106 
MR, W. R. BOUSFIELD: IONIC SIZE IN RELATION TO 
and NaCl solutions could be expressed by one formula with the same constants for 
both substances, viz. :— 
D = 1-86 + 3-3 8V„ 
where §V S is the change in the solution volume for different dilutions. This formula 
would also include a non-electrolyte such as sugar, in which SV s is almost zero. 
Furthermore, by correcting the concentration for the amount of water abstracted 
by and combined with the solute, a constant value was obtained for the effective 
molecular freezing-point depression of KC1 solutions down to a concentration of twice 
decinormal. 
The equivalent refractive power of IvCl and NaCl solutions was similarly shown to 
be a linear function of the ionic volume, and, consequently, also of the solution 
volume. 
In the sketch of the course of the present paper we have passed over some matters 
arising incidentally which may call for mention. But it must be observed that the 
main purpose kept in view throughout is to show the interpenetration of the theory 
of ionic sizes with the theories of the various phenomena of solutions, and to test the 
theory as far as possible in its relation to such phenomena. In this process various 
side avenues have been opened up, which we have forborne to follow if they carried 
us too far from the main track. 
One incidental matter of importance is the correction of the coefficient of ionisation 
which is afforded by the theory, according to which the true value of the coefficient 
of ionisation is 
\ Tj 
““ A' 1 + B/r* 3 ' 
The result of this correction is to make the values of a for KC1 and NaCl very 
nearly the same for equimolecular solutions, down to twice decinormal concentration. 
In confirmation of this it was found that the observed densities of mixtures of 
normal solutions of KC1 and NaCl were correctly given by the “law of mixtures” 
within very narrow limits. 
The conductivities of these mixtures were also determined and a mixture law for 
the conductivities based upon the consideration of viscosities and ionic sizes was 
formulated. The agreement thus obtained between observed and calculated values 
also tends to show that the hypothesis upon which the calculations were made is 
correct. 
The increase in the ionic volumes which takes place with increasing dilution must 
be nearly proportional to the increasing volume of water combined with the ion. The 
development of this matter quantitatively, so as to determine the number of molecules 
of water in combination with the ion under different circumstances, is a matter of 
great importance, but it cannot be dealt with in this paper. The effect of water 
