BLACK GLASS JEWELLERY FROM BRACARA AUGUSTA
Mario da Cruz
what evidences of local glass production really
exist?
There is in Bracara Augusta a well documented
production of adornment objects in opaque black
glass from the Late Roman period. Due to the lack
of data from the excavations, it is for the time
being not quite possible to forward more precise
data as to the chronology of the different jewels.
Recent studies in archaeometry (Gratuze, 2001a;
Gratuze, 2001b), however, are demonstrating that
some of the black jewellery, for long time
classified as Iron Age or even Bronze Age is in
fact Roman.
The modern Portuguese town of Braga is an
Augustan foundation and capital of one of the three
Roman conventus of the northwest of the Iberian
Peninsula. Under Diocletian these three conventus
were detached from the province of Tarraconensis
to form the new province of Callaecia with
Bracara Augusta as its capital. This promotion
started a period of great development and
prosperity that explains the stimulus given to
commerce and craft activities in the city, among
which stands out the local glass production, of
which black glass jewellery is a small part.
The opaque black glass is a clear imitation
of jet; a semiprecious material very popular in the
region in Roman times. As far as we know, the
black glass from Braga was exclusively destined to
manufacture small jewellery objects, along with a
marginal production of gaming counters, hence
there was not a single black vase recorded Cosyns
and Hanut (2005) makes clear that the production
of vessels in black glass ended around the middle
of the 3rd century AD, about a generation before
the start of the Braga hey-day). This is not true
vice versa, however. Indeed, although the great
majority of jewels found in Braga are of black
glass, there are also a few examples in other
colours, mainly finger rings and beads. Opaque
coloured glass, like yellow and blue, was also used
for decorative details.
But, before describing the Braga material,
PRODUCTION
In the Southwest part of Bracara Augusta, in an
area commonly known as Fujacal, production
waste containing black glass can easily be
identified within the bulk of cullet. Preliminary
studies point in the direction of a profound urban
restructuring in this area, at the time of the Late
Roman wall construction by the last third of the
3rd century AD. With the exception of the
workshop, it is possible to identify all the
evidences of black glass jewellery production.
Starting from the raw material, there are
several lumps of black opaque glass (Fig. 1) that
preserve the shape and even chips of the crucible.
Fig. 1. Lumps of black glass.
The so-called “opaque black glass” is in
reality neither opaque nor black when watched
7
AIHV Annales du 17e Congrès, 2007
closely (J. Bayley in Guido 1999; Cosyns et al.
2006). To be considered as true opaque the
presence of opacifying agents in the form of
crystals is required which is not the case here. On
the other hand, if we hold a thin walled chip
against the light, we can see some colour and light
coming through. In the case of the black glass
used in Bracara Augusta it takes on dark green
tones, but purple, blue and brown tones are also
known (Cosyns, 2005, 113). The opaque black
appearance is given by a saturation of colouring
agents. Comparative analysis of the chemical
composition of black glass and natural coloured
blue-green glass from the excavations in Fujacal,
Braga (Table 1) revealed that the colouring agent
used was iron oxide (Fe2O3), as in most of other
known cases (Gratuze, 2001a; Gratuze, 2001b).
Some elements, like Copper (CuO) and Lead
(PbO), show an unusual high representations and
others, like calcium (CaO) and alumina (Al2O3),
are of lower content.
The different composition of black glass
from Braga and S. Vito suggests that the colouring
of black glass was done locally.
Dealing with small quantities, the fusion was
made in small conical crucibles found with black
glass in its interior (Fig. 3, 0.0.1).
The jewels are of simple crafting. Be it in
the case of bracelets, or rings, the technique
consisted of dragging and enrolling more or less
broad strings like 0.2.1 and 0.2.3 (Figs. 2 and 3),
around a stick with the desired diameter, fixing
their ends and diluting this seam. In bracelet 1.1.3
and finger rings 2.2.1 and 2.2.3 this seam is very
clear. The square beads would follow a similar
principle, in which the string is bent and palmed
around two little irons at the origin of the two
perforations, as you can see in 3.1.3 and 3.1.5.
Table 1. Comparative chemical compositions for black
and blue-green glass from Braga and a black grooved
bracelet from S. Vito al Tagliamento, Italy (Gratuze,
2001a, no. 23). Samples BRA26 and BRA29 from
Braga analysed by Bernard Gratuze in IRAMAT,
Institut de Recherche sur les Archéomatériaux, Centre
Ernest Babelon, C.N.R.S., Orléans. Our special thanks
to them.
Na2O
MgO
Al2O3
SiO2
P2O5
Cl
K2O
CaO
MnO
Fe2O3
CuO
PbO
TiO2
Black
BRA26
16.7%
0.79%
2.30%
60.4%
0.10%
1.01%
0.39%
5.06%
1.01%
11.4%
0.19%
0.16%
0.11%
Blue-green
BRA29
14.8%
0.54%
3.12%
70.4%
0,07%
1.11%
0.30%
8.13%
0.71%
0.39%
0.0025%
0.0006%
0.0723%
Black
N. 23
15.3%
0,.7%
2.65%
64.0%
0.20%
0.97%
0.52%
7.35%
0.76%
7.28%
0.056%
0.030%
0.08%
In the case of the spiral-like cylindrical
beads, the process was even simpler and consisted
of rolling a thread around an iron. In the case of the
remaining beads, the technique would be pulling
little “pips” or drops, like 0.2.3 and 0.2.4 (Figs. 2
and 3), which would later be perforated and
crafted.
There are basically two decorative
techniques, both hot crafting: the first one
consisted of printing simple furrows or stamps; the
second was the application of threads of the same
or a contrasting colour.
THE JEWELS
A catalogue of all jewellery in black glass
excavated in Braga are listed at the end of this
paper. A selection of only the most representative
material is illustrated in Fig. 3.
The bracelets represent, with 70 % of the
registered jewels, the essential part of the
production, and were of daily and widespread use.
Among these the plain ones are the more abundant
and variable in size, closely followed by the
grooved ones.
This is certainly due to its
Fig. 2. Glass production waste: drops and strings.
8
Roman Glass
Fig. 3. (Scale ½) Black glass jewellery finds: 0.01 - 0.3.4 evidences of glass production; 1.1.1 - 1.4.1 bracelets; 2.2.1 2.2.6 finger rings; 3.1.1 - 3.5.2 beads; 4.1.1 - 4.2.2 pendants/amulets; 6.1 rough gemstone (for further information go
to the list of black glass jewellery finds).
simplicity and easy crafting. Less abundant and
more elaborate are the stamped ones like the two in
Fig. 4. These examples are the only intact ones
found in Braga but unfortunately have disappeared
(Sousa, 1973, fig. 2).
If the daily use of these fragile bracelets is in
itself already bizarre, although irrefutable because
of the intense patterns of wear shown, the use of
finger-rings may have had a more sporadic and
ritual character, associated with burials.
Among the various complete beads, the
square ones are the most original, not so much with
regard to their form and decoration but to the fact
that they present two longitudinal perforations. In
addition to the evidence that they only show
Fig. 4. Bracelets and pin. Unknown whereabouts.
patterns of wear on the decorated external side,
allows us to argue that these are segments of
9
AIHV Annales du 17e Congrès, 2007
bracelets and not collar beads. Parallels for these
square beads, as for bracelets, are found in
different parts of the Roman Empire but not always
correctly classified or dated (see jet material
Allason-Jones 1999; Hagen 1937).
The spherical plain beads most likely did not
have much success, despite their simplicity and
proven local production. The cylindrical beads,
plain or spiral-formed, still pose an enigma. If on
the one hand they are relatively simple forms of
easy crafting, on the other hand they are limited to
a single piece of each variant. Putting into relation
those facts, together with the lack of wear and their
shared origin in Fujacal, we might be in the
presence of prototypes which were not
commercialized.
The spherical pendants as well as the
elongated ones can be classified as amulets. The
spherical ones probably represented fruits: grapes,
apple, pomegranate, etc. As to the elongated
trailed amulets, it is difficult to determine which
object they imitate, or even their exact function.
Their use as pendants of a collar is not very
probable, due to the fact that they show a sole
decentralized loop, in which they differ from
known parallels (Nolen, 1994, vi-129 and vi-130).
The whereabouts of the only known pin,
which is merely registered by the picture in Fig. 4,
is unknown, but the colouring and the link with the
two bracelets situate it indisputably in the sphere of
production of black glass from Braga. What was
argued referring to glass rings is also true for this
unique piece, which imitates from close up
common bone pins.
tesserae and also used for decorating cups and jars
with glass trails and threads.
It is obvious that a production of false gems
only makes sense in Braga if there was a market
for these fakes (a “cluster”), i.e., one or more
workshops of gem cutters and engravers. We think
that this workshop could be situated on the Insula
of Carvalheiras, from where three out of five
imitations of nicolos were found, as well as other
intaglios. Curiously, it is from here the major
quantity and variety of jet objects ever found in
Braga comes. Most likely we are in the presence
of a cutting and engraving workshop of jet,
gemstones and their imitations in glass, acquired
locally.
Hence, without the rough gem, imitating
nicolos (Figs. 5 and 3, 6.1) found in Fujacal all this
would be mere speculation. This important find
can be seen as a real “missing link” connecting the
glass production from Fujacal, and the engraving
of gemstones from Carvalheiras. To strengthen
this idea, there emerged, once again in Fujacal, one
single genuine nicolos intaglio (Fig. 4, bottom
left). Could it have served at any moment as a
model?
GLASS IMITATION OF NICOLOS
If the production of black glass in Braga was
aiming at the imitation of jet, it is only natural that
the “falsification” temptations did not end here and
were extended to precious stones, an intent which
after all can not be dissociated from the appearance
of glass itself. It is also natural that one of the
stones to be imitated was nicolos (a variety of
agate), due to the following reasons. First of all
because it was one of the most popular and
appreciated stones by the Romans, and then
because it is relatively simple to forge. In the case
of Braga, the two colours necessary for their
execution already existed in form of glass lumps:
the opaque black of the remaining jewels and the
opaque blue of the ones used for the cutting of
10
Fig. 5. From top to bottom: rough gem and similar
crafted, genuine nicolo and similar glass imitation.
REFERENCES
Allason-Jones, L., 1999. The Roman Jet in the
Yorkshire Museum. York.
Bayley, J., 1999.
Appendix notes on the
composition of coloured glasses. in: Guido, M.
The glass beads of Anglo-Saxon England c. AD
Roman Glass
Gratuze, B., 2001a. Etude de la composition de
bracelets protohistoriques en verre du Museo
Civico “Federico de Roco” (S. Vito al
Tagliamento, Italie). Orléans, C.N.R.S. (PDF
file)
Gratuze, B., 2001b. Etude de la composition d’un
écarteur de collier en verre de l´Aveyron
(Collection B. Tournier, Musée de Gap).
Orléans, C.N.R.S. (PDF file)
Hagen, W., 1937. ‘Kaiserzeitliche Gagatarbeiten
aus dem rheinischen Germanien’.
Bonner
Jahrbücher 142, 77-144.
Nolen, J., 1994. Cerâmicas e vidros de Torre de
Ares. Lisboa, IPM.
Sousa, J.J.R., 1973. ‘Subsídios para a carta
arqueológica de Braga’. Studia Archeologica
23, 7-29.
400-700. A preliminary visual classification of
the more definitive and diagnostic types.
London, The Society of Antiquaries of London,
89-93.
Cosyns, P. and Hanut, F., 2005. ‘Black glass of
second to third-century date in northern Gaul: a
preliminary survey’.
in: Annales du 16e
Congrès, de l’AIHV, 113-118.
Cosyns, P., Janssens, K., Schalm, O. and Van der
Linden, V., 2006. ‘Black glass in the Roman
Empire: a work in progress’. in: Creemers, G.,
Demarsin, B. and Cosyns, P. (eds.) Roman
glass in Germania Inferior.
Interregional
comparisons and recent results. Proceedings of
the International conference (Tongeren 2005),
Atuatuca 1.
Tongeren, Provinciaal GalloRomeins Museum, 30-41.
BLACK GLASS JEWELLERY FINDS
N. / Form - Inventory no. (site reference) Measurements (thickness = T.; length = L.; width = W.; diameter = D.) /
Remarks
0 Evidences of glass production
0.1 Crucible:
0.1.1 - 0959/00 (BRA98FUJ Q1 =141=) Small conical crucible with added black glass
0.2 Waste:
0.2.1 - 0495/01 (BRA98FUJ M,N/25,26 =12=) O-shaped section rectilinear plastic cord
0.2.2 - 0498/01 (BRA98FUJ Q15 =141=) O-shaped section rectilinear plastic cord with a central perforation in the
thickest end
0.2.3 - 0449/01 (BRA98FUJ Q20 =16=) No shape nugget/drop with sand marks
0.2.4 - 0474/01 (BRA98FUJ I20,21,22,23 =59=) No shape nugget/drop with sand marks
0.3 Rejections:
0.3.1 - 0475/01 (BRA98FUJ K21,22,23,24=30=) Flat square bead with tooling marks and an obstructed perforation
0.3.2 - 0503/01 (BRA00FUJ Q18 =34=) Misshapen flat square bead with an obstructed perforation
0.3.3 - 0007/06 (BRA96CSF 16 =3=) Unfinished spherical bead
0.3.4 - 0493/01 (BRA98FUJ M,N/25,26 =11=) Distorted finger ring with tool mark
1 Bracelets (all have D-shape section)
1.1 Plain:
1.1.1 - 1162/04 (BRA94SG 405 =3=) D.: 66 mm; T.: 7x15 mm. Intense wear
1.1.2 - 0966/95 (BRA95CSC 6 =1=) D.: c.60 mm; T.: 3.5x6 mm. Little wear
1.1.3 - 0350/94 (BRA93FUJ H12 =67=) D.: c.60 mm; T.: 5x7 mm. Intense wear. Protuberance from the seam
1.1.4 - 1808/92 (BRA92CARV 60 s/c) D.: 60 mm; T.: 5x6.5 mm. Intense wear. Seam visible inside
1.1.5 - 0128/96 (BRA95MIS F13 =13=) D.: c.60 mm; T.: 4x10 mm. Intense wear
1.1.6 - 0954/97 (BRA78SH 2) D.: c.60 mm; T.: 4x14.5 mm. Worn
1.1.7 - 1422/97 (BRA77P 4a) D.: c.70 mm; T.: 5x7 mm. Intense wear
1.1.8 - 0076/00 (BRA96ECS 35 «3») D.: c.70 mm; T.: 5x6 mm. Intense wear
1.1.9 - 0953/97 (BRA78S 1) D.: c.60 mm; T.: 4x7 mm. Intense wear
1.1.10 - 0951/97 (BRA77Q 1), D. indeterminable; T.: 6.5x11 mm. Intense wear
1.1.11 - 0944/97 (BRA77A 6d) D.: 70 mm; T.: 7x8 mm. Soft wear. O section, not D
1.1.12 - 0064/00 (BRA91CVL 91) D.: 70 mm; T.: 6.5x10 mm. Worn
1.1.13 - 0077/00 (BRA77A 7B) D.: 70 mm; T.: 5x7 mm. Worn
1.1.14 - 0078/00 (BRA77E 7) D.: c. 50.5 mm; T.: 5x8 mm. Little wear
1.1.15 - 0078/97 (BRA91CARV31 «fossa») D.: 80 mm; T.: 5x7 mm. Worn
1.1.16 - 1385/96 (BRA92CARV 71) D.: 70 mm; T.: 6x11 mm. Intense wear
1.1.17 - 0150/97 (BRA95T 19b =0=) D.: 70 mm; T.: 6x15 mm. Intense wear
11
AIHV Annales du 17e Congrès, 2007
1.1.18 - 0004/06 (FALPERRA70I 3E6) D.: 75 mm; T.: 6x15 mm. Worn
1.1.19 - 0222/92 (BRA92CARV 78 =2=) D.: 55 mm; T.: 2.5x3 mm. Lack of wear. Extremely thin
1.1.20 - 1384/96 (BRA91CVL x164y156) D.: 65 mm; T.: 5x7 mm. Little wear
1.1.21 - 1459/98 (BRA97ALB E7 =156=) D.: 65 mm; T.: 5x6,5 mm. Little wear
1.1.22 - 0071/00 (BRA97FUJ Q21 =109=) D.: 60 mm; T.: 7x10 mm. Intense wear
1.1.23 - 0496/01 (BRA97FUJ Q21 =65=) D.: 65 mm; T.: 6x7 mm. Little wear
1.2 Grooved:
1.2.1 - 0072/00 (BRA84CARV x152y132=1391=) D.: c. 60 mm; T.: 5x9 mm. Intense wear
1.2.2 - 0075/00 (BRA66MAX X) D. Indeterminable; T.: 5x10 mm. Little wear
1.2.3 - 1204/96 (BRA96SST K20 =25=) D.: 80 mm; T.: 5.5x9-11.5 mm. Worn. Irregular thickness
1.2.4 - 1205/96 (BRA96SST K20 =38=) D.: 70 mm; T.: 6x12 mm. Soft wear. C shaped arch grooves
1.2.5 - 0945/97 (BRA77A 6G) D.: 60 mm; T.: 5x9.5 mm. Worn. Spaced deep grooves
1.2.6 - 0955/97 (BRA88CVL x136y160 =2=) D.: 50 mm; T.: 6.5x12 mm. Worn. Deep grooves
1.2.7 - 0950/97 (BRA77B 16) D.: 70 mm; T.: 4.5x7,5 mm. Intense wear. Thin oblique grooves
1.2.8 - 1278/97 (BRA79Pb 10d «2») D.: 45 mm; T.: 4x8 mm. Intense wear. Closed and shallow grooves
1.2.9 - 0065/00 (BRA77A 25C) D.: 60 mm; T.: 7,5x13 mm. Intense wear. Grooves with a tool perforation mark
1.2.10 - 0066/00 (BRA98T 118 =2=) D.: 60 mm; T.: 5x9 mm. Worn
1.2.11 - 1012/04 (BRA04TR UE500) D.: 65 mm; T.: 5x11 mm. Worn
1.2.12 - 0946/97 (BRA77A 10A s/c) D.: 60 mm; T.: 5x10 mm. Worn. Silky inside surface
1.2.13 - 1421/97 (BRA77P 4A s/c) D.: 65 mm; T.: 4.5x8 mm. Worn. Shiny surface
1.2.14 - 0074/00 (BRA78C 1 s/c) D.: 50 mm; T.: 5.5x9,5 mm. Intense wear. Deep closed grooves
1.2.15 - 0004/02 (BRA88PAÇO E2 =323=) D.: 65 mm; T.: 4x9.5 mm. Intense wear. Soft irregular grooves
1.3 Stamped:
1.3.1 - 0073/00 (BRA94SG203 =3=) D.: 60 mm; T.: 6.5x11,5 mm. Worn. Oval stamping
1.3.2 - 0127/96 (BRA95MIS F13 =13=) D.: 65 mm; T.: 10x14 mm. Intense wear. Oval stamping
1.3.3 - Not found (BRA X) Whole stamped bracelet. Unknown whereabouts (Fig. 4)
1.3.4 - Not found (BRA X) Whole stamped bracelet with central depression. Unknown whereabouts (Fig. 4)
1.4 Trails:
1.4.1 - 0220/92 (BRA92CARV 74 =7=) D.: 80 mm; T.: 8x10.5 mm. Worn. Fused opaque yellow sneak-thread
trails
1.4.2 - 0472/01 (BRA98FUJ I20,21,22,23 =59=) D.: 75 mm; T.: 5x9 mm. Worn. Fused opaque green sneak-thread
trails
2 Finger rings
2.1 Simple:
2.1.1 - 0541/97 (BRA96SE V31 «sep. 6») D.: 18 mm; T.: 1.5x3 mm. Little wear. Complete
2.1.2 - 0926/97 (BRA97Tb10I =2=) D.: indeterminable; T.: 3 mm. O-shaped section, irregular thickness
2.2 – With a central setting:
2.2.1 - 0003/00 (BRA00MAX Sep1) D.: 16 mm; T.: 2x4.5 mm. D-shaped section, central setting showing a rosette
shape stamp
2.2.2 - 0002/ 00 (BRA00MAX Sep1) D.: 17 mm; T.: 2x4.5mm. Soft wear. D-shaped section, central setting
missing
2.2.3 - 0484/01 (BRA98FUJ L13 =102=) D.: 17 mm; T.: 2x4 mm. Intense wear. D-shaped section, small and rude
central setting
2.2.4 - 0952/97 (BRA78Pb 1 «01») D.: 15 mm; T.: 3 mm. Worn. O-shaped section, small round central setting
2.2.5 - 2154/99 (BRA97TB 10I) D. indeterminable; T.: 3.5 mm. Little wear. O-shaped section, small round central
setting
2.2.6 - 0067/00 (BRA77Q 2) D.: 15 mm; T.: 3x7 mm. D-shaped section, oval central setting
3 Beads
3.1.Flat square with double perforation:
3.1.1 - 0276/97 (BRA76GE 451) W.: 11 mm; L.: 11 mm; T.: 4 mm. Little wear. Net of 4x2 grooves
3.1.2 - 0202/0 (BRA93CARV98) W.: 11 mm; L.: 11 mm; T.: 5 mm. Worn. Net of 7x1 grooves
3.1.3 - 0453/01 (BRA97FUJ G22 =22=) W.: 17 mm; L.: 17 mm; T.: 5 mm. Worn. Net of 3x6 grooves
3.1.4 - 0490/01 (BRA98FUJ M,N25,26 =11=) W.: 15 mm; L.: 14 mm; T.: 4.5 mm. Lack of wear. Net of 5x2
grooves
3.1.5 - 0088/03 (BRA02CARV Leste25 503) W.: 15 mm; L.: 13 mm; T.: 4 mm. Worn. Net of 7x2 grooves
3.2 Plain spherical:
3.2.1 - 1243/96 (BRA96CSF 12 =6=) W.: 13 mm; L.: 13,5 mm. Little wear
3.2.2 - 0227/92 (BRA92CARV 82 =1=) W.: 8,5 mm; L.: 5 mm. Little wear
12
Roman Glass
3.3 Decorated spherical:
3.3.1 - 0502/01 (BRA00FUJ Q18 =36=) W.: 12.6 mm; L.: 14 mm. Split surface. Opaque yellowish white zigzag
3.3.2 - 0016/02 (BRA X) W.: 19 mm; L.: 14 mm. Split surface. White and light blue eyes
3.3.3 - 0942/97 (BRA MAX) W.: c.15 mm; L.: 8 mm. Fused irregular opaque white trail
3.4 Cylinder:
3.4.1 - 0501/01 (BRA00FUJ Q18 =34=) W.: 4 mm; L.: 10.8 mm. Little wear. Square section
3.4.2 - 0494/01 (BRA98FUJ M,N/25,26 =11=) W.: 5.5 mm; L.: 10 mm. Lack of wear. O section
3.5 Thread cylinder:
3.5.1 - 0491/01 (BRA98FUJ M,N/25,26 =11=) W.: 6 mm; L.: 9.5 mm. Lack of wear. Spiral twisted thread
3.5.2 - 0492/01 (BRA98FUJ M,N/25,26 =11=) W.: 4.5 mm; L.: 11 mm. Lack of wear. Thin spiral twisted thread
4 Pendants/Amulets
4.1 Spherical pendants:
4.1.1 - 0008/06 (BRA93FUJ H12 =51=) W.: 20 mm; L.: c .30 mm. Worn. Broken loop
4.1.2 - 0073/05 (BRA X) W.: 24 mm; L.: c.32 mm. Lack of wear. Broken loop
4.1.3 - 0009/06 (BRA96SST H14 =8=) W.: 18 mm; L.: c.23 mm. Intense weathering. Broken loop, navel
4.1.4 - 0309/06 (BRA06TR.PSN303UE712) W.: 19 mm; L. 23 mm. Worn yet intact
4.2 Trailed amulets:
4.2.1 - 1065/96 (BRA96SST H12 =8=) W.: 11 mm; L.: 44 mm. Worn. Split spiral opaque yellow trails
4.2.2 - 0930/97 (BRA95SG 402 =7=) W.: 7 mm; Preserved L.: 30 mm. Little wear. Spiral trails.
5 Pin
5.1 - Not found (BRA X) Pin with faceted head. Unknown whereabouts (Fig. 4)
6 Man-made rough gemstone
6.1 - 0482/01 (BRA98FUJ L13 =101=) W.: 10 mm; T.: 3.5 mm. Black and opaque blue rough gem, imitating
nicolos.
MARIO DA CRUZ
Largo de Santa Barbara 4, 3 esq., PT-1150-287 Lisbon, Portugal
e-mail: mariodacruz@hotmail.com
13