126 Mr. W. W. J. Nicol on the Molecular 
a constant quantity. The volumes, therefore, of the above 
elements and groups of elements are independent of the 
manner in which they may be combined together, provided 
only they be determined in solution in water and under the 
same conditions. Now such is not the case in the solid state. 
Taking the molecular volumes of the above salts, calculated by 
Schroder* from the most reliable determinations of the specific 
gravity of the solid salts, we have the following : — 
(K-Na)Cl =37-4-27-1 =10-3, 
CK-Na)N0 3 =48-5-37-6 =10-9, 
(K -Na)(^) = ( 32 - 8 ; 26 ' 7 ) = 3-l. 
K^Cl-^) =37-4-16-4=21, 
*(a-90 
27-1-13-3 = 13- 
K(N0 3 -C1) =48-5-37-4 = ll-l, 
Na(N0 3 -Cl) = 37-6-27-1 = 10-5. 
All attempts to extend the experiments of Kopp on liquids 
to the case of solid salts have, I believe, failed, or at the best 
been attended by only partial success, owing doubtless to the 
impossibility of obtaining these under such conditions that 
their molecular interspaces are even approximately compa- 
rable. This, however, is not the case with some isomorphous 
salts examined by Thorpe f and others ; but I shall return to 
this point later on. Meanwhile I believe that I am justified 
in concluding, from my experiments above, that, u'hen salts 
are dissolved in water, the molecular interspaces in various 
solutions are approximately coextensive. Hence it is possible 
to ascertain the relative molecular volumes of the salts them- 
selves with a degree of accuracy equal, at least, to that with 
which the molecular volumes of liquids have been determined. 
I will now consider the causes which lead to the variations 
in the difference between the molecular volumes of two salt- 
solutions, and which thus affect the accuracy of the above 
determinations. Putting the errors of experiment on one side, 
we have the main cause in the effect of concentration on the 
volumes of the solutions compared, or, what is the same thing 
(as I hope to be able to show), in the different solubilities of the 
two salts. Among the best-marked instances given above, we 
find two cases in which A increases with the concentration, 
* PoggendorfFs Annalen, cvi. p. 242. 
t Journal of Chemical Society, 1880, p. 102. 
