54 J. E. Marr and R. H. Adie — The Lakes of Snoicdon. 



leaves the lake at the north end of the railway tunnel), and the 

 alluvium is just above lake-level. There is no stream of any im- 

 portance in this valley, and it is difficult to account for its existence 

 unless we suppose that the main river ran through it, and that it 

 became filled with drift, causing the formation of the lake, which 

 drained over what was formerly a low col, situated at the present 

 exit. Near the point where this old valley joins the present one, 

 are several large pools in the present valley surrounded by alluvium. 

 They are probably "kettle-holes" in the drift. The floor of the 

 Seiont Valley is occupied by drift all the way from the lake to the 

 sea, so that it is quite possible that the floor of the lake may be below 

 sea-level and yet that the lake may not be in a rock-basin. 



We now proceed to consider the two lakes of the Vale of Gwynant. 

 Each is about three-quarters of a mile in length, and has its longer 

 axis running in the direction of the valley. Llyn Gwynant is the 

 higher of the two, and the bottom of the valley is occupied by drift 

 between its foot and the head of Llyn-y-ddinas, At the head of 

 the latter lake, the river has partly cut its valley through this drift, 

 but without reaching the rock. The exit from this lake is apparently 

 over rock, though no actual rock is seen in the stream at the outlet, 

 but the width of the area devoid of rock is here very small. A bold 

 rocky eminence lies between the foot of the lake and the road ; the 

 road is carried along a drift-filled depression west of the present 

 exit for a short distance, and to the north of this the drift-filled 

 depression is seen north of the road, and joins the lake a few 

 hundred yards above the exit. This depression is about thirty yards 

 wide in its narrowest part. To the south-west it joins the drift- 

 covered bottom of the valley, and this valley-bottom is covered with 

 drift to some distance below Beddgelert. It may be noted that 

 Beddgelert stands on an alluvial flat which was once an old lake, 

 and a barrier of rock runs across the stream at the entrance to the 

 Pass of Aberglaslyn, but a drift-filled depression is seen to the west 

 of this, which joins the main valley a short way down the pass. 



The lakes in the Gwrfai Valley present points of some interest. 

 Llyn Cwellyn is 464 feet above sea-level. An alluvial flat runs for 

 half a mile from the foot of the lake, and no doubt marks a former 

 portion of the lake filled up by the sediment brought down by the 

 small streams from Moel Eilio and Mynydd Mawr. The water then 

 runs over rock forming the cascade at Nant Mill, and the head of 

 this cascade is only about five feet below the level of Cwellyn. At 

 this point a barrier of rock extends right across the valley in such 

 a way as to forbid the existence of any drift-filled depression, which 

 could account for the lake. We lay special stress upon this point, 

 for it might be urged that a possible drift-filled channel could be 

 indicated in the case of all lakes, owing to the large extent of 

 ground where live rock is not seen as compared with that which 

 shows the rocks in situ. One of us has had considerable experience 

 in the examination of the exits of lakes, and has found so many 

 (like Llyn Peris and Llyn-y-ddinas) where there is only one possible 

 exit, that he felt sure that if rock-basins exist with any frequency in 



