OF PLANETARY ATMOSPHERES. 
21 
maguitude of the times which have to be taken into account in tracing the past 
history of our planet under conditions more or less similar to those prevailing in 
present or geological times. 
The times calculated in the })revious paragraph range up to far higher orders of 
magnitude. 
We may therefore safely draw the following conclusions 
1. The Earth’s attraction is cajDable, according to the kinetic theory, of retaining a 
gas of twice the weight of hydrogen in the form of a (practically) permanent atmo¬ 
sphere of uniform temperature, as high as any temperatures commonly existing in 
its present atmosphere. 
2. The vapour of water is similarly capable, according to the kinetic theory, of 
existing on Mars in the form of a (practically) permanent atmosphere of uniform 
temperature, at any ordinary temperature. Hence it follows tliat either 
{a) Helium df)es not escape from our atmosphere and water does not escape from 
that of Mars ; 
Or {h) The escape takes place under far more favourable conditions, such as far 
higher temperatures thantliose assumed in previous investigations, or than we should 
be naturally led to assume from our knowledge of the conditions prevailing in those 
regions of our atmosphere that have been explored ; 
Or (c) The escape is due to translational movements of molecules other than 
those investigated by the methods of the kinetic theory of gases. 
[Postscript added August, 1900; revised October, 1900. 
Since the above paper was read, Dr. Johnstone Stoney has written several papers 
in which he maintains that helinm does escape from our atmosphere. If this view 
be adopted, the present investigation must be regarded as a proof that the escape, 
instead of being attributable to tlie motions which the kinetic theory assigns to 
the molecules of a gas under ordinary conditions as to temperature, &c., must Ije 
due either (1) to the existence of other causes or conditions, or (2) to a divergence 
between the law of distribution which forms the basis of our commonly accepted 
kinetic theory and that occurring in an actual atmos})here. 
Owing to the vastness of the subject, it appears only desirable for me in the 
present note to touch briefly on a few of the main points. 
In connection with (l) Dr. Stoney points out that the actual distribution of mole¬ 
cules of a gas in any particular problem is dependent on two functions, tt and 8, of 
which he supposes the first to determine the normal distrllmtion, wliile the second 
represents the deviations due to disturbing causes. He is of opinion tliat the effect 
of the 8 function is to increase the probability of a molecule acquiring a velocity con¬ 
siderably in excess of the mean velocity. In tliat case it must also correspondingly 
decrease the probability of the molecule jiossessing a velocity not considerably in 
