582 G. K. GILBERT 



of this relation that I avail myself of Johnson's hypothesis. Each 

 glacier which receives more snow on one side than on the other 

 adjusts its cross profile to a condition of equilibrium by moving away 

 from the region of greater supply to the region of lesser supply. This 

 lateral motion is, of course, combined with the general, and more 

 rapid, forward motion of the glacier, but it is nevertheless competent 

 to produce at the side of the glacier phenomena quite similar to those 

 at the head. Fig. 3 shows diagrammatically the ground plan of a 

 glacier flowing southward. The greatest snow accumulation is in 

 the cirque, A A, where precipitation is at a maximum, where depletion 

 through solar influence is at a minimum, and where circling cliffs pro- 

 tect against removal by the wind. There is great accumulation also 

 along the west margin, BB, where the snow drifted by the westerly winds 

 comes to rest in the shelter of the west wall of the glacier trough. 

 Along the eastern border, CC, the snow deposit is comparatively 

 small, because of exposure to the westerly winds. The resulting 

 lines of ice-flow are as drawn.- Moving directly away from the walls 

 of the cirque, the glacier makes and annually renews the bergschrund, 

 ah. Moving obliquely away from the west wall of the trough, the 

 glacier similarly produces the minor bergschrund he, and this minor 

 bergschrund leads to sapping and the production of a cliff, just as 

 the major bergschrund causes the cirque cliff. Thus the west wall 

 of the trough is kept steep, and is thereby contrasted not only with 

 the east wall of the same trough, but with the east wall of the adjacent 

 trough, so that the rock crest between the two troughs is not symmetric. 



Usually in viewing a cirque it is possible to trace about its wall a 

 somewhat definite line separating a cliff or steeper slope above from 

 a gentler, usually scalable, slope below. This line I conceive to mark 

 the base of the bergscrhund at a late stage in the excavation of the 

 cirque basin. I have called it in my notes "the schrund line." It 

 can usually be traced for some little distance beyond the cirque, and 

 sometimes for several miles on one wall or other of the glacier trough. 

 Advancing along the trough wall, it descends gradually with a slope 

 which may be assumed to represent the gradient of the ice surface, 

 that surface having been somewhat higher than the schrund line. 

 Its expression outside a cirque may be seen in Figs. 2, 5, and 6. 



At a somewhat lower level than that to which the preceding para- 



