108 



Mr. W. B. Hardy. On the Mechanism of 



that brief heating to red heat destroyed the reabsorptive powers of the 

 gel of silica.* 



There is one binary system in which gelation is an irreversible 

 process {i.e., coagulation) which can be readily studied under the 

 microscope. The hydrosol is a ternary system composed of water, a 

 minute trace of free acid or alkali, and the modification of egg albumen 

 which is produced by heating it to 100°. Coagulation occurs when the 

 free acid or alkali is removed. As the coagulation point is neared the 

 proteid particles in the hydrosol increase in size, so that spheres 0-75 

 to 1/x in diameter are formed. These become arranged in rows which 

 anastomose so that an open net with regular polygonal meshes is 

 formed.! In this case the process of gel building is the same as 

 that which can be followed so easily in ternary mixtures, and in both 

 cases a definite surface separates the phases. It is probable that the 

 hydrogel of silica is formed in the same way, since Picton and Linder 

 have shown by optical tests that, as the point of coagulation is 

 approached, larger and larger particles of silica form in the hydro- 

 sol. J These particles may be solid solutions of -water in silica, or 

 they may-be large molecular aggregates of silica free from water. I 

 incline to think that the latter is the more probable assumption, since^ 

 if they were solutions, it is difficult to see why the process should be 

 irreversible. 



In the case of the reversible systems agar-water, or gelatine-water- 

 alcohol, the particles seem to be of the nature of solid solutions. 



The system agar and water consists of two components, and, 

 therefore, a nonvariant system should be defined by four coexistent 

 phases. Since the gel stage consists of three phases, namely two 

 solutions and a vapour phase, it should be a monovariant§ system. That 

 is to say, the composition of the phases should be fixed by fixing 

 either the temperature or the pressure of the vapour phase. The 

 experiments show that this is not the case. The composition of the 

 fluid and solid phases is not constant for a given temperature. This 

 result might be regarded as being due to the passive resistance to 

 change in the system which is introduced by the formation of a soHd 

 phase. On this view if the velocity of the reaction were known, the 

 phases would be fixed if the element of time were introduced and 

 accorded a definite finite value. This is the method which Bancroft sug- 

 gests for dealing with such cases ;1| it is, however, possible that there are 



* ' Zeits. f. Anorg. Chem.,' vol. 13, 1896, p. 289. 



f The process is described in detail in an earlier paper by the author in the 

 * Journal of Physiology,' vol. 24, 1899, p. 182, and the information which the 

 microscope affords as to the manner in which irreversible gels are built is discussed 

 there. 



:J: ♦ Journal of the Chemical Society,' vol. 61, 1892, p. 148. 

 § That is to say, a system having one degree of freedom. 

 11 ' The Phase Rule,' p. 234. 



