ICE-EROSION IN THE CUILLIN HILLS, SKYE. 231 



dykes, and sheets eroded down to a common level and figuring upon a single smooth 

 flat rock-surface. Throughout the interior of the mountain area indeed, all the 

 conspicuous features—precipices, ridges, barriers, basins, etc. — are carved out of the 

 rock-complex in a fashion wholly irrespective of lithological differences or geological 

 structure. Only on the outskirts of the mountains do we find exceptions, which thus 

 o-o to emphasize the rule. This is seen, for instance, on the southern face of Gars- 

 bheinn, and still more clearly in the form of the corries to the east of that mountain 

 (see Ordnance Map), where the juxtaposition of gabbro and basaltic lavas has given 

 rise to some bold escarpments. Here, where glacial erosion has played a less dominant 

 part in shaping the existing land-surface, geological structure has asserted itself in 

 the usual mauner. 



(vi.) Forms of the Valleys, and Relation of Tributaries to Principal Streams. 



We may now proceed to recall some of the more obvious differences between ice 

 and water which may be expected to aid us in discriminating the effects produced by 

 these two agents of erosion. Running water in a valley concentrates its direct action in 

 oreat part upon certain narrow channels, viz., the main stream and its tributaries. The 

 courses of the tributaries, and even of the main stream in different parts of its course, 

 make various angles with the general direction of the valley. If now we suppose the 

 valley to become occupied by ice moving down it, the conditions are greatly changed. 

 A glacier fills a large part of the width of the valley ; an ice-cap, as in the Cuillins at 

 the principal glaciation, more than fills the whole valley. Moreover, we vatxy, as a first 

 approximation to the truth, consider this body of ice as moving down-stream as a ivhole. 

 The differential movement which takes place within the mass imports a considerable 

 modification of this broad view, but does not destroy its validity for our argument : 

 though ice is not a ' rigid ' substance, it is rigid in comparison with water. 



This different manner in which the eroding force is applied must produce results 

 which can in part be foreseen. We must expect a tendency to simplification of the 

 form of the valley in ground-plan and in cross-section (the longitudinal profile falls 

 under other rules). Lateral erosion — unfettered here by any consideration of ' base 

 level' — will come into play to reduce or destroy projecting spurs, to straighten curved 

 reaches, to plane away the subsidiary ridges which separate adjacent minor tributaries, 

 etc. ; and the result of such action, if continued, will be to widen the floor of the 

 valley and to straighten and steepen its walls. 



The valleys of the Cuillins are straight in ground plan,* and this straightness 

 extends also in very great measure to the slopes which bound them. Transversely 

 they show a flattening of the floor and a steepness of the bounding walls, which give 



* Harta Corrie is the only exception, and here we have already seen that, at the maximum glaciation, the ice 

 when it reached the curve no longer followed the direction of the valley. 



