Molecules of a Gas in a Field of Force. 



643 



The expression on the right involves T through the three 

 last factors only. As to w 2 — wi it is probable that its value 

 is almost independent of T, but if it has a linear temperature 

 coefficient the effect will be to multiply the coefficient of 



const. 



e t~ by a constant factor. To investigate the value of i 

 ns a function of T, it is clearly necessary to consider the 

 behaviour of the expression 



. . . (35) 



^*> a U£)'- 



when T is varied. As to this, it is to be remembered that 

 C and x x are not independent, but that a\ is determined 

 when C is given. G itself involves both V : and T, being 

 proportional to TVi 3 . In dealing with metals it is probable, 

 from a consideration of the probable numerical values of the 

 quantities occurring, that Y 1 will not vary much with T even 

 at moderately high temperatures. At any rate, as a first 

 approximation we can treat Y 1 as independent of T, and C 

 as directly proportional to T. The following table exhibits 

 a series of corresponding values of 0, x^ log e g{G, Xi) and T. 



I. 



II. 



III. 



IV. 



c. 



x v 



log^CUO- 



T( o K) ^^/0N\t 

 M ; 2MR\4ttVJ U - 



(V 1 =4-16X10 3 ). 



•09 



•08 



•07 



•06 



•05 



•04 



•03 



•02 



•015 



•01 



o 



22 

 24 

 2-65 



304 

 3 64 

 4-44 



6-42 

 8-60 

 12-6 



•940 

 1-200 

 1-493 

 1-856 

 2-338 

 3-002 

 3-968 

 5-92 

 7-94 

 11-78 



1800 



1600 



1400 



1200 



1000 



800 



600 



400 



300 



200 



The values of x\ have been obtained from the corresponding 

 values of C by the graphical method already described. The 

 values of T are, strictly speaking, not values of T at all, but 



arc va 



lues of 5^(Jj^j f C, where V 1 = 4-16x10^' 



CCS. 



