HVPOTHESTS OF CAUSE OF GLACIAL PERIODS 563 
sources has proceeded more rapidly at or about the periods of 
crustal disturbance. This, as we shall see, coincides practically 
with the periods of depletion from atmospheric action on the 
surface, and hence, so far forth, the two processes tend to neu- 
tralize each other and preserve the constancy of the earth’s 
atmosphere. The hypothesis must therefore recognize that it 
was only when one agency fell behind the other in its com- 
petency that its specific results became manifest, and then only 
by the difference in their respective effects. 
Varying vates of loss.—The rate of chemical action of the 
atmosphere on the surface of the rocks is believed to have been 
intimately connected with the extent and height of the land area, 
considering the earth as a whole. There were qualifying con- 
ditions, as we shall see, but notwithstanding, this is regarded as an 
important law. It is madea fundamental postulate of the hypoth- 
esis, and the vitality of the hypothesis as a working instrument 
of investigation hangs very largely upon it. It is obvious that 
theyereater the suriace-area of rock exposed ‘to the effective 
action of the atmosphere, the more rapid will be the rate of 
disintegration, other things being equal, and the more rapid the 
consumption of carbon dioxide. * 
The rate of carbonation of the rock is dependent upon 
elevation as well as superficial area. The disintegration of rock 
is the most active by far in the zone lying between the surface 
and the level of permanent underground water, technically the 
water table. It is in this zone that the atmosphere and the 
moisture of the earth combine to give the greatest chemical 
t Oxygen is also consumed in the decomposition of average rock, but in less 
amount than carbonic acid, and as the amount of oxygen in the air is very much 
larger than that of carbon dioxide, the part consumed is far less critical. In an 
exhaustive study of the constitutional history of the atmosphere, the loss and gain of 
oxygen must be considered, and certain very interesting and important phases of 
atmospheric variation are probably connected with the production and consumption 
of the oxygen, but, as indicated, they are much less immediate and critical in the 
consideration of thermal problems with which our hypothesis is more especially con- 
cerned, and for the sake of simplicity the oxygenation of the rocks may be tempora- 
rily neglected; and for like reasons the many minor reactions may also be ignored 
and attention confined to the carbonation. 
