A complete cast copper-alloy dodecahedron dating to the Roman period (c. AD 43-410). Type 1b.
This object was discovered during a controlled archaeological investigation by a local History and Archaeology Group and Allen Archaeology and was recovered from a pit described by the excavators as a quarry infilled with debris as a midden. Other finds include a box-flue tile fragment, grey-ware pottery, roof tile debris and animal teeth. Photographs and information were kindly provided by Lorena Hitchens who is currently undertaking a PhD on the topic of dodecahedrons. The object has not been handled by the recorder. Photographs are the copyright of Lorena Hitchens.
Description
Dodecahedra are characterised by their openwork (hollow) 12-sided form. Each face is a pentagon with equilateral sides, pierced with a different-sized circular hole Sometimes, faces are decorated with markings surrounding the holes or on the faces or edges. Each vertex, or corner where three sides meet, has a semi-spherical knob, totalling 20 when completely intact. All are roughly the size of an adult human fist, which, it should be noted, can vary widely, from 5cm to 11cm in diameter.
The casting is of high quality, with no cracks, gaps or voids from manufacturing are visible.
This example closely aligns with "1b" style (based on the 1907 typology developed by de Saint-Venant, later refined by Grenier in 1996, then Guggenberger in 1999). Type 1b is defined by a varying number of concentric rings per face, with up to 12 faces decorated; the Norton Disney object is decorated on all 12 faces. Face A, with the largest hole, has one ring. Face J, the face with the smallest hole, has three rings; all other faces have two rings. There are no other markings or stamps inside or outside the
object. The holes on the faces are graduated with slight differences in size
Measurements
Height: 80 mm, Height (without knobs): 70 mm; Width: 86 mm (without knobs) 75 mm; Weight: 254g. Side length of faces: 27 mm.
Discussion
This is the first example to have been found in the Midlands (stratified or unstratified) and is an exceptionally large example, although it is thought that a larger example is located in the British Museum, although this has been resized on three occasions. It has an unusually large lead content. Lead is required to make it flow into the mould, but this amount makes it fairly impractical for use.
This is the 10th example recorded on the national PAS database - with it being the first for Lincolnshire and the Midlands. For similar complete or near complete examples see YORYM-41CD72 and BH-692011
What are they?
The function of these enigmatic forms is still unclear and no firm conclusions have been reached. Suggestions as to their function include: candleholders, polygonal dice, rangefinders, sceptre heads and surveying instruments.
XRF testing data
Gerry McDonnell tested the object with pXRF on 14 Sept 2023. Details regarding the device, settings, targets, calibration, etc., can be found in the separate specialist report from Gerry McDonell Archaeometals Consulting. The following results were communicated verbally to the author:
● Copper (Cu) 63%
● Lead (Pb) 26%
● Tin (Sn) 8%
● Zinc (Zn) 0.2%
● Iron (Fe) trace
These elements are typical of late Romano-British 'leaded gunmetal' cast objects made of mixed recycled copper-alloy metal objects. The lead proportion is high even for typical leaded gunmetal. This may be a random result from the recycling of other metal objects/scrap or an intentional decision by the maker to add more lead. Lead helps the liquid metal "flow" well into nooks and crannies of 3-dimensional object moulds (Bayley and Butcher 2004:15). Decreasing zinc availability from the mid-1st century AD onward could indicate, if
cross-referenced with other dating methods, a later date for this object (Bayley, Crossley, and Ponting 2008:49). By the latter half of the Empire, zinc was nearly unavailable in Roman
Britain, with zinc only present in copper-alloys due to the recycling of earlier brass objects. Iron can exist in trace amounts in other metal elements, or the iron trace detected by XRF may stem from surface contamination. In either case, the iron detected here is unlikely to result from any intentional alloying.
References
Bayley, J., and Butcher, S. (2004) Roman brooches in Britain: a technological and typological study based on the Richborough Collection. London: Society of Antiquaries of
London.
Bayley, J., Crossley, D. and Ponting, M. (ed.s) (2008) Metals and Metalworking: a research framework for archaeometallurgy. London: Historical Metallurgy Society Occasional
Publication No 6.
Guggenberger, M. (1999) Die römischen Dodekaeder. Eine Gesamtdarstellung (Dipl.Universität Innsbruck).