262 The Earl of Berkeley on Solubility and 



BPi/dc 2 = 0, we have also BP 2 /B^i = 0, hence the horizontal 

 tangents to the conjugate have their-foints of contact on the 

 same ordinates as K and L. The fact that we have to pro- 

 long the curves through the limits of solubility shows that 

 supersolubility is always theoretically possible ; and I would 

 suggest that the ordinates through the points of contact 

 of the horizontal tangents represent the absolute limit 

 of super saturation, for after these ordinates are passed 

 ^Pj/^c 2 is negative, and by paragraph (4) instability must 

 set in *. 



(16) Referring to the mechanism by which a liquid mixture 

 separates into two liquid phases. There are three possible 

 ways of bringing this about, and, in all three, changes in 

 osmotic pressure are involved. 



(a) By changing the temperature of the liquid. In all 



cases examined the first appearance of the new phase 

 is heralded by a cloudiness throughout the liquid. 

 This cloud is supposed to be due to minute globules of 

 the new phase f. 



(b) By altering the pressure on the liquid. I am not 



aware of any record dealing with the first appearance 

 of the new phase under these conditions. Doubtless 

 a cloudiness is also apparent. 



(c) By altering the concentration. This method is analo- 



gous to evaporating off one of the components at 

 constant temperature ; and it is realizable in the 

 system shown in fig. 5 by decreasing the pressure q 1 

 and thus causing solvent to flow from left to right. 

 Here again no experimental evidence of the formation 

 of globules is available, but judging by the evaporation 

 analogy we are justified in assuming that they would 

 appear. 



We will discuss method (e) as it has the advantage, so far 

 as our diagram is concerned, of involving no change in the 

 temperature and pressure of the solutions. 



If we leave out of account what may be called the acci- 

 dental setting in of instability, the globules, on the hypothesis 

 postulated in paragraph (15), will appear only when the 

 mixture has reached the point K. Probably they can only 

 be visible when the separate entities have acquired the 



* Instability may set in before these points are reached ; for, apart 

 from the presence of " nuclei," there may be intermolecular forces whose 

 effect we are unable to forecast. 



t See Donnan, Ohem. News, vol. xx. p. 139 (1904) ; Rothmund, Zeit. 

 Phys. Chem. vol. lxiii. p. 76 (1908) ; von Lepkowski, Zeit. Phys. Chem. 

 vol. lxxv. p. 608. 



