More information : (NY 90021240) Roman Marching Camp (R) (Site of) (1)
The Roman Marching Camp at Reycross is of quadrilateral form, the rampart being constructed of surface material, with a ditch at the NE angle. The existing gates, some of which had returns on either side, were covered by tutuli; the street lines have vanished, but by joining opposite gates the layout form suggests square castramentation that preceeded the tertiary plan. The kink in the Roman road (RR 820) at the east gate suggests the camp pre-dated the road, indicating a very early date (1st c AD). A slight square mound (NY 90001235) where pottery of an unspecified date was found (a) is of unknown age or purpose. Plan. Mainly as described and planned by Richmond. (2)
Surveyed at 1:10 000. (3)
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.
One of the most impressively defended camps in Britain was constructed at 447 m above OD on the summit of the road over the Stainmore Pass, the route across the Pennines, now the A66, that links Teesdale with the Eden valley. The main Roman road from York (Eburacum) to Carlisle (Luguvalium) and into western Scotland used this pass, linking the local forts at Bowes (Lavatris) and at Brough under Stainmore (Verteris). The defences were thrown up on the shoulder of Buzzard Hill which extends southwards, creating at this point a steep slope down to the N bank of the River Greta. The W defences of the camp were laid out along the crest of this shoulder so as to take full advantage of the unrestricted view over the water shed, 2 km to the W. Similarly, the S defences occupy the crest of the rocky hillside above the river, eliminating any dead ground on this side. The position of these two sides and the length of the S one were dictated by the topography (contra Richmond and McIntyre 1934, 51 (see auth 2)) but there were no such strong factors determining the alignment of the N and E portions of the perimeter. As a result, none of the corners of the camp is a right angle; the N side was perhaps intended to be parrallel to that on the S, but on the E this was not even attempted. The E side was probably only prevented from being at right angles to the other two by the need to avoid some marshy ground; there was no disadvantage in this, however, and the position gives extensive views to Bowes and beyond. To the NE and N the ground is relatively level, descending only gradually into the head of Deep Dale and, farther away, into Baldersdale.
The camp encloses a comparatively level area of 1.8 ha (20 acres), which slopes only very gently from the NW quarter towards the S and E. To the S of the modern road the ground is more broken and there are gullies and ribs of rock within the interior. The rampart forming the perimeter is unusually massive, still standing up to 1.8 m high in the centre of the S side and measuring as much as 11 m wide across its base. It is evident that a substantial quantity of stone was used in its construction; surface indications suggest that the material was not, in the main, derived from a conventional outer ditch but may have been gathered or scraped up from either side.
However, the excavation in 1990 of the short length of rampart that was due to be destroyed in the construction of the S carriageway of the modern road revealed the presence of a berm about 1 m wide and a substantial but very irregular ditch, up to 2 m wide and 0.8 m deep, of which there was no trace on the surface. MacLauchlan (1849, 351 (3a)), writing on the contemporary quarrying, said that a ditch had not been identified, but this view may have been based on surface observations alone. Only along part of the N side and next to the most northerly of the E gates are there any surface signs of an external ditch, here no more than 0.4 m deep. In 1989, probing through the peat at the NW angle by RCHME suggested that there might have been a short length of external ditch there, perhaps as much as 0.8 m deep, its centre being about 4.7 m from that of the rampart.
Each of the nine gates that can still be identified was defended by a traverse, the mounds of which are oval rather than bar shaped. These now vary in height; those on the S are 1.4 m high, and the northernmost example on the W side survives to a height of 1.6 m; outside the E and W central gates they are 1.2 m and 1.0 m high respectively, but at the N gate on the E and along the N side the traverses are only between 0.4 m and 0.6 m high. These latter measurements may well be affected by the presence of the raised peat bog which is here at least 0.4 m thick. None of the traverses now exhibits any sign of an attendant outer ditch. It is likely that the N carriageway of the modern A66, shown on the plan before the construction of the S carriageway began in 1991, and which follows the line of its Roman predecessor, entered and left the camp through two further gates. Excavation, however, has suggested that the easterly of these gates may have been slightly farther to the S or may have been widened subsequently. Some cobbling, perhaps that of a road, and some small post-holes seem to have been stratigraphically later than the rampart (Robinson forthcoming (3b)). With the exception of the poorly preserved W example along the N side, each gate is flanked by inturned rampart-terminals, unusual features which seem to belong exclusively to forts and camps constructed in the 1st century AD (Jones 1975, 118 (3c)).
The rampart shows little variation in form, although it is noticeably broader at the NW and SE angles; in the centre of the S side and close to the central gate on the W it also has a markedly flat top, 2.0 m across. In the W third of the N side the rampart is only visible intermittently through the peat bog. Although this ground in itself may have provided some natural defence, especially in wet weather, it is nevertheless surprising that probing does not suggest the survival here of any substantial features below the surface. This is particularly unexpected when contrasted with the proportions of the rampart elsewhere.
Along the W portion of the S side the rampart is impressive, standing 1.5 m high internally and forming a false crest above the Greta valley. Despite th evidence for a ditch elsewhere, the material here seems to have been scraped up from the interior, an impression accentuated by the presence of a sinuous, natural scarp immediately to the N. However, some of this rampart may not be artificial: it seems that the constructors may have chosen to cut back the N side of another natural rock spine to form the rearward scarp. This may also be the most likely explanation of what appears, at first glance, to be a single course of outer facing-stones, 4.0 m long, which again is probably a natural rock outcrop. The traverse mound of the W gate here seems to have been carved out of another such spine of rock.
Knowledge of the interior of the camp is limited. Blanket bog has invaded much of the NW and as much as a quarter of the interior has been removed by surface quarrying for limestone, apparently in the early 19th century between the visits made by Roy (1793, 73-4, pl xvii (3d)) and MacLauchlan (1849, 350-1 (see auth 3a)). However, when small portions of the area to the S of the modern road were excavated in 1990 some randomly disturbed stake-holes were identified, associated with 4th century pottery (Robinson forthcoming (see auth 3b)). In the 19th century some pottery was found by the quarrymen in the vicinity of the mound on the summit, to the N of the Rey Cross itself, but the character of the sherds was not recorded (MacLauchlan 1849, 351 (see auth 3a)). The irregular shape of the camp and the provision of only two gates on the topographically restricted S side must have made the internal planning awkward.Richmond and McIntyre (1934, 54-6 (see auth 2)) attempted to resolve this presumed problem, but their suggestions are not readily verifiable and the calculations do not seem to take account of the broken ground along the S side. It is unlikely that this area was ever intended to be occupied, although it had to be included within the defences in order to place the perimeter on the most advantageous topographical line.
Despite the massive size of the defences, the camp does seem originally to have been for temporary occupation only. This is suggested by its carefully chosen, albeit awkward, defensive position and the fact that it lies halfway between the forts at Bowes and Brough, each less than 12 km away. In Richmond's phrase, this was 'a superb example of castra aestivalia, placed at the gate of Cumbria by an army intending to march deep into that territory the next day' (Richmond and McIntyre1934, 52 (see auth 2)). It would appear that this initial occupation took place early in the Roman conquest of the North, probably in the Flavian period. This is suggested by the inturned terminals of the rampat at the gates, which seem to be a Flavian characteristic (Jones 1975, 118 (see auth 3c)), and by the realignment of the Roman road (noted by Richmond and McIntyre 1934, 57 (see auth 2)) at the point where it crossed the E rampart. This realignment indicates that the camp already existed when the road, conventionally thought to be late 1st-century in date, was surveyed and constructed. In addition, because sufficient land exists close to the forts at Bowes and Brough for a camp of this size, it would seem that this position, high up on the pass, would only be preferable in the period before either of those forts was established, sometime in the AD 70s or 80s.
Whatever the initial date of the camp, there is little doubt that the defences, astride this important Roman road, were reused. Late 3rd or 4th-century pottery was found in the upper fill of the ditch and within the interior in 1990 (Robinson forthcoming (see auth 3b)), and some of the gates -those on the S, and the N ones on the E and W sides - show signs of having been blocked. This is most obvious on the W, where the N gate has been closed by a low bank standing 0.5 m high and consisting mainly of stone. At the other gates that seem to have been blocked, only a small number of boulders survive in position, continuing the line of the rampart acroos each gap. The date of this work is unknown.
Three features within the interior should be mentioned briefly. The first is Rey Cross itself (NY 81 SE 4) a simple shaft reduced to a stump. Its history as a boundary stone, from at least as early as the late 13th century, has been set out by Collingwood (1927 (3e); Robinson forthcoming (see auth 3b)) but the name itself is Scandinavian, and therefore the cross is presumably much older (Smith 1928, xlvi (3f)). With the building of the dual carriageway, the cross was moved some distance to the E, outside the area of the Roman camp. A few metres to the N, on the opposite side of the line of the Roman road and on the summit, is what appears to be a square mound, 0.8 m high and about 15 m across. It is surrounded by quarrying on all sides except the SE. Recorded by Hutchinson (1776, 15 (3g)) and by Roy (1793, 74 (see auth 3d)), it seems to have been respected by quarrymen. Its origin is uncertain. It could have been sepulchral but the most closely comparable earthwork in the vicinity is the signal station adjacent to the camp on Bowes Moor, 3 km to the E (Annis forthcoming (3h)). If there was a chain of signal stations over the pass (Farrar 1980 (3i)) it is conceivable that this mound, which when seen from Bowes Moor is on the skyline, was a part of that system.
In the NE quarter of the camp is an irregular group of recumbent stones which, it has been suggested (Vyner in lit (see auth 3g)), may have been a stone circle. It is possible that this is what MacLauchlan (1849, 350 (see auth 3a)) thought was a 'tumulus'; he gave no description of it but if the identification was correct then the material within the kerb has been robbed. Its significance here is that any comparatively early structure in this position could have interrupted the line of the internal road, running from gate to gate, in between the tents of the Roman camp (see NY 91 SW 13). Full information is included in the NMR Archive. (4)
The condition of the site was assessed by RCHME following the widening of the A66 trunk road. Two short lengths of the N-S ramparts had been lost. (5)
Additional references. (6-7) |