150 
MESSRS. W. R. BOUSFIELD AND C. ELSPETH BOUSFIELD 
Comparing the values of n in Table XVII. with those in Table XVI., the values 
for solutions I., II. and III., are of course those from which the respective constants 
were derived. The values of n for solutions IV., V., VI. and VII. agree well with 
one another considering that for solution VII. we are in both cases extrapolating to 
a dilution ten times as great as that of solution III. The values of n for the four 
last solutions are also in fair agreement with the values calculated in the former 
paper (‘Trans. Chem. Soc.,’ vol. 105, p. 1821, 1914) from freezing points and vapour 
pressures. The value of n for h = 236 derived in the former paper from the freezing 
point value came out as 6'86. The present values for h = 221 at 20° C. are, as they 
should be, a little lower. 
These concordant results appear to justify the hypothesis that the specific heat 
lowering and the specific volume lowering of the free water may without serious 
error be taken as proportional to the percentage concentration, with the corollary 
that the specific heat and the specific volume of the liquid solute may be treated as 
constant for all dilutions, and also the specific heat and specific volume of the 
combined water with perhaps somewhat less accuracy. One is inclined to think that 
the slight divergence in the values of n in Tables XVI. and XVII. may be due to 
the diminished accuracy of the latter hypothesis. It may be noted that the ratio of 
the constant C for the specific heat lowering of the free water to the constant B for 
the specific volume lowering of the free water is C/B = 0'02/0'0008536 = 23'4, 
which is nearly equal to L. 
In a previous paper (‘Boy. Soc. Proc.,’ A, vol. 88, p. 167, 1913) the value of 
Er 0 for NaCl came out as about 21 c.c. instead of 25 c.c. But at that time no 
method had been devised for treating the variation of the constitution of the 
free water with concentration, and the calculation was made on the assumption 
that the specific volume of the free water was the same as that of pure water, or 
in other words the constant B was taken as zero. To obtain a closer approximation 
it was necessary to make some allowance for the variation of the specific volume 
of the free water with the varying concentration of the solute, as has been done 
above. 
The values of the various constants have been worked out for KC1 as well as for 
NaCl, and although the vapour pressure data and the resulting values of n for KC1 
are not as accurate as those for NaCl it may be well to note the comparative 
results. # 
* Too much stress must not be laid on these figures, since there still exists some uncertainty as 
to the exact values of n both for NaCl and for KC1. This would affect the values of all the constants 
except L. The value of L does not in any way depend on the value of E% nor the L relations at 
the end of Section 19. More exact vapour pressure data are required to settle the values of n. 
