•J 14 



At coiicciit I'atidiis hiii'lu'r tliaii tliosc (if tlic al)()\'(' iiicii- 

 lioiK'd raiiiiv, a devil rificalioii ot" a (lilTci'dil !>"])(' (»l)taiiis, 

 wliicli wi' sliall describe farther on. 



Siiiee devitrification is a clianj^c of stale c()m])aral)le, in 

 many respects to other cliani'cs of state like melting and 

 boilini>', it is i)ro])a])le that, like theni, it takes ])lace at a 

 temperature which is a function of the chemical structure 

 of the substance under investigation. One should expect 

 a rise in tlu' devitrification temperatures as one passes, 

 within the sami' sei'ies, from compounds of simple molec- 

 uhir structure to those with a more c()mi)lex one, as is the 

 case with the boiling point, which rises when one passes, 

 for example, from methane to pentane. We could estab- 

 lish the existence of such a relation. In a series of sugars 

 we obtained (duration of devitrification, 5 minutes)' : 



Glucose 2M C, H,,0„ -40.6° 



Sucrose 2M Ci,H,,(),, -31.8° 



Kaffinose IM CisH.s.O,, 5H,0 - 27.2° 



Dextrin 2M x {C, H,„0.-,). - 9.4° 



The other high-molecular-weight substances which we have 

 studied, such as, gelatine, albumin, gums, dextrin, have 

 high devitritication temperatures (about -10°). Sub- 

 stances whose chemical composition is more or less similar 

 to that of the sugars, but which have much lower molecular 

 weights, such as glycerine, ethylene glycol and formalde- 

 hyde, have been found, in preliminary experiments, to 

 have devitrification temperatures in the neighborhood of 

 -60° to -70°. 



However, the relation which we just noted cannot be 

 simply compared to that between the molecular complexity 

 of a substance and its melting or boiling point. In the one 

 case, it is question of solutions, in the other of definite 

 compounds. 



It is to be expected that certain molecular groupings in 

 the solute have a specific effect on the devitrification tem- 



1 Liiyt't, R. J., ii:iiKM- in jircss in ./. P/(.(/.v. Chrm. 



