ON THE THEORY OF SOLUTION. g20 
that the difference here noted does not depend upon whether these binary 
compounds are electrolytes or not. 
In the third place there is a great deal of evidence that the molecular 
refraction of a substance is the same whether it be deduced from its solu- 
tion in alcohol, ether, benzene, bisulphide of carbon, or any other solvent 
that does not act chemically upon it. The same rule applies in some 
instances to solution in water; thus the molecular refraction of ammonia 
in alcohol, or in different quantities of water, was found to be about 8-96. 
The value for gaseous ammonia, as deduced from Dulong’s observations, 
is 8°60. 
A notable exception is hydrochloric acid. Very early in the history 
of refraction equivalents it was recognised that this acid in aqueous 
solution gave a value much larger than the gas itself, or than what would 
be obtained by adding together the values for chlorine and hydrogen in 
combination, as deduced from other sources. Dr. Perkin found a similar 
great increase of magnetic rotation in an aqueous solution of hydrochloric 
acid, but on dissolving the gas in isoamyloxide and examining the solution 
he found it rotated the plane of polarisation to very little more than the 
theoretical amount. The speaker therefore determined the refraction of 
this solution, and found the hydrochloric acid in it to have practically the 
theoretical value. 
HCl, theoretical value = : F “ : « 12 or 11:3 
HCl, in water . : F “ A . - . about 14-4 
HCI, in isoamyloxide . - . . . - . 11:36 
It would not be safe to use this increase of refraction of hydrochloric 
acid in aqueous solution as an evidence either of dissociation or of the 
formation of a hydrate. For the sum of the molecular refractions of free 
hydrogen and free chlorine, as determined by Dulong or Mascart, would 
be only 10°3, rather less than the theoretical, instead of more, as might be 
expected on the dissociation hypothesis,! while, on the other hand, the 
addition value of H,O in recognised hydrates (such as crystallised alums) 
seems to be the same as that of pure water, namely, 5°93. 
The general inference drawn by the speaker from the accumulated 
evidence was that the old conclusion is substantially correct; that mole- 
cular refraction and dispersion may be safely deduced from substances in 
solution where the solvent is chemically inactive, but that in the case of 
water there is some profound change effected upon the constitution of 
hydracids, haloid salts, and probably some other compounds by the act of 
solution. What this change may be cannot at present be inferred from 
optical analysis. 
Dr. James WALKER read the following translation of a communication 
from Dr. ARRHENIUS :— 
‘In the “Journ. Chem. Soc.” for 1890, p. 355, Mr. Pickering writes :— 
“Tt is indeed surprising that van ’t Hoff, Arrhenius, and others should 
not have recognised that every known deviation from the so-called normal 
depression, when induced by increase of strength of the solution, is in 
exactly the opposite direction to that which it should be if the law of 
‘ These numbers would have been brought more closely together if the calculation 
had been made by means of Lorenz’s formula “ ae x instead of the simpler oat 
+ 
With liquids and solids it is practically unimportant which formula is employed. 
x 2 
