560 T. C. CHAMBERLIN 
crystalline rocks, but it is impossible to say how far it was 
absorbed from terrestrial sources. It has even been suggested 
that all the gases both of meteorites and of the crystalline rocks 
were simply absorbed from the air and not brought in from with- 
out; but their proportions are so different from those of the 
air that it would be necessary to assume an extraordinary 
selective power to make this possible ; for oxygen must be wholly ~ 
rejected as a gas; nitrogen, though greatly preponderant, must 
be almost neglected ; carbon dioxide must be absorbed in great 
quantities relatively ; carbon monoxide, though very rare in the 
atmosphere, must be taken in abundantly; while the hydrogen, 
which is scarcely detectable in the atmosphere, must be absorbed 
in superlative amounts. It is difficult to conceive how a mete- 
orite passing rapidly through the air can have absorbed many 
times its volume of an element which does not appear in the air 
in detectable quantities. However, I am unable to say that the 
analyses were made sufficiently soon after the fall of the mete- 
orites to make this point conclusive. But, at any rate, the 
hypothesis of selective absorption is confronted with grave 
difficulties and the alternate hypothesis that the gases are 
brought to the earth in the meteorites seems the more probable. 
The emanations from comets support the view that meteorites 
are charged with gases in extra-terrestrial regions. Here again 
inquiry is needed and experimental tests are obviously sug- 
gested. 
If gases are brought in with meteorites, it is probable that 
independent molecules are flying through space and are caught 
up by the earth. The modern doctrine of molecular velocities, 
which holds that gases are liable to escape, and presumably are 
escaping constantly, from planetary bodies, carries the presump- 
tion that individual molecules are flying through space with some 
degree of frequency. Astronomical phenomena, to be sure, seem 
to indicate that the quantitative value of these cannot be very 
great, but as definite data are yet wanting and we are dealing 
with vast lapses of time and slow processes of depletion, making 
need for slow processes of accretion only, this agency may 
