it is impossible to say which was the original.
Dr. Wainwright opened up four 15 foot squares along the South and West sides of the Rath, on the assumption that any buildings would shelter here under the ramparts, protected from the prevailing winds. He found two distinct periods, Roman on top, and Iron age underneath, separated by some 10 inches of fallow earth. In each of the boxes he found post-holes— most of them huge, up to three feet in diameter and three feet deep. These tended to be in rows of three, radiating out from the rampart. This rather suggested that they might form the supports for some sort of gallery, running round the rath (compare Dun Ardtreck and the Scottish Brochs, with their galleries; could there be a possible relationship? It is this intriguing possibility that makes the further excavation so important).
All over this Iron age level there are traces of metal-working, mostly bronze, but with some traces of iron slag. Pieces of the actual crucibles for the bronze-working were found, one of which can probably be reconstructed; it was circular, about six inches in diameter, and six inches deep. It also had a lid, which is an Irish feature—only one other has been found in Wales. There was also some pottery, which Dr. Wainwright says is similar to the pottery he found in the promontory-fort of Coygan which he excavated in 1963-5, and shortly to be published as a monograph by the Cambrian Archaeological Association.
This Iron age occupation was followed by a considerable period of desertion—long enough to allow
Top : Dr . Wainrigh t surveys some o f th e post-holes
Below: This shows th e height o f th e surviving bank round th e Rath
Note the gentle slope of the hill, the typical situation of the Pembroke Raths
10 inches of soil to accumulate. When it was re-occupied during the Roman period all traces of the previous structures must have perished, for some traces of stone walls were found, in some places overlying the post-holes. These stone walls will be traced in future excavations. The bulk of the pottery was of the late third and fourth centuries, though some might have been late second. There were also one or two sherds which Dr. Wainwright intends submitting to Leslie Alcock to see if they are of his sub-Roman types; subRoman occupation must, therefore, remain for the time being one of the many problems to be solved in future years.