Beeswax in metalworking in Viking Period Gotland - ra...

Beeswax in metalworking in Viking Period GotlandGustafsson, Ny Björnhttp://kulturarvsdata.se/raa/fornvannen/html/2016_097Fornvännen 2016(111):2 s. 97-101Ingår i samla.raa.se

Trade in commodities plays an important role inmost works on Early Medieval Scandinavia. Wea-pons, furs, timber, beeswax, glass beads, textilesand slaves were among the goods that, based oncontemporary accounts, were traded by andamong Scandinavians. As pointed out by HeleneClarke (1985, p. 118), very few of them are likelyto have survived archaeologically. As an exampleshe highlighted the discrepancy between writtenaccounts and archaeological finds in England,where 13th and 14th century taxation records forincoming goods are preserved. They include a largenumber of import goods that, had it not been forthe accounts, would have been unknown to us. Ifthis situation is transferred to earlier trade, forwhich no detailed accounts exist, it is obvious thatthe scholarly perspective on trade risks focusingtoo much on luxury goods that are mentioned incontemporary accounts and underestimating theimportance of more basic goods that are not men -tioned. This is important as most of the tradegoods, basic and precious alike, have rotted away,corroded beyond recognition or simply becomevaporised. In the following I will focus on a tradegood from the latter group: beeswax.

Early trade in beeswax is mainly known through

eastern sources that mention it being exportedout of present-day Russia (Warnke 1987; Andro-shchuk 2013, p. 51). Given that beeswax, in laterMedieval sources, is mentioned in connectionwith westbound trade from the Baltic Sea (cf.Galton 1971a, p. 60; 1971b) it is generally assum-ed to have been traded westward during earlierperiods as well. From the 10th century onward,wax was generally connected to the massive needof candles in churches and other religious institu-tions (Crane 1999, p. 599 f). In Scandinavia, thispractise was less important before the mid-11thcentury, though a number of 10th century waxcandles have been found: e.g. in burials at Jelling(Krogh & Leth-Larsen 2007, p. 68 pp) andMammen (Leth-Larsen 1991) in Denmark andGniozdovo in Russia (Avdusin & Puskina 1988,p. 22). Instead it is probably safe to say that themost important use for wax in the period wasmetalworking, specifically casting in the lost-wax technique (à cire perdue).

Beeswax in Gotlandic metalworkingBeeswax is one of the dark horses of early metal-working. Its plasticity and low melting point makeit ideal for transforming a positive representation

Fornvännen 111 (2016)

Beeswax in metalworking in Viking Period GotlandBy Ny Björn Gustafsson

Gustafsson, N.B. 2016. Beeswax in metalworking in Viking Period Gotland. Forn-vännen 111. Stockholm.

Beeswax was an important trade good in the Early Middle Ages. Due to its primaryuses though, for metalworking and in candles, very little of it survives in the archaeo-logical record. This paper discusses the use of beeswax on Gotland in non-ferrousmetalworking and presents two recently dated major finds of beeswax from theisland. Radiocarbon has shown that both belong to the Viking Period.

Ny Björn Gustafsson, Irisdalsgatan 82, SE–621 43 [email protected]

Art. Gustafsson 97-101_Layout 1 2016-05-24 10:37 Sida 97

Fig. 1. Fish-head pendantfrom Öja parish on Got-land (SHM 2370). Draw-ing by Carl Fredrik Lind-berg, after Montelius 1874.

of an object into a negative mould imprint. Thewax is easily sculpted in three dimensions andjust as easily melted out when the mould is firedand made ready for casting. This is, of course, alsothe central problem: wax used in metalworkingis melted and re-used or vaporised in the process.It has recently been argued that wax played amuch lesser part in casting than assumed in ear-lier works on metalworking – instead, mastermodels or extant objects are said to have beenused to create the negative imprints needed inmould making (cf. Söderberg 2001 w. refs). Tosome extent this is clearly true: reduplication wasan important principle in the Viking Period.Some types of objects, like dress jewellery, caneven be considered as mass-produced.

This does not necessarily exclude the use ofwax though. Making certain objects actuallyrequired it. This is especially evident in the pro-duction of hollow objects. One example is castfish-head pendants (Sw. fiskhuvudformiga hän-gen). In the 10th century such pendants (fig. 1)were largely mass produced – a normal set of pen-dants typically seems to have included 15–20 pen-dants mounted together as a semi-circular collar(Thunmark-Nylén 2006, p. 199 ff). Thanks to afind from Klints in Othem on north-eastern Got-land (SHM 34663:61), along with extant pen-dants in various stages of completion, the line ofproduction can be reconstructed.

The Klints piece was first reported as half apendant, but further study showed that it is actu-ally a master model for the front face (Gustafsson2011a, p. 95). As the finished pendants are hol-low, the master simply cannot have been used tocreate moulds on its own; the hollow centre wascreated by means of a clay core. Several extantpendants show how this was achieved: the corepiece, made from mould loam, was fixed by meansof a number of strategically placed iron spacers.These held the core in place and allowed moltencopper alloy to enter in between the outer mould,which carried the negative imprint of the master,and the core. This may sound simple in theory,but in practice it demanded wax. The master mo d-el could not, as with masters for flatter objectswithout under-cut details, simply be pressed intosoft mould-loam to create a casting cavity. In-stead it must have been placed on a flat surfacewhere soft, fine-grained clay was pressed downover it. The master was then removed and thenegative imprint in the clay coated with a thinlayer of wax, probably by pouring molten wax inthe cavity and allowing it to solidify brieflybefore pouring the excess out. The result wouldbe a copy of the master, a front half of a pendant– in beeswax.

The wax model was then removed from theclay, either by peeling the latter off or simply byputting it in water where the unfired clay woulddissolve. (The Klints master probably came witha master for the back of a pendant. Unfortunate-ly it has not been recovered; the site at Klints wasbadly damaged by illegal landscaping during theextension of a golf course, and most finds wererecovered by metal-detection of bulldozed topsoil – Pettersson 2005.) The next step was to fitthe front and back wax halves together, creatinga complete wax version of a pendant. This wasfollowed by the first stage in the production ofthe physical mould: filling the hollow centre ofthe wax model with mould loam to form a core.If the wax pendant had been embedded in mouldloam at this stage the resulting mould would not,however, have been suitable for casting. To accomp-lish this, the aforementioned iron spacers had tobe added. They were probably applied to the waxpendant right after the core had been inserted.Normally, one short spacer was placed at the bot-

98 Ny Björn Gustafsson



tom of a pendant and one, longer, through themiddle. The spacers were made long enough toprotrude outside the wax model. This allowedthem to be embedded in the outer layer of mouldloam which was used to cover the wax positive.An un-fired pendant mould thus included, fromthe inside out: A) clay core, B) iron spacers, C)wax positive and D) outer mould (fig. 2). Whenthe mould was fired the spacers kept the core inplace and the cavity left by the lost wax was thusopen to receive the molten metal. This recon-structed line of production is supported by thefact that many pendants retain spacers lodged inthe surface. Sometimes it is also possible to ob-serve remains of fired core clay adhering to thelower inside corners of damaged pendants. Thisserves to illustrate the line of production of onetype of object where beeswax was necessary forthe desired result.

Another Gotlandic example is a set of fivecast yet unfinished sword pommels from Mäster-by (Gustafsson 2011b). Through 3D scanning ithas been established that one of the pommels wasused as a master model to create wax positives forthe other four. Since they are all hollow the sametechnique as with the pendants had been used,but due to thermodynamic shrinkage in wax,mould and metal (cf. Lønborg 1994, p. 154) thecopies are all c. 5–6% smaller than the originalpommel. In the 3D rendition of the pommels it isalso quite clear that they were cast from wax pos-itives, as the seams between what was once twohalves are clearly visible.

Two finds of Viking Period beeswaxAfter the above, it seems safe to assume that therewas great demand for beeswax among Gotland’smetalworkers, and that the wax was largely re-used or vaporised in the line of production. How-ever, there are two major finds of beeswax fromthe island. The first is in the Swedish HistoryMuseum (SHM 11120), and measures when re-fitted 195 by 105 mm at its widest and is 30 mmthick. The present weight of the piece is abouthalf a kilo, but it must once have formed part of amuch larger circular cake, about 400 mm indiameter. It was found in Halla parish in Starrar,a drained fen on central Gotland, before 1900when it was added to the collections.

99Beeswax in metalworking …


Fig. 2. Schematic cross-section of a mould for a fish-head pendant: a) vertical, b) horizontal.


The second find, in the Gotland County Muse-um (GF C9903), is a much larger cylindrical blockwith a diameter of 330 mm and a height of 110mm, weighing 11 kg. According to the museuminventory it was found in Mästermyr fen on south-ern Gotland in 1951. This fen was once a vastwetland, large enough to stretch through sevenparishes. Sadly, the brief entry on the find loca-tion and the year of recovery is all that is knownabout the find. It should be noted that in 1936Mästermyr fen also yielded a rich and intriguingtool hoard (Arwidsson & Berg 1983). It is tempt-ing to associate the two finds with each other, butdue to our lack of contextual information on thewax find there is nothing to imply a connection.Why beeswax was originally deposited in wet-lands is unknown. If the usual ritual perspectiveis left aside, a more practical reason may havebeen to safeguard the material from mice andother pests.

The two Gotlandic wax finds are both singlefinds and thus they cannot be dated by context.Still, catalogue entries in the Swedish HistoryMuseum refers to the find from Starrar as “po ten-tially Bronze Age or High Medieval”. The Mäs -ter myr block is displayed in the High Medievalsection of the Museum’s exhibitions. As a com-parison, there is another block of beeswax fromSweden, weighing 13 kg. It was found in Lund,Scania in 1938 (KM 39273; Mårtensson & Wahlöö1970, p. 112). It has not been dated but was found

in a charcoal-rich layer together with sherds of Bal-tic ware dating from the 11th or 12th centuries(Kjellberg 1947, p. 43 p).

Beeswax is eminently suitable for radiocar-bon dating as it has a low intrinsic age; it is sec-reted by bees in a hive and the included 14Cimmediately starts to decay. Additionally, eventhough the amount of wax in the two Gotlandicpieces must represent a large number of hives, itis probably safe to assume that it was collectedover a short period. I had the Gotlandic findssampled for radiocarbon dating in 2014. Theresults were surprisingly similar: with over 95%likelihood both wax blocks were produced in theperiod 890–1030 cal AD (Possnert 2015). Theblock from Mästermyr (Ua-50647, 1068±30 BP)might be 5–6 years older than the one from Star-rar (Ua-50648, 1062±30 BP), but both are with-out doubt from the Viking Period. According tothe analyses they either date from 900–920 calAD (10.3% and 6.2% probability for the Mäs -termyr block and the Starrar piece respectively)or 970–1020 AD (57.9% and 62%, respectively).

Regardless, they both originated in a periodwhen non-ferrous metalworking thrived andlarge amounts of beeswax are likely to have beenconsumed on Gotland (cf. Gustafsson 2013).They thus offer important insights, both intohow the wax used by Viking Period artisans oncemay have reached the island (the intact blockfrom Mästermyr) and how it was further dividedas it was used up (the Starrar piece). As to theirgeographical origin, further laboratory analysesmight help to answer that question. Many writ-ten accounts point, as mentioned above, towardsthe Baltic States and western Russia as centres ofearly beekeeping.

Funding for the radiocarbon dating was generouslyprovided by the Wilhelmina von Hallwyl Gotland foun-dation.

ReferencesAndroshchuk, F., 2013. Vikings in the East – essays on Con-

tacts along the Road to Byzantium (800–1100). StudiaByzantina Upsaliensia. Uppsala.

Arwidsson, G. & Berg, G., 1983. The Mästermyr find – aViking age tool chest from Gotland. Stockholm.

Avdusin, D.A. & Puskina, T.A., 1988. Three ChamberGraves at Gniozdovo. Fornvännen 83. Stockholm.

Clarke, H., 1985. English and Baltic trade in the mid-dle ages – an evaluation of the evidence. Lindquist,S-O. (ed.). Society and trade in the Baltic during theViking Age. Acta Visbyensia VII. Visby.

Crane, E., 1999. The World History of Beekeeping andHoney Hunting. New York.

Galton, D., 1971a. Survey of a thousand years of beekeepingin Russia. Bee Research Association. London.

– 1971b. Beeswax as an import in Medieval England.Bee World 52:2. London.

Gustafsson, N.B., 2011a. Beyond Wayland – thoughtson early medieval metal workshops in Scandi-navia. Historical Metallurgy 45:2. London

– 2011b. A Viking Period Metalworking Hoard fromAlvena in Mästerby parish, Gotland. Fornvännen106.

– 2013. Casting Identities in Central Seclusion. Aspects ofnon-ferrous metalworking and society on Gotland in theEarly Medieval Period. Theses and Papers in Scien-tific Archaeology 15. Stockholm University.

Kjellberg, S.T., 1947. Något om vax. Kulturen1946. Lund.Krogh, K.J. & Leth-Larsen, B., 2007. Hedensk og Kris-

tent. Fundene fra den Kongelige Gravhøj i Jelling.Vikinge¬kongernes Monumenter i Jelling 2.Copenhagen.

Leth-Larsen, B., 1991. The wax candle from Mammen.Iversen, M. et al. (eds). Mammen – grav, kunst ogsamfund i vikingetid. Højbjerg.

Lønborg, B., 1994. Fremstilling af vikingetidens skål-formede fibler. Kuml 1991–92. Århus.

Montelius, O., 1874. Sveriges forntid – försök till framställ-ning af den svenska fornforskningens resultat. Atlas 2.Jernåldern. Stockholm.

Mårtensson, A.W. & Wahlöö, C., 1970. Lundafynd. Enbilderbok. Lund.

Pettersson, A-M., 2005. Arkeologisk efterundersökning:fyndplats för vikingatida silverskatt och boplatslämning,Klints 1:16, Othem socken, Gotlands kommun. Läns-museet på Gotland. Visby.

Possnert, G., 2015. Resultat av 14C-datering av bivaxoch obränt ben från Gotland. Analysrapport frånÅngströmlaboratoriet, Uppsala universitet.

Söderberg, A. 2001. Scandinavian Iron Age and EarlyMedieval ceramic moulds – lost wax or not orboth? Tulp, C. et al. (eds). Experimental and Educa-tional aspects on Bronze Metallurgy, Wilhelminaoord18–22 October 1999. Leiden.

Thunmark-Nylén, L., 2006. Die Wikingerzeit GotlandsIII:1. Text. Stockholm.

Warnke, C., 1987. Der Handel mit Wachs zwischenOst- und Westeuropa im frühen und hohen Mittel-alter. Düwel, K. et al. (eds). Untersuchungen zu Han-del und Verkehr der vor- und frühgeschichtlichen Zeit inMittel- und Nordeuropa IV. Der Handel der Karo-linger- und Wikingerzeit. Göttingen.

100 Ny Björn Gustafsson



Summary

Beeswax is one of the commodities that are oftenmentioned in discussions of Early Medieval tradein Scandinavia. Due to the nature of its main uses,in candles and metalworking, beeswax is rarelyfound archaeologically. In Scandinavia, prior toc. AD 1050, most wax was presumably used forthe latter of the two purposes, in preparation ofcasting moulds according to the so-called lost-wax technique (à cire perdu). The plasticity of bees-wax made it well suited for sculpting the intend-ed objects, and its low melting point meant thatit was also easy to melt it out of the castingmoulds when these were fired and made ready foruse.

The role of beeswax in non-ferrous metal-working has to some extent been disputed. Simp-ler objects may well have been produced without

the use of wax. For some object types however itwas essential, in particular for the casting of hol-low items. Early Medieval Gotland was a centreof non-ferrous metalworking, and its artisansmust have been dependent on a steady supply ofimported wax. Two major finds of beeswax havein fact been made on the island. The age of thesefinds, one from the drained fen Starrar in Hallaparish on central Gotland and one from an un -certain location in the once vast Mästermyr fenin the south, has been unclear. Radiocarbon dat-ing has now established that both are from theViking Period (890–1030 cal AD, 2 sigma). Thegeographical origin of the wax is uncertain, butperiod sources point to the western parts of pres-ent-day Russia as a core area for early beekeep-ing.

101Beeswax in metalworking …



Beeswax in metalworking in Viking Period Gotland - ra...

Documents

Transcript of Beeswax in metalworking in Viking Period Gotland - ra...