COXÖTITUTIÜN OF ULYCOCOLL AND ITS DERIVATIVES. 97 



the latter, then, décomposes, under the ronditions of the experiment, 

 into glycollic acid, nitrogen, and water : 



liX-xiis.xu., ru:-oii 



I = I + X., + H.,(J. 



OC-OH OC-011 



13. Melallic dcrivaUces. Amono- the deri^■ati^■es of n'lvcocoJl one. wliich 

 is best known and most characteristic, is the copper compound. The 

 deep bhie colour of this bodv and its sohiljilitv in alkalies distinüTiish 

 it from ordinary carboxylic salts of copper, and lead us to the con- 

 clusion that it is most probably a cuprammonium compound, a con- 

 ception which can be readily expressed by the following scheme : 



H.,C - XH3 H.C - XH., - Cu - H.,X - CH., 



2 I I 4- CuO = I 1 II +H.,0 . 



OC-0 OC-0 0-CO 



If we attempted to represent the copper compound as an ammoniated 

 derivtitive by hel[) of the open formula and gave it the constitution 

 Cu(XH.CH2.COOH)2, we should have to make the baseless assumption 

 that the hydrogen of the carboxyl group remains unreplaced by copper 

 even in the presence of an excess of cupric oxide. If, on the other 

 hand, we regarded the copper derivative as (H2X.CH._,.COO)2C*u, look- 

 ing upon glycocoll as a carboxylic compound (an acid) because it 

 dissolves oxide of copper and some (^ther metallic oxides, we might 

 argue that ammonium chloride or even anmionia itself is also an 

 acid ! 



Reference must here be made to two important papers bearing 

 upon the ([uestion : one by Curtius and Goebel (/. pralt. Clieiii., 37, 

 150 [18.S.S]), the other by Kraut (Lieh. Ami., 266, i^9l> [1891]). By 

 digestion of the hydrochloride of an amido-acetic ester with an excess 

 of freshly precipitated cupric oxide, Curtius and Goebel obtained a 



