DIFFUSION IN DEFORMED GELS Zy 



such concentrations as can conveniently be handled must be 

 continued for hours or even days. I had, however, shown 

 previously * that optical anisotropy persists in deformed 

 gels kept at ordinary temperature (io° to i5°C.) for many 

 days after the complete disappearance of stress, which appeared 

 to justify even prolonged experiments. 



The deformation employed in the first instance was the 

 simplest conceivable. A rectangular prism of gelatin gel was 

 compressed between two parallel plates, in such a manner 

 that it could expand freely in the two directions at right angles 

 to the axis of compression, to 80 or 70 per cent, of its original 

 height. Into one of the faces parallel to the direction of com- 

 pression a solution diffused through a small circular opening. 

 The diffusion zone in an isotropic medium is (very approxi- 

 mately) a hemisphere and its intersection with the gel surface 

 is the equator of the former. Assuming the gel to become aniso- 

 tropic, the diffusion velocities in the direction of maximum 

 compression and in that at right angles to it — in which the gel 

 undergoes maximum elongation — would be different, in which 

 case the intersection of the diffusion zone with the surface 

 would no longer be a circle. 



Simple as this arrangement is in principle, considerable 

 difficulties were encountered in realising it experimentally. 

 The first was the choice of a suitable diffusing solution. As 

 comparative measurements of the diffusion zone were required, 

 a sharp boundary was necessary, which naturally could not 

 be secured by using, e.g., a coloured solution only, as this pro- 

 duces a gradient to zero concentration. It is therefore essential 

 to produce a reaction between the diffusing solution and some 

 sort of indicator contained in the gel, which gives a sharp 

 line of demarcation. Although at first sight the choice of 

 suitable solutions and indicators seems almost unlimited, it 

 proved in practice to be very narrow if all sources of error 

 were to be avoided. Thus any solution having a lyotropic 

 action was at once ruled out, as it would distort the results 

 completely. If, to give a striking example, a very dilute 

 acid is allowed to diffuse into gelatin gel (oxalic acid gives a 

 sharp boundary even without indicator owing to the presence 

 of calcium salts in all gelatins), the diffusion zone swells markedly 

 and its modulus decreases correspondingly. The diffusion 

 zone therefore yields to the compression more than the surround- 

 ing gel and becomes elliptical. No valid conclusions regarding 

 the diffusion velocity can, however, be drawn from this — rather 

 tempting' — result : this can easily be demonstrated by allowing 

 acid to diffuse into an uncompressed prism and compressing 

 it subsequently. The diffusion zone becomes an ellipse, the 



1 Ibid. 



