290 
Proceedings of the Royal Society of Edinburgh. [Sess. 
From this comparison it will be seen that there is good agreement 
between the two sets of values. The accuracy of Tammanu's figures 
depends largely on the accuracy of the density observations which he 
used ; as these latter were taken from various sources they differ consider- 
ably. This can be seen, for example, in the case of NaOH above, where 
the values separated by the dotted line are from different sources. Again, 
it should be mentioned that the results quoted are in general mean 
values for four or five different temperatures ranging from 5° C. to 30° C. 
I referred at the outset of this paper to a list of compressibilities by 
Rontgen and Schneider, and also to the fact that the pressures they worked 
with never exceeded a few atmospheres. Tammann tested his formula by 
their results, and showed that calculation and observation agreed to within 
about 3 per cent. ; this is the same degree of accuracy as I find to hold at 
higher pressures also. 
Rontgen and Schneider’s table of compressibilities is interesting also 
from another point of view. It is indicated in their paper how, from their 
observations on compressibility, it is possible to calculate what they call 
the relative molecular compressibility. The numbers representing the 
relative molecular compressibility tell us how much the volume of a certain 
mass of water changes when a given number of water molecules are replaced 
by the same number of molecules of a dissolved substance. Collecting their 
values, we have the following table : — 
Relative Molecular Compressibility. 
H. 
NH 4 V 
K. 
Li. 
Na. 
I 
•960 
•913 
•918 
•892 
NO s 
•980 
•953 
•901 
•893 
•878 
Br 
•972 
•951 
■894 
•887 
•870 
Cl 
•954 
•933 
•872 
•868 
•849 
OH 
1-000 
1-009 
•779 
•782 
•761 
S0 4 
•942 
•808 
•682 
co 3 
•669 
•660 
•644 
If, then, it is a general law that the compressibility is inversely pro- 
portional to the internal pressure, we would expect to get a similar table 
for the contraction on solution, since the greater the contraction the greater 
the internal pressure. To test this I have chosen the quantity which 
Kohlrausch calls the molecular volume of the body in solution. The greater 
this molecular volume is, the smaller is the contraction on solution. The 
molecular volume is given by 
Q 
1000 — — where At 
m 
is the equivalent 
