*238 Ml! ALFRED HAEKEB ON 



is the stable form, is for ice-erosion unstable, since any departure from it leads to a 

 further departure. There must, of course, be a limit to the action described, viz., when 

 the lower layers of the ice begin to be ponded in the lee of a strong feature, and the 

 upper layers slide over them. 



These remarks receive striking illustration in the Cuillin Hills. Excluding the 

 three main interior valleys, which in no part reach more than a very moderate altitude, 

 all the glens show a very remarkable configuration. The longitudinal profile consists 

 of two or three stretches of moderate slope divided by relatively steep drops, over 

 which the water cascades. Where two such drops occur, as in all the longer valleys, the 

 upper one is both higher and steeper than the lower. The heads of some of the glens, 

 such as Coire na Banachdich, Coir' a Ghrunnda, and Coire nan Laogh, are almost 

 inaccessible from below ; and to an observer coming into the district for the first time 

 this stepped or storeyed form of the valleys is one of its most conspicuous peculiarities. 

 The more or less precipitous drops, which constitute the steps and separate the 

 successive storeys, are in no instance connected with anything in the geological 

 structure of the ground, and there is no correspondence as regards levels between 

 even adjacent valleys. In typical examples the descent is usually 200 to 400 feet, 

 with an average gradient which varies in different cases from 30 to about 70 vertical 

 in 100 horizontal. The best example of a valley divided into three parts by two steep 

 drops is the one made up of Lota Corrie and the upper part of Harta Corrie.* Coir' a' 

 Ghrunnda (fig. 6) illustrates a different case, the upper part of the Corrie being cut off 

 by two sharp drops close together, giving a total fall of about 850 feet in a horizontal 

 distance of 2100 feet. 



Apart from superficial accumulations which may complicate the conditions — a case 

 with which we are not concerned in the Cuillins — a negative or reversed gradient in 

 the ' Thalweg ' implies, of course, a rock-basin, and is, as Ramsay long ago pointed out, 

 a result which cannot be arrived at by aqueous erosion. There is, in the opinion of 

 the present writer, some danger of attaching too much weight to rock-basins as 

 phenomena indicative of ice-action, with the result of diverting attention from other 

 phenomena equally characteristic and often of a larger order. That the floor of a 

 valley is lower at a certain place than at another place farther down-stream is of 

 interest because it is an absolute criterion, not one of degree ; but to dwell unduly 

 upon this is to treat the rock-basin as an isolated phenomenon instead of what it is, 

 viz., an integral part of the valley. It is an incident, depending not merely on glacial 

 erosion, but on glacial erosion operating under certain local conditions. The requisite 

 conditions may be realised in more than one way ; and in a classification of the 

 characteristic forms of ice-erosion according to their origin and essential significance 

 different rock-basins would fall under different heads. 



Of the three main interior valleys of the Cuillins, two, viz., those of Coruisk and 

 Camasunary (Strath na Creitheach), contain elongated rock-basins. The determining 



* See section, Oeol. May. 1 899, p. 1 97. fig. 2. 



