Subscribe now for full access and no adverts
In 2013, a team from ULAS (University of Leicester Archaeological Services) were excavating the remains of an Iron Age settlement near the village of Glenfield, on the edge of Leicester, when they made an unexpected discovery. The settlement itself was unusual, with a dense area of overlapping roundhouses and enclosures created through successive phases of occupation throughout the Iron Age, but the composition of the finds that the site produced was unique, representing one of the largest collections of Middle Iron Age metalwork yet recovered from the county, and featuring a range of objects that had never been seen before in the area (see CA 335). Together with an eclectic array of iron tools, dress items, and weapons, the most intriguing aspect was a collection of 11 complete or near-complete cauldrons fashioned from iron and copper-alloy – a remarkable and rare assemblage.
The Glenfield cauldrons are one of only two large groups of Iron Age/early Roman cauldrons to be found in the UK to-date (a bigger assemblage was discovered in 2004, at Chiseldon in Wiltshire; see CA 214), and such discoveries can offer illuminating insights into the life of a prehistoric community. Cauldrons are large enough to contain substantial quantities of food and drink, and are associated with gatherings and feasts. The Glenfield settlement was probably a host site for such events, but what happened at the end of these vessels’ lives is also instructive. The many and varied ways in which the Glenfield cauldrons were eventually disposed of – placed in Iron Age ditches across the site, with a group of eight buried in one ring-ditch – also suggest that it was a place where unusual depositional acts took place, setting it apart from many contemporary settlements in Leicestershire.
Glenfield’s story of Iron Age occupation and unusual artefacts will feature in a new temporary display to show this significant new discovery, as part of the refurbished Jewry Wall: A Real Roman Experience museum in Leicester, when it opens on 26 July. As well as highlighting the cauldrons’ journey from creation to burial as revealed by investigative conservation and scientific analysis, the display will highlight how this information fed into the creation of a stunning full-sized replica cauldron by archaeological ironworker Hector Cole. Here, we will give CA readers a flavour of this process, and what we learned along the way.
Investigative conservation
Excavating the Glenfield cauldrons was a painstaking process. They were all extremely weak, heavily corroded, and deformed due to the long-term effects of being buried in heavy clay soil. In order to recover the delicate objects without further damage, they were all carefully lifted in soil blocks, wrapped in plaster bandage supports, and were transported off-site like this for more detailed investigation. Each block was then CT-scanned to create a record of them in situ and as a guide to their examination. This was followed by micro-excavation, further X-radiography, investigative cleaning of selected parts, and detailed recording by conservators at MOLA (Museum of London Archaeology). The aim of this work was to examine and record structural details of the cauldrons – their dimensions, any decoration, patching, and mineral-preserved organics – and included sampling of metal elements and residues for scientific analysis. Ultimately, we hoped to uncover evidence for how the cauldrons had been manufactured, used, and repaired during their active lives, and how they had been treated in deposition – and we were not disappointed.
Most of the soil blocks held single cauldrons, apart from two examples where the remains of two vessels were uncovered, overlapping each other in close proximity. Some were placed rim uppermost, others inverted, but all of the cauldrons had become crushed and distorted in the ground. Our task was to make sense of these fragile finds, and through micro-excavation it became clear that they all belonged to a typology known as Iron Age/Early Roman Group II, representing the ‘globular composite’ type described by Jody Joy in 2014 (see ‘Further reading’ below). Each comprised iron and copper-alloy sheet-metal riveted together using a similar construction method, formed from four main sections: an iron rim attached to an iron band (A), copper-alloy band (B), and copper-alloy base (C), with the lower edge of A overlapping the upper exterior of B, and the upper edge of C overlapping the lower exterior of B.
Among the ironwork there was some metallic survival but, in most cases, heavy corrosion had resulted in complete mineralisation among the ring handles or iron bands, and we were only able to discern suitability for metallurgical analysis through CT and X-ray imaging. Despite this variable preservation, however, all of the cauldrons retained at least part of their rims, and the majority had surviving parts of their iron band, with at least one ring-handle remaining. We were able to measure each cauldron’s diameter at the join of the lower edge of the iron band A and the upper edge of copper-alloy band B (termed the A/B join) or, at least, to estimate this where the rim and band only partially survived. The vessels varied considerably in size, from a diameter of 360mm (14.2 inches) at the smallest to 560mm (22 inches) at the largest – a span that is roughly consistent with the range of sizes at Chiseldon and within the typical size range of other Group II cauldrons.
In contrast to the sheeting, the copper-alloy rivets securing the structural parts survived relatively well and could be seen clearly on the CT-scan images. Even they were subject to corrosion, however, and in some cases had been completely lost. From those that remained, we could record their sizes, form, and spacing, which gave us indication of the quantity of rivets that Hector would need for his reconstruction, as well as a choice of riveting styles that he could consider.
Catering clues
What can be said about how the cauldrons were used? Their copper-alloy parts were eggshell thin, with each band B measuring just 0.2-0.3mm thick at the B/C join. As every vessel’s band B was crushed and distorted, it was not possible to tell definitively whether they are globular or U-shaped, as was possible with the Chiseldon group. However, by assuming that the cauldrons are roughly hemispherical (in accordance with Joy’s Group II typology), we were able to estimate an approximate volume for each example. This calculation suggested a capacity of between c.15-57 litres (3.3-12.5 gallons) if filled to the rim, or c.10-42 litres (2.2-9.2 gallons) if filled to the bottom of the iron band, depending on the size of each vessel. The latter measurement would seem more practical for serving or if the cauldron and its contents needed transporting without the risk of spillage.
As for how many people each cauldron could have fed, this is tricky to estimate, as it depends on many factors, such as the size of serving, the number of adults versus children present, the number of cauldrons being used at one time, the consistency of the contents themselves, and the extent to which they were filled. Modern stewing pots are sold on the assumption of a portion of 1 litre (0.2 gallon) stew per person, which perhaps offers some indication of the number of adults that the Glenfield group might cater for, if they were all in use simultaneously for this purpose.
What was on the menu at Glenfield? Thankfully, Iron Age people do not appear to have been fastidious in their washing up, which has allowed us to explore this in more detail. It was clear from the sampling of a thin, dark residual layer, as well as matter that had become caught in crevices and between patched layering, that the cauldrons had not been thoroughly cleaned before they went into the ground. In fact, given some of the build-up, they may not have been regularly washed during their active use – a choice that might have helped to keep them well sealed against leakage. Luckily for their archaeological study, this means that sampling and analysis of organic residues to look at the types of food cooked in the cauldrons has been possible: the results will be presented in a forthcoming publication.
Other insights into the cauldrons’ use-history come from the fact that, where their fabric survived more completely, we could see that they (like those at Chiseldon) had been extensively repaired. This is perhaps unsurprising given the thinness of the copper-alloy elements, and from the locations of patches, and the sampling and metallurgical analysis of some of these, we have indications that repairs were made both during manufacture and as the cauldrons were used over time. These small patches and washers were not replicated in Hector’s cauldron, but they are nonetheless important for understanding the significance of these objects to their Iron Age owners. The fact that they were thoroughly repaired, probably on numerous successive occasions, rather than discarded when damaged, suggests that they were highly valued and carefully maintained. The manner of the patching also indicates that they were made to be used, as they needed to be watertight, rather than being purely ornamental or symbolically made for burial. Many patches appeared to be impulsive and functional in design, but others may have been deliberately conspicuous, such as one with a decorative rivet that could even have been a signature mark left by the repairer.
Making an Iron Age cauldron
When Hector set out to replicate one of the Glenfield cauldrons, the intention was that it should broadly represent the making processes for the whole group, while being based on the dimensions of one chosen example. This was an intriguing exercise to see how the recorded dimensions and features of the metalwork, and the predicted sequence of construction, worked on a practical level. So, what did we find?
The replica cauldron band and base were created from a bronze sheet measuring 700mm (27.6 inches) in diameter by 1.2mm (0.05 inch) thick. The size of the disc needed was calculated from records that had been created during the initial investigative conservation, providing enough to make a cauldron 385mm (15.2 inches) in diameter, with a little extra to allow for trimming when the bowl reached its final shape. The method used for forming the bowl is known as hollowing or sinking – where the flat sheet is beaten into a hollowed-out wooden block with a domed hammer or mallet, starting from the rim of the disc and working to the centre, while constantly rotating it. Bronze becomes less malleable when it is beaten, so the sheet had to be regularly annealed (a process of heating and cleaning) to keep it as soft as possible for working. This is consistent with observations made by Jody Joy, who noted that that this technique would be essential to achieve the extreme thinness of the sheeting that we see in archaeological examples.
Once the shape of the bowl was completed, the surface was smoothed out by planishing, where the bowl is put upside-down on to a domed stake and hammered all over to create a smooth surface. Different types of hammer leave different signatures, and as the surfaces of the Glenfield cauldrons were relatively unmarked, a hide mallet was used for the replica in order to obtain as smooth a finish as possible. Finally, the upper rim of the bowl was levelled by trimming to a measured and marked-out line using a pair of shears.
The bottom of the shaped bronze was cut away and a replacement part, representing bowl C, was riveted on to match the typical format of the cauldrons. A key concern during the shaping process was the base splitting – an ever-present risk while Hector was thinning it out – and it appears that Iron Age metalworkers also had to contend with this. Strategic sampling during conservation and metallurgical analysis of the Glenfield cauldrons suggested that the bowl C was indeed sometimes added during manufacture, and that it had a repair and reinforcement function rather than being a purely pre-planned design feature.
Forging ahead
The upper parts of the cauldron were all made from sheet and billets of wrought iron, which were forged into various sections: the upper band A, the rim, the wrought-iron handles, handle mounts and their washers, and the handle plates. The upper band was cut in two sections from wrought-iron sheet, and the two halves were then formed round to fit the circumference of the bowl and riveted together. Once the wall had been fitted to the bowl, the holes for fixing the two parts together were marked out, following details from the original cauldron. There was a remarkable consistency to the diameter of the rivet holes in the Iron Age example, indicating that they must have been punched using a special tool, and an estimation of the type of implement used was created for the replica project. Fixing the iron band to the copper bowl required 100 holes to be created through the two layers, and as many rivets; this join alone took roughly five to six hours to achieve.
Decorative handle plates were also incorporated, replicating a ‘vegetal’ art design seen on one of the Glenfield vessels. These were forged from thin wrought-iron sheet using the repoussé method to form the raised decorative parts: a mould of the design was created in lead, and the shape was formed in the sheet using a variety of punches, before each plate was cut to shape and riveted to the iron band. Hector found that the rim was easiest to fit right at the end of the process, with much of the cauldron complete. The Glenfield group represents a number of rim and band types, and for the replica a solid rim was forged from three pieces of wrought-iron bar and joined by scarf-welding before it was shaped to the cross-section of the original.
The final stage was to shrink the rim on to the top of the cauldron wall, which Hector did by bringing the rim to a dull red heat, placing the cauldron on to the rim, and then rapidly cooling it with water. Downward pressure was applied to the cauldron to hold it in place on the rim during the cooling process – but after several attempts the rim refused to stay put, and the decision was made to secure it with custom-made straps. A similar arrangement was observed on one of the Chiseldon cauldrons, suggesting that Iron Age metalworkers had to tackle similar challenges. Overall, the replica took more than 100 hours to make, equating to roughly 14 skilled metalworking days.
Valued vessels
The Glenfield cauldrons are an incredible group of objects that have opened a new window on to Iron Age society in the English Midlands. As centrepieces for large gatherings at feasts, they would have been highly important objects to the communities who used them, helping to bring people together from far and wide. It is evident from the patching preserved on their surfaces, and the care that was clearly taken to keep them in use, that these vessels were valued and had long histories that tied them to the settlement. The conservation and analysis of this collection has revealed a remarkable amount of new detail about Iron Age cauldrons, and the replica has helped to bring some of this data to life, creating a wonderful, attention-grabbing object. One can only imagine people’s reactions to seeing a cauldron suspended above a fire, with its contents bubbling away inside – for many, these cauldrons must have made a deep impression and fostered a sense of occasion.
Just as the cauldrons represent a spirit of community and coming together, the Glenfield project has from start to finish benefited hugely from a team of specialists collaborating across different organisations to reveal the fascinating story of a long-lived Iron Age settlement and its unusual possessions. Conservation work was carried out in close consultation with Alexandra Baldwin and Jody Joy, who undertook the study and conservation of the Chiseldon group, while members of the British Museum’s scientific team who took part in the Chiseldon project were able to build on this with analysis of the metallurgy, mineral-preserved organics, and residues from Glenfield. This consistency has allowed for a comparative study of these rare and important objects, greatly enhancing the results and aiding the understanding and interpretation of future discoveries. We are very much looking forward to showcasing this work in the new museum display, and sharing the results in full when the project is published as part of the Leicester Archaeology Monograph series in 2026.
Source:
John Thomas is Deputy Director at ULAS.
Liz Barham is a Senior Conservator at MOLA.
Hector Cole is a Master Arrowsmith and archaeological ironworker (http://www.hectorcoleironwork.com).
We are grateful to Wilson Bowden Ltd for funding the entirety of this project.
Further reading:
• J Joy (2014) ‘“Fire burn and cauldron bubble”: Iron Age and early Roman cauldrons of Britain and Ireland’, Proceedings of the Prehistoric Society 80: 327-362.
• A Baldwin and J Joy (2017) A Celtic Feast: the Iron Age cauldrons from Chiseldon, Wiltshire (British Museum Research Publication 203).
All Images: University of Leicester Archaeological Services (ULAS), unless otherwise stated
