66 ME. CHARLES TOMLINSON ON SUPERSATURATED SALINE SOLUTIONS. 
many solutions of the same, and of various degrees of supersaturation. There is also a 
difference in the mode of action of a speck, compared with that of a film. A speck or 
mote or point acts by determining the crystallization downwards at right angles to the 
surface, or radially, in closely packed crystalline lines ; whereas in the case of a pure 
film crystallization proceeds from every part of its lower surface simultaneously, so as 
often to produce large well-shaped crystals parallel with the film. 
(3) Why should a Jilin act as a nucleus , and a lens of the same oil not so act? — It can 
be shown experimentally that a clean inactive rod immediately becomes active simply 
by drawing it through the hand ; so also a clean surface, such as the inside of a flask, 
smeared with a little oil becomes powerfully nuclear ; but if the smear on the surface 
of the rod or on the inner wall of the flask can gather itself up into globules, these are 
not nuclear. Phenomena of this kind seem to be explained by reference to the surface- 
tension of liquids. In the case of a film formed by depositing a drop of an oil on the 
surface of a supersaturated saline solution, the surface-tension of such solution being 
diminished at the point touched by the oil, the surface of the solution surrounding the 
point touched exerts a tractive force on the oil, and spreads it into a film, with adhesion 
to the surface and a nuclear action on the solution ; because being brought into contact 
with the saline particles, but not with the aqueous (seeing that oil and water do not 
combine, while adhesion readily takes place between oil and salt), the saline molecules 
are separated from the aqueous, and the action once begun is propagated. When, 
on the other hand, an oil deposited on the surface of the solution assumes the form 
of a well-shaped double convex lens, or even a disk, there is no adhesion properly so 
called, and no nuclear action, since the lens is separated from the solution by surface- 
tension, and is not nuclear because it is really not in contact with the solution. Even 
when the solution is shaken so as to break up the lens into globules and to submerge 
them, each globule is as completely separated by surface-tension from contact with the 
solution as if it were outside the flask, at least so far as any nuclear action is concerned. 
Whether a drop of oil form a film or a lens depends on several variable circumstances, 
such as the temperature of the solution, its degree of supersaturation, its surface visco- 
sity, and so on. 
(4) Is it possible to deposit a film on a supersaturated saline solution without its acting 
as a nucleus ? — In a recently published account of a series of experiments, it is stated 
that an oil, such as citronella, was dissolved in ether (approximately 1 part oil to 20 of 
ether), and a drop of this solution was allowed to fall upon a supersaturated solution of 
Glauber’s salt ; as the ether evaporated, a film of the oil was left on the surface of the 
solution, and it did not act as a nucleus. Iu repeating this experiment, I found that in 
most cases the ether itself, or its vapour acting at a distance, formed a film on the sur- 
face of the solution, and produced crystallization ; but when this effect was not brought 
about, and the ether evaporated, the oil was left in the form of a multitude of minute 
globules or lenses, which, as already stated, are not nuclear. But granting the result 
stated above to have been obtained, and that a film was actually deposited on the sur- 
