places both are lost, without any reason to suppose that the loss is due to effacement by agricultural operations. General Pitt-Rivers conjectures that these places may have been formerly occupied by forest, and that it was there easier to make an abattis of felled trees.
In all cases the excavations were continued down to the undisturbed chalk beneath the ditch and the mounds. They revealed, both in the rampart and on the old surface under it, pieces of Samian ware, cleats, and other objects of iron, and, in the case of Bokerly Dyke, Roman coins, which proved that both dykes were erected during, or subsequent to, Roman times. In what circumstances, or during what war, however, the dykes were built is still undetermined. The object evidently was the defence of the south-western corner of the island from enemies coming from the north and east. But who were the enemies, or who the defenders, is a problem that further researches have yet to make manifest.
But, however interesting the problems connected with the dykes may be, the student of folk-lore will naturally turn rather to the village at Woodyates. This is the third ancient village discovered in the course of the author's excavations. It will be remembered that the race who had occupied the villages described in the former volumes averaged, the men 5 feet 2.6 inches, and the women 4 feet 10.9 inches in height. The village called Woodyates, from the name of a modern cluster of buildings a short distance to the south-west of the site, was occupied by a people answering to a similar description. Bokerly Dyke runs through it at the point where the dyke crosses the Roman road from Badbury Rings to Old Sarum. The portion of the settlement examined is chiefly on the outside of the dyke; and how much of it was inside, or how much more outside, is yet unknown. The village, as indicated by the turn just here of the Roman road, appears to have been in existence before the road was made; but some of the drains bear evidence of having been cut