201 



reported that wlicii llic lempci-ature was lowered slowly 

 to -5° the subcooling- point of the material a])proached 

 that temperature, while if cooling- to - 5° was rajjid, sub- 

 cooling ceased near -3°. 



Wright and Taylor (1921) confirmed the observations 

 of the last mentioned authors by showing that with a cool- 

 ing temperature of -9° a subcooling point of -6.5° was 

 obtained (on potato), while with a cooling temperature of 

 -12.9° subcooling ceased at -5°. 



As to the factor, solute concentration, a work of Jatfe 

 (1903) would indicate that more concentrated solutions 

 would be more sensitive to crystallization by shock. 



Voigtliinder (1909) attempted to investigate the influ- 

 ence of the osmotic concentration on the degree of sub- 

 cooling, in plants. He worked with 20 different species 

 and made at least fifty determinations on each. No rela- 

 tion was found; the subcooling point varied at random 

 from -4.19° to -11.07°, and the maximum subcooling was 

 about the same in plants isotonic with 1% and in plants 

 isotonic with 4.6 y^ KNO3. Unfortunately, in the same 

 work, strange results were obtained for the freezing points, 

 there being no relation between the latter and osmotic 

 pressure; for example, materials isotonic with 2.6yc and 

 4A% KNO3 were found to have the same freezing point. 

 However, since at subcooling there is a reversal of the 

 direction of the freezing curve, which can be more accu- 

 rately determined than the simple slowing of the curve at 

 freezing {Cf. Fig. 7, curve 5), Voigtlander's subcooling 

 maxima might be sufficiently well observed to be taken in 

 consideration. 



That a higher concentration favors subcooling would 

 result from an observation of Moran (1925) who exposed 

 chicken eggs to evaporation until they lost about l^f of 

 their weight and noticed, then, that subcooling took place 

 more readily. 



Weigman (1936) says that he obtained deep subcoolings 

 with snails {helix) possessing an operculum, while uncov- 

 ered snails presented only slight subcoolings. Since it is 



