152 
ME. W. K. BOUSFIELD: IONIC SIZE IN RELATION TO 
Table XXIX.—KC1. 
m. 
log h. 
r 3 
'k • 
a* 3 
'Cl • 
I». 
A n/m 
observed. 
A n/m 
calculated. 
_ 
Difference. 
0-03 
3-26629 
3-474 
3-331 
6-805 
0-0104 
0-0104 
+ 
0-1 
2-74257 
3-222 
3-065 
6-287 
0-01026 
0-01030 
-4 
0-311 
2-24734 
2-775 
2-601 
5-376 
0-01013 
0-01014 
-1 
0-5 
2-03874 
2-508 
2-328 
4-836 
0-01005 
0-01004 
+ 1 
1-234 
1-63710 
1-866 
1-687 
3-553 
0-00980 
0-00980 
± 
Table XXX.—NaCl. 
m. 
log h. 
n* 3 
'Na • 
r r 3 
; ci • 
I». 
A n/m 
observed. 
A n/m 
calculated. 
Difference. 
0-02 
3-44259 
11-798 
3-385 
15-18 
0-01049 
0-01048 
+ 1 
0-03 
3-26646 
11-694 
3-331 
15-03 
0-01046 
0-01046 
+ 
0-05 
3-04444 
11-532 
3-240 
14-77 
0-01043 
0-01043 
+ 
o-i 
2-74304 
11-211 
3-066 
14-28 
0-01035 
0-01036 
- 1 
0-2 
2-44131 
10-725 
2-813 
13-54 
0-01026 
0-01027 
- 1 
0 -25 
2-34404 
10-521 
2-712 
13-23 
0-01023 
0-01023 
+ 
0-3 
2-26451 
10-336 
2-622 
12-96 
0-01020 
0-01020 
+ 
0-5 
2-04105 
9-711 
2-331 
12-04 
0-01008 
0-01008 
+ 
0-75 
1-86290 
9-089 
2-062 
11-15 
0-00994 
0-00997 
-3 
1-0 
1-73590 
8-578 
1-852 
10-43 
0-00982 
0-00988 
-6 
Table XXX. shows that the relation in the case of NaCl down to half normal 
solutions is particularly accurate. 
In Part IV. ( b ) it was shown that the ionic volume might be calculated from the 
radions of the separate ions with fair accuracy down to nearly normal solutions. It 
is obvious, however, that, as the proportion of un-ionised solute becomes greater, it 
becomes more and more inaccurate to take the ionic volume of a molecule as the sum 
of the volumes of a pair of ions. It is probable that the divergence of the calculated 
values in Table XXX. for the last two observations tabulated is due to this error in 
the calculation of the ionic volumes. Now, as in the case of freezing-point depression, 
so here, having shown previously that solution volume is a linear function of ionic 
volume, we deduce that the equivalent refractive power, which is shown above to be 
a linear function of ionic volume, must therefore also be a linear function of solution 
volume. The series of values given by Dinkhauser for KC1 is not sufficiently good 
to make this comparison useful. But the series for NaCl shows clearly the remark¬ 
able fact that the equivalent refractive power is accurately a linear function of the 
solution volume right through the range of dilute and concentrated solutions. 
In Table XXXI. are set out the solution volumes, calculated from the density data 
