OF PLANETARY ATMOSPHERES. 
‘23 
removed from the field of study of the kinetic theory, and the velocities necessary for 
the escape of gases are largely dependent on the velocities of the jets as a whole. 
But if disturbing’ causes of any kind have to be invoked in order to account for the 
escape of gases, the Boltzmann-Max well doctrine being abandoned as insufficient, 
I do not see how the arguments used by Dr. Stoney, in his 1 897 paper, can be 
regarded as conclusive. In that paper the condition for escape is made to depend 
on the ratio of the velocity necessary for escape to the velocity of mean square. 
This implies that the velocities of the molecules of a gas are distributed about the 
mean according to some definite fixed law, such as “ Maxwell’s Law,” so that (as 
stated on pp. 310, 314 of his paper) a velocity of say nine times the velocity of 
mean square is sufficiently frequeiit to give rise to a marked escape of gas, while a 
velocity of 20 times the velocity of mean square occurs so seldom as to have no 
appreciable effect on the progress of events. But directly external disturbing causes 
are brought to bear on the question, there is no longer any necessary fixed relation 
between the velocities these are capable of producing, and the velocities of mean 
S(piare of the molecules on which they act. 
For example, if the disturbing causes take the form of jets or prominences, they 
will have the effect of impressing on all tlie molecules affected the common velocity 
of the jet. If several different gases occur in the same jet, the changes of velocity 
will bear no fixed relation to the velocities of mean square, but will be independent 
of the latter. Again, consider the effects of tide generating force. If on two 
different planets, one with a satellite and one without, the conditions were equally 
favourable to the permanence of a certain gas, the tide-generating force due to the 
satellite might remove the gas from one planet while it was retained on the other. 
Or again, a certain gas on one planet might, owing to the smallness of the disturbing 
causes, so rarely attain a velocity of 10 times tlie velocity of mean square that 
such occurrences had no appreciable influence. On anotlier planet the disturbing 
causes might become so great as to frequently give the molecules a velocity of 
20 times the velocity of mean square. 
In connection with (2), wlien reading Dr. Stoney’s 1897 paper, I naturally Imagined 
(as I believe others liave done) that Maxwell’s Law was tacitly assumed as the 
basis of his investigations, and my present calculations were undertaken in order to 
place the question on a statistical basis, in the expectation that the conclusions would 
confirm Dr. Stoney’s. I assumed that the object of the d 2 ^osteriori method was to 
overcome a difficulty I had long felt, of drawing a hard and fast line between gases 
wliich do escape, and those which do not. The nearness of the values of the critical 
velocity-ratios found by Dr. Stoney for helium in our atmosphere and water on 
Mars, interpreted in the light of the kinetic theory, naturally justified Dr. Stoney’s 
inference that if the former gas escapes the latter will also escape. My calculations, 
however, show that on the assumed hypotheses neither gas escapes. 
We are now told, however, that Dr, Stoney abandoned “Maxwell’s Law” 
