DIFFUSION IN DEFORMED GELS 93 



and the method of stressing is best described in the author's 

 own words : " Pour a bichromate gel into a cardboard tray 

 to a depth of about ^ in. Support a glass, or xylonite, plate 

 provided with a small central hole, in contact with the surface 

 of the gel until the latter has completely set. The transparent 

 plate must fit the tray exactly. When setting is complete, cut 

 away two opposite sides of the tray, supporting the other pair 

 by wooden blocks. Press the plate on to the surface of the 

 gel and apply weights. The gel is now free to move in one direc- 

 tion, but hindered in the other. It is therefore compressed 

 against the remaining sides of the tray. Now place a drop of 

 strong silver nitrate over the central hole and leave in a damp 

 atmosphere to diffuse. 



" In all cases where differential straining of the gel was 

 successful, the diffusing silver nitrate reached an elliptical 

 boundary." 



No independent criterion regarding the strains in the gel 

 appears to have been applied : examination in polarised light 

 would have shown the distribution of strain unambiguously, 

 but is not mentioned and could not have been used with a 

 cardboard tray as container. The strain must have been of a 

 complicated character, as the gel adhered to the cardboard, 

 and also very probably to the glass plate, at least at first ; 

 as far as I know, gelatin gel does not adhere to xylonite without 

 some special precautions. Assuming the gel to adhere to both 

 the tray and the cover plate, compression would set up a com- 

 plicated internal strain and of course bulging round the edges ; 

 there would also be a strong tendency for the gel to detach itself 

 from the glass and to shift in the direction in which it is free 

 to escape, viz. where the opposite sides of the tray have been 

 removed. The tendency of the gel to detach itself from the 

 glass is much increased by the formation in it of a dense precipi- 

 tate and by the well-known contraction which takes place 

 when the concentrated silver nitrate diffuses into a fairly 

 dilute gelatin gel. In view of my own results, it therefore 

 appears probable that the ellipticity of the reaction zone is 

 due, not to differences in the diffusion velocity, but to a gradual 

 yielding of the gel which mimics this effect. In the experi- 

 ments with cylinders I have found it one of the greatest diffi- 

 culties to prevent the upper plate from detaching itself as the 

 reaction progresses, and the consequent spurious differences 

 in the reaction zones at opposite ends of the same diameter. 



From the results obtained so far with sufficient precautions 

 it is not possible to do more than draw a negative conclusion 

 as regards the problem of gel structure : the conclusion that 

 those elements which on deformation give rise to optical aniso- 

 tropy and to elasticity of shear are either without influence 



