Page:Folk-lore - A Quarterly Review. Volume 4, 1893.djvu/248

 240 no images, no funeral urns were found—nothing to enable us to

that occupied the minds and looked through the eyes wherewith they surveyed the dark wet forest and upland clearing around their rude homes of wattle and clay in those far-off times.

General Pitt-Rivers' new volume is marked by the same admirable characteristics as the previous ones. As before, we are impressed with his minute accuracy, his anxiety to lay before the reader all the facts, independently of any theory, so as to put him in a position to judge for himself on the questions disputed, his careful reasoning, and his wide anthropological learning. The volume is chiefly concerned with explorations of Bokerly Dyke (a rampart about four miles long, which yet throughout the greater part of its length forms the boundary between Dorsetshire and Wiltshire, and runs in a south-easterly direction), and of Wansdyke at places not very far from Silbury. Both these ramparts have been thrown up for purposes of defence against the north and north-east. The frontier defended by Wansdyke seems to have run along the valley of the Avon to a point above Bath, where the dyke crosses the river and appears to join the Roman road from Bath to Marlborough, running continuously with the latter until it reaches the valley of the Kennet. Before the road enters the valley the dyke parts company with it, and continues along the heights through Savernake Forest to the borders of Berkshire, where it turns to the south and is lost. In point of construction,, both Bokerly and Wansdyke are similar, consisting of a ditch with a small external mound and a higher rampart within. They are not of uniform height; and in