Historic England Research Records

(SS 733071) An earthwork formerly occupied the summit of the ridge on which Bury Barton (SS 70 NW 1) now stands. It was presumably a small hill-top camp of the type usually found in the late Iron Age. No trace of it remains today. (1)

Bury Barton is BERIA in 1086 and BURY in 1503. From burh. -This may have been some stronghold guarding the passage of the River Yeo here. (2)

A new fort first identified by J K St Joseph in 1975 above the west bank of the river Yeo. The remains consist of a fort platform of 4.5 acres (1.9ha), measuring 170m by 112m over the ramparts and part of a larger enclosure which may extend to c 21 acres (9 ha). Excavation directed by M Todd and S Goddard in 1983, trenched the north bank of
the smaller work to reveal a clay rampart with turfwork surviving at the rear. Behind this a pit filled with burnt material yielded South Gaulish Samian and other sherds. (3)

Further excavation in July 1984 showed the northern rampart to be about 4m. wide and surviving to almost a metre in height. Behind the defences, rubbish deposits included first century pottery and metalworking residues. Later Roman material, notably second or third century pottery, together with the fact that the defences were not systematically levelled on abandonment in the first century suggests, that the site may have been reused at a later period. The end of a first century timber building was identified in the interior. Other finds included artifacts in Beer flint one or two of which may be Neolithic and post-medieval sherds from gullies. (4)

A rapid examination of air photography (6) shows this Roman fort, visible as an earthwork, at Bury Barton. Parchmarking of the northeast corner of the fort indicates that it was double ditched. The remains of the larger enclosure are also visible, being ploughed between 1985 and 1989; its exact purpose is unclear. (5-6)

Bury Barton. Two distinct Roman military works. The smaller fort is about 1.9 hectares (4.5 acres) in area, and is entirely within the the larger, which is either a camp or campaign-base up to 8 hectares (19.2 acres) in extent. The south side can be easily traced for 100 metres, but beyond that point it has been ploughed out and the air photographs do not reveal its course. The west side can be traced along its entire length, except where it is overlain by farmbuildings and modern terraces. Immediately north of these and the lane its followed by a modern hedge for a short distance and it is thereafter clear for a further 90 metres to the site of the northwest angle. The northern side has all but vanished, it seems to have run along the very edge of the pronounced scarp which falls to the River Yeo beneath. The position of the northeast and eastern side is not yet known. The smaller enclosure is bounded by a striking earthwork, traceable around all the circuit except where it is overlain by the Mediaeval farm. The northeast angle is clearly evident in pasture and the northern side is easily traced to the northwest angle of a modern hedge lying along the line of the ditches for more than half of its couse. At the mid-point of this side, the rampart is apparently interrupted by a gate-opening. There is no interruption in the well-preserved south side, suggesting that no gate existed here. If this is the case, then this work would appear to have faced north, with its porta praetoria in the northern side. There is no surface indication of the site of the east gate, while the likely position of the west gate is inaccessible. There is little to report as yet about the detailed chronology of these two military works. The pottery so far recovered from the fort indicates an occupation in the Claudio-Necronian period, but to little material is available to fix the opening date with any precision. Abandonment early in the Flavian period is to be expected, probably at a date not far removed from the mid-seventies when the legionary fortress at Exeter was given up. The presence of a small quantity of later Roman period probably of second or early third century is interesting, particularly as it is evident that the defences were not systematically destroyed at the close of occupation in the first century. This may indicate that the site performed some role as an official installation in the second or early third century. Its defences appear unusually large for a temporary camp and looks to be more appropriate to a campaign-base.
One of the places listed in the Ravenna Cosmography, `Nemetostatio', includes the element `Nemeton' and should be clearly sought in this region. This has commonly been identified with the Roman fort at North Tawton, (SX69NE2), but the case is weak. This site lies on the Taw(Tavus?) not the Nymet(now known as Yeo) and on the margin of the known distribution of the Nymet place-names. Bury Barton must be accounted a much better candidate, lying as it does on the Nymet and well within the enclave of the names in Nymet. The presence of material later than the first century here adds a modicum of support to the identification. A `Statio' which served offical purposes should be expected to produce evidence of occupation later even though this may not have been intensive. Although not amounting to proof, these several indications combine to make this the most likely for `Nemetostatio' yet known to us. (7)

Parts of the features described above were recorded as cropmarks on HIstoric England 2018 oblique aerial photographs. (8)