More information : [SX 9235 9532] An earthwork, on the crest of Stoke Hill, was discovered in the routine examination of air-photographs by Archaeology Division, in 1953. Field examination showed that the only extant remains were part of an outer earthwork, consisting of banked inner ditch, c. 450 feet in diameter. There was no trace of the squarish inner enclosure, 130 feet wide. The area was strewn with Md. and later potsherds but nothing Roman was found. Two small trial-holes in the central earthwork revealed no sign of wall-footings (1) (2) The site was excavated by Fox and Ravenhill (3) in 1956/7 by sectioning the defences and stripping most of the inner area. Finds included a coin of Carausius, a few Roman potsherds and 18/20th c. rubbish. The clay sub-soil rendered excavation inconclusive but the excavators accept the earthwork to be a Roman signal station which either lacked a watch-tower or the tower was a simple timber structure which has left no traces. The site commands extensive views in every direction. (1-3) The site is under plough and all intelligible remains of the earthwork have been destroyed. (4) A fortlet, used as a look-out post and signal station, on Stoke Hill. It supplemented the 1st C. fort at Exeter. (5) The double-ditched enclosure at Stoke Hill was first thought to be a late Roman signal station, but then comparison with Martinhoe and Old Burrow led to the suggestion of a 1st century date. Another possibility is that the site is pre-Roman; doubled-ditched enclosures are common round the Exe valley, and although their date has not yet been determined, the type is likely to originate in the Early Iron Age. The absence of 1st cent. finds from the site favours this interpretation. (6) Excavation by H Miles in 1971 in advance of building in the area between the inner and outer ditches revealed no trace of any occupation. A trench cut through the outer bank confirmed that its outer face had been stone-revetted. (7) The presence of a rampart immediately within the outer ditch would be unusual in a Roman military structure and it may well be of prehistoric date. Excavation yielded only a coin of Carausius (286-93) and a third century potsherd, both unstratified; but the character of the earthwork so closely resembles the mid-first-century fortlets of Old Burrow and Martinhoe that Stoke Hill is best accepted as of comparable date. Alternatively a date in the late third century would suggest use in connection with naval defence of that period. (8)
A rapid examination of air photography (9a) shows the cropmarks and faint earthworks of this probable Roman signal station, which is similar to the probable signal station near Ide (SX 88 NE 13). The regularity of its construction reduces the likelihood of it being an Early Iron Age enclosure. (9) |