Compositional distinctions between 16th century 'façon-de-Venise ...

Compositional distinctions between 16th century ‘façon-de-Venise’ and
Venetian glass vessels excavated in Antwerp, Belgium†
I. De Raedt,a K. Janssens*a and J. Veeckmanb
aDepartment of Chemistry, University of Antwerp, Universiteitsplein 1, B-2610 Antwerp, Belgium.
E-mail: koen@uia.ua.ac.be
bExcavation Department of the City of Antwerp, Godefriduskaai 36, B-2000 Antwerp, Belgium
Received 29th October 1998, Accepted 10th December 1998
At the beginning of the 16th century, Antwerp was one of the most important ports in Europe. Among the
archaeological finds that are dated to this period of great wealth, an important number of fragments from luxurious
glass vessels have been found. In view of the high price of finished Venetian glassware, it is unlikely that all the
Antwerp glass finds are Venetian products. The aim of this investigation was to determine whether there was a
significant local production of ‘façon-de-Venise’ glass vessels in Antwerp in the 16th century and to elucidate how
this local production can be distinguished from ‘real’ (i.e., imported) Venetian ware. The analysis results for 122
‘façon-de-Venise’ glass objects excavated in Antwerp were considered. Clustering of the Antwerp data revealed five
compositional groups. Two groups of glass objects were found that feature a typical Venetian composition, one with
a ‘Cristallo’-like composition and the other closely resembling ‘Vitrum Blanchum’ glass. The other three
compositional groups contain glasses that probably were fabricated locally. The composition of a waste glass
marble coincides with one of these groups. The differences between these three groups may indicate that different
types of alkali were in use or possibly mixing of alkali types by the glass workers. The same non-Venetian
composition was found in London, Amsterdam and various locations in France, which may be explained by the
export of locally manufactured ‘façon-de-Venise’ vessels from Antwerp to these nearby cities or by the use of similar
fluxes/flux mixtures to those in Antwerp.
Introduction
the soda-lime type of glass represents the most important
At the beginning of the 16th century, Antwerp was one of the
most important ports in Europe; its wealthy citizens could
afford to purchase luxury items in large numbers, an important
fraction of which was presumably imported from the
Mediterranean area. Examples of such luxury products are
glazed earthenware and glassware (both originating from
various locations in Northern Italy and brought to Antwerp
by Genovese and Venetian traders), of which a significant
number of remains have been excavated at various archaeological
sites in the city centre. Many objects were found in
cesspits, which in addition to refuse, were also used for the
group, probably because this glass composition was employed
locally for the manufacture of both common table ware and
luxury vessels.
With the aim of attempting to make a distinction between
locally fabricated vessels and imported objects, of establishing
some insight into which raw materials were used for the (local )
glass manufacture, the major composition of about 100 clear
and colourless ‘façon-de-Venise’ objects was determined by
electron probe X-ray microanalysis (EPXMA). Objects made
from other types of glass were not included in the present study.
disposal of broken household goods such as glass vessels. In
a later phase, the city of Antwerp itself became a large
production centre for imitation Venetian or ‘façon-de-Venise’
Overview of published compositions of 16–17th
century Venetian/façon-de-Venise vessel glass
glass.1 The earliest documents referring to Italian glassmakers In the literature, only a limited number of papers has been
taking up residence in Antwerp date from 1537. Until the devoted to the analysis of Venetian and ‘façon-de-Venise’ glass
middle of the 17th century, different glassmakers and mer- objects so only a very fragmentary picture of the possible
chants of both Italian and local origin were active in the city compositional differences between genuine Venetian pro-
and were granted monopolistic privileges by the city council duction and its look-alikes is available. Verità and Toninato3,4
or higher authorities to ensure the success of their business. have papers in which the presence of two grades of Venetian
In a previous paper, we reported on the different composition glass, called ‘Cristallo’ (high-quality glass) and ‘Vitrum
categories found among the Antwerp archaeological glass of Blanchum’ (standard quality) (10 and 18 samples analysed,
the 15–17th centuries.2 The study involved the analysis of 96 respectively), with clearly distinct chemical compositions is
representative objects. In addition to the ‘façon-de-Venise’ discussed (see Tables 1 and 2 later). ‘Cristallo’ glass is a
glass, important amounts of green ‘forest’ glass (rich in CaO) transparent and colourless soda-glass that was fabricated from
and colourless ‘fern’ glass (with K as the major cation) were the middle of the 15th century in Venice. Its name was derived
also excavated in the historical centre of the city. Nevertheless, from rock crystal, which it resembled in purity, brightness and
among the archaeological remains dated to the 16th century, homogeneity. The composition of ‘Cristallo’ glass was reputed
to remain very constant with time and was kept a secret. It is
†Presented at the Fifteenth International Congress on X-ray Optics
and Microanalysis (ICXOM), Antwerp, Belgium, August 24–27, 1998.
characterized by a fairly high Na O content (16.9% on average)
2
and low MgO, K O and CaO contents (1.8, 2.8 and 4.9%,
2
J. Anal. At. Spectrom., 1999, 14, 493–498 493

espectively, on average) and was made with clear quartz either the use of soda-ash of variable quality or of two ash
pebbles and ash imported from Syria (the ‘Levant’), which sources (e.g., of Levatine and Spanish origin) and one similar
was washed prior to use, thereby decreasing the abundance of to ‘Vitrum Blanchum’. By combining most of the above-
Fe O , MgO and CaO in the resulting glass. The other group
2 3
of Venetian glass is ‘Vitrum Blanchum’ and dates from the
mentioned analysis results, Bronk7 suggested the existence of
six types of soda-glass, depending on the origin of the ash
14–16th centuries. This is also soda-glass but of inferior (Levantine soda-ash, Spanish soda-ash or soda-ash mixed with
quality, probably owing to the lower purity of the raw materials K-rich wood-ash) and on whether the ash was washed prior
employed for its production.3 ‘Vitrum Blanchum’ contains to use. According to Ashtor,11 Levantine soda ash contains
significantly less Na O (13.3% on average) and more of
2
divalent metal oxides (MgO 3.4%, CaO 9.5%) since it is made
41% Na CO whereas the ashes of saline plants which grow
2 3
in Spain, such as those from the region of Alicante, Cartagena
with the unwashed soda ash. The K O content is almost the
2
same as that of ‘Cristallo’. Both of these compositions were
and Malaga, contain 25–30% Na CO . Another type of soda
2 3
ash, called soda of vareque, may be produced from plants
also found by McCray5 in colourless glass fragments from growing on the coast of southern France and contain about
various locations in the Venetian lagoon (24 samples analysed,
the majority of which are similar to ‘Vitrum Blanchum’) and
10–15% Na CO .
2 3
From the above summary of published analysis results, one
Amsterdam (seven samples excavated at the Keizersgracht, can conclude that glass vessels with a ‘Vitrum Blanchum’ or
fragments of vessels and beads along with fragments resulting similar composition occur frequently, but that the ‘Cristallo’
from production processes, of which all have a ‘Vitrum composition is almost exclusively found in Venice itself.
Blanchum’-like composition). Mortimer6 presented results on With the aim of establishing whether in Antwerp (as in
the analysis of glass samples from the first half of the 17th most of the other sites mentioned above with the exception of
century (finished or partly formed beakers, a bottle, cups, Venice), only ‘Vitrum Blachum’ and not ‘Cristallo’ composi-
canes glass waste, glass remains in a crucible) excavated in tions are found and also to investigate in greater detail whether
London (the City and Old Broad Street) and found the 19 the Antwerp-specific composition(s) identified previously2 are
analysed pieces all to have a composition similar to ‘Vitrum indeed significantly different from all literature data, the
Blanchum’. Half of the objects, however, showed a higher number of analysed 16th century vessels which typologically
K O content (5–7%) that the Venetian compositions (2–3%).
2
Bronk7 also analysed fragments from Amsterdam (five fragfall
within the Venetian or ‘façon-de-Venise’ category and
which have a soda-glass composition, was increased from 33
ments made of soda-glass) and Venice (15 soda-glass goblets) to 122. Below, a brief description of the typology of these
and found all of the Amsterdam samples to consist of ‘Vitrum Antwerp samples and of the sites where they were found is
Blanchum’ except that the K O content was double that in
2
the Venetian glass (4.4–5.6%). Of the 15 Venetian samples,
given, after which the sample preparation procedures and the
analytical method are discussed. Next, the various major
only one sample showed a ‘Cristallo’ composition; six frag- element groups that could be distinguished within the series
ments had a ‘Vitrum Blachum’ composition and two others
featured an intermediate composition; three samples showed
are described and discussed in relation to the literature data.
a higher K O (5–6%) and a lower Na O (10–13%) content
2 2
than ‘Vitrum Blanchum’ and two other samples showed
Glass samples analysed
unusually high Na O (19–21%) and very low K O (1–1.5%)
2 2
contents. In their comprehensive survey of glass in France
The 122 vessels which were analysed were found at different
excavation sites in the historical centre of Antwerp. The largest
from the 10th to the 18th century (486 fragments analysed), number of samples originate from the palace of the bishops
Barrera and Velde8 encountered soda-glass originating from of Antwerp and from the medieval fortress ‘Het Steen’. Fifteen
various sites in France (Paris, Orléans, Angers, Rouen, Meaux, of the samples were excavated at an embankment wall. The
Châlon-sur-Marne and Chambaran), about 100 pieces in rest of the samples originate from excavations in the city
total. This series consists of different types: goblets, tumblers, parking area (Groenplaats), Suikerrui, Kaasstraat and
footed cups, waste glass, flasks and vases. They noted that Haverstraat. At each excavation site, different typological
Venetian/‘façon-de-Venise’ vessels were originally made out of categories of glass were found; these finds include beakers and
soda-glass but from 1450 onwards imitations were present goblets with or without decoration, in plain or filigree glass.
made of K-rich glass. Of the totality of the northern-French Most of the investigated vessels are dated to the first or second
soda-glasses, about half (45 samples) are K-rich and the other half of the 16th century. In the 17th century in Antwerp,
half K-poor (50 samples). Ulitzka9 analysed Venetian glass primarily ‘façon-de-Venise’ vessels made out of K-rich glass
from the Collection of the Veste Coburg (63 samples) and are found.2 Typologically it is difficult to distinguish genuine
made a distinction between nine vessels (possibly manufac- Venetian products from those of local origin; the same
tured in Hall, near Innsbruck, Austria) having a composition production methods, decoration techniques and forms were used
very similar to Venetian ‘Vitrum Blanchum’, three vessels with as in Venice.1 For example, many fragments of filigree glass,
a higher K O level (as found in London6 and Amsterdam7)
2
and two groups of vessels which were assumed to have been
originally made in Murano in the second quarter of the 16th
century but also blown in Antwerp in the second half of the
made with different types of Spanish soda-ash (12 and five 16th century, were found. Various types of filigree glass can
objects, respectively, containing about 10–11% Na O).
2
However, as pointed out by Bronk,7 one of these ‘Spanish’
be distinguished, such as ‘vetro-a-fili’ (with threads of white
or coloured glass that do not cross one another), ‘vetro-acompositions
might also have been obtained by mixing retorti’ (with twisted threads of glass, usually combined ‘vetro-
‘Cristallo’ with K-rich potash glass in a 251 mixture. Of the a-fili’) and ‘vetro-a-reticello’ (with a network of twisted glass
numerous Antwerp finds, up to now only Van Der Wee10 has threads). These objects usually are decorated with masks (lion
published the analysis results for three soda-ash glass objects or Neptune heads) or rosettes. In addition to the filigree
(similar to ‘Vitrum Blanchum’ but with higher K O and lower
2
CaO contents), whereas in our previous paper,2 on the basis
objects, glasses with relief-blown decorations that date from
the second half of the 16th century were also found. Another
of the analysis of 33 vessels of various types, we indicated the type is the conical beaker, decorated with checkered spiral
existence of two or possibly three compositional groups among threads and sometimes additionally adorned with rosettes or
the Antwerp soda-lime vessels, of which two groups showed a pearled prunts. Bossed beakers have a mould-blown decoration
constant abundance of total alkalimetal-oxide content of drops or tears. The meshwork pattern (very popular in
(Na O+K O#20–21%, CaO#10%, MgO#3%), indicating
2 2
Venice in the 15th century and brought to the Antwerp region
494 J. Anal. At. Spectrom., 1999, 14, 493–498

Fig. 1 Illustrations of some of the frequently occurring Venetian and
‘façon-de-Venise’ vessel types.
Results and discussion
Compositional grouping among the Antwerp vessels
During the cluster analysis, the Na O, K O, CaO and MgO
2 2
concentrations were used to differentiate between the different
objects. Fig. 2 shows a dendrogram providing an overview of
the structure of the Antwerp data set. Five major clusters can
be distinguished. In Fig. 3(a) the K O content of the analysed
2
objects is plotted against their CaO content, indicating the
validity of the cluster separation shown in Fig. 2; the group
separation appears straightforward when use is made of the
CaO (above and below 7%) and the K O content (below 3%,
2
between 3 and 6% and above 6%). In the corresponding Na O
2
versus MgO plot [Fig. 3(b)], the cluster separation is less
evident, although in the literature the MgO level is suggested
to distinguish between the use of washed and unwashed sodaash
for glassmaking.7 The average concentrations of the abovementioned
oxides in each cluster are given in Table 1. Also in
in the 16th century) is another method which was used to
decorate glass. Another category comprises the so-called
‘comet beakers’, adorned with blue decorations resembling
comets. A last category of glass that was analysed includes
very luxurious, refined objects called winged and serpentstemmed
glasses; both types of goblets were first made in Fig. 2 Dendrogram obtained by hierarchical cluster analysis of the
Venice before 1550 but are also assumed to have been made data set. Only the Na O, K O, CaO and MgO concentrations were
2 2
locally by glasshouses in the Antwerp region.1,12 Fig. 1 shows used for the clustering.
the most frequently occurring vessel types.
Experimental
Small fragments of glass of about 1 mm2 were cut from the
excavated fragments and embedded in an acrylic resin. Samples
were taken only from broken vessels. The cross-section of the
glass mounted in the resin was then ground and polished with
diamond paste to a 1 mm grain size. Finally, the resin blocks
were carbon-coated to prevent charging of the non-conducting
glass surface.
Analysis was performed using a Jeol 6300 scanning electron
microscope equipped with an energy dispersive X-ray fluorescence
(EDXRF) Si(Li) detection system. The following instrumental
settings were used: 20 kV accelerating voltage, 1 nA
beam current, a magnification setting of about 800 and a 100 s
analysis time. These conditions ensure that no migration of
Na+ takes places during the analysis.
The net elemental X-ray intensities were calculated with the
program AXIL (analysis of X-rays by iterative least squares)
and a standardless ZAF program was used to calculate the
composition.13 Various National Institute of Standards and
Technology (NIST) reference materials were used to validate
the quantification procedure. The quantitative accuracy of the
method is discussed in more detail elsewhere;2 in brief, one
can state that the major element oxide concentrations above
1% m/m can be determined with an accuracy of about 5%
relative. Cluster analysis, a multivariate data analysis technique,
was used for the interpretation of the analytical data
for the ‘façon-de-Venise’ glasses. Hierarchical clustering
(HCA) was used with the purpose of finding objects within
the data set with similar composition and to attempt to
distinguish local production from imported wares. The IDAS Fig. 3 (a) K O versus CaO scatter plot and (b) Na O versus MgO
2 2
software package was utilised for this purpose.14
scatter plot for the Antwerp glass vessel series.
J. Anal. At. Spectrom., 1999, 14, 493–498 495

Table 1 Mean values and standard deviations of some major element oxides within each of the Antwerp clusters
Number of objects in each cluster
Average composition (% m/m)
Relief
Winged Filigree blown Comet
Cluster Na O
2
K O
2
CaO MgO Goblet glass glass beakers beaker
Venetian ‘Cristallo’ 16.9±1.7 2.8±0.3 4.9±0.7 1.79±0.43
Antwerp cluster 1 15.3±1.4 4.4±1.5 4.9±0.8 1.65±0.25 7 8 3 5 —
Antwerp cluster 2 11.1±0.5 8.6±0.8 5.1±0.6 2.66±0.44 — 1 1 1 3
Antwerp cluster 3 15.1±1.1 4.6±0.5 9.1±0.9 3.11±0.30 13 — 12 3 —
Antwerp cluster 4 13.8±1.5 6.8±0.6 10.3±0.6 3.54±0.39 15 — 4 — —
Antwerp cluster 5 14.6±1.1 2.2±0.4 10.0±1.1 1.87±0.27 2 — 13 — —
Venetian ‘Vitrum Blanchum’ 13.0±1.1 2.5±0.6 9.9±1.3 3.38±0.82
Table 1, the composition of these clusters is compared with values may indicate that different types of alkali were in use
the composition of the two genuine Venetian glass types.3 or that a mixture of ash types or ash qualities was used.
Cluster 1 includes all the Antwerp objects with an average Considering the cost of imported soda-ash, it may be
composition of 15.3% Na O, 4.3% K O and 4.9% CaO. This surmised that the Antwerp glassworkers mixed K-rich ash
2 2
composition is very similar to the composition of ‘Cristallo’ with soda-ash to lower the price. However, since cluster 2 is
glass, although the average K O content (4.4±1.5%) is higher not so abundantly present, it appears that this did not take
2
than the group average reported by Verità3 (2.8%). This group place very frequently. Whether the use of this mixed ash is
comprises different types of objects: winged goblets, goblets, situated in the transition period from soda-glass to potash-
filigree objects and beakers with decorations such as checkered glass (roughly at the turn of the 16th century) is not clear.
spiral trails, ribs and droplets, all of which can be assumed to
have been imported from Venice. It should be mentioned that Analysis of Antwerp production waste
this constitutes a substantial group (43 objects).
Another piece of evidence for the production of ‘façon-de-
On the other side of the dendrogram, cluster 5 contains
Venise’ glassware in Antwerp is the fact that a few glass
glass objects with a composition similar to that of ‘Vitrum
‘marbles’, i.e., waste products from the fabrication process of
Blanchum’ glass. A significant difference is noted in the MgO
filigree objects, have been excavated (Fig. 4).12,15 Until now,
content, which is lower than in the Venetian reference group.
the existence of these objects was the only direct indication of
It is interesting that only one filigree object and two goblets
‘façon-de-Venise’ glass production in Antwerp; however, the
belong to this group and that the total number of objects (15)
presence of these few objects does not provide clues on the
is significantly lower than the number of Antwerp ‘Cristallo’
importance of this activity relative to the import of finished
objects. One can assume with some degree of certainty that
Venetian goods. By slightly polishing the side of one of the
also the objects in cluster 5 have been imported; the relative
marbles and carbon coating it, it was possible to determine
limited number of ‘Vitrum Blanchum’ vessels in Antwerp may
the major composition of the marble without sampling or
be explained by the existence of locally produced vessels of
destroying its integrity. In Table 2 the composition is compared
equivalent quality. Indeed, clusters 2, 3 and 4 show no close
with individual compositions of Venetian and Antwerp
resemblance to either of the two Venetian comparison groups.
‘Cristallo’ and ‘Vitrum Blanchum’ and to the composition of
Although cluster 3 features a high concentration of Na O
2 a typical object from cluster 3. The composition of the glass
(about 15%) and K O and CaO concentrations similar to
2 marble clearly coincides with that of cluster 3 members and
those in ‘Vitrum Blanchum’, it has a too high K O content.
2 differs significantly from the vessels (assumed to be) of Venetian
Filigree glasses and goblets with or without decorations
(applied meshworks, droplets, checkered spiral trails and ribs)
belong to this cluster. The composition of this cluster is
indicative of the use of unwashed soda-ash for the glassmaking
and resembles that of the objects found in London6 and
Amsterdam,7 which also showed a higher K O content than
2
the Venetian ‘Vitrum Blanchum’. In cluster 4 the concentration
of Na O is lower and the K O content is higher than those in
2 2
cluster 3, although the MgO and CaO contents of both clusters
are similar. This suggests the use of a different kind of ash for
the glass manufacture. This group consists mainly of undecorated
goblets (a few have a highly decorated stem) and of four
filigree fragments. Cluster 2 has a composition completely
different from that of the Venetian glass; these objects contain
significantly less Na O and contain more K O than CaO,
2 2
indicative of the use of mixed ash type (containing both sodaash
and potash). This group comprises objects of various types
such as comet beakers, a winged glass, a ribbed beaker and a
filigree glass. Although featuring about the same Na O content
2
as the two ‘Spanish’ ‘façon-de-Venise’ groups proposed by
Ulitzka,9 the CaO, K O and MgO contents do not agree.
2
Probably the glass objects from clusters 1 and 5, with
similar compositions to ‘Cristallo’ and ‘Vitrum Blanchum’,
were imported. The other clusters contain objects that may Fig. 4 Formation of a glass waste marble during the manufacture of
have been fabricated locally. The variation in the K O/Na O ‘vetro-a-fili’ beakers.
2 2
496 J. Anal. At. Spectrom., 1999, 14, 493–498

Table 2 Examples of individual Venetian and Antwerp glass compositions (in % m/m) compared with that of the Antwerp glass waste marble
‘Cristallo’ ‘Cristallo’ ‘Vitrum Blanchum’ ‘Vitrum Blanchum’ Goblet cluster 3 Glass marble
Constituent Venice3 Antwerp Venice3 Antwerp Antwerp Antwerp
Na O
2
MgO
15.5
2.35
15.1
1.29
12.5
3.05
13.6
3.79
14.9
2.15
15.7
1.90
Al O
2 3
SiO
2
P O
2 5
SO
3
Cl
0.48
72.7
0.13
0.30
0.75
1.38
71.3
0.21
0.19
0.69
1.40
68.6
0.37
0.28
0.88
1.19
66.9
0.36
0.29
0.67
1.43
65.67
0.35
0.18
0.79
1.58
65.4
0.44
0.13
0.90
K O
2
CaO
2.85
4.30
3.83
5.26
2.90
9.05
2.29
9.91
4.08
9.57
3.40
9.36
MnO 0.28 0.53 0.50 0.27 0.36 0.49
Fe O
2 3
BaO
0.19
n.d.
0.26
0.02
0.38
n.d.
0.32
0.02
0.43
0.05
0.50
0.02
PbO n.d. 0.01 n.d. 0.29 0.01 0.10
origin. The objects in cluster 3 can therefore in a direct way from Amsterdam belong to this group. For both clusters
be linked to local Antwerp production.
Na O+K O#20±2%. Cluster c and c have a similar com-
2 2 1 2
position to that of Antwerp clusters 3 and 4 (see Fig. 2), i.e.,
Comparison with literature data
with a composition similar to ‘Vitrum Blachum’ but with
Since in the Antwerp data one encounters both compositions
which are similar and those which are significantly different
from those found for the Venetian wares, it is relevant to
compare these data with the other published compositions of
Venetian/‘façon-de-Venise’ glass. In order to do so, the set of
122 Antwerp compositions was augmented with published
data series from regions geographically close to Antwerp.
Hierarchical cluster analysis was then performed on the large
data set. The first additional data set consisted of the 19
compositions published in Ref. 6 (Old Broad Street and City
of London); the second series consisted of seven soda-glasses
from Amsterdam5 and the third series are the French glass
compositions published by Barrera and Velde8 (about 100
pieces). In addition, the data published by Verità and
Toninato3,4 on truely Venetian glass were also included so as
to bring the total number of composition in the data matrix
to 254.
Fig. 5 shows the dendrogram obtained by clustering the
entire data set (again only the Na O, K O, CaO and MgO
2 2
concentration values were used). Table 3 lists the average
values of these oxide concentrations for each major cluster
shown in Fig. 5 (dotted lines). Clusters a and b include all the
‘Cristallo’ glass objects, the majority of the Venetian samples
belonging to cluster a with high Na O (17%) and low K O
2 2
(3%) content; cluster b contains less Na O, more K O and
2 2
about the same amounts of CaO and MgO. The Antwerp
glass objects with a ‘Cristallo’-like composition (see Fig. 2,
cluster 1) are divided over these two clusters, as are 30 French
vessels of similar composition. None of the London glasses
were found within these two clusters and only two objects
elevated K O (5–7% instead of 2–3%) The 19 Antwerp objects
2
of cluster 3, however, are now redistributed over clusters c
1
and e . Next to the Antwerp objects, one Venetian ‘Vitrum
2
Blanchum’ glass, most of the London glasses and 25 French
glasses belong to cluster c. Since this group also contains the
Antwerp glass marble, cluster c may be labelled ‘Antwerp
production’. The data in Table 3 again suggest that in London,
Amsterdam, various sites in northern France and Antwerp,
the same raw materials or bulk glass of the same origin were
used in the local glassworks or that glass vessels produced in
Antwerp were exported to these regions. Cluster d correspond
to the Antwerp cluster 2 in Fig. 1; also in France, a few objects
(five) with this high K O and relatively low Na O composition
2 2
are found. This mixed-ash group clearly corresponds to ‘non-
Venetian’ glass production. Whether this type of glass was
produced in Antwerp, however, cannot be established.
Clusters e and e show a fairly similar composition and
1 2
contain all the Venetian ‘Vitrum Blanchum’ glasses and the
Antwerp glass objects from clusters 3 and 5 (see Fig. 2).
Cluster e contains most of the Venetian ‘Vitrum Blanchum’
1
glasses, two London objects and 11 French glasses. Most of
the objects from Amsterdam belong to cluster e in addition
2
to some from London and a large group of French glasses.
When the dendrogram of the Antwerp glasses (Fig. 2) is
compared with that of the entire data set (Fig. 5), it appears
that largely the same data structure is found, i.e., in both
cases, in addition to both Venetian groups, a K O-rich ‘Vitrum
2
Blanchum’ variety and a small mixed-ash group is found.
It is also striking that both for ‘Cristallo’ and ‘Vitrum
Blachum’, two subgroups are formed (clusters a/b and e /e ),
1 2
where most of the Venetian wares are concentrated in one
(clusters a and e ) whereas the majority of the non-Venetian
1
finds tend to cluster together in the other subgroup (clusters
b and e ).
2
Conclusions
In view of the high price of finished Venetian products, it is
unlikely that all the Antwerp glass finds are Venetian products.
Either soda or sodium-rich ash (and possibly the other raw
materials for glassmaking), or soda-glass in bulk or finished
glass vessels must have been imported. In view of the presence
of the port, the cheapest and safest transportation route from,
e.g., Venice, Alicante or the Middle East to Antwerp, would
Fig. 5 Dendrogram obtained by hierarchical cluster analysis of the undoubtedly have been the sea. The aim of this investigation
total data set comprising analysis results from Antwerp, various sites was to determine whether there was a significant local pro-
in northern France, Amsterdam and London. duction of ‘façon-de-Venise’ glass vessels in Antwerp in the
J. Anal. At. Spectrom., 1999, 14, 493–498 497

Table 3 Mean values and standard deviations of some major element oxides within each of the clusters found in the large data set, comprising
compositions from Antwerp, various sites near Antwerp and Amsterdam, London, France and Venice
Average composition (% m/m) Number of objects in each cluster
Cluster Na 2 O K 2 O CaO MgO Antwerp Amsterdam London France Venice
Cluster a 17.2±1.1 2.9±0.5 4.8±0.8 1.68±0.38 9 — — 7 8
Cluster b 14.6±0.9 4.3±1.0 5.6±0.9 1.70±0.37 15 2 — 21 2
Cluster c
1
Cluster c
2
Cluster d
13.3±1.6
12.2±0.7
11.4±0.8
5.3±0.9
6.6±1.1
8.2±0.9
9.0±1.2
10.7±1.2
5.8±1.3
2.80±0.54
3.01±0.41
1.88±0.41
19
9
8
—
—
—
6
7
—
14
11
5
1
—
—
Cluster e
1
Cluster e
2
13.4±1.1
14.9±1.2
2.5±0.7
3.3±0.9
10.6±0.9
9.3±0.9
3.32±0.76
2.60±0.46
10
24
—
5
2
4
11
37
12
5
16th century and to elucidate how this local production can 2 K. H. Janssens, I. Deraedt, O. Schalm and J. Veeckman,
be distinguished from ‘real’ (i.e., imported) Venetian wares.
The analysis results for 122 ‘façon-de-Venise’ glass objects
excavated in Antwerp were considered. Clustering of the
Antwerp data revealed five compositional groups. Two groups
of glass objects were found that feature a typical Venetian
3
4
5
6
Microchim. Acta, 1998, 15, 315.
M. Verità, Rivi. Staz. Sper. Vetro, 1985, 1, 17.
M. Verità and T. Toninato, Rivi. Staz. Sper. Vetro, 1990, 4, 169.
J. Baart, personal communication, results obtained by P. McCray.
C. Mortimer, in Trade and Discovery: the Scientific Study of
Artefacts from Post-Medieval Europe and Beyond, Occasional
composition, one with a ‘Cristallo’-like composition and the Paper 109, ed. D. R. Hook and D. R. M. Gaimster, British
other closely resembling ‘Vitrum Blanchum’ glass. These Museum Press, London, 1995.
objects (58, or about half of the investigated vessels) were
probably imported from Venice as finished products. The three
other compositional groups contain glasses that were probably
fabricated locally. The composition of a waste glass marble
coincided with one of these groups. The differences between
these three groups may indicate that different types of alkali
7
8
9
H. Bronk, Chemisch-analytische Untersuchungen Fruhneuzeitlicher
Glaeser Mittel- und Sudeuropas unter Anwendung einer Quasi-
zerstorungsfreien Mikroprobenahmetechnik, Mensch und Buch,
Berlin, 1998.
J. Barrera and B. Velde, Archéol. Médiévale, 1989, 19, 81.
S. Ulitzka, in Venezianisches Glas der Sammlung der Veste Coburg,
ed. A.-E. Theuerkauff-Liederwald, Luca, Lingen, 1994, pp. 40–42.
were in use or it may represent a mixing of alkali types by the 10 P. Van der Wee, Bulletin van de Antwerpse vereniging voor Bodem-
glass workers. Since the ‘local products’ represent (slightly en Grotonderzoek, 1987, 3, 53–62.
more than) 50% of the investigated series of vessels, one can
conclude that an important fraction of the glass vessels have
a local origin. However, a profound study and comparison of
the trace composition of the Antwerp and Venetian glass
vessels is necessary for a definite answer concerning the origin
11
12
13
E. Ashtor and G. Cevidalli, J. Eur. Hist., 1983, 12, 475.
S. Denissen, Overzicht van de glasblazersfamilies te Antwerpen
tijdens de 17de eeuw, Bull. Antwerpse Vereni. Bodem-
Grotonderzoek, 1985, 5,9.
O. Schalm, K. Janssens and F. Adams, presented at EMAS’97,
11–15 May 1997, Torquay.
of the Antwerp glass vessels. The same non-Venetian composi- 14 I. Bondarenko, B. Treiger, R. Van Grieken and P. Van Espen,
tion was found in London, Amsterdam and various locations Spectrochim. Acta, Part B, 1996, 51, 441.
in northern France, which may be explained by export of
locally manufactured ‘façon-de-Venise’ vessels from Antwerp
to these nearby cities or by the use of similar fluxes or flux
mixtures to those in Antwerp.
15 H. E. Henkes, Glass without gloss: utility glass from five centuries
excavated in the low countries, 1300–1800, Rotterdam Papers 9,
Rotterdam, 1994.
References
1 L. Engen, Het Glas in Belgïe, Mercatorfonds, Antwerp 1989,
pp. 121–133.
498 J. Anal. At. Spectrom., 1999, 14, 493–498
Paper 8/08385A