136 



^vlli(•ll ronncd "lut'inhriiiU's" ahoiil ()..") micron in thickness, 

 \vliil(' llic other whicli srciiicd ('nch)s('(l within these mem- 

 ])r;nies, consisted of much broader layers. Outside of the 

 rings, solidificntion continued in the form of rays (Fig. 11, 

 A), the rings l)ulgin,<!: into the l)ase of tlie rays. The rays 

 themselves were dixidcd transversely into com])ai'tnieiits 

 l)y tlie curved nienihi'aiies. At the ti]) of ;i ray, or at tlie 



Fig. 11. Formation of "rays" at tlu' in'riidiory of crvstalliiK' discs in 

 frozen gelatin gels. (From Hardy, 1926.) An entire ray is shown in A, and 

 the terminal portion of a ray, at higher magnification, in B. 



outer edge of a circle, one could see a series of tine etched 

 lines (Fig. 11, B). 



Moran explains the concentric structure by assuming 

 that the heat developed in the formation of ice in a center 

 of crystallization keeps the local temperature higher at 

 that point for a time and prevents a further freezing. The 

 gelatin in the immediate neighborhood becomes more con- 

 centrated on account of the withdrawal of the water trans- 

 formed into ice and it cannot freeze. When the local tem- 

 perature is lowered again, a new layer of ice will be formed. 

 The concentric shells will result from such an alternation 

 of warmings and coolings. 



Hardy claims that, while the '^ membrane" phase con- 

 sists of dehydrated gel, the layers between the membranes 

 are not made of ice, as Moran thought, but of a solid solu- 

 tion of ice and gelatin. He bases his conclusions on the 

 optical properties of the various phases of the system (in 

 polarized light) and on the behavior of the material on 

 thawing. 



Disseminated F r e e z i n g. With weaker con- 

 centrations or lower temperatures, for example, with 88% 

 watei- at -19° or in liquid air, a third type of freezing oc- 

 curred. Moran described it as follows: "The interior of 



