1. A cliff-top enclosure, defined by a series of three banks and ditches to the north and north-east; traces of a 1.0m wide rubble wall have been noted on the crest of the inner bank, thought also to have been revetted internally; there is a causewayed entrance at the south-eastern end of the defences: the internal area is at present only 40m by 62m.
(source Os495card; SM70NE1)

2. Watery Bay (also called Marloes Sound Rath) does not lie on a distinct promontory, but occupies a plateau location on the edge of a high cliff. It is a multivallate site, with the arrangement of the ramparts indicating two-phase construction, although the sequence of construction is not clear. The inner rampart is sufficiently distinct from the outer ramparts to indicate a separate phase of construction. It is curved, extending for c. 85m and stands 3m above the interior of the fort and 5m above the ditch external to it. It now terminates sharply at the cliff edge at its western end and at the fort's entrance at its eastern end. The entrance is a simple gap, and the bank continues to the south of it, albeit in a much reduced form, c. 0.7m high, along the edge of the cliff. The outer ramparts are more complex and consist of two main lines of bank and ditch, with a hint of a third line between them. They are not as massive as the inner rampart, but nevertheless they are substantial with the outer bank rising to over 3m above the external ditch. Although these outer ramparts have a unified appearance, the plan shows that the lines of bank and ditch diverge towards the west, perhaps indicating that these, too, are of more than one phase. A small stream runs immediately to the east of the entrance. On the other, east, side of the stream a large bank, c.30m long and up to 2m high with traces of a mostly in-filled ditch on its eastern side provides additional protection for the entrance. There is a hint that this bank continues to the north, but this may be a natural feature augmented by stone clearance. The flat, featureless, internal area is roughly square and measures c. 60m by 60m. As noted above a low bank along the cliff edge defines the western side of the internal area. There is a similar, but much lower, bank along part of the eastern side of the site, again along the cliff edge. If these two banks were part of the defences then it seems likely that only the southwestern and southern sides of the interior have been lost to the sea, if, indeed, any has been lost at all. It is unknown how much of the defences (if any) have been lost at their western end. Analyses of historic remote sensing data (maps and aerial photographs) did not provide information on loss. The main threat to the site seems to be from visitor erosion, in particular there is scarring where the Pembrokeshire Coast Path crosses the defences on the western side of the fort.

Reference:
Page, M., Barker, L., Driver, T. and Murphy, K. 2008. Remote sensing and the Iron Age coastal promontory forts of Pembrokeshire, Archaeology in Wales 48, 27-38.

3. Field visit by T. Driver and L. Barker on 22nd Oct 2009 confirmed survival of an earthwork bank along the inner edge of the cliff slope suggesting that we are seeing the original form and limits of the interior; coastal erosion appears slowed or negligible due to the lie of the bedding planes of rock. The stream on the south-east side of the fort also appears to have been enclosed within the defences of the fort with the addition of a free-standing length of bank and ditch beyond the stream to the south-east.
T. Driver & L. Barker, RCAHMW, 16th November 2009.

T. Driver, RCAHMW