﻿Virial of a Mixture of Ions. 



577 



The extrapolation of these curves to the axis is certainly a 

 rather violent one, but it may be done with some confidence 

 by observing the following points : — (1) the first curve must 

 be everywhere below the limit and the second one everywhere 

 above it; (2) the two curves must extrapolate to the same 

 point on the axis; (3) from the fact that the virial of the 

 ions external to the second arrangement is zero, it may be 

 inferred that the second curve for large values of m approaches 

 the limit much more closely than the first one, and that it 

 probably tends to become parallel to the axis as m~i becomes 



zero. Taking these facts into account, I estimate the limit 

 (see curve) as <£(A)=— 0'66. This gives for the average 

 Phil. Mag. S. 6. Vol. 23. No. 13G. April 1911. 2 Q 



