264: The Earl of Berkeley on Solubility and 



(18) I would suggest the following explanation as involving 

 the least assumptions. No solution is truly homogeneous, 

 consequently, in a liquid at the limit of supersaturation, a 

 small departure from the mean concentration in the direction 

 of the other phase will surely take place. This new aggregate 

 of molecules, having dPi/dc 2 or ~dV 2 l~dci negative as the case 

 may be, will, by paragraph (4), go on increasing in concen- 

 tration until a new state of equilibrium is reached. This 

 aggregate, or incipient globule as we may call it, will 

 itself be unstable and will tend to throw out the other 

 phase, thus contributing to the separation into two phases. 

 The conditions being unstable it is impossible to follow the 

 events taking place in the globule, the more so as when the 

 property of surface tension is acquired there will be pressure* 

 and small concomitant temperature changes f involved. But, 

 although at first sight equilibrium should be reached instan- 

 taneously, yet, as the supply of molecules for further growth 

 can only be obtained through the relatively slow process of 

 diffusion, we can state that a finite time will be required 

 before the final concentration is reached. 



During this interval ol time the globules will tend to one 

 or the other of the two final states of equilibrium which are 

 represented by equations (1) and (3) respectively. In the 

 first case the separating phase is a pure solid, in the second 

 it is a liquid mixture. 



(19) An example will show this more clearly : suppose we 

 have a saturated solution of water in ether, under constant 

 pressure as in fig. 5, with the solvent, pure ether, in the right- 

 hand compartment. Assume the temperature such that on 

 the withdrawal of a small quantity of the ether, ice is formed 

 — equation (1) has been operative on the contents of the 

 globules. If, however, the temperature had been very 

 slightly higher, the liquid phase, water saturated with ether, 

 would have come out, and in that case equation (3) was 

 operative. It is evident, had we assumed an unsaturated 

 solution of water in ether to start with, the withdrawal of 

 ether, at the right temperature, could only lead to the for- 

 mation of ice. Here again I would suggest that the first 

 change which takes place is the formation of globules richer 

 in ether, which having ftP/fcc negative are unstable, and tend, 

 under the influence of their own internal pressure, to the only 

 equilibrium possible, namely that represented by equation (1). 



* A droplet of water of - 000f5 cm. diameter has an internal pressure, 

 due to surface forces, of 6*4 atmos : it still contains 10 8 molecules. 



t Owing to the small size of the globules these temperature changes 

 will be negligibly small. 



