
Escape of Gases from Atmospheres. 699 
All that is then necessary for a molecule to escape is that 
one of the two that have encountered shall have the direction 
of its flight outward, that it shall have sufficient speed, and that 
it shall escape other encounters. If the chance that these 
events shall happen befalls each molecule in the penultimate 
stratum of the helium atmosphere as often as once in several 
days, there would probably be an abundant outflow of 
helium from the earth to account for the observed rate of 
its escape. 
Here, however, we are on debatable ground. We can 
only follow events in detail with probability, not with 
certainty. But on the other hand, when we trust to the 
inductive argument based on the ascertained behaviour of 
helium, as stated in an earlier paragraph, we are on secure 
ground. We may rely on the conclusion to which it leads 
viz., that helium is escaping from the earth’s atmosphere, 
and that the rate of escape is the same as the rate of the net 
inflow from the earth into the atmosphere. By the net 
inflow is meant the supply after deducting something like a 
1/6000 or 1/3000 part of the whole, in order to allow for 
the very minute quantity of helium that had been washed 
out of the atmosphere by rain and which is being restored 
to it. 
There are other matters, too, which would need to be 
understood and allowed for before we should be entitled to 
trust the deductive method of proof. Thus, the internal 
events that go on within the molecules of matter are of more 
than one kind, and in gases stand differently related to the . 
translational motions. This is revealed to us by phospho- 
rescence and other phenomena. An attempt to make a 
preliminary classification of these internal events has been 
made by the present writer in a memoir on the Kinetic 
Theory of Gas*. But without going into these and other 
matters, enough has been said to show how inadequate the 
deductive method is—at least as hitherto handled—to be—a 
safe guide in dealing with the matters with which it has been 
made to grapple. Thisof course also shows that objections based 
on investigations of this character, have no weight against the 
testimony about the rate at which gases do actually escape 
from atmospheres which is given by such facts as the absence 
of atmosphere from the moon and the behaviour of helium 
upon the earth. 
* “Of the Kinetic Theory of Gas regarded as illustrating Nature.” 
By G. Johnstone Stoney, F.R.S. See Scientific Proceedings of the Royal 
Dublin Society of June 1895, vol. viii. p. 356; or Phil. Mag. for October 
1895, p. 362. 
