Corbridge, Northumberland, evaluation of a Roman bridge and Saxon watermill

An evaluation was carried out on 2 structures beside the River Tyne under threat from erosion. These were the southern abutment of the Roman bridge and a structure on the northern side, which can now be identified as an Anglo-Saxon watermill. The work, which was jointly funded by EH and Northumberland County Council and undertaken by Tyne and Wear Museums, established these as monuments of national importance and assessed the extent of the threat from erosion. The study area lies at the bottom of the escarpment occupied by the Roman site of Corbridge. The present village of Corbridge is c 1km to the east, and was founded in the Anglo-Saxon period; it is assumed that the collapse of the Roman bridge caused the settlement shift to a location where the river could be forded.

The bridge carried Dere Street, which approached from the south-east. The line of the bridge is marked by the remains of 6 stone piers on the river bed, and the drought conditions of 1995 allowed a brief appraisal to be made of these. Although the position of the northern abutment is unknown, it is likely that the bridge crossed the original course of the river at right angles. Subsequent changes in course have been considerable and the remains of the abutment, occupying an area 11.2m by 6.4m and built of large dressed stone blocks, now lie in shallow water. In recent years masonry has also been exposed on the bank, as walkers have eroded the turf.

Downstream from the point where the line of the bridge meets the bank is a level shelf of boulders, cobbles, and gravel, covered by shallow water, extending out from the bank for a distance of 1220m before dropping steeply into the main channel. At the eastern end of the shelf, c 90m downstream, are the remains of the watermill. The edge of the shelf probably represents the waterline in Anglo-Saxon times, with the watermill standing on a low terrace, now flooded. The submerged remains occupy an area 18m by 7m and consist of a platform of large dressed blocks derived from the Roman bridge and large timbers for which radiocarbon dates in the 9th and 10th centuries have previously been obtained.

Survey and limited excavation of the bridge identified the masonry on land as a road ramp, in use with the abutment, although not bonded into it. The core of the ramp was mainly of rubble but included large worked blocks. A retaining wall at the south-east side was stepped up into the bank and ran roughly at right angles to the line of the bridge. Four courses survived, composed of worked blocks, the average size of which was 1m by 0.5m by 0.3m. The base of the revetting had been undermined by scouring, causing the masonry to slide forward; a line of blocks, originally joined by iron bar clamps (since robbed), lay tilted at an angle of 45 degrees. Masonry fallen beyond the edge of this revetment included 2 large moulded blocks with slots in their upper faces, probably to carry a balustrade, and an octagonal moulded plinth, 0.8m across, interpreted as the support for a statue base. These must represent the topmost courses of the ramp, the first to collapse when their metal clamps were robbed. Stone from the lower courses would have been removed by later stone-robbers. Planning of the abutment in the water revealed important structural details not recorded by previous survey in the 1960s, as well as indicating the results of continuing erosion.

The watermill on the northern side of the river was of the horizontal-wheel type, in which the wheel is located in a basement, connected by a shaft to the mill-stone in an upper room. Water from a leat is channelled through a timber chute into the wheel-room, the floor of which must be substantial to withstand scouring. The only other example of this date, at Tamworth in Staffordshire, was entirely timber-built; however, the western end of the Corbridge mill comprises a wheel-house floor of dressed blocks derived from the Roman bridge, each roughly 1m long and 0.6m wide. Bounding the east and west sides were heavy timbers, the largest 7m long and 0.4m in section, others 5.5m in length. There was a well-preserved pattern of slots in the timbers for waterchute emplacements and uprights for the timber superstructure. To the east lay another stone floor, a large concentration of tumbled blocks derived from the Roman bridge and a further area of paving, showing the structure had been complex, possibly of several phases. Limited excavation revealed a line of stakes running under the present river bank, possibly the timber side of a millpool. A timber waterchute was discovered, no longer in situ, but lying on the edge of the shelf of boulders, c 20m from the river bank, one end tilted into deep water. It closely resembles an example from Ireland, but is much larger. Diving carried out in May 1996 identified it as a single massive timber, 10.3m long, with a flared top 0.7m wide, and a hollowed-out channel down its full length. Although both sites have been the subjects of limited study or observation in the past, the new information gathered through this evaluation has emphasised their archaeological importance. The masonry revealed on the south bank belongs to a massive bridge, its elaborate decoration possibly including statuary. Construction techniques of the abutment parallel the 2nd-century stone bridge at Chesters on Hadrian's Wall so closely as to suggest being part of the same building programme. However, the remarkable state of preservation of the remains so far revealed suggests that much more masonry is still in situ than at Chesters.

The watermill is unique in the northern region and only the second well-preserved example of this period in England; the evidence suggests that a millpool and associated structures such as leats are likely to be well- preserved beneath the present river bank. The waterchute is apparently unique in this country. Such a substantial structure has important implications for the status of Corbridge in the Anglo-Saxon period. Furthermore, with the exception of princely sites such as Yeavering, nothing is known of the secular aspects of settlement sites in the north-east before the 11th century