ICE-EROSION IN THE CUILLIN HILLS, SKYE. 249 



ground mainly by a vast number of erratics scattered over the tracks of the glaciers 

 and lying thickly along certain lines and in certain places. In attempting to map 

 these out accurately, considerable difficulty might be experienced in some parts in 

 separating the later from the earlier deposits without suspicion of possible error. 

 Where, however, characteristic rocks are available for tracing out the lines of move- 

 ment no doubt can exist. For instance, as we emerge from Glen Sligachan, we find 

 that the surface erratics are of granite to the right and of gabbro to the left, with only 

 a narrow belt of intermingling, and we can draw a line accordingly, as already done 

 for the earlier drift. We find that, instead of curving away westward, this line runs 

 straight on towards Portree, only .a little on the east side of the highroad ; showing 

 that these erratics were brought down at a time when the Scottish ice-sheet had 

 withdrawn, or at least had ceased to press heavily upon the coast of Skye. 



To the west of this line and of the mountains, gabbro is constantly the principal 

 element in these accumulations, and remains so to all distances. There are, however, 

 basaltic lavas, from the patches enclosed in the gabbro mass and from the lower slopes, 

 and representatives of the doleritic and other rocks of the minor intrusions of the 

 Cuillins ; besides peridotites and other locally distributed types, which are found along 

 lines leading from their several places. The blocks are not planed or scored. Many 

 of them are of considerable size, and some very large ones are found in the lower part 

 of Coire Labain, in Coire na Creiche, near the mouth of Harta Corrie, and elsewhere. 

 The largest are usually of gabbro, picrite, and other massive rocks ; the more jointed 

 rocks of the dykes and sheets have broken into smaller blocks ; the laminated rhyolite, 

 with a slate-like fracture, is usually represented by small fragments. 



Only in a few places do these later glacial accumulations assume anything like 

 the form of typical moraines. There is, however, one remarkable exception, which is 

 of sufficient interest to demand notice. It occurs opposite the mouth of Coir' a' 

 Ghrunnda, and was noticed by J. D. Forbes in his paper already mentioned.* Here 

 we have a perfect crescentic moraine, measuring 900 yards from horn to horn and 800 

 yards from that line to the front. The material consists in great part of large blocks, 

 chiefly of gabbro, but including also picrite and the other rocks of Coir' a' Ghrunnda. 

 The front portion is a curved ridge 150 yards across and 50 feet in height ; but 

 towards the two horns the height diminishes and the width increases, until the ridge 

 is represented only by a belt of closely scattered blocks. The moraine lies partly upon 

 a patch of hummocky drift, which has preserved much of its characteristic surface 

 relief. 



To the later glaciation must be attributed the perched blocks which are conspicuous 

 objects on the bare slopes of some of the Cuillin valleys, as noticed by Sir A. Geikie. 

 Good examples are seen on the western side of Sgurr na Stri, towards Lochs Coruisk 

 and Scavaig, and on the lower slopes of Sgurr a' Coir' an Lochain. In the latter place 



* On Forbes' small sketch-map Coir' a' Ghrunnda appears as "Bottomless Corry." The moraine is marked 

 with the letter E on our map ; see also fig. 6. 



