330 PROF. W. MORRIS DAVIS ON [-^-Ug. 1909, 



basin as compared with, the erosion farther down the valley. It has 

 been urged that, given a basin in the path of a glacier, the retar- 

 dation of the bottom part of the ice in the depression, and the faster 

 motion of the ice over the outlet would result in the obliteration of 

 the basin ; and, if no other factors entered into the problem, this 

 conclusion might well be accepted. But other factors do enter, 

 especially in the way of variations of rock-resistance and of ice- 

 pressure and velocity; hence it is perfectly reasonable to believe that, 

 if an eroding glacier found no basins in the pre-Grlacial valley of 

 which it took possession, it would tend to produce basins wherever 

 the rocks were more easily excavated, or wherever the ice-motion 

 was accelerated. Naturally, a balance would in time be struck 

 between the forces of excavation of the basin and erosion of the 

 outlet ; and then erosion would proceed at essentially the same rate 

 in the basin and at the basin-outlet. But no one knows what 

 measure of inequality the bed of a glacial channel may have when 

 such a balance is reached. Longitudinal profiles drawn on a true 

 scale are instructive in this connexion ; they usually show a much 

 less inequality of valley- (ice-channel) floor where lakes occur than 

 is indicated in freehand profiles of the same district. This is 

 particularly true in a district where the lakes are so small as in 

 North Wales. Indeed, if glaciers did any erosive work in North 

 Wales, the lake-basins are so small a part of it that they need hardly 

 be considered in the present discussion. Even if the valley-floors 

 had not been here and there deepened in excess so as to hold lakes, 

 there would still be an abundance of abnormal forms in the Snowdon 

 district, more indicative of strong glacial erosion than are the lake- 

 basins. 



Yalley-floors : rock-steps. — Abrupt rock-steps extending 

 all across the broad floors of wide opened valleys are common in the 

 Snowdon district. They would be well illustrated by true-scale 

 profiles through Snowdon summit ; but the construction of such 

 profiles is not easy, because the Ordnance Survey map has a contour- 

 interval of 250 feet at altitudes over 1000 feet. One profile might 

 pass from the eastern to the western cwm ; the other from the 

 north-western to the southern cwm. The eastern profile in Cwm 

 Dyli would show three strong steps ; and a fourth might be 

 counted if the series began with a small uppermost cwm, not 

 shown on the 1-inch Ordnance Survey map. Lakes (Glaslyn and 

 Llydaw) occupy rock-basins between the upper steps, and these 

 once supplied a dashing torrent on the lower step ; but most of the 

 water is now taken down in large pipes to an electric power- 

 station, the electric current being carried thence by many wires to 

 the slate-quarries at Ffestiniog, Llanberis, and Nantlle. The north- 

 western (Cwm Dur-arddu) and the western (Cwm Clogwyn) pro- 

 files would each include two well-defined steps ; but the steps do not 

 agree with each other in altitude or in height, and they necessarily 

 disagree with the three steps in the eastern profile. The southern 



