in the Kinetic Theory of Gases. 87 



Already the perfect symmetry with reference to the YZ- 

 plane shows that the equal oppositely-directed momentum from 

 left to right is carried through the YZ-plane, and that 

 therefore the coefficient of friction is 



V = 2M : B. 



The integrations are effected by putting 



p=— ucos#, q = v sin 6 cos \, r=vsm0sin\. 



Integrating first with reference to \ from to 2tt, then with 

 reference to % it follows that 



/p.r. .., Cl ja . ./2cos 2 v cos 2 sin 2 0f\ 

 r] = 2Fn^/— \ &er**dv \ fflzmOi — j -^ J -Y 



Integration with reference to 9 gives 



V!p— (}-f)> 



and the partial integration of the negative term 



7)=-—Vn\/ — \ -^e~ hv dv. 

 lo V 7tJ / 



If we observe that, if s denotes the diameter of a molecule, 



/=™Vf[r- + (^ + ?)J/- b *]' 



it follows that 



8P f" zh-^dz 



'4 R 



h Jo ze- 



15tts 2 \//i Jo ze~ z * + (2z 2 + 1)J 'er*dz 



This is exactly the expression found by 0. E. Meyer. I 

 will not assert that it is really more accurate than that 

 calculated by Prof. Tait and myself, since by other methods 

 one would obtain other numerical results ; but it is at least 

 of equal authority. This value is smaller than that calculated 

 by Maxwell for the coefficient of friction by about one twen- 

 tieth, whilst the value calculated by Prof. Tait is nearly as much 

 greater. 



The remarks here made at length with reference to gaseous 

 friction of course apply equally to Prof. Tait's method of 

 calculating diffusion and conduction of heat. 



Two other points connected with the theory of gaseous 

 friction may be mentioned here : — 



(1) If a gas move as a whole, with a constant very small 



