HYPOTHESES BEARING ON CLIMATIC CHANGES 655 
that every portion of it has once been a constituent of the 
atmosphere and may be again. In atmospheric studies it must 
be recognized as a potential atmosphere. According to the best 
data at command, the ocean holds in solution about eighteen 
times as much carbon dioxide as the atmosphere. But even this 
reserve supply if fully available leaves the perpetuity of atmos- 
pheric conditions congenial to life very short, viewed geologically. 
This threat of disaster is not, however, a scientific argument, 
whatever function it may have in awakening interest and neutral- 
izing inherited prejudice. 
A broad comparison between the atmosphere of Palzozoic 
times and that of Cenozoic times fails, I think, to give proof of 
any radical difference in the constitution of the two atmospheres. 
The magnolia flora in North Greenland in Tertiary times indi- 
cates a scarcely less wide distribution of warm climate than the 
life of the same regionin Paleozoic times. Glaciation in northern 
Norway announced by Reusch and confirmed by Strahan, in 
times apparently just preceding the Paleozoic era, is as sugges- 
tive of atmospheric poverty as anything that introduces the Cen- 
ozoic times. The signs of glaciation at the close of the Palzo- 
zoic erain India, Australia, and South Africa, reaching within 20° 
of the equator, indicate a thermal depression even more mar- 
velous than that which closed the Cenozoic era. The salt 
deposits of the middle latitudes in Paleozoic times, notably 
those of Michigan and New York, in areas where the great 
basins now overflow voluminously, seem to imply an aridity quite 
comparable to anything which has succeeded. The extensive 
terranes of hematite-stained rocks, contrasted with the limonite- 
stained terranes, while their interpretation is more problematic, 
make suggestions of concurrent import. 
A comparison of early with later life, stripped of theoretical 
presumptions, does not seem to me clearly to imply any great 
difference in the content of carbon dioxide. Air-breathing 
life, to be sure, has left no certain record earlier than the middle 
Palzozoic, but these earliest forms afford no clear proof that 
they were determined by non-susceptibility to an excess of carbon 
