some other Constants of the Inactive Gases. 



807 



This miscellaneous collection of figures shows very plainly 

 that even with the simple monatomic gases, the kinetic theory 

 is not altogether satisfactory. A similar set of figures for 

 helium and argon has been published by Jeans (I. c. and 

 Phil. Mag. (6) viii. p. 692 (1904)), but, as far as my know- 

 ledge goes, the molecular constants of the other gases have 

 not been calculated before. 



Part II.— Calculation of N — the Number of 



Molecules in Unit Gas Volume. 



* 



From the equation L = — /cj wj , w e can easily derive 

 the relation N 



320L 3 V : 



where V= - , No- 3 as before, 

 b 



Using the values of V obtained from the coefficient b of 

 van der Waals's equation, from the dielectric constant K, and 

 from the refractive index //,, we can obtain a series of values 

 of N for helium and argon, for which gases L is known 

 from the viscosity coefficient. The following table gives the 

 results : — 





Values of V X 10s. 



I 

 Values of N x 10 19 . 





From b. 



From K. 



From ft. 



From b. 



From K. 



From fj.. 



He 



A 



1-8 

 33-7 



247 



2-31 

 18-61 



0415 

 2-7 



22 



25-2 



8-87 





The agreement 

 L 



is not good, but we must remember that 

 y error in ij or V becomes much magnified in the final 

 result owing to the fact that these quantities occur as cube 

 and square respectively. Still this alone will hardly explain 



(Ann. d. Phjs. xxv. p. 894 (1908)) 

 L 3 V 2 for a large number of gases 

 helium and argon, and found all 



the discrepancy. Sirk 

 calculated the values of 

 and vapours, including 

 sorts of values ranging between 1*7 and 2S'2 ( x 10 -23 ), 

 which would give values of N ranging from 18'4 x 10 19 to 

 1*10 x 10 19 . The abnormally high values for helium Sirk 

 ascribes to experimental errors in the determination of L and 

 fi, but this I am not inclined to endorse, as both /j, and n are 

 now known with quite considerable accuracy. But I cannot 

 see any other obvious explanation, especially as hydrogen 

 o-ives a value of about 9 x 10 19 calculated from //,. 



312 



