ISSN 1357-4442Editor: Simon Denison

Issue no 31, February 1998


Ridgeway hillforts reveal their little differences

A hillfort is a hillfort, right? Same name, same function? Not so, according to new work on hillforts on the Ridgeway in Oxfordshire.

Over recent years, interpretations of hillforts have changed. Once they were seen as largely defensive, but more recently they have been seen as places for ritual and communal activities of various types. These interpretations have generally been thought to apply to all hillforts equally.

However, excavations near the Uffington White Horse, led by Oxford University archaeologists Gary Lock and Chris Gosden, have shown that different hillforts were put to completely different uses in late prehistory, however similar they look today.

Uffington Castle, close to the White Horse, was built on a site with ancient sacred associations, with a Neolithic long barrow and early Bronze Age round barrows. The White Horse itself was carved in the late Bronze Age, and the hillfort was sited at the end of a Bronze Age territorial boundary.

Iron Age use of the hillfort was intermittent, with a few pits and no structures, but there was a concentrated ritual re-use of the site in the late 3rd and early 4th centuries AD. Coins, pottery, animal bones, and a small oven were found inside the fort; an enclosure with burials was built outside the entrance, and the long barrow was re-used for burial. The impression, according to Dr Lock, is of ‘a temple or shrine’, where eating and drinking took place, and money-offerings were made within ritual ceremonies.

‘This was a long serving ritual complex looking back to the past through mythology and perhaps genealogical linkages to the landscape, ’ he said. It recalls other prehistoric sites around Britain that seem to have been re-used in the Roman period for ritual (see BA, November).

Segsbury Camp, 12km from Uffington, however, was a straightforward settlement, densely occupied by roundhouses throughout the Iron Age. Although pits contain deposits that suggest ‘personal’ ritual, the site was largely domestic in function. Nothing Romano-British has been found at all.

Between these two hillforts is Rams Hill, excavated in the 1970s, which also seems to have been largely a settlement, although the extent of occupation – from the late Bronze Age right through to the Romano-British period – far exceeds that of Segsbury.

A fourth hillfort is Hardwell Camp, just 2km from Uffington – unexcavated, little known, but curiously different in location from the other hill-top forts. Lying halfway down the Ridgeway slope, it is tucked away in a curve and invisible from most angles. This odd positioning suggests its builders had some specialist function in mind – although exactly what that may have been remains a mystery.

Return to Table of Contents | Return to CBA Homepage

First Dark Age settlement in mid-Wales

The first Dark Age settlement site in mid-Wales, dating from the late Roman period to around the 6th century, has been found on a hill near Welshpool in Powys. All previously known settlements of this date in Wales lie around the coastline.

The site, defended by ramparts, contains imported early medieval French pottery and continental glass, such as ‘cone beaker’ drinking vessels which had to be drained by the drinker as they had no base to allow them to stand upright. Imported glassware and pottery of this type is known at contemporary coastal sites such as Dinas Powys near Cardiff and Tintagel in Cornwall. The discovery a few years ago of cone beaker glass at Much Wenlock Priory, however, in Shropshire, and now at this Welsh site, may indicate that the Severn – which is navigable as far as Welshpool – was used as a trading channel in the period.

The site, excavated by Chris Arnold of the University of Wales Aberystwyth and Jeremy Huggett of Glasgow University, contains late Roman pottery, glass, and some datable coins, as well as evidence for early medieval metalworking. The overall size of the settlement and its status, however, are as yet unclear. Imported pottery and glassware have traditionally been associated with sites of high status, according to Dr Arnold, but that may be misleading because no sites of the period have been found without them. ‘Instead of a hierarchy of settlements, there may rather have been a hierarchy of people within each settlement, ’ he said.

Return to Table of Contents | Return to CBA Homepage

Villas sited ‘away from major towns’

A survey of Roman villas in south-east England suggests that the majority did not cluster around major towns – as has commonly been assumed. Almost all, however, were sited close to waterways, facing south in fertile valleys, and most lay fairly near major Roman roads.

The survey, conducted by the London based archaeologist Harvey Sheldon and four colleagues, and published in the latest issue of London Archaeologist, covers all known villas in Bedfordshire, Buckinghamshire, Hertfordshire, Essex, and Middlesex. It follows an earlier survey by Mr Sheldon and colleagues of villas south of London in Kent, Surrey and Sussex (published in the same journal in 1993), and reaches broadly similar conclusions.

Of 298 possible sites examined, 61 were judged as certain or probable villas – following at least partial excavation that allowed the form of the building to be established.

Of major Roman towns in the area, only St Albans seems to have been a focus for villas, with 16 certain/ probable sites within 20km (four hours travel, there and back, on a horse). Colchester had six within the same radius, Chelmsford four, and London only two. By contrast, two thirds of the villas lay within 20km of a minor town. The reason for so few villas near London, the survey suggests, may have been because of the area’s poor quality soils.

All but one villa lay close to a river or stream – useful not only for domestic, agricultural and industrial reasons, but also for transport. It is generally assumed that most villas derived their wealth from the sale of agricultural goods to market, and the costs of river transport are estimated to have been five times less than those by road. According to the survey, the vast majority of sites would have been accessible by shallow craft, and two produced evidence for a wharf or quay.

Access to major roads seems to have been less of a priority. Two thirds lay within 10km of a major road (compared to 80 per cent in the south-of-London survey); while only a quarter of villas in both surveys lay within 1km of a road. These, it is suggested, may have doubled up as mansiones, or travellers’ lodgings. Most villas were sited in river valleys and, where the alignment is known, 90 per cent faced south. A third of the villas are known to have been built in the 1st century, and, where dating evidence exists, most were occupied more or less continuously through the 2nd and 3rd centuries. Some continued into the 4th.

London Archaeologist is published quarterly. Details from 8 Woodview Crescent, Hildenborough, Tonbridge TN11 9HD.

Return to Table of Contents | Return to CBA Homepage

In brief

Date of Ridgeway
Excavations at the Ridgeway hillforts have produced evidence for the date of the Ridgeway itself at Uffington Castle. The compacted chalk of the path was found to run over a filled-in late Bronze Age ‘linear ditch’, or territorial boundary, suggesting (not conclusively) that the path, at this point, dates from the early Iron Age at the earliest.

Stirling Castle
The site of the medieval private royal chapel at Stirling Castle, where King James IV prayed for forgiveness for plotting the death of his father James III, appears to have been confirmed by the discovery of six skeletons, all aligned east–west, under a flagstone floor in a former army cookhouse in the castle. The existence of the chapel was known from documents, but not its location. The skeletons are thought to have belonged to members of the governor’s household. The Historic Scotland excavation also produced the remains of a medieval tiled floor and sherds of pottery. The skeletons will be radiocarbon dated, and then reburied.

Medieval bridge
Parts of a 12th century stone bridge have been found 10ft below street level in the centre of King’s Lynn in Norfolk, once one of the country’s busiest ports. The bridge spanned the River Purfleet, and was made of high-quality, white Caen stone, quarried in France, which was also used in a number of cathedrals, including Norwich and Canterbury.

Mesolithic pet bear
A bear seems to have been kept as a pet, or as a performer, by an itinerant Mesolithic group in central Europe. A 7,000-year-old brown bear’s jawbone found in a cave on the French/Swiss border seems to have been fitted with a muzzle, perhaps of leather, from an early age to the bear’s death at the age of six. According to the Swiss archaeozoologist Louis Chaix, writing in the latest Journal of Archaeological Science, the bear’s relatively early death suggests it may have been ritually slain.

Dark Age fish-traps
The number of early medieval fish-traps known in Britain is increasing. In Essex, a series of large, timber fish-traps around the coastline were long regarded as medieval, but two have now produced mid-Saxon radiocarbon dates of the 7th–9th centuries. The traps are in the Blackwater Estuary and near Bradwell-on-Sea.

Meanwhile, up the Thames in London, a mid-Saxon ‘wharf’ originally thought to be part of King Offa’s palace at Chelsea, and given wide national publicity last year (see BA, September), has on further examination been re-interpreted – as yet another Saxon fish-trap.

Essex Archaeology, an annual round-up of archaeological news from Essex, is now available free from the county council’s archaeology section (send an A4 SAE).

NEWS is compiled by Simon Denison

Return to the British Archaeology homepage

Return to the CBA homepage

Council for British Archaeology, 1998