Alfred Harker — Stibnerial Erosion of Sktje. 



487 



distinctive characteristics of ice-drainage as contrasted with water- 

 drainage. Thus, a glacier occupies the whole wi.lth of its valley, 

 while a river and its tributaries mostly follow narrow and defined, 

 paths: hence the peculiar U-shape of the glaciated valley and the 

 precipitous descent by which a tributary glen joins the main valley 

 (Fig. 1). Again, the amphitheatre form of the head of a glaciated 

 valley connects itself with the fact that, while erosion near the 

 sources of a river is very feeble, a glacier, on the other hand, 

 springs into being, like Athene, fully armed. The dependence of 

 glacial erosion upon the pi-essure, and so upon the thickness of the 

 ice at any point, seems to offer an explanation of other characteristic 

 features, such as the exaggeration of slight inequalities of gradient 

 in a valley into precipitous drops (Fig. 2) and the excavation of 

 small rock-basins. 



riG.2 



Fig. 2. — Longitudinal profile of valley occupied by glacier. Owing to inequality of 

 gradient, the thickness of the ice at AB is greater than either above or below 

 that place. The consequent more rapid erosion at B has the effect of 

 exaggerating the iuequahty. 



As distinguished from the erosion due directly to ice action, we 

 have to note further the disintegration caused by frost in the later 

 stages of the Glacial period. Nothing in the scenery of the Guillins 

 is more striking than the sharp contrast of the splintered and jagged 

 summit-ridges with the rounded and polished surfaces of the corries 

 and the les^ser heights. The first impression is that the higher peaks 

 have stood out above the ice, and so escaped the moulding and 

 smoothing which have left their impress on the lower slopes ; but 

 in attempting to follow out this idea in detail many difficulties are 

 encountered. Moreover, the movements of the ice, as traced out in 

 the course of a complete survey, and the history of the glaciation of 

 the district as a whole, almost compel us to the conclusion that, at 

 the stage of maximum glaciation, the whole district was covered, 

 even to the highest summits. Later, when the ice was reduced to 

 gradually shrinking valley glaciers, the uncovered summit-ridges 

 would be subjected during a long time to conditions highly favourable 

 to the shattering and splintering of which they give evidence. In 

 no other way can we account for the huge talus, often composed of 

 large unworn blocks, which chokes the upper part of so many of the 

 mountain valleys, being specially well displayed in the southern 

 branch of Coireachau Euadha and along the whole extent (a mile 



