209 



is transferred from air at 20° to a liquid bath at - 200°, the 

 temperature drops about 200 degrees duriii<>' the first sec- 

 ond. It is generally admitted that, with liquid air as a 

 cooling- bath, the drop in temperature is somewhat slower 

 because of the formation, around the object, of the pro- 

 tective air mantle mentioned above ; however, the cooling 

 velocity obtainable w4th liquid air amply suffices to vitrify 

 objects whose thickness is of the order of 0.1 millimeter. 



When the mass of material to be cooled is too large, only 

 the most external layer vitrifies. The inner parts, wdiich 

 lose their heat too slowly, freeze. 



Whenever crystallization begins at some point on an 

 object, the conditions for vitrification of the rest are im- 

 paired. The p;ortion which crystallizes liberates heat ; 

 this spreads to the surrounding parts and maintains their 

 temperature at the freezing point. Hence, the whole mass 

 might freeze. 



In our research on the vitrification of colloids we have 

 found, moreover, that their water content determines the 

 possibility or impossibility of vitrification. In general, 

 with 50^ gelatin solutions, w^e have been able to vitrify 

 layers 0.3 mm. in thickness (by the method of immersion 

 in liquid air), while, wdth solutions containing 90% water, 

 W'C could vitrify only smears a few micra thick. 



Attempts at vitrifying pure w^ater have been made by 

 a few investigators. Burton and Oliver (1935) obtained, 

 from steam, some solid water in wdiich X-ray analysis did 

 not reveal any crystalline structure. 



Previously, Haw^kes (1929) had mentioned an experi- 

 ment in which a drop of solid amorphous water was ob- 

 tained, by chance, during rapid cooling. 



The difficulty experienced in vitrifying pure water has 

 been attributed, in general, to the high velocity of crystal- 

 lization of that substance. Walton and Judd (1914) who 

 measured this velocity, found 65 mm. per second (an 

 exceptionally high velocity). 



According to Callow (1925), the addition of 3^/( gelatine 

 to w^ater reduces its velocity of crystallization to 1 350 of 

 its value. 



