318 REPORT—1890. 
From the instances above mentioned some answer may be obtained to 
the third question, whether the molecular depression is independent of 
the nature of the dissolved substance. The values obtained with these 
four substances, taking solutions of a strength corresponding to that of 
their gases, are :— 
Calcium chloride . a A 5 5 A 5 i - 2° 850 
Calcium nitrate . : A : = ; 4 c oe OTE 
Sulphuric acid . ‘ : - : ; 3 o) 2 vol 
Alcohol - ‘ ; 5 F ¢ : ‘ ‘ . 2180 
a variation of 30 per cent., which must give an emphatic denial to the 
idea of absolute constancy; and if we take instances from other sub- 
stances, where the data available refer to solutions of somewhat greater 
strength, we find that the very substances on which the idea of constancy 
was originally founded show variations reaching 60 per cent. (‘ Phil. 
Mag.’ 1890, vol. i. p. 492), while in other cases, which I have quoted else- 
where (loc. cit. p. 493),’ the variation attains the still larger dimensions 
of 260 per cent. 
To every one, therefore, of the three test questions as to constancy and 
regularity, the experimental results give an unhesitating negative. 
In the instances quoted above the depression actually found for alcohol 
has been doubled in order to simplify the comparison of it with the 
other substances. Alcohol belongs to that class of bodies which give just 
half the value in water that the majority do, and of which there are some 
instances in the case of every solvent yet examined. The explanations 
which the supporters of the chemical and physical theories give of these 
half values differ so radically from each other that it is hopeless to attempt 
to arrive at any agreement as to the nature of solution till this difference 
is settled. The chemists say that these half values are in all cases the 
abnormal ones, just as Raoult did originally, and explain them by repre- 
senting the molecules of the dissolved substances which give them to 
consist of two fundamental molecules. The physicists give exactly the 
same explanation in the case of every solvent except water, but in this 
case they say that the smaller values are the normal ones, and the larger 
the abnormal, the double magnitude of these being caused by the disso- 
ciation of the dissolved molecule into its two ions, whereby two mole- 
cules or acting units are formed from every one originally added. 
If Raoult’s views as to the consistency of the molecular depression 
can be maintained, the data themselves are conclusive against making 
this exception in the case of water ; for, since the substances which give 
the lower values are supposed to act normally, it is evident that, if the 
values given are in any way abnormal, this abnormality must be due to 
the solvent. Now the values certainly are abnormal ; they are about 
1°-03, whereas the normal value for one molecule dissolved in 100 mole- 
cules of other solvents is 0°63, and the excess can, therefore, only be 
explained by assuming that the molecules of water are more complex than 
those of other solvents in the proportion of 1:03 to 0:63, or 1} to 1; in 
other words, the water molecules must be 1}H,O. This view cannot be — 
reconciled with the atomic theory. 
Indeed the theory of dissociation into ions is altogether unintelligible 
to the majority of chemists. It seems to be quite irreconcilable with our 
1 The depression produced by H,O in 100H,SO, is 1°-07 instead of 0°-07 as there 
given. 
