It is this geological accident that gives the site its character. Whoever occupied this promontory fort did not need to excavate a deep trench or pile up an earthwork to guard their landward flank; the sea had already done the most dramatic part of the job, carving a gap through the rock that would have made any approach awkward and exposed.

The fort was recorded in detail by the antiquarian T. J. Westropp in 1921. Writing at a time when the defences were more legible than they are today, he described a fosse, or defensive ditch, on the landward side of the arch, followed by dry stone walling with foundations roughly twelve feet thick. Inside that lay a square terraced area of about eighty feet to each side, itself enclosed by a further wall around six feet thick. When surveyors revisited the site, the broad outline Westropp described was still recognisable: a stone wall, now around 0.8 metres wide, still follows the line he recorded, and the interior terraced platform survives much as he saw it. What has gone, or at least is no longer visible, is the fosse on the outer side. On the north-east and north-west edges of the interior, only faint traces of an earthen bank remain. The ground within the enclosure drops steeply northward toward the bay.