64 SCIENCE PROGRESS 



the view that a single network may not account for the elastic 

 properties of different gels. 



As was pointed out by Hatschek [19 14] the lens-shaped 

 form of gas-bubbles occurring in gels must have some relation 

 to their structure. He found that the bubbles place themselves 

 at right angles to pressure applied to the gel and parallel to 

 tension. By measuring the angles between pairs of bubbles 

 in 10 per cent, gelatin Hatschek found certain values approxi- 

 mately repeated a number of times, although other measure- 

 ments varied enormously, so that whether the angles have much 

 to do with the crystalline structure is doubtful. However, planes 

 of cleavage are suggested and the simplest structure that would 

 conform approximately to the conditions is that of piled shot 

 or a brick-stack. In this relation it is remarkable that ultra- 

 microscopic examination of gelatin gels of from 1 to 6 per cent., 

 from which the liquid has been expressed, shows flocks of 

 separate grains of irregular shape with clear spaces separating 

 the flocks. Increasing concentration reduces the size of the 

 grains, as would be required by v. Weimarn's theory, and, at 

 the same time, the empty spaces gradually disappear, until, 

 at about 6 per cent, the solid phase fills the whole field and the 

 single grains can no longer be differentiated. Agar and silicic 

 acid show a similar globulitic structure. [Bachmann, 191 1; 

 Zsigmondy and Bachmann, 191 2.] 



During experiments on banded precipitates in gels the slow 

 motion of solid particles through gels was frequently observed. 

 Moreover, many precipitates appeared gradually to settle 

 down, leaving the gel apparently unchanged above. This 

 suggests rather that the individual gel particles are easily 

 separated than that they are joined fast in a network. 



It may be noted in this connection that the solid particles 



deposited by the chemical reaction of solutions diffusing into 



gels are frequently spherical in shape, resembling natural 



spherulites. Their form may be attributed to the adsorption, 



by the granules, of the reaction components from the diffusing 



p 

 solutions so that the value of v. Weimarn's j- at the surface of 



the granules is greater than in the surrounding gel, and, owing 

 to the diffusion of further supplies of nutrient matter, becomes 

 larger than the original number of condensation centres. The 

 grains are therefore unable to grow as regular crystals and 



