Ysgol yr Hendre, Llanbeblig, Caernarfon

An early medieval cemetery and Roman construction camp

Introduction | Excavation Results | Neolithic Pit | Roman Pit Ovens and Possible Construction Camp | Early Medieval Cemetery

Medieval Features | 19th and 20th Century | Acknowledgements and Further Reading

Roman Pit Ovens and a possible Construction Camp for Segontium Fort

Scattered across the site, within and beyond the later cemetery, were eighteen roughly figure-of-eight shaped features formed from two conjoining pits. In each case one of the pits had orange-red heat-affected soils along the base and sides and must have held a fire, while the other, although containing charcoal, generally lacked traces of burning. All these features ranged in length from 2.98m to 1.40m, in breadth from 2.0m to 0.65m and in depth from 0.58m to 0.12m. The evidence suggested that the fire pits were roofed possibly with turves supported on branches, creating an enclosed space that could be used for cooking. The other pit was used to rake out the remains of the fire.

Plan showing location of ovens (in black)


The features have been interpreted as ovens. It is usual in clay or pit ovens to remove the ash from the oven once it has reached temperature, insert the food, seal up the oven and leave to cook. The design of the oven is not one intended for long term use, and from the amount of charcoal in the raking pit and the intensity of burning in the ovens it is suggested that each oven was only used a few times.

Most of the ovens were separated from each other by a distance of 15-20m, though some were paired, and others formed a rough line, but there was little coherent pattern to their distribution. The very few small finds from these features included a corroded nail, tiny fragments of burnt bone, very occasional flint flakes and a single abraded piece of pottery.

This last is difficult to diagnose, but is almost certainly a small piece of Roman pottery. None of the finds seemed to be directly related to the use of the ovens. Analysis of the charcoal revealed that the fuel used in the ovens was mainly oak, with some hazel, ash and willow or poplar. Elm was also occasionally used.

Examples of the ovens

A few charred cereal grains were found including wheat, barley and oats, but not very many and it is possible that the grain was introduced with straw to light the fire.

The lack of finds made it difficult to date the ovens so charred fuel and cereal grains from seven of the ovens was radiocarbon dated. Statistical analysis of the results suggests that the ovens were in use between cal AD 25–80 and cal AD 60–120, over a period lasting no longer than 80 years and probably for only 1–30 years. The dates are actually consistent with all the ovens being used at the same time and it is the statistical error on the dates that give the potential range. The fragile nature of the ovens suggests that each one was used for only a short period, and their layout over the site could be consistent with them being used by small groups of people at roughly the same time.

Section through one of the raking out pits showing the layers of charcoal
and burnt soil from each use

A pair of ovens close together

Though it is not possible to prove with certainty that all the ovens were in use simultaneously, it is worth examining the hypothesis, and looking again at the statistical analysis. By combining all the dates, a much closer estimate of use can be arrived at which gives a time span of cal AD 65-80; a precise enough range to be compared to historical events. It has been suggested by the excavators of Segontium Roman fort, using historical evidence and pottery and coin evidence, that ‘the site is Agricolan, and dates to AD 77 or shortly after’ (Casey and Davies, 1993, 10). The Roman writer Tacitus tells us that Iulius Agricola, who was governor between AD 77 and AD 83, subdued the local tribe, which he called the Ordovices, and went on to attack Anglesey.

The idea that these might be Roman military field ovens is supported by the discovery of many very similar ovens in a Roman marching camp at Kintore, Aberdeenshire. There it was suggested that the ovens represent the position of tents of individual contubernia (groups of 8 men), and this interpretation would fit with the distribution of ovens at Ysgol yr Hendre, where the estimated date suggests they were in use during the construction of the fort at Segontium. This may, therefore, have been the site of a camp for the soldiers building the fort. Its location some 300m from the fort is not a problem, as other known construction camps are further from their forts, but a defensive ditch around the camp would be expected, and none were found during excavation. However, elsewhere in Britiain there is evidence that suggests that not all temporary camps had ditches. The use of tribuli, defences built from stakes lashed together, or other similar devices may have provided sufficient defence without the need of ditches.

Tribuli created by a re-enactment group
Copyright: Sean Richards, Legio IX Hispana, California

If the ovens do represent a construction camp for Segontium it is possible to imagine each contubernium camped at fairly regular distances apart, all with a tent and most with an oven. Pieces of leather, interpreted as panels from Roman army tents, were found in Roman wells near Segontium in 1920 and 1977. Just such tents could have been used at Ysgol yr Hendre and such a camp would have left few archaeological traces other than the ovens.

One of the Roman ovens fully excavated,
showing the sides and base of the fire pit burnt red