318 PROP. W. MORRIS DAVIS ON [Aug. I9O9, 



XXI. Special Features of the Theory of 

 Glacial Erosion - . 



As in the theory of glaciers as protective agencies, there are 

 certain special features in the theory of glaciers as destructive 

 agents which need special mention. One of the most significant 

 points to consider is the varying duration of glacial action. 



(a) If the Glacial Period were short, the ice might disappear 

 before the cwms were enlarged enough to destroy much of the 

 pre-Glacial mountain-form, and before the glacial channels were 

 well excavated beneath the pre-Glacial valleys ; then the sides and 

 floors of the channels would be left ragged and uneven, as a result 

 •of the more rapid work of erosion on the weaker parts and the less 

 rapid work on the more resistant parts ; and, so long as the young 

 glaciers were actively deepening their channels, the channel-sides 

 would be steep aud cliff-like, with a U-shaped cross-profile. 



(b) But, if the Glacial Period were long, the channels might be 

 worn to such a depth that further deepening would be slow; then 

 the floor and sides would be smoothed and the channels would be 

 widened so as to lose their U-shaped cross-profile and gain that 

 of a round-bottom V. 



(c) Wherever a considerable mass of weaker or more jointed 

 rocks occurred in the graded pre-Glacial valley-floor, young ice- 

 streams would deepen their channels more rapidly than on harder 

 or less jointed rocks next up-stream ; thus rock-steps would be 

 produced, and once produced they would ' retreat ' up-stream by 

 the removal of rock-blocks from their face. Similarty (d), the 

 channels would be eroded to a greater depth in belts of weak rocks 

 than in belts of hard rocks next down-stream, and thus rock-basins 

 would be formed in the channel-beds. 



If (<?) the Glacial Period were very long, some of the cwms ex- 

 cavated in the valley-heads might grow so large as to reduce 

 and consume the spurs between them, and thus destroy all traces 

 of pre-Glacial form, leaving only sharpened peaks and ridges 

 between the cwms, as in fig. 19 (p. 316): then (/) the channels 

 would be so well perfected that their sides and floors would be 

 nearly or quite smooth, and few rock-steps would remain in their 

 course, as in fig. 21 (p. 317). 



But (g) rock-steps would still necessarily occur at the junction of 

 small and large glacial channels, however long the Glacial Period 

 endured; for it has already been shown that a discontinuity of 

 channel-bed must persist at the junction of a small tributary with a 

 a larger trunk stream, whether the stream be of ice or of water, 

 however maturely the channel beds are eroded to an even slope 

 elsewhere. 



(h) The hanging lateral valleys would, however, not be syste- 

 matically arranged in sets related to rock-steps farther up-stream 

 in the main valley ; they would be irregularly placed as to altitude, 

 for the depth of the floor of a hanging valley below the upper limit 

 of glaciation would ordinarily be related to the depth of the main 



