66 SCIENCE PROGRESS 



p 



surface. From his formula, N — Ky, a sol or a gel results 



according as a large number, N, of crystallisation centres are 

 produced from a large excess concentration, P(= C — L), or a 

 small solubility L. Intermediate values of P and L give rise to 

 micro or macro crystals. If his formula applies to substances 

 for which this intermediate stage vanishes, by reason of their 

 slow velocity of crystallisation, the transition point between the 

 sol and gel stages should correspond to the act of crystallisation. 

 From the general behaviour of sols, increase of concentration 

 beyond a limiting value does lead to precipitation. Since, in 

 the case of the lyophile sols, there is considerable affinity 

 between the two phases, a temperature variation is indicated, 

 as is confirmed by Hardy's figures for the liquid expressed from 

 agar. And the transition or gelation point should in general be 

 a function of the concentration and temperature of the sol. 

 The influence of concentration is well exemplified by mastic. 

 Dilute alcoholic solutions of mastic poured into water produce 

 sols, while a greater concentration of mastic gives fairly per- 

 manent gels. Whether or not the precipitate forms a coherent 

 gel or merely sinks to the base of the vessel would depend on its 

 mass and degree of dispersity. If the precipitate is sufficiently 

 voluminous it would drink up the liquid and form a coherent 

 gel. 



The gels of camphorylphenylthiosemicarbazide investigated 

 by Hatschek [19 12] led him to suspect a crystalline structure 

 formed by growth of amicrons into a network : a conclusion 

 which he hesitated to accept on account of the mathematical 

 difficulties. These gels appear to be typically elastic, and are 

 produced (1) when an up to about 5 per cent, hot alcoholic 

 solution of the substance is suddenly cooled, (2) from the gradual 

 or rapid cooling of toluene solutions of similar concentration, 

 or (3) by pouring a concentrated alcoholic solution into petro- 

 leum. The latter method gives gels with a concentration as 

 small as 1 : 400. The solubilities in alcohol, toluene and petro- 

 leum are respectively 0*5 per cent., 1 per cent, and very small. 

 P. P. v. Weimarn's theory easily supplies an explanation of 

 the first and third cases, though the second is more difficult 

 to understand, for the slow cooling of an alcoholic solution leads 

 to the production of macrocrystals, though the substance is less 

 soluble in alcohol than in toluene. The explanation may lie 



