THE ULTERIOR BASIS OF TIME DiVISIONS 401 
retained at the surface of the earth, the water vapor enters less 
into the atmosphere and low temperature and aridity are the 
consequences. 
If these considerations are valid, the history of the earth has 
been marked by periods of relative cold and aridity resulting 
from stages of rapid rock disintegration, alternating with periods 
of warmth and moisture correlated with periods of limited rock 
disintegration and of carbonic acid accumulation. These stages 
are genetically connected with periods of continental elevation 
and rapid subaérial degradation, on the one hand, and with 
slight degradation and sea incursion, on the other. It will be 
observed that continental elevation as a purely topographical 
condition contributes to cold and aridity, while continental 
degradation correlated with oceanic extension contributes to 
equalization of temperature and to warmth. We have, there- 
fore, the conjoint action of topographic agencies with atmos- 
pheric constitution in producing alternations of cold and aridity 
with warmth and moisture. The aridity is thought to express 
itself in salt and gypsum deposits and in the red sediments with 
which these are habitually associated; the cold, in glaciation; 
the warmth and moisture, in the polar extension of tropical life. 
Now these atmospheric influences are strictly simultaneous 
for all parts of the globe, latitudinal effects, of course, being 
neglected, for the diffusion of the atmosphere is such as to ren- 
der its constitution practically uniform for all parts of the globe. 
In so far, therefore, as atmospheric conditions of a constitutional 
nature affect the progress of terrestrial phenomena, they affect 
them universally, and if these influences are pronounced and can 
be identified they furnish an additional basis for the strict corre- 
lation of transoceanic action and for the division of geological 
history into its natural epochs. 
In summation, therefore, I rest in a somewhat confident hope 
that under continued study adequate natural bases for the more 
important divisions of geologic time and for a stable and fitting 
nomenclature will be found (1) in simultaneous internal read- 
justments alternating with intervals of relative quiescence, (2) 
