334 



PROF. "W. MOEEIS DAVIS ON 



[Aug. 1909, 



up stream ; and again in the valley of Afon Gwyrfai on each side 

 of Llyn Cwellyn. In hoth these localities little streams from short, 

 valleys in the uplands rush down the even valley-side slopes with a. 

 very slight entrenchment ; yet they are already actively at work in 

 the establishment of a better relation with the main stream. 



Under the theory of ice-protection, the smooth troughs are 

 the work of normal erosion during a part of the Glacial Period 

 when the ends of the local glaciers stood at the top of the next 

 up-stream rock-step ; but it may be asserted with much confidence 

 that the trough-like valleys around Snowdon could not possibly 

 have been produced under such conditions. A trough-valley with 

 a catenary cross-profile may doubtless be sometimes produced by 

 normal erosion in non-glaciated regions, provided appropriate 

 structures compel the form ; but, in the Snowdon district, trough- 

 valleys occur in various stratified and massive structures of various 

 attitudes, which are not adapted to their production by normal 

 erosion. The usual features of a normal, nearly mature valley in 

 which the side-walls still retain steep or eliffed upper slopes (AB, 

 fig. 29) are the thin sheet of retreating talus (BC) beneath the cliff,. 



and the shallow alluvium of 



Fig. 29. — Cross-profiles of a normal 

 mature valley and of a glaciated 

 valley. 



the flood-plain (CD) beneath 

 the talus. A glaciated trough- 

 valley, in which little post- 

 Glacial waste has gathered,, 

 may have steep upper slopes- 

 (A'B') below which some 

 growing talus (B'C) may 

 already have accumulated ; 

 but this may not suffice 

 completely to conceal the 

 curving rock - slope (CD'} 

 which reaches to the stream. 

 The flood-plain element is. 

 here not essential, although 

 often in process of formation 

 by post-Glacial aggradation. 

 Furthermore, if the trough-valleys of the Snowdon district had 

 been widened by normal erosion, the side-streams would have- 

 eroded side-valleys of much greater depth than the trifling little 

 5- or 50-foot gulleys of to-day, and thus the trough-like continuity 

 of the valley-sides would have been interrupted. 



Under the theory of ice-erosion, a trough -like channel is a 

 natural expectation ; for the heavy and sluggish ice-stream would 

 tend to deepen and reshape any other kind of channel until 

 its cross-profile assumed a catenary curve. When thus maturely 

 reshaped, the channel — which we call a ' valley ' te-day, in the 

 absence of the glacier that once occupied it — would have rather 

 smooth sides, because the ice would have trimmed away all salients 

 and spurs ; and because the presence of the ice would have 



