Iron Age defended promontory; medieval watch station. The site is protected by an earthen bank and ditch across the landward side, and on the north side by precipitous cliffs on the north and east. The bank is 7m wide with a maximum inner height of 1.5m and outer height of 2.2m. A ditch to the landward of this is 6m wide on average and 1m deep, though it has suffered plough damage.


Excavation by PS Gelling in 1957 showed that the rampart was faced with dry walling inside and out and was 3.8m wide. The excavator believed that both faces were part of a reconstruction, for they stood on a layer of slatey rubble and building stone, which was thought to be Iron Age in origin. The rebuilding of the rampart probably took place before the construction of the rectangular longhouse within the enclosed promontory. Prior to investigation, there were surface indications of the building, but excavation proved it to be orientated NW-SE, with bowed walls and internal measurements of 7.5m by 4m. There was a possible entrance 2m wide in the south-east corner, but no trace of stonework. Apart from a few postholes for support for the roof and entrance, little else survived to indicate the construction of the walls, which seem to have been of earth and turf. There was a rudimentary raised bench along each long wall and at the west gable, formed by cutting down into the bedrock in the centre of the building; here there was also a hearth.


The excavator believed that the house represented Medieval reoccupation of the site. More recently it has been argued that the building was used as a watch-house as part of a system of watch and ward, and did not serve a domestic purpose.