534 JOSE RAAT CONTE, 
Bay to have been the center of elevation, of ice accumulation 
and of radiating run-off: Suppose farther, that the elevation 
commenced about the end of the Pliocene and was the cause of 
the cold. Then considering the previous warmth of the Tertiary 
times, it is probable that the elevation would go on for a long 
time and reach a considerable degree before there would be any 
snowfall at all in this region, and still much longer time before 
there would be annual excess over waste by evaporation and 
melting, z. ¢., before there would be perpetual snow. After per- 
petual snow was reached, under the effect of the ice-sheet motion 
tending ever to bring about equilibrium, the increase of the 
thickness of the ice sheet must have been extremely slow. In 
fact, it is evident that no such thickness as actually occurred could 
Fic. 2.—Diagram showing supposed relation of land elevation to ice accumula- 
tion, as now revised. 
have been attained at all, but for the subsidence of the earth-crust 
under the weight of the ice. The increased thickness was con- 
ditioned upon and waited on the subsidence. I believe, there- 
fore, that an increase of one inch per annum would be an extrav- 
agant estimate. At this rate it would take 150,000 years to 
make a thickness of 12,000 feet, which is the estimate of Dana. 
To this must be added the much greater time before the per- 
petual snow was formed at all. 
Under the light of these estimates and especially of these 
new views of a long Ozarkian epoch preceding the glacial epoch, 
I would therefore modify my previous diagram, making the begin- 
ning of the ice accumulation much later and making both the 
Glacial and the Champlain much shorter than before, as shown in 
the revised diagram, Fig. 2. The greater distinctness of the 
Ozarkian from the Glacial is shown 
