163 



hydrogels of alumina, of silica and of ferric oxide. While 

 freozino- was completed at -4° in the sand mixture, it 

 required a temperature of -28° to be terminated in the 

 lampblack mixture. In most of the other substances 

 studied, it was observed that at - 20° some water was still 

 crystallizing out. 



Moran (1926), who had previously observed that, when 

 gelatin gels in form of discs are frozen slowly, ice forms 

 at the surface, determined the quantity of water left in 

 the core of the discs, that is, the quantity of non-frozen 

 water, after exposure to various low temperatures. The 

 curve of Figure 24, constructed from his data, represents 

 the amount of water found in the cores of discs exposed to 

 temperatures from - 3° to - 19°. At this last temperature, 



/o 



45 



40 



35 



0° -5° -10° -15" -20 



o 



Fig. 24. Proportion of water which remains unfrozen in gelatin gels ex- 

 posed to various temperatures. (Calculated from Moran 's data, 1926.) Ab- 

 scissae: temperatures; ordinates: unfrozen water in percent of the total 

 weight. The horizontal dotted line represents the quantity of water which 

 does not freeze at any temperature. 



