'2V2 



When the tomporatiire is sufficioiitly liii;h, devitrification 

 ])rocoods witli an easily measnrahle velocity. For exam- 

 ple, a IM solntion of sncrose, vitrified in a thin layer, de- 

 vitrities within 10 seconds when exposed to a temperature 

 of -26^'; it devitrities in a minute at about - oO% while it 

 does not devitrify at all, even within an hour, at -35^. 

 We call - 26°, - 30°, etc., ''temperatures of devitrification," 

 but it is evident that one must indicate the time required 

 for devitrification at each temperature if one wishes to 

 give to the notion of devitrification temperatures a precise 

 meaning. 



We have undertaken to establish the devitrification tem- 

 peratures, or better, the time-temperature curves, for 

 aqueous solutions of various organic (Luyet, 1939) and 

 inorganic substances. The method employed was as fol- 

 lows: A small drop of the solution to be vitrified was 

 placed between two glass strips, each about 0.1 mm. thick, 

 kept apart by two bits of glass 0.1 mm. in thickness. This 

 preparation was first immersed in liquid air ; thereafter it 

 was placed in an isopentane bath maintained at a constant 

 temperature, and the time necessary for complete crystal- 

 lization was determined. The opacity of a frozen prepa- 

 ration served as a term of comparison to show when the 

 devitrification was complete. 



The curves obtained seem to indicate that, contrary to 

 the general belief, devitrification does not take place at all 

 at very low temperatures. If one extrapolates the curve 

 of Figure 31, for example, it becomes parallel to the time 

 axis at a temperature a little below that at which devitri- 

 fication occurs within a minute. The hypothesis of a very 

 slow rate of devitrification, requiring hundreds of years 

 for producing an observable result, seems, therefore, to be 

 out of the question, at least for certain substances like 

 those studied here. 



AVithin a rather large range of concentrations, for ex- 

 ample, from 0.9M to 2.2M for a sucrose solution, the de- 

 vitrification temperatures change but little. Thus, lAL and 

 2M sucrose solutions have devitrification temperatures 



