212 THOMAS C. CHAMBERLIN AND ROLLIN T. CHAMBERLIN 



exposed portions of the face, and by expansion in the joints are 

 likely to rive the wall rock and to detach fragments from it. It 

 seems probable from the nature of the case that the exit of such 

 internal drainage takes place more largely at or near the foot of 

 the cirque wall than at higher levels. Such a localization of the 

 action is specially fitted to promote basal sapping. If the sapping 

 at the base of the cirque wall be thus made in some notable part 

 dependent on the seepage of water from the mountain mass at 

 or near the base of the wall, it will not perhaps seem strange that 

 the sapping should proceed somewhat downward as well as back- 

 ward, following in reverse the direction that the waters of seeps 

 and springs usually take in issuing, and thus give rise to the impor- 

 tant fact observed by Johnson that the floor of the cirque fre- 

 quently inclines 'somewhat toward the cirque wall. 1 



In this view, the sapping is not made in any radical way depend- 

 ent on diurnal changes of temperature due to the openness of 

 the schrund above; rather it presumes that the base of the schrund 

 will often be filled with snow, ice, or rock fragments fallen from 

 above and that it will not be freely exposed to the briefer class 

 of changes of temperature that affect the outer air. It does 

 presume, however, that the mean temperature at the base of the 

 schrund, and at the base of this part of the glacier generally, favors 

 freezing whenever tension aids, and that it is favorable to glacial 

 growth rather than glacial wastage, as this is the general fact in 

 this part of a glacier. The periodicities of seasonal temperature 

 and the variations attending the cyclonic movements of the atmos- 

 phere extending over some days seem to us more probably effective 

 in the sapping at the base of the cirque wall than daily changes. 



V. GLACIAL STEPS 



Conditions somewhat similar to those of the cirque, save in the 

 matter of the beginnings of motion, are often found at other points 

 along the length of the glacier below the cirque. They do not 

 seem to be in any way dependent on the existence or the absence of 

 a declared cirque at the head of the glacier. If a glacier takes 

 origin in a sharp gulch or in a pointed valley, a typical cirque 



1 Willard D. Johnson, Jour, of Geol., XII, 576. 



