| Summary: | Possible villa site set in the slopes of a combe overlooking the mouth of the River Otter. Bulk of fieldwalking finds come from an area 1000sq m north of the coastal path. A single piece of tile was found on the path in 1988. The excavations of 1989 produced evidence of at least two buildings, an oven and evidence of land management. The west end of a building (corner, referred to above) 9.1m wide was excavated. This was built largely of coursed local calcrete rubble, sometimes set in herring-bone pattern, which stood 1.1m tall and 0.75m wide upon a 1m wide foundation of unmortared stone in a foundation trench. The north west corner of the building had dressed sandstone quoins. Eavesdrip drain gullies ran outside the wall. The interior had a simple beaten earth floor with animal bone, sea shell, slate and pottery trodden into it. Some time after its construction the south wall had been faced with gravelly mortar. Abandonment is indicated by an accumulation of sand and slate fragments followed by a layer of collapsed roof slates and then mounds of wall rubble. Mounds of similar rubble were found c17m to the west at a depth of 2m, which may indicate that the building was over 9m long or that there was a further building nearby. The simple floor and poor quality slate roof suggest that this building was an ancillary structure. 20m to the south a further building was identified at the foot of the scarp rising from the coast path to the cliff top. This was indicated by a mound of rubble containing unabraded pottery and tile fragments including roof tiles and pila tiles from an underfloor heating system. This building may have been lost to cliff erosion, or may have been terraced into the scarp, the rubble tumbling down to rest beside the coast path. This building was of higher status than the first. 20m to the west of this part of a shallow straight sided trench, 0.4m deep and over 0.77m wide, was excavated. A few stones in the bottom suggest it was a robber trench. There was one piece of Roman tile in the fill. This may represent a small or narrow building set in a terrace which is no longer visible. Circa 5m north east of the heated building, the remains of an oven were excavated. This was 0.6m in diameter and cut into an outcrop of sandstone. There were signs of burning around its outer edge and flue and it was filled with a charcoal and loam deposit containing a Roman pot sherd. This area had been plough damaged. Drainage gullies had been dug along the upper slopes of the combe, in order to carry away surface water. These had infilled with dark, charcoal-rich loam with sea shells, animal bone and pot sherds which may indicate dumping of domestic refuse or the manuring of adjacent fields. One gully also produced tile fragments, iron nails and slag. Terraces, with drystone revetting walls, were also cut along the lower slopes of the combe, though much of the revetting had been robbed. These walls were out of use before the final collapse of the first stone building, as a 0.5m deposit of colluvium over the combe floor was overlain by collapsed roof material and was also cut by a pit 1.9m diam and 0.5m deep. The collapsed buildings were covered by up to 0.8m of colluvium, which produced a medieval pot sherd, suggesting that medieval or later cultivation of the slopes continued causing further hillslope erosion. The Roman pottery evidence suggests occupation concentrated on the 3rd century, but may have begun in the late 2nd century. It is possible that an earlier, timber, structure underlay the first stone building, but this was only represented by a compacted surface containing fragments of pot, slate and sea shells, but this may equally well have been an old ground surface. Elsewhere the deep hillwash deposits may overlie further structures. |
|---|