GELATION OF THE NATURAL EMULSOIDS 65 



develop as starlike growth figures, globular aggregates, or as 



spherical crystals [Bradford, 191 7]. It is evident that the 



slower the velocity of crystallisation of a particle the more 



p 

 easily will the value of y- increase in this way. Now in the 



case of large unwieldly molecules, like those of the natural 

 emulsoids, the directive force of crystallisation may very well 

 be small, and for this reason, coupled with the inertia of the 

 molecules and the viscosity of the medium, fresh molecules might 

 take a considerable time to fit themselves into the crystal 

 structure. Consequently molecules or aggregates might arrive 

 faster than needed for the regular growth of crystal faces 

 and result in the production of globular crystals, or even in 

 extreme cases of aggregates piled up without regard to orienta- 

 tion. Perrin's large-grained emulsions of mastic and gamboge 

 exhibit this spherical shape clearly, and Moore observed chains 

 of globules in dilute silicic acid [1915]. 



On the basis of such a formation it is perhaps a little difficult 

 to understand the coherence of gels and to explain exactly their 

 stress-elongation curves. The coherence is to some extent 

 affected by the films which form on the surface during setting. 

 And there may be amicroscopic filaments which come into play. 

 The existence of an elastic surface layer around the disperse 

 particles might also contribute to the effect. It will also be 

 remembered that the liquid expressed from a 5 per cent, gelatin 

 gel contains from o'i6 to 0*46 per cent, gelatin [O. Biitschli, 

 1896], while that from a 2*23 per cent, agar gel varies from 

 0*09 per cent, at 5 to 0*47 per cent, at 36 [Hardy, 1900]. 

 The viscosity of this liquid would be appreciably greater than 

 that of water. 



The above figures show that the concentration of the liquid 

 in equilibrium with the solid phase varies with the temperature. 

 And, since at a few degrees higher temperature, the whole mass 

 was fluid it is difficult to resist the conclusion that the excess has 

 a crystallised cut on cooling. This is in accordance with the 

 influence of concentration on the temperature of gelation [Rona, 

 1902 ; Rohloff, 1907 ; Levites, 1908]. 



A similar inference may be made from another point of 



view. P. P. v. Weimarn's researches show how to prepare a 



given substance in any desired degree of dispersity, and indicate 



that the disperse particles are crystalline however great the specific 



5 



