HYPOTHESIS OF CAUSE OF GLACIAL PERIODS 551 
has been shown that the carbon dioxide and water vapor of the 
atmosphere have remarkable power of absorbing and temporarily 
retaining heat rays, while the oxygen, nitrogen, and argon of the 
atmosphere possess this power in a feeble degree only. It fol- 
lows that the effect of the carbon dioxide and water vapor is to 
blanket the earth with a thermally absorbent envelope. Their 
absence would leave the surface of the earth essentially exposed 
to the free impact and free radiation of the solar rays, measura- 
bly, though of course not entirely, as if the earth were devoid of 
atmosphere. A reduction or an increase in these constituents 
would produce corresponding partial effects. The general 
results assignable to a greatly increased or a greatly reduced 
quantity of atmospheric carbon dioxide and water may be sum- 
marized as follows :? 
a. An increase, by causing a larger absorption of the sun’s 
radiant energy, raises the average temperature, while a reduction 
lowers it. The estimate of Dr. Arrhenius, based upon an elabo- 
rate mathematical discussion of the observations of Professor 
yLangley, is that an increase of the carbon dioxide to the amount 
of two or three times the present content would elevate the 
average temperature 8° or g° C. and would bring on a mild cli- 
mate analagous to that which prevailed in the Middle Tertiary 
age.* On the other hand, a reduction of the quantity of carbon 
dioxide in the atmosphere to an amount ranging from 55 to 62 
per cent. of the present content, would reduce the average tem- 
perature 4° or 5° C., which would bring on a glaciation compa- 
rable to that of the Pleistocene period. 
The amount of moisture in the atmosphere, other things being 
equal, is directly dependent upon the temperature, and the tem- 
perature in turn is dependent upon the amount of moisture in 
the atmosphere. This reciprocal dependence renders the aqueous 
t The action has two phases that in an exhaustive exposition would need separate 
statement, (1) the absorption of solar rays before they reach the earth, and (2) the 
absorption of the rays radiated by the earth. The latter have longer wave-lengths on 
the average and are relatively much more affected by the constitution of the atmos- 
phere. 
2 Dr. Arrhenius, loc. cit. 
