ICE-EROSION IN THE CUILLIN HILLS, SKYE. 247 



within the area occupied by the smooth ; they do not extend so far from the mountains, 

 and on the other hand they do not enter the mountain-tract itself except along the 

 floors of the wide and level interior valleys. This type of drift is indeed confined to 

 open places, usually near the mountains, in parts where the slope of the surface is 

 gentle, and it is found in such places especially where the flow of the ice has been 

 obstructed or checked by extraneous interference. The finest development is in the 

 Red Hills tract, and more particularly at such places as Luib and Strollamus, where 

 considerable branches of the native ice abutted directly upon the flank of the Scottish 

 ice-sheet. In the part of the island which we have chiefly considered, the largest area 

 of hummocky drift is that near Sligachan, extending from Harta Corrie to near 

 Garadubh, on the Portree road, a distance of ten miles. Arms of this extend up the 

 valley of Allt Dearg Mbr and over into the Drynoch valley for distances of 1 mile and 

 1^ miles respectively, and there is a detached area in Coire Keidh na Loch. Patches 

 occur again in Strath na Creitheach and the open part of Coire Riabhach, between 

 Druim nan Ramh and Druim an Eidhne, and others on the west side of the Cuillins, 

 viz., south of Coir' a' Ghrunnda and between Allt Coire La bain and Allt Coire na 

 Banachdich. The most remote isolated occurrence observed is a small patch at the 

 head of the Talisker valley. 



The phenomena in the Skye district, viewed generally, seem to find their simplest 

 explanation on the supposition that the waxing and waning of glacial conditions have 

 been controlled less by secular changes in the mean temperature than by variations in. 

 the amount of precipitation over the area. This question could not be profitably 

 discussed as a local one. But whatever the causes which terminated the maximum 

 glaciation of central Skye, it cannot be doubted that the ice -cap shrank away from the 

 mountains as well as from its margin. As the steep summit-ridges emerged, there 

 was added a new element of ice-transport by the accumulation of material upon the 

 upper surface of the ice ; an element increasing in relative importance as the erosive 

 effect at the lower surface became feebler. We interpret the hummocky drift as 

 the materia] — superglacial, englacial, and infraglacial — finally deposited by stranded 

 portions of the confluent glaciers, cut off from their supply behind and melting as they 

 stood. All the features of its composition, no less than its distribution, seem to accord 

 best with this view. The absence of all ordinary morainic accumulations referable to 

 this epoch suggests that the disappearance of the ice was a somewhat rapid event. It 

 is clear, from the dispersal of boulders, that no change in the direction of the ice- 

 drainage took place from the maximum glaciation to the close of this phase, such lines 

 as that already noticed marking the mid-stream of the Sligachan branch being traceable 

 uninterruptedly through the smooth and the hummocky drift alike. 



