Super solubility from the Osmotic Standpoint. 263 



property of surface tension ; the circumstances determining 

 this are not yet ascertained, but we can assert with con- 

 siderable plausibility, that a separating surface is not possible 

 until the number of molecules contained in the incipient 

 globule is large. 



Fig. 5. 



/ — 



f 



*-&•/>-% 



MEMBRANE 



(17) It is necessary to consider the question of the con- 

 centration inside the globules at some length — for, although 

 we are about to deal with unstable conditions, yet some light 

 may be thrown on the cause of the difference between 

 solutions of solids in liquids and solutions of imperfectly 

 miscible liquids. It might be thought that as a solution is 

 only a collection of two sets of molecules, these in virtue of 

 their mobilities will, given time enough, have all possible 

 relative concentrations at some point or other in the mass, 

 and the separation into two phases occurs when the concen- 

 tration has reached, at one of these points, a value represented 

 by the ordinate through F of fig. 4. It can readily be shown 

 that this explanation cannot be a general one ; for take the 

 case of benzene and water, whose mutual solubility is very 

 small, and suppose we have a litre of a homogeneous saturated 

 solution of benzene in water, and calculate, on ordinary 

 probability data (preferential molecular attractions are 

 excluded), the time it would require for the concentration at 

 any point in the mass to be that at F (i.e. one molecule of 

 water to 6000 molecules of benzene), we shall find that 

 10 11000 years would have to elapse before this state were 

 reached ! 



A more plausible hypothesis is the following : assume 

 that at some point or other in the mixture, which is about to 

 separate, a few molecules of the pure solute come together 

 by hazard (on probability data this is quite possible) and that 

 these have a preferential attraction for their own species ; it 

 is easy to see that growth would continue until separation 

 ensued. The main objection to this explanation is that dif- 

 fusion (a direct function, in this case, of a nearly infinite 

 osmotic pressure ; point M of figure) must be acting all the 

 time against the further growth of the aggregate *. 



* This hypothesis may be strengthened by assuming that the pre- 

 ferential attraction of the pure solute molecules ceases when a certain limit 

 is reached r the aggregate then taking in, by diffusion, solvent molecules 

 ation re quired by point F of diagram. 



