RECENT ADVANCES IN SCIENCE 207 



nickel — having been described by Sabatier and Senderens in 

 1905. The preparation of tertiary amines from the halogen 

 derivatives of hydrocarbons may, according to Matter (D.R.P. 

 301450 and 301832) be effected by means of sodamide. Thus 

 if chlorobenzene is heated with sodamide at no-120 tri- 

 phenylamine is formed, while benzyl chloride under the same 

 conditions yields tribenzylamine. 



About a year ago Emmet Reid (Journ. Amer. Chem. Soc. 

 191 7, 39, 124 and 304) recommended the use of p-nitro benzyl- 

 bromide for the identification of acids and phenols. He found 

 that when this reagent is boiled in alcoholic solution with the 

 sodium salts of acids, esters are readily formed which can be 

 identified by their melting-points ; in the same way sodium 

 or potassium phenoxides gave solid phenolic esters. Brewster 

 (Journ. Amer. Chem. Soc. 191 8, 40, 406) now states that good 

 yields of this reagent may be obtained by gradually adding 

 bromine, dissolved in carbon tetrachloride to a boiling solu- 

 tion of p-nitrotoluene, containing a trace of iodine, and ex- 

 posing to sunlight. 



Although the titration of glycine by means of formaldehyde 

 is now a well-known and frequently employed method, the 

 reaction between these two substances has not hitherto been 

 explained. In a paper by Krause (Berichte, 191 8, 51, 136) it 

 is shown that the reaction is somewhat complex, resulting in 

 the formation of a condensation compound of the formula 



HO . CH(CH 2 . NH . CH 2 COOH) 2 . 



When ordinary formalin is employed the methyl alcohol 

 usually present in this substance takes part in the reaction as 

 follows : 



2 NH,CH 2 COOH + 2CH 2 + CH 3 OH = C 7 H u 5 N 2 + 2 H 2 0. 



But if pure formaldehyde is employed, relatively more of this 

 substance is used up and formic acid is produced at the same 

 time, as shown by the following equation : 



I 2NH 2 CH 2 COOH + 4CH 2 = C 7 H u 5 N 2 + HCOOH + H 2 0. 



The salts of glycine behave in a similar manner to the 

 free acid. 



A further contribution to the vexed question of the con- 

 necting link in carbohydrate formation during assimilation 



