of Gases from Atmospheres. 511 



thousand millions per second : and it is an easy inference 

 from what he says that it is only within limits which can be 

 defined that it is permissible to treat the distribution of 

 velocities within his model as sufficiently approximating to 

 the distribution within a gas ; as, for example, when investi- 

 gating phenomena, like wind or viscosity or diffusion, in 

 which the succession of events to be explained is enormously 

 shiver than the above rapidity of molecular encounters. 

 Moreover, he is careful to point out that there are other 

 respects in which his model fails to represent actual gases, and 

 speaking of one such shortcoming Maxwell says — u This 

 result seems decisive against the unqualified acceptation of 

 the hypothesis that gases are such systems of hard elastic 

 particles" (see Maxwell's Scientific Papers, vol. i. p. 409). 



I am anxious to emphasise this because I knew Clerk 

 Maxwell, with whom no one could be brought into contact 

 without being impressed by the marvellous grasp and accuracy 

 of his physical insight ; and I always feel regret when I see 

 mistakes of subsequent writers represented as having been 

 his. It is precisely by ' the unqualified acceptation ' of 

 Maxwell's hypothesis which Maxwell deprecates, and by mis- 

 applying it to gas under the conditions which prevail in the 

 penultimate stratum of an atmosphere, that the results come 

 out which Mr. Cook has given. 



The present writer arrived at what were practically the 

 same numerical results either 36 or 37 years ago, at an early 

 stage of his investigation, by applying Maxwell's formula in 

 the same way. But there was this marked difference between 

 Mr. Cook's treatment of the subject and mine. I was all 

 through aware that 1 was not dealing with gas but with 

 Maxwell's model, while Mr. Cook represents himself as 

 thinking that Maxwell's formula gives the true distribution 

 of velocities in gases, and under all circumstances. When 

 therefore I found that Maxwell's formula did not explain the 

 escape of carbon dioxide from the Moon, I searched for the 

 cause of its insufficiency, with the result that I found those 

 several respects in which it is insufficient which are enumerated 

 in my paper in last June's Philosophical Magazine * (see 

 also Astrophysical Journal for July 11)04, p. 69). 



I regret to see that Mr. Cook is under the impression that 

 his data have been put in a false light in a passage which he 

 quotes from my paper of last June : in this I think and hope he 



* In the footnote at the end of that paper (see p. 700) readers are 

 referred to an earlier paper "On the Physical Constitution of the Sun 

 and Stars " in the Proceedings of the Royal Society for 1^98. This should 

 have been in the Proceedings of the lioyal Society for 18G8. 



