Subscribe now for full access and no adverts
It is a running joke in archaeology that the most significant finds tend to emerge right at the end of an excavation – and this was certainly the case during an ongoing community project near Norton Disney in Lincolnshire. In the summer of 2023, Norton Disney History and Archaeology Group were working with Allen Archaeology to investigate a possible Roman settlement at Potter Hill, close to the known location of a villa (which was dug in the 1930s) and the line of the Fosse Way (the Roman road between Exeter and Lincoln). As that season of digging drew towards its close, they had opened a final trench over a large Roman sand-quarrying pit, which had later been reused as a handy place to dump refuse. Its fill contained a wealth of familiar finds, from pieces of pottery to ceramic building materials – but towards the top was something more unusual: a copper- alloy dodecahedron.
Around 130 examples of these enigmatic objects are known from across the Roman Empire, with previous discoveries scattered across modern Austria, Belgium, Britain, France, Germany, Hungary, Luxembourg, the Netherlands, and Switzerland. Despite this wide geographic spread, it is striking that they all come from the north-western provinces of the Empire (and predominantly from Gaul), with none known from the imperial heartlands of Italy or the wider Mediterranean area, and none from Roman territories in Africa. The Norton Disney example is the 33rd to emerge from Britain, but it is the first from the Midlands.
Each dodecahedron has common features – they are hollow, 12-sided objects cast from copper-alloy, with a hole in each face and a round knob at each vertex – but no two are exactly alike, and they vary greatly in size, ranging from the dimensions of a golf ball to a tennis ball. What is clear, however, is that they would have been very complex to create, requiring great skill and a lot of metal to cast. The effort and expense involved in making a single dodecahedron suggests that they may have held some kind of special significance – but no written reference to them is known from any surviving Roman text, and no depiction has been identified in mosaics or wall paintings. For now, their purpose remains obscure, though diverse theories have been put forward.
One option posits that they are a kind of surveying tool, though the fact that the dodecahedra are not of a standard size and lack numbering or other marks that could help with the taking of measurements casts doubt on this. Other interpretations include a candleholder, a knitting dolly, or even a child’s toy, though such mundane uses seem odd when clay or wooden equivalents could have been created much more cheaply to achieve the same result. By contrast, some suggest that dodecahedra had no practical function at all, but were instead a ‘test piece’ intended to demonstrate the skill of a particular artisan. Alternatively, perhaps we should fall back on the archaeological catch-all of ‘ritual’. It has been proposed that dodecahedra could have played some kind of role in divination rites, or might, with their 12 faces, be linked to the signs of the Zodiac, though there is no obvious way to distinguish one side from another or apply any specific symbolism to each.
The discovery of the Norton Disney dodecahedron has not solved this mystery, but it represents a particularly well-preserved example, with many others existing in a damaged or very fragmentary state. At 8cm tall, it is one of the largest dodecahedra yet found in Britain, and is surprisingly heavy at 245g. Archaeometallurgical analysis revealed the reason: its alloy includes a strikingly high proportion of lead, at 27 per cent (together with 63 per cent copper and traces of other metals), which was possibly intended to help the metal flow more easily into its mould to create the object’s intricate design. This dodecahedron is also unusual because it was found during a modern excavation (many others are antiquarian finds: the first-known British example was discovered in 1739) and within a datable archaeological context. This wider story is explored in an exhibition currently running at the University of Nottingham Museum (see ‘Further information’ below), which places the dodecahedron within the much longer story of human activity at Potter Hill, which has seen millennia of occupation.
Seeking the settlement
Norton Disney History and Archaeology Group and Allen Archaeology have been investigating the archaeology of Potter Hill for a number of years. Following geophysical surveys near the villa site in 2020, their 2021-2022 excavations explored the remains of an Iron Age enclosure, a roundhouse, and evidence of iron smelting pre-dating Roman occupation of the area. Then, in 2023, the team moved to a new area located about 700m from the villa, where an earlier survey in 2019 had indicated a possible Roman settlement. A large anomaly identified in these surveys proved to be the sand-quarrying pit mentioned above, and the dodecahedron was discovered just as the 2023 dig was drawing to a close. The team returned in 2024 to learn more about the circumstances in which the dodecahedron had been buried, carefully quadranting the almost 2m-deep pit and examining the diverse detritus that had been dumped into it.
The story of Potter Hill is coming into clearer focus with every investigation, but the dodecahedron represents a particularly intriguing episode.
The quarry pit proved to be full of broken pottery, animal bone, a single hobnail, and pieces of flat roof-tile (tegulae), as well as box-flue tiles from Roman heating systems. Helpfully, a radiate of Claudius II (r. 268-270) was found close to the top of the fill, above the dodecahedron, indicating that the pit could not have been completely filled before this date, while the pottery also included locally made Swanpool ware which is known to have been produced south-west of Lincoln in the 4th century. Chris Clay, a Director at Allen Archaeology, whose company supervised the dig, noted that there was nothing about the deposit to suggest anything structured or deliberate. Similarly, the dodecahedron had not been placed at the base of the pit (which might hint at some kind of foundation act) nor at the very top of the fill (to suggest a closing gesture). Instead, the elaborate object appears to have been discarded like the more everyday rubbish surrounding it. Could its presence at Potter Hill indicate that there was something special about the site during the later Roman period? The presence of roof tiles and box flue within the same pit speak of a substantial, high-status building in the vicinity, and the villa is too far away to be the likely source.
A second trench opened about 10m to the east also produced quantities of domestic material, further illuminating the Roman settlement (and keeping the finds team very busy processing over 2,500 fragments of pottery). The exhibition includes a near-complete cooking pot as well as part of a mortarium, or grinding dish, which preserves a stamp revealing the name of its maker: G Attius Marinus, who is thought to have been in operation c.AD 80-100. As well as more pieces of Swanpool ware (easily identifiable with its distinctive ring-and-dot design), other locally made ceramics included simple grey wares spanning the 2nd to 4th centuries. Most of these would have been used for food storage or preparation, but the base of one Trent Valley vessel had been drilled through to create a spindlewhorl, recycling a broken pot into a useful weaving tool.
Other ceramics appear to have been imported from further afield, among them Nene Valley colour-coated wares from the area around Peterborough, and fine Samian tablewares from Gaul, some of them elaborately decorated (one fragment in the exhibition is adorned with images of a tree and a bird). This settlement was clearly well-connected to the wider Roman world, as well as locally. In addition to the proximity of the Fosse Way, Potter Hill stands between the Rivers Trent and Witham, and the settlement itself occupies high ground with sweeping views across the surrounding valleys.
This advantageous position might explain, too, why the site had been an attractive place to settle long before the Roman road was created. The displays feature a number of other finds from the excavation that pre-date the Roman period: Mesolithic and Neolithic worked flints, a copper-alloy blade (found across two years and in two pieces, now reunited) dating to the Bronze Age, and pieces of hand-moulded Iron Age pottery. The Roman phase of the site’s life was long as well, spanning the 1st to 4th centuries, and included both domestic and industrial activity. The project team hope to tease out more of this history in future excavations, and the independently financed group are already fundraising for next year. While the story of Potter Hill is coming into clearer focus with every investigation, the dodecahedron represents a particularly intriguing episode – and the site clearly has many more secrets to reveal.
Antiquarian interest
The first person to describe the archaeology of Potter Hill in detail was the Lincolnshire antiquarian William Stukeley, who visited on 7 September 1722 and created the sketch shown here. In his Itinerarium Curiosum, he writes that ‘there is a high barrow or tumulus called Potters Hill, where they say was a Roman pottery’. Subsequent scholarship was sceptical of this theory, but the Norton Disney team believe that he may have been right after all: they recently commissioned a re-cataloguing of the villa finds, which revealed that some of the ceramics might in fact represent kiln furniture. This year’s excavation close to the dodecahedron pit also revealed a possible kiln site, later used as a dumping ground, which produced pottery wasters, kiln furniture, hundreds of sherds from various wares, and some large, ‘Iron Age style’ triangular loom weights. Stukeley also featured in the 2022 season, when the team held a tea party on 7 September to mark the 300th anniversary his visit – including a cake that appeared as ‘Edible Archaeology’ in CA 393.
Source:
• Richard Parker is Secretary of the Norton Disney History and Archaeology Group, and curator of The Norton Disney Dodecahedron: its discovery and excavation.
• Chris Clay is a Director of Allen Archaeology Ltd.
• Dr Clare Pickersgill is Keeper of the University of Nottingham Museum.
Further information:
The Norton Disney Dodecahedron: its discovery and excavation will be at the University of Nottingham Museum until 18 January 2026. The museum is open noon-4pm Thursday-Sunday; entry is free. See http://www.lakesidearts.org.uk/exhibition/the-norton-disney-dodecahedron-exhibition for more details.
References:
Hitchens, L (2024) ‘Appendix 7 Roman Dodecahedron’, in Archaeological Evaluation Report, Norton Disney, Lincolnshire, pp. 34-41. Allen Archaeology.
McDonnell, G. (2024) ‘Appendix 8 Hand-Held X-Ray Fluorescence Analysis of a Copper Alloy Romano-British Dodecahedron’, in Archaeological Evaluation Report, Norton Disney, Lincolnshire, pp. 44-48. Allen Archaeology.
