21(1 



tliat only water cryslnlli/A's duriiii'' llic successive devitri- 

 fications; l)iit tliis i)oint is to be investigated.) 



The existence of three ranges of concentration witli the 

 properties just described has been observed in a great 

 number of solutions. Thus, solutions of gum arabic^ do 

 not devitrify if their concentration exceeds 68''/. They 

 devitrify in the form of tufts between 6S'/( and G-^Vc , and 

 in the amber-color form below 64%. 



Devitrifications at more than two ranges of tempera- 

 ture were noticed with solutions of substances such as 

 urea.- 



Witli sohitions of sodium chloride, two devitrifications 

 of the anilxM-color type occur at different temperatures, 

 one of them in the neighborhood of -28°. When the de- 

 vitrified material is warmed up, a partial melting takes 

 place at about -21°, the well-known freezing temperature 

 of the eutectic mixture. 



We have also applied the method of devitrification to 

 the study of the mode of binding of water in substances 

 which set as gels. A preliminary work'' has shown that a 

 sugar solution to which pectin has been added and which 

 has been allowed to set, devitrifies at temperatures slightly 

 below those at which an identical but alkaline solution, 

 which does not set, devitrifies. 



From what we have said on the higher devitrification 

 temperatures of solutions of substances such as gelatine, 

 dextrin, the gums, etc., it follows that these solutions can 

 freeze only from about zero to 10 or 12 degrees below zero. 



4. Yitrofusion. The direct passage from the vitreous 

 to the liquid state can be effected by a rapid warming. 

 The conditions foi- assuring this ])assage and avoiding the 

 intermediate crystalline state are fundamentally the same 

 as those required for vitrification : 1. The greatest possible 

 temperature difference between the warming bath and the 

 object ; 2. The smallest possible mass and the greatest pos- 



1 Eesearch made in collaboration with Mr. J. Fnltoii. 



2 Studied in collaboration with Mr. H. Noe. 

 a In collalioratioii witli Mr. W. Sclimiesing. 



