J. E. Marr and R. H. Adie — The Lakes of Snotcdon. 65 



Britain, there must be some where proof is obtainable that there is 

 no possible dritt-fiUed exit. Cwellyn illustrates this : tliere is no 

 possible exit at the foot, and if this lake were backed by high hills 

 towards the head, the existence of a rock-basin could be proved here ; 

 but, as we shall now proceed to point out, the physiographical features 

 at the head of the lake are compatible with the existence of a drift- 

 tilled depression in this direction ; and, indeed, some of the phenomena 

 exhibited above the head of Cwellyn are extremely dilEcult to 

 explain, unless such a depression exists there. 



An alluvial flat extends above Cwellyn for a quarter of a mile up 

 the Grwrfai. This river runs over solid rock at Khyd-ddu between 

 Cwellyn and Llyn-y-gader, but a drift-filled depression is traceable 

 from the head of the alluvial flat, firstly up the main stream, then up 

 a tributary joining it a few yards south-east of Cwellyn Slate Quarry; 

 it crosses a low watershed at a height of 740 feet (i.e., nearly ^00 feet 

 above the surface of Cwellyn, and nearly loO feet higher than that 

 of Llj'n-y-gader), just east of Ffridd Slate Quarry, after which it 

 follows another small stream and joins Llj'ii-y-gader close to the 

 prominent crag, which stands out of the alluvium on the east side of 

 the lake. This is the onlj' possible exit in this direction, and its 

 resemblance to a drift-filled gorge is very striking. The depth of 

 Cwellyn is, so far as we are aware, not known ; but assuming it is 

 nearly 50 feet deep, the gorge, which is about 80 yards wide at the 

 narrowest part, would require to be 350 feet deep at this point, in 

 which case it would be comparable with some of the Alpine gorges. 

 Such gorges might well be cut by the water issuing from a glacier, 

 and highly charged with sediment, and the nature of the ground is 

 favourable for the formation of one at this point, for it is occupied 

 by a well-jointed basic intrusive rock. Furthermore, the rock comes 

 to the surface here so extensively, that there is no approach to 

 any similar drift-filled depression ; in fact, where the depression 

 crosses the col, it is a conspicuous feature owing to the rocky ridge 

 above and below it, and it is difficult to understand why this con- 

 tinuous tract of drift-covered ground occurs here, except on the 

 supposition that a valley lies below. If the possibility of the 

 existence of the gorge be admitted, there is no difficulty connected 

 with the introduction of the drift, for the locality is just beneath the 

 great cwm on the west side of Snowdon, which must have been the 

 gathering-ground of a large glacier, and also lies below the drift- 

 covered tract on the upland plateau on which Maen Bras stands. 

 1'here is another remarkable feature which requires explanation, but 

 which is readily understood if it be supposed that a drift-filled valley 

 exists here. To the south-east of Llyn-y-gader is a peaty moor, which 

 slopes gently to the watershed separating the Gwrfai from the 

 Colwyn. Viewed from the detached mass known as Pitt's Head, the 

 watershed appears as a level line, and is apparently composed of an 

 alluvial deposit. It strikingly resembles one which one of us has 

 l^reviously described at the head of Wet Sleddale in Westmoreland 

 (GrEOL. Mag., Dec. IV, Vol. I, p. 539), and as in the case of the Wet 

 Sleddale watershed, may be accounted for on the supposition that 



