R. M. Deeley — Erosion hy Rivers and Glaciers. 393 



bottoms so deeply cut into that they are frequently occupied by 

 a succession of lakes, many of which rest in true rock basins. 

 Attention has recently been called by Dr. Wallace to a number of 

 the peculiarities of these valleys and lakes. Indeed, upon the 

 question of the origin of lake basins the discussion of the erosive 

 power of ice has largely turned. 



The opponents of Ramsay's theory of glacier erosion base their 

 opposition almost wholly upon theoretical considerations of a more 

 or less unsatisfactory kind, holding that ice can only polish and 

 grind, and could not have shaped the valleys and excavated the lake 

 basins in the time we can reasonably expect the glaciers to have 

 been active. Even here, however, their notions of chronology may 

 be woefully at fault. But there are those also who hold that the 

 rate of erosion at and beyond the terminal portion of a glacier is 

 greater than beneath it, and that, consequently, a lake basin can 

 never be formed by erosive action beneath the ice. But this is 

 quite contrary to the teaching of the phenomena to be seen in such 

 positions. 



We have pointed out that the handiwork of glaciers in the 

 form of masses of Boulder-clay, gravel, and sand, is so extensive 

 and stretches from the mountains into the lowlands to such great 

 distances, that it is difficult, reasonably, to resist the conclusion that 

 the erosive power of glaciers is very great indeed. Seeing this, 

 some popular writers have attempted to revive the ancient theory 

 of vast floods and great earth convulsions. 



Dr. W^allace shows that small lateral valleys seldom, if ever, enter 

 the main valleys at such heights as to cause a lake to branch into 

 them. When such branching occurs it is due to the confluence 

 of two important valleys or to the flow of a glacier over 

 a neighbouring " col." But these features are by no means confined 

 to vallej's occupied by lakes. The valley of Chamouni may be well 

 instanced. Here, on one side, the valleys in which the Glacier des 

 Bossons, the Mer de Glace, and the Glacier d'Argentiere rest, 

 terminate abruptly on the valley-side at a great height above the 

 Arne, and the ice from them is precipitated in great cascades down 

 the slope. On the opposite side the upland valleys, although not 

 occupied by glaciers, present the same features, a steep climb 

 bringing one suddenly into an upland basin-shaped valley. The main 

 valley really appears to have been cut like a huge trench through 

 everything. 



In everything, however, that relates to ice and its work contradictory 

 phenomena present themselves. In the first place it is a brittle 

 substance, and yet when fragments of it are pressed together, at or 

 near the melting-point, they readily weld together. And, again, 

 0. Mugge ^ has proved that a crystal of ice is not at all viscous, i.e., 

 it will not deform even slowly under the strains produced by small 

 stresses, and is only plastic like moist clay in a direction at right 

 angles to the optic axis ; j^et glacier ice, which consists of countless 



^ " Ueber die Plasticitat der Eiskrystalle " : Jahrbuch fiir Minera'.ogiej 1895, 

 p. 211. 



