58 ME. CHAELES TOMLINSON ON SUPEESATXJEATED SALINE SOLUTIONS. 
contact. When, however, by a suclclen jerk the globules are flattened against the wall 
of the flask and form films, the solution instantly becomes solid. 
A body may also be said to act as a nucleus when the cohesion of its own particles is 
weaker than its affinity for one or other of the constituents of the solution. In a solid 
or in a liquid globule the cohesion of its own particles is greater than that affinity ; but 
in a film (which has extension only, length and breadth, the thickness being small 
enough to be disregarded) cohesion acts but laterally and not in every direction, and 
therefore the substance, coming into actual contact with the solution, yields to the 
attraction of one or other of the constituents of the solution for which it has the greater 
affinity, that attraction acting at right angles to the plane of the cohesive action of the 
film, and consequently meeting little or no resistance. Separation of the constituents of 
the solution then sets in. For example, the affinity in the case of a film of alcohol is for 
the water of the solution, and in the case of an oil-film for its salt, &c. 
But if the film have no preferential affinity for one or other of the constituents of 
the solution no separation results, as when glycerine diffuses through the solution 7 . 
This would suggest that form and affinity were the elements in nuclear action. If a 
film of glass, for example, could be reduced to a sufficient state of tenuity, would it act 
as a nucleus 1 But at any rate it may be supposed that contact and affinity are the sole 
elements. 
It will be gathered from the foregoing details that the distinction between unclean 
and clean resolves itself into the condition oi filmy and globular in the case of liquids ; 
and when solids act as nuclei, they are more or less contaminated with films of matter 
foreign to their composition. 
Although in general the surface-tension of those liquids that form lenses, instead of 
films, on the surface of the solutions may be less than that of the solutions, yet the 
viscosity of the solutions retards their spreading. But if a globule of oil, such as castor- 
oil, instead of being allowed to drop from the end of the glass rod upon the surface, be 
delivered to it so that it does not for an instant become an independent globule, it 
may spread on the surface and form the usual cohesion-figure ; under such a condition 
it instantly acts as a nucleus, and the solution becomes solid. 
But not only the viscosity of the oil, but a similar property on the part of supersatu- 
rated saline solutions interferes with those strict relations between the surface-tensions of 
liquids which several distinguished physicists are just now endeavouring to establish. 
The surface-viscosity greatly diminishes when the supersaturated saline solutions have 
been kept from twelve hours to a couple of days, and the more watery parts seem to 
come up to the surface, so much so that I had long supposed that horizontal sections of 
the solution contained in a wide tube 10 or 12 inches in length were unequally rich in 
salt ; but on drawing off portions of the solution from various parts into tared cruci- 
7 It lias often been stated that a solution of Glauber’s salt acts as a nucleus to a supersaturated solution of 
tbe same salt; but I have shown (Chemical News, February 4, 1870) that if the conditions of chemical purity 
be attended to, the solution merely diffuses through the supersaturated solution, without acting as a nucleus. 
