of the Phenomena of Dissociation in Gases. 29 



If we put N 2 — in this equation, the value of Bj is ob- 

 tained as follows: — Let J 1 be the value of J when N 2 = 0. 

 It is easy to show that 





+ (2x 2 + \)C e-*\lx 



xe~ x2 



(13) 



•JO 



In exactly the same way that Clark calculated the table given 

 in Prof. Tait's paper previously quoted, I have found as 

 numerical value of the integral in equation (13), 



0-0605022. 



Hence B x = 4*2706 NjR^ 2 a ; while previously (from equa- 

 tion (3), putting ^ = 0*886 since N 2 = 0) we found for B, the 

 value 4 , 2584N 1 R 12 2 a. In a similar manner B 2 could be cal- 

 culated. It will therefore be seen that the former method 

 gives fairly correct results. 



There is one other method of calculating /a, which I will 

 again apply to the case N 2 = 0. From the equation (6) of 

 § 4, we can show that 



T! -«'K I Bi w ] . . . . (14 ) 



* \ JL e -*v*i* dv 



Jo ^1 



The denominator is equal to J x and is therefore known. 

 On writing Y = ax, we obtain 



V ^B i _^--+(2^-l)j" 



B x 3V 



+ (2^+1) (" ' e-*dx 



(15) 



xe 

 whence 



e~ x2 dx 



V 2 



-if , •"-"( -, [i + '• • ~ ... ]*■ tu» 



Jo Jo 



Applying Clark's method, I have found for the numerical 

 value of the integral (16), 0*088139, which gives 



V* = 0*67776a 2 . 



