Supersaturation of Salt-Solutions. 457 
hundred water-molecules; this last is the strongest solution 
that can be conveniently worked with. In the case of both 
salts the molecular volume constantly increases with the con- 
centration, very rapidly at first, and then more slowly, just as it 
does in the case of an ordinary solution of an anhydrous salt, 
such as potassium nitrate. The chief point of importance is 
the absence of any abrupt break above the ordinary satura- 
tion-point ; there is no discontinuity, the steady increase is 
unbroken ; there is nothing in the numbers for the molecular 
volumes to indicate a difference in the constitution of the 
solution. I ought to add that, owing to the unavoidable loss 
of water during the heating necessary for the solution of the 
salt, it is not to be expected that the density determinations 
should be as accurate as those of ordinary solutions; not 
much reliance can be placed, I fear, on the fourth figure ; not 
by any means so much as on the fifth figure in other cases: the 
effect of this is to affect the molecular volume in the first 
decimal to the extent of +5. 
It would seem, then, that these solutions were simply solu- 
tions of the anhydrous salt which are yet unsaturated. I 
have already expressed my conviction that a salt in sclution 
parts with its water of crystallization *, and attempted to 
prove it by experiment. If this were correct, and the super- 
saturated solution such as I conceived it to be, it followed 
that anhydrous sodium sulphate should dissolve in a so-called 
supersaturated solution ; in other words, a solution saturated 
by contact with crystals of a hydrated salt should be able to 
dissolve that salt when dehydrated. 
To test this a wide-mouthed bottle was nearly filled with 
erystals of Na,S,O35 aq., and placed in boiling water. The 
quantity of salt was nearly 60 grammes ; a small, thin glass 
bulb containing less than a gramme of the dehydrated salt, 
and sealed, was also placed in the bottle. When the thiosul- 
phate was completely fused, the stopper was put in, and, after 
cooling slightly, the bottle was placed in water at about 
15°C. Thiosulphate of sodium fuses at 48° to 50° C., yet 
the solution thus obtained was perfectly stable ; no crystalli- 
zation was caused by gentle shaking. By a smart blow of 
the bottle on the knee the small bulb was broken, and the 
dehydrated salt brought in contact with the solution; no 
crystallization took place; on the contrary, solution was com- 
plete ina few seconds. The experiment was repeated with 
from 4 to 4:5 grms. of Na,8,0, to 40 grms. of Na,8,Oz 5 aq. 
In this case solution was not complete at the ordinary tem- 
perature, but was readily effected by heat. When the solu- 
* Phil, Mag. Sept. 1884, p. 181. 
