226 MR ALFRED HARKER ON 



basin. In following this course part of the lower layers of the ice would have to face 

 an upward gradient of about 1 in 4 for a distance of nearly three-quarters of a mile. 

 Such behaviour would be incomprehensible in a valley glacier, but it is intelligible on 

 the supposition that the Sligachan valley near Loch Dubh watershed was occupied by 

 ice extending higher than the summits of the Cuillins. 



Immediately outside the mountain area proper, a new and potent factor came 

 into play, viz., the pressure of the Scottish ice-sheet coming from the east. Hence we 

 find a sharp westward diversion of the great ice-streams from the Sligachan, Cama- 

 suuary, and Coruisk valleys and of the other branches, such as that from the east side 

 of Blath-bheinn. The interior ice-drainage in turn bore back that from the outer side 

 of the Cuillins, as alread}^ remarked, so that, with increasing distance from its source, 

 the movement of the native ice conformed more and more to that of the great Scottish 

 ice-sheet. In the Red Hills the relations must have been more peculiar, for there 

 some important valleys opened directly upon the flank of the invading Scottish ice. 

 It is probable that at the climax of glaciation no point in the Isle of Skye rose above 

 the ice. It is certain, at least, that for many miles from the Cuillins hill and valley 

 were alike buried, and the form of the ground exerted only a very partial control over 

 the direction of flow. The mid-stream line of the ice from Grlen Sligachan crossed three 

 watersheds in its nearly semicircular course from Sligachan to Loch Eynort ; while the 

 left wing of the same ice-stream found its way over the pass Bealach a' Mhaim, to the 

 north of Bruach na Frithe, part of the base of the stream rising for this purpose about 

 1000 feet in a distance of three miles. 



As already remarked, the direct evidence concerning the direction of flow of the 

 ice is applicable to the lower portion only. It is not improbable that the upper layers 

 followed in some places a somewhat different course. In the mountain area their 

 movement would be less directly influenced by the configuration of the land-surface, and 

 might more nearly realise the ideal radial outflow. Elsewhere the direction of move- 

 ment might cross a valley occupied by ice either stagnant or flowing down the valley. 

 There is reason to believe that something of this kind occurred at one time in the case 

 of Loch Sligachan. Since in discussing glacial erosion we are concerned with the basal 

 portion of the ice only, it is not necessary to pursue this question. 



The foregoing account concerns the principal glaciation of the area only. Here, as 

 in some other parts of Britain, there was a later and minor glaciation, taking the form, 

 not of an ice-cap, but of glaciers occupying the valleys. At this stage the obstruction 

 offered by the Scottish ice-sheet had been removed, and the Skye ice was free to follow 

 a course more in accordance with the local topography, as is partly shown by the second 1 

 set of arrows on the map. 



