Vol. 65.] GLA.GIAL EROSION IN NORTH WALES. 329 



Valley-floors: lakes. — The lakes which occur abundantly 

 ■about Snowdon are all of moderate or small area and depth. They 

 are well treated in a paper by Dr. Jehu (1902), from which it appears 

 that Llyn Llydaw, in the upper part of Cwm Dyli, where a sounding 

 of 190 feet was made, is one of the deepest. This lake, occupying 

 a high-standing basin in the floor of the great eastern cwm of 

 Snowdon, has, like many of the other small lakes hereabouts, every 

 appearance of being contained in a true rock-basin ; if any drift- 

 barrier does occur at the outlet, its depth cannot be nearly as much 

 as 190 feet. Llyn Cwellyn is a valley-floor lake, west of Snowdon, 

 with a depth of 122 feet. Its lower end is half a mile above a rather 

 pronounced rock -step in the floor of its broadly rounded valley ; it 

 is against all reasonable possibility that a drift-filled gorge, 122 feet 

 deep, should exist in this barrier down which Afon Gwyrfai cascades ; 

 hence Cwellyn also must occupy a rock-basin. The two lakes of the 

 Gwynant Valley have the appearance of being rock-basins : they 

 gain interest, because of association with picturesque rocky hills by 

 which the valley-floor is interrupted ; and these, I take it, are so 

 many examples of unfinished channel-bed smoothing. Where the 

 valley widens, east of Llyn Dinas, the hummockyform of the bed of the 

 great glaciers that there scoured a wide channel is very pronounced 

 and characteristic. Some of the smallest lakes in the cwm-floors 

 are, on the other hand, pretty surely enclosed by morainic barriers, 

 at least in good part. 



At Beddgelert the valley of Afon Glaslyn is broad and flat-floored : 

 about a mile to the south, the valley is narrowed to a rock-walled 

 gorge, the Pass of Aberglaslyn, where the stream has a rapid fall : 

 it would seem as if the open valley by Beddgelert were of rock-basin 

 form, and that it may well have held a lake for a time ; but the 

 lake is now destroyed by alluvial aggradation of the basin, and by 

 erosion of the outlet-gorge. Similar examples are well known in 

 the Alps. 



It is evident that the rock-basin lakes cannot be the result of 

 pre-Glacial erosion, for normal erosion does not produce such rock- 

 basins. Under the theory of ice-protection some special process 

 must be invoked to account for the basins. Drift-barriers, 

 frequently appealed to, are as a rule not admissible here. Warping, 

 still often accepted as a means of accounting for the great piedmont 

 lakes of the Alps, is not admissible there, for the reasons that 

 Wallace clearly set forth some years ago. As to a recent warping 

 of the Snowdon district, it would be difficult to prove or disprove. 

 If it occurred, it might well account for various rock-basins, but it 

 ought to have other recognizable consequences also : for example, a 

 warping that would produce Llydaw or Cwellyn ought to deform, 

 or crush, or rend some of the neighbouring ridges ; but no signs of 

 such disturbance are noted in relation to the rock-basins. 



Under the theory of ice-erosion, rock-basin lakes are not to be 

 regarded, as they too often have been, as marking the full measure 

 of erosion, but as indicating simply the excess of ice-erosion in the 



