(Centred at NY 84939651) Camp. (1)

The temporary camp at Bagraw is situated on a long narrow shelf between Dere Street and Bagraw Burn. This situation made it necessary to adopt unusual proportions. There are two divisions, the whole work being sub-divided at a later date to make a smaller north camp.
The south angles of the camp are visible, but the south side has been obliterated by the dykes of a field leading to Dere Street. The east side is 1497 feet long, the west 1575 feet and the north 539 feet. The original north and south gates have disappeared but the south gate of the reduced work, covered by a traverse, survives in a central position. There is an east gate with traverse 340 feet south of the north front. At a corresponding point on the west side there is a gap with a suggestion of a traverse in front of it. It seems likely that the first work, like the second, had four gates only.
The area first enclosed is large enough to accommodate a legion, while the later division would take a force about half as big. (2)

(Consulted, mentioned, no additional information). (3)

Description by Authy 2 generally correct except that there seems to have been originally double ramparts at the front of the larger camp. These are still visible although badly mutilated and partly robbed. See sketch survey. (4)

No change from report of 17 7 56.
Published survey (25") revised. (5)

Resurveyed at 1:2500. (6)

Surveyed by RCHME. (7)(8)

This camp has been re-assessed in connection with RCHME's survey and publication of Roman Camps in England. The following descriptive account is taken from the published text.
The camp at Bagraw in Redesdale, approximately equidistant from the forts at Blakehope and High Rochester (Bremenium, is situated at 200 m above OD, along a relatively flat-topped but locally prominent ridge, aligned NW to SE. The camp overlooks the valley of the Rede immediately to the W, and is protected by the steep sided ravine of the Bagraw Burn on its SE flank. The modern A696 road cuts obliquely across the SW quarter; its Roman predecessor, Dere Street, runs parallel with the SW rampart of the camp and only 5 m outside it, following the line of the ridge. To the S of the modern road, the line of Dere Street survives as a terrace in the hillside, with a single scarp 0.5 m high; to the N it is a cambered linear mound up to 11 m wide and 0.4 m high. The relative chronology of the camp and the Roman road cannot be demonstrated from fieldwork alone, since the alignments of both are determined by the topography. The likelihood is that the road preceded the camp although the absence or poor survival of entrances also makes this difficult to prove.
The earthworks are extremely unusual, consisting of two elements, set end to end. The layout of these two parts, although superficially rectangular, is most irregular. The E and w sides of each part lie parallel to one another but are realigned by about three degrees close to their midpoint. Of the three short sides, aligned approximately NE to SW, the S and central ones are approximately parallel but the N side is markedly askew. Thus the N element of the earthworks is a trapezium, whereas the S one is a parallelogram. This difference in layout may be significant in the determination of function and relative chronologies.
The N part occupies an area of 3.6 ha (8.9 acres). Its N side has an abraded outer ditch 0.3 m deep; the rampart here is up to 0.4 m high but towards its W end it is reduced to a single scarp. The N part of the E side is better preserved within a plantation, the rampart standing 1.1 m above the bottom of the ditch which is as much as 0.7 m in depth. The S side, the medial feature of the earthworks as a whole, has been disturbed by tracks but the rampart survives to a maximum height of 0.8 m high and the ditch is 0.3 m deep. On the W, however, no ditch survives alongside Dere Street, but the rampart stands up to 1.1 m high externally and 0.5 m internally. A later bank and ditch cut across the defences and the line of the road; their function is unclear.
Two gates are clearly visible: on the E, to the N of the central point of this side, there is a narrow gap, now only 3 m wide, guarded by a traverse mound 0.4 m high; on the S, a traverse mound of similar height, with a fragment of an outer ditch, 0.4 m deep, protects a gate 6.5 m wide. This gate is about 5 m off-centre and from it the low agger of a road, no more than 9 m wide and 0.3 m high, runs northwards at a slight angle to the axis of the camp. The point on the N side towards which this road is heading has been severely scoured and abraded and no gate can now be identified on the surface. The line of the W rampart is broken approximately opposite to the gate on the E side but no unequivocal Roman features survive here.
The S part of the earthworks, enclosing 3.5 ha (8.6 acres), is more difficult to interpret. The E side has been eroded by the steep slopes above the burn but where it survives at all the ditch is as much as 0.4 m deep; the rampart stands 1.4 m high externally and 0.4 m high internally. The fragment surviving at what appears to be the SE angle is not on the same alignment as the larger portion to the N, perhaps suggesting that an adjustment had to be made to position the earthworks on the crest of the natural slopes. The W side of the enclosure, a bank 1.0 m high, is featureless. As in the N part, no W entrance is visible, but it is conceivable that the whole of this side may have been reused as a later boundary bank, thus blocking any possible Roman gates. The passage of the modern road through the short S side, slightly to the E of its midpoint, may suggest the former presence of a break in the earthworks here. Otherwise, however, this portion of the perimeter is somewhat confusing. It seems that a drain has been cut inside what appear to be Roman defences and the upcast deposited on its S side. To the S stand the broken remains of what seems to be the Roman rampart, still up to 1.0 m in height.
The junctions between the two parts of the earthworks have been much disturbed and thus their relative chronology is in doubt. MacLauchlan's proposal (1825a, 32 (9b)) that the camp was double in size is not entirely satisfactory, for the traverse outside the 'medial' defences is unlikely to have survived if the S half was subsequently occupied. It is possible, however, that the S part of the earthworks played a subsidiary role, forming an annexe to a contemporary or slightly earlier camp to the N, the semi-permanent status of which may be indicated by its well-developed quasi-axial roadway. The irregularities of layout, especially the difference in alignment between the E and W sides, rule out the suggestion (Richmond 1940, 120 (9c)) that this was one elongated camp which was subsequently divided into two. Full information is included in the NMR Archive. (9a)

General association with HER 12392 (Dere Street). (9)

Ramparts of Bagraw Roman Camp (actually 2 camps, one inside the other) are clearly visible on lidar imagery, but nothing noted to add to what is already known of the site through previous survey. (10)