40 THE ROYAL SOCIETY OF CANADA 



This extremely slight reaction given by amides is of interest, 

 for, comparing this class of compounds with amino acids, it is the latter 

 which are much the more resistant to hydrolysis to produce ammonia; 

 and yet it is precisely that class which reacts the more readily with 

 triketohydrindene hydrate. Thus the non-reactivity of the amides 

 supports the idea that the reactivity of the amino acids is due to their 

 dissociation. 



Amines, however, have been shown to react with triketohydrin- 

 dene hydrate by Neuberg and this result we have confirmed in the 

 case of ethylamine. A one per cent, solution of ethylamine gives, 

 on heating with "ninhydrin," and subsequent dilution a strong blue 

 color readily identified as the ammonium salt of diketohydrindylidene- 

 diketohydrindamine. When the concentration is such that 1 c.c. = 

 0-05 mgs. nitrogen, there is no reaction; but in presence of pyridine, 

 a strong coloration is produced which is more in amount than that given 

 by ammonium salts. The few secondary and tertiary amines which we 

 have examined give only the faintest of reactions. These facts we 

 regard as significant; but we have examined too few a number of bases 

 as yet to warrant the drawing of any definite conclusions. An exam.- 

 ination of a large number of bases of varied types is in contemplation 

 and we wish to reserve this part of the subject. 



THE "ninhydrin" REACTION WITH GLYCEROL, ETC. 



The results of Halle, Loewenstein and Pribram who obtained 

 a blue coloration by heating anhydrous glycerol with ninhydrin 

 are of the greatest importance, for if substantiated they would form 

 quite a new development of the chemistry of the test. At first we 

 were unable to repeat their results, but this, we found later, was due 

 to our glycerol containing too large a proportion of water. Since 

 then, and using anhydrous glycerol, we have been able to obtain a 

 bluish shade of color on boiling it with triketohydrindene hydrate for 

 thirty seconds. We soon recognized however, that dififerent samples 

 of glycerol gave very different amounts of color; one, a specimen of 

 Schuchardt's crystallisable glycerol, only gave the faintest trace 

 of pinkish coloration, while another marked "anhydrous" gave quite 

 a strong blue color. This color was identified as the ammonium salt 

 of diketohydrindylidene-diketohydrindamine by the three previously 

 mentioned tests. This of course instantly led us to the conclusion 

 that the coloration given by glycerol was due to traces of nitrogenous 

 impurity and to extend that explanation to the reactions given by 

 glycol, aldehydes, ketones and sugars, for which claims to a positive 

 ninhydrin reaction had been made by Halle, Loewenstein and Pribram. 

 We tested this hypothesis in the following way. One c.c. of ammonium 



