215 



peratuix's of llic solution. In prelimiiiaiy investigations 

 on this problem' we have observed that dextrose and levu- 

 lose, wliich have the same empirical formula but differ in 

 the structural arrangement of their atoms, have the same 

 devitrification temperatures, while sucrose and lactose, 

 which are likewise isomers, exhibit a slight difference 

 therein. 



There also exists a relation between devitrification tem- 

 peratures and water of hydration or, more generally, 

 between devitrification temperatures and the wat<'r-binding 

 capacity of the solute (Luyet, 1939). With sucrose solu- 

 tions we found that, when the concentration w^as higher 

 than that corresponding to ten molecules of water per one 

 of sucrose, devitrification did not take place at all, no 

 matter what the temperature was. With lower concentra- 

 tions, corresponding to one molecule of sugar for every 

 11, 12 or 13 of \vater, devitrification occurred at tempera- 

 tures of about - 50°, and the frozen mass exhibited a par- 

 ticular type of crystallization which we designate by 

 ''crystallization in tufts" there being tufts of crystalline 

 needles in the vitreous mass. With lesser concentrations, 

 a different type of devitrification set in : the preparation 

 assumed an amber color which gradually darkened to com- 

 plete opacity. The devitrification temperatures were 

 about 15 degrees higher in this case. It is this type of 

 devitrification which we have studied particularly and 

 which has furnished the numerical data given above. One 

 can, it seems, interpret the results observed in these three 

 ranges of concentration by supposing that the first 10 

 molecules of water which come in contact with a sugar 

 molecule attach themselves so firmly to it that they cannot 

 be torn loose by the forces of crystallization. The next 

 three molecules would be bound by a different type of 

 bond, and, when more than 13 molecules of water are 

 present per molecule of sugar, this excess water would be 

 held only by the forces of solution and would be free to 

 solidify. (In making this hypothesis, we are supposing 



1 In collaboration with Dr. C. and Miss M. .Jordan. 



