326 



SCIENCE 



[N. S. Vol. XXXII. No. 819 



hardness, siich as limestones and shales, its 

 sides exhibit a succession of terrace walls 

 and shelving banks, while a marked dip 

 and other dominant structures produce 

 their own modifications. It is also agreed 

 that a valley excavated or greatly enlarged 

 by a glacier should be U-like in section. 

 But an Alpine valley, especially as we ap- 

 proach its head, very commonly takes the 

 following form. For some hundreds of 

 feet up from the torrent it is a distinct V ; 

 above this the slopes become less rapid, 

 changing, say, from 45° to not more than 

 30°, and that rather suddenly. Still higher 

 comes a region of stone-strewn upland val- 

 leys and rugged crags, terminating in 

 ridges and peaks of splintered rock, pro- 

 jecting from a mantle of ice and snow. 

 The V-like part is often from 800 to 1,000 

 feet in depth, and the above-named au- 

 thors maintain that this, with perhaps as 

 much of the more open trough above, was 

 excavated during the glacial epoch. Thus 

 the floor of any one of these valleys prior 

 to the ice age must often have been at least 

 1,800 feet above its present level.^^ As a 

 rough estimate we may fix the deepening 

 of one of the larger pennine valleys, tribu- 

 tary to the Rhone, to have been, during the 

 ice age, at least 1,600 feet in their lower 

 parts. Most of them are now hanging val- 

 leys ; the stream issuing, on the level of the 

 main river, from a deep gorge. Their trib- 

 utaries are rather variable in form; the 

 larger as a rule being more or less V- 

 shaped; the shorter, and especially the 

 smaller, corresponding more with the 

 upper part of the larger valleys ; but their 

 lips generally are less deeply notched. 

 Whatever may have been the cause, this 



"The amount varies in different valleys; for 

 instance, it was fully 2,880 feet at Amsteg on the 

 Reuss, just over 2,000 feet at Brieg in the Rhone 

 Valley, about 1,000 feet at Guttanen in the Aare 

 Valley, about 1,550 feet above Zermatt and 1,100 

 feet above Saas Grund. 



rapid change in slope must indicate a cor- 

 responding change of action in the erosive 

 agent. Here and there the apex of the 

 V may be slightly flattened, but any ap- 

 proach to a real U is extremely rare. The 

 retention of the more open form in many 

 small, elevated recesses, from which at the 

 present day but little water descends, sug- 

 gests that where one of them soon became 

 buried under snow,'^^ but was insignificant 

 as a feeder of a glacier, erosion has been 

 for ages almost at a standstill. 



The V-like lower portion in the section 

 of one of the principal valleys, which is all 

 that some other observers have claimed for 

 the work of a glacier, can not be ascribed 

 to subsequent modification by water, be- 

 cause ice-worn rock can be seen in many 

 places, not only high up its sides, but also 

 down to within a yard or two of the pres- 

 ent torrent. 



Thus valley after valley in the Alps 

 seems to leave no escape from the follow- 

 ing dilemma : Either a valley cut by a 

 glacier does not differ in form from one 

 made by running water, or one which has 

 been excavated by the latter, if subse- 

 quently occupied, is but superficially modi- 

 fied by ice. This, as we can repeatedly see 

 in the higher Alpine valleys, has not suc- 

 ceeded in obliterating the physical features 

 due to the ordinary processes of erosion. 

 Even where its effects are most striking, as 

 in the Spitallamm below the Grimsel Hos- 

 pice, it has not wholly effaced those fea- 

 tures ; and wherever a glacier in a recent 

 retreat has exposed a rock surface, that 

 demonstrates its inefficiency as a plough. 

 The evidence of such cases has been pro- 

 nounced inadmissible, on the ground that 



^ My own studies of mountain aistricts have led 

 me to infer that on slopes of low grade the action 

 of snow is presers'ative rather than destructive. 

 That conclusion was confirmed by Professor Gar- 

 wood in a communication to the Royal Geograph- 

 ical Society on June 20 of the present year. 



