Yol. 65.] GLACIAL EROSION IN NORTH WALES. 333 



formations of the Snowdon district are relatively resistant, as has 

 already been stated. Second, that in many non-glaciated districts 

 of massive rocks, young water-streams may show numerous little 

 falls between short, roughly -graded reaches ; and that in such cases 

 one may fairly assume that the streams have discovered differences 

 of resistance which the eye and the hammer cannot detect. No one 

 makes a mystery of the work of young cascading water-streams 

 because such differences of resistance are not apparent ; and similarly 

 in the case of young glaciers, it is the most natural thing in the 

 world that, if they erode at all, they should discover differences in 

 the resistance of the bed-rocks, which may not be immediately 

 apparent to us. The differences may be due to greater or less 

 resistance to crushing or scouring, to greater or less frequency of 

 .jointing ; to the attitude in which planes of bedding, schistosity 

 and jointing stand in relation to the ice- motion. Moreover, it must 

 be borne in mind that, in the work of a glacier as in that of a 

 stream, the element of time enters largely. Abundant opportunity 

 is afforded to the eroding agent for the discovery of every element 

 of weakness possessed by the underlying rocks. In view of all this, 

 the surprising thing would be to find an immaturely glaciated 

 district in which no rock-steps broke the slope of the valley-floors. 

 ,Bnt, as in the case of cwms, just how the ice works to produce rock- 

 steps is not yet made clear. It appears reasonable to me to accept 

 the possibility of plucking as an important part of the process ; but 

 for the present no full explanation of it can be given. 



A modification of the theory of ice-erosion ascribes a large part 

 'of the erosion of trough-like glaciated valleys to the action of sub- 

 glacial streams. I propose to consider this theory in detail else- 

 where, and take occasion here only to point out that it is difficult 

 to believe that an agency, which did not cut down the rock-steps in 

 the valley- floors and at the mouths of the lateral hanging valleys, 

 •could greatly deepen the valleys themselves. 



Yalley-sides. — Many of the valleys in the Snowdon district 

 have, when considered in cross-section, what is commonly called a 

 U-shaped cross profile ; but the term U-shape generally overdoes 

 the matter, for the valley-sides are not vertical. Catenary seems 

 a better term with which to describe the cross-profile of the glaciated 

 valleys in the Snowdon district ; but it is not intended thereby to 

 imply that the cross-profiles have exactly a catenary curve. When 

 considered longitudinally, the valleys around Snowdon are frequently 

 rather even-sided or trough-like for significant distances, without 

 advancing spurs or receding embayments ; but sometimes the valley- 

 sides are of a peculiarly uneven, ragged, craggy form, pleasingly 

 picturesque to the eye, yet perplexingly difficult to describe in 

 words ; they are occasionally steep and rugged. These several 

 forms will be considered in succession. 



The smooth-sided trough of catenary cross-profile is well shown 

 in the Gwynant Valley, on either side of Llyn Gwynant and farther 



