1878.] 335 [Niles. 



the ice more rapidly than the rock, but it must be remembered that 

 the ice is constantly renewed by the motion of the glacier. It 

 will be readily seen that such streams, by the peculiarities of their 

 situation and action must exercise an influence in determining the 

 precipitous character which the sides of glacial valleys so often have. 

 If it is objected that such water-worn surfaces are rarely met with 

 upon the sides of valleys from which the glaciers have retreated, it 

 must be remembered that the ice above the streams and the atmos- 

 pheric agencies modify these surfaces after they have been left by the 

 streams, hence the rocks have the features which they received from 

 the last agent which acted upon them. 



Still lower and quite underneath the side of the glacier there are 

 larger and often much longer lateral streams, which are much more 

 important agents in the excavation and formation of the valleys. 

 These, flowing in channels of their own formation in the rock and 

 quite below the ice, tend to deepen the valley along its edges and to 

 give it that canon-like form so often seen. 



Sometimes the aqueous erosion under the sides of a glacier is 

 greater than it is under the medial portion, and when this has been 

 continued for long periods the edges of the valley have become the 

 deepest portions, and when the lower end of the glacier has receded 

 to this part of the valley it is often bifurcated, its terminations being 

 upon opposite sides of the rocky eminence left in the central part of 

 the valley. Such knolls or hills occur in the valleys of ancient gla- 

 ciers, as, for example, in the valley of the Rhone at Sion, and they 

 have always been a puzzle to the advocates of a purely glacial origin 

 of such valleys. If, however, we duly recognize the power of sub- 

 glacial streams, the hills which are sometimes left in positions where 

 they have been fully exposed to the action of glaciers appear as 

 a normal and not as an anomalous result of the agencies which 

 have excavated such valleys. 



With many other glaciers and often with other parts of the same 

 glacier, the medial stream is the most important one in volume and 

 power, and then it tends to make that part of the valley the deepest, 

 and the glacier assumes a corresponding form. 



In conclusion I will state that the observations of three summers 

 among the glaciers of the Alps have led me to estimale the relative 

 agency of glaciers and sub-glacial streams in the erosion of valleys 

 as follows: viz., that the sub-glacial streams are of primary impor- 

 tance in working in advance of the ice in deepening and enlarging 



PROCEEDINGS B. S. N. H. — VOL. XIX. 22 JULY, 1878. 



