Vol. 65.] 



GLACIAL EROSION IN NORTH WALES. 



313 



independent glaciers of such neighbouring lateral valleys to end in 

 a line of almost uniform altitude, on which the mouths of a set of 

 hanging valleys must, according to this theory, stand. Further- 

 more (*), the climate of the Glacial Period as a whole must, under 

 this theory, be peculiar in advancing through several long main- 

 tained pauses to its climax, and then in moderating either with 

 corresponding pauses or without significant pause to its close. 

 (j) When the ice finally disappears, the valley-heads will be 

 revealed, essentially unchanged from their pre-Glacial form. But 

 certainly the most peculiar feature of the theory of glacial pro- 

 tection is (1c) its fundamental postulate that glaciers do not erode, 

 in view of the erosive action easily observed, even though slow, under 

 the sides and ends of mountain-glaciers all the world over. 



Nevertheless, if, in the face of these special and peculiar require- 

 ments, the theory that glaciers are protective should lead to 

 consequences which correspond to observed facts, while the theory 

 that glaciers are destructive, even if apparently simpler in require- 

 ments, should lead to consequences that are contradicted by 

 observed facts, we should necessarily discard the second theory 

 as inadmissible, however plausible it might have seemed at first * 

 and we should have to accept the first theory as to all appearance 

 competent, even though some of its requirements were such as to 

 make it at the outset seem improbable of realization. 



XX. Glaciers as Destructive Agencies. 



The general discussion of glaciers as destructive agencies need 

 not include a consideration of special stages of glacial advance, 



but may pass at once 

 Fig. 14.— Diagram of a large eroding glacier to the case of a well- 

 in the valleys of a subdued mountain. developed glacier, as 



indicated in fig. 14. 

 Erosion is assumed to 

 take place under all 

 parts of the glacier 

 here shown, but the 

 erosion may well be a 

 slow process ; the only 

 requirement is that it 

 should be significantly 

 more rapid than nor- 

 mal erosion. It may 

 here be pointed out 

 that the theories of 

 erosion under the lead 

 of water-streams and 

 under the lead of ice- 

 streams involve certain resemblances and certain contrasts. As 

 to resemblances, in each case the established stream occupies a 

 channel at the bottom of its valley, and the banks of the channel 



