SOLUBILITY OF SALTS IN WATER AT HIGH TEMPERATURES. 25 
distillation of water from one part to another will take place. A paraffin bath 
appeared to offer greater probability of uniformity than an air bath, and was con¬ 
sequently employed in most of the best experiments.* 
3. Further, it is obvious that the part of the tube not occupied by salt and 
solution must be filled with vapour of water, which, on cooling, will condense to the 
liquid state, and mix partly with the undissolved residue and partly with the 
solidified solution which has been drained away. Unfortunately, too, this circumstance 
is complicated by the fact that the tension of such water vapour is not that which 
would be given by pure water, but is the smaller tension given by the saline solution 
contained within the tube. 
By ascertaining the volume of this vapour, approximate correction of the results 
can be effected with the aid of the tables of vapour tension of salt solutions, published 
by Wullner (Poggend. Ann., ciii., 529 ; cx., 564). 
In any case the error on this account could not be great, but we nevertheless 
thought it advisable to make a few experiments with the object of testing directly 
its probable magnitude. 
A number of bent glass tubes containing crystals of sodium sulphate, Na. 2 SO 4 .10HoO, 
in one limb, were sealed up and heated in a paraffin bath to temperatures ranging 
from 115° to 150°, the experiment being conducted in all respects as if a solubility 
were to be determined, except that the solution was not decanted. They were then 
allowed to cool, cut in two at the bend, and the end remote from the salt was at 
once closed by a stopper and weighed. It was then dried and re-weighed. Sub¬ 
sequently its capacity was ascertained. The following were the results :— 
Experiment. 
Weight of Water. 
Difference. 
No. 
Calculated. 
Found. 
grm. 
grm. 
grm. 
1 
•0103 
•0153 
+ •0050 
2 
•0102 
•0116 
+ •0014 
3 
•0141 
•0129 
-•0012 
4 
•0067 
•0061 
-•0006 
5 
•,0151 
•0137 
-•0014 
6 
•0106 
•0173 
+ •0067 
As the quantity of solution dealt with was never less than 4 grms., and was 
generally about 8 grins., it is obvious that the greatest error to be expected on 
account of occasional slight distillation from inequality of temperature, or on account 
of the water vapour always present, is inappreciable. When the metal tube is used 
the error is still less, because its greater mass and superior conducting power are 
favourable to the maintenance of a constant temperature. 
Attempts were made to employ a metal tube provided with valves of various kinds, 
* The bath was in fact carefully tested, and the constancy of its temperature ascertained. 
MDCCCLXXXIV. 
E 
