Ceglia, Andrea, P. Cosyns,
K. Nys, H. Terryn, H. Thienpont
& W. Meulebroeck 2015: Late
antique glass distribution and
consumption in Cyprus: a
chemical study, Journal of
Archaeological Science, 61,
213–222.
Ceglia, Andrea, P. Cosyns,
N. Schibille & W. Meulebroeck
2019: Unravelling provenance
and recycling of Late Antique
glass with trace elements,
Archaeological and Anthropological
Sciences, 11, 279–291.
Conte, S., T. Chinni, R. Arletti
& M. Vandini 2014: Butrint
(Albania) between eastern and
western Mediterranean glass
production: EMPA and
LA-ICP-MS of late antique and
early medieval finds, Journal of
Archaeological Science, 49, 6–20.
Corbella, F., 2017: The Lent
Merovingian glass beads
µ-Computed Tomography, XRF
and SEM analysis of complex glass
beads, Unpublished Internship
report at TU Delft for École
Polytechnique, Université
Paris-Saclay, s.l., s.n., 64 pp.
Crocco, R., H. Huisman,
Y. Sablerolles, J. Henderson,
B. van Os & A. Nieuwhof 2021:
Hunting colours: origin and
reuse of glass tesserae from
the Wierum terp, Archaeological
and Anthropological Sciences,
13(9), 1–22. https://doi.
org/10.1007/s12520–021–
01391–4
De Bruin, J., 2018: Living in
Oegstgeest 575–725 AD, in:
M. Kars, R. van Oosten,
M.A. Roxburgh & A. Verhoeven,
Rural riches & royal rags? Studies
on medieval and modern
archaeology, presented to Frans
Theuws, Zwolle: SPA-Uitgevers,
20–25.
De Bruin, J., C. Bakels &
F. Theuws (eds) 2021:
Oegstgeest. A riverine settlement
in the early medieval world system,
Bonn: Habelt Verlag.
De Koning, J., 2012: Het
aardewerk, in: J. Dijkstra (ed.),
Het domein van de boer en de
ambachtsman. Een opgraving op
het terrein van de voormalige
fruitveiling te Wijk bij Duurstede:
een deel van Dorestad en de villa
Wijk archeologisch onderzocht,
ADC monografie 12, Amersfoort,
117–235.
De Koning, J., 2015: Onder het
stuifzand. Overstoven vroegmiddeleeuwse nederzettingen bij
Bloemendaal de opgravingscampagnes Groot Olmen 2005:
2006 en 2007, Zaandijk.
De Sigoyer, S.D.B., C. Peters,
S. Mathieu & C. Fontaine
2005: Vestiges de fours de
verriers d’epoque Mérovingien
à Huy aux Ruelles (Belgique):
Aperçu des trouvailles, Bulletin
de l’Association Française pour
l’Archéologie du Verre, 2005,
29–33.
Degryse, P. & J. Schneider
2008: Pliny the Elder and Sr–Nd
isotopes: tracing the provenance
of raw materials for Roman
glass production, Journal of
Archaeological Science, 35(7),
1993–2000.
Degryse, P., I. Freestone,
S. Jennings & J. Schneider
2010: Technology and provenance
study of Levantine plant ash glass
using Sr–Nd isotope analysis,
Römisch-Germanisches
Zentralmuseum.
Dekówna, M., 1978: Les verres
de Haithabu. Rapport
préliminaire, Annales du
7e Congrès de l’Association
Internationale pour l’Histoire du
Verre, Liège, 167-188.
117
—
Dekówna, M., 1980: Methods
of examining ancient glasses,
in: J. Schild (ed.), Unconventional
Archaeology, Wroclaw:
Ossolineium, Wydawnictwo
Polskiej Akademii Nauk,
213–233.
Dell’Acqua, F., 1997: Ninthcentury window glass from the
monastery of San Vincenzo al
Volturno (Molise, Italy), Journal
of glass studies, 33–41.
DePaolo, D.J. & G.J. Wasserburg
1976: Nd isotopic variations
and petrogenetic models,
Geophysical Research Letters, 3(5),
249–252.
Dijkman, W., 2013: Artisanal
Activities in Merovingian
Maastricht, Medieval and Modern
Matters, 4, 23–39.
Dijkstra, J. (ed.) 2012: Het
domein van de boer en de
ambachtsman. Een opgraving op
het terrein van de voormalige
fruitveiling te Wijk bij Duurstede:
een deel van Dorestad en de villa
Wijk archeologisch onderzocht,
Amersfoort (ADC-monografie
12).
Dijkstra, M.P.F. 2011: Rondom
de mondingen van de Rijn en
Maas. Landschap en bewoning
tussen de 3e en de 9e eeuw in
Zuid-Holland, in het bijzonder
de Oude Rijnstreek, (PhD thesis
University of Amsterdam).
Dijkstra, M., Y. Sablerolles &
J. Henderson 2011: A traveller’s
tale. Merovingian glass bead
production at Rijnsburg, the
Netherlands, in: C. Theune,
F. Biermann, R. Struwe &
G.H. Jeute, Zwischen Fjorde und
Steppe, Festschrift für Johann
Callmer zum 65. Geburtstag,
Rahden/Westf.: Leidorf, 175–199
(Internationale Archäologie;
Studia honoraria; Band 31).
Dodt, M., A. Kronz & K. Simon
2021: Production of early
medieval glass in Cologne, in:
A. Willemsen and H. Kik (eds),
Dorestad and its networks,
Proceedings of the third
‘Dorestad Congress’ held at the
National Museum of Antiquities,
Leiden, the Netherlands 12–15
June 2019, 179–191.
Evison, V.I., 1972: Glass Cone
Beakers of the “Kempston”
Type, Journal of Glass Studies,
48–66.
Fontaine, S. D. & D. Foy 2007:
L’ épave Ouest-Embiez 1, Var; le
commerce maritime du verre
brut et manufacturé en
Méditerranée occidentale dans
l’Antiquité, Revue Archéologique
de Narbonnaise, 40, 235-268.
Foster, H.E. & C.M. Jackson
2009: The composition of
‘naturally coloured’ late Roman
vessel glass from Britain and
the implications for models of
glass production and supply,
Journal of Archaeological Science,
36(2), 189–204.
Foy, D. & M.-D. Nenna 2001:
Tout feu tout sable: mille ans de
verre antique dans le Midi de la
France, Edisud, Marseilles.
Foy, D., M. Picon, M. Vichy &
V. Thirion-Merle 2003:
Caractérisation des verres de la
fin de l’Antiquité en
Méditerranée Occidentale:
l’émergence de nouveaux
courants commerciaux, in:
D. Foy & M.-D. Nenna (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, juin
2001, 41–85, Éditions Monique
Mergoil, Montagnac.
Freestone, I.C., 2015: The
recycling and reuse of Roman
glass: analytical approaches,
Journal of glass studies, 29–40.
Freestone, I.C., Y. Gorin-Rosen
& M.J. Hughes 2000: Primary
glass from Israel and the
production of glass in late
antiquity and the early islamic
period, in: N. Marie-Dominique
(ed.), La route du verre. Ateliers
primaires et secondaires du second
millénaire av. J.-C. au Moyen Âge,
Maison de l’Orient et de la
Méditerranée, Jean Pouilloux,
Lyon, 65–83.
Freestone, I.C. & M.J. Hughes
2006: The origins of Jarrow
glass, in: R. Cramp, Wearmouth
and Jarrow Monastic sites, Vol. 2,
English Heritage, 147–155.
Freestone, I.C., R.E. Jackson-Tal,
O. Tal & I. Taxel 2015: Glass
production at an early Islamic
workshop in Tel Aviv, Journal of
Archaeological Science 62, 45-54.
Freestone, I.C., P. Degryse,
J. Lankton, B. Gratuze &
J. Schneider 2018: HIMT, glass
composition and commodity
branding in the primary glass
industry, in: D. Rosenow,
M. Phelps, A. Meek &
I. Freestone (eds), Things that
Travelled: Mediterranean Glass in
the First Millennium AD, UCL
Press, 159–190. https://doi.
org/10.2307/j.ctt21c4tb3.14
Gai, S., 2005: Vitres et vitraux
du palais impérial de
Charlemagne à Paderborn, in:
D. Foy (ed.), De transparentes
spéculations. Vitres de l’Antiquité
et du haut Moyen Âge (OccidentOrient), Exposition temporaire
en liaison avec les 20e rencontres
de l’AFAV sur le thème du verre
plat, Bavay, Musée-site
d’archéologie, 83–85.
Gallo, F., A. Silvestri,
P. Degryse, M. Ganio,
A. Longinelli & G. Molin 2015:
Roman and late-Roman glass
from north-eastern Italy:
The isotopic perspective to
provenance its raw materials,
Journal of Archaeological Science,
62, 55–65.
Gam, T., 1990: Prehistoric Glass
Technology – Experiments and
Analyses, Journal of Danish
Archaeology, 9, 203–213.
Ganio, M., M. Gulmini,
K. Latruwe, F. Vanhaecke &
P. Degryse 2013: Sasanian glass
from Veh Ardašīr investigated
by strontium and neodymium
isotopic analysis, Journal of
archaeological science, 40(12),
4264–4270.
118
—
Gaut, B., 2011: Vessel Glass and
Evidence of Glassworking in
Skre, D. (ed.). Things from the
town: artefacts and inhabitants in
Viking-age Kaupang, Kaupang
Excavation Project publication
series volume 3, Aahus
University Press, 169-258.
Gaut, B. & J. Henderson 2011:
Compositional analysis.
Appendix 9.1 in Skre, D. (ed.).
Things from the town: artefacts
and inhabitants in Viking-age
Kaupang, Kaupang Excavation
Project publication series
volume 3, Aahus University
Press, 262-271.
Genga, N., M. Siciliano,
A. Tepore, A. Mangone,
A. Traini & C. Laganara 2008:
An archaeometric approach
about the study of medieval
glass from Siponto (Foggia,
Italy), Microchemical Journal, 90,
(1), 56-62.
Gliozzo, E., E. Braschi,
F. Giannetti, A. Langone &
M. Turchiano 2019: New
geochemical and isotopic
insights into the Late Antique
Apulian glass and the HIMT1
and HIMT2 glass productions—
the glass vessels from San
Giusto (Foggia, Italy) and the
diagrams for provenance
studies, Archaeological and
Anthropological Sciences, 11,
141–170.
Gorin-Rosen Y., 2000:
The ancient glass industry in
Israel—summary of the finds
and new discoveries, in:
M.-D. Nenna (ed.), La route du
verre Ateliers primaires et
secondaires du second millénaire
av. J.C. au Moyen Âge, Maison de
l’Orient Méditerranéen,
Jean Pouilloux, Lyon, 49–63.
Gratuze, B. & J.-N. Barrandon
1990: Islamic glass weights and
stamps, Archaeometry, 32,
155–162.
Gratuze, B., D. Foy, J. Lancelot
& F. Téreygeol 2003: Les
“lissoirs” carolingiens en verre
au plomb: mise en evidence de
la valorisation des scories
issues du traitement des
galènes argentifères de Melle
(Deux-Sèvres), in: D. Foy &
M.-D. Nenna (eds), Échanges et
commerce du verre dans le monde
antique, Actes du colloque de
l’AFAV, Aix-en-Provence et
Marseille, 7–9 June 2001,
Montagnac, France: Éditions
Monique Mergoil, 101–108.
Haevernick, T.E., 1979:
Karolingisches Glas aus
St Dionysius in Esslingen,
Forschungen und Berichte der
Archäologie des Mittelalters in
Baden-Württemberg, 6, 157–171.
Heaser, S., 2018: Anglo-Saxon
Glass Beadmakers, s.l. 19 pp.
Heck, M., T. Rehren &
P. Hoffmann 2003: The
production of lead–tin yellow
at Merovingian Schleitheim
(Switzerland), Archaeometry,
45(1), 33–44.
Hemminga, M. & T. Hamburg
2006: Een Merovingische
nederzetting op de oever van de
Oude Rijn. Opgraving (DO) en
Inventariserend Veldonderzoek (IVO)
Oegstgeest - Rijnfront zuid 2004,
Leiden (Archol Rapport 69).
Hemminga, M., T. Hamburg,
M. Dijkstra, C. Cavallo,
S. Knippenberg, S.M.E. van Lith,
C.C. Bakels & C. Vermeeren
2008: Vroegmiddeleeuwse
nederzettingssporen te Oegstgeest.
Een Inventariserend Veldonderzoek
en Opgraving langs de Oude Rijn,
Leiden (Archol Rapport 102).
Henderson, J., 1985: The raw
materials of early glass
production, Oxford Journal of
Archaeology, 4(3), 267–291.
Henderson, J., 1985: The raw
materials of early glass
production, Oxford Journal of
Archaeology, 4, 267-291.
Henderson, J., 1989: The
scientific analysis of ancient
glass and its archaeological
interpretation, in: J. Henderson
(ed.), Scientific analysis in
archaeology and its interpretation,
Oxford University Committee
on Archaeology Monograph
no. 19, UCLA Institute of
Archaeology Research Tools 5,
Oxford: Oxbow Books, 30–62.
Henderson, J., 1991a: Chemical
characterisation of Roman
glass vessels, enamels and
tesserae, in: P.B. Vandiver, J.
Druzik & G.S. Wheeler (eds),
Materials issues in art and
archaeology II, Research Society
Symposium Proceedings, vol.
185, Pittsburgh, PA: Materials
Research Society, 601–607.
Henderson, J., 1991b: The
glass, in: P. Armstrong, D.
Tomlinson & D.H. Evans (eds),
Excavations at Lurk Lane Beverley
1979–82, Sheffield Excavation
Reports 1, Sheffield, England:
J.R. Collis, 124–30.
Henderson, J., 1993: Aspects of
Early Medieval glass production
in Britain, Annales du 12e Congrès
de l’Association Internationale
pour l’Histoire du verre, Vienna
26–31 August 1991, Amsterdam:
AIHV, 247–259.
Henderson, J., 1999: Scientific
analysis of the glass and glassbearing artefacts: Technique,
raw materials used and
archaeological interpretation,
in: J.C. Besteman, J.M. Bos,
D.A. Gerrets, H.A. Heidinga &
J. de Koning, The Excavations at
Wijnaldum, reports on Frisia in
Roman and Medieval times,
volume 1, Rotterdam: Balkema,
287–297.
Henderson, J., 2000: The science
and archaeology of materials,
Routledge: London and
New York.
Henderson, J., 2002: Tradition
and experiment in 1st millennium
ad glass production – the
emergence of early Islamic
glass technology in late
antiquity, Accounts of Chemical
Research, 35, 594–602.
Henderson, J., 2012:
De Glasvondsten, in: Het domain
van de boer en de ambachtsman.
Een opgraving op het terrein van
de voormalige fruitveiling, te Wijk
bij Duurstede: een deel van
Dorestad en de villa Wijk
archeologisch onderzocht,
Amersfoort (ADC Monografie
12), 293–354.
Henderson, J., 2013: Ancient
glass, an interdisciplinary
exploration, New York and
Cambridge: Cambridge
University Press.
119
—
Henderson, J., 2022:
The politics of production,
glass provenance and social
context on the early Islamic
Silk Roads, Journal of Islamic
Archaeology 8.2, 203-237.
Henderson, J., 2023: Susteren
glass: scientific evidence for
raw material use, supply and
provenance, Stoepker, H. (ed.),
2023: Sporen van Susteren,
archeologische vondsten uit een
Karolingisch klooster en een
adellijk vrouwenstift, de
basispublicatie, Venlo,
1357-1382. https://doi.org/10.
17026/dans-xsf-nypw
Henderson, J., K. Challis,
S. O’Hara, S. McLoughlin,
S. Gardner & G. Priestnall
2005a: Experiment and
innovation: early Islamic
industry at al-Raqqa, Syria,
Antiquity, 79, 1–15.
Henderson, J., J.A. Evans,
H.J. Sloane, M.J. Leng, &
C. Doherty 2005b: The use of
oxygen, strontium and lead
isotopes to provenance ancient
glasses in the Middle East,
Journal of Archaeological
Science, 32, 665-673.
Henderson, J., S. Chenery,
E. Faber & J. Kröger 2016:
The use of Electron Probe
Microanalysis and Laser
Ablation-Inductively Coupled
Plasma-Mass Spectrometry for
the investigation of 8th–14th
century plant ash glasses from
the Middle East, Microchemical
Journal, 128, 134–152.
Henderson J., S. Chenery,
E.W. Faber & J. Kröger 2021:
Political and technological
changes, glass provenance and
a new glass production model
along the west Asian Silk Road,
in: Florian Klimscha (ed.), Berlin
Studies of the Ancient World, 67,
Berlin: edition Topoi.
Henderson, J. & I. Holand
1992: The Glass from Borg, an
early Medieval Chieftain’s Farm
in Northern Norway, Medieval
Archaeology, XXXVI, 39–58.
Henderson, J. & R. Ivens 1992:
Dunmisk and glass-making in
Early Christian Ireland,
Antiquity, 66, 52–64.
Henderson, J., J. Evans &
Y. Barkoudah 2009: The roots
of provenance: glass, plants
and isotopes in the Islamic
Middle East, Antiquity, 83,
414–429.
Henderson, J., H. Ma &
J. Evans 2020: Glass production
for the Silk Road? Provenance
and trade of Islamic glasses
using isotopic and chemical
analyses in a geological context,
Journal of Archaeological Science,
119, 105–164. doi.org/10.1016/j.
jas.2020:105164
Henderson, J. & Y. Sablerolles
2020: An overview of Dutch
Early Medieval glassworking,
published chemical and isotopic
analyses of glass beads and
vessels, raw, material
provenance of beads and
vessels, changes in raw material
use over time and a plan for
future scientific research.
Henderson, J., T. Sode &
Y. Sablerolles 2019: Early
medieval tesserae from
Scandinavia and the
Netherlands: a case for re-use
and recycling, in: L. Van Wersch,
L. Verslype, S. Strivay &
F. Theuws (eds), Early medieval
tesserae in Northwestern Europe,
Habelt, Bonn, 68–95.
Henton, A., 2020: Aux portes
de l’abbaye d’Elnone – SaintAmand. Premiers résultats de
la fouille préventive de la
Grand-Place de Saint-Amandles-Eaux (Nord, Hauts-deFrance), 43e Colloque Archaeologia
Mediaevalis, Mars 2020: Namur,
Belgique, 54–58. hal-03229653
Huisman, D.J., J. Van der Laan,
G.R.Davies, J.H.van Os,
N. Roymans, B.Fermin &
M. Karwowski 2017: Purple
haze: Combined geochemical
and Pb-Sr isotope constraints
on colourants in Celtic glass,
Journal of Archaeological
Science 81, 59-78
Huisman, H., M. Aarts,
M. Kars, F. Mulder, D. NganTillard & B. van Os, 2019,
Maken en handelen:
Merovingische kralen uit het
Sittard-Kemperkoul grafveld
geanalyseerd, Paleoaktueel 30:
65-74
Hulst, R.A., 1992: Opgravingen
Rijksarchief te Maastricht:
Merovingische afvalkuilen en
laat-middeleeuwse
verdedigingswerken,
Archeologie in Limburg, 42,
89–94.
Isings, C., 1957: Roman glass
from dated finds, Groningen:
Archaeologica Traiectina, 2.
Isings, C., 1978: Glas, Spiegel
Historiael, 13, 4 (Dorestadnummer), 260–262.
Isings, C., 2009: Glas, in:
M. Nokkert, A.C. Aarts &
H.L. Wynia (eds), Vroegmiddeleeuwse bewoning langs de
A2. Een nederzetting uit de zevende
en achtste eeuw in Leidsche Rijn,
Utrecht (Basisrapportage
Archeologie 26 Gemeente
Utrecht), 246–251.
Isings, C., 2015: Glass, in:
W.A. van Es & W.J.H. Verwers,
Excavations at Dorestad 4, The
settlement on the river Bank Area,
Amersfoort: Nederlandse
Oudheden 18, 422–446.
Isings, C., G. Rauws, H. Lägers
& R. de Kam 2009: Schitterend!
Twintig eeuwen glas uit Utrechtse
bodem, Utrecht.
Jackson, C.M., C.A. Booth &
J.W. Smedley 2005: Glass by
design? Raw materials, recipes
and compositional data,
Archaeometry, 47: 781–95.
Jankowiak, M., 2021: Dirham
flows into northern and eastern
Europe and the rhythms of the
slave trade with the Islamic
world, in: J. Gruszczyński,
M. Jankowiak & J. Shepard
(eds), Viking-Age Trade: Silver,
Slaves and Gotland, Routledge:
Abingdon and New York.
https://doi.org/10.4324
/9781315231808
Jezeer, W. & S. Jongma 2002:
Valkenburg-De Woerd. Werkput
510 – De Geul. Een studie naar
vroegmiddeleeuws aardewerk uit
Valkenburg (ZH), Amsterdam
(unpublished thesis, University
of Amsterdam).
120
—
Kamber, B.S., A. Greig &
K.D. Collerson 2005: A new
estimate for the composition
of weathered young upper
continental crust from alluvial
sediments, Queensland,
Australia, Geochimica et
Cosmochimica Acta, 69(4),
1041–1058.
Kaspers, A., 2020: WijnaldumTjitsma revisited. Testing the
potential value of fieldsurveying terp sites, in: A.
Nieuwhof (ed.), The excavations
at Wijnaldum, Volume 2:
Handmade and wheel-thrown
pottery of the first millennium AD,
Groningen: Groningen
Archaeological Studies 38,
University of Groningen &
Barkhuis Publishing, 193–240.
Kato, N., Y. Nakai & Y. Shindo
2009: Change in chemical
composition of early Islamic
glass excavated in Raya, Sinai
Peninsula, Egypt: on-site
analyses using a portable X-ray
fluorescence spectrometer,
Journal of Archaeological Science,
36, 1698–1707.
Keller, C., 2012:
Karolingerzeitliche
Keramikproduktion am
rheinischen Vorgebirge, in:
L. Grunwald, H. Pantermehl &
R. Schreg (eds), Hochmittelalterliche Keramik am Rhein. Eine
Quelle für Produktion und Alltag
des 9. bis 12. Jahrhunderts, Mainz:
RGZM-Tagungen 13, 209–224.
Kind, T., A. Kronz & K.H.
Wedepohl 2002: Karolingerzeitliches Glas und verschiedene
Handwerksindizien aus dem
Kloster Fulda. Aufarbeitung der
Altfunde Joseph Vonderaus
von 1898–99, in: Zeitschrift für
Archäologie des Mittelalters, 31,
2003, 2004: 61–93.
Koch, U., 1977: Das
Reihengräberfeld bei Schretzheim,
Berlin: Mann.
Koch, U., 1987: Der runde Berg
bei Urach VI, Die Glas- und
Edelsteinfunde aus den
Plangrabungen 1967–1983,
Heidelberger Akademie der
Wissenschaften, Komm.
Alamannische Altertumskunde
12, Sigmaringen.
Kouwatli, I., H.H. Curvers,
B. Stuart, Y. Sablerolles,
J. Henderson & P. Reynolds
2008: A pottery and glass
production site in Beirut (BEY
015), Bulletin d’Archéologie et
d’Architecture Libanaise, 10,
103–129.
Kronz, A., V. Hilberg, K. Simon
& K.H. Wedepohl 2016: Glas
aus Haithabu, Zeitschrift für
Archäologie des Mittelalters, 43,
39–58.
Krüger, I. & K.H. Wedepohl
2003: Composition and shapes
of glass of the early medieval
period (8th to 10th century AD) in
Central Europe, Échanges et
commerce du verre dans le monde
antique, 93–100.
Lahlil, S., I. Biron, M. Cotte,
J. Susini & N. Menguy 2010:
Synthesis of calcium antimonate
nano-crystals by the 18th dynasty
Egyptian glassmakers, Applied
Physics A, 98(1), 1–8.
Langbroek, M., 2021a: Beads
and beadmaking in the early
medieval settlement at
Oegstgeest, in: J. De Bruin,
C. Bakels & F. Theuws (eds),
Oegstgeest. A riverine settlement
in the early medieval world system,
Bonn, 2021, 278–293.
Langbroek, M.B., 2021b: Beads
from Dorestad, in: A. Willemsen
& H. Kik (eds), Dorestad and its
networks, Proceedings of the
third ‘Dorestad Congress’ held
at the National Museum of
Antiquities, Leiden, the
Netherlands 12–15 June 2019:
55–99.
Lassaunière, G., I. Pactat,
B. Gratuze & É. Louis 2016:
L’artisanat du verre au haut
Moyen Âge dans le nord de la
France (Nord et Pas-de-Calais),
Bulletin de l’Association Française
pour l’Archéologie du Verre,
76–82.
Later, C., 2010: Von Trinkgläsern
und Hängelampen, Gedanken
zur Funktion frühmittelalterlicher Sturzbecher, in:
G. Suhr & K. Hösch (eds),
Bajuwarenhof Kirchheim – Projekt
für lebendige Archäologie des
frühen Mittelalters, Jahresschrift
2009, München 2010, 65-76.
Louis, E., 2015: Les indices
d’artisanat dans et autour du
monastère de Hamage (Nord),
Bulletin du centre d’études
médiévales d’Auxerre, BUCEMA
[on-line], Hors-série n° 8, mis
en ligne le 28 janvier 2015:
consulté le 16 décembre 2021.
https://doi.org/10.4000/
cem.13684.
Lund Feveile, L., 2006:
Hulsglasskår fra markedspladsen
i Ribe, ASR 9, Posthuset, in:
C. Feveile (ed.), Ribe Studier.
DetÆldste Ribe. Udgravninger på
nordsiden af Ribe Å 1984-2000,
Bind 1.1 (Jysk Arkologisk Selskabs
skrifter 51). Moesgård, 195-278.
Lundström, P., 1981: De kommo
vida ... Vikingars hamn vid Paviken
på Gotland, Stockholm
(Sjöhistoriska museets
rapportserie 15).
Magendans, J.R. & J.A.
Waasdorp 1989: Franken aan de
Frankenslag. Een vroegmiddeleeuwse nederzetting in
’s-Gravenhage, ’s-Gravenhage
(VOM-reeks 1989-2).
Matin, M., 2019: Tin-based
opacifiers in archaeological
glass and ceramic glazes: a
review and new perspectives,
Archaeological and Anthropological
Sciences, 11(4), 1155–1167.
Matin, M., M. Tite &
O. Watson 2018: On the origins
of tin-opacified ceramic glazes:
New evidence from early
Islamic Egypt, the Levant,
Mesopotamia, Iran, and
Central Asia. Journal of
Archaeological Science, 97,
42–66.
Matthes, C., 1998: Die
Glasperlen des merowingerzeitlichen Gräberfeldes Griesheim,
Magisterarbeit Berlin.
Meharg. A. A., K. J.Edwards,
J. E. Scholfield, A. Raab,
J. Feldmann, A. Moran,
C. L. Bryant, B.Thornton &
J.J. C. Dawson 2012: First
comprehensive peat
depositional records for tin,
lead and copper associated
with the antiquity of Europe’s
largest cassiterite deposits,
Journal of Archaeological Science,
39, 717-727.
121
—
McCormick, M., 2001: Origins
of the European economy.
Communications and commerce,
AD 300–900, New York:
Cambridge University Press.
Meek, A., J. Henderson &
J. Evans 2012: The isotopic
analysis of English Forest glass
from the Weald and
Staffordshire, Journal of
Analytical Atomic Spectroscopy,
27, 786–795.
Miedema, M., 1983:
Vijfentwintig eeuwen bewoning in
het terpenland ten noordwesten
van Groningen, (PhD
dissertation, Vrije Universiteit,
Amsterdam).
Mirti, P., M. Pace, M.M. Negro
Ponzi & M. Aceto 2008:
ICP-MS analysis of glass
fragments of Parthian and
Sasanian epoch from Seleucia
and Veh Ardašīr (central Iraq),
Archaeometry, 50, 429–50.
Mirti, P., M. Pace, M.
Malandreno & M. Negro Ponzi
2009: Sasanian glass from Veh
Ardašīr: new evidences by
ICP-MS analysis, Journal of
Archaeological Science, 36,
1061–1069.
Nenna, M.-D., M. Vichy &
M. Picon 1997: L’atelier de
verrier de Lyon du Ier siècle
après J.-C., et l’origine des
verres “Romains”, Revue
d’archéométrie, 21 (1997), 81–87.
Nenna, M.-D., 2014: Egyptian
glass abroad. HIMT glass and
its markets, in: J. Bayley,
C. Jackson, D. Keller & J. Price
(eds) Neighbours and successors
of Rome. Traditions of glass
production in use in Europe and
the Middle East in the later first
millennium AD, Oxbow Books,
Oxford, 178–193.
Nenna, M.-D., 2015: Primary
glass workshops in GraecoRoman Egypt: preliminary
report on the excavations of
the site of Beni Salama, Wadi
Natrun (2003: 2005–9), in:
Justine Bayley, I. Freestone &
C. Jackson (eds), Glass of the
Roman world, Oxbow Books,
Oxford, 1–22.
Neri, E., B. Gratuze &
N. Schibille 2018: The trade of
glass beads in early medieval
Illyricum: towards an Islamic
monopoly, Archaeological and
Anthropological sciences 11,
1107-1122. https://doi.
org/10.1007/s12520-0170583-5.
Nicolay, J.A.W., 2014: The
splendour of power. Early medieval
kingship and the use of gold and
silver in the southern North Sea
area (5th to 7th century AD),
Groningen: Groningen
Archaeological Studies 28:
Barkhuis & University of
Groningen Library.
Nieuwhof, A., 2006: De wierde
Wierum (provincie Groningen). Een
Archeologisch Steilkantonderzoek,
Groningen: Groningen
Archaeological Studies 3.
Nieuwhof, A. (ed.) 2020: The
Excavations at Wijnaldum, Vol. 2:
Handmade and wheel-thrown
Pottery of the first Millennium AD,
Groningen: Groningen
Archaeological Studies 34.
Nokkert, M., A.C. Aarts &
H.L. Wynia 2009:
Vroegmiddeleeuwse bewoning
langs de A2. Een nederzetting uit
de zevende en achtste eeuw in
Leidsche Rijn, Utrecht
(Basisrapportage archeologie
26 Gemeente Utrecht).
Norde, E.H.L.D., 2019:
Nederzettingsresten uit de vroege
middeleeuwen in het plangebied
Leeuwesteyn Noord in Leidsche
Rijn, gemeente Utrecht,
Archeologisch onderzoek: een
opgraving, Weesp: RAAPRapport 3855.
Nyst, C.L., 2003: Karolingisch
Glas van Dorestad. Een
glasinventarisatie van vier
opgravingen , (unpublished
BA thesis, University of
Amsterdam).
Pactat, I., M. Guérit, L. Simon,
B. Gratuze, S. Raux & C. Aunay
2017: Evolution of glass recipes
during the Early Middle Ages in
France: analytical evidence of
multiple solutions adapted to
local contexts, in: S. Sophie &
A. Pury-Gysel (eds), Annales du
20e congrès de l’Association
Internationale pour l’Histoire
du Verre: Fribourg/Romont,
7–11 septembre 2015: Association
Internationale pour l’Histoire
du Verre, 334–340
Pactat, I., 2021: L’activité verrière
en France entre le viiie et le xie
siècle. Résilience et mutations
d’une production artisanale,
Thèse de doctorat en
archéologie, sous la direction
de Philippe Barral et de Danièle
Foy, soutenue le 3 septembre
2020 à l’Université de
Bourgogne Franche-Comté,
Laboratoire Chronoenvironnement/MSHE C.-N.
Ledoux. Bulletin du centre
d’études médiévales d’Auxerre|
BUCEMA, (25.1). https://doi.
org/10.4000/cem.18365
Paynter, S., 2008: Experiments
in the reconstruction of Roman
wood-fired glassworking
furnaces: waste products and
their formation processes,
Journal of Glass Studies, 50,
271–290.
Peake, J.R. & I.C. Freestone
2012: Cross-craft interactions
between metal and glass
working: slag additions to early
Anglo-Saxon red glass, in: W,
Meulebroeck, K. Nys, D.
Vanclooster & H. Thienpont
(eds) Integrated approaches to the
study of historical glass, IA1S 12,
Proceedings of SPIE vol. 8422,
1-12, International Society for
Optics and Photonics.
Peake, J.R. & I.C. Freestone
2014: Opaque yellow glass
production in the early
medieval period: new
evidence, Neighbours and
successors of Rome: traditions of
glass production and use in Europe
and the Middle East in the later 1st
millennium AD. Oxford: Oxbow
Books, 15–21.
122
—
Phelps, M., I.C. Freestone,
Y. Gorin-Rosen & B. Gratuze
2016: Natron glass production
and supply in the late antique
and early medieval Near East:
The effect of the ByzantineIslamic transition, Journal of
archaeological science, 75: 57–71.
Philippsen, B., C. Feveile,
J. Olsen, J. & S. M. Sindbæk
2021: Single-year radiocarbon
dating anchors Viking Age
trade cycles in time, Nature.
https://doi.org/10.1038/s41586021-04240-5
Pion, C., 2014: Les perles
mérovingiennes. TypoChronologie, fabrications et
fonctions, Unpublished PhD
Thesis, Université libre de
Bruxelles, Brussels.
Pol, A., 1999: Munten, in:
J. Plumier-Torfs, S. PlumierTorfs, M. Regnard &
W. Dijkman, Mosa nostra.
De Maasvallei van Verdun tot
Maastricht in de Merovingische
periode (5e-8e eeuw), Namur:
Carnet du Patrimoine 28, 11.
Preiß, F., 2010: Tesserae and
glass drops, in: A. Willemsen
& H. Kik (eds), Dorestad in an
international framework. New
research on centres of trade and
coinage in Carolingian times,
Turnhout: Proceedings of the
first ‘Dorestad Congress’ held
at the Museum of National
Antiquities Leiden, The
Netherlands, June 24–27,
123–134.
Raux, S., B. Gratuze,
J.-Y. Langlois & E. Coffineau
2015: Indices d’une production
verrière du Xe siècle à La
Milesse (Sarthe), Bulletin de
l’Association Française pour
l’Archéologie du Verre, Actes
des 29e Rencontres, Paris
(Nov. 2014), 66–70.
Redknap, M., 1988: Medieval
pottery production at Mayen:
recent advances, current
problems, in: D.R.M. Gaimster,
M. Redknap & H.-H. Wegner
(eds), Medieval and later pottery
from the Rhineland and its
markets, Oxford: BAR Int. Series
440, 3–37.
Risom, T., 2013: Perlemageren fra
Ribe, Historien, Materialerne,
Teknikkerne, Ribe.
Rooksby, H.P., 1964: Yellow
cubic lead-tin oxide opacifier in
ancient glasses, Physics and
Chemistry of Glasses, Section B
of J. Soc. Glass Technol., 5,
20–25.
Rossi, R., 2009: Il vetro grezzo
e le altre materie prime del reli
tto romano di Mljet (Meleda),
Croazia Quaderni Friulani di
Archeologia, 19, 193-202.
Sablerolles, Y., 1992: Het glas
van Gennep. De glasvondsten
van een nederzetting uit de
volksverhuizingstijk te Gennep
(Zuid-Limburg). Doctorate
thesis, University of
Amsterdam.
Sablerolles, Y., 1993: A
Dark-Age glass complex from a
Frankish settlement at Gennep
(Dutch Limburg), Proceedings of
the International Association for
the history of glass, Vienna 1991,
Amsterdam: AIHV, 197-206.
Sablerolles, Y., 1999: Beads of
glass, faience, amber, baked
clay and metal, including the
production waste from glass
and amber bead making, in:
J.C. Besteman, J.M. Bos,
D.A. Gerrets, H.A. Heidinga &
J. de Koning (eds), The excavations
at Wijnaldum. Reports on Frisia in
Roman and Medieval Times I,
Rotterdam: Brookfield,
253–287.
Sablerolles, Y., 2019: Een
smeltkroes voor glas uit de
vroege middeleeuwen, in:
E.H.L.D. Norde (ed.), Nederzettingsresten uit de vroege
middeleeuwen in het plangebied
Leeuwesteyn Noord in Leidsche
Rijn, deel 1, Gemeente Utrecht,
Archeologisch onderzoek: een
opgraving (RAAP Rapport 3855),
136–138.
Sablerolles, Y., 2023:
Glasproductie in het klooster
van Susteren, Stoepker, H.
(ed.), 2023: Sporen van Susteren,
archeologische vondsten uit een
Karolingisch klooster en een
adellijk vrouwenstift, de
basispublicatie, Venlo,
1249-1279. https://doi.
org/10.17026/dans-xsf-nypw
Sablerolles, Y., J. Henderson &
W. Dijkman 1997: Early
medieval glass bead making in
Maastricht (Jodenstraat 30),
the Netherlands. An archaeological and scientific
investigation, in: U. von
Freeden & A. Wieczorek (eds),
Perlen. Archäologie, Techniken,
Analysen, Bonn: Akten Internat.
Perlensymposium Mannheim
1994: Koll. Vor- und
Frühgeschichte 1, 295–313.
Sablerolles, Y. & J. Henderson
2012: De glasvondsten, in:
J. Dijkstra (ed.), Het domein van
de boer en de ambachtsman. Een
opgraving op het terrein van de
voormalige fruitveiling te Wijk bij
Duurstede: een deel van Dorestad
en de villa Wijk archeologisch
onderzocht, Amersfoort, (ADC
monografie 12), 293–350.
Sablerolles, Y. & J. de Koning
2015: Kralen in het zand: de
glasvondsten, in: J. de Koning,
Onder het stuifzand – Overstoven
vroegmiddeleeuwse nederzettingen
bij Bloemendaal. De opgravingscampagnes Groot Olmen 2005:
2006: en 2007: Inclusief een
uitgebreide re-interpretatie van de
opgraving Wijk aan Zee-Hoogovens
uit de jaren zestig van de vorige
eeuw, Zaandijk, 307–320.
Sanke, M., K.H. Wedepohl &
A. Kronz 2002: Karolingerzeitliches Glas aus dem Kloster
Lorsch, Zeitschrift für Archäologie
des Mittelalters, 30, 37–75.
Schibille, N. & I.C. Freestone
2013: Composition, Production
and Procurement of Glass at
San Vincenzo al Volturno: An
Early Medieval Monastic
Complex in Southern Italy,
PLoS One 8(10), e76479.
https://doi.org/10.1371/journal.
pone.0076479
Schibille, N., A. SterrettKrause & I.C. Freestone 2016.
Glass groups, glass supply and
recycling in Late Roman
Carthage, Archaeological and
Anthropological Sciences, 9,
1223–1241.
123
—
Schibille, N., A. Meek,
B. Tobias, C. Entwistle,
M. Avisseau-Broustet,
H. Da Mota & B. Gratuze 2016:
Comprehensive chemical
characterisation of Byzantine
glass weights. PLoS One 11(12),
p.e0168289.
Schibille, N., B. Gratuze,
E. Ollivier & É. Blondeau 2019:
Chronology of early Islamic
glass compositions from Egypt,
Journal of Archaeological Science,
104, 10–18.
Schibille, N., J. de J. Ares,
M.T.C. García & C. Guerrot
2020: Ex novo development of
lead glassmaking in early
Umayyad Spain, Proceedings
of the National Academy of
Sciences 117 (28) 16243-16249;
https://doi.org/10.1073/
pnas.2003440117
Schibille,N., V. Amorós Ruiz,
J. De Juan Ares & S. Gutiérrez
Lloret 2022: Rare alkali
elements as markers of local
glass working in medieval
Tolmo de Minateda (Spain),
ChemPlusChem 87, e202200147.
Shortland, A.J., 2004: Evaporites
of the Wadi Natrun: Seasonal
and annual variation and its
implication for ancient
exploitation, Archaeometry, 46,
4: 497–517.
Sier, M.M., J. van Doesburg &
W.J.H. Verwers (eds) 2004:
Wijk bij Duurstede-Frankenweg/
Zandweg, Amersfoort
(ADC-rapport 282).
Silvestri, A., G. Molin &
G. Salviulo 2008: The colourless
glass of Iulia Felix, Journal of
archaeological science, 35(2),
331–341.
Silvestri, A., S. Tonietto,
G. Molin & P. Guerriero 2012:
The palaeo-Christian glass
mosaic of St. Prosdocimus
(Padova, Italy): archaeometric
characterisation of tesserae
with antimony- or phosphorusbased opacifiers, Journal of
archaeological science, 39(7),
2177–2190.
Simpson, S.J., 2014: Sasanian
glass: an overview, in: D. Keller,
J. Price, & C. Jackson (eds),
Neighbours and Successors of
Rome. Traditions of Glass Production
and Use in Europe and the Middle
East in the Later 1st Millennium AD,
Oxford: Oxbow Books,
200–231.
Siu, I., J. Henderson, D. Qin,
Y. Ding, J. Cui & and H. Ma
2020: New light on plant ash
glass found in Africa: Evidence
for Indian Ocean Silk Road
trade using major, minor, trace
element and lead isotope
analysis of glass from the
15th—16th century AD from
Malindi and Mambrui,
Kenya, PLoS One, 15, p.
e0237612.
Smith, T., J. Henderson &
E.W. Faber 2016: Early
Byzantine glass supply and
consumption: the case of
Dichin, Bulgaria, in: Guxi Gan,
Julian Henderson & Qinghui Li
(eds), Recent research in the
scientific investigations of ancient
glass, World Scientific, 207–232.
Sode, T., 2004: Glass bead
making technology, in:
M. Bencard, A.K. Rasmussen
& H.B. Madsen (eds), Ribe
excavations 1970–76 (Vol. 5),
Moesgard: Jutland Archaeological Society, 83–102.
Standish, C.D., S.W. Merkel,
Y.-T. Hsieh, & J. Kershaw 2021:
Simultaneous lead isotope
ratio and gold-lead-bismuth
concentration analysis of silver
by laser ablation MC-ICP-MS,
Journal of Archaeological Science
125, 105299.
Steppuhn, P., 1998: Die
Glasfunde von Haithabu, Berichte
über die Ausgrabungen in
Haithabu 32, Neumünster.
Stoepker, H., 2021: Het klooster
van Susteren (714–1802). Archeologisch onderzoek van een
Karolingische abdij en een adellijk
vrouwenstift, Amersfoort
(Nederlandse Archeologische
Rapporten 73).
Szöke, B.M., K.H. Wedepohl &
A. Kronz 2004: Silver-stained
windows at Carolingian
Zalavar, Mosaburg
(southwestern Hungary),
Journal of Glass Studies 46,
85–104
Tal, O., E.E. Jackson-Tal & I.C.
Freestone 2004: New evidence
for glass production at Lat
Byzantine Apollonia-arsuf,
Israel, Journal of Glass Studies,
51, 51–65.
Thedéen, S., 2009: Who’s that
girl? The cultural construction
of girlhood and the transition
to womanhood in Viking Age
Gotland, Childhood in the Past
1(1), 78–93.
Van Dockum, S.G., 1997:
Parkeerplaats AH, in: Archeologische Kroniek Provincie Utrecht
1992–1993, Utrecht, 120–121.
Van Doesburg, J., 2004:
Glas, in: M.M. Sier, J. van
Doesburg & W.J.H. Verwers
(eds), Wijk bij DuurstedeFrankenweg/Zandweg, Amersfoort
(ADC-rapport 282), 42.
Van Es, W.A. & W.J.H. Verwers
1980: Excavations at Dorestad 1.
The harbour: Hoogstraat 1,
Amersfoort: Nederlandse
Oudheden 9.
Van Lith, S.M.E. & Y. Sablerolles
1995: Verres du IVe et du Ve
siècle des sites d’habitat aux
Pays-Bas, in: F. Foy (ed.) Le verre
de l’antiquité tardive et du haut
moyen âge, Association Française
pour l’Archéologie du Verre,
huitième rencontre, Guiry-enVexin, 18-19 Novembre 1993,
41-50.
Van Os, B.J.H., R.M. Vogelzang,
J.W. de Kort, D.J. Huisman,
M. Kars, D.J.M. Ngan-Tillard,
W. Verwaal & E. Meijvogel
2014: Kralen van glas en
barnsteen, in: R.C.G.M. Lauwerier
& J.W. de Kort (eds), Merovingers
in een villa 2, Romeinse villa en
Merovingisch grafveld Borgharen
– Pasestraat Onderzoek,
Amersfoort (Rapportage
Archeologische Monumentenzorg 222).
Van Wersch, L.F. 2013:
Merovingian ceramic and glass
in the early towns of the middle
Meuse valley, Medieval and
modern matters 4, 133-151.
124
—
Van Wersch, L., F. Mathis &
P. Hoffsummer 2009: Étude
des verres du Haut Moyen Âge
découverts dans l’église des
saints Hermès et Alexandre à
Theux, in: Vitrail, Verre et
Archéologie entre le Ve et le XIIe
siècle, Actes de la table ronde
réunie au Centre d’Études
Médiévales d’Auxerre, 15–16 juin
2006, Paris, 161–171.
Van Wersch, L., I. Biron,
B. Neuray, F. Mathis, G. Chêne,
D. Strivay & C. Sapin 2014:
Les vitraux alto-médiévaux de
Stavelot (Belgique), ArchéoSciences
38, 219–234.
Van Wersch, L., C. Loisel,
F. Mathis, D. Strivay & S. Bully
2015: Analyses of Early Medieval
Stained Window Glass from
the Monastery of Baume-lesMessieurs (Jura, France),
Archaeometry, 58, 930–946.
Van Wersch, L., B. Gratuze,
F. Mathis, M. Bonnin, D. Strivay,
H. Da Mota Rocha & C. Sapin
2018: Glass tiles from SaintSauveur (Burgundy, France),
Journal of Glass Studies, 60,
163-181.
Van Winkelhoff, A.M., 2021:
Light in the Dark Ages: a
conceptual approach to
the role of glass vessels in the
Merovingian burial rite, (RMAthesis Leiden University,
Leiden).
Wedepohl, K.H., 2008:
Mittelalterliches HolzascheGlas, in: H. Flachenecker,
G. Himmelsbach & P. Steppuhn
(eds), Glashüttenlandschaft
Europa, Regensburg: Schnell
and Steiner.
Verhulst, A., 2002: The
Carolingian Economy, Cambridge
Medieval Texbooks,
Cambridge: Cambridge
University Press.
Wedepohl, K. H. &
A. Baumann 1997: Isotope
composition of Medieval lead
glasses reflecting early silver
production in Central
Europe, Mineralium
Deposita 32, 292–295.
Vollgraff, C.W. & G. van Hoorn
1934: Opgravingen op het Domplein
te Utrecht, Wetenschappelijke
Verslagen II, De opgravingen in
juni en juli 1933: Haarlem.
Vrielynck, O., F. Mathis &
C. Pion 2018: Vers une typochronologie des perles
mérovingiennes en Gaule du Nord,
Namur.
Wedepohl, K.H., 2003: Flachund Gefäßglas der Karolingerzeit,
Nachrichten der Akademie der
Wissenschaften Göttingen, II.
Mathematische–Physikalische,
Klasse, 3, 115–156.
Wedepohl, K.H. & K. Simon
2010: The chemical composition
of medieval wood ash glass
from Central Europe,
Geochemistry, 70(1), 89–97.
Wedepohl, K.H.,
W. Winkelmann &
G. Hartmann 1997: Glasfunde
aus der karolingischen Pfalz in
Paderborn und die frühe
Holzasche-Glasherstellung,
Ausgrabungen und Funde in
Westfalen Lippe, 9A, 41–53.
Willmott, H. & K. Welham
2013: Late seventh-century
glassmaking at Glastonbury
Abbey, Journal of Glass Studies
55, 71–83.
Willmott, H. & K. Welham
2015: Saxon Glass Furnaces, in:
R. Gilchrist & C. Greene (eds)
Glastonbury Abbey: archaeological
investigations 1904–79, Society
of Antiquaries of London,
London, 218–238.
Wynia, H., 2013: Graven op het
Domplein, in: DompleinMagazine,
Special I: Het Domplein in Utrecht,
Utrecht, 9–15.
Yener, K.A., F. Kulakoğlu,
E. Yazgan, R. Kontani,
Y.S. Hayakawa, J.W. Lehner,
G. Dardeniz, G. Öztürk,
M. Johnson, R. Kaptan &
A. Hacar 2015: New tin mines
and production sites near
Kültepe in Turkey: a thirdmillennium BC highland
production model, Antiquity, 89,
596-612.
Appendices
Appendix I
sample list
Appendix II
major and minor chemical compositions of samples analysed by electron probe microanalysis
Appendix III
trace element chemical compositions of samples analysed by LA-ICP-MS
Appendix IV
photos of the samples from Maastricht and Utrecht
125
—
126
—
Appendix I
sample list
The list of the samples from the different sites used in this study: Maastricht (Jodenstraat and Mabro sites), Gennep, Wijnaldum,
Utrecht, Wijk bij Duurstede (Dorestad), Susteren and Deventer.
Site
Sample
Find number
Feature
number
Feature type
Begin date
feature (AD)
End date
feature (AD)
Category object
Form/object
Maastricht-Jodenstraat
1
1-1-7
1/7
pit
580/90
610/20
production waste
rod
Maastricht-Jodenstraat
2
1-1-7
1/7
pit
580/90
610/20
production waste
rod
Maastricht-Jodenstraat
3
1-1-7
1/7
pit
580/90
610/20
production waste
rod
Maastricht-Jodenstraat
4
01-01-2007
1/7
pit
580/90
610/20
production waste
rod
Maastricht-Jodenstraat
5
1-1-7
1/7
pit
580/90
610/20
production waste
thread
Maastricht-Jodenstraat
6
1-1-7
1/7
pit
580/90
610/20
production waste
rod
Maastricht-Mabro
7
3-4-0
0
pit
-
-
production waste
crucible
Maastricht-Mabro
8
3-OA-55
3/55
pit
-
-
production waste
crucible
Maastricht-Mabro
9
1-3-51
1/51
pit
-
-
production waste
crucible?
Maastricht-Mabro
10
3-OA-1
3/1
pit
-
-
production waste
crucible
Maastricht-Mabro
11
1-5-OA
1/OA
pit
-
-
production waste
crucible
Maastricht-Mabro
12
3-AA'-400
3/400
pit
-
-
production waste
crucible
Maastricht-Mabro
13
2-2-18
2/18
pit
-
-
production waste
crucible
Maastricht-Mabro
14
3-5-24
3/24
pit
-
-
production waste
crucible
Maastricht-Mabro
15
3-5-24
3/24
pit
-
-
production waste
crucible
Maastricht-Mabro
16
3-4-12
3/12
pit
-
-
production waste
crucible
Maastricht-Jodenstraat
17
1-1-7
1/7
pit
580/90
610/20
production waste
crucible
Maastricht-Jodenstraat
18
1-1-7
1/7
pit
580/90
610/20
production waste
crucible
Maastricht-Jodenstraat
19
1-1-7
1/7
pit
580/90
610/20
production waste
crucible
127
—
Key weathering
Key archaeological periods
x=slightly weathered xx = moderately weathered xxx= badly weathered
1= 450-550 AD 2= 550-650 AD 3= 650-750 AD 4= 750-850 AD 5= 850-1000 AD
Type of
fragment
Typology
(Isings, Koch,
Callmer, Ribe
or Pion)
Description object
Colour
Transparancy
Decoration
Weathering
Archeological
period
Publication
section
-
rod, red with weathered
surface
-
opaque
-
x
2
Sablerolles et al.
19997
section
-
rod, red with weathered
surface
-
opaque
-
x
2
Sablerolles et al.
19997
section
-
white twisted rod
-
opaque
-
-
2
Sablerolles et al.
19997
section
-
white rod
-
opaque
-
-
2
Sablerolles et al.
19997
section
-
white rod (thin, c 2-3 mm)
-
opaque
-
-
2
Sablerolles et al.
19997
section
-
yellow rod > 5 mm
-
opaque
-
-
2
Sablerolles et al.
19997
rim
-
crucible rim fragment with
white material (frit?) on both
sides
-
opaque
-
-
2
Sablerolles et al.
19997
body
-
crucible base fragment with
colourless glass on both
sides, white on the inside
-
opaque/
translucent
-
-
2
Sablerolles et al.
19997
?
-
small crucible fragment with
green and weathered
opaque yellow glass
-
translucent/
opaque
-
x
2
Sablerolles et al.
19997
rim
-
crucible rim with colourless
glass on inside
-
transparent
-
-
2
Sablerolles et al.
19997
rim
-
crucible rim fragment with
thick white material (frit?)
on both sides; colourless
glass on the inside
-
opaque/
transparent
-
-
2
Sablerolles et al.
19997
base
-
crucible base fragment with
opaque yellow glass on
inside. colourless
vitrification on the outside.
-
opaque/
transparent
-
-
2
Sablerolles et al.
19997
ceramic
-
small fragment of red
ceramic with (natural?)
green glass on both sides
-
translucent
-
-
2
Sablerolles et al.
19997
base
-
red ceramic pot base with
yellow (outside) and yellowcolourless (inside) glass
-
opaque/
transparent
-
-
2
Sablerolles et al.
19997
ceramic
-
small ceramic fragment with
deep translucent glass on
inside
-
transparent
-
-
2
Sablerolles et al.
19997
base
-
thick red ceramic base with
deep translucent green glass
on both sides, partially
weathered glass on both
sides
-
transparent
-
x
2
Sablerolles et al.
19997
base
-
crucible base of reddish-grey
ceramic with opaque yellow
glass on inside
-
opaque
-
-
2
Sablerolles et al.
19997
base
-
crucible base of grey
ceramic, with weathered
opaque yellow glass on
inside.
-
opaque
-
x
2
Sablerolles et al.
19997
base
-
crucible base of red ceramic,
weathered opaque yellow
glass on inside
-
opaque
-
x
2
Sablerolles et al.
19997
128
—
Appendix I sample list
Site
Sample
Find number
Feature
number
Feature type
Begin date
feature (AD)
End date
feature (AD)
Category object
Form/object
Maastricht-Jodenstraat
20
1-1-7
1/7
pit
580/90
610/20
production waste
crucible
Maastricht-Jodenstraat
21
1-1-7
1/7
pit
580/90
610/20
production waste
crucible
Maastricht-Jodenstraat
22
1-1-7
1/7
pit
580/90
610/20
production waste
crucible
Maastricht-Jodenstraat
23
1-1-7
1/7
pit
580/90
610/20
production waste
crucible
Maastricht-Jodenstraat
24
1-1-7
1/7
pit
580/90
610/20
production waste
crucible
Maastricht-Jodenstraat
25
1-1-7
1/7
pit
580/90
610/20
production waste
crucible
Maastricht-Jodenstraat
26
1-1-7
1/7
pit
580/90
610/20
production waste
crucible
Maastricht-Jodenstraat
27
1-1-7
1/7
pit
580/90
610/20
production waste
crucible
Maastricht-Jodenstraat
28
1-1-7
1/7
pit
580/90
610/20
production waste
crucible
Maastricht-Jodenstraat
29
1-2-3
1/3
pit
580/90
610/20
production waste
crucible
Maastricht-Jodenstraat
30
1-1-7
1/7
pit
580/90
610/20
production waste
brick/tegula?
Maastricht-Jodenstraat
37
1-1-7
1/7
pit
580/90
610/20
production waste
drop
Maastricht-Jodenstraat
38
1-1-7
1/7
pit
580/90
610/20
production waste
drop
Maastricht-Jodenstraat
39
1-1-7
1/7
pit
580/90
610/20
production waste
ingot?
Maastricht-Jodenstraat
40
1-1-7
1/7
pit
580/90
610/20
production waste
ingot?
Maastricht-Jodenstraat
41
1-1-7
1/7
pit
580/90
610/20
production waste
punty?
Maastricht-Jodenstraat
42
1-1-7
1/7
pit
580/90
610/20
production waste
punty?
Maastricht-Jodenstraat
43
1-1-7
1/7
pit
580/90
610/20
production waste
punty?
Maastricht-Jodenstraat
44
1-2-3
1/3
pit
580/90
610/20
window
flat quite thick
window glass;
one side
worked with
grozing (sp.)
iron?
Maastricht-Jodenstraat
45
1-2-3
1/3
pit
580/90
610/20
window
thin window
glass; two sides
been worked
with grozing
(sp.) iron?
129
—
Key weathering
Key archaeological periods
x=slightly weathered xx = moderately weathered xxx= badly weathered
1= 450-550 AD 2= 550-650 AD 3= 650-750 AD 4= 750-850 AD 5= 850-1000 AD
Type of
fragment
Typology
(Isings, Koch,
Callmer, Ribe
or Pion)
Description object
Colour
Transparancy
Decoration
Weathering
Archeological
period
Publication
base
-
crucible base of red ceramic,
layer of opaque yellow glass
on inside
-
opaque
-
-
2
Sablerolles et al.
19997
base
-
crucible base of thin grey
ceramic, colourless glass
-
translucent
-
-
2
Sablerolles et al.
19997
base
-
crucible base with opaque
yellow glass and white
material
-
opaque
-
-
2
Sablerolles et al.
19997
base
-
crucible base with opaque
yellow glass and white
material
-
opaque/
translucent
-
-
2
Sablerolles et al.
19997
base
-
crucible base of red ceramic,
opaque yellow and
brownish red vitrification
-
opaque
-
-
2
Sablerolles et al.
19997
base
-
crucible base of red ceramic,
opaque yellow and redishbrown vitrification
-
opaque
-
-
2
Sablerolles et al.
19997
base
-
crucible base of red ceramic,
opaque yellow glass
-
opaque
-
-
2
Sablerolles et al.
19997
base
-
crucible base of red ceramic,
opaque yellow glass
-
opaque
-
-
2
Sablerolles et al.
19997
base
-
crucible base of red ceramic,
opaque yellow glass
-
opaque
-
-
2
Sablerolles et al.
19997
base
-
crucible base of red ceramic,
deep translucent green
glass, esp. thick on bottom
-
translucent
-
-
2
Sablerolles et al.
19997
-
-
possible furnace brick
fragment with weathered
opaque yellow glass
-
opaque
-
x
2
Sablerolles et al.
19997
complete
-
blue glass drop, c. 1 cm
-
translucent
-
-
2
Sablerolles et al.
19997
complete
-
blue glass drop, c. 1 cm
-
translucent
-
-
2
Sablerolles et al.
19997
crushed
-
small (crushed?) blue glass
fragments
-
translucent
-
-
2
Sablerolles et al.
19997
crushed
-
small (crushed?) blue glass
fragments
-
translucent
-
-
2
Sablerolles et al.
19997
undiagnostic
-
fragment of opaque red
glass
-
opaque
-
-
2
Sablerolles et al.
19997
undiagnostic
-
fragment of opaque red
glass
-
opaque
-
-
2
Sablerolles et al.
19997
undiagnostic
-
fragment of blue-green
glass
-
translucent
-
x
2
Sablerolles et al.
19997
-
-
flat quite thick translucent
yellowish window glass; one
side worked with grozing
(sp.) iron?
-
translucent
-
-
2
Sablerolles et al.
19997
-
-
thin amber window glass;
two sides been worked with
grozing iron
-
translucent
-
-
2
Sablerolles et al.
19997
130
—
Appendix I sample list
Site
Sample
Find number
Feature
number
Feature type
Begin date
feature (AD)
End date
feature (AD)
Category object
Form/object
Maastricht-Jodenstraat
46
1-2-3
1/3
pit
580/90
610/20
window
moderately thin
window glass.
One side
rounded.
Maastricht-Jodenstraat
47
1-1-7
1/7
pit
580/90
610/20
production waste
rod
Maastricht-Jodenstraat
48
1-1-7
1/7
pit
580/90
610/20
production waste
drop
Maastricht-Jodenstraat
49
1-1-7
1/7
pit
580/90
610/20
production waste
drop
Maastricht-Jodenstraat
50
1-1-7
1/7
pit
580/90
610/20
production waste
drop
Maastricht-Jodenstraat
51
1-1-7
1/7
pit
580/90
610/20
production waste
drop
Maastricht-Jodenstraat
52
1-1-7
1/7
pit
580/90
610/20
production waste
drop
Maastricht-Jodenstraat
53
1-1-7
1/7
pit
580/90
610/20
production waste
drop
Maastricht-Jodenstraat
54
1-1-7
1/7
pit
580/90
610/20
production waste
rod
Maastricht-Jodenstraat
55
1-1-7
1/7
pit
580/90
610/20
production waste
drop
Maastricht-Jodenstraat
56
1-1-7
1/7
pit
580/90
610/20
production waste
rod
Maastricht-Jodenstraat
57
1-1-3
1/3
pit
580/90
610/20
production waste
rod
Maastricht-Jodenstraat
58
?
-
pit
580/90
610/20
production waste
rod
Maastricht-Jodenstraat
59
?
-
pit
580/90
610/20
production waste
rod
Maastricht-Jodenstraat
60
1-2-5
1/5
pit
580/90
610/20
vessel
cone
Maastricht-Jodenstraat
61
1-2-5
1/5
pit
580/90
610/20
bead
bead
Maastricht-Jodenstraat
62
1-1-7
1/7
pit
580/90
610/20
production waste
drop
Maastricht-Jodenstraat
63
1-1-7
1/7
pit
580/90
610/20
production waste
drop
Maastricht-Jodenstraat
64
1-1-7
1/7
pit
580/90
610/20
production waste
?
Maastricht-Jodenstraat
65
1-1-7
1/7
pit
580/90
610/20
production waste
?
Maastricht-Jodenstraat
66
1-1-7
1/7
pit
580/90
610/20
production waste
rod
Maastricht-Jodenstraat
67
1-1-7
1/7
pit
580/90
610/20
production waste
rod
Maastricht-Jodenstraat
68
1-1-7
1/7
pit
580/90
610/20
vessel
ribbed bowl
Maastricht-Jodenstraat
69
1-1-7
1/7
pit
580/90
610/20
production waste
punty
131
—
Key weathering
Key archaeological periods
x=slightly weathered xx = moderately weathered xxx= badly weathered
1= 450-550 AD 2= 550-650 AD 3= 650-750 AD 4= 750-850 AD 5= 850-1000 AD
Type of
fragment
Typology
(Isings, Koch,
Callmer, Ribe
or Pion)
Description object
Colour
Transparancy
Decoration
Weathering
Archeological
period
Publication
-
-
moderately thin pale green
window glass. one side
rounded.
-
translucent
-
-
2
Sablerolles et al.
19997
section
-
thin opaque olive green rod
-
opaque
-
-
2
Sablerolles et al.
19997
-
-
opaque green drop
-
opaque
-
-
2
Sablerolles et al.
19997
-
-
irregular green drop with
soil fused to it
-
translucent
-
-
2
Sablerolles et al.
19997
-
-
drop of weathered opaque
yellow glass
-
opaque
-
-
2
Sablerolles et al.
19997
-
-
irregular drop of reddish
glass
-
opaque
-
-
2
Sablerolles et al.
19997
-
-
irregular drop of deep
translucent (‘black’) glass
-
translucent
-
x
2
Sablerolles et al.
19997
-
-
stretched pale opaque blue
piece of rod?
-
opaque
-
-
2
Sablerolles et al.
19997
section
-
thin opaque red rod with
weathered exterior
-
opaque
-
x
2
Sablerolles et al.
19997
-
-
irregular red drop,
weathered surface
-
opaque
-
-
2
Sablerolles et al.
19997
section
-
double rod; yellow with
slightly greenish tint
-
opaque
-
-
2
Sablerolles et al.
19997
section
-
blue-green rod
-
opaque
-
-
2
Sablerolles et al.
19997
section
-
twisted opaque white rod
-
opaque
-
-
2
Sablerolles et al.
19997
section
-
white rod
-
opaque
-
-
2
Sablerolles et al.
19997
base
-
base of translucent green
cone
-
translucent
-
-
2
Sablerolles et al.
19997
half
-
fragmented tapering cobalt
blue bead
-
opaque
-
-
2
Sablerolles et al.
19997
drop
-
drop; naturally coloured
yellowish
-
translucent
-
-
2
Sablerolles et al.
19997
drop
-
irregular drop; naturally
coloured yellowish
-
translucent
-
-
2
Sablerolles et al.
19997
undiagnostic
-
thin opaque yellow
fragment
-
opaque
-
-
2
Sablerolles et al.
19997
undiagnostic
-
thin opaque yellow
fragment
-
opaque
-
-
2
Sablerolles et al.
19997
section
-
rod, red with weathered
surface
-
opaque
-
x
2
Sablerolles et al.
19997
section
-
rod, red with weathered
surface
-
opaque
-
x
2
Sablerolles et al.
19997
ribbed bowl
-
quite thick blue green glass
fragment with a rib (roman?)
-
translucent
-
-
2
Sablerolles et al.
19997
-
-
“punty glass” yellow thinwalled fragment
-
opaque
-
-
2
Sablerolles et al.
19997
132
—
Appendix I sample list
Site
Sample
Find number
Feature
number
Feature type
Begin date
feature (AD)
End date
feature (AD)
Category object
Form/object
Maastricht-Jodenstraat
70
1-1-7
1/7
pit
580/90
610/20
production waste
punty
Maastricht-Jodenstraat
71
1-1-7
1/7
pit
580/90
610/20
production waste
punty
Maastricht-Jodenstraat
72
1-1-7
01-jul
pit
580/90
610/20
production waste
punty
Maastricht-Jodenstraat
73
?
?
pit
580/90
610/20
production waste
rod
Maastricht-Jodenstraat
74
?
?
pit
580/90
610/20
production waste
rod
Maastricht-Jodenstraat
75
?
?
pit
580/90
610/20
production waste
rod
Maastricht-Jodenstraat
76
?
?
pit
580/90
610/20
production waste
rod
Gennep
GE 41
4481
35/2
sunken hut
400
550
vessel
cone
Gennep
GE 42
4060
28/8
sunken hut
400
550
vessel
cone
Gennep
GE 43
1749
13/34
sunken hut
400
550
vessel
cone
Gennep
GE 44
2115
20/37
sunken hut
400
550
vessel
cone
Gennep
GE 45
3079
19/26
sunken hut
400
550
vessel
cone
Gennep
GE 46
2527
6/10
sunken hut
400
550
vessel
cone
Gennep
GE 47
2557
7/120
sunken hut
400
550
vessel
cone
Gennep
GE 48
1313
8/54
sunken hut
400
550
vessel
cone
Gennep
GE 49
2278
6/1
sunken hut
400
550
vessel
cone
Gennep
GE 50
1399
8/35
sunken hut
400
550
vessel
cone
Gennep
GE 51a
2795
10/38
well
400
550
vessel
cone
Gennep
GE 51b
1397
8/35
sunken hut
400
550
vessel
cone
Gennep
GE 52
-
-
-
400
550
vessel
cone
Gennep
GE 53
2598
10/1
sunken hut
400
550
vessel
bowl
Gennep
GE 54
4254
27/13
sunken hut
400
550
vessel
bowl
133
—
Key weathering
Key archaeological periods
x=slightly weathered xx = moderately weathered xxx= badly weathered
1= 450-550 AD 2= 550-650 AD 3= 650-750 AD 4= 750-850 AD 5= 850-1000 AD
Type of
fragment
Typology
(Isings, Koch,
Callmer, Ribe
or Pion)
Description object
Colour
Transparancy
Decoration
Weathering
Archeological
period
Publication
-
-
“punty glass” yellow thinwalled fragment
-
opaque
-
-
2
Sablerolles et al.
19997
-
-
“punty glass” yellow thinwalled fragment
-
opaque
-
-
2
Sablerolles et al.
19997
-
-
“punty glass” yellow thinwalled fragment
-
opaque
-
-
2
Sablerolles et al.
19997
section
-
turquoise green rod
fragment
-
opaque
-
-
2
Sablerolles et al.
19997
section
-
turquoise green rod
fragment
-
opaque
-
-
2
Sablerolles et al.
19997
section
-
green rod fragment
-
opaque
-
-
2
Sablerolles et al.
19997
section
-
green rod fragment
-
opaque
-
-
2
Sablerolles et al.
19997
base
Koch 1987 III?
-
yellow-green
translucent
-
-
1
Sablerolles 1992,
cat. 121.1
base
Koch 1987 III I
-
olive brown
-
-
-
1
Sablerolles 1992,
cat. 85.1
base
Koch 1987 III?
-
olive green
-
-
-
1
Sablerolles 1992,
cat. 134.2?
rim
Koch 1987 III H
-
olive green
-
selfcoloured
spiral
-
1
Sablerolles 1992,
cat. 81.6
rim
Koch 1987 III?
-
light bluegreen
-
selfcoloured
spiral
-
1
Sablerolles 1992,
cat. 127.1
rim
Koch 1987 III I
-
yellow-green
-
selfcoloured
spiral
-
1
Sablerolles 1992,
cat. 86.3
rim
Koch 1987 III?
-
pale yellow/
colourless
-
selfcoloured
spiral
-
1
Sablerolles 1992,
cat. 124.1
rim
Koch 1987 III?
-
pale green/
colourless
-
white spiral
-
1
Sablerolles 1992,
cat. 123.1
rim
Koch 1987 III
I?
-
pale green/
colourless
-
selfcoloured
spiral
-
1
Sablerolles 1992,
cat. 155.2
rim
Isings 1957
106b2?
-
yellow-green
-
brown
spiral
-
Roman
Sablerolles1992,
cat. 2.3
rim
Isings 1957
106c2?
-
yellow-green
-
brown
spiral
-
Roman
Sablerolles 1992,
cat. 5.1
body
Isings 1957
106c2?
-
yellow-green
-
brown
arcades
-
Roman
Sablerolles 1992,
cat. 5.2
rim
Koch 1987 III?
-
light yellowgreen
-
selfcoloured
spiral
-
1
Sablerolles 1992,
cat. -
rim
Koch 1987 IV K
-
light yellowgreen/
colourless
-
-
-
1
Sablerolles 1992,
cat. 141.1
rim
Koch 1987 IV
M
-
light greenyellow
-
selfcoloured
spiral
-
1
Sablerolles 1992,
cat. 17
134
—
Appendix I sample list
Site
Sample
Find number
Feature
number
Feature type
Begin date
feature (AD)
End date
feature (AD)
Category object
Form/object
Gennep
GE 55
-
-
-
400
550
vessel
bowl
Gennep
GE 56
1915
16/38
sunken hut
400
550
vessel
bowl
Gennep
GE 57
-
-
-
400
550
vessel
bowl
Gennep
GE 58
-
-
-
400
550
vessel
bowl
Gennep
GE 59
-
-
-
400
550
vessel
bowl
Gennep
GE 60
4549
35/2
sunken hut
400
550
vessel
bowl
Gennep
GE 61
4549
35/2
sunken hut
400
550
vessel
bowl
Gennep
GE 62
4464
35/2
sunken hut
400
550
vessel
bowl
Gennep
GE 63
4563
35/2
sunken hut
400
550
vessel
bowl
Gennep
GE 64
4549
35/2
sunken hut
400
550
vessel
bowl
Gennep
GE 65
4901
41/3
post hole
sunken hut
400
550
vessel
bowl
Gennep
GE 66
4257
32/20
pit
400
550
vessel
bowl
Gennep
GE 67
1316
8/54
sunken hut
400
550
vessel
bowl
Gennep
GE 68
1512
8/8
sunken hut
400
550
vessel
bowl
Gennep
GE 69
1500
8/8
sunken hut
400
550
vessel
bottle
Wijnaldum
WIJ 1
9782
2351
occupation
surface
450
500
vessel
cone
(Kempston)
Wijnaldum
WIJ 2
6794
1346
waste
deposit?
450
500
vessel
bowl
Wijnaldum
WIJ 3
10802
2565
sunken hut/
waste deposit
600
700
vessel
bowl
Wijnaldum
WIJ 4
6242
801
ditch
450
500
bead
small glob
Wijnaldum
WIJ 5
1356
625
area with
metal waste/
waste deposit
575
625
bead
small glob
Wijnaldum
WIJ 6
1428
608
well 9
575
625
bead
small glob
135
—
Key weathering
Key archaeological periods
x=slightly weathered xx = moderately weathered xxx= badly weathered
1= 450-550 AD 2= 550-650 AD 3= 650-750 AD 4= 750-850 AD 5= 850-1000 AD
Type of
fragment
Typology
(Isings, Koch,
Callmer, Ribe
or Pion)
Description object
Colour
Transparancy
Decoration
Weathering
Archeological
period
Publication
rim
Koch 1987 IV
M
-
pale green
-
selfcoloured
spiral
-
1
Sablerolles 1992,
cat. -
rim
Koch 1987 IV
M
-
pale bluegreen
-
selfcoloured
spiral
-
1
Sablerolles 1992,
cat. 172.1
rim
Koch 1987 IV
M
-
pale green/
colourless
-
selfcoloured
spiral
-
1
Sablerolles 1992,
cat. 173.2
body
Koch 1987 IV
M
-
pale green
-
selfcoloured
spiral
-
1
Sablerolles1992,
cat. -
base
Koch 1987 IV
M
-
light yellowgreen
-
selfcoloured
spiral
-
1
Sablerolles 1992,
cat. -
rim
Koch 1987 IV
B?
-
light yellow/
colourless
-
-
-
1
Sablerolles 1992,
cat. 136.2
base
Koch 1987 IV
B?
-
light yellow/
colourless
-
-
-
1
Sablerolles 1992,
cat. 136.3
body
Koch 1987 IV K
-
light yellowgreen/
colourless
-
white
festoons
-
1
Sablerolles 1992,
cat. 147.1
rim
Koch 1987 IV L
-
light greenblue/
colourless
-
white
feather
-
1
Sablerolles 1992,
cat. 151.5
rim
Koch 1987 IV L
-
light greenblue
-
white
feather
-
1
Sablerolles 1992,
cat. 151.5
rim
Koch 1987 IV L
-
light greenyellow/
colourless
-
white
feather
-
1
Sablerolles 1992,
cat. 154.1
body
Koch 1987 IV L
-
light bluegreen
-
white
feather
-
1
Sablerolles 1992,
cat. 150.1
base
Koch 1987 IV ?
-
light bluegreen
-
-
-
1
Sablerolles 1992,
cat. 162
body
Koch 1987 IV ?
-
light bluegreen
-
-
-
1
Sablerolles 1992,
cat. 161
body
Isings 1957 101
-
yellow-green
-
red streaks
-
Roman
Sablerolles 1992,
cat. 1.3
body
Koch 1987 III N
-
pale bluish
green
transparent
selfcoloured
loops
x
1
Sablerolles 1999
VESSEL cat. 7
rim
Koch 1987 IV L
-
pale bluish
green
transparent
white
feather
x
1
Sablerolles 1999
VESSEL fig. 1.9
rim
Koch 1987 IV
-
light bluish
green
transparent
white spiral
x
1
Sablerolles 1999
VESSEL cat. 10
complete
Pion 2014
B1.1-2b
-
yellow
opaque
-
-
1
Sablerolles 1999,
BEAD fig. 5.13
-
Pion 2014
B1.1-2a
-
yellow
opaque
-
-
2
Sablerolles 1999
BEAD cat. 14
-
Pion 2014
B1.1-2b
-
yellow
opaque
-
-
2
Sablerolles 1999
BEAD cat. 15
136
—
Appendix I sample list
Site
Sample
Find number
Feature
number
Feature type
Begin date
feature (AD)
End date
feature (AD)
Category object
Form/object
Wijnaldum
WIJ 7
1462
1167
sod layer
600
700
bead
small glob
Wijnaldum
WIJ 8
3901
574
ditch
575
625
bead
small glob
Wijnaldum
WIJ 9
3901
574
ditch
575
625
bead
small glob
Wijnaldum
WIJ 10
3695
2605
truncated
layer/waste
deposit
800
850
vessel
funnel
Wijnaldum
WIJ 11
7359
1636
truncated
layer/waste
deposit
800
900
vessel
funnel
Wijnaldum
WIJ 12
7507
3296
truncated
layer/
occupation
surface
775
850
vessel
funnel
Wijnaldum
WIJ 13
7877
3358
truncated
layer/
ooccupation
surface
800
850
vessel
funnel
Wijnaldum
WIJ 14
6704
1233
ditch
550
600
bead
small glob
Wijnaldum
WIJ 15
10906
340
occupation
surface
770
900
vessel
funnel
Wijnaldum
WIJ 16
1526
1114
sod layer
700
750
vessel
jar?
Wijnaldum
WIJ 17
5812
2098
occupation
surface/waste
deposit
450
550
bead
small glob
Wijnaldum
WIJ 18
6712
1330
occupation
surface
450
550
bead
small annular
Wijnaldum
WIJ 19
5534
1233
ditch
550
600
bead
small glob
Wijnaldum
WIJ 20
7448
1233
ditch
550
600
bead
small glob
Wijnaldum
WIJ 21
2655
625
met/wd
575
625
bead
small glob
Wijnaldum
WIJ 22
5632
975
cultivation
layer
550
560
bead
small glob
Wijnaldum
WIJ 23
6704
6704
ditch
550
600
bead
small glob
Wijnaldum
WIJ 24
1461
548
occupation
surface/waste
desposit
550
600
bead
short cylindrical
Wijnaldum
WIJ 25
5666
1232
ditch
550
650
bead
short cylindrical
Wijnaldum
WIJ 26
7448
1233
ditch
550
600
bead
short cylindrical
Wijnaldum
WIJ 27
2454
558
cultivation
layer
500
550
bead
short cylindrical
137
—
Key weathering
Key archaeological periods
x=slightly weathered xx = moderately weathered xxx= badly weathered
1= 450-550 AD 2= 550-650 AD 3= 650-750 AD 4= 750-850 AD 5= 850-1000 AD
Type of
fragment
Typology
(Isings, Koch,
Callmer, Ribe
or Pion)
Description object
Colour
Transparancy
Decoration
Weathering
Archeological
period
Publication
-
Pion 2014
B1.1-2a
-
yellow
opaque
-
-
2 or 3
Sablerolles 1999
BEAD cat. 16
-
Pion 2014
B1.1-2a
-
yellow
opaque
-
-
2
Sablerolles 1999
BEAD cat. 18
-
Pion 2014
B1.1-2a
-
yellow
opaque
-
-
2
Sablerolles 1999
BEAD cat. 19
rim
Lund Feveil
2006, rim type
e
-
light (blue-)
green
translucent
-
x
4
Sablerolles 1999
VESSEL fig. 1.20
rim
Lund Feveil
2006, rim type
e
-
almost
colourless
transparent
-
x
4
Sablerolles 1999
VESSEL fig. 1.21
rim
Lund Feveil
2006, rim type
d
-
yellowish
green
translucent
-
x
4
Sablerolles 1999
VESSEL fig. 1.22
rim
Lund Feveil
2006, rim type
e
-
almost
colourless
transparent
-
x
4
Sablerolles1999
BEAD fig. 1.23
-
Pion 2014
B1.1-02b
-
yellow
opaque
-
-
2
Sablerolles 1999
BEAD cat. 25
rim
Lund Feveil
2006, rim type
g
-
dark blue
translucent
incalmo rim
x
4
Sablerolles 1999
VESSEL fig. 1.26
rim
-
-
blue-green
translucent
yellow
spiral, white
arcade
-
3/4
Sablerolles 1999
VESSEL fig. 1.30
-
Pion 2014
B1.1-3b
-
red
opaque
-
-
2
Sablerolles 1999
BEAD cat. 46
complete
Pion 2014
B1.1-3b
-
red
opaque
-
-
2
Sablerolles 1999
BEAD fig. 5.47
-
Pion 2014
B1.1-3b
-
red
opaque
-
-
2
Sablerolles 1999
BEAD cat. 49
-
Pion 2014
B1.1-3b
-
red
opaque
-
-
2
Sablerolles1999
BEAD cat. 53
-
Pion 2014
B.1.1-4a
-
white
opaque
-
-
2
Sablerolles 1999
BEAD cat. 56
-
Pion 2014
B.1.1-4a
-
white
opaque
-
-
2
Sablerolles 1999
BEAD cat. 57
complete
Pion 2014
B1.1-4a
-
white
opaque
-
-
2
Sablerolles 1999,
fig. 5.58
complete
Pion 2014
B1.4-1a
-
yellow
opaque
-
-
2
Sablerolles 1999,
fig. 5.66
-
Pion 2014
B1.4-3a
-
white
opaque
-
-
2
Sablerolles 1999
BEAD cat. 73
-
Pion 2014
B1.4-3a
-
white
opaque
-
-
2
Sablerolles 1999
BEAD cat. 78
-
Pion 2014
B1.4-2a
-
red
opaque
-
-
2
Sablerolles 1999
BEAD cat. 80
138
—
Appendix I sample list
Site
Sample
Find number
Feature
number
Feature type
Begin date
feature (AD)
End date
feature (AD)
Category object
Form/object
Wijnaldum
WIJ 28
10824
1384
truncated
layer/waste
deposit
650
750
bead
short cylindrical
Wijnaldum
WIJ 29
11090
2546
ditch
650
750
bead
short cylindrical
Wijnaldum
WIJ 30
6884
575
well 8
770
850
bead
short cylindrical
Wijnaldum
WIJ 31
7448
1233
ditch
550
600
bead
short cylindrical
Wijnaldum
WIJ 32
1024
1079
sod layer
640
750
bead
biglobular
Wijnaldum
WIJ 33
6704(2)
1233
ditch
550
600
bead
biglobular
Wijnaldum
WIJ 34
6562(1)
2064
sod layer
500
550
bead
irregular spiral
Wijnaldum
WIJ 35
3316(2)
3532
pit?
775
850
bead
segmented,
‘gold’ foil
Wijnaldum
WIJ 36
6562(1)
2064
sod layer
500
550
bead
segmented,
gold foil
Wijnaldum
WIJ 37
3326(1)
3542
ditch/
occupation
surface?
875
900
bead
segmented,
silver foil
Wijnaldum
WIJ 38
9737(1)
2341
occupation
surface/sod
layer
450
500
bead
segmented,
silver foil
Wijnaldum
WIJ 39
10608(1)
100
well 6
750
850
bead
segmented,
layered
Wijnaldum
WIJ 40
10786
514
ditch
750
770
production waste
tessera
Wijnaldum
WIJ 41
3829
696
occupation
suface
425
500
production waste
rod
Wijnaldum
WIJ 42
4601
2817
truncated
layer/
occupation
surface
750
800
production waste
punty glass
Wijnaldum
WIJ 43
2943
625
area with
metal waste
575
625
production waste
furnace?
Utrecht-Domplein
31
1933-77-36
?
?
-
-
production waste
crucible
Utrecht-Domplein
32
1933-77-53
?
?
-
-
production waste
crucible
Utrecht-Domplein
33
1933-zn3
?
?
-
-
production waste
crucible
Utrecht-Domplein
34
1933-234
?
?
-
-
production waste
crucible
139
—
Key weathering
Key archaeological periods
x=slightly weathered xx = moderately weathered xxx= badly weathered
1= 450-550 AD 2= 550-650 AD 3= 650-750 AD 4= 750-850 AD 5= 850-1000 AD
Type of
fragment
Typology
(Isings, Koch,
Callmer, Ribe
or Pion)
Description object
Colour
Transparancy
Decoration
Weathering
Archeological
period
Publication
-
Pion 2014
B1.4-2a
-
red
opaque
-
-
3
Sablerolles 1999
BEAD cat. 81
-
Pion 2014
B1.4-2a
-
red
opaque
-
-
3
Sablerolles 1999
BEAD cat. 82
-
Callmer 1977
A135?
-
red
opaque
-
-
4
Sablerolles 1999
BEAD cat. 83
-
Pion 2014
B1.4-2a
-
red
opaque
-
-
2
Sablerolles 1999
BEAD cat. 84
-
Pion 2014
B1.2-1b
-
yellow
opaque
-
-
3
Sablerolles 1999
BEAD cat. 97
-
Pion 2014B1.21b
-
yellow
opaque
-
-
2
Sablerolles 1999
BEAD cat. 99
-
Pion 2014
B1.8-01
-
black/dark
blue
opaque
-
-
1
Sablerolles 1999
BEAD cat. 102
complete?
Callmer 1977
E140?
-
yellowish?
transparent
-
-
4
Sablerolles 1999
BEAD, fig. 5.115
-
Pion 2014
A4.1-1
-
colourless
transparent
-
-
1
Sablerolles 1999
BEAD cat. 116
complete
Callmer 1977
E140?
-
colourless
transparent
-
-
4
Sablerolles 1999
BEAD, fig. 5.118
-
Pion 2014
A4.2-1
-
colourless
transparent
-
-
1
Sablerolles1999
BEAD cat. 119
complete
Pion 2014
A3.1-7
-
red on
colourless
opaque
-
-
4
Sablerolles1999
BEAD, fig. 5.122
-
-
-
yellow
opaque
-
-
Roman
Sablerolles 1999,
fig. 4, cat. 216
-
-
-
greenish
white
opaque
-
-
1
Sablerolles 1999,
fig. 4, cat. 217
-
-
-
turquoise
opaque
-
-
4
Sablerolles 1999,
fig. 4, cat. 218
-
-
-
yellow
opaque
-
x
2
Sablerolles 1999,
fig. 3, cat. 219
body?
-
crucible fragment (one of
two), thin-walled grey
ceramic with a thin layer of
green glass
-
translucent
-
-
4
-
body?
-
crucible fragment, thinwalled beige ceramic with a
thin (cracked) layer of green
glass.
-
translucent
-
-
4
-
base
-
crucible base fragment, thick
grey with red glass, overlain
by a think layer of cracked
green glass.
-
opaque/
translucent
-
-
4
Vollgraff & van
Hoorn 1934
rim
-
crucible rim fragment, thick
grey - pink fabric with a thick
layer of striped red and
green glass. weathered
areas and a white glassy
material under the rim.
-
opaque/
translucent
-
x
4
Vollgraff & van
Hoorn 1934
140
—
Appendix I sample list
Site
Sample
Find number
Feature
number
Feature type
Begin date
feature (AD)
End date
feature (AD)
Category object
Form/object
Utrecht-Domplein
35
1933-zn2
?
?
-
-
production waste
crucible
Utrecht-Domplein
36
1933-77-84
?
?
-
-
production waste
crucible
UtrechtOudwijkerdwarsstraat
77
6-1-170
170
pit
-
-
production waste
undiagnostic
UtrechtOudwijkerdwarsstraat
78
5-1-135
135
pit
-
-
production waste
drop
UtrechtOudwijkerdwarsstraat
79
5-1-135
135
pit
-
-
vessel
sherd
Wijk bij Duurstede
(Dorestad)
LM 16
-
-
-
-
-
vessel
lamp base
Wijk bij Duurstede
(Dorestad)
LM 17
-
-
-
-
-
vessel
funnel beaker
Wijk bij Duurstede
(Dorestad)
LM18
-
-
-
-
-
vessel
funnel beaker
Wijk bij Duurstede
(Dorestad)
LM 19
-
-
-
-
-
vessel
funnel beaker
Wijk bij Duurstede
(Dorestad)
LM 20
-
-
-
-
-
vessel
funnel beaker
Wijk bij Duurstede
(Dorestad)
LM 21
-
-
-
-
-
vessel
funnel beaker
Wijk bij Duurstede
(Dorestad)
LM 22
-
-
-
-
-
vessel
funnel beaker
Wijk bij Duurstede
(Dorestad)
LM 23
-
-
-
-
-
vessel
funnel beaker
Wijk bij Duurstede
(Dorestad)
LM 24
-
-
-
-
-
vessel
bowl
Wijk bij Duurstede
(Dorestad)
LM 25
-
-
-
-
-
vessel
possible
unguentarium
Wijk bij Duurstede
(Dorestad)
LM 26
-
-
-
-
-
vessel
possible
unguentarium/
bowl
Wijk bij Duurstede
(Dorestad)
LM 27
-
-
-
-
-
vessel
funnel beaker
Wijk bij Duurstede
(Dorestad)
LM 28
-
-
-
-
-
vessel
bell beaker
Wijk bij Duurstede
(Dorestad)
LM 29
-
-
-
-
-
vessel
bell beaker
Wijk bij Duurstede
(Dorestad)
LM 30
-
-
-
-
-
vessel
rim?beaker
Wijk bij Duurstede
(Dorestad)
LM 31
-
-
-
-
-
vessel
rim?beaker
Wijk bij Duurstede
(Dorestad)
LM 32
-
-
-
-
-
vessel
jar
Wijk bij Duurstede
(Dorestad)
LM 33
-
-
-
-
-
vessel
trail decorated
rim
141
—
Key weathering
Key archaeological periods
x=slightly weathered xx = moderately weathered xxx= badly weathered
1= 450-550 AD 2= 550-650 AD 3= 650-750 AD 4= 750-850 AD 5= 850-1000 AD
Type of
fragment
Typology
(Isings, Koch,
Callmer, Ribe
or Pion)
Description object
Colour
Transparancy
Decoration
Weathering
Archeological
period
Publication
rim
-
crucible rim, grey fabric with
green glass and weathered
surface.
-
translucent
-
x
4
Vollgraff & van
Hoorn 1934
rim
-
crucible rim fragment (1 of
4), thin pitted pink-grey
ceramic, green and red glass
attached.
-
translucent/
opaque
-
-
4
-
undiagnostic
-
fragments (crushed?)
-
translucent
-
-
3
-
-
-
irregular drop, pale green
modern
-
translucent
-
-
3
-
sherd
-
small green sherd
-
translucent
-
-
3
-
base
-
-
green
translucent
-
-
4
-
body
-
-
green
translucent
-
-
4
-
body
-
-
green
translucent
-
-
4
-
base
-
-
green
translucent
-
-
4
-
body
-
-
green
translucent
-
-
4
-
body
-
-
green
translucent
-
-
4
-
body
-
-
green
translucent
-
-
4
-
body
-
-
olive green
translucent
-
-
4
-
body
Isings 1957,
type 24?
-
green
translucent
-
-
Roman
-
body
Isings 1957,
type 10
-
green
translucent
-
-
Roman
-
body
Isings 1957,
type 10 or 20
-
pale blue
translucent
-
-
Roman
-
body
-
-
green
translucent
-
-
4
-
body
-
-
green
translucent
-
-
3
-
body
-
-
green
translucent
-
-
3
-
rim
-
-
green
translucent
-
-
4?
-
rim
-
-
turquoise
translucent
-
-
4?
-
body
-
-
green
translucent
-
-
3
-
rim
-
-
turquoise
translucent
-
-
4
-
142
—
Appendix I sample list
Site
Sample
Find number
Feature
number
Feature type
Begin date
feature (AD)
End date
feature (AD)
Category object
Form/object
Wijk bij Duurstede
(Dorestad)
DOR 53
-
-
-
-
-
vessel
gold glass
decorated
?beaker
Wijk bij Duurstede
(Dorestad)
DOR 61
-
-
-
-
-
vessel
red trailed
beaker
Wijk bij Duurstede
(Dorestad)
DOR 66
-
-
-
-
-
vessel
blue rimmed
beaker
Wijk bij Duurstede
(Dorestad)
DOR 90
-
-
-
-
-
vessel
blue rimmed
beaker
Wijk bij Duurstede
(Dorestad)
DOR 91
-
-
-
-
-
vessel
sub-sample:
body of 90
Wijk bij Duurstede
(Dorestad)
DOR 96
-
-
-
-
-
vessel
beaker
Wijk bij Duurstede
(Dorestad)
DOR 97
-
-
-
-
-
vessel
funnel beaker
Wijk bij Duurstede
(Dorestad)
DOR 98
-
-
-
-
-
vessel
funnel beaker
Wijk bij Duurstede
(Dorestad)
DOR 100
-
-
-
-
-
vessel
jar
Wijk bij Duurstede
(Dorestad)
DOR 101
-
-
-
-
-
vessel
palm cup
Wijk bij Duurstede
(Dorestad)
DOR 102
-
-
-
-
-
vessel
palm funnel
series
Wijk bij Duurstede
(Dorestad)
DOR 103
-
-
-
-
-
vessel
palm funnel
series
Wijk bij Duurstede
(Dorestad)
DOR 104
-
-
-
-
-
vessel
palm cup
Wijk bij Duurstede
(Dorestad)
DOR 105
-
-
-
-
-
vessel
base
Wijk bij Duurstede
(Dorestad)
DOR 106
-
-
-
-
-
vessel
palm cup or
funnel
Wijk bij Duurstede
(Dorestad)
DOR 107
-
-
-
-
-
vessel
palm cup
Wijk bij Duurstede
(Dorestad)
DOR 108
-
-
-
-
-
vessel
palm cup
Wijk bij Duurstede
(Dorestad)
DOR 109
-
-
-
-
-
vessel
palm cup
Wijk bij Duurstede
(Dorestad)
DOR 110
-
-
-
-
-
vessel
palm funnel
Wijk bij Duurstede
(Dorestad)
DOR 111
-
-
-
-
-
vessel
palm funnel
Wijk bij Duurstede
(Dorestad)
DOR 112
-
-
-
-
-
vessel
palm funnel
Wijk bij Duurstede
(Dorestad)
DOR 113
-
-
-
-
-
vessel
gold foil
decoarted palm
funnel
Wijk bij Duurstede
(Dorestad)
DOR 115
-
-
-
-
-
vessel
funnel beaker
base
Wijk bij Duurstede
(Dorestad)
DOR 116
-
-
-
-
-
vessel
funnel beaker
143
—
Key weathering
Key archaeological periods
x=slightly weathered xx = moderately weathered xxx= badly weathered
1= 450-550 AD 2= 550-650 AD 3= 650-750 AD 4= 750-850 AD 5= 850-1000 AD
Type of
fragment
Typology
(Isings, Koch,
Callmer, Ribe
or Pion)
Description object
Colour
Transparancy
Decoration
Weathering
Archeological
period
Publication
body
-
-
pale green
translucent
-
-
4
-
body
-
-
pale green or
colourless
translucent
-
-
4
-
rim
-
-
pale green
translucent
-
-
4
-
rim
-
-
blue
translucent
-
-
4
-
body
-
-
pale green
translucent
-
-
4
-
body
-
-
pale green
translucent
-
-
4?
-
body
-
-
pale green
translucent
-
-
4
-
body
-
-
pale green
translucent
-
-
4
-
body
-
-
pale green
translucent
-
-
3
-
body
-
-
pale green
translucent
-
-
3
-
body
-
-
mid green
translucent
-
-
4
-
body
-
-
yellow-green
translucent
-
-
4
-
body
-
-
mid green
translucent
-
-
3
-
base
-
-
red and
colourless
opaque and
transparent
-
-
4
-
body
-
-
pale green
translucent
-
-
3
-
body
-
-
pale green
translucent
-
-
3
-
body
-
-
pale green
translucent
-
-
3
-
body
-
-
pale green
translucent
-
-
3
-
body
-
-
pale green
translucent
-
-
3
-
body
-
-
pale green
translucent
-
-
3
-
body
-
-
pale green
translucent
-
-
3
-
body
-
-
pale green
translucent
-
-
4?
-
base
-
-
yellow-green
iridescent
translucent
-
-
4
-
body
-
-
mid green
translucent
-
-
4
-
144
—
Appendix I sample list
Site
Sample
Find number
Feature
number
Feature type
Begin date
feature (AD)
End date
feature (AD)
Category object
Form/object
Wijk bij Duurstede
(Dorestad)
DOR 117
-
-
-
-
-
vessel
funnel beaker
Wijk bij Duurstede
(Dorestad)
DOR 118
-
-
-
-
-
vessel
funnel beaker
Wijk bij Duurstede
(Dorestad)
DOR 119
-
-
-
-
-
vessel
funnel beaker
Wijk bij Duurstede
(Dorestad)
DOR 120
-
-
-
-
-
vessel
funnel beaker
Wijk bij Duurstede
(Dorestad)
DOR 121
-
-
-
-
-
vessel
funnel beaker
Wijk bij Duurstede
(Dorestad)
DOR 122
-
-
-
-
-
vessel
funnel beaker
Wijk bij Duurstede
(Dorestad)
DOR 123
-
-
-
-
-
vessel
funnel beaker
with bulge
Wijk bij Duurstede
(Dorestad)
DOR 124
-
-
-
-
-
vessel
funnel beaker
with bulge
Wijk bij Duurstede
(Dorestad)
DOR 125
-
-
-
-
-
vessel
funnel beaker
Wijk bij Duurstede
(Dorestad)
DOR 126
-
-
-
-
-
vessel
funnel beaker
Wijk bij Duurstede
(Dorestad)
DOR 127
-
-
-
-
-
vessel
funnel beaker
Wijk bij Duurstede
(Dorestad)
DOR 128
-
-
-
-
-
vessel
funnel beaker
Wijk bij Duurstede
(Dorestad)
DOR 129a
-
-
-
-
-
vessel
vessel with
applied blue
thread
Wijk bij Duurstede
(Dorestad)
DOR 129b
-
-
-
-
-
vessel
sub-sample
thread
decorating 129a
Wijk bij Duurstede
(Dorestad)
DOR 130a
-
-
-
-
-
vessel
funnel beaker
Wijk bij Duurstede
(Dorestad)
DOR 130b
-
-
-
-
-
vessel
vessel?
Wijk bij Duurstede
(Dorestad)
DOR 131
-
-
-
-
-
vessel
funnel beaker
Wijk bij Duurstede
(Dorestad)
DOR 132
-
-
-
-
-
vessel
funnel beaker
Wijk bij Duurstede
(Dorestad)
DOR 133
-
-
-
-
-
vessel
?funnel beaker
Wijk bij Duurstede
(Dorestad)
DOR 134
-
-
-
-
-
vessel
vessel?
Wijk bij Duurstede
(Dorestad)
DOR 135
-
-
-
-
-
vessel
beaker trail
below tim
Wijk bij Duurstede
(Dorestad)
DOR 136
-
-
-
-
-
vessel
funnel beaker
base
Wijk bij Duurstede
(Dorestad)
DOR 137
-
-
-
-
-
vessel
funnel beaker
Wijk bij Duurstede
(Dorestad)
DOR 138
-
-
-
-
-
vessel
funnel beaker
145
—
Key weathering
Key archaeological periods
x=slightly weathered xx = moderately weathered xxx= badly weathered
1= 450-550 AD 2= 550-650 AD 3= 650-750 AD 4= 750-850 AD 5= 850-1000 AD
Type of
fragment
Typology
(Isings, Koch,
Callmer, Ribe
or Pion)
Description object
Colour
Transparancy
Decoration
Weathering
Archeological
period
Publication
body
-
-
acqua
translucent
-
-
4
-
body
-
-
pale green
translucent
-
-
4
-
body
-
-
pale green
translucent
-
-
4
-
body
-
-
pale green
translucent
-
-
4
-
body
-
-
pale green
translucent
-
-
4
-
body
-
-
pale green
translucent
-
-
4
-
body
-
-
pale green
translucent
-
-
4
-
body
-
-
pale green
translucent
-
-
4
-
body
-
-
colourless
transparent
-
-
4
-
body
-
-
pale green
translucent
-
-
4
-
body
-
-
colourless
transparent
-
-
4
-
body
-
-
colourless
transparent
-
-
4
-
rim
-
-
pale green
translucent
-
-
4
-
rim
-
-
cobalt blue
translucent
-
-
4
-
body
-
-
blue
translucent
-
-
4
-
body
-
-
cobalt blue
translucent
-
-
4?
-
body
-
-
pale green
translucent
-
-
4
-
body
-
-
aqua mid
green
translucent
-
-
4
-
body
-
-
pale green
translucent
-
-
4
-
body
-
-
mid green
translucent
-
-
4?
-
rim
-
-
pale green
translucent
-
-
4
-
base
-
-
yellow-green
translucent
-
-
4
-
body
-
-
pale green
translucent
-
-
4
-
body
-
-
mid green
translucent
-
-
4
-
146
—
Appendix I sample list
Site
Sample
Find number
Feature
number
Feature type
Begin date
feature (AD)
End date
feature (AD)
Category object
Form/object
Wijk bij Duurstede
(Dorestad)
DOR 139
-
-
-
-
-
vessel
bowl
Wijk bij Duurstede
(Dorestad)
DOR 140
-
-
-
-
-
vessel
funnel beaker
Wijk bij Duurstede
(Dorestad)
DOR 141
-
-
-
-
-
vessel
funnel beaker
applied cable
Wijk bij Duurstede
(Dorestad)
DOR 142
-
-
-
-
-
vessel
funnel beaker
trail decoarted
Wijk bij Duurstede
(Dorestad)
DOR 143
-
-
-
-
-
raw chip
raw chip
Wijk bij Duurstede
(Dorestad)
DOR 144
-
-
-
-
-
tessera
tessera
Wijk bij Duurstede
(Dorestad)
DOR 145
-
-
-
-
-
tessera
tessera
Wijk bij Duurstede
(Dorestad)
DOR 146
-
-
-
-
-
tessera
tessera
Wijk bij Duurstede
(Dorestad)
DOR 147
-
-
-
-
-
tessera
tessera
Wijk bij Duurstede
(Dorestad)
DOR 148
-
-
-
-
-
tessera
tessera
Wijk bij Duurstede
(Dorestad)
DOR 149
-
-
-
-
-
rod
rod
Wijk bij Duurstede
(Dorestad)
DOR 150
-
-
-
-
-
linen smoother
linen smoother
Wijk bij Duurstede
(Dorestad)
DOR 151
-
-
-
-
-
linen smoother
linen smoother
Susteren
SUST 1
V12-053-GL-09
S12/067
water course
4310
800
1200
bead
annular
Susteren
SUST 2
V07-216-GL-01
S07/148
cistern
600
900
bead
biconical
Susteren
SUST 3
V09-205-GL-01
S09/200
posthole
700
1000
bead
conical
Susteren
SUST 4
V01-304-GL-01
S01/212
water course
4200
1000
1300
bead
conical
Susteren
SUST 5
V08-190-GL-17
S08/171
water course
4302
700
1000
bead
conical
Susteren
SUST 6
V04-245GL-04
S04/244
water course
4250
1000
1350
bead
cylindrical
Susteren
SUST 7
V09-129-GL-01
S09/100
grave 58
900
1100
window
irregular
Susteren
SUST 8
V04-194-GL-01
S04/199
water course
4400
700
1200
window
triangle?
Susteren
SUST 9
V08-190-GL-10
S08/171
water course
4302
700
1000
window
trapezium
Susteren
SUST 10
V09-273-GL-01
S09/179
grave 67
800
900
window
leaf?
147
—
Key weathering
Key archaeological periods
x=slightly weathered xx = moderately weathered xxx= badly weathered
1= 450-550 AD 2= 550-650 AD 3= 650-750 AD 4= 750-850 AD 5= 850-1000 AD
Type of
fragment
Typology
(Isings, Koch,
Callmer, Ribe
or Pion)
Description object
Colour
Transparancy
Decoration
Weathering
Archeological
period
Publication
body
-
-
mid green
translucent
-
-
3 or 4
-
body
-
-
yellow green
translucent
-
-
4
-
rim
-
-
pale green
translucent
-
-
4
-
rim
-
-
pale green
translucent
-
-
4
-
chip
-
-
cobalt blue
translucent
-
-
4?
-
whole
-
-
cobalt blue
translucent
-
-
4?
-
whole
-
-
cobalt blue
translucent
-
-
4?
-
whole
-
-
turquoise
opaque
-
-
4?
-
whole
-
-
turquoise
opaque
-
-
4?
-
whole
-
-
opaque mid
green
opaque
-
-
4?
-
incomplete
-
-
opaque
yellow
opaque
-
-
4?
-
incomplete
-
-
dark green
translucent
-
-
4
-
incomplete
-
-
dark green
translucent
-
-
4
-
half
Koch 1977
Group O
-
black
translucent
3 white
zigzags, 4
blue trails
-
1
-
complete
Callmer 1977
B546?
-
bluish green
translucent
3 yellow
zigzags, 4
red trails
-
4?
-
complete
?
-
blue-green
translucent
yellow
feather, 2
red trails
x
4?
-
half
?
-
bluish green
translucent
white
feather, 2
yellow
bands
x
4?
-
half
?
-
greenish
translucent
yellow
festoons
x
4?
-
fragment
?
-
greenish
translucent
white and
orange
festoons,
yellow
bands
x
4?
-
complete?
-
-
dark blue
translucent
-
x
4?
-
complete?
-
-
dark blue
translucent
-
-
4?
-
almost
complete
-
-
dark green
translucent
-
x
4?
-
almost
complete
-
-
colourless
transparent
-
-
4?
-
148
—
Appendix I sample list
Site
Sample
Find number
Feature
number
Feature type
Begin date
feature (AD)
End date
feature (AD)
Category object
Form/object
Susteren
SUST 11
V04-133-GL-01
S04/162
water course
4400
700
1200
window
rectangle?
Susteren
SUST 12
V08-138-GL-01
S08/155
water course
4302
700
1000
window
undiagnostic
Susteren
SUST 13
V06-158-GL-01
S06/150
water course
4200
700
1300
window
semi-circle
Susteren
SUST 14
V08-214-GL-01
S08/236
water course
4301
700
1000
window
undiagnostic
Susteren
SUST 15
V08-190-GL-12
S08/171
water course
4302
700
1000
window
undiagnostic
Susteren
SUST 16
V09-190-GL-11
S08/171
water course
4302
700
1000
window
undiagnostic
Susteren
SUST 17
V12-053-GL-10
S12/067
water course
4310
800
1200
production waste
crucible
Susteren
SUST 18
V12-053-GL-11
S12/067
water course
4310
800
1200
production waste
crucible
Susteren
SUST 19
V08-190GL-02
S08/171
water course
4302
700
1000
vessel
funnel
Susteren
SUST 20
V08-190-GL-07
S08/171
water course
4302
700
1000
vessel
bowl
Susteren
SUST 21
V08-190-GL-03
S08/171
water course
4302
700
1000
vessel
funnel
Susteren
SUST 22
V05-194-GL-01
S05/219
water course
4400
700
1000
vessel
(palm)funnel
Susteren
SUST 23
V04-232-GL-01
S04/199
water course
4400
700
1200
vessel
funnel
Susteren
SUST 24
V04-166-GL-01
S04/171
water course
4302
700
1000
vessel
bowl
Susteren
SUST 25
V12-053-GL-01
S12/067
water course
4310
800
1200
vessel
funnel
Susteren
SUST 26
V08-219-GL-01
S08/218
pit
800
900/1000
vessel
funnel
Susteren
SUST 27
V12-053-GL-02
S12/053
water course
4310
800
1200
vessel
(palm)funnel
Susteren
SUST 28
V07-148-GL-01
S07/023
pit
900
1000
vessel
funnel
Susteren
SUST 29
V07-148-GL-02
S07/023
pit
900
1000
vessel
funnel
Deventer
DEV 1
434/16203
15050
cesspit 278
850
900
production waste
slag, hollow
Deventer
DEV 2
434/10479
11011
floor level 24
850
900
bead
globular
Deventer
DEV 3
434/99302
10239
cesspit 228
850
900
vessel
funnel
Deventer
DEV 4
434/99721
12507
cesspit 142
850
900
vessel
undiagnostic
Deventer
DEV 5
434/21098
12507
cesspit 142
850
900
vessel
undiagnostic
Deventer
DEV 6
434/99025
290
waste pit 56
850
900
vessel
funnel
149
—
Key weathering
Key archaeological periods
x=slightly weathered xx = moderately weathered xxx= badly weathered
1= 450-550 AD 2= 550-650 AD 3= 650-750 AD 4= 750-850 AD 5= 850-1000 AD
Type of
fragment
Typology
(Isings, Koch,
Callmer, Ribe
or Pion)
Description object
Colour
Transparancy
Decoration
Weathering
Archeological
period
Publication
almost
complete
-
-
bluish green
translucent
-
xx
4?
-
fragment
-
-
bluish green
translucent
-
-
4?
-
complete
-
-
bluish green
translucent
-
x
4?
-
fragment
-
-
bluish green
translucent
-
x
4?
-
fragment
-
-
dark bluegreen
translucent
-
-
4?
-
fragment
-
-
dark bluegreen
translucent
-
-
4?
-
fragment
-
-
dark blue
translucent
-
xx
4?
-
fragment
-
-
light (bluish)
green
translucent
-
-
4?
-
base
-
-
bluish green
translucent
-
xx
4
-
body
Lund Feveile
2006, rim
type a
-
almost
colourless
transparent
yellow
reticella,
spiral
x
4
-
rim
Lund Feveile
2006, rim
type e
-
bluish green
transparent
-
-
4
-
base
-
-
bluish green
translucent
-
-
3/4
-
base
-
-
light bluegreen
translucent
-
-
4
-
body
-
-
dark blue
translucent
yellow
spiral
-
4
-
rim
Lund Feveile
2006, rim
type d
-
yellow-green
translucent
-
-
4
-
rim
Lund Feveile
2006, rim
type g
-
almost
colourless
transparent
blue
incalmo rim
-
4
-
base
-
-
blue-green
transparent
-
-
3/4
-
body
-
-
bluish green
transparent
white
reticella
-
4
-
rim
Lund Feveile
2006, rim
type e
-
light bluish
green
translucent
-
-
4
-
complete
-
-
grey-green
opaque
-
-
5
-
complete
-
-
dark blue
translucent
-
-
5
-
body
-
-
bluish green
translucent
-
-
5
-
body
-
-
colourless
-
x
5
-
body
-
-
colourless
transparent
-
x
5
-
body
-
-
bluish green
translucent
optic blown
ribs
-
5
-
150
—
Appendix I sample list
Site
Sample
Find number
Feature
number
Feature type
Begin date
feature (AD)
End date
feature (AD)
Category object
Form/object
Deventer
DEV 7
434/99302
10239
cesspit 228
850
900
vessel
funnel?
Deventer
DEV 8
312/29024
21911
wastepit 60
900
925
bead
globular
Deventer
DEV 9
312/29064
22458
cesspit 74
900
925
undiagnostic
undiagnostic
Deventer
DEV 10
312/29089
22638
cesspit 74
900
925
vessel
funnel?
Deventer
DEV 11
312/29089
22638
cesspit 74
900
925
vessel
funnel?
Deventer
DEV 12
312/29090
22638
cesspit 74
900
925
vessel
funnel?
Deventer
DEV 13
434/12394
12765
cesspit 235
900
900
vessel
funnel/conical
beaker?
Deventer
DEV 14
312/29028
21911
wastepit 60
900
925
vessel
undiagnostic
Deventer
DEV 15
312/29028
21911
wastepit 60
900
925
production waste
raw glass
Deventer
DEV 16
312/29090
22638
cesspit 74
900
925
window
window
Deventer
DEV 17
312/29057
22457
cesspit 74
900
925
production waste
raw glass
Deventer
DEV 18
312/29028
21911
wastepit 60
900
925
production waste?
undiagnostic
(trail?)
Deventer
DEV 19
434/10380
10879
layer 19
900
950
vessel
funnel
Deventer
DEV 20
434/99116
2301
wastepit 107
900
950
vessel
cup
Deventer
DEV 21
434/99144
2538
wastepit 174
900
950
production waste
raw glass
Deventer
DEV 22
434/99289
6623
house 9
900
950
vessel
beaker?
Deventer
DEV 23
434/99289
6623
house 9
900
950
vessel
undiagnostic
Deventer
DEV 24
434/99598
7160
cesspit 130
900
950
window
window
Deventer
DEV 25
434/99578
7120
cesspit 128
925
950
vessel
funnel/beaker?
Deventer
DEV 26
434/99139
2535
cesspit 172
900
950
window
window?
Deventer
DEV 27
434/99154
2583
cesspit 116
900
950
production waste
raw glass
Deventer
DEV 28
434/10638
11039
cesspit 256
890
925
vessel
funnel?
Deventer
DEV 29
434/99154
2583
cesspit 116
900
950
production waste
raw glass
Deventer
DEV 30a
434/7682
7214
cesspit 133
900
950
vessel
bottle?
Deventer
DEV 30b
434/7682
7214
cesspit 133
900
950
vessel
bottle?
Deventer
DEV 31
312/20975
20356
wastepit 30
900
950
window?
window
Deventer
DEV 32
312/22981
20765
wastepit 39
950
1000
window
window
Deventer
DEV 33
312/22985
20765
wastepit 39
950
1000
window
window
Deventer
DEV 34
312/22986
20765
wastepit 39
950
1000
bead?
undiagnostic
Deventer
DEV 35
434/99423
14377
cesspit 327
950
1050
vessel
undiagnostic
Deventer
DEV 36
434/99923
15616
cesspit 286
950
1050
vessel
undiagnostic
151
—
Key weathering
Key archaeological periods
x=slightly weathered xx = moderately weathered xxx= badly weathered
1= 450-550 AD 2= 550-650 AD 3= 650-750 AD 4= 750-850 AD 5= 850-1000 AD
Type of
fragment
Typology
(Isings, Koch,
Callmer, Ribe
or Pion)
Description object
Colour
Transparancy
Decoration
Weathering
Archeological
period
Publication
rim
-
-
light green
translucent
-
x
5
-
complete
-
-
light bluegreen
translucent
-
x
5
-
splinter
-
-
yellowbrown
translucent
-
-
5
-
rim
-
-
blue-green
translucent
opaque
white spiral
-
5
-
rim
-
-
blue-green
translucent
-
-
5
-
rim
-
-
light bluish
green
translucent
-
-
5
-
body
-
-
bluish green
translucent
-
-
5
-
body/base
-
-
undiagnostic
?
-
xxx
5
-
chip?
-
-
undiagnostic
?
-
xxx
5
-
fragment
-
-
-
?
xxx
5
-
chip?
-
-
undiagnostic
?
-
xxx
5
-
fragment
-
-
bluish green
translucent
-
-
5
-
body
-
-
bluish green
translucent
optic blown
ribs
-
5
-
rim
Isings 1957,
type 96a
-
yellow-green
translucent
-
x
5
-
chunk
-
-
bluish green
translucent
-
-
5
-
body
-
-
colourless
transparent
selfcoloured
trail
x
5
-
body, curved
-
-
pale pink
translucent
-
xx crizzled
5
-
fragment
-
-
pale bluegreen
translucent
-
xx
5
-
body
-
-
colourless
transparent
2 white
trails
-
5
-
fragment
-
-
deep
turquoise
translucent
-
xx
5
-
chip
-
-
light green
translucent
-
xx
5
-
body
-
-
yellowish
green
translucent
-
xx
5
-
chip
-
-
bluish green
translucent
-
xx
5
-
neck?
-
-
pale green
translucent
-
xx
5
-
neck?
-
-
red purple
streak
translucent
-
xx
5
-
fragment
-
-
blue-green
translucent
-
-
5
-
fragment
-
-
green?
?
-
xx
5
-
fragment
-
-
green?
?
-
xxx
5
-
fragment
-
-
amber
translucent
-
-
5
-
rim, thick
-
-
greenish
translucent
-
xxx
5
-
body
-
-
greenish
translucent
-
xx
5
-
152
—
Appendix I sample list
Site
Sample
Find number
Feature
number
Feature type
Begin date
feature (AD)
End date
feature (AD)
Category object
Form/object
Deventer
DEV 37
434/12362
12915
house 12
950
1050
vessel
funnel?
Deventer
DEV 38
434/99923
15616
cesspit 286
950
1050
window
window
Deventer
DEV 39
312/20678
20306
wastepit 34
950
1050
window
window
Deventer
DEV 40a
312/29048
22270
cesspit 80
950
1050
production waste?
melted, 2 layers
Deventer
DEV 40b
312/29048
22270
cesspit 80
950
1050
production waste?
melted, 2 layers
Deventer
DEV 41
312/29063
22425
layer 8
950
1050
window?
window/inlay?
153
—
Key weathering
Key archaeological periods
x=slightly weathered xx = moderately weathered xxx= badly weathered
1= 450-550 AD 2= 550-650 AD 3= 650-750 AD 4= 750-850 AD 5= 850-1000 AD
Type of
fragment
Typology
(Isings, Koch,
Callmer, Ribe
or Pion)
Description object
Colour
Transparancy
Decoration
Weathering
Archeological
period
Publication
body
-
-
light bluegreen
translucent
-
x
5
-
rectangle?
-
-
bluish green
translucent
-
x
5
-
fragment
-
-
greenish
translucent
-
xx
5
-
chip
-
-
red
opaque
-
-
5
-
chip
-
-
turquoise
translucent
-
-
5
-
fragment
-
-
turquoise
translucent
-
-
5
-
154
—
Appendix II
Element oxide
Na2O
major and minor chemical
compositions of samples
analysed by electron probe
microanalysis
MgO
Al2O3
SiO2
P2O5
SO3
Cl
K2O
Gennep glass samples
GE 41
13.69
0.86
2.68
67.89
0.08
0.33
-
0.76
GE 42
15.23
0.85
2.97
68.84
-
0.4
-
0.54
GE 43
15.18
0.77
2.94
69.55
0.03
0.37
-
0.43
GE 44
15.38
1
3.37
71.06
-
0.32
-
0.44
GE 45
14.74
0.98
3.1
68.21
0.05
0.31
-
0.55
GE 46
14.98
0.78
2.87
68.99
-
0.35
-
0.44
GE 47
15.91
0.62
2.7
74.73
-
0.33
-
0.69
GE 48
13.63
0.77
2.91
71.17
0.05
0.37
-
0.72
GE 49
14.34
0.85
2.91
70.65
-
0.36
-
0.76
GE 51
13.93
0.94
2.94
70.37
0.02
0.34
-
0.72
GE 52
14.44
0.89
2.4
70.83
0.01
0.33
-
0.56
GE 53
15.1
0.73
2.7
71.4
-
0.4
-
0.68
GE 54
15.85
1.11
2.81
67.76
0.01
0.42
-
0.61
GE 55
16.15
1.11
2.9
67.59
0.05
0.42
-
0.6
GE 56
16.36
1.22
2.8
70.4
-
0.31
-
0.41
GE57
14.04
0.92
2.84
68.36
-
0.41
-
0.86
GE 58
13.97
1
3.01
68.04
0.03
0.42
-
0.8
GE 59
15.86
0.99
2.77
68.66
0.01
0.44
-
0.68
GE 60
16.12
1.19
2.93
66.78
0.04
0.46
-
0.69
GE 61
13.78
0.69
2.74
72.39
0.01
0.26
-
0.72
GE 62
14.39
0.78
2.88
70.89
0.06
0.29
-
0.71
GE 63
13.6
0.68
2.85
72.72
0.01
0.23
-
0.73
GE 64
14.15
0.76
2.96
70.94
0.04
0.26
-
0.81
GE 65
15.05
0.74
2.86
71.65
-
0.3
-
0.61
GE 66
14.19
0.82
2.99
72.26
0.01
0.26
-
0.57
GE 67
14.57
0.83
2.85
72.54
0.02
0.27
-
0.61
GE 68
12.36
0.69
2.93
73.37
-
0.25
-
0.74
GE 69
14.95
1.04
3.02
71.02
-
0.25
-
0.42
11.77
1.15
2.43
53.26
0.13
0.31
0.63
0.62
Joden 2
12.43
1.01
2.44
60.89
0.15
0.24
0.64
0.72
Joden 3
16.42
1.21
2.56
67.05
0.09
0.35
0.79
0.57
Joden 4
14.1
1.13
2.67
64.19
0.15
0.33
0.78
0.44
Maastricht-Jodenstraat (MAJO) glass samples
Joden 1
Joden 5
14.49
1.24
2.79
62.07
0.18
0.27
0.8
0.47
Joden 6
10.49
0.92
2.23
58.09
0.19
0.23
0.67
0.48
Joden 37
16.04
0.72
2.25
69.97
0.05
0.34
0.96
0.55
Joden 38
17.03
0.88
2.35
70.36
0.1
0.39
0.84
0.44
Joden 39
16.54
0.76
2.35
70.14
0.05
0.31
0.95
0.46
Joden 40
16.87
0.73
2.26
71.34
0.02
0.3
1
0.3
Joden 41
15.49
1.14
2.6
62.42
0.18
0.22
0.7
0.84
Joden 42
14.15
0.72
2.38
61.92
0.1
0.23
0.84
0.66
Joden 43
16.14
1.16
2.58
61.94
0.2
0.25
0.66
0.87
155
—
CaO
TiO2
MnO
FeO
CoO
Sb2O5
SnO2
CuO
PbO
Total
7.42
0.12
0.84
1.49
0.01
0.66
-
0.53
2.99
100.41
7.43
0.32
1.53
2.14
-
0.19
0.02
0.13
0.35
101.04
6.94
0.31
2.26
2.04
0.03
0.02
0.03
-
0.02
100.97
6.44
0.32
1.59
1.99
0.04
0.04
-
0.03
0.06
102.16
7.75
0.31
2.4
1.99
-
0.06
-
0
0.11
100.6
6.76
0.31
2.25
1.61
-
0.03
-
-
0.02
99.47
7.6
0.11
0.93
0.6
-
0.35
0.02
0.03
0.04
104.74
8.2
0.16
0.53
0.92
0.04
0.18
‘-
0.03
0.21
99.9
8.58
0.2
1.03
0.95
-
0.15
0
0.12
0.38
101.37
7.33
0.19
1.06
1.26
-
0.54
0.01
0.41
1.37
101.48
7.07
0.16
0.86
1.14
0.01
0.36
-
0.29
1.51
100.96
8.18
0.14
1.52
0.69
0.02
0.08
0.04
0.06
0.09
101.9
8.84
0.14
1.49
0.68
-
0.08
-
-
-
99.88
8.9
0.18
1.51
0.78
-
0.09
-
-
0.02
100.32
6.98
0.25
1.78
0.86
0
0.01
0.01
-
0
101.44
9.2
0.15
1.4
0.92
-
0.08
-
-
0.08
99.35
9.23
0.18
1.34
1.09
0.03
0.08
-
-
0.02
99.28
9.06
0.15
1.3
0.85
-
0.07
-
0.04
0.01
100.96
9.46
0.15
1.54
0.86
0
0.08
-
-
0.03
100.37
7.79
0.09
0.66
0.73
-
0.27
-
0.02
0.06
100.28
7.67
0.15
0.98
0.78
-
0.19
0.01
0.02
0.14
99.96
7.6
0.11
0.9
0.6
-
0.24
-
-
0.09
100.43
7.84
0.13
0.99
0.84
0.02
0.14
0.01
0.04
0.2
100.17
7.05
0.16
1.09
0.8
0.02
0.23
-
0.08
0.39
101.1
7.67
0.15
0.95
0.8
0.02
0.15
0.01
0.01
0.09
101.01
7.44
0.17
1.11
0.75
-
0.14
-
0.06
0.1
101.5
8.28
0.1
0.74
0.64
-
0.15
0
0.02
0.07
100.45
6.32
0.42
1.96
1.23
-
0.02
0.01
0.03
0.04
100.78
5.11
0.11
1.21
2.66
0.01
-
0.65
0.45
7.13
87.61
6.53
0.12
1.18
2.51
0.01
-
0.63
0.58
6.13
96.22
7.07
0.12
1.5
0.79
-
-
0.53
-
0.37
99.42
6.41
0.16
1.45
1.24
-
-
3.08
0.2
4.8
101.12
6.26
0.19
1.63
1.39
-
-
2.86
-
4.69
99.32
5.25
0.09
1.36
0.71
-
-
1.72
-
19.16
101.58
5.55
0.1
0.05
0.86
0.01
-
0.18
0.02
0.67
98.31
6.18
0.12
0.13
0.57
0.01
-
-
0.02
0.27
99.69
5.62
0.11
0.04
0.56
-
0.07
-
-
0.3
98.27
5.42
0.11
0.06
0.68
0.02
-
-
-
0.53
99.64
6.9
0.14
1.27
1.3
-
0.04
0.14
1.54
3.13
98.03
6.28
0.1
0.69
1.55
-
0.2
0.17
1.7
7.32
99
6.74
0.14
1.45
0.86
0.01
-
0.59
2.77
0.71
97.08
156
—
Appendix II major and minor chemical compositions of samples analysed by electron probe microanalysis
Element oxide
Na2O
MgO
Joden 44
19.39
Joden 45
16.58
Joden 46
17.46
Joden 47
11.07
Joden 48
15.82
Joden 49
Joden 50
Joden 51
Al2O3
1.45
SiO2
P2O5
SO3
Cl
K2O
2.62
63.2
0.18
0.28
0.88
0.78
0.95
2.65
0.99
2.68
66.86
0.22
0.22
0.87
0.65
65.6
0.05
0.21
0.97
0.43
0.83
2.17
47.47
1.1
2.68
63.17
0.17
0.17
0.44
0.54
0.14
0.24
0.78
0.88
16.84
1.22
2.65
64.27
0.17
0.31
0.83
0.97
12.05
1.18
2.25
56.27
0.13
0.3
0.67
7.99
12.65
1.17
2.38
54.17
0.16
0.28
0.67
0.64
Joden 52
16
1.18
2.37
58.7
0.19
0.36
0.66
0.79
Joden 53
14.31
0.72
2.57
70.63
0.07
0.2
0.88
0.4
Joden 54
12.14
1.08
2.52
56.33
0.2
0.29
0.71
0.69
Joden 55
9.66
0.74
2
38.02
0.14
0.17
0.37
1.69
Joden 56
10.53
0.96
2.09
44.95
0.12
0.21
0.56
0.46
Joden 57
13.1
1
2.39
60.96
0.17
0.26
0.75
0.64
Joden 58
14.88
1.2
2.45
60.2
0.19
0.34
0.78
0.49
Joden 59
15.05
1.11
2.53
61.74
0.18
0.29
0.78
0.72
Joden 60
16.81
0.8
2.72
67.87
0.21
0.14
0.77
0.84
Joden 61
15.49
0.72
2.38
69.34
0.14
0.26
0.92
0.58
Joden 62
16.14
1.29
2.64
67.06
0.13
0.31
0.73
0.93
Joden 63
16.76
1.55
2.67
64.02
0.18
0.34
0.76
1.77
Joden 64
9.8
0.72
1.89
39.52
0.06
0.15
0.46
0.32
Joden 65
9.96
0.86
1.91
37.42
-
0.09
0.42
0.41
Joden 66
11.7
1.02
2.49
55.59
0.17
0.25
0.64
0.74
Joden 67
14.48
1.04
2.42
61.18
0.15
0.23
0.69
0.78
Joden 68
15.95
0.52
2.53
69.67
0.14
0.17
1.04
0.58
Joden 69
10.01
0.83
2.17
47.3
0.2
0.22
0.49
0.37
Joden 70
11.24
1.06
2.36
56.05
0.16
0.25
0.61
0.53
Joden 71
10.41
0.9
2.29
46.31
0.19
0.19
0.57
0.58
Joden 72
10.8
0.88
2.18
50.67
0.15
0.24
0.69
0.5
Joden 73
14.28
1.28
2.38
60.58
0.14
0.31
0.76
0.74
Joden 74
14.8
1.26
2.4
59.42
0.28
0.34
0.77
0.74
Joden 75
13.08
0.9
2.38
58.4
0.15
0.23
0.67
0.66
Joden 76
14.26
1.3
2.44
58.18
0.2
0.31
0.69
0.73
Maastricht-Jodenstraat (MAJO) crucibles
Joden 19 (lead glass)
0.31
0.35
4.07
24.57
0.1
0.69
-
1.01
Joden19 (yellow residue)
0.62
0.54
8.15
27.46
0.14
0
0.01
1.62
Joden 20 (lead glass)
0.41
0.68
2.53
22.74
0.03
0.07
0.1
1.03
Joden 20 (yellow residue)
0.406
0.675
2.533
22.737
0.1
-
0.029
1.034
Joden 21 (natron glass)
12.29
0.91
10.56
69.24
0.13
0.07
0.18
2.49
Joden 22 (lead glass)
0.26
0.59
4.39
26.65
0.21
0.17
0.04
0.85
Joden 23 (white melt)
0.11
0.39
2.47
12.45
0.04
0.04
-
-
157
—
CaO
TiO2
7.61
MnO
FeO
CoO
Sb2O5
SnO2
0.16
1.9
0.66
6.91
0.14
1.02
5.85
0.34
1.67
5.14
0.15
0.98
0.96
0.01
-
6.29
0.16
1.35
0.79
0.02
0.03
7.43
0.15
1.55
0.84
0.07
0.1
4.11
0.15
1.31
0.99
-
-
5.02
0.11
1.22
2.19
0.05
-
CuO
PbO
Total
-
0
-
-
0.05
99.16
0.95
-
0.08
-
0.08
0.02
98.21
1.08
0.03
0.04
-
0.12
-
97.52
3.71
-
23.75
97.55
0.31
1.84
3.14
98.72
-
0.08
0.04
97.54
1.16
0.12
2.48
91.15
0.81
0.75
8.63
90.9
6.59
0.12
1.36
0.78
0.05
-
0.17
4.16
0.27
93.74
4.86
0.14
0.05
0.74
0.04
0.03
0.05
0.17
0.62
96.48
5.74
0.13
1.24
3.94
-
-
1.5
1.1
7.69
95.31
3.16
0.1
0.87
1.96
0.06
-
1.02
0.36
36.54
96.86
5.5
0.11
1.13
0.9
0.02
-
3.13
0.07
26.27
96.99
7.42
0.14
1.22
0.72
0.05
-
0.24
2.64
3.71
95.42
5.96
0.12
1.59
0.7
0.02
-
5.68
0.07
2.21
96.89
6.53
0.13
1.56
0.8
0.07
-
2.12
0.08
3.13
96.81
6.63
0.15
0.68
0.8
0.01
0.21
-
0.11
0.35
99.11
5.89
0.09
0.41
0.74
0.04
1.72
-
0.15
0.33
99.19
7.53
0.16
1.76
0.73
0.02
0.06
-
0
-
99.48
7.96
0.16
1.8
0.78
-
0.09
-
-
-
98.85
3.05
0.09
0.87
0.7
-
-
3.04
0
42.07
102.75
2.82
0.09
0.59
0.56
0.01
-
1.71
-
27.16
84
5.69
0.14
1.44
2.7
-
-
1.57
1.42
10.78
96.35
6.65
0.17
1.29
2.55
-
0.1
0.13
2.33
3.27
97.44
6.92
0.07
0.46
0.42
0.03
0.12
-
0.08
0.08
98.8
4.27
0.09
0.86
1.07
0.01
-
3.8
0.01
24.46
96.15
5.58
0.14
1.1
0.89
-
-
1.77
0.01
17.23
98.98
4.98
0.08
1.01
1.25
0.07
-
2.81
0.09
26.72
98.44
4.95
0.12
1.37
0.7
-
-
1.46
0.04
24.48
99.22
8.23
0.13
1.58
0.84
0.06
-
0.28
2.57
3.96
98.13
7.23
0.14
1.64
0.89
0.08
-
1.8
2.89
4.2
98.88
8.83
0.11
1.09
0.84
0.03
-
0.67
2.59
5.86
96.48
6.88
0.12
1.47
0.82
-
0.03
0.39
2.47
3.9
94.19
0.68
0.26
0.01
1.55
-
-
0.49
0.06
61.53
95.66
3.84
0.51
0.04
2.43
0.03
-
9.48
-
42.98
97.84
0.86
0.27
0.05
2.48
-
-
0.54
-
62.48
94.26
0.86
0.272
0.051
0.483
0.028
-
9.536
-
62.477
101.22
2.17
0.48
0.33
3.17
0.02
0.03
-
-
-
102.07
1.64
0.24
0.05
1.5
0
0.06
1
0.1
56
93.74
0.54
0.11
0.01
0.87
-
-
61.58
0.01
22.83
101.45
158
—
Appendix II major and minor chemical compositions of samples analysed by electron probe microanalysis
Element oxide
Na2O
MgO
Al2O3
SiO2
P2O5
Joden 25 (lead glass)
0.15
0.67
6.24
28.15
Joden 27 (lead glass)
0.28
0.44
3.44
Joden 27 (yellow residue)
0.05
0.05
1.05
Joden 28 (lead glass)
0.31
0.34
Joden 28 (yellow residue)
0.09
0.15
Joden 29 (natron glass)
17.08
1.1
3.24
0.49
0.36
2.18
Joden 30 (white melt)
SO3
Cl
0.03
-
35.43
0.17
11.12
0.25
4.26
25.04
1.71
15.39
K2O
0.05
0.84
0.01
0.03
1.38
0.01
0.04
-
0.07
0.7
0.05
0.95
0.03
-
0.05
-
67.71
0.16
0.24
0.65
1.02
19.78
0.02
-
0.01
-
Maastricht-Mabro crucibles
Mabro 7
9.87
1.33
7.24
65.88
0.38
0.06
0.04
7.17
Mabro 8
15.79
0.72
9.87
69.38
0.08
0.13
0.38
0.85
Mabro 9
3.49
0.98
7.54
67.21
0.18
0.03
-
13
Mabro 10
12.43
1.49
2.74
65.69
0.24
0.11
0.01
2.99
Mabro 11
11.2
0.82
12.48
62.9
0.16
0.05
0.01
2.99
Mabro 12
0.02
0.25
0.65
15.64
0.06
-
0.06
0.84
Mabro 13
13.46
1.21
3.57
69.98
0.21
0.18
0.52
2.36
Mabro 14
0.49
0.59
6.27
28.43
0.09
-
0.05
0.86
Mabro 15
11.64
1.65
2.76
69.88
0.27
0.12
0.05
2.62
Mabro 16
16.67
0.99
3.48
68.99
0.1
0.19
0.38
1.27
WIJ1
16.97
0.92
2.57
69.16
0.1
0.26
0.79
0.73
WIJ2
17.69
0.95
2.62
68.26
0.1
0.23
0.92
0.74
WIJ3
16.44
0.94
2.69
68.43
0.13
0.28
0.78
0.85
WIJ4
10.24
0.23
1.03
40.69
0.01
0.13
0.72
0.19
WIJ5
10.86
0.89
2.39
37.21
0.11
0.15
0.41
0.57
WIJ6
10.76
0.73
2.18
39.23
0.11
0.14
0.48
0.45
WIJ 7
9.29
1.21
1.86
38.89
0.24
0.15
0.22
0.95
WIJ 8
8.72
0.38
2.35
40.37
0.09
0.01
0.48
0.4
Wijnaldum glass samples
WIJ 9
6.96
0.3
2.24
35.31
0.04
0.01
0.42
0.31
WIJ 10
14.93
0.71
2.47
68.16
0.11
0.12
1
0.61
WIJ 11
16.48
0.95
2.57
69.18
0.16
0.21
0.73
1.14
WIJ 12
15.99
0.88
2.65
70.57
0.11
0.22
0.99
0.77
WIJ 13
16.46
0.74
2.55
71.35
0.09
0.18
0.98
0.62
WIJ 14
7.05
0.46
1.97
26.42
0.07
0.11
0.37
0.39
WIJ 15
14.65
0.64
2.57
69.7
0.14
0.1
1
0.41
WIJ 16 low lead
15.72
0.84
2.79
69.24
0.16
0.19
0.83
0.91
WIJ 16 high lead
10.98
0.48
1.9
44.92
0.09
0.12
0.52
0.51
WIJ 17
6.58
0.58
4.13
48.65
0.34
0.03
0.27
1.23
WIJ 18
13.37
1.07
2.79
59.73
0.2
0.22
0.76
0.91
WIJ 19
13.5
1.28
2.78
58.87
0.21
0.3
0.88
2.69
WIJ 20
14.87
1.45
2.76
61
0.29
0.26
0.74
1.11
WIJ 21
15.01
1.42
2.74
64.76
0.23
0.29
0.92
0.79
159
—
CaO
TiO2
0.34
MnO
0.23
FeO
0
CoO
Sb2O5
SnO2
2.2
0.02
-
CuO
PbO
0.47
0.12
Total
55.03
94.55
1.36
0.21
0.03
1.64
0.02
-
0.44
0.09
51.62
96.58
0.3
0.25
0.02
0.52
0
-
23.03
0.04
65.34
102.06
0.77
0.24
-
1.61
-
-
0.75
-
60.24
95.33
0.25
0.16
0.03
0.46
0.01
-
17.85
0.09
59.97
96.24
7.23
0.18
1.34
0.91
0.01
0.05
-
0.21
0.1
101.22
0.76
0.12
0.02
0.78
0.01
-
55.81
0.49
22.5
103.33
4.95
0.33
0.55
2.65
0
0.2
-
-
-
100.66
2.44
0.38
0.09
2.22
0.01
0.02
-
0.08
0.05
102.48
-
0.52
0.1
2.71
0.15
0.56
-
0.01
-
96.46
7.58
0.18
0.38
1.17
0.01
0.17
-
0.11
-
95.29
2.52
1.36
0.7
4.79
-
0.11
-
0.02
0.09
100.19
0.47
0.29
0.09
1.16
0.01
-
12.75
0.27
59.24
91.8
6.74
0.23
0.39
0.99
0
0.15
-
0.36
0.08
100.43
0.66
0.31
0.06
2.48
0.01
-
1.1
0.1
53.77
95.26
7.52
0.14
0.43
1.16
-
0.11
-
0.03
0.05
98.42
5.82
0.25
1.05
1.17
-
0.04
-
-
0.05
100.45
7.15
0.15
1.04
0.85
0.02
0.13
0.01
0.1
0.15
101.1
7.07
0.16
1.18
0.9
0.02
0.14
0.03
0.05
0.17
101.22
7.79
0.18
1.07
0.99
0.01
0.1
0.06
0.14
0.18
101.05
2.76
0.04
0.01
0.3
0
-
4.85
-
40.26
101.47
3.41
0.11
0.33
1.32
0.01
-
3.69
0.05
38.82
100.32
3.39
0.12
0.33
0.91
0
-
5.12
0.06
36.76
100.77
5.72
0.15
1.1
1.27
0.03
-
3.11
-
36.64
100.83
3.76
0.08
0.01
0.53
0.02
-
5.11
0.09
36.84
99.23
2.76
0.06
0.02
0.43
0.03
-
6.18
0.06
44.35
99.47
9
0.27
0.2
0.82
0.02
0.07
-
0.13
0.02
98.64
6.93
0.17
0.83
0.77
0.02
0.31
-
0.14
0.17
100.76
6.6
0.18
0.98
0.79
0.01
0.28
-
0.1
0.23
101.34
7.02
0.11
0.93
0.64
0
0.39
-
0
0.1
102.18
1.81
0.1
0.04
0.83
0.01
-
4.86
0.15
56.58
101.23
9.48
0.29
0.23
1.21
0.13
0.02
0
0.11
0.11
100.81
7.38
0.1
0.54
1.05
0.02
0.35
0
0.38
0.37
100.86
4.29
0.07
0.41
0.75
0.02
-
2.38
0.15
33.93
101.53
2.69
0.29
1.72
7.46
0
-
0.14
1.81
23.21
99.12
5.99
0.16
1.12
4.36
0.03
-
0.79
1.99
6.62
100.11
5.49
0.17
0.27
3.84
0.04
-
1
1.02
7.61
99.96
7.23
0.18
1.25
5.63
0.02
-
0.7
0.57
2.89
100.93
6.55
0.14
1.29
2.7
0.01
-
0.98
-
1.49
99.32
160
—
Appendix II major and minor chemical compositions of samples analysed by electron probe microanalysis
Element oxide
Na2O
MgO
WIJ 22
14.07
WIJ 23
16.18
WIJ 24
11.23
WIJ 25
15.05
WIJ 26
17.07
WIJ 27
13.38
WIJ 28
14.09
WIJ 29
13.16
Al2O3
0.69
SiO2
P2O5
SO3
Cl
K2O
2.5
67.49
0.13
0.18
0.95
0.51
1.57
2.49
68.22
0.25
0.27
0.81
1.03
0.98
2.49
51.57
0.15
0.2
0.7
0.56
1.25
2.39
62.26
0.22
0.27
0.79
1.5
1.25
2.67
63.85
0.17
0.38
0.83
0.78
1.59
3.7
62.36
0.2
0.26
0.77
1.01
1.32
2.97
59.2
0.26
0.33
0.76
0.95
2.14
2.12
52.46
0.77
0.47
0.63
1.71
WIJ 30
14.47
1.33
2.29
62.44
0.22
0.21
0.78
0.88
WIJ 31
15.09
1.43
3.59
62.15
0.18
0.28
0.81
0.97
WIJ 32
11.1
0.64
2.08
45.68
0.1
0.21
0.66
0.51
WIJ 33
9.9
0.55
1.89
33.09
0.09
0.14
0.41
0.38
WIJ 34
0.24
0.19
6.26
35.54
0.5
0.01
0.01
1.66
WIJ 35
12.67
5.45
1.38
70.23
0.11
0.07
0.57
2.42
WIJ 36
17.07
1.31
2.72
65.1
0.17
0.28
0.91
0.71
WIJ 37
12.18
5.43
1.19
70.77
0.09
0.23
0.64
2.33
WIJ 38
16.01
1.11
2.65
69.88
0.07
0.31
0.55
0.62
WIJ 39 low lead
16.76
1.24
2.79
66.28
0.2
0.29
0.53
1.05
WIJ 39 high lead
11.07
1.2
2.78
58.87
0.4
0.22
0.54
1.19
WIJ 40
11.19
0.52
2.19
63.51
0.06
0.23
0.88
0.47
WIJ 41
16.97
1
2.82
67.28
0.15
0.23
0.93
0.68
WIJ 42
15.21
0.86
3.15
66.88
0.06
0.13
0.92
0.39
LM 25
16.85
0.66
2.55
70.62
0.1
0.2
0.99
0.75
LM 26
16.87
0.72
2.77
66.61
0.17
0.27
0.86
0.74
LM 27
15.02
0.67
2.34
65.33
0.14
0.23
0.61
1.04
LM 28
16.71
0.71
2.49
67.94
0.08
0.26
0.91
0.73
LM 29
15.44
0.8
2.42
68.03
0.14
0.24
0.87
0.93
LM 30
15.37
1.15
2.37
68.57
0.05
1.14
-
1.4
Wijk bij Duurstede (Dorestad) glass samples
LM 31
16.21
0.79
2.39
68.51
0.06
1.12
-
1.14
LM 33
15.54
0.92
2.75
67.2
0.02
1.17
0.01
1.03
DOR 53
17.01
0.73
2.63
69.39
0.13
0.18
0.85
0.89
DOR 61
16.44
0.79
2.85
68.86
0.16
0.17
0.79
0.87
DOR 66
17.72
0.67
2.34
71.22
0.03
1.23
-
0.54
DOR 90
15.95
0.77
2.61
68.89
0.01
1.19
0.04
1.1
DOR 91
18.09
0.61
2.48
70.9
0.02
0.28
-
0
DOR 95
1.3
1.53
12.35
64.2
0.25
0.01
-
4.32
DOR 100
15.62
0.64
2.9
70.82
0.03
1.12
0.01
1.09
DOR 101
16.4
0.69
2.66
71.25
0.07
1.25
0.01
0.83
DOR 102
15.7
0.66
2.61
71.92
-
1.25
-
0.91
DOR 103
1.48
7.14
2.37
61.68
0.01
0.99
0.01
8.56
161
—
CaO
TiO2
MnO
FeO
CoO
Sb2O5
SnO2
-
CuO
PbO
Total
0.58
0.08
0.56
0.85
0.01
3.44
0.02
3.19
95.25
0.32
0.18
0.63
0.97
0.02
0.3
0.14
1.29
1.43
0.02
-
1.7
0.02
1.06
95.7
-
1.56
-
23.37
95.99
6.18
0.16
1.24
2.04
-
-
1.51
-
5.12
99.99
6.42
0.1
1.27
1.78
0.01
-
0.54
0.14
3.62
100.87
6.05
0.17
0.55
3.81
0.02
-
0.11
1.57
4.22
99.77
6.38
0.14
0.61
4.09
0.01
-
0.27
2.45
5.64
99.46
8.44
0.17
0.22
3.32
0.01
1.57
0.63
11.39
1.75
100.95
5.85
0.21
0.89
3.9
0.01
0.03
0.03
2.22
3.9
99.66
6.09
0.21
0.51
3.62
-
0.01
-
1.12
4.3
100.36
3.61
0.07
0.21
0.86
-
-
2.41
-
32.37
100.51
2.22
0.13
0.09
0.83
-
-
3.26
0.28
46.8
100.04
0.22
0.41
4.24
9.09
0.04
0.01
0.17
0.25
40.15
98.97
6.64
0.1
0.64
0.6
0.01
0.1
-
-
0.05
101.05
9.19
0.16
1.85
1.29
0.01
0.05
-
0.11
-
100.93
6.21
0.02
0.64
0.42
-
0.08
-
0.01
0.02
100.25
7.97
0.14
0.49
0.86
0.02
0.04
-
0.02
-
100.72
8.65
0.11
1.09
0.94
0.04
0.04
-
0.16
0.29
100.46
6.1
0.22
1.1
1.56
0.02
-
3.25
1.85
8.29
98.67
3.86
0.11
0.52
1.81
-
1.33
-
-
13.27
99.94
6.65
0.29
1.13
1.35
0
0.26
-
0.06
0.36
100.16
8.89
0.08
0.03
0.45
0.03
0.01
0.01
2.26
1.61
100.96
6.61
0.12
0.66
0.59
-
0.48
-
-
-
101.18
8.66
0.13
0.59
0.73
0.02
0.29
-
0.34
0.32
100.08
7.03
0.14
-
0.01
0
0.29
0.02
0
0.39
93.26
6.64
0.11
-
-
0.01
0.32
0.01
-
0.27
97.17
7.81
0.15
0.01
0.01
0
0.16
-
-
0.26
97.26
7.44
0.11
-
0.73
0.05
0.64
-
0.11
-
99.11
6.54
0.11
0.06
0.54
0.01
0.21
-
1.25
-
98.93
7.07
0.21
-
0.87
0.01
0.17
-
2.24
-
99.2
7.25
0.15
0.67
0.67
-
0.41
-
0.07
0.24
101.26
7.46
0.15
0.61
0.84
-
0.33
0.02
0.16
0.69
101.19
5.94
0.1
0.02
0.51
-
0.52
-
0.04
1.26
102.13
7.28
0.15
-
0.79
0.02
0.22
0
0.37
-
99.41
5.99
0.12
-
0.6
0.03
0.58
0.03
0.02
1.19
100.93
2.78
0.55
0.06
11.94
0.02
0.2
-
0.1
0.02
99.65
7.2
0.14
-
0.89
0.04
0.3
-
0.16
0.31
101.27
6.65
0.12
-
0.85
0.03
0.62
-
0.33
0.49
102.26
6.82
0.13
-
0.91
-
0.37
-
0.11
0.03
101.42
13.65
0.18
-
0.71
0.04
0.33
-
-
-
97.16
162
—
Appendix II major and minor chemical compositions of samples analysed by electron probe microanalysis
Element oxide
Na2O
MgO
DOR 104
15.86
DOR 105 colourless
DOR 105 op red
Al2O3
SiO2
P2O5
0.71
2.86
71.87
15.94
1.11
2.66
68.89
15
0.64
2.73
68.24
SO3
-
Cl
K2O
1.21
0.01
1.01
0.03
1.19
0.01
1.11
0.06
1.18
-
0.91
DOR 106
16.28
0.62
2.82
71.63
0.01
1.26
0.02
0.81
DOR 107
16
0.82
2.8
71.4
-
1.19
0.01
0.95
DOR 108
16.47
0.73
2.69
72.34
-
1.14
-
0.7
DOR 109
15.92
0.75
2.8
72.3
-
1.14
0.02
0.91
DOR 110
16.56
0.81
2.78
68.65
0.15
0.21
0.85
1.02
DOR 111
14.73
1.11
3.07
65.81
0.22
0.19
0.64
1.08
DOR 112
16.35
0.82
2.63
69.85
0.15
0.2
0.93
1.02
DOR 113
16.25
0.86
2.7
69.08
0.13
0.27
0.88
0.93
DOR 115
16.57
1.33
2.21
66.95
0.21
0.28
0.81
1.43
DOR 116
16.55
0.84
2.71
67.86
0.13
0.28
0.89
0.75
DOR 117
14.8
0.98
2.77
69.81
0.19
0.21
0.82
1.08
DOR 118
16.19
0.81
2.85
69.74
0.15
0.17
0.86
0.94
DOR 119
15.78
0.81
2.75
68.22
0.14
0.18
0.69
0.85
DOR 120
16.51
0.85
2.6
68
0.17
0.19
0.84
1.11
DOR 121
17.27
0.86
2.63
69.12
0.1
0.24
0.97
0.78
DOR 122
15.13
1.01
2.3
67.69
0.11
0.23
1
1
DOR 123a
16.92
0.84
2.91
67.41
0.09
0.22
1
0.61
DOR 123b
17.28
0.85
2.69
68.44
0.08
0.21
0.97
0.69
DOR 124
15.07
0.71
2.98
68.85
0.22
0.15
0.72
0.86
DOR 125
17.6
0.77
2.38
71.25
0.09
0.2
1.09
0.49
DOR 126
16.11
0.7
2.48
70.13
0.08
0.24
1.15
0.57
DOR 127
17.21
0.7
2.67
69.84
0.05
0.16
0.99
0.62
DOR 128
17.17
0.69
2.51
70.15
0.08
0.17
1.1
0.6
DOR 129
16.51
0.87
2.83
66.66
0.15
0.17
0.86
0.97
DOR 130a
14.07
0.94
2.81
69.25
0.19
0.19
0.72
0.93
DOR 130b
16.2
0.81
3.44
69.36
0.19
0.19
0.49
1
DOR 131
16.97
0.72
2.77
69.13
0.11
0.19
0.72
0.62
DOR 132
16.29
0.87
2.94
67.68
0.24
0.17
0.71
1.31
DOR 133
15.4
0.74
2.8
68.69
0.19
0.24
0.69
0.98
DOR 134
15.33
0.81
2.63
68.6
0.1
0.23
1.04
0.69
DOR 135
16.54
1.02
2.94
69.59
0.25
0.2
0.35
1.14
DOR 136
1.21
8.09
1.07
63.76
2.06
0.08
0.03
7.51
DOR 137
15.96
0.8
2.74
68.3
0.19
0.21
0.78
0.95
DOR 138
14.71
0.86
2.91
69.39
0.25
0.19
0.78
0.87
DOR 139
15.84
0.72
2.82
69.19
0.16
0.17
0.78
1.02
DOR 141
15.74
0.87
2.8
68.08
0.22
0.19
0.81
1.1
DOR 142
15.95
0.85
2.77
68.54
0.15
0.21
0.66
0.85
DOR 143
17.47
0.9
2.7
67.55
0.1
0.28
1.06
0.54
163
—
CaO
TiO2
MnO
FeO
CoO
7
0.12
-
0.67
7.29
0.16
-
6.8
0.14
-
Sb2O5
SnO2
CuO
PbO
0.01
0.29
-
0.02
0.72
-
0.28
0.01
2.14
0.01
0.2
-
Total
0.27
101.9
0.15
-
99.54
1.11
0.38
99.53
6.93
0.12
-
0.84
0.03
0.12
0
0.09
0.63
102.22
6.81
0.14
0.02
0.84
-
0.39
-
0.2
0.77
102.34
6.97
0.14
-
0.66
0.02
0.27
-
0.06
1.09
103.28
7.16
0.12
-
0.8
0.02
0.22
0.01
0.05
-
102.22
7.25
0.11
0.56
0.83
0
0.25
-
0.06
0.41
100.48
7.03
0.2
0.63
1.86
0.02
0.08
0.25
0.55
3.31
100.79
7.27
0.15
0.57
0.74
0.01
0.21
-
0.16
0.35
101.41
7.29
0.16
0.77
0.79
-
0.35
-
0.08
0.39
100.91
7.89
0.15
1.18
0.98
0.01
0.19
-
0.1
0.14
100.43
7.6
0.17
1
0.78
0.01
0.22
-
0.21
0.38
100.39
7.6
0.18
0.77
0.86
0.02
0.34
-
0.08
0.35
100.87
7.18
0.09
0.73
0.72
-
0.26
-
-
0.33
101.01
7.12
0.1
0.53
0.72
0.01
0.3
-
0.29
0.63
99.11
7.78
0.14
0.71
0.75
-
0.27
-
0.06
0.31
100.28
6.96
0.15
0.89
0.73
0.02
0.25
-
0.1
0.29
101.37
9.49
0.25
1.27
0.82
-
0.11
-
-
0.09
100.5
7.27
0.2
1.16
0.84
0.04
0.2
-
0.1
0.11
99.9
7.65
0.15
1.08
0.85
-
0.22
-
-
0.02
101.18
6.78
0.18
0.42
2.49
0.02
0.34
-
0.03
0.9
100.7
6.22
0.13
0.49
0.5
-
0.41
-
-
0.08
101.7
6.21
0.13
0.57
0.55
0.02
0.59
-
0.07
0.16
99.76
7.02
0.11
0.87
0.42
0.01
0.3
-
0.02
-
100.99
6.81
0.09
0.67
0.5
-
0.35
-
-
-
100.89
6.9
0.1
0.61
0.82
0.02
0.42
-
0.34
1.16
99.37
6.95
0.11
0.63
0.82
0
0.37
-
0.27
0.29
98.55
6.56
0.2
0.51
1.14
0.04
1.07
-
-
0.37
101.57
6.58
0.15
0.38
0.74
0.01
0.63
-
0.03
0.19
99.95
7.97
0.15
0.58
0.89
0.01
0.23
0.03
0.25
0.46
100.77
7.11
0.1
0.58
0.74
-
0.39
0.12
0.34
1.03
100.13
6.92
0.13
0.63
0.98
-
0.68
-
0.53
1.36
100.66
7.74
0.13
0.77
0.73
-
0.17
-
-
0.03
101.6
13.21
0.07
0.93
0.52
-
0.33
-
-
0.09
98.95
7.17
0.14
0.65
0.78
0.03
0.44
-
0.08
0.43
99.63
7.26
0.15
0.63
0.85
0.02
0.33
-
0.12
0.85
100.15
7.4
0.12
0.43
0.91
0.02
0.27
0.21
0.01
0.12
100.18
7.37
0.15
0.66
0.85
0.03
0.41
0.05
0.19
1.05
100.56
7.1
0.15
0.51
0.88
-
0.74
-
0.33
1.07
100.76
7.13
0.16
0.7
1.09
0.03
0.74
-
0.29
0.09
100.83
164
—
Appendix II major and minor chemical compositions of samples analysed by electron probe microanalysis
Element oxide
DOR 144
Na2O
MgO
Al2O3
15.19
0.56
DOR 145
14.25
DOR 146
16.97
DOR 147
DOR 148
DOR 149
SiO2
P2O5
SO3
Cl
K2O
2.33
71.88
0.08
0.2
1.11
0.5
0.65
2.56
69.38
0.08
0.29
0.88
0.61
0.42
2.02
67.79
0.05
0.36
1.38
0.4
15.32
0.68
2.62
66.99
0.09
0.31
0.73
0.57
11.36
0.54
2.54
62.79
0.19
0.18
0.8
0.62
10.79
0.44
1.99
41.2
0.05
0.1
0.49
0.47
DOR 149 mineral
1.52
0.03
0.81
9.1
-
-
0.14
-
DOR 150
1.49
1.87
7.54
43.81
1.29
0.02
0.02
3.66
DOR 151
1.2
1.82
6.38
43.41
1.12
0.01
0.01
3.7
Deventer glass samples
DEV 1
3.1
1.1
7.81
64.21
0.25
0.1
0.11
9.97
DEV 2
16.95
0.73
2.56
69.18
0.06
0.34
0.92
0.53
DEV 3
16.19
0.72
2.81
70.23
0.16
0.25
0.96
0.92
DEV 4
7.2
1.91
1.38
63.86
0.15
0.16
0.38
13.77
DEV 5
10.25
1.81
1.26
64.4
0.47
0.13
0.5
10.44
DEV 6
2.83
5.98
2.99
59.23
2.28
0.05
0.55
12.68
DEV 7
3.25
6.04
3.48
59.99
2.29
0.14
0.58
11.47
DEV 8
0.4
4.16
0.95
60.66
2.07
0.08
0.08
11.62
DEV 9
12.87
1.56
1.93
73.24
0.04
0.06
0.03
0.58
DEV 10
14.55
0.97
2.84
69.44
0.24
0.12
0.9
1.57
DEV 11
14.97
0.69
2.89
70.25
0.18
0.14
0.94
0.61
DEV 12
16.06
0.84
2.74
69.76
0.12
0.21
0.89
0.9
DEV 13
16.76
0.95
2.23
70.38
0.04
0.19
1.19
1.39
DEV 14
0.99
4.39
2.05
59.15
2.97
0.15
0.37
11.85
DEV 15
1.11
3.91
3.29
53.43
4.38
0.53
0.18
17.01
DEV 16 weathered
0.55
0.15
5.85
75.71
-
0.13
0.03
5.38
DEV 17 weathered
0.05
0.03
1.47
82.61
-
0.01
0.04
1.46
DEV 18
8.27
3.48
2.17
60.23
2.47
0.15
0.46
9.17
DEV 19
16.8
0.94
2.76
69.01
0.12
0.21
0.85
1.1
DEV 20
15.98
1.03
2.5
71.68
0.08
0.22
1.22
0.75
DEV 21
0.54
3.54
2.6
59.36
1.87
0.06
0.04
9.52
DEV 22
7.84
1.91
1.24
63.73
0.18
0.19
0.38
14.54
DEV 23 weathered
2.7
0.88
1.08
78.46
1.22
0.13
0.05
2.97
DEV 24
2.92
3.82
1.61
56.1
3.89
0.15
0.39
4.1
DEV 25
13.63
1.84
1.22
69.37
0.17
0.19
0.43
8.03
DEV 26
1.47
3.59
1.24
55.88
2.76
0.16
0.15
17.79
DEV 27
1.84
5.2
3.19
54.63
3.64
0.17
0.17
13.09
DEV 28
3.29
6.25
2.99
58.73
2.35
0.03
0.49
12.73
DEV 29
0.4
3.78
2.24
56.24
2.18
0.1
0.02
10.59
DEV 30a
1.26
3.75
2.82
50.15
1.92
0.21
0.26
18.18
DEV 30b
0.28
0.64
6.05
86.49
-
0.03
0.08
1.57
165
—
CaO
TiO2
5.96
MnO
0.09
FeO
0.59
CoO
Sb2O5
0.76
0.15
SnO2
CuO
PbO
Total
1.16
-
0.2
-
100.76
6.55
0.12
0.37
0.75
0.02
2.61
-
0.09
0.62
99.83
5.77
0.06
0.02
0.43
0.01
2.89
-
2.5
0.14
101.2
6.9
0.1
0.55
0.54
-
1.98
-
0.94
1
99.31
6.91
0.11
0.97
0.5
-
0.59
-
1.79
10.66
100.56
3.78
0.14
0.31
0.49
0.01
-
3.01
0.29
39.12
102.66
0.99
0.04
0.08
0.36
-
-
26.72
0.26
62.64
102.68
12.28
0.22
0.18
2.3
0
0.66
-
-
21.95
97.29
13.75
0.24
0.22
2.27
0.02
0.69
-
-
22.12
96.95
10.95
0.47
0.03
2.29
0.02
0.54
-
-
-
100.95
6.95
0.1
0.87
0.96
0.06
0.05
-
-
0.11
100.38
6.75
0.17
0.59
0.66
0
0.32
-
0.08
0.18
100.99
5.77
0.11
1.08
0.49
0.01
0.77
-
0.11
2.83
99.98
8.6
0.07
1.24
0.47
0
0.54
-
-
0.08
100.24
11.09
0.24
0.75
0.76
-
0.54
-
0.06
0.01
100.03
10.59
0.29
0.65
0.9
-
0.59
-
0.2
-
100.45
17.98
0.11
0.72
0.31
0.02
0.56
-
-
-
99.72
10.43
0.03
0.01
0.31
-
0.05
-
0.04
0.03
101.2
8.46
0.21
0.48
0.94
0.01
0.13
-
0.13
0.05
101.04
9.02
0.25
0.17
0.88
-
0.05
-
0.17
0.01
101.22
6.73
0.16
0.91
0.73
-
0.24
-
0.01
0.07
100.37
6.51
0.09
1.09
0.98
-
0.02
-
0.01
0.01
101.83
15.24
0.29
0.72
1
-
0.54
-
0.12
-
99.82
13.73
0.2
0.38
0.93
0.02
0.79
-
0.15
0.03
100.06
1.93
0.24
0.04
0.44
-
0.19
-
-
0.06
90.7
0.81
0.15
0.01
0.12
0.02
0.2
-
0.01
-
86.99
11.5
0.17
0.64
0.96
0.02
0.5
-
0.07
0.09
100.33
7.06
0.16
1.09
0.99
0.01
0.19
-
-
0.21
101.48
6.08
0.08
1.08
1.02
0.02
0.04
-
0.05
0.02
101.84
17.78
0.36
0.87
0.85
0
0.61
-
-
-
98
5.13
0.08
0.49
0.23
-
0.74
-
-
2.42
99.09
1.33
0.11
0.28
0.36
0.02
0.12
-
0.1
0.09
89.9
23.59
0.14
1.23
0.55
0.04
0.26
-
-
0.01
98.79
4.36
0.12
0.49
0.26
0.01
0.37
-
-
0.01
100.49
11.95
0.07
0.42
0.42
-
0.86
-
2.46
0.06
99.28
14.31
0.3
0.79
1.25
-
0.54
-
0.23
-
99.35
10.79
0.18
0.72
0.84
0.02
0.62
-
0.02
0.13
100.18
21.15
0.27
0.74
0.71
-
0.54
-
-
-
98.97
17.56
0.12
0.95
0.61
0.02
0.93
-
-
-
98.72
0.99
0.1
0.01
0.26
0.13
0.06
-
0.05
0.07
96.8
166
—
Appendix II major and minor chemical compositions of samples analysed by electron probe microanalysis
Element oxide
Na2O
MgO
DEV 31
14.97
DEV 32
0.89
DEV 33
0.84
DEV 34 weathered
Al2O3
0.59
SiO2
P2O5
SO3
0.15
Cl
0.1
K2O
2.6
70.13
1.11
0.71
5.4
0.74
54.94
2.77
0.21
0.11
12.71
4.78
0.88
57.22
2.08
0.06
0.54
17.03
-
-
-
-
-
-
-
-
2.09
5.14
0.98
52.27
3.08
0.19
0.23
13.77
DEV 36
1.18
3.72
1.24
56.07
3.15
0.2
0.13
17.51
DEV 37
18.06
1.8
1.91
63.98
0.53
0.04
1.25
4.77
DEV 38
1.38
3.52
1.02
58.08
3.2
0.23
0.27
16.1
DEV 39
0.74
4.46
2.41
56.75
3.12
0.04
0.27
14.09
DEV 40a
14.95
0.62
2.61
67.97
0.12
0.16
0.87
0.82
DEV 40b
14.93
0.67
2.7
70.23
0.14
0.13
0.91
0.71
DEV 41
14.55
1.08
2.62
67.19
0.41
0.2
0.92
1.97
Sust 1 (bead body)
16.71
0.83
2.76
63.08
0.18
0.1
0.82
0.83
Sust 2 (bead body)
18.01
0.81
2.75
68.22
0.08
0.18
0.88
0.84
Sust 2 (decoration)
16.31
0.74
2.61
62.49
0.2
0.19
0.81
0.95
Sust 3 (bead body)
15.85
3.4
2.79
65.12
0.25
0.27
0.53
1.9
Sust 3 (decoration)
8.98
0.39
1.5
36.15
0.08
0.03
0.44
0.51
Sust 4 (bead body)
16.88
5.75
2.06
64.47
0.23
0.18
0.68
1.86
DEV 35
Susteren glass samples
Sust 4 (decoration)
16.81
1.36
2.72
67.7
0.15
0.15
0.85
1.57
Sust 5 (bead body)
20.03
0.78
2.49
66.7
0.1
0.23
1.17
0.68
Sust 5 (decoration)
18.5
0.65
2.52
67.29
0.1
0.6
1.15
0.7
Sust 6 (bead body)
16.96
0.75
3.05
69.97
0.06
0.16
0.92
1.21
Sust 6 (decoration)
17.67
0.85
2.79
67.06
0.13
0.22
0.91
1
Sust 7
18.89
0.71
2.47
67.15
0.08
0.32
1.05
0.64
Sust 8
18.99
0.78
2.44
66.01
0.09
0.37
1.08
0.58
Sust 9
17.24
0.84
2.95
64.99
0.1
0.16
0.62
0.78
Sust 10
19.29
0.83
2.6
65.33
0.09
0.29
0.84
0.66
Sust 11
1.9
5.6
2.46
59.06
2.53
0.06
0.47
14.21
Sust 12
17.99
0.99
2.72
66.44
0.21
0.22
0.98
0.84
Sust 13
16.98
1.98
2.66
64.89
0.21
0.22
0.8
1.48
Sust 14
16.18
1.17
2.64
65.83
0.31
0.19
0.82
1.27
Sust 15
18.29
1.07
2.66
62.45
0.15
0.28
1.23
0.74
Sust 16
15.07
1.41
3.6
66.99
0.31
0.16
0.74
1.54
Sust 17
14.26
0.66
2.33
65.14
0.18
0.24
0.67
5.2
Sust 18
13.96
0.88
7.74
63.34
0.31
0.04
0.03
3.3
Sust 19
8.87
3.69
2.23
61.74
1.5
0.16
0.47
8.5
Sust 20
19.04
0.86
2.58
64.85
0.1
0.35
1.26
0.64
Sust 21
16.48
1.05
2.68
67.22
0.19
0.18
0.83
1.1
Sust 22
16.56
1.08
2.69
67.45
0.27
0.27
0.81
1.24
Sust 23
18.06
1.01
2.6
66.25
0.37
0.22
0.96
0.84
167
—
CaO
TiO2
MnO
9.33
0.3
20.26
0.09
14.37
0.1
FeO
0.12
CoO
Sb2O5
SnO2
CuO
PbO
Total
0.86
0.02
-
-
0.19
0.07
101.25
0.53
0.31
0.01
0.64
-
0.07
-
99.68
0.57
0.41
-
0.71
-
-
-
99.58
-
-
-
-
-
-
-
-
-
0
17.69
0.11
2.36
0.44
0.01
0.5
-
-
0.09
98.94
14.31
0.11
0.39
0.44
0
0.72
-
0.1
-
99.27
7.55
0.12
0.08
0.46
0.01
0.21
-
-
-
100.77
14.59
0.05
0.35
0.31
-
0.66
-
0.06
-
99.81
15.19
0.26
0.4
0.83
-
0.61
-
0.08
0.03
99.28
8.89
0.23
0.18
0.88
0
0.06
-
2.29
0.13
100.78
8.95
0.27
0.17
0.88
0.02
0.07
-
0.09
0.29
101.14
7.01
0.11
0.53
0.56
-
0.24
-
2.57
0.02
99.97
7.08
0.17
0.63
5.09
0.01
0.27
0.23
0.26
1.35
100.4
7.1
0.12
0.62
0.86
0
0.29
0.12
0.27
0.71
101.86
6.74
0.13
0.57
4.59
0.01
0.34
0.28
1.74
1.69
100.38
7.03
0.18
0.54
1.29
0.1
0.2
0.32
0.14
0.24
100.14
3.5
0.12
0.43
0.54
0
0
2.72
0.13
44.29
99.8
7.28
0.1
0.61
0.63
0
0
0
0.12
0.12
100.95
7.68
0.12
0.67
0.95
0.01
0.25
0.14
0.16
0.44
101.73
6.66
0.22
0.84
0.63
0.02
0.34
0
0.03
0.09
101.01
6.28
0.09
0.44
0.69
0
1.79
0
0.16
0.68
101.62
7.15
0.13
0.25
0.38
0
0
0.19
0.39
0.21
100.96
7.97
0.13
0.71
0.82
0.01
0.32
0
0.17
0.22
101.76
6.56
0.11
0.32
0.88
0.07
1.58
0
0.3
0.22
101.35
6.55
0.08
0.37
0.94
0
1.7
0
0.27
0.3
100.53
6.28
0.18
0.54
1.38
0.02
0.48
0.22
3.06
1.3
101.12
7.07
0.13
0.75
0.44
0
0.35
0
0
0.2
98.89
11.02
0.34
0.59
1.01
0
0
0
0.04
0
99.28
7.13
0.2
0.81
0.91
0
0.36
0.14
0.31
0.84
101.09
7.39
0.15
0.55
0.89
0
0.3
0.14
0.28
0.57
99.48
8.89
0.24
0.66
1.05
0.03
0.16
0
0.26
0.36
100.05
6.46
0.18
0.31
1
0
0.69
0.2
0.6
2.73
99.01
7.49
0.2
0.51
1.31
0
0.17
0
0.17
0.58
100.25
7.08
0.1
0.43
1.04
0.07
1.5
0.08
0.25
0.68
99.91
5.81
1.07
0.47
1.35
0.02
0.17
0.06
0.79
0.05
99.37
9.73
0.16
0.7
0.89
0.01
0.25
0.09
0.26
0.46
99.69
6.94
0.11
0.87
0.92
0.01
0.33
0
0.29
0.65
99.79
8.39
0.16
0.6
1.1
0.01
0.18
0
0.2
0.42
100.8
8.28
0.23
0.71
1.01
0.02
0.09
0.13
0.35
0.71
101.89
7.46
0.1
0.43
0.9
0
0.38
0.13
0.28
0.75
100.74
168
—
Appendix II major and minor chemical compositions of samples analysed by electron probe microanalysis
Element oxide
Sust 24
Na2O
MgO
Al2O3
SiO2
P2O5
17.92
0.9
2.8
66.41
Sust 25
16.61
0.76
2.62
Sust 26
19.09
1.16
2.87
Sust 27
16.62
1.2
Sust 28
15.38
4.2
Sust 29
17.79
0.84
16.45
0.92
SO3
Cl
K2O
0.22
0.24
0.89
0.85
67.02
0.08
0.28
0.82
0.62
65.79
0.08
0.24
1.03
0.68
2.61
66.22
0.28
0.17
0.84
1.23
2.03
64.61
0.27
0.15
0.66
2.48
2.73
66.47
0.15
0.26
0.92
0.76
2.65
68.09
0.18
0.35
1.1
0.39
Utrecht glass samples
Utr 77
Utr 78 modern
-
-
-
-
-
-
-
-
13.5
0.41
1.7
69.87
0.03
0.21
0.03
0.39
Utr 31
8.77
0.57
12.6
66.05
0.03
0.01
0.1
4.73
Utr 32
15.54
1.66
4.65
67.07
0.06
-
0.3
1.9
Utr 33
0.37
0.44
36.89
54.29
-
0.04
0.04
2.26
Utr 79
Utrecht crucibles
Utr 34
14.39
1.02
2.82
63.59
0.56
0.15
0.2
1.45
Utr 35
13.08
0.76
7.41
69.7
-
0.05
0.24
2.24
Utr 36
8.21
0.84
7.68
66.9
-
0.16
0.07
5.4
169
—
CaO
TiO2
MnO
FeO
CoO
Sb2O5
SnO2
7.5
0.18
0.61
0.88
0.03
1.97
7.28
0.09
0.34
0.86
0.09
6.27
0.4
1.79
1.26
0.01
7.88
0.16
0.65
0.88
0
0.25
7.1
0.16
0.6
0.7
0.03
0.11
7.24
0.18
0.74
0.95
0
0.5
6.29
0.11
0.08
0.74
-
CuO
PbO
Total
0
0.24
0.61
102.25
0.24
0
0.24
0.85
98.79
0
0
0
0.08
100.74
0
0.24
0.59
99.82
0
0.06
1.64
100.16
0.14
0.53
1.4
101.6
-
-
0.06
0.57
97.98
-
-
-
-
-
-
-
-
-
-
10.39
0.03
0.03
0.12
0.08
0.04
-
0.06
0.07
96.97
2.48
0.64
0.14
3.17
0.11
0.45
0.02
0.03
0.09
99.98
5.78
0.21
0.22
1.53
0.14
0.11
0.02
0.01
-
99.19
0.34
2.22
0.02
2.88
0
0.12
0.02
0.01
0
99.93
6.82
0.15
0.73
1.52
0.13
0
-
0.62
4.91
99.07
4.87
0.59
0.37
1.36
0.06
0.29
0.03
1.12
-
102.16
3.29
0.45
0.3
1.59
0.03
0.5
-
0.39
-
95.81
170
—
Trace
elements
Appendix III trace element chemical
compositions of samples
analysed by LA-ICP-MS
Li
B
V
Cr
Co
Ni
Cu
Zn
As
Rb
Sr
Y
Zr
Nb
Mo
Sn
Sb
Gennep glass samples
GE41
7.4
187.9
27
22
54
27.9
5177.3
258.9
59.5
9.1
484
7.5
88.8
3.3
2.4
2546
4570
GE42
8
351.5
78
133
22.1
27.6
1194.5
31.6
19.4
7.3
522
12.2
181
4.5
2.4
308
1081
GE43
5.1
107
74
40
11.5
22
114.6
32.6
11
6.1
500
10.1
142.7
3.8
3.7
10
24
GE44
6.5
199.3
74
40
12
25.5
231.8
37.5
11.7
6.8
501
10.6
139.1
3.8
3.6
125
163
175
GE45
7.2
151.3
76
47
16
25.3
358.6
44.7
15.8
7.4
552
11.5
185.5
4.6
3.1
67
GE46
4.9
114.7
75
40
11.9
21.9
120.6
37.2
10.8
6
504
10.9
151.8
3.9
3.7
25
68
GE47
7.2
201.3
26
19
37.1
18.2
3408.8
105.9
44.5
8.9
536
7.6
86.1
2.8
2.5
2639
3559
GE48
9.3
158.1
16
12
6.7
7.7
131.3
37.2
10.6
9.2
450
6.5
52.7
1.8
1.3
72
1741
GE49
11.2
165.5
25
18
14.4
11.7
474.3
69.3
10.1
10.7
542
7.5
81
2.6
2.1
608
1021
GE51
10.9
156.4
26
18
14.8
12.3
724.9
69.3
10.7
9.2
575
7.3
82.3
2.8
2.2
571
872
GE52
13.5
187
26
21
58.9
30.5
6044.5
293.6
64.9
11.9
508
8.4
93.3
3.6
2.6
1640
4274
GE53
17.7
142.2
33
17
13.9
13.6
382.5
49.5
6.1
8.8
697
7.7
83.6
2.7
3.6
220
322
GE54
9
208.2
31
18
7.7
16.7
66.5
35.5
5
8.2
734
8.5
99.5
3.2
3.2
445
331
GE55
7.3
203.2
32
17
7.2
11.6
66.9
24.4
4.7
7.5
746
8.4
95.9
3.2
3.6
25
300
GE56
5.7
236.4
34
30
8.7
14.2
34.9
27.8
2.5
5.6
562
8.3
141.4
3.7
4.6
4
1
GE57
11
183.6
32
22
7.5
11
83
25.2
5.2
9.5
693
8.3
89.4
3
2.6
456
267
GE58
11
177.2
31
17
7.8
11.3
85.3
37.8
5.4
10.1
686
8.3
91.2
3.2
2.3
395
340
GE59
9.2
194.8
27
15
6.6
16
217.4
121.2
22.1
22.9
773
8.2
89.4
2.8
3.2
24611
237
GE60
8.1
208.7
33
17
7
12.2
73.4
21.7
4.6
7.4
770
8.4
97
3.3
3.2
9
314
GE61
12.2
174.8
19
14
7.4
9.4
135.8
32.7
11.9
11.3
503
6.9
62.6
2.2
1.5
203
1998
GE62
15.1
165.8
25
18
12.7
11.8
397
76.7
11.2
10.4
545
7.6
77.9
2.6
2.4
704
1469
GE63
11.6
158.2
23
18
8.7
10.5
209.7
40.9
8.1
11
522
7.9
83.2
2.6
2.3
300
1123
GE64
16.4
180.7
26
17
11.7
12.4
242
104.4
9.2
14.2
577
7.8
77.6
2.7
2.3
1315
1148
GE65
12.4
163.3
27
18
17.1
12.3
1042.8
64
13.2
9.1
530
7.6
82.7
2.6
2.7
337
1474
GE66
8.8
159.8
19
14
9
10
337.9
39.1
10.4
8.8
488
6.9
56.8
2
2.7
76
1007
GE67
10.6
169
27
20
14.1
12
456.6
64.3
9.7
8.8
490
7.5
83.2
2.6
3
150
1042
GE68
8.4
170.2
20
14
10.1
10.3
217
43.1
6.4
10.4
506
7.3
59.9
2.1
1.7
254
845
GE69
7.3
173.9
39
46
12
15.2
83.6
33.7
4.6
6
514
9.6
199.7
4.9
5.2
34
8
Maastricht-Jodenstraat (MAJO) glass samples
Joden 1
4.5
179.3
26.38
27.36
9.4
41.4
15637.3
4231.5
51.3
18.1
479
10.2
113.3
3.6
3.7
11459
118
Joden 2
7.4
145.9
27.23
15.95
10.9
37.7
15536.7
2119.5
43.5
10.6
615
7.3
81.4
2.8
4.7
6426
206
Joden 3
9
169.1
31.11
16.8
8.4
16.7
90.1
36.6
5.6
9.1
745
7.8
85.2
3.1
4.3
18692
165
Joden 4
8
155.1
50.54
18.22
11.7
22.6
91.6
36
9.9
8.2
649
9.4
97.4
3
3.5
52964
38
Joden 5
6.7
164
30.69
15.05
9.9
23.1
201.4
45.2
11.7
9
631
7
77.9
2.9
5.3
53150
84
Joden 6
11.6
138.8
32.97
19.4
8.7
22.1
302.2
74.7
30.1
17
714
8.9
91.3
3.5
4.2
78067
216
Joden 37
4.5
161.2
13.79
14.41
259.2
82.9
357
23
12.2
14.4
436
6.3
78.5
2.2
3.3
9097
3
Joden 38
4.9
189.7
15.62
12.67
204.7
69.3
823.9
17.6
9.1
6.6
541
6.2
73.8
2.2
2.6
81
6
Joden 39
4.4
163.1
15.09
14.03
311.2
89.9
1070.8
18
9
7.1
491
6.7
78.3
2.2
3.7
33
4
Joden 40
4.3
174.5
14.93
14.02
308.6
97
1059.1
19.3
12.7
7.4
491
6.2
76.3
2.1
4.1
55
4
Joden 41
8.4
166.3
21.8
16.27
35.4
48.2
20139.9
1286.1
65.9
12.5
521
7.9
66.8
2.3
1.7
2182
1513
Joden 42
9.3
153.8
20.55
12.56
32.4
47.9
20130.9
1804.3
72.1
12.9
486
7.7
63.1
2.2
1.9
2413
1648
Joden 43
9.9
158.2
30.69
15.57
13.9
74.5
31204.5
3655.4
53.3
10.4
691
7.8
84.4
3
4.9
8394
252
Joden 44
4.6
121.5
21.88
9.93
4.3
14.7
38.8
18.5
2.1
5.4
572
5
54.8
2
4.4
6
34
Joden 45
15.2
153.9
28.82
18.13
10.7
16.7
152.4
53.1
7.4
10.6
568
7.8
83.1
2.7
3.2
190
684
Joden 46
15.4
157.4
36.32
36.73
9.3
12.6
138.6
57.1
5.5
6.8
470
8.6
157.3
4
3.1
24
357
Joden 47
11.3
166
37.74
17.88
10.8
25.8
277.7
68.8
13.8
12.3
740
9.2
98.3
3.5
4.9
50418
247
Joden 48
10.1
170.8
33.58
16.66
8.9
18.8
1287.7
44.4
3.9
10.5
755
7.6
92.2
3.1
5.3
215
189
Joden 49
9.2
404.3
52.39
28.12
7.6
268.3
5825
2077.4
2.4
41.7
164
20.2
286.3
8.4
0.3
195
66
171
—
Cs
Ba
La
Ce
Pr
Nd
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
Hf
Ta
Pb
Th
U
0.27
327
7.7
13.4
1.7
6.9
1.5
0.4
1.2
0.2
1.2
0.3
0.7
0.1
0.7
0.1
2
0.2
25821
1.3
1.2
0.13
353
11.9
15
2.6
10.8
2.2
0.6
2.4
0.4
2.2
0.4
1.1
0.2
1.3
0.1
4.2
0.3
3265
1.8
1.5
0.09
385
9.9
14.3
2.3
9.9
2
0.5
1.8
0.3
1.8
0.4
1.1
0.2
1.1
0.1
3.4
0.2
125
1.5
1.3
0.08
386
10.7
14.8
2.4
10
2.2
0.6
1.8
0.3
1.7
0.4
1.1
0.1
1
0.2
3.4
0.2
647
1.5
1.3
0.08
464
10.9
16.2
2.6
11.4
2.5
0.6
2.4
0.3
2
0.4
1.1
0.2
1.3
0.2
4.3
0.3
922
1.8
1.4
0.08
396
10.1
14.7
2.4
9.7
1.9
0.6
2
0.3
1.8
0.4
1.1
0.2
1.1
0.2
3.6
0.2
192
1.6
1.3
0.2
321
7.8
13.7
1.8
7.4
1.6
0.4
1.3
0.2
1.3
0.2
0.8
0.1
0.6
0.1
2.2
0.2
18040
1.3
1.3
0.18
262
6.5
11
1.5
5.7
1.3
0.4
1.2
0.2
1
0.2
0.6
0.1
0.5
0.1
1.3
0.1
441
1.1
1
0.21
352
7.3
13.4
1.7
7
1.4
0.4
1.3
0.2
1.2
0.3
0.8
0.1
0.7
0.1
2.1
0.2
2036
1.3
1.2
1.2
0.18
360
7.1
13.4
1.7
7.3
1.4
0.4
1.3
0.2
1.2
0.2
0.8
0.1
0.6
0.1
2.1
0.2
3371
1.3
0.34
367
8.7
14.5
1.9
7.8
1.7
0.4
1.5
0.2
1.3
0.3
0.8
0.1
0.9
0.1
2.3
0.2
16358
1.6
1.2
0.12
439
7.4
12.9
1.8
7.2
1.5
0.3
1.5
0.2
1.4
0.3
0.8
0.1
0.7
0.1
2
0.2
1407
1.3
1.4
0.14
359
8.2
14.7
1.9
7.6
1.8
0.4
1.6
0.2
1.5
0.3
0.9
0.1
0.8
0.1
2.6
0.2
140
1.7
1.4
0.13
361
8.1
14
1.9
7.9
1.6
0.4
1.6
0.2
1.4
0.3
0.8
0.1
0.8
0.1
2.5
0.2
121
1.5
1.6
0.07
415
7.9
14.5
1.9
7.4
1.4
0.4
1.5
0.2
1.4
0.3
0.8
0.1
0.9
0.1
3.4
0.2
16
1.6
1.4
0.17
339
8.4
14.3
2
7.7
1.8
0.4
1.5
0.2
1.5
0.2
0.9
0.1
0.7
0.1
2.2
0.1
262
1.4
1.4
0.18
330
8.3
14.6
2
7.6
1.6
0.5
1.5
0.2
1.5
0.3
0.9
0.1
0.7
0.1
2.3
0.2
356
1.5
1.4
0.22
383
7.7
13.5
1.8
7.7
1.6
0.4
1.7
0.2
1.4
0.3
0.8
0.1
0.7
0.1
2.2
0.2
25043
1.3
1.4
0.11
368
8.3
14.7
1.9
8.3
1.7
0.4
1.9
0.2
1.4
0.3
0.8
0.1
0.9
0.1
2.3
0.2
113
1.5
1.6
0.19
276
6.6
12.4
1.5
6.4
1.3
0.4
1.3
0.2
1.1
0.2
0.7
0.1
0.6
0.1
1.5
0.1
498
1.2
1.1
0.24
367
7.6
13.7
1.7
7.2
1.5
0.4
1.4
0.2
1.3
0.3
0.7
0.1
0.6
0.1
1.9
0.2
1581
1.3
1.1
0.18
342
7.4
12.8
1.7
7.2
1.5
0.4
1.4
0.2
1.3
0.2
0.7
0.1
0.7
0.1
2.1
0.2
713
1.3
1.1
0.27
367
7.7
13.3
1.8
7.1
1.4
0.4
1.5
0.2
1.3
0.3
0.7
0.1
0.7
0.1
1.8
0.1
1361
1.3
1.2
1.3
0.22
425
7.6
12.9
1.7
7.2
1.5
0.4
1.4
0.2
1.2
0.3
0.9
0.1
0.7
0.1
2
0.2
3237
1.3
0.16
348
6.9
12.3
1.6
6.2
1.4
0.4
1.3
0.2
1.2
0.3
0.7
0.1
0.6
0.1
1.4
0.1
803
1.1
1
0.19
387
7.2
13.1
1.7
6.6
1.4
0.4
1.3
0.2
1.2
0.2
0.7
0.1
0.7
0.1
2.1
0.2
1267
1.2
1.1
0.16
303
6.9
12.6
1.6
6.5
1.3
0.4
1.2
0.2
1.3
0.2
0.7
0.1
0.7
0.1
1.6
0.1
700
1.1
1.1
0.07
453
9.1
16.2
2.2
9.1
1.9
0.5
1.9
0.2
1.7
0.3
1.1
0.1
0.9
0.1
4.8
0.3
211
2
1.5
0.26
386
9.9
16.9
2.3
9.5
1.7
0.4
1.8
0.3
1.6
0.3
1
0.1
0.9
0.1
2.8
0.2
88534
1.9
1
0.26
274
7.5
14
1.8
7
1.5
0.4
1.3
0.2
1.3
0.3
0.8
0.1
0.7
0.1
2.1
0.2
61933
1.4
1.1
0.24
339
8
14.1
1.8
7.8
1.6
0.5
1.5
0.2
1.4
0.3
0.9
0.1
0.8
0.1
1.9
0.2
2319
1.4
1.2
0.31
350
9.3
13.7
2.1
8.4
1.7
0.5
1.8
0.3
1.7
0.3
0.9
0.1
0.9
0.1
2.3
0.2
47497
1.5
1.1
0.46
268
7.1
12.7
1.6
6.6
1.3
0.4
1.4
0.2
1.2
0.2
0.7
0.1
0.6
0.1
2
0.2
316647
1.3
1.3
1.05
239
9.5
17.5
2.2
8.9
1.8
0.4
1.8
0.3
1.5
0.3
0.9
0.1
0.9
0.1
2.4
0.2
780143
2.2
1.1
0.06
151
6.1
11.1
1.4
5.7
1.2
0.4
1.3
0.2
1.1
0.2
0.7
0.1
0.6
0.1
2
0.1
5767
1.1
1.1
0.08
159
5.9
10.8
1.5
5.8
1.3
0.3
1.2
0.2
1
0.2
0.6
0.1
0.6
0.1
1.7
0.2
3529
1
1.2
0.1
155
6.3
11.3
1.4
5.9
1.2
0.3
1.1
0.2
1.2
0.2
0.7
0.1
0.7
0.1
2.1
0.1
5276
1.1
1
0.07
155
6.3
11.3
1.4
6.3
1.2
0.4
1
0.2
1
0.3
0.8
0.1
0.6
0.1
1.9
0.1
5394
1.1
1
0.39
298
7.8
13.1
1.7
6.9
1.4
0.5
1.3
0.2
1.3
0.2
0.9
0.1
0.6
0.1
1.5
0.2
75113
1.3
1
0.54
270
7.3
13.1
1.7
7.3
1.4
0.4
1.3
0.2
1.3
0.3
0.8
0.1
0.8
0.1
1.6
0.1
84741
1.4
1.2
0.22
297
7.8
14.2
1.8
7.8
1.7
0.4
1.6
0.2
1.2
0.3
0.8
0.1
0.8
0.1
2.2
0.2
7289
1.3
1.1
0.08
217
5
9.1
1.2
4.7
1
0.2
1
0.1
0.9
0.2
0.6
0.1
0.4
0.1
1.4
0.1
40
0.9
0.8
0.24
346
7.5
13.2
1.8
7.3
1.8
0.4
1.4
0.2
1.3
0.2
0.7
0.1
0.7
0.1
2.1
0.2
338
1.3
1.1
0.08
1529
8.1
14.4
2
7.2
1.4
0.6
1.5
0.2
1.6
0.3
1
0.1
1
0.1
3.9
0.3
178
1.6
1.2
0.55
314
9.6
15.6
2.2
9.1
1.8
0.4
1.9
0.3
1.6
0.3
0.9
0.1
0.9
0.1
2.4
0.2
407431
1.7
1.2
0.18
343
8
14.2
2
7.7
1.6
0.4
1.6
0.2
1.3
0.2
0.8
0.1
0.7
0.1
2.3
0.2
835
1.5
1.2
1.69
234
19.9
38.5
4.3
16.3
3.6
0.7
3.6
0.5
3.8
0.7
2.5
0.3
1.9
0.4
7.9
0.6
297
5.9
2
172
—
Appendix III trace element chemical compositions of samples analysed by LA-ICP-MS
Trace
elements
Li
B
V
Cr
Co
Ni
Cu
Zn
As
Rb
Sr
Y
Zr
Nb
Mo
Sn
Sb
Joden 50
5.1
138.1
41.79
18.01
10.6
22.8
94.2
71.2
12.9
72.6
482
7.8
85.6
2.9
9.3
124317
122
Joden 51
6.8
139.1
25.17
13.94
12
38.4
7657.8
1092
29.2
8.4
608
5.8
65.9
2.4
4.6
6446
188
Joden 52
7.2
149.9
28.97
14.81
12.1
75.3
42945.3
5559.9
74.4
8.6
666
6.9
84.1
2.8
5.1
1830
159
Joden 53
4
160.5
17.07
15.45
178.8
47.2
805.3
22.5
7.6
6.7
365
5.9
84.2
2.5
2.5
3897
3
Joden 54
7.9
143.4
34.27
15.11
24.7
42.3
9609.1
1721.7
32.7
10
548
7
70.7
2.7
3.3
11562
170
Joden 55
8.1
88.9
21.38
13.91
9.3
68.3
5565.2
6530.8
85.4
10.9
426
5.5
58.4
2.3
2.7
10115
92
Joden 56
6.6
114.8
23.68
12.84
6.2
15.2
126
35
10
8.6
561
6.1
65.8
2.4
4
30841
169
Joden 57
6
159.1
25.4
14.58
20.7
88.3
26978.6
1474.6
43.2
8.6
602
7.1
74.6
2.7
4.4
5930
110
Joden 58
5.6
165.1
32.57
14.07
9.3
21.5
57.6
38.4
9.7
7.9
660
6.5
70
2.5
5.5
41978
60
Joden 59
7.5
157.7
31.1
14.33
9.5
16.9
112.2
36.8
7.7
9.1
659
7.2
77.2
2.6
4.3
27321
212
Joden 60
0.1
237
0.34
29.13
1.3
14.9
82
68.6
0.9
0.2
7
0.2
0.7
-
-
376
11
Joden 61
12.5
195.2
18.66
11.61
378.8
22.6
2057.6
59.8
43.4
13.1
441
6.4
60.7
2.2
2.1
124
11536
Joden 62
11.1
160.3
32.81
17.08
7.1
16.1
51.1
29
5.1
8.7
788
7.9
100.7
3.2
5.4
10
99
Joden 63
6.3
166.7
30.21
16.38
7.5
16.1
7.3
26.6
3.3
10.8
860
7.5
84.6
2.9
5.6
684
179
Joden 64
5.2
99.1
21.46
12.13
4.6
15.1
121.4
32.3
17.1
8.5
401
5
59.9
2.3
3.3
39547
98
Joden 65
5.6
93.9
18.73
10.93
5.2
10.9
136.2
36.5
7.7
7.1
481
4.7
51.3
1.9
3.3
23702
121
Joden 66
7.3
161
33.09
17.2
14.2
53.5
14623.8
2929.4
46.4
10.4
655
7.2
81.4
3
4.9
14412
113
Joden 67
10.6
150.4
31.43
16.77
13.5
65.5
26901
4409.2
60.3
11.8
648
8.1
84.9
2.8
4.6
1585
215
Joden 68
3.6
133.6
12.57
32.33
15.6
9.5
154.7
18.7
2.1
9.1
389
6.1
45.5
1.5
1.5
34
592
Joden 69
7.2
105.3
43.54
15.87
10.1
23.2
183.8
39.1
11.4
8.5
473
7.8
67.1
2.5
2
104859
109
Joden 70
8.6
137.9
30.49
15.75
8.7
19.7
182.4
52.5
6.7
9.6
631
7.3
78.6
2.8
4
24800
153
Joden 71
6.8
100.4
40.71
12.9
8.7
21.7
238.1
36
11.6
8.9
450
7.2
62.3
2.3
2.4
39902
86
Joden 72
6.2
136
24.71
13.51
8.2
18.2
116.7
37.4
10.6
9.1
496
6.1
65.6
2.4
4.4
37442
70
Joden 73
6.5
154
28.16
15.2
19.3
96.5
28045.7
1778.3
50.9
7.9
737
6.9
86.6
2.7
5.3
18725
183
Joden 74
6
149.3
26.78
15.54
19.3
89.8
27398.5
1726.3
47.4
7.5
717
6.8
80.7
2.7
5
5641
154
Joden 75
8.5
136.6
34.06
14.53
12.2
62.7
28328.3
3250.1
63.5
9.3
582
7.2
70.6
2.4
3.2
50978
294
Joden 76
6
150.4
27.77
14.63
18.9
90.7
26781
1682
46
7.5
719
7.2
80.1
2.6
5.2
5800
154
Wijnaldum glass samples
WIJ1
10.1
163.4
28.18
15.89
9.8
11.2
230.1
54.6
7.1
9.6
631
8.6
79.7
2.5
2.1
402
626
WIJ2
19
174.3
27.85
18.31
11.3
11.7
576.3
70.2
8.8
10.4
598
8.3
89.6
2.7
2.3
542
989
WIJ3
15
168.2
28.67
17.59
12.1
12.4
573.4
144.8
9.9
11.5
625
8.9
86.4
3
2
997
803
WIJ4
2
104.8
4.39
4.29
1.2
7.9
224.2
19.2
45
4
215
3.3
24
0.8
0.1
23001
3049
WIJ5
7.6
97.8
32
14.76
9.6
23.1
155.4
42.3
12.6
15.2
363
9.2
59.6
2.4
1
43112
117
WIJ6
6.2
95.2
28.12
17.75
9.6
20.9
214.4
32
23.6
7.7
343
8.2
71.5
3.2
1.1
59853
112
173
WIJ 7
8.4
93.3
32.74
18.13
18.1
35.6
199.4
83.6
8
13.2
565
7.8
74.4
2.8
4.3
43807
WIJ 8
6.8
49.2
12.74
12.58
2.2
12.7
190.3
15.2
2.6
11.1
240
5.2
29.1
1.5
0.4
37252
1
WIJ 9
8
41
10.88
11.2
2.1
13.4
263.5
29.2
4.3
12
212
4.7
25.6
1.6
0.3
46669
1
WIJ 10
9.2
90.9
21.33
26.39
8.7
8.3
375.9
56.6
3.2
8.2
235
7.1
184.7
3.9
0.5
82
156
WIJ 11
55.6
145.4
20.82
17
8.8
10.7
698.3
181.8
14.6
16.3
440
7
75.2
2.5
1.7
221
1802
WIJ 12
8.1
164.2
23.24
21
6.5
9
123
38.3
13.3
9.8
458
7.5
95.5
2.6
2.2
35
1679
WIJ 13
8.8
163.8
22.45
17
6.5
9.1
74.6
29.9
11.3
8.9
485
7.2
73.6
2
2
19
2258
WIJ 14
8
60.9
17.15
17
4.8
29.6
290.8
54.8
18
15.8
136
4.7
57.6
3.4
0.3
63145
53
WIJ 15
4
98.5
20.28
26
822.5
25.1
934.3
2256.1
11.2
4.2
177
6.6
169.1
3.9
2.6
97
7
WIJ 16
6.5
76.7
9.91
9
7.6
6.8
1172.2
42.8
9.7
8.3
283
4.7
38.9
1.2
0.9
359
828
WIJ 17
20.2
76.1
50.3
22
51.6
118.8
19996.3
2905.8
89.5
46.2
287
14
197.5
6.7
3.6
2095
145
WIJ 18
13.1
151.6
52.76
22
25.4
48.4
18941.6
2536.3
65.1
14.4
528
10.9
99.5
3.5
3.4
7748
551
173
—
Cs
Ba
La
Ce
Pr
Nd
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
Hf
Ta
Pb
Th
U
0.8
304
8
12.9
1.9
7.7
1.4
0.4
1.5
0.2
1.3
0.3
0.8
0.1
0.6
0.1
2
0.2
23851
1.3
0.16
238
6.2
11.2
1.4
5.8
1.1
0.3
1.1
0.2
1
0.2
0.6
0.1
0.6
0.1
1.6
0.2
78875
1.1
0.9
1
0.15
280
6.9
12.3
1.6
6
1.6
0.3
1.3
0.2
1.2
0.2
0.7
0.1
0.6
0.1
2
0.1
3115
1.3
1.1
0.08
159
6.4
12
1.6
5.8
1.2
0.3
1.2
0.2
1.1
0.2
0.6
0.1
0.7
0.1
2.1
0.2
4798
1.1
1
0.23
276
7.6
12.8
1.7
6.9
1.3
0.4
1.3
0.2
1.2
0.3
0.7
0.1
0.7
0.1
1.7
0.1
71856
1.3
1.1
0.57
259
7.1
11.2
1.4
6
1.2
0.2
0.9
0.2
1
0.2
0.5
0.1
0.6
0.1
1.5
0.1
354763
1.3
1.7
0.39
186
6.3
11.2
1.5
6.6
1.2
0.3
1.1
0.2
1
0.2
0.7
0.1
0.5
0.1
1.6
0.1
277719
1.3
0.9
0.18
275
7.1
12.7
1.7
6.9
1.6
0.4
1.6
0.2
1.2
0.3
0.6
0.1
0.8
0.1
1.8
0.2
37801
1.3
1
0.18
317
6.8
11.4
1.5
6.2
1.3
0.3
1.3
0.2
1.2
0.2
0.8
0.1
0.7
0.1
1.7
0.1
22283
1.1
1.2
0.19
270
6.8
12.4
1.7
6.7
1.3
0.4
1.4
0.2
1.1
0.3
0.7
0.1
0.7
0.1
2
0.2
28303
1.2
1.3
0.11
9
0.3
2.4
0
0.1
0
0
0.1
0
0
0
0
0
0
0
0.1
0
2484
0
0
0.41
261
6.9
11.7
1.5
5.8
1.3
0.3
1.2
0.2
1.1
0.2
0.6
0.1
0.6
0.1
1.5
0.1
2011
1.2
1.1
0.13
293
7.7
14.4
1.8
7.8
1.5
0.5
1.6
0.2
1.3
0.3
0.9
0.1
0.8
0.1
2.4
0.2
88
1.4
1.2
0.1
258
7.3
13.3
1.7
7.1
1.4
0.4
1.5
0.2
1.3
0.3
0.7
0.1
0.7
0.1
2.2
0.2
205
1.3
1.2
0.64
167
5.2
8.8
1.2
4.6
1.1
0.2
0.8
0.1
0.9
0.2
0.5
0.1
0.5
0.1
1.5
0.2
468522
1.1
0.7
0.34
150
4.9
8.9
1.2
4.6
0.9
0.2
0.9
0.1
0.8
0.2
0.5
0.1
0.5
0.1
1.4
0.1
221741
0.9
0.7
0.3
291
7.6
14.2
1.8
7.3
1.5
0.4
1.6
0.2
1.4
0.3
0.7
0.1
0.7
0.1
2.1
0.5
123336
1.5
1.3
0.24
318
7.8
15.2
1.9
7.7
1.7
0.4
1.5
0.2
1.4
0.3
0.8
0.1
0.8
0.1
2.1
0.2
38984
1.5
1.1
0.15
224
5.9
10.6
1.4
5.1
0.9
0.3
1.1
0.2
0.9
0.2
0.6
0.1
0.5
0.1
1.2
0.1
182
1
0.9
0.6
261
7.8
12.3
1.8
7.4
1.4
0.3
1.5
0.2
1.3
0.3
0.8
0.1
0.7
0.1
1.8
0.2
423852
1.3
0.9
0.34
249
7.6
13.4
1.7
7.3
1.5
0.3
1.4
0.2
1.4
0.2
0.7
0.1
0.6
0.1
2
0.2
206637
1.3
0.9
0.43
265
7.1
10.8
1.6
6.3
1.1
0.3
1.5
0.2
1.3
0.3
0.8
0.1
0.6
0.1
1.6
0.1
288546
1.1
0.8
0.44
221
6.5
11.8
1.4
6.1
1.3
0.3
1.2
0.2
1.1
0.2
0.6
0.1
0.5
0.1
1.5
0.1
294751
1.2
1.1
0.22
254
7.2
13.3
1.7
6.9
1.5
0.3
1.3
0.2
1.4
0.2
0.7
0.1
0.7
0.1
2.2
0.2
40403
1.3
1.1
0.18
252
7
12.5
1.6
6.5
1.2
0.3
1.3
0.2
1.1
0.2
0.7
0.1
0.6
0.1
1.9
0.2
38669
1.3
1
0.5
283
7.9
12.9
1.8
7.3
1.5
0.4
1.5
0.2
1.4
0.3
0.7
0.1
0.8
0.1
1.7
0.2
59580
1.3
0.9
0.24
249
7.2
12.7
1.6
6.6
1.4
0.4
1.2
0.2
1.2
0.2
0.7
0.1
0.7
0.1
2
0.2
38566
1.3
1.1
1.2
0.14
320
8.1
13.8
1.9
7.6
1.7
0.5
1.6
0.2
1.4
0.3
0.7
0.1
0.9
0.1
2
0.2
1149
1.3
0.14
399
8.2
13.9
2
7.8
1.5
0.4
1.5
0.2
1.3
0.3
0.8
0.1
0.9
0.1
2.3
0.2
1409
1.4
1.1
0.21
357
8.4
14.8
2
8.3
1.7
0.5
1.4
0.2
1.3
0.2
0.8
0.1
0.8
0.1
2.4
0.2
2525
1.4
1.2
0.1
85
3.1
5.4
0.7
3
0.8
0.2
0.6
0.1
0.5
0.1
0.3
0.1
0.3
0
0.7
0
328951
0.4
0.5
0.83
177
9.6
15.4
2.3
8.7
1.8
0.4
2
0.3
1.6
0.3
1
0.1
0.9
0.1
1.6
0.2
496383
1.7
0.9
0.29
150
7.8
12.2
1.7
7.1
1.5
0.3
1.2
0.2
1.4
0.3
0.9
0.1
0.9
0.1
1.9
0.2
477936
1.6
0.9
0.66
222
8.8
13.4
2.1
8
1.5
0.4
1.6
0.2
1.5
0.3
0.8
0.1
0.8
0.1
1.9
0.2
508491
1.6
1
0.62
156
5.3
10.3
1.2
5.2
1
0.3
0.9
0.1
0.9
0.2
0.5
0.1
0.4
0
0.8
0.1
440657
1.1
1
0.68
139
4.9
9.8
1.2
4.6
1
0.3
0.9
0.1
0.7
0.2
0.5
0.1
0.4
0.1
0.7
0.1
545641
1.2
0.5
0.1
199
7.3
13.9
1.8
7
1.6
0.4
1.3
0.2
1.2
0.2
0.7
0.1
0.7
0.1
4.5
0.3
584
1.6
1.1
0.25
316
7.4
13
1.6
7
1.3
0.3
1.3
0.2
1.2
0.3
0.7
0.1
0.6
0.1
1.9
0.1
1698
1.3
0.9
0.16
397
7.3
13
1.7
6.8
1.5
0.4
1.2
0.2
1.2
0.2
0.7
0.1
0.6
0.1
2.4
0.1
725
1.3
1
0.13
348
6.7
12
1.6
6.6
1.2
0.4
1.4
0.2
1.2
0.2
0.7
0.1
0.7
0.1
1.9
0.1
241
1.1
1
0.82
74
6.5
12
1.5
5.1
1.2
0.2
0.9
0.1
0.9
0.2
0.5
0.1
0.4
0.1
1.6
1
707917
2.2
0.7
0.09
169
6.8
13.1
1.7
6.8
1.1
0.3
1.4
0.2
1
0.2
0.7
0.1
0.7
0.1
3.9
0.3
354
1.4
1.2
0.18
163
4.2
7.9
1
3.7
0.8
0.2
0.8
0.1
0.6
0.2
0.5
0.1
0.4
0.1
1
0.1
1445
0.8
0.6
2.34
405
17.5
61.5
4.3
16.3
3.1
0.5
2.8
0.4
2.6
0.5
1.4
0.2
1.5
0.2
5.4
0.5
270987
7.3
2
0.48
462
11.1
17.4
2.7
11.5
2.1
0.5
1.9
0.3
1.9
0.4
1.2
0.2
1.2
0.2
2.7
0.2
73142
2.1
1.2
174
—
Appendix III trace element chemical compositions of samples analysed by LA-ICP-MS
Trace
elements
Li
B
V
Cr
Co
Ni
Cu
Zn
As
Rb
Sr
Y
Zr
Nb
Mo
Sn
Sb
WIJ 19
9.6
155.5
28.19
17
23.8
38.2
8869.3
2160.2
42.2
12.2
487
15
85.6
2.9
0.9
12300
59
WIJ 20
8.1
180.2
75.45
19
28.4
44.1
7731.9
711.9
34.9
11.1
621
13.2
99.1
3.4
2.7
7222
68
WIJ 21
6.3
157.7
61.66
17
20.2
35.3
112.1
51.2
15.8
8.2
638
12.4
91.2
3.2
2.7
20435
56
WIJ 22
5.7
168.8
21.93
12
8.6
12.9
206.2
56.2
15.6
8.5
449
6.9
53.1
1.9
1.2
32949
1976
WIJ 23
5.8
165.5
27.84
21
9.5
23.3
46.6
29.1
5.5
7.5
616
7.1
101.9
3.2
1.3
120874
95
WIJ 24
7.4
139.9
45.88
16
20.5
43.1
289.3
63.5
22.1
9.7
602
10.6
89.9
3.2
4.2
33842
130
WIJ 25
8.7
157.5
45.22
15
23.5
35.1
97.9
50.5
17.2
8.6
670
10.1
81.9
2.7
3.6
37943
248
WIJ 26
7.1
185.7
41.15
15
18.4
34
94.2
46
12.4
7.7
684
10.7
87.8
2.9
2.9
31212
63
WIJ 27
9.3
162.9
45.98
20
16.1
46.2
12288.1
2793
40.3
17.2
564
17.3
90
3.5
1.8
1600
157
WIJ 28
8.4
150
40.79
18
14.8
64.6
20931.5
11130.1
68.3
12.9
593
13.4
92.4
3.4
1.8
6040
265
15816
WIJ 29
6.1
183.4
19.29
14
23
86.6
111970.5
1077.6
159.6
7.2
805
5.5
57.2
3.1
2.2
10200
WIJ 30
5.7
165.5
29.89
21
8.7
53.2
21945.1
8389.6
47.7
7.3
502
7.7
103.3
3
1.4
822
72
WIJ 31
8.1
175.3
41.19
18
22.3
45.1
11701
8078.2
134.4
17.2
544
15.6
86.9
3.5
1.3
984
142
WIJ 32
6.8
120.9
19.43
16
6.3
17.5
124.6
39.6
23.4
11.4
325
6.9
77
2.6
0.6
39588
66
WIJ 33
8.2
77.1
17.88
18
5.9
24.2
4382.9
1309.7
97.2
11.4
204
6
76.7
2.6
0.5
67882
36
WIJ 34
34.9
21.4
79.96
6
88.4
30.5
2231.8
177.9
51.9
79.4
49
26
187
9.3
1.3
1548
74
WIJ 35
22.9
70.6
10.21
39
55.5
24.6
570.1
33
4.8
13.8
356
4.3
58.8
1.5
1.7
18
0
WIJ 36
6.8
182.4
43.01
17
7.9
23.1
66.5
31.4
8.3
7.1
800
9.5
86.5
3.2
5.5
7
170
WIJ 37
20.4
68.3
10.25
21
3.9
13
22.1
152.4
1.4
13.2
370
3.1
32.5
1.2
2
1
0
WIJ 38
5.9
159.2
16.92
13
3.6
5.6
16.6
11.8
2.9
6.1
536
6.8
73.7
2.5
0.7
2
51
WIJ 39
6.2
149.3
36.77
19
125.2
248.9
12078.9
6260.3
277.6
8.7
654
7.8
89.1
3.5
6
21208
298
WIJ 40
4.9
202.6
17.46
13
3.9
7.7
80
27
25
11.5
419
5.7
77.7
2.4
0.8
483
10121
WIJ 41
10.7
164.8
34.66
25
20
16.6
1208.3
126.5
19.2
9.7
466
8.4
106.8
3.1
2.2
513
1650
WIJ 42
3.5
48.3
6.99
11
3
22
21236.9
1192.3
37.7
7.1
480
7.1
40.7
1.6
0.1
5913
50
7.6
274.8
29.4
20.4
10.7
450
6.5
66.9
2.1
1.7
51
3575
Wijk bij Duurstede (Dorestad) glass samples
LM 25
5.9
203.3
18
13
6.1
LM 26
17.1
165
23
18
28.6
12.7
1143.8
107.5
11.4
20.6
467
7.3
75.3
2.5
2.1
295
1785
LM 27
12.8
144.2
19
14
13
25.7
33870.8
7164.7
202.2
17.6
408
6.2
66.1
2.2
1.8
558
2057
1734
LM 28
8.4
149.2
19
15
8.7
8.5
412.8
97.5
12.4
10.4
425
6.3
70.1
2.2
1.4
96
LM 29
12.5
124.9
20
17
16.9
12.4
2112
104.1
12.7
14.1
376
6.5
86.3
2.6
1.6
275
975
LM 30
18.7
130.4
17
15
580.5
25.6
1833.1
883.4
24.2
18.7
430
6.1
73.2
2.2
2.5
157
5703
LM 31
13.6
148
19
14
28
28.2
55184.7
15768.8
605.7
28.3
398
6.9
58.2
2.2
1.6
333
1859
LM 33
36
136.8
20
17
20.2
21.7
25129.6
2969.8
111.6
17.2
318
6.1
79.8
2.8
1.3
1994
1130
DOR 53
24
159.1
17.7
13.82
13
9.9
1708.3
90.6
14.1
14.1
469
6.6
59.5
2
1.7
230
2737
DOR 61
9.6
466.3
2.48
20.05
27.1
2.7
60.8
3693.2
58
70.3
25
3.6
138.9
1.2
0.7
18
495
DOR 66
15.2
145.1
18.73
15.28
22.3
11
1308.1
82.8
13
13
461
7
63.4
2.2
1.9
906
2122
DOR 90
16.8
153.8
22
17
17.4
13.6
3662.3
378.5
25
17.9
434
6.8
77.6
2.4
2
417
1352
DOR 91
5.6
199.8
17
13
6.6
8.6
199.6
30.3
22.6
10.1
450
6.3
64.9
2.1
1.8
57
3610
DOR 95
51
85.8
88.14
83.52
24.5
69
26.5
81.6
6.8
142.7
189
28.1
180.5
15.5
1.3
3
0
DOR 100
28.3
159.2
21
16
27.1
14.8
1789.2
115.7
13.7
18.3
463
7.5
76.6
2.6
1.9
977
2214
DOR 101
10.5
185.5
21
14
29.4
13.5
3542.6
101.6
28.6
15.3
470
6.6
65.2
2.4
1.8
579
4680
DOR 102
17.6
171.8
20
14
33.1
14.7
1332.6
92.1
18.1
16.6
465
6.7
62.4
2.3
1.9
674
2941
DOR 103
14.6
230.2
13
10
8.3
12
69.1
289
7.9
144.6
519
7.6
77.3
3.5
0.9
286
18
DOR 104
16.5
154.1
23
15
18.9
14
1765
691.8
14
16.7
482
7.4
68.2
2.6
2
495
1804
DOR 105
15.8
138.3
19
15
9.5
15.6
11995.6
3754
52.1
14.6
412
6.6
65.1
2.3
1.5
1614
1542
DOR 106
17.3
156.5
16
12
152
14.5
1005.1
62.5
9.1
11.5
482
6.9
51.7
1.9
2.6
1261
1451
175
—
Cs
Ba
La
Ce
Pr
Nd
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
Hf
Ta
Pb
Th
U
0.38
197
12.5
20.1
3
11.5
2.3
0.6
2.5
0.4
2.5
0.5
1.4
0.2
1.4
0.2
2
0.2
76928
1.6
1.3
0.23
247
13.6
16.3
3.2
13.3
2.6
0.7
2.9
0.4
2.3
0.5
1.4
0.2
1.3
0.2
2.3
0.2
30393
1.6
1.3
0.23
245
13.4
15
2.9
12.3
2.6
0.6
2.5
0.4
2.3
0.4
1.5
0.2
1.3
0.2
2.2
0.2
18301
1.5
1.3
0.18
215
7
11
1.7
6.8
1.6
0.3
1.1
0.2
1.2
0.2
0.7
0.1
0.6
0.1
1.4
0.1
17846
1
1.1
0.19
299
7.5
13.5
1.8
6.8
1.1
0.4
1.1
0.2
1.2
0.3
0.7
0.1
0.7
0.1
2.3
0.2
15019
1.4
1.3
0.38
306
11.4
14.6
2.4
9.6
1.9
0.5
2.2
0.3
1.8
0.4
1
0.1
1.2
0.2
2.3
0.2
269522
1.6
1.3
0.2
279
11
14
2.3
10.2
1.7
0.5
2.1
0.3
1.8
0.4
1.1
0.1
0.8
0.1
2.1
0.2
51806
1.3
1.3
0.11
288
11.3
14.7
2.5
10.9
2.2
0.6
2.2
0.3
1.9
0.4
1.2
0.2
1.1
0.1
2.1
0.2
38168
1.5
1.6
1.6
0.8
272
15.3
24.8
3.7
15.2
3.2
0.9
3.5
0.5
2.9
0.6
1.7
0.2
1.5
0.2
2.2
0.2
45749
2.3
0.43
296
11.7
18.3
2.7
11.8
2.6
0.6
2.6
0.4
2.4
0.4
1.3
0.2
1.1
0.2
2.4
0.2
58770
1.9
1.5
0.07
230
6.1
11.8
1.4
6.5
1.2
0.3
1.1
0.2
1.1
0.2
0.5
0.1
0.6
0.1
1.5
0.2
16703
1.4
0.7
0.17
325
7.2
12.9
1.7
7
1.3
0.3
1.3
0.2
1.3
0.3
0.8
0.1
0.9
0.1
2.7
0.2
43264
1.4
1.4
0.76
253
14
23.2
3.4
14.4
3.2
0.6
3.1
0.5
2.7
0.5
1.5
0.2
1.4
0.2
2.1
0.2
46027
2.1
1.5
0.52
143
7
12.9
1.6
7.4
1.5
0.4
1.4
0.2
1.3
0.3
0.7
0.1
0.7
0.1
2
0.2
416164
1.6
1
0.78
110
6.4
12.3
1.6
6.1
1.6
0.3
1.3
0.2
1
0.2
0.6
0.1
0.6
0.1
2.1
0.2
653466
1.6
0.8
5.6
1561
41.8
234.4
10.9
39.7
7.7
1.9
6.9
1
5.7
1
2.8
0.4
2.3
0.5
4.4
0.8
429979
12
3.3
0.13
125
4.5
7.7
1.1
3.7
0.4
0.1
0.7
0.1
0.7
0.1
0.5
0.1
0.4
0.1
1.3
0.1
344
1.2
0.5
0.1
263
9.4
14.1
2.2
9
1.9
0.5
2.1
0.3
1.7
0.4
1.2
0.1
0.8
0.1
2.2
0.2
74
1.5
1.2
0.23
121
3.2
6
0.7
3
0.4
0.1
0.7
0.1
0.4
0.1
0.4
0
0.4
0
1
0.1
2
0.9
0.6
0.03
186
6.9
11.7
1.5
7
1.2
0.3
1.2
0.2
1.1
0.2
0.6
0.1
0.6
0.1
1.8
0.1
23
1.2
1.2
0.14
325
7.6
13.8
2
7.1
1.4
0.5
1.3
0.2
1.6
0.2
0.7
0.1
0.7
0.1
2.1
0.3
71968
1.4
1.1
0.25
159
6.4
11.7
1.4
6.4
1.3
0.3
1
0.1
1.1
0.2
0.6
0.1
0.6
0.1
1.9
0.1
71267
1.5
1.1
0.22
362
8.5
14
2
8.3
1.8
0.4
1.5
0.2
1.3
0.3
1
0.1
0.7
0.1
2.4
0.2
4933
1.3
1.1
0.11
232
6.3
13.2
1.5
6.3
1.3
0.4
1.3
0.1
1.1
0.3
0.7
0.1
0.5
0.1
1.1
0.1
17115
0.8
0.6
0.16
264
6.8
11.8
1.5
6.2
1.2
0.3
1
0.2
1.1
0.2
0.7
0.1
0.5
0.1
1.7
0.1
1033
1.2
1.1
0.45
384
8.1
14.3
1.8
7.4
1.5
0.4
1.2
0.2
1.2
0.2
0.7
0.1
0.7
0.1
2
0.2
2119
1.3
1.1
0.39
284
6.5
12
1.6
6.3
1.2
0.4
1.1
0.2
1.1
0.2
0.6
0.1
0.5
0.1
1.8
0.1
3722
1.1
1
0.21
268
6.3
11.6
1.5
6.2
1.1
0.3
1.2
0.2
1.1
0.2
0.6
0.1
0.6
0.1
1.8
0.2
959
1.1
1
0.27
274
6.9
13
1.7
6.4
1.6
0.4
1.2
0.2
1.1
0.2
0.6
0.1
0.6
0.1
2.2
0.1
2182
1.3
1
0.31
237
6.7
12.4
1.5
6.1
1.3
0.3
1.2
0.2
1.2
0.2
0.6
0.1
0.6
0.1
1.6
0.1
1685
1.2
1
0.61
257
7.2
13.3
1.6
6.9
1.3
0.4
1.4
0.2
1.2
0.2
0.6
0.1
0.6
0.1
1.7
0.1
4782
1.3
1
2
260
6.7
13.2
1.6
6.3
1.3
0.4
1
0.2
1.1
0.2
0.7
0.1
0.5
0.1
2
0.2
4907
1.3
1.1
0.35
300
6.6
11.9
1.6
6.1
1.3
0.3
1.3
0.2
1.2
0.2
0.7
0.1
0.6
0.1
1.6
0.1
1775
1.2
1
4.6
0.29
114
3.4
6.9
0.8
2.8
0.5
0.1
0.6
0.1
0.5
0.1
0.4
0.1
0.4
0.1
3.8
0.1
5355
1.8
0.27
289
6.9
12.5
1.7
6.5
1.1
0.4
1.2
0.2
1.2
0.2
0.7
0.1
0.8
0.1
1.6
0.1
7806
1.3
1
0.31
336
7.3
13.4
1.7
6.8
1.2
0.4
1.4
0.2
1.1
0.2
0.7
0.1
0.6
0.1
2
0.2
2824
1.3
1
0.2
267
6.8
11.9
1.5
6
1.2
0.3
1.1
0.2
1
0.2
0.6
0.1
0.6
0.1
1.5
0.1
1130
1.1
1
10.25
603
34.5
69.7
8.3
30.7
6.3
1.5
5.8
0.8
5.1
1
2.7
0.4
2.6
0.4
4.9
1.1
19
11.2
2.8
0.4
328
7.6
14
1.8
7.4
1.5
0.4
1.3
0.2
1.3
0.2
0.8
0.1
0.7
0.1
2.1
0.1
6199
1.4
1.1
0.47
337
7.7
14.6
1.7
7.4
1.5
0.4
1.6
0.2
1.3
0.2
0.7
0.1
0.6
0.1
1.5
0.1
6311
1.4
1
0.35
305
7.1
12.9
1.6
7.2
1.4
0.4
1.2
0.2
1.1
0.2
0.7
0.1
0.5
0.1
1.6
0.1
5839
1.3
1
3.54
1506
71.5
93.2
8.8
27.5
2.6
0.5
2.2
0.3
1.4
0.2
0.7
0.1
0.6
0.1
2.2
0.3
1192
2.8
1
1.1
0.37
324
7.8
13.9
1.8
7.6
1.3
0.4
1.3
0.2
1.1
0.2
0.8
0.1
0.7
0.1
1.8
0.1
3801
1.4
0.27
310
6.7
12.9
1.7
6.6
1.4
0.3
1.2
0.2
1.2
0.3
0.7
0.1
0.6
0.1
1.7
0.2
11967
1.3
1
0.2
266
6.7
12.2
1.6
6.2
1.3
0.4
1.3
0.2
1.2
0.2
0.6
0.1
0.6
0.1
1.3
0.1
9092
1
1.1
176
—
Appendix III trace element chemical compositions of samples analysed by LA-ICP-MS
Trace
elements
Li
B
V
Cr
Co
Ni
Cu
Zn
As
Rb
Sr
Y
Zr
Nb
Mo
Sn
Sb
DOR 107
15.3
149.8
20
15
24.6
12.4
2092
120
18.2
15.9
451
7
67
2.5
1.7
752
DOR 108
12.6
164.1
20
14
12.2
10.5
345.4
66.8
9.2
10.4
468
6.7
62.1
2.2
2.2
50
2788
1336
DOR 109
21.5
146.5
17
14
15.5
10.4
669.5
76.9
9.3
12.9
463
6.9
56.5
2.1
1.5
271
1426
DOR 110
17.6
144.5
17.71
14.35
15.8
10.9
851.1
83.4
9.8
14.4
458
6.6
54.6
2.2
1.5
531
1796
DOR 111
14.7
139.6
23.41
18.74
36
39.8
4143.2
668.7
68
16.8
472
7.7
77.9
3.5
1.7
2441
1804
DOR 112
16.4
152.1
19.05
15.41
23.3
12.6
1613.8
111.5
15.6
13.5
464
7.1
66.6
2.3
1.5
732
2187
DOR 113
17.6
152.4
21.56
17.15
18.3
13.8
881
92
15.3
15.5
482
7.5
72
2.3
2
512
2115
DOR 115
39.9
152.2
23.55
19.11
9.1
12.3
634.5
126.6
9.6
24.9
479
7.2
85.6
2.6
2
306
1313
DOR 116
21.7
144.7
25.15
20.71
18.6
13
861.5
61.8
16
11.7
475
7.5
79.3
2.5
2.3
495
1313
DOR 117
30.1
154.2
21.16
18.55
28.7
13.1
1149.1
122.7
16.1
17.7
478
7.8
74.4
2.3
2
430
1973
DOR 118
9.6
150.4
15.37
14.09
9.3
8.5
460.5
46.6
10.2
13
446
6.4
56.2
1.9
1.4
116
1523
DOR 119
15.4
141.8
18.77
14.63
31.9
12.5
1215.8
100.6
17.5
13.9
455
6.7
61.6
2.2
1.8
746
2566
DOR 120
13
136.6
22.1
18.95
15.5
10.8
1130.2
202.8
17.8
17.1
410
6.9
82
2.3
1.7
604
1082
DOR 121
14.3
161
22.41
16.88
13.6
11.2
580.3
57.3
13.3
11.3
464
7
73.3
2.4
2.4
223
1934
DOR 122
17.9
99.9
20.57
24.32
6.7
10
273.7
97
4
8.8
267
6.8
153.2
3.5
0.8
28
285
DOR 123a
6.1
146.4
27.59
22.58
15.1
13.1
608.7
47.5
13.2
8.3
474
7.7
90
2.6
2.7
136
1456
DOR 123b
8.7
153.3
26.53
21.88
8.7
12.2
564.6
42.7
10.8
11.8
494
7.8
86.8
2.4
2.8
46
1200
DOR 124
7.3
141.2
20.48
13.55
7.4
9.7
240.7
41.2
8.1
13.8
436
7.4
52.4
1.9
2.2
34
1128
DOR 125
5.4
173.6
15.38
11.48
5
7.5
225.5
30.1
24
10
464
6.3
57.4
1.8
0.9
48
2856
DOR 126
4.8
196.4
17.37
13.52
5.6
7.5
216.6
29.9
22.4
12.4
422
6.5
69.1
2.1
1.6
51
4506
DOR 127
7.9
155.6
18.86
12.44
5.7
8.6
105.5
28.2
12.4
8.4
468
6.7
48.4
1.7
1.5
31
1853
DOR 128
9.7
176.7
18.05
11.65
5.8
7.9
128.5
28.6
15.4
8.5
443
6.4
52.8
1.6
1.7
40
2250
DOR 129
14.6
176.9
18.23
14.19
23.9
14
2142
130.5
20.6
14.4
429
6.4
60
2.2
1.6
903
2840
DOR 130a
17.7
192.8
16.53
12.69
7.9
8
417.3
48.5
12.8
15.4
372
6.4
61.5
1.9
1
115
2686
DOR 130b
16.3
191.7
20.18
15.53
305.1
28.4
1656
114.4
26.8
16.7
404
6.4
59.5
3.1
2.9
243
6909
DOR 131
8.1
198.5
11.05
10.1
7.2
5.8
673.5
44.2
23.6
8.5
398
5.7
46.5
1.7
0.5
164
3303
DOR 132
30.4
145.6
18.79
14.78
23.4
13.5
1045
104.5
9.8
18.6
441
8.2
66.9
2.4
1.6
1399
1418
DOR 133
13.9
169.3
18.33
13.75
19.8
13.4
1909.5
130.6
21.6
14.4
447
6.9
61.9
2
1.5
819
2486
DOR 134
6.2
191.7
19.53
13.34
22.8
19.4
6169.7
271
47
13.8
443
6.4
60.8
2.1
1.8
885
4673
DOR 135
14.7
167.2
20.54
16.2
18.2
12.4
1454
122.4
13.6
15.4
450
7.2
69.8
2.3
1.6
496
1658
DOR 136
8.9
350.6
6.87
8.9
8.5
5.6
45
240.3
16
132.1
443
6.5
69.6
2.6
0.4
2
16
DOR 137
16.4
165.9
19.38
14.69
21.3
11.6
1573.3
98.1
16.8
16.7
469
6.6
65.4
2.2
1.7
508
2935
DOR 138
15.1
161.9
19.26
15.45
22.7
13.7
2174.1
115.8
16.1
14.1
460
7.1
65.7
2.3
1.6
997
2447
DOR 139
29.6
168.7
19.4
15.6
5.9
9.2
372.2
84.3
13.1
19.1
395
7.4
72.5
2.3
1.2
220
2481
DOR 141
14.6
148.2
19.37
15.81
27.7
13.8
1490.6
108.9
16.6
15.9
471
6.9
64.4
2.3
1.6
1045
2746
DOR 142
12.5
166.1
18.76
14.43
31.9
12.2
4528.1
160.9
30.1
13.5
421
6.1
62.5
2.1
1.4
1076
5916
DOR 143
4.6
167.3
19.55
16.85
370.2
68.4
2440.5
99.4
27.5
9
451
6.6
70.8
2.2
5
276
4311
DOR 144
4.1
153.3
16.09
10.52
831.3
31.3
954
27.3
8.7
7.1
373
5.6
42.7
1.5
2
8
4735
DOR 145
4.3
160.8
15.54
11.71
336.5
20.6
1168.7
65
29.4
12.3
447
6.5
58.3
2
1.3
111
21027
DOR 146
5
139.9
9.01
8.26
11.8
8.9
24074.2
1279.3
102.4
6.4
383
5.6
35.8
1.6
0.2
960
15011
DOR 147
5.1
118
25.49
11.83
6.3
12.9
10658.5
65.9
37.9
14.2
480
6.7
50.4
1.9
1.9
105
13203
DOR 148
7
87.9
23.95
12.8
8.5
20.3
18352.8
79.7
32.5
11.2
536
7.4
44.4
1.8
3.3
772
5691
DOR 149
12.3
67.8
16.78
12.47
20.4
16.5
1582.8
81.7
20.9
13.1
305
5.4
50.6
2.3
1.2
36161
2250
DOR 150
31.7
75.9
35.26
32.64
8.3
15.4
331.4
1700.2
29
132.9
823
14.9
107.7
6.2
0.8
33
4175
DOR 151
66.4
88
36.72
31.51
8.4
17.3
171.2
1933.9
25.7
128.3
1203
17.3
143.7
8.8
0.8
9
4493
177
—
Cs
Ba
La
Ce
Pr
Nd
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
Hf
Ta
Pb
Th
U
0.33
289
7
13.6
1.6
7
1.4
0.4
1.2
0.2
1.2
0.2
0.7
0.1
0.6
0.1
1.8
0.1
7015
1.4
0.2
322
6.9
12.4
1.6
6.3
1.3
0.4
1.2
0.2
1.1
0.2
0.6
0.1
0.6
0.1
1.6
0.1
627
1.1
1.1
1
0.29
267
6.8
13.5
1.7
6.8
1.4
0.4
1.2
0.2
1.2
0.2
0.7
0.1
0.6
0.1
1.4
0.1
1405
1.2
1
0.27
268
6.7
12
1.7
6.6
1.2
0.4
1.2
0.2
1.1
0.2
0.6
0.1
0.6
0.1
1.4
0.1
3937
1.1
1
0.57
309
7.9
14.3
1.8
6.9
1.5
0.4
1.5
0.2
1.3
0.2
0.7
0.1
0.6
0.1
1.9
0.2
27064
1.6
1.1
0.3
290
7.2
12.4
1.5
6.8
1.4
0.5
1.3
0.2
1.2
0.2
0.7
0.1
0.7
0.1
1.6
0.2
5615
1.3
1.1
0.32
320
7.7
13.2
1.8
7.4
1.6
0.4
1.3
0.2
1.3
0.3
0.8
0.1
0.7
0.1
1.9
0.2
3616
1.3
1.1
0.4
405
7.7
13.5
1.8
7.2
1.4
0.4
1.2
0.2
1.3
0.3
0.6
0.1
0.7
0.1
2
0.2
1307
1.2
1
1
0.15
358
7.2
12.9
1.7
6.9
1.4
0.4
1.3
0.2
1.1
0.3
0.7
0.1
0.8
0.1
2
0.1
4322
1.2
0.34
370
7.3
13.7
1.7
7.2
1.2
0.4
1.5
0.2
1.2
0.3
0.7
0.1
0.7
0.1
2
0.1
3351
1.4
1
0.22
269
6.4
12
1.7
6.2
1.2
0.4
1.4
0.2
1.2
0.2
0.6
0.1
0.6
0.1
1.5
0.1
754
1
0.9
0.32
290
7.2
12.9
1.7
6.5
1.5
0.4
1.4
0.2
1
0.2
0.7
0.1
0.5
0.1
1.6
0.1
5727
1.3
1.1
0.29
299
7.5
13.4
1.7
6.7
1.4
0.3
1.2
0.2
1.1
0.2
0.6
0.1
0.6
0.1
2
0.2
2914
1.2
1.1
0.27
334
7.2
12.3
1.6
6.7
1.4
0.3
1.2
0.2
1.2
0.3
0.7
0.1
0.7
0.1
1.7
0.1
2149
1.1
0.9
0.08
277
7.1
13.5
1.6
6.8
1.2
0.4
1.5
0.2
1.1
0.2
0.6
0.1
0.8
0.1
3.6
0.2
461
1.5
1.2
0.17
408
7.5
12.8
1.7
6.7
1.2
0.4
1.5
0.2
1.3
0.3
0.7
0.1
0.8
0.1
2.1
0.1
1325
1.2
1
0.39
392
7.3
13.2
1.7
6.8
1.6
0.4
1.4
0.2
1.3
0.3
0.7
0.1
0.8
0.1
2.1
0.2
473
1.2
1
0.24
271
7.2
12.7
1.7
6.8
1.3
0.4
1.3
0.2
1.4
0.2
0.7
0.1
0.7
0.1
1.4
0.1
538
1.2
1
0.16
238
6.2
11.5
1.5
6.3
1.2
0.3
1.2
0.1
1.1
0.2
0.6
0.1
0.5
0.1
1.4
0.1
680
1.1
1.1
0.24
294
6.5
12.2
1.5
6.2
1.1
0.4
1.1
0.2
1.1
0.2
0.5
0.1
0.6
0.1
1.7
0.1
1915
1.4
1.2
0.22
300
6.1
11.3
1.5
6
1.1
0.4
1.3
0.2
1.1
0.2
0.7
0.1
0.7
0.1
1.1
0.1
509
0.9
0.9
0.22
253
6.3
11.1
1.4
5.9
1.2
0.3
1.1
0.2
1
0.2
0.5
0.1
0.5
0.1
1.4
0.1
422
0.9
0.9
0.31
262
6.3
11.7
1.5
6.1
1.1
0.3
1.2
0.2
1.1
0.2
0.6
0.1
0.6
0.1
1.5
0.2
8405
1.2
0.9
0.36
221
6.8
12.4
1.5
6.4
1.3
0.3
1.2
0.2
1.1
0.2
0.7
0.1
0.6
0.1
1.6
0.1
1437
1.3
1
0.34
230
7.3
13.1
1.7
6.6
1.2
0.3
1.1
0.2
1.1
0.2
0.7
0.1
0.6
0.1
1.6
0.2
2794
1.4
1
0.15
190
5.6
9.7
1.3
5.6
1.3
0.3
0.9
0.2
0.9
0.2
0.6
0.1
0.4
0.1
1.2
0.1
1315
0.9
0.8
0.48
292
8
14.4
1.9
7.5
1.4
0.3
1.6
0.2
1.4
0.3
0.8
0.1
0.7
0.1
1.7
0.2
4020
1.4
1
0.38
335
6.4
12.1
1.5
6.6
1.5
0.3
1
0.2
1.1
0.2
0.6
0.1
0.6
0.1
1.6
0.1
9054
1.2
1
0.33
267
6.9
12.4
1.5
6.5
1.5
0.3
1.3
0.2
1.2
0.2
0.7
0.1
0.5
0.1
1.5
0.1
10898
1.6
1
0.31
324
6.9
12.4
1.6
6.5
1.1
0.4
1.3
0.2
1.3
0.3
0.6
0.1
0.6
0.1
1.8
0.1
3840
1.2
1
2.24
1099
68
83
8.1
23.2
2.1
0.4
1.7
0.2
1
0.2
0.5
0.1
0.5
0.1
1.7
0.2
701
1.5
0.6
0.34
312
7.2
13
1.7
6.8
1.1
0.3
1.4
0.2
1.1
0.2
0.6
0.1
0.5
0.1
1.8
0.2
4250
1.4
1.1
0.31
302
7.1
13.3
1.7
6.7
1.3
0.4
1.4
0.2
1.3
0.2
0.7
0.1
0.5
0.1
1.6
0.1
8335
1.4
1
0.39
240
7.5
14.4
1.9
7.2
1.7
0.4
1.4
0.2
1.1
0.3
0.7
0.1
0.7
0.1
1.8
0.1
485
1.5
1.1
0.37
304
7.4
13.1
1.8
6.8
1.3
0.4
1.3
0.2
1.2
0.2
0.6
0.1
0.5
0.1
1.6
0.1
9728
1.3
1
0.31
259
6.7
12.2
1.5
6.3
1.3
0.3
1.2
0.2
1.1
0.2
0.6
0.1
0.6
0.1
1.6
0.2
12180
1.4
1.1
0.21
279
6.5
12
1.6
6.7
1.4
0.4
1.3
0.2
1.1
0.2
0.7
0.1
0.6
0.1
1.7
0.2
2844
1.1
1
0.1
212
5.4
10
1.2
5.5
1.2
0.4
1
0.1
1
0.2
0.6
0.1
0.5
0.1
0.9
0.1
177
0.9
1
0.32
223
7.1
13.3
1.7
6.6
1.4
0.4
1.4
0.2
1.1
0.2
0.5
0.1
0.6
0.1
1.3
0.1
5671
1.6
1.1
0.1
174
6.4
9.9
1.4
5.9
1.1
0.3
1.1
0.1
0.9
0.1
0.5
0.1
0.5
0.1
1
0.1
1335
1.1
0.9
0.29
272
7.5
13.3
1.7
6.5
1.5
0.4
1
0.2
1.2
0.2
0.6
0.1
0.6
0.1
1.4
0.1
2305
1.6
1.1
0.36
309
7.9
13.6
1.7
7.7
1.5
0.4
1.5
0.2
1.2
0.2
0.9
0.1
0.7
0.1
1.2
0.1
121308
1.3
1.1
0.42
206
5.9
10.3
1.3
4.9
1.1
0.3
1.2
0.2
0.9
0.2
0.6
0.1
0.5
0.1
1.3
0.2
488078
1.3
0.9
8.43
25021
25.9
41.3
5.4
20.8
3.7
1.2
2.9
0.4
2.4
0.5
1.4
0.2
1.3
0.2
2.7
0.6
263859
6
6
8.58
22165
34.2
58.8
7.1
26.5
5
1.6
4.2
0.5
3.1
0.6
1.6
0.2
1.5
0.2
4.4
1
288731
12.8
7.4
178
—
Appendix III trace element chemical compositions of samples analysed by LA-ICP-MS
Trace
elements
Li
B
V
Cr
Co
Ni
Cu
Zn
As
Rb
Sr
Y
Zr
Nb
Mo
Sn
Sb
Deventer glass samples
DEV 1
52.5
415.5
59.57
79.22
12.8
41.8
29.8
68.5
18.6
190.5
402
18.4
180
12.5
1.6
2
1
DEV 2
5.9
114
22.93
14.77
411.8
44.1
724.6
26.7
5.1
7
580
7.1
76.3
2.5
3.7
17
5
DEV 3
10.3
155
17.24
12.94
9.4
8.3
844.9
294.1
20.6
15.2
429
6.5
62.4
2
1.6
127
1669
DEV 4
10.5
43.3
12.72
11.98
19.3
19.7
85.8
76.1
29.1
31.3
382
3.3
56.2
2.5
1.7
1
1050
DEV 5
17.7
118.7
11.71
6.42
14.5
27.8
303.8
127
63.5
65.1
466
4.9
33.4
1.3
2.3
13
2
DEV 6
14.5
262.3
15.7
17.37
6.9
9.8
45.8
329.1
2.8
179.1
410
10.4
140.7
4.8
0.3
1
4
DEV 7
12.9
226.4
19.46
24.05
5.3
11.6
44.5
260.2
3
191.7
377
11.4
192
6
0.2
1
3
DEV 8
13.8
215.6
7.21
8.92
6
8.1
56.7
332.9
1.1
119.5
682
3.9
71.3
2.7
1.7
1
1
DEV 9
6.9
44.4
5.55
168.08
2.9
4.5
8.9
27.9
24.6
10.9
182
2.6
107.8
1.5
0.2
6
3
DEV 10
31.4
137.6
19.35
19.95
12.1
11.1
1017.9
108.2
5.5
23.5
335
6.5
113.8
2.8
1.1
323
759
DEV 11
6.2
89.5
21
24.08
10.1
8
487.9
82.6
6.4
7.7
253
6.9
158.4
3.8
0.4
121
117
DEV 12
11.4
155.8
21.24
17.56
8.7
9.7
508.5
45.9
12.7
13
429
6.9
73.8
2.4
2
74
1314
DEV 13
49.9
174.2
21.57
14.67
8.5
31.8
112.3
411.4
74.2
295
916
10.6
149.5
4.4
1.7
90
45
DEV 14
28.3
190.4
19.9
18.97
7.3
20.8
412.1
161.3
2
206.3
269
11.1
189.7
5.4
0.5
29
0
DEV 15
18.1
123.5
17.77
26.31
4.1
24.8
121.7
442.1
0
174
281
9
145.8
4.6
3.4
10
0
DEV 16
weathered
0
118.2
5.27
20.08
6.5
23.7
382.9
274.7
6.5
65.9
113
9.8
143
3.7
0.3
35
1
DEV 17
weathered
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
DEV 18
52.3
169.2
16.22
17.27
10.9
18.2
1619.7
191.8
6.3
124.5
416
8.9
122.9
3.9
2
261
595
DEV 19
20.5
143.2
22.87
19.78
16.2
13.5
677.5
72.9
11.1
14.5
427
7.1
91.9
2.8
2.4
389
1091
DEV 20
12.2
182.1
33.84
14.79
5.2
15.1
46.2
28.7
5.5
6
456
7.8
64.6
2.1
2.1
2
38
DEV 21
10.9
191.1
19.34
27.27
6.1
17.2
57
257.7
1.8
96.1
606
9.8
192.5
6.9
1.8
1
0
DEV 22
14
58.3
7.32
10.66
7.8
7.4
40.8
68.8
14.4
34.1
371
3.6
54.6
2.3
1
1
194
DEV 23
weathered
13.7
46.1
7.96
8.16
4.5
4.4
34
43.1
5.1
16.3
102
3.1
88.5
5.7
0.5
2
34
DEV 24
24.3
245.4
10.25
10.85
42.5
46.2
87.1
462.9
2.2
38.9
782
4.9
70.4
3.7
0.6
1
0
DEV 25
10.8
63.1
8.6
7.29
3.2
6.9
95.3
87.2
6.9
22.2
260
3.1
66.7
2.3
1.1
3
121
DEV 26
35.4
120.1
6.71
6.54
1.5
9.5
21901.8
169
7
362.4
455
2.7
70.4
1.4
1
55
2
DEV 27
15.3
178.3
22.89
26.95
7.1
37.6
75.2
244.8
1.7
233.4
373
11.8
190.9
6.1
4.1
5
1
DEV 28
13
260.9
15.65
17.05
7
11.1
45.8
324
2
181.2
410
9.5
124.6
4.3
0.2
1
0
0
DEV 29
28
172.3
17.78
21.22
7.7
19.6
59.2
295.6
1.1
109.5
892
7.5
142.2
5.5
1.7
1
DEV 30a
16.2
173
8.8
6.08
49.1
38.6
85.4
321
48.3
354.7
1058
6.6
161.6
3.3
0.5
7
1
DEV 30b
0.2
18.2
15.31
9.69
40.7
40.3
115.8
716.8
41.7
112.9
240
11.3
273.5
5.3
2.3
11
2
DEV 31
50.1
92.2
19.82
25.1
7.7
9
372.5
48.9
2.3
10.6
238
6.4
155.6
3.6
0.3
111
130
DEV 32
9.3
228.3
6.21
10.14
4.7
16.4
85.7
206.8
0.6
190.3
360
3.5
89.6
2
2.9
11
0
DEV 33
8.1
251.9
7.93
13.57
5.1
14.8
69.4
164.4
0.8
170.6
218
3.5
65.3
1.7
2.5
2
0
29.4
1951.6
14.89
2391.84
0
14.9
7.4
1669.2
89.8
1.3
1
0
0
0
0
126
36
DEV 35
15.8
263.3
4.19
6.02
3.3
10.6
59.5
268.5
1.4
431.8
993
2.8
71.7
1.1
0.8
1
0
DEV 36
43.7
166.1
7.79
10.58
3.3
12
780.8
213.6
2.3
331.5
417
3.2
76.4
1.7
2.7
37
1
DEV 37
11.7
107.1
9.99
9.01
2.7
7.7
25.5
76.5
4.7
10.2
264
5.7
129.9
2.5
0.9
5
3
DEV 34
weathered
DEV 38
19.9
160.3
5.31
5.98
1
4.6
158.3
187.8
2.1
306.4
583
1.8
56.7
1.3
0.9
5
1
DEV 39
22.8
168.4
17.97
19.81
5.7
14.7
49.2
196.2
1.4
167.7
339
9.3
127.6
5
8.2
1
0
DEV 40a
19
87.6
18.93
23.8
8.1
11.5
21560.9
80.5
5.3
12.6
235
6.1
145.3
3.4
0.5
156
175
DEV 40b
10
89.1
20.87
25.05
8.9
8.4
837.6
71.3
5.1
11.2
269
6.8
157.3
3.7
0.3
258
137
12.4
127.1
15.87
15.39
12.7
15.4
21451.4
113
35.3
36.3
366
6.6
96
2.8
1.1
201
928
DEV 41
179
—
Cs
Ba
La
Ce
Pr
Nd
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
Hf
Ta
Pb
Th
U
2.5
370
27.2
67.8
6.2
23.6
4.5
1.1
4.8
0.6
3.6
0.8
2
0.3
1.8
0.3
5.2
0.7
12
10.9
2.6
0.09
280
6.8
11.6
1.6
6.3
1.4
0.4
1.2
0.2
1
0.2
0.6
0.1
0.7
0.1
1.9
0.2
442
1.1
1.1
0.22
287
7.6
12.6
1.7
6.7
1.4
0.3
1.3
0.2
1.2
0.2
0.6
0.1
0.6
0.1
1.5
0.1
1323
1.2
1
0.38
960
5.1
9.6
1.1
4.2
0.8
0.3
0.8
0.1
0.6
0.1
0.4
0.1
0.3
0.1
1.4
0.2
31684
0.9
3
0.39
582
4.8
6.6
1
4.3
1.2
0.2
0.9
0.1
0.8
0.1
0.5
0.1
0.4
0.1
0.9
0.1
70
1.1
2
1.24
958
45.1
58.7
6.5
20.6
3
0.5
2
0.3
1.8
0.4
1.1
0.1
1.1
0.1
3.7
0.3
5
3
0.9
1.22
902
38.6
50.8
5.5
18.9
2.7
0.5
2.5
0.4
2.1
0.5
1.3
0.2
1.2
0.2
5.4
0.4
8
4
1.1
0.99
1647
10.4
16.9
1.6
5.3
0.7
0.2
0.6
0.1
0.6
0.1
0.5
0.1
0.4
0.1
1.9
0.2
3
1.1
0.4
0.22
251
3.9
10.8
0.7
2.6
0.4
0.1
0.5
0.1
0.3
0.1
0.3
0
0.2
0.1
2.8
0.1
60
0.8
0.6
0.53
341
10.3
16.3
1.8
7.9
1.5
0.3
1.3
0.2
1.1
0.2
0.8
0.1
0.6
0.1
2.9
0.2
1349
1.6
1
0.23
182
7.4
13.5
1.7
6.5
1.3
0.4
1.1
0.2
1.1
0.3
0.8
0.1
0.7
0.1
3.8
0.3
842
1.5
1.1
0.19
392
7.6
12.7
1.7
6.7
1.4
0.3
1.1
0.2
1.2
0.3
0.6
0.1
0.5
0.1
1.7
0.1
650
1.2
1
1.82
4829
12.2
24.9
2.9
10.9
2.2
0.8
1.8
0.3
1.6
0.4
1.1
0.2
1.1
0.1
3.9
0.3
394
4.5
2.5
1.1
1.2
1075
39.5
54.3
5.4
17.5
2.7
0.4
2.1
0.3
1.8
0.4
1.3
0.2
1.3
0.1
4.8
0.4
16
4.2
1.93
1082
13.8
24.1
2.9
10.6
1.7
0.5
1.9
0.2
1.7
0.3
1
0.2
1
0.1
3.7
0.4
88
3.5
1
0.45
359
16
30.7
3.6
12.6
2.2
0.6
1.9
0.3
2
0.5
1.5
0.1
0.8
0.1
3.4
0.4
29
5.3
0.6
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
0.9
1.19
863
29.7
34.8
4.5
14.8
2
0.4
2
0.3
1.5
0.3
1
0.1
0.9
0.1
3.3
0.3
984
2.4
0.26
428
7.6
13.2
1.8
6.7
1.3
0.4
1.5
0.2
1.3
0.3
0.8
0.1
0.7
0.1
2.3
0.2
2035
1.3
1
0.06
235
7.6
11.5
1.7
7.1
1.4
0.3
1.5
0.2
1.4
0.3
0.8
0.1
0.8
0.1
1.5
0.1
37
1.1
0.9
1.6
2243
20.7
36.2
3.7
12.8
1.8
0.5
1.7
0.2
1.6
0.4
1
0.2
1.1
0.2
4.9
0.5
6
3.2
1.1
0.33
419
4.4
7.7
0.9
3.9
0.8
0.2
0.7
0.1
0.5
0.1
0.3
0
0.4
0.1
1.3
0.1
28453
0.9
0.9
0.26
190
6.1
11.3
1.3
4.7
0.8
0.2
0.6
0.1
0.6
0.1
0.4
0
0.4
0.1
2
0.3
70
1.6
0.8
0.34
3273
10.2
14.8
1.6
5.7
1
0.4
0.9
0.2
0.8
0.2
0.4
0.1
0.4
0
1.9
0.2
4
1.8
0.6
0.52
321
4.4
7.5
0.9
3.5
0.6
0.2
0.5
0.1
0.4
0.1
0.3
0.1
0.3
0
1.8
0.2
138
1.3
1
0.3
3.55
1273
3.6
7
0.8
2.8
0.5
0.2
0.4
0.1
0.4
0.1
0.3
0
0.3
0
1.7
0.1
521
1
1.55
1243
36.2
39.5
5.8
19
2.6
0.6
2
0.3
2.1
0.5
1.2
0.2
1.2
0.2
5.1
0.4
26
3.4
1.1
1.22
939
45.7
57.7
6.5
22
2.8
0.6
2.5
0.3
1.5
0.3
1
0.1
0.8
0.2
3.3
0.3
4
2.7
0.9
0.8
1.28
1761
14.3
26.2
2.7
8.7
1.3
0.4
1.2
0.2
1.2
0.3
0.8
0.1
0.8
0.2
3.9
0.3
4
2.4
1.52
5423
12.2
32.3
2.8
11.5
2.2
0.8
1.6
0.2
1.2
0.3
0.7
0.1
0.6
0.1
4.3
0.2
38
4.6
1.8
0.48
3473
20.8
55.4
4.8
17.6
3.1
0.9
3.2
0.4
2.1
0.4
1
0.2
1.1
0.1
7.9
0.4
68
7.4
2.8
0.22
189
6.6
13.1
1.5
6.6
1.3
0.4
1.4
0.1
1
0.2
0.7
0.1
0.6
0.1
3.8
0.2
795
1.4
1.1
1
1124
21.4
17.9
2.3
7.5
0.8
0.2
0.6
0.1
0.6
0.1
0.3
0.1
0.3
0.1
2.5
0.1
3
1.1
0.3
0.65
663
60.5
40.1
6.6
20.6
1.7
0.2
1.1
0.1
0.6
0.1
0.3
0
0.3
0.1
1.5
0.1
3
1
0.3
0
0
0
0.4
0
0
0
0
0
0
0
0
0
0
0
0
0
0.5
3
0.8
0
4.31
2718
6.2
9.5
1
3.5
0.4
0.3
0.4
0.1
0.5
0.1
0.3
0
0.3
0
1.9
0.1
4
0.9
0.3
2.75
1332
14.5
15.4
1.9
7.2
0.6
0.2
0.7
0.1
0.6
0.1
0.4
0.1
0.4
0.1
2.1
0.1
155
1.2
0.4
0.16
106
9.3
16.5
2.1
7.3
1.6
0.3
1.4
0.2
1.1
0.2
0.6
0.1
0.5
0.1
3.2
0.2
58
2.4
0.7
3.31
1350
3.2
6.1
0.6
2.1
0.4
0.2
0.3
0.1
0.3
0.1
0.2
0
0.3
0
1.3
0.1
50
0.8
0.3
1.16
605
24.5
44.7
4.3
15.1
2.2
0.4
2
0.3
1.5
0.3
0.9
0.1
0.9
0.1
3.2
0.3
3
2.6
0.8
0.29
186
7
13.2
1.6
6.2
1.2
0.3
1.3
0.2
1.2
0.2
0.7
0.1
0.7
0.1
3.4
0.3
931
1.4
1.1
0.23
203
7.2
13.9
1.6
6.5
1.3
0.3
1.3
0.2
1.1
0.2
0.8
0.1
0.6
0.1
3.8
0.2
1681
1.6
1.2
0.34
350
9.6
16
1.9
7.2
1.3
0.3
1.2
0.2
1.2
0.2
0.7
0.1
0.7
0.1
2.4
0.2
1913
1.5
1
180
—
Appendix III trace element chemical compositions of samples analysed by LA-ICP-MS
Trace
elements
Li
B
V
Cr
Co
Ni
Cu
Zn
As
Rb
Sr
Y
Zr
Nb
Mo
Sn
Sb
Susteren glass samples
Sust 1 (bead
body)
6.5
139
11.6
11.3
138.7
24.5
9010
1638
22.3
8.7
461
6.6
47.3
1.7
1.2
1947
3471
Sust 2 (bead
body)
10
150.7
17.8
14.1
16.3
12.3
1879.3
99.3
15
12.3
417
7.2
65.1
2.2
1.8
684
1988
Sust 2
(decoration)
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Sust 3 (bead
body)
17.5
120.7
18.7
46.4
974
34
1180
3990
23.5
13.8
506
5.7
98.6
2.4
2.3
3223
505
Sust 3
(decoration)
14.6
147
25.2
21.5
20.1
28.3
1580.7
114.2
19.1
14.2
432
8.4
91.7
3.7
2.2
96733
3357
Sust 4 (bead
body)
16.1
142.6
19.3
19.7
50
13.7
1279.7
247.6
12.4
14.7
445
7
74.1
2.2
1.7
570
1635
Sust 4
(decoration)
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Sust 5
(bead body)
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Sust 5
(decoration)
10
166.9
21.7
18.7
13.5
13.1
947.3
81.2
16.1
12.6
428
7.6
80.6
2.6
2.1
9195
2488
Sust 6
(bead body)
15.8
159.3
20.2
16.5
14.8
11.7
1569
94.1
13.4
14.9
471
7.4
75.6
2.4
2
625
2127
Sust 6
(decoration)
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Sust 7
6.4
187.4
13.9
11.9
327
28.2
2173.3
64.8
33.4
9.2
404
6.1
57.9
1.9
2.4
148
10020
Sust 8
5.5
183.7
13.2
11.2
322.3
25.7
1887.7
57.1
32.9
7.5
399
6
52.8
1.7
2.3
131
10253
Sust 9
11.6
169.7
22.9
17.8
51.9
24
22100
339.7
28.6
17.9
406
7.2
74.9
3.4
2.2
1690
3597
Sust 10
10.6
163
17
14.2
5.1
7
117.7
45.9
13.9
11.5
432
6.6
68.5
2
1.7
162
2492
Sust 11
14.2
189.9
25.4
24
11.3
16.1
82.9
257.9
2
398.3
303
12.2
226.6
7.6
0.4
11
4
Sust 12
12.6
132.1
22.7
21.5
22.9
13.4
1867.3
168.3
13.2
18.1
411
6.5
93.1
3.1
1.6
944
1072
Sust 13
14.2
138.9
19
25.1
35.9
15.5
1888.3
127.3
15.9
15.3
451
5.9
72.7
2.2
1.8
484
2078
Sust 14
16.6
132.2
21.2
21
22.2
12.8
2057
114.4
11
18.1
386
7.4
99.2
2.9
1.7
394
1198
Sust 15
6.6
154.1
18.4
13.8
53.9
14.5
6090
118.6
29
13.4
394
6.8
60.8
2.3
1.5
1082
6810
Sust 16
22.2
138.3
22.2
25.2
24.9
16
1159.7
106
13.5
24.5
402
8.3
97
3.2
1.6
557
1682
Sust 17
14.6
171.5
15.5
12.3
419
35.4
2139
94.6
29
61.1
431
5.9
51.5
1.9
2.9
498
9703
Sust 18
17.7
141
74.4
38
14.2
21.1
4906.7
190.8
8.2
106.6
290
11.2
169.6
19.2
2.3
668
1276
Sust 19
12.1
179.2
16.5
14.2
27
13.3
2070
190
13.5
152.3
396
7.1
90.5
3.3
1.1
628
1990
Sust 20
5
156.3
19.2
12.7
13.9
11.6
2463.3
66.3
21.8
12.2
443
5.9
53.4
2
1.9
279
2423
Sust 21
15.4
127.9
22.3
21
22.9
13.8
1722.3
136.1
11.4
16.2
407
7.1
90.6
2.7
1.5
642
1221
Sust 22
9.4
161
18.6
14.8
30.1
12.2
2264
116.3
17.9
12.3
460
6.4
58.8
2.2
1.6
597
2990
Sust 23
11.5
147.2
19.3
15.3
22.6
12.7
1597.7
119.3
15
13.2
455
7.1
63.6
2.2
1.7
670
2095
Sust 24
6.5
154.5
13.3
11.4
513.3
32.2
1585.7
64.8
31.1
11.9
421
6.4
45.5
1.7
2.4
559
13767
Sust 25
30.7
160.7
40.7
41
11
16.3
111.7
44.1
5.4
15.5
495
9.4
161.4
4.4
4
61
260
Sust 26
12.1
160.5
22.8
18
20.6
14.2
2322
127
20.1
12.3
461
6.7
71.5
2.4
2.2
673
2499
Sust 27
13.7
144.1
20.4
18.8
23.7
13.5
1282.3
107.7
11
19.1
451
7.2
71.7
2.3
1.9
563
1723
Sust 28
22.1
115.6
17.3
46.4
25.7
19.4
415.3
126.1
25.4
13.9
503
4.9
92.9
2.1
1.9
252
426
Sust 29
9.3
165.4
21.5
16.4
19.3
15.8
3163.3
187.2
41.2
13.6
456
6.6
66.2
2.5
2
876
3287
2.3
151.7
14.98
14.55
246.3
80.4
904
22.3
8.5
6.3
406
7.3
72.9
2.7
3.3
404
70
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
9.8
105
12.93
863.06
0.8
13.2
3.6
11.8
15.3
12.2
58
8
115.9
1.6
0.4
2
0
Utrecht glass samples
Utr 77
Utr 78
modern
Utr 79
181
—
Cs
Ba
La
Ce
Pr
Nd
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
Hf
Ta
Pb
Th
U
0.12
224
6.5
11.3
1.4
5.8
1.3
0.3
1.1
0.2
1.1
0.2
0.7
0.1
0.6
0.1
1.2
-
6687
0.9
0.7
0.28
274
6.8
12
1.5
6.2
1.2
0.3
1.3
0.2
1.2
0.3
0.7
0.1
0.6
0.1
1.7
-
5260
1.2
0.9
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
0.35
201
7.6
14.4
1.7
6.3
1.2
0.3
1.1
0.2
1.1
0.2
0.6
0.1
0.5
0.1
2.6
-
2246
1.7
0.9
0.77
430
8.7
16
2
7.8
1.6
0.4
1.5
0.2
1.4
0.3
0.9
0.1
0.8
0.1
2.5
-
868333
1.8
1.1
0.3
293
7.3
12.8
1.6
6.4
1.3
0.3
1.3
0.2
1.1
0.2
0.7
0.1
0.6
0.1
1.9
-
3800
1.4
1
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
0.35
357
7.2
12.6
1.6
6.4
1.4
0.4
1.3
0.2
1.2
0.3
0.7
0.1
0.6
0.1
1.9
-
175333
1.3
1.1
0.31
312
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
7.3
13.3
1.7
7
1.2
0.4
1.3
0.2
1.3
0.2
0.7
0.1
0.7
0.1
1.9
-
5003
1.3
1.1
0.15
205
6.7
11
1.4
5.9
1.2
0.3
1.1
0.1
1.1
0.2
0.6
0.1
0.5
0.1
1.5
-
1862
1.1
1
0.14
194
6.5
10.9
1.4
5.4
1.1
0.3
1.1
0.2
1
0.2
0.6
0.1
0.5
0.1
1.4
-
2150
1.1
1
1.1
0.83
255
8.2
14.9
1.7
7
1.5
0.4
1.3
0.2
1.1
0.2
0.7
0.1
0.6
0.1
1.9
-
11550
1.8
0.28
252
6.4
11.5
1.5
5.8
1.2
0.3
1.3
0.2
1.1
0.2
0.6
0.1
0.6
0.1
1.7
-
614
1.1
1
1.75
1107
37.3
64.3
5.9
19
2.9
0.5
2.4
0.3
2.1
0.4
1.2
0.2
1.2
0.2
5.8
-
28
5
1.5
1.1
0.37
358
8.3
14.9
1.9
7
1.3
0.4
1.3
0.2
1.1
0.2
0.7
0.1
0.7
0.1
2.3
-
4760
1.4
0.35
291
7.1
13.3
1.6
6.4
1.2
0.3
1.1
0.2
1
0.2
0.6
0.1
0.5
0.1
1.9
-
4397
1.3
1
0.32
349
8.5
14.9
1.8
7.3
1.3
0.4
1.3
0.2
1.1
0.3
0.7
0.1
0.7
0.1
2.5
-
3657
1.4
1.1
0.45
235
7.8
13.5
1.6
6.7
1.3
0.3
1.3
0.2
1.1
0.2
0.6
0.1
0.6
0.1
1.5
-
12260
1.3
0.9
0.87
310
9.5
18.5
2.2
8.7
1.7
0.4
1.6
0.2
1.4
0.3
0.9
0.1
0.8
0.1
2.5
-
5157
2.2
1.1
0.54
276
6.7
11.9
1.5
6
1.2
0.3
1
0.2
1
0.2
0.6
0.1
0.5
0.1
1.3
-
6350
1.1
1
2.97
357
13.8
27
2.7
8.9
1.7
0.5
1.6
0.3
1.8
0.4
1.3
0.2
1.3
0.2
4.6
-
302
5.8
2.9
1.68
803
28.5
41.2
4.2
13.4
1.9
0.4
1.4
0.2
1.2
0.3
0.7
0.1
0.6
0.1
2.3
-
4973
2.3
1
0.3
320
6.7
12.2
1.5
6.1
1.2
0.3
1.1
0.2
1.1
0.2
0.6
0.1
0.5
0.1
1.3
-
4217
1.2
1
0.38
318
8
14
1.8
7
1.4
0.4
1.4
0.2
1.1
0.3
0.7
0.1
0.7
0.1
2.2
-
3250
1.4
1.1
0.3
275
7
12.4
1.6
6.3
1.4
0.4
1.2
0.2
1.1
0.2
0.6
0.1
0.5
0.1
1.5
-
6760
1.2
1
0.35
288
7
12.9
1.7
6.4
1.3
0.3
1.2
0.2
1.2
0.2
0.6
0.1
0.6
0.1
1.7
-
4950
1.3
1.1
0.26
226
6.9
10.8
1.4
5.7
1.2
0.3
1.2
0.2
1
0.2
0.6
0.1
0.5
0.1
1.2
-
6077
1
0.9
0.34
1736
8.9
16.3
2.1
8.5
1.8
0.5
1.6
0.3
1.6
0.3
1
0.1
0.9
0.1
4
-
315
1.8
1.3
0.32
351
7.2
13
1.6
6.4
1.4
0.4
1.2
0.2
1.1
0.2
0.7
0.1
0.6
0.1
1.8
-
5890
1.3
1.1
0.43
345
7.6
13.4
1.7
6.8
1.5
0.3
1.3
0.2
1.2
0.2
0.6
0.1
0.6
0.1
1.8
-
3856
1.3
1
0.25
207
6.9
13.3
1.6
6.2
1.1
0.3
0.9
0.1
0.8
0.2
0.5
0.1
0.5
0.1
2.4
-
10567
1.4
0.8
0.37
299
7.2
12.8
1.6
6.4
1.3
0.4
1.3
0.2
1.1
0.2
0.7
0.1
0.6
0.1
1.7
-
11330
1.3
1
0.06
183
6.7
12.1
1.5
6.1
1.3
0.3
1
0.2
1.1
0.2
0.6
0.1
0.6
0.1
1.9
0.2
4035
1.2
1.3
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
0.36
318
5.5
12.3
1.3
5.2
1.3
0.3
1.5
0.2
1.4
0.2
0.9
0.1
0.8
0.1
2.9
0.1
93
2.9
0.6
Appendix IV photos of the samples from
Maastricht and Utrecht
1:1
0
1:2
5cm
Figure appendix IV.1 Photo MABRO 1, Mabro, Maastricht, rim fragment
of crucible with white ‘frit-like’ material adhering.
0
Figure appendix IV.4 Photo MABRO 4, Mabro,Maastricht, rim fragment
of crucible with colourless glass on inside.
1:2
0
1:2
10cm
Figure appendix IV.2 Photo MABRO 2, Mabro, Maastricht, Crucible fragment
with colourless and white/yellow material attached.
1:2
0
10cm
10cm
Figure appendix IV.3 Photo MABRO 3, Mabro, Maastricht, small crucible
fragment with green and weathered opaque yellow pigment.
0
10cm
Figure appendix IV.5 Photo MABRO 5, Mabro,Maastricht, rim fragment
of crucible with colourless vitreous material on inside white ‘frit-like’
material adhering to both sides.
183
—
184
—
1:1
1:2
0
10cm
Figure appendix IV.6. Photo MABRO 6, Mabro, Maastricht, base fragment
of crucible with opaque yellow pigment adhering.
0
5cm
Figure appendix IV.9 Photo MABRO 9, Mabro,Maastricht, small crucible
fragment with deep translucent glass on the inside.
1:1
0
1:1
5cm
Figure appendix IV.7 Photo MABRO 7, Mabro,Maastricht, small red fragment
of crucible with natural green glass adhering to both sides.
1:2
0
10cm
Figure appendix IV.8 Photo MABRO 8, Mabro,Maastricht, red fragment
of crucible base with opaque yellow outside and colourless-green glass
adhering on the inside.
0
5cm
Figure appendix IV.10 Photo MABRO 10, Mabro,Maastricht, red base
fragment of crucible with green vitrification on lower side.
185
—
1:2
0
1:2
10cm
Figure appendix IV.11 Photo MAJO 1, Jodenstraat, Maastricht, red base
fragment of crucible with opaque yellow pigment adhering (MAJO=
Jodenstraat, Maastricht).
0
10cm
Figure appendix IV.14. Photo MAJO 4, Jodenstraat, Maastricht, base
fragment of crucible with white residue adhering (MAJO= Jodenstraat,
Maastricht).
1:2
0
10cm
Figure appendix IV.12. Photo MAJO 2, Jodenstraat, Maastricht, crucible base
of grey fabric with weathered opaque yellow pigment and translucent
residue adhering (MAJO= Jodenstraat, Maastricht).
1:2
0
10cm
Figure appendix IV.13 Photo MAJO 3, Jodenstraat, Maastricht, base fragment
of crucible with weathered opaque yellow pigment adhering (MAJO=
Jodenstraat, Maastricht).
Figure appendix IV.15 Photo MAJO 5 (inside), Jodenstraat, Maastricht, base
of crucible with opaque yellow glass adhering (MAJO= Jodenstraat,
Maastricht).
186
—
1:2
0
10cm
Figure appendix IV.16 Photo MAJO 5 (outside), Jodenstraat, Maastricht,
base of crucible.
1:2
0
10cm
Figure appendix IV.19 Photo MAJO 8, Jodenstraat, Maastricht,
base of crucible with dark translucent glass adhering (MAJO= Jodenstraat,
Maastricht).
1:2
0
10cm
Figure appendix IV.17 Photo MAJO 6, Jodenstraat, Maastricht, base fragment
of crucible with opaque yellow glass on inside (MAJO= Jodenstraat,
Maastricht).
1:2
0
10cm
Figure appendix IV.20 Photo MAJO 9, Jodenstraat, Maastricht,
possible brick fragment with opaque yellow and white residue adhering
(MAJO= Jodenstraat, Maastricht).
1:2
0
10cm
Figure appendix IV.18 Photo MAJO 7, Jodenstraat, Maastricht,
base of crucible with opaque yellow glass adhering (MAJO= Jodenstraat,
Maastricht).
187
—
1:1
0
2:1
5cm
Figure appendix IV.21 Photo MAJO 10, Jodenstraat, Maastricht,
blue fragments of glass (MAJO= Jodenstraat, Maastricht).
2,5cm
Figure appendix IV.22 Photo MAJO 11, Jodenstraat, Maastricht, scrap
of red glass (MAJO= Jodenstraat, Maastricht).
2:1
0
2,5cm
Figure appendix IV.24 Photo MAJO 13, Jodenstraat, Maastricht, scrap of
green glass (MAJO= Jodenstraat, Maastricht).
1:1
2:1
0
0
2,5cm
Figure appendix IV.23 Photo MAJO 12, Jodenstraat, Maastricht, scrap of
red glass (MAJO= Jodenstraat, Maastricht).
0
5cm
Figure appendix IV.25 Photo MAJO 14, Jodenstraat, Maastricht, yellow-green
window glass (MAJO= Jodenstraat, Maastricht).
188
—
1:1
0
2:1
5cm
Figure appendix IV.26 Photo MAJO 15, Jodenstraat, Maastricht, yellow-green
window glass (MAJO= Jodenstraat, Maastricht).
5cm
Figure appendix IV.27 Photo MAJO 16, Jodenstraat, Maastricht, pale yellowgreen window glass (MAJO= Jodenstraat, Maastricht).
1:1
0
2,5cm
Figure appendix IV.29 Photo MAJO 18, Jodenstraat, Maastricht, weathered
yellow drop (MAJO= Jodenstraat, Maastricht).
2:1
1:1
0
0
5cm
Figure appendix IV.28 Photo MAJO 17, Jodenstraat, Maastricht, thin green
glass rod (MAJO= Jodenstraat, Maastricht).
0
2,5cm
Figure appendix IV.30 Photo MAJO 19, Jodenstraat, Maastricht, red glass
drop (MAJO= Jodenstraat, Maastricht).
189
—
2:1
0
2,5cm
Figure appendix IV.31 Photo MAJO 20, Jodenstraat, Maastricht, dark green
glass drop (MAJO= Jodenstraat, Maastricht).
1:1
0
5cm
Figure appendix IV.34 Photo MAJO 23, Jodenstraat, Maastricht, twisted
opaque white rod (MAJO= Jodenstraat, Maastricht).
2:1
0
2,5cm
Figure appendix IV.32 Photo MAJO 21, Jodenstraat, Maastricht, milky blue
pulled rod (MAJO= Jodenstraat, Maastricht).
1:1
0
5cm
Figure appendix IV.35 Photo MAJO 24, Jodenstraat, Maastricht, green beaker
base (MAJO= Jodenstraat, Maastricht).
1:1
0
5cm
Figure appendix IV.33 Photo MAJO 22, Jodenstraat, Maastricht, thin red rod
(MAJO= Jodenstraat, Maastricht).
190
—
2:1
0
1:1
2,5cm
Figure appendix IV.36 Photo MAJO 25, Jodenstraat, Maastricht, blue punty
glass (MAJO= Jodenstraat, Maastricht).
0
Figure appendix IV.39 Photo DOM 1, Domplein, Utrecht, Body fragment
of crucible with colourless glass adhering (DOM=Domplein, Utrecht).
1:1
0
1:1
5cm
Figure appendix IV.37 Photo MAJO 26, Jodenstraat, Maastricht, blue-green
flat ribbed fragment (MAJO= Jodenstraat, Maastricht).
2:1
0
5cm
2,5cm
Figure appendix IV.38 Photo MAJO 27, Jodenstraat, Maastricht, green glass
rod fragments (MAJO= Jodenstraat, Maastricht).
0
5cm
Figure appendix IV.40 Photo DOM 2, Domplein, Utrecht 2 Body fragment
of crucible with colourless glass adhering (DOM=Domplein, Utrecht).
191
—
1:1
1:2
0
10cm
Figure appendix IV.41 Photo DOM 3, Domplein, Utrecht 3 Base fragment of
crucible with pale green and red glass adhering (DOM=Domplein, Utrecht).
5cm
Figure appendix IV.42 Photo DOM 4, Domplein, Utrecht 4 Rim fragment of
crucible with pale green and red glass adhering (DOM=Domplein, Utrecht).
Figure appendix IV.44 Photo DOM 6, Domplein, Utrecht 6 Body fragment of
crucible with pale green and red glass adhering (DOM=Domplein, Utrecht).
0
2,5cm
Figure appendix IV.45 Photo OUDWIJ 1 (sample 78), A lump of melted pale
green glass from Utrecht (OUDWIJ=Oudwijkerdwarsstraat, Utrecht).
2:1
1:2
0
5cm
2:1
1:1
0
0
10cm
Figure appendix IV.43 Photo DOM 5, Domplein, Utrecht 5 Rim fragment
of crucible with green glass adhering (DOM=Domplein, Utrecht).
0
2,5cm
Figure appendix IV.46 Photo OUDWIJ 2 (sample 79), A chunk of pale green
glass from Utrecht (OUDWIJ=Oudwijkerdwarsstraat, Utrecht).
This monograph brings together for the first time comprehensive combined archaeological, technological and scientific investigations
of early medieval glass production in the Netherlands. The relationships between scientific results, archaeological contexts, sample
dates, object types, colour, changes in glass technologies over time, as well as the social, economic and political factors affecting glass
supply, and glass production, are discussed. We have selected samples from nine key sites, dating to between the late 4th and
11th centuries. Trace element and isotopic results for early medieval glasses have provided new and significant insights. They show that
most glass in use was recycled and that there is a greater proportion of imported ‘pristine’ Egyptian glass in the Merovingian period
than in the Carolingian period. A small proportion of wood ash glass was added to imported Carolingian glass found in the
Netherlands; in contemporary northern Italian and Spanish glasses Levantine glass was added as part of the recycling process instead.
We highlight the international importance of evidence for the production of yellow and white glass tin-based colorants in Maastricht
and their use to make monochrome beads there. A wider range of glass technologies was in use after the 9th century following an
important technological transition.
This scientific report is intended for archaeologists, as well as for other professionals and amateur enthusiasts involved in archaeology.
The Cultural Heritage Agency of the Netherlands provides knowledge and advice to give the future a past.