Reviews of the Ice Age. 227 
Hudson and James bays, a mer de glace flowed outward in all 
directions, and the high mountains of northern Labrador pro- 
jected above its surface. Everywhere the drift is largely de- 
rived from the local rock formations. Its considerable thick- 
ness and varied deposits imply a long and diversified history, 
but definite division of the period into glacial and interglacial 
epochs has not been fully determined. Probably moderate 
oscillations of the ice border may be adequate for the expla- 
nation of such beds as are known in Canada between deposits 
of till. Deep preglacial rock decay was followed by a great 
amount of glacial erosion. Before the Ice age the land was 
much elevated, but when the ice-sheet retreated it was lower 
than now, so that f ossilif erous marine beds were spread above 
the till along the St. Lawrence and Ottawa valleys and about 
Hudson bay. On the higher parts of the country stratified 
gravel, sand, and clay were deposited by the waters discharged 
from the melting ice. 
Prof. T. C. Chamberlin estimated the length of the princi- 
pal interglacial epoch in the northern United States to be much 
longer than the postglacial epoch, as shown by the erosion of 
rock gorges of the Delaware, Susquehanna, and upper Ohio 
rivers subsequent to the deposition of the early glacial grav- 
els of the highest terraces. The pebbles and cobbles of crys- 
talline rocks contained in gravels of the Mississippi valley 
near Natchez, which are referred to glacial derivation from the 
Archa?an regions of the upper Mississippi and lake Superior 
by those who appeal to great uplift of the continent as the 
cause of its glaciation, may instead have come from nearer 
areas of such rocks in the Ozark district of Missouri. The 
successive times of ice accumulation and advance appear to 
have been much longer than the times of its retreat. The 
most important interglacial stage of ice retreat, perhaps un- 
covering all the northern United States, was attended b} r the 
deposition of the greater part of the loess. With the re-advance 
of the ice-sheet, and during the stages of halt in its later de- 
parture, the remarkable marginal moraines which cross the 
northern states and Canada were formed. These number 
twelve to twenty in order from south to north, so Ear as 
already explored, and others will doubtless be found beyond 
these. Between the times of moraine accumulation, the lobes 
