574 THE JOURNAL OF GEOLOGY. 
perhaps, the twofold division, appear to pass from side to side 
of the glacier. This may be verified by comparing the east, the 
frontal and the west sides as shown in Figs. 31, 35 and 33. 
The lamine are generally plane or slightly undulatory and 
essentially horizontal, but occasionally they are warped and 
crumpled; sometimes they are faulted. 
The major planes of stratification seem to be essentially par- 
allel with the bottom of the glacier. In detail this is not always 
true, and theoretically it is probably not accurately true as a gen- 
eralization—but in a broad sense it appears that the attitude of 
the layers is controlled by the bottom and not by the top of the 
glacier. The profile views illustrate this. Apparently the white 
ice thickens rapidly as we go back from the edge, while the dis- 
colored ice at the base probably remains approximately constant — 
in thickness. This seems to me to agree with theoretical pre- 
sumptions. The melting of the glacier is chiefly at the sur- 
face. There is little ground for believing that any considerable 
amount of interior melting takes place that is not compensated 
by weireezine. Hence the upper layers must rnin out as mat 
approaches the edge of the glacier, while the lower layers remain 
approximately constant. The ratio, therefore, of the débris- 
bearing ice to the white ice above it, as seen on the edge, cannot 
be regarded as applicable to the thicker portions of the glacier 
back from the edge. At the extreme margin of the Bryant gla- 
cier the débris-bearing ice constitutes two-thirds of the section. 
It certainly would lead us far astray to assume that the débris is 
distributed through two-thirds of the vertical section of the ice 
on the summit of Redcliff peninsula. 
The drainage of Bryant glacier is essentially the same as that 
of the two northern glaciers previously described. The waters 
produced by melting are shed from the back of the glacier in 
rivulets, very few of which find an opportunity to descend to the 
bottom through crevasses because of the scantiness of these. The 
rivulets, however, instead of cascading down the sloping face of 
the glacier as they were permitted to do in the preceding instances, 
are forced to project themselves from the overhanging edge as 
