300 PROCEEDINGS: BOSTON SOCIETY NATURAL HISTORY. 
i 
effective erosion is not antagonistic to a belief in effective deposition 
in the case of glaciers any more than in the case of rivers. In each 
case the action varies appropriately to its place in the drainage 
system and to its stas;e in the cvcle. But there will be a later sta 2 :e, 
when the wasting of the superglacial slopes reduces them to mod¬ 
erate declivity, so that the waste delivered from them decreases in 
quantity; then the outflowing water stream at the end of the glacier 
may become a degrading agent; the altitude of the end of the gla¬ 
cier may be slowly lessened; and a very slow and long-continued 
deepening of the whole glacial channel will take place, without 
requiring a departure from an essentially graded condition. 
As the general denudation of the region progresses, the snow fall 
must be decreased and the glacial system must shrink somewhat, 
leaving a greater area of lowland surface to ordinary river drainage. 
When the upland surface is so far destroyed that even the hill tops 
stand below the 200-foot contour, the snow fields will be represented 
only by the winter snow sheet, and the glaciers will have disap¬ 
peared, leaving normal agencies to complete the work of denudation 
that they have so well begun. 
If a snow line at sea-level be assumed, glaciation would persist 
even after the land had been worn to a submarine plain of denuda¬ 
tion at an undetermined depth beneath sea-level. The South Polar 
regions offer a suitable field for the occurrence of such a surface. 
Whether glaciers of the Norwegian or of the Alpine type shall 
occur, is dependent partly on initial conditions, partly on the stage 
of advance through the cycle of denudation. If the initial form 
offer broad uplands, separated by deep valleys, snow fields of the 
Norwegian type may have possession of the uplands during the 
youth of the glacial cycle; but when maturity is reached, the up¬ 
lands will be dissected, and the original confluent snow field will be 
resolved into a number of head reservoirs, separated by ridges. On 
the other hand, as the later stages of the cycle are approached, the 
barriers between adjacent reservoirs will be worn away, and they 
will tend to become confluent, here and there broken onlv by 
Nunatuker. If the snow line lay low enough, a completely con¬ 
fluent ice and snow shield would cover the lowland of glacial denu¬ 
dation when old age had been reached. If the glacial conditions of 
Greenland preceded as long as they have followed the glacial period 
over the rest of the North Atlantic region, who can say how far the 
ice of the Greenland shield has modified the forms on which its 
work began! 
