478 MR. S. CHAPMAN ON THE KINETIC THEORY OF A GAS 



The expression obtained in Part I. for the coefficient of diffusion between two 

 gases is 



1/2 /m + m'V /2 1 



\hmrn') ( 

 Since 



v + v p+p' Pt ' 



where p t is the total pressure, and since v and v do not occur elsewhere in the 

 expression, it follows that, according to the present theory, the coefficient of diffusion 

 is independent of the relative proportions of the gases in the mixture. On this point 

 STEFAN and MAXWELL were at issue with MEYER,* whose theory predicted a large 

 change in the coefficient of diffusion as the proportions of the gases were varied. 

 Several experiments! have recently been made at Halle to test the rival theories. 

 The pairs of gases taken were 0-H, O-N, H-N, A-He, C0 2 -H. The ratio of the 

 components was varied from 1 : 3 to 3 : 1 (roughly), and the results showed systematic 

 differences in the values of D 12 which reached 8 per cent, in extreme cases. Though 

 part of this variation may be due to experimental errors, it is undoubtedly the case 

 that appreciable variations of D ia exist, or, at any rate, variations in the values of D J2 

 as derived from the accepted theory of the experiments. This theory is of rather an 

 elaborate nature, and takes no account of the boundary conditions ; any attempt at a 

 revision of the theory of diffusion should include an examination of the theory of the 

 experiments by which the diffusion is measured ; and it is by no means impossible 

 that faults in the accepted theory of the experiment may account for the small 

 variations of 5 per cent, or so in the determinations of D 12 . At the same time it must 

 be remembered that the value of D 12 here arrived at theoretically is based on the 

 assumption that the square of the velocity of diffusion, together with small deviations 

 from MAXWELL'S law of distribution, are negligible. If these had been taken account 

 of (and it is easy to do so) part or the whole of the outstanding discrepancy would 

 probably be explained. As the variations are so small, however, I have not troubled 

 to carry out these calculations. 



The experiments emphatically disprove MEYER'S theory, which predicts changes of 

 20 per cent, or more in D 12 , and there can be no doubt that the general principles of 

 the theory as laid down by STEFAN and MAXWELL are much nearer the truth than 

 those advocated by MEYER. A modification of MEYER'S theory by GROSS,! which 

 reduced the amount of the variations predicted by MEYER, has also been shown to be 

 incorrect, since the actual' variations of Di 2 , though similar in amount to the theoretical 

 values deduced by GROSS, are in the opposite direction. 



* See his ' Kinetic Theory,' Chapter VIII. 



t R. SCHMIDT, 'Halle Diss.,' 1904, and 'Ann. d. Phys.,' 14, 1904, p. 801 ; R. DEUTSCH, ' Halle Diss.,' 

 1907 ; JACKMANN, ' Halle Diss.,' 1906 ; and LONIUS, ' Halle Diss.,' 1909. All these results are summarized 

 by LONIUS, ' Ann. d. Phys.,' 29, 1909, p. 664. 



J G. GROSS, ' WIED. Ann.,' 40, p. 424, 1890 



