474 Proceedings of Royal Society of Edinburgh. [sess. 
using dry crystals of sodium thiosulphate (lSra 2 S 2 0 3 ,5H 2 0) ; also, 
that supersaturated solutions can be easily prepared in the cold, 
and therefore that supersaturation is not a property conferred on 
the solution by heat, as many supposed; and, further, that, by 
evaporation at the ordinary temperature, supersaturated solutions 
yielded crystals, not of the ordinary hydrate, but of a lower one 
containing less water. Other experiments — for an account of 
which reference must be made to the original papers — showed con- 
clusively that the physical properties of a solution underwent no 
abrupt change when it passed into the region of supersaturation, 
either by fall of temperature or by increase in its concentration. 
The question whether supersaturation ever occurs in the solutions 
of salts which crystallise without water is generally regarded as 
undecided, while the many well-marked examples to be found 
amongst organic compounds are passed over in silence, inasmuch 
as the conditions preclude the possibility of the formation of 
hydrates or compounds analogous to them, and consequently an 
explanation which is fairly satisfactory in the case of hydrated 
salts is here inapplicable. 
Shortly after the publication of the experiments on supersatura- 
tion referred to above, my attention was drawn to the work of 
Frankenheim* on the crystalline form of potassium nitrate. I had 
found in the course of other work that this salt was abnormally 
soluble in solutions of sodium nitrate, and it struck me that this 
was probably due to the solubility of the unusual, hexagonal modi- 
fication of potassium nitrate isomorphous with sodium nitrate 
being greater than that of the usual rhombic modification. I there- 
fore repeated Frankenheim’s experiment : a small drop of strong 
solution of potassium nitrate was placed under a low-power micro- 
scope, and allowed to evaporate spontaneously at the ordinary 
temperature ; as evaporation proceeded, rhombohedral crystals 
made their appearance, and in some cases continued to grow in 
size and number until the water had entirely evaporated. In 
other cases, more usually, rhombic prisms put in an appearance 
after a time, growing with great rapidity, thus indicating super- 
saturation. It was further found that if a drop from which a 
* Pogg. Ann., 1837, 40, p. 447 ; 1854, 92, p. 354, 
