J. E. Marr and R. S. Adie — The Lakes of Snowdon. 53 



westerly aspect of tlie slopes causing- much vegetation to grow on 

 the south-west faces of the hills and giving rise to peat-mosses, 

 thereby allowing the waters to discharge gently instead of running 

 oif at once. In connection with this, we may notice that the 

 streams on the north-east side of the Snowdon range are gradually 

 cutting their way back into the hills, thus shifting the watershed 

 to the south-west of its general trend where these notches are formed. 

 This is specially well seen in the case of the passes of Maes Cwm 

 and Cwm Brwynog between Snowdon and Moel Eilio. 



The literature dealing with the Snowdonian Lakes is not ex- 

 tensive. No doubt all geologists have read the masterly account 

 of " The Old Glaciers of Switzerland and North Wales," by Professor 

 Eamsay, which originally appeared in the" series of " Peaks, Passes, 

 and Glaciers," and was afterwards published separately in 1860, 

 Kecently Mr. W. W. Watts contributed " Notes on some Tarns near 

 Snowdon," which will be found in the Report of the British 

 Association for 1895 (p. 683) and also in this Magazine (Dec. IV, 1895, 

 Vol. II, p. 565). During last Easter Vacation we paid some attention 

 to the lakes and tarns of Snowdon, and believe that our observations 

 may be of some use as a small contribution to the subject of the 

 origin of lakes. In the first place we will deal with the lakes of 

 the larger valleys of the Seiont, Gwrfai, and Gwynant, and then 

 notice the lakes and lakelets which lie embosomed in the upland 

 hollows of Snowdon. 



The Seiont, rising at the Pass of Llanberis, flows in a north- 

 westerly direction through the two lakes of Llanberis, Llyn Peris 

 and Llj'n Padarn. These lakes were once one, and are now separated 

 by the alluvial strip near Dolbadarn Castle. This strip is part of the 

 delta brought down by the stream which descends from the heights 

 of Snowdon and Moel Eilio, and it is of interest to notice that at 

 one time a considerable bay must have extended up the valley 

 occupied by this stream, for alluvium extends some way towards the 

 waterfall Ceuuant Mawr. In Llyn Peris there is a bay on the 

 south-west side of the lake, which is continued landward as a valley 

 of some importance, down which a stream of minute proportions 

 runs, and consequently the sediment which it has borne down has 

 been insufficient to fill up the bay. Several similar bays at one 

 time marked the south-west shores of Llyn Padarn, but as the lakes 

 of Llanberis are now used as receptacles for slate-rubbish, the 

 primary features of their shore-lines are almost obliterated. Llyn 

 Peris once extended much further up the valley, as shown by the 

 strip of alluvium extending to Gwastadnant. 



The present exit of the river from Llyn Padarn is between two 

 rocky masses, but there is a considerable width of ground, on which 

 the bridge at Cwm-y-glo is built, which shows no rock ht sitii, though 

 it probably exists at no great depth beneath the sui'face. A depression 

 occupied by alluvium leaves the lake on its western side, about a 

 quarter of a mile south of the present exit, and curving round a rocky 

 knoll joins the present stream opposite tlie village of Cwm-y-glo. 

 It is somewhat over 100 yards wide iu its narrowest part (where it 



