1896 - 97 .] Mr W. W. J. Nicol on Super saturation. 
475 
number of rhombohedra had crystallised out, which was therefore 
undoubtedly saturated to that form, was stirred with a needle 
which had been cleaned by heating to redness, no crystallisation 
of the rhombic form was induced, but if the needle was touched 
with a rhombic crystal and again brought in contact with the 
solution, instant separation of the rhombic variety ensued, and the 
rhombohedra, when touched by the prismatic crystals, became 
opaque and developed prisms shooting out on all sides. In short, 
the solution exhibited all the phenomena of supersaturation. 
This may be taken as the crucial experiment, the one which 
gives us the key to all examples of supersaturation, enabling us to 
state the proximate cause of all such cases, as well as those of 
superfusion, as follows : — Whenever under the conditions of experi- 
ment two allotropic modifications of the dissolved, or fused substance 
can exist , then supersaturation or superfusion , as the case man be, 
is also possible. The term allotropic is here used in a somewhat 
wider sense than usual, as including not only amorphous and 
different crystalline forms of the same substance but also different 
crystalline forms due to the presence or absence of foreign mole- 
cules, such as water of crystallisation. The above is the funda- 
mental condition which must be fulfilled if supersaturation is to 
occur at all. Two minor conditions limit the range within which 
such supersaturation can exist : these are concentration and tem- 
perature of the solution, and are interdependent, though it does 
not follow that we are able to attain either limit. Thus, in the 
case of temperature, the lower limit may lie below the so-called 
cryohydrate point, where the conditions are altered by the exist- 
ence of new phases ; but for every concentration there is a critical 
temperature, and for every temperature a critical concentration, 
outside the limits of which the supersaturation breaks down, not 
by the separation of the less stable form but the formation of the 
usual stable modification. 
It is not my intention in this communication to enter fully into 
the experimental evidence on which the foregoing conclusions are 
based. I had intended to delay the publication until I had 
accumulated a mass of facts sufficient to place the conclusions 
derived from them beyond the possibility of doubt, but the appear- 
ance of a paper by Ostwald in the Zeitschrift fur physikalische 
