328 Stonsy—Of Atmospheres upon Planets and Satellites. 
and where e, the coefficient of elasticity, depends on the amount of the kinetic 
energy which is expended on internal events during the collision. It is therefore 
necessarily a proper fraction; so that e, in the case of solid particles, cannot 
exceed 1, whereas, in the encounters between molecules, it may have any value 
whether above or below 1. This is because, during an encounter between 
molecules, energy is in some cases imparted to, and in other cases withdrawn 
from, the motions of the molecules along their free paths, whereas, in a mere 
collision, energy is always withdrawn. In fact, the cnternal events of individual 
molecules are in communication with heat motions in the ether, and interchange 
energy with it. A molecule may thus absorb energy from the ether during the 
whole of the long flights which it makes, when near the top of an atmosphere, 
between its encounters; and any excess of energy thus acquired will be shared 
with the motions of translation of the molecules when the next encounter takes 
place. Accordingly, the value of e will vary from one encounter to another, 
and, near the boundary of an atmosphere, there may be changes in the velocities 
of the molecules which are more abrupt than in situations where the gas 
is denser. 
The effect here spoken of would be more marked in the case of helium, water, 
nitrogen, or oxygen, than in that of hydrogen, inasmuch as solar rays of the kind 
that hydrogen can absorb reach the Earth in a feebler state than those which 
the other gases absorb, owing to the partial absorption by hydrogen which has 
already taken place in the hot outer atmosphere of the Sun. On this account the 
rays that can affect hydrogen are the relatively feeble radiations from Fraunhofer 
lines, whereas the molecules of the other gases are exposed on the confines of our 
atmosphere to the glare of full sunshine. This is evidenced by the Earth-lines of 
the solar spectrum, especially those due to oxygen and aqueous vapour. 
These considerations were taken into account in fixing on — 66° C. as the 
maximum temperature to. be attributed to the outer layer of our atmosphere. No 
doubt it would, in some slight degree, improve the investigation to use a rather 
lower temperature in the solitary case of hydrogen; but it was not thought 
necessary to make a distinction of this kind in an investigation which, from the 
nature of the case, could only be approximate. The only effect of introducing 
the refinement would have been to show that the facility with which hydrogen 
escapes from an atmosphere is not quite so much in excess of the facility with 
which helium escapes as the numbers in Chapter IV. indicate. This is almost 
certainly true to some small extent; but it leaves our main conclusions undis- 
turbed. Accordingly, the simpler mode of inquiry, in which these and other 
small differences are ignored, has been an adequate investigation for our 
purpose. 
