HVPOTHESES BEARING ON CLIMATIC CHANGES 673 
gressive development the familiar phenomena due to superficial 
waters would appear. The surface would soon lose its extremely 
fragmentary condition and take on the terrestrial form; the sub- 
terranean frozen zone would disappear and the vulcanism assume 
the terrestrial type. 
This hypothesis, it will be observed, departs radically from 
the familiar view in that it initiates the atmospheric history by 
a tenuous envelope which continued to slowly increase. By the 
hypothesis, as thus far sketched, the atmosphere was derived 
from the interior. After the earth reached the requisite size the 
collection of wandering gases would supplement it. The compe- 
tency of this external source is almost wholly a matter of con- 
jecture and its vague possibilities need not be discussed here. 
It need only be remarked that the hypothesis of molecular dis- 
charge involves the peopling of space with flying molecules. 
The measure of competency of the interior to supply an 
atmosphere is obviously a critical question. Unfortunately we 
are almost entirely without specific quantitative data bearing on 
the subject. We know that there is not a little atmospheric mate- 
rial in the interior as demonstrated in volcanic action and in the 
content of the minute pores of the hypogene rocks, but we do 
not know how far this was derived from the surface. If the 
moon never has had an appreciable external atmosphere its 
explosive eruptions were not due to surface infiltration and the 
implications of its numerous and vast craters are very sug- 
gestive. We can also draw inferences from meteorites which 
/ sometimes contain several times their volume of gases, as well 
as solid matter susceptible of conversion into atmospheric con- 
stituents. But at best we can only form very vague quantitative 
notions. On the other hand, we are liable to overestimate the 
amount required. The atmosphere and the ocean combined are 
little more than ;,),, of the mass of the earth. To be com- 
petent, the ingathered matter need therefore only contain about 
sy Of I per cent. of atmospheric and aqueous material, plus an 
added factor for what may have been lost and what still remains 
in the interior. This percentage does not seem large enough 
