BEAD FURNACES IN ANGLO-SAXON BRITAIN An investigation into the likely sources of heat used by beadmakers in the 5th to 7th centuries CE

Abstract

While many thousands of glass beads have been excavated from early medieval and Anglo-Saxon sites in Britain, no complete furnaces or
bead workshops have so far been found. This paper investigates the findings of many studies on the subject of ancient furnaces and applies an understanding of hot glass beadmaking techniques to suggest the likely heat sources that would have been used for beadmaking. Illustrations of the types of furnace studied give a clear view of the tools and techniques of the ancient beadmakers. A case is made that Anglo-Saxon beadmakers used mainly lampwork techniques to make their finely decorated polychrome beads and these required small heat sources such as volcano furnaces and open hearths. Larger furnaces are used for beads made in quantities and
many imported beads from Europe and the Near East are of this type.

Full text

1
BEAD FURNACES IN
ANGLO-SAXON BRITAIN
An invesgaon into the likely sources of heat used by beadmakers
in the 5th to 7th centuries CE
Sue Heaser
2022
© Sue Heaser, Text, illustraons and photographs
Independent Archaeological Researcher
The Farthings, Long Green, Wortham, Suolk IP22 1PU UK
sueheaser@live.co.uk
Tel: 01379 898479
www.sueheaser.com
ABSTRACT
While many thousands of glass beads have been excavated from early medieval and Anglo-Saxon sites in
Britain, no complete furnaces or bead workshops have so far been found. This paper invesgates the
ndings of many studies on the subject of ancient furnaces and applies an understanding of hot glass
beadmaking techniques to suggest the likely heat sources that would have been used for beadmaking.
Illustraons of the types of furnace studied give a clear view of the tools and techniques of the ancient
beadmakers. A case is made that Anglo-Saxon beadmakers used mainly lampwork techniques to make their
nely decorated polychrome beads and these required small heat sources such as volcano furnaces and
open hearths. Larger furnaces are used for beads made in quanes and many imported beads from Europe
and the Near East are of this type.

2
INTRODUCTION
One of the essenal requirements of hot glass beadmaking is a source of heat to melt
glass. Soda lime glass of the type used by ancient beadmakers has a working
temperature of between 750°C and 1000°C so any heat source needs to reach these
temperatures. This is lower than the temperatures needed for glassblowing (see table
below).
The queson is, what type or types of furnace would the Anglo-Saxon beadmakers
have used? No specic beadmaking furnaces have so far been found in Brish
excavaons of Anglo-Saxon sites but there are several examples of glass furnaces that
were probably used for glassblowing. It is important to realise that beadmaking and
glass vessel making are very dierent cras and require quite dierent skills even
though they use the same material. Therefore a beadmaker may not use the same
design of furnace as a glassblower making vessels. Bead furnaces could be much
smaller aairs and experiments have shown that a beadmaker can even work over
an open hearth (Krzyżanowska, M and Frankiewicz, M., 2015). This does not mean that beadmakers would not have
used larger furnaces, but suggests that beadmaking heat sources could be more improvised, varied, and smaller.
SIGNIFICANT TEMPERATURES OF SODA LIME GLASS
Glass making temperatures c. 1675°C
Glassblowing temperatures 1000°C to 1700°C
Glass melng 1000°C
Glass soening 700°C
Lampwork glass working temperature range 750°C – 1000°C (930°c TYPICAL)
Glass annealing temperature 540°C
Blacksmith’s forge for working iron c. 1000°C
Temperature to melt bronze 913°C
TERMINOLOGY
There is some confusion in archaeological literature as to the denion of the words “furnace”, “kiln” and “oven”. Here
is a brief claricaon:
Furnace: A container that is heated to high temperature so that substances placed inside will melt or burn. Furnaces
usually have openings or glory holes so that the hot chamber within can be accessed while it is hot. Used for melng
metals and for glass working.
Kiln: an oven, furnace, or heated enclosure used for processing a substance by burning, ring, or drying. Usually used
for clay or poery. Kilns tend to be closed before heang and only opened again aer the process is nished and the
kiln has cooled down.
Oven: In English an oven is a more general term and can also mean a domesc cooking oven. Best avoided when
referring to glass working or ceramics. Translaons from European languages into English oen make this mistake. A
glass furnace or poery kiln is never called an oven in English.
Tools:
Mandrel: a pointed iron tool on which beads are wound.
Gathering iron: a long iron tool used to gather molten glass from a furnace.
Fig. 1. A twisted cable bead from
Eriswell, Suolk - a type found
only in Britain.
Courtesy of Cotswold Archaeology, Suolk

3
1. Lampwork winding
This ancient technique is used all over the world and has a long history. It is also called
“lampworking” or “ame working”. The heat source used is tradionally an open ame (or
lamp). The beadmaker uses rods of glass or small pieces such as tesserae which are heated by
holding them over an exposed ame unl they become molten. The melted glass is then
wound round a mandrel to make the bead. Further shaping is achieved by returning the bead
into the ame frequently to keep it at a semi-molten consistency, then moving it just out of
the ame to shape and decorate it.
Tools for lampwork beadmaking are simple and few: mandrels which can double as a
gathering iron, tongs, tweezers, a needle tool and a knife are all that are needed.
THE TWO MAIN METHODS OF MAKING WOUND BEADS
The clue to the type of heat source may be found in the beads themselves. Many early medieval beads were made by
winding and there are likely to be two main heat source methods used for this, lampwork winding and furnace
winding. Kanungo (2001) describes these two dierent methods of making beads in India and these ancient techniques
are sll pracsed in tradional bead workshops around the world. These techniques would also have been used in
early medieval mes because it is a case of the material driving the arsan and the material (the glass) has hardly
changed.
Other types of beads give further clues to the likely heat source—drawn beads and mosaic beads that use complex
milleori canes and rolled pad techniques are more likely to require a furnace because components are usually
produced in quanty or require larger gathers of hot glass. These types are considered to be imported into Britain
from workshops in the Near East.
Fig. 3. A hooked tool is used to pull up a peak of the molten glass.
The peak will stand proud of the surface of the glass for several
seconds to enable the next step.
Fig. 4. Before the peak slumps back into the mass of hot glass, a
hot mandrel is pressed against the side of the peak and turned so
that the glass is wrapped round the mandrel. Liing the mandrel
will pull the glass thread thin and sever it.
2. Furnace winding
Here the glass is melted in crucibles or on slabs inside an enclosed furnace. The beadmaker reaches a hooked tool or
gathering iron into the furnace from an opening in the side and draws up a peak of melted glass. Before the peak
subsides back into the molten mass, a heated mandrel is pressed against the side of the peak and rotated to wind the
glass round it to create a bead. It can then be further shaped and decorated as required and this usually happens in the
mouth of the opening in the furnace (called the glory hole) so that the heat can be controlled by moving the bead into
and out of the heat. This technique is associated with larger workshops because furnaces require more fuel and
manpower to keep them fed. These larger furnaces can have several workstaons arranged around the outside of the
furnace to maximise beadmaking.
Beadmaking tools used with a larger enclosed furnace are much the same as those used for lampwork beadmaking
with the excepon of the mandrel and gathering iron. These tools will need to be much longer. The heat from a
furnace will be greater than with a volcano furnace or open hearth and the beadmaker will need to work further away
from the heat.
Fig. 2. Lampwork
technique: Winding a bead
held above an open ame.

4
TYPES OF BEADMAKING FURNACES
A. ENCLOSED FURNACES
1. Examples from present day bead workshops
Examples of this type of furnace are found in beadmaking workshops in India, Nepal and Turkey. It is considered highly
likely that these simple furnaces have been used for hundreds of years and can give us valuable clues as to how beads
were made in anquity.
The furnaces are usually domed, with a hole for the fuel at the boom and openings in the side (called glory holes) for
accessing molten glass inside the furnace and for making beads. The glass is usually in the form of pre-made coloured
broken chunks that are placed into large crucibles or tanks on a ledge inside the furnace. When the furnace is red, the
glass melts.
Fig. 5. A beadmaking furnace from Nazarkoy, Turkey. Three work staons
and annealing ovens in the protrusions. One of the recesses will contain
the fuelling opening. (From Nenna and Pichot 2008.)
Fig. 6. Beadmaking furnace from Nepal. Four work staons.
(From Garborieau 1977.)
Anglo-Saxon: No traces of furnaces with associated nds of beadmaking waste have so
far been found in Britain from Anglo-Saxon excavaons. This means that we need to rely
on nds from glassblowing sites, mostly of later date; descripons in early literature; and
Roman and Iron Age furnace evidence. Several furnace bases have been excavated at
Glastonbury in Somerset and these have been dated to c. 680 CE (Wilmo and Welham
2013). Crucibles were found with the furnaces but the waste glass suggests that they
were used for making vessels and not beads but they can sll give useful informaon on
how bead furnaces might have been made and used.
Iron Age, Hellenisc and Roman: Archaeological excavaons of furnaces
such as at the Iron Age site of Culduthel, Scotland (Hatherley and
Murray 2021) are usually limited to the base levels so that the upper
structures and chimney types have to be conjectural. A relief of
glassblowers on rst century Roman clay lamps found in Croaa, Italy
and Slovenia gives a clue to the shape (Fig. 8). Hellenisc ceramic kilns
are more complete and give more informaon although they are for a
dierent cra (Hasaki 2006).
Other sources: Illustraons from medieval manuscripts from 1000 CE
onwards (Fig. 7) show furnaces for glassblowing with workstaons
arranged around the perimeter for mulple arsans to work at the same
me. While the shapes are similar to those used today in beadmaking workshops in
Turkey, India and Nepal, the furnaces are much larger than those used for making beads.
Fig. 7. Medieval glassblowing
furnace from Agricola 1556
Fig. 8. Roman
lamp showing
glassblowing
2. Archaeological and Historical Evidence

5
3. Replica Furnace Experiments
There are many ongoing experiments in Britain and Europe with
experimental archaeologists combining research of ancient examples
with the designs of modern furnaces from arsan workshops in Asian
countries where resources are limited. The resulng furnaces are
tested by glass working arsans. Hill and Taylor (2008) have built
experimental Roman furnaces and these give valuable informaon
about furnaces that are likely to be similar to those used by the Anglo-
Saxon glass workers who followed. Other experiments are on-going in
Germany (Wiesenberg 2018).
The results suggest that the furnaces had a re chamber below with a stoke
hole for adding fuel; ledges built inside for holding crucibles or slabs for melng glass; and openings in the side for
the glass workers to access the molten glass (see Fig. 9). They would have been fuelled with large quanes of wood.
This type of furnace could be used by a beadmaker who would melt dierent colours of glass in crucibles inside the
furnace and then using the glory hole opening in the side of the furnace, dip into them as required. The glory hole
would also provide a workstaon for creang the beads. Other smaller openings could be used as annealing ovens.
Fig. 9. A replica Roman glass-blowing
furnace from Hill and Taylor 2008.
B. OPEN HEAT SOURCES
The lack of evidence of larger kilns from the Anglo-Saxon period suggests that the beadmakers may have used more
ephemeral heat sources. Using lampwork techniques, beads can be made without an enclosed furnace using a
source of heat from below. Modern beadmakers use gas-fuelled blowtorches for this technique. The glass is melted
in the open ame as required for each bead and not pre-melted inside a furnace. The hot ame is then used to
maintain the glass in a molten state while the bead is made and decorated. Recent years have seen a variety of
experiments of this type of beadmaking and they are described below. All have been shown to work for beadmaking.
1. Open Hearth
Experiments with recreang open hearths to use for
beadmaking have shown that this is a likely alternave to a
Volcano furnace for small scale arsan producon
(Krzyżanowska and Frankiewicz 2015). The hearths built were
based on nds at Ribe and were a clay rectangle of about
50cms x 30cms. A pair of bellows was required with charcoal
as fuel and two workers were necessary: one to make the
beads and the other to tend the bellows and refuel. A
temperature of over 1000°C was achieved and a variety of
beads were made. The remains of an open hearth heat source
for beadmaking would leave a very small footprint that could look
like a domesc hearth.
Fig. 11. Open Hearth experimental furnace aer
Krzyżanowska and Frankiewicz
Pung the evidence together
Figure 10 shows the cross secon of a possible internal
arrangement of a small domed furnace used for making
glass beads. The boom is made from slabs of stone or
le and the dome and sides are made of clay but it could
be made purely from clay built over wooden former that
is then removed. A ledge inside the kiln is for placing
pots (or crucibles) of crushed glass and the glory hole
allows the beadmaker to reach inside to scoop up the
molten glass on a gathering iron or mandrel. This would
be called a “pot furnace” today.
Fig. 10. Cross secon of a bead furnace

6
3. Volcano furnaces
Giberson (1995) describes a Volcano furnace that is a small clay
cone shape which is stoked with fuel from below and has an
aperture in the top where the intense heat exits and beads can
be made. Similar furnaces that are sll used today have been
described at Bida in Nigeria (Dubin 1987/2009) and in Nepal
(Garborieau 1977). This type of furnace could have been used at
Ribe in Denmark where the traces of small hearths of about
50cms diameter have been found amongst beadmaking debris
dang from 600-700 CE (Risom 2013). Bellows are somemes
used (Bida) but are shown not to be essenal to raise the
temperature to the required level for beadmaking.
Wiesenberg (2018 and ongoing research) has created
reconstructed furnaces at Borg and Kempten in Germany and
shown that charcoal and the constant use of bellows is
unnecessary to create the heat required for bead working.
Charcoal requires the constant use of bellows to achieve
sucient ame but wood alone produces the right condions
without the hard work of using bellows. Wiesenberg esmates
that about 50kg of wood is sucient to run this type of
beadmaking furnace for a day. Wiesenberg’s furnace design has
a volcano-type opening at the top with a short tunnel opening
or collar at the base of the furnace as the feed hole for the
wood. Deep slots in the opening at the top of the furnace allow
mandrels to be supported in the ame and this protects the
hands from the heat.
Fig. 13. Volcano Furnace reconstrucon from Risom 2013.
Fig. 14. Replica beadmaking furnace fuelled by wood. From
Wiesenberg 2018.
Fig. 12. Replica Viking Furnace with bellows
FUEL
As seen above, both charcoal and wood have been proposed as the fuel used for bead furnaces and the least labour-
intensive is wood which also gives the beer ame for beadmaking (Wiesenberg ibid). Bead furnaces have dierent
requirements to glassblowing: the furnaces can be smaller and can be heated up and cooled down in a shorter me.
My producon of replica Anglo-Saxon beads has shown that even an elaborate decorated polychrome bead can be
made in less than 20 minutes so heang a furnace during the day and then allowing it to cool down overnight with
nished beads annealing inside is a likely scenario. This would save on collecon and preparaon of wood
considerably.
2. Small Furnaces with bellows
There has been much experiment with Viking age
furnaces and these are usually made as small domed clay
furnaces, red with charcoal, and use a large pair of
bellows to provide a draught. The bellows are worked by
a helper, to keep a constant draught of air entering the
furnace and raise the heat while the beadmaker works at
the opening in the top. These have been trialled by re-
enactors at Viking fesvals and experimental archaeology
locaons and have been shown to work well for
beadmaking. The assumpon that bellows were used
comes from images in Viking carvings and illustraons of
forges used for metalworking, mostly of later date.

7
IDENTIFYING THE FURNACES USED
Can the furnace type be idened from beads?
Kanungo (2004) has described how beads that have a black deposit of ferrous oxide
inside the bead hole were made in a furnace while those without this deposit were
made by lampworking with an open heat source. However, other studies have
shown that the dark deposit could be caused by using salt as a bead resist instead
of clay slip or by using no resist at all (Grimbe 2010). The conclusion is that at
present we cannot tell which type of heat source was used for a parcular bead.
Perhaps analysis of the deposits inside beads may throw light on this.
But there are other clues from making replica beads. Using a blow torch as a heat
source has shown that the narrow ame of the blow torch is a handicap to some of
the techniques. Simple winding and decorang were not a problem but making
milleori canes and drawn beads was more challenging as a larger area of heat was
needed to keep the mass molten. The wider opening of the volcano-type furnaces
would have been suitable for both wound polychrome beads and small-scale milleori cane making. Larger
producon of drawn beads and milleori canes would require a larger furnace with a glory hole so that bigger
gathers of glass could be kept molten for drawing.
CONCLUSIONS
Anglo-Saxon cemeteries in Britain have produced large numbers of glass beads so the lack of any archaeological
evidence of bead furnaces can suggest several things:
1. The beads were not made in Britain.
This has to be rejected because many types of beads found in Anglo-Saxon cemeteries are only found in Britain.
2. Beads were made in Britain but the furnaces used were small and ephemeral so lile trace of them can be
found in excavaon.
This suggests that beads were not made in large workshops employing groups of beadmakers but were probably
made by inerant beadmakers, working alone or in a small family group. They could travel the country, carrying their
stock of coloured glass and simple tools and would build a new small furnace or open hearth as required to pracse
their cra. Furnaces of this size have been shown to deteriorate quickly and weather
away leaving only a hearth as a trace. (Risom 2013; Wiesenberg 2018).
3. Beads were made in workshops using large furnaces at a few locaons only and to
date none of these have been found.
It is possible that both opons 2 and 3 existed at the same me but the types of bead
made in Anglo-Saxon Britain suggest opon 2 is the most likely. I have made replicas of
all the dierent types of glass bead that are only found in Britain and all can be made
using lampwork and simple heat sources. Many imported beads such as drawn and
segmented beads, “recella” beads, mosaic beads (Fig. 16) with complex milleori
paerns and very large glass beads such as the spindle whorl types would be easier to
make with a furnace.
4. Beadmakers made use of glassblowers’ furnaces or metalworkers’ furnaces.
This is a further possibility. Glassblowing furnaces are probably too large for beadmaking but small furnaces used for
melng and pouring bronze would be useful for a beadmaker. Söderberg (1999) has shown how simple such
furnaces could be, very like the open hearth furnace described above. Mixed waste of beadmaking glass and non-
ferrous metals has been found at Garannes Hill fort in Co. Cork, Ireland. At Culduthel (Hatherley & Murray 2021)
furnaces, bead waste and metalworking nds were found and although this is an Iron Age site, it shows further that
crasmen could share furnaces.
Fig. 15. Fine polychrome beads like
these would need lampwork
techniques. Morningthorpe, Norfolk.
Courtesy of Norwich Castle Museum.
Fig. 16. Mosaic bead from Norfolk.
These are rare nds in Anglo-Saxon
graves and are probably imported.
Courtesy of NPS archaeology.

8
REFERENCES
Agricola, Georg. 1556. De re metallica. Gutenberg Project.
hps://www.gutenberg.org/les/38015/38015-h/38015-h.htm
Boulogne, S. 2009. La producon de bijoux de verre dans l’espace islamique médiéval et tardif: la queson d’un
arsanat spécialise. 17ème Congrès de l’Associaon Internaonale pour l’Histoire du Verre, Sep 2006, Anvers,
Belgium. pp.261-268
Charleston, R. J. 1978. Glass Furnaces through the Ages. Journal of Glass Studies , Vol. 20, pp. 9-33. Corning Museum
of Glass
Dubin, L. 1987/2009. The History of Beads, pp. 123. Thames and Hudson, London.
Gaborieau, M. 1977. Bracelets et grosses perles de verre. Fabricaon et vente en Inde et au Népal. Objets et Mondes
17, 3, pp. 111-127.
Grimbe, J. 2010. Making Glass Beads from the Past. EuroREA 7/2010. EXARC
Hasaki, E. 2006. The Ancient Greek Ceramic Kilns and their Contribuon to the Technology and Organizaon of the
Poers’ Workshops. P. Tasios and C. Palyvou (eds.), Proceedings of the 2nd Internaonal Conference on Ancient
Greek Technology. Athens, 221–227
Hatherley, C. and Murray, R. 2021. Culduthel: An Iron Age Craworking Centre in North-East Scotland. Edinburgh:
Society of Anquaries of Scotland. hps://doi.org/10.9750/9781908332202
Kanungo, A. K. 2001. Glass Beads in India: Lamp Winding and Moulding Techniques. Man and Environment Vol. 26
(2), pp 99-108, Pune, India
Kanungo, A. K. 2004. Glass Beads in Ancient India and Furnace-Wound Beads at Purdalpur: An Ethnoarchaeological
Approach. Asian Perspecves, Vol. 43, No. 1 (Spring 2004), pp. 123-150
Krzyżanowska, M. and Frankiewicz, M. 2015. An Archaeological Experiment with Early Medieval Glass Bead
Producon in an Open Hearth – The Results. Slavia Anqua LVI, pp. 109 – 127
Nenna, M.D. and Pichot, V. 2008. Le Village de l’oeil, les verriers de Nazarköy, CNRS (USR 3134) - entre d’Etudes
Alexandrines, Harpocrates Publishing, Egypte, 14”00, hps://www.youtube.com/watch?v=DtdQml1n_iY.
Taylor, M. and Hill, D. 2008. Experiments in the Reconstrucon of Roman Wood-Fired Glassworking Furnaces. Journal
of Glass Studies, Vol. 50, pp. 249-270 Corning Museum of Glass
Risom, T. 2020. The Bead Maker From Ribe: The Story of a Viking Crasman. Forlaget Trælår. Sydvestjyske Museer,
Rolland, J. 2018. Gestes, Savoir-faire et Décisions: regards technique sur l’évoluon de la producon de parures en
verre au second âge du Fer. La spécialisaon des producons et les spécialistes, Séances de la Société préhistorique
française. Paris. Pp 85-107
hp://www.prehistoire.org/515_p_56208/seance_16.html
Söderberg, A. 1999. Metal Casng. Blowing New Life in Old Technology-Viking Age Metal casng.” Viking Heritage
Newsleer Number 6 1999 pg. 13
Wiesenberg, F. Glasperlenherstellung am holzbefeuerten Lehmofen. Experimentelle Archaologie in Europa. EXARC
Jahrbuch 2018
Wilmo, H. and Welham, K. 2013. Late Seventh-Century Glassmaking at Glastonbury Abbey. Journal of Glass Studies,
Vol. 55, pp. 71-83. Corning Museum of Glass

9
EXPERIMENTAL ARCHAEOLOGY FURNACE PROJECTS
Park Roman Villa Borg, Im Meeswald 1, 66706 Perl-Borg, Germany
Many ongoing furnace projects
hp://www.glasofenexperiment.de/
Mark Taylor and David Hill, The Glassmakers
Unit 16, Project Workshops, Lains Farm, Quarley, Andover, Hampshire SP11 8PX, UK
hp://www.theglassmakers.co.uk/index.htm
ACKNOWLEDGMENTS
I would like to thank Frank Wiesenberg and Thomas Risom who have both provided invaluable advice on making and
using small furnaces for beadmaking. Also the members of Historical Beadmakers, Facebook, who have shared their
discoveries with me.
I would also like to thank the following for permission to reproduce beads from their collecons: Cotswold
Archaeology, Needham Market, Suolk; Norwich Castle Museum; NPS Archaeology, Norwich.
Sue Heaser
2022
www.sueheaser.com