244 The Loch Urr Crannog. 



length, and partially submerged, communicating with the main 

 island ; (4) the main island, with ruins of outer wall and four 

 inner enclosures. 



We come now to speak of the earthwork. If the O.S. map 

 is referred to it will be seen that a point of land, bearing the 

 significant name "White Isle," juts out from the Dunscore side 

 of the loch and forms a peninsula some six acres or thereby in 

 extent. This portion of ground takes a somewhat conical 

 form, and it is upon the neck of land which connei^ts this knoll 

 with the mainland that the earthwork is found. 



It consists of a well-defined rampart and ditch, and appears 

 to have been a work of some importance. The length of the 

 rampart may be stated approximately at 500 feet. On the west 

 side, where the rampart is most entire, it rises above the present 

 ground level of the trench in front to a height of between eleven 

 and twelve feet. At the rear the height is about four feet. Two 

 sections were cut here ; one through the rampart itself ; another 

 through the ditch in front. The rampart was found to be com- 

 posed of a stiff white clay, mixed with rough gravel. This 

 overlies the original soil to a depth of six feet six inches, 

 measuring from the centre of the rampart. Xo stone larger than 

 coarse gravel was found here; but it may be mentioned that 

 stones show through the surface at two points on the face of the 

 rampart, and at one at least of these points the stones have 

 evidently been arranged so as to overlap one another. I observe 

 from the copy of the " Transactions ' ' lately placed in my hands 

 that Mr Barbour refers to a similar arrangement as occurring at 

 Birrenswark. In the rear of the rampart a rough stone pavement 

 ten feet in width was found. As this pavement lies near the 

 surface, it has the effect of altering the character of the vege- 

 tation, and the pavement can be traced in this way along the 

 greater part of the line of rampart. It should be mentioned 

 that a wide gap of between eighty and ninety feet occurs almost 

 in the centre of the earthwork. It seems probable that this 

 portion has been broken down, but at present I have no proof 

 to offer in support of this suggestion. The trench in front of 

 the rampart was found to be silted up to a depth of five feet 

 ten inches. This fact is important, for it proves, almost con- 

 clusively, that the point of land which we now know as a 

 peninsula was at one time an island. The name "White Isle," 



