326 PROF. W. MORRIS DAVIS ON [Aug. I909, 



pluck the rock-wall and bed, and it is believed that freezing and 

 thawing aid in prying off large blocks in the bergschrund around 

 the neve-head, as far down as weather-changes can penetrate. 

 But, as the examination of what is now going on under the ice 

 of glacier-filled cwms is limited to the moderate depth to which 

 hardy observers can descend in the marginal crevasses, the actual 

 behaviour of the ice in the enlargement of a cwm is not yet well- 

 determined by direct observation. Hence such support as is gained 

 for the supposition that glacier-heads can erode retrogressively 

 while they grind down their beds, must come for the present chiefly 

 from the degree of success with which the assumption of glacial 

 erosion can explain the whole series of abnormal forms in glaciated 

 mountains. 



If we now, for a moment, provisionally accept the idea that cwms 

 are the work of glaciers, there remains an interesting enquiry to 

 make regarding the particular pre-Glacial forms that favoured the 

 first accumulation of snow as the embryo of the glacier that was to 

 be developed later. The suggestion may be made that, when all the 

 cwms of such a region as North Wales, for example, are studied out 

 and catalogued as to altitude and dimensions, it may be possible to 

 find on the lower outskirts of the higher and more heavily glaciated 

 district, some small incipient cwms, in which the pre-Glacial form 

 is little changed because the ice-action there was short-lived and 

 weak. By careful classification it may be possible to arrange all 

 the cwms in a series, from these slightly developed forms to the 

 maturely excavated cwms of Snowdon itself. Passing then from 

 North Wales to a more severely glaciated region, such as the Sierra 

 Nevada of California, or Norway, cwms in a still later stage of 

 ■development than those of Snowdon may be found. When this 

 lias been done, it will probably be possible to determine the kind 

 •of pre-Glacial valley-heads that favoured the formation of a local 

 glacier and the excavation of a cwm ; at the same time, it will be 

 possible to frame a systematic method of describing cwms in terms 

 of their stage of development, as well as in terms of their size ; and 

 thus a decided advance will be made in the study of this class of 

 land forms. Still more : as experience broadens, it will probably 

 become possible to classify cwms not only in terms of their own 

 development, but also in terms of the stage of pre-Glacial develop- 

 ment of the mountains and valleys in which the cwms have been 

 excavated ; and, when this is accomplished, we shall have an 

 approach to a really philosophical method of treating these curious 

 forms. 



As a small contribution to this enquiry, mention may be made, 

 first of a faintly developed cwm on the south-western face of 

 Mynydd Mawr ; it is merely a slight recession of the mountain-base 

 and a moderate increase in the steepness of its side slope, without 

 the development of distinct head-cliffs. Later in the series is the 

 well-developed Cwm Du, on the northern face of the same mountain ; 

 here the cwm-floor is well scoured out, and the cliffs at the head and 



