More information : (NY 5655 7460) Roman Fort (GS) (1)
The Roman name was probably 'Banna' (Rudge Cup, altar inscription at Birdoswald and physical site - Banna = spur or headland (2)) [ NB now Banna is believed to be Birdoswald, see source 6]. It was apparently an outpost to control the district and the natural corridor into Cumberland, not a halting-place on a military way; there is no acceptable evidence for the continuation past Bewcastle of the road from the station at Birdoswald. (2-3)
Excavations in 1937 showed that the wooden buildings in the Hadrianic multangular fort were replaced in the third century by stone structures of the conventional type. These were looted and burnt, probably in AD 297, and rebuilt, the occupation ending with the disaster of AD 367. Roman objects found in and near the fort include an almost perfect Samian bowl (Form 37) discovered in 1934; two altars, found in the churchyard and in the excavations of 1937; two silver plaques with an embossed relief of Cocidius, and various inscribed stones, pottery, etc. (4-5)
Originally a sub-hexagonal fort, only the remains of two sides in the west now survive, the remainder having been destroyed by water erosion in the south and east, and the moat of Bew Castle and modern farm buildings at the NE angle. The remaining sides consist of a much mutilated earthen rampart the outward-facing scarp of which attains a height of about 4m. There appears to be no original gate, and any accompanying ditch has been obliterated partly by an old field bank, and partly by the modern road. The internal area covers some 6 acres and is largely occupied by St Cuthbert's Church and graveyard; farm buildings and field boundaries, but there are a number of obvious turf-covered foundations. It is impossible to say whether they are Roman or associated with the medieval Bew Castle.
Published Survey revised.
Name 'FANVM COCIDII'? accepted for 4th. edition R.B. Map. Banna is now thought to be Birdoswald. (6)
NY 565 747. Three phases of Roman occupation were revealed when excavations were carried out, in advance of the construction of a barn, in the N part of the hexagonal Roman fort at Bewcastle. The first phase was of timber buildings, possibly of Hadrianic date. These were replaced in the second half of the second century with stone buildings. The third phase, of late 3rd century date, included the destruction of a store room which was replaced by a barrack block. Other features uncovered during the excavation included the rampart which was composed of dumped sand contained by a rear stone curb. Occupation during the 4th century was attested by the discovery of unstratified 4th century pottery sherds. (See plan). (7)
Roman Fort. Grade 1. (8)
Centred NY 5653 7459. Earthwork remains of Roman fort much eroded, especially on the east side where the remains could be confused with old field banks. Earthworks also destroyed by later castle (NY 57 SE 4), church and farm. The original fort perimeter is clearest on the NW, SW and SE sides. The interior shows numerous earthwork banks and features, but these may be later. Rectilinear earthworks, possibly building foundations, lie just to the west of the fort at NY 5639 7461. (9)
The fort at Bewcastle, centred at NY 5655 7460, was surveyed by RCHME in 1984-5. The fort, probably FANUM COCIDI (10a), occupying a knoll of glacial sand and gravel, is an outpost fort of Hadrian's Wall. It is irregularly hexagonal in shape, presumably to maximise the natural defensive potential of the site. Much of it is destroyed or obscured by erosion, and also by the superimposition of later buildings both within it and on the rampart, including St Cuthbert's Church and churchyard, the former rectory gardens, Demesne Farm and outbuildings, and the moated Bew Castle (see NY 57 SE 4). The only area which is at all well-preserved is in the SW where the surviving rampart follows the crestline of the natural scarp, with a single external ditch, found by excavation to be about 5.0m below the level of the present crest. Full information is available in the NMR archive. (10)
The results of the RCHME survey have been published. (11)
The results of the excavations in 1949, 1954 and 1956 of the internal bath house were published in 1993. (12)
Excavations uncovered four phases;
Phase 1 was a turf fort dated to c.AD 122-139/42 with at least one stone built gateway with internal buildings of timber and stone. An inscription suggests the fort was built by the Cohors I Dacorum. After a short period of abandonment coinciding with the move into Scotland the fort was reoccupied in c.AD 163.
Phase 2 saw the construction of a stone fort and the replacement of timber buildings with stone ones. A building dedication suggests part of Legio VI Victrix formed the garrison at this time.
Phase 3 saw a drastic reorganization of the fort's interior during the late 2nd/early 3rd century. Many buildings were rebuilt or altered and new barracks built. The garrison may have been the Cohors I Nerva Germanorum.
Phase 4 Towards the latter quarter of the 3rd century the fort was again subjected to drastic remodelling. A new wall was built on the west and north sides, considerably reducing the area of the fort indicating a much reduced garrison in the late 3rd/early 4th century. The internal buildings were also altered, the bath house may have been converted into a barracks.
Coin evidence suggests abandonment in the early 4th century AD. (13) |