Trans Dumfriesshire Galloway Natur Hist Antiq Soc 53

NY 185785. Pollen analysis showed Neo interference with the vegetation, but there is no evidence of permanent Neo occupation. A single palisade could be as early as LBA. The main rampart followed, enclosing 7ha in 6th or 7th century BC, originally timber revetted at front and rear; a stone face was added some distance in front. Two gateways were re-examined, and four superimposed houses were revealed, with no occupation necessarily later than first half of 2nd century AD. The Antonine fortlet on the S slope presumably precluded civilian occupation. The original ramparts were already degraded when turned over to Roman target practice, and the old theory of a siege at 155-8 AD is abandoned, especially as the alleged 'circumvallation' is unconvincing. Limited excavation of the fortlet and south Roman camp also took place and pottery, missiles, a sword etc are catalogued.

A fortuitous section through the motte at Roberton (NS 940270) gave evidence for the mound make-up and summit structures of timber; other mottes in the valley were surveyed. The distribution of five certain and two possible mottes within eleven original parishes points to the documented Flemish settlement of the 1150s and '60s.