OP ARTS AND SCIENCES. Ill 



which can he easily separated by fractional distillation under re- 

 duced pressure : the former suhstance was then identified by its 

 hoiling point (137°), hut the formanilide obtained is never pure, 

 hut contains a considerable quantity of acetanilide from which it 

 can he separated by recrystallization from ligroine and benzene. 

 This is due to the fact that formanilide, as shown by a special 

 experiment, is converted almost entirely in half an hour at 90° by 

 acetic anhydride into acetanilide. 



A new experiment was therefore tried, using formic acid instead 

 of acetic acid, in the hope of obtaining thereby formic anhydride 

 which has not yet been isolated. 



On mixing phenylisocyanide with pure formic acid (mpt. 8°. 6) 

 at 0°, a very violent reaction takes place, with much evolution of 

 gas : the phenylisocyanide was therefore slowly poured into well 

 cooled formic acid (2 molecules), whereby immediately at 0° a 

 steady evolution of pure carbonic oxide takes place. On distilling 

 the residue, only formanilide was obtained, which boils at 283°- 

 285° almost without decomposition, and, recrystallized from ligro- 

 ine, melts at 48°. The product obtained is identical in every re- 

 spect with a formanilide made according to the method of Wallach 

 and Wiisten.* This experiment shows that formic anhydride, 

 which probably is formed as an intermediate product, 



H 

 CO 



H 

 C 6 H 5 -N-C= + ggSo = C ° H ^= C OH + °( co 



H 



H 



is not capable of existence at 0°, but decomposes spontaneously 

 into 2 CO and H 2 0. Gerhardt f has already attempted, without 

 success, to obtain this substance from benzoylchloride and sodic 

 formiate. The following further observation was then made in 

 this direction. On pouring phosphorus oxychloride (1 molecule) 

 over potassic formiate (4 molecules), a violent reaction takes place 

 at 0°, and much carbonic oxide is evolved. 



On adding anhydrous oxalic acid to phenylisocyanide, reaction 

 takes place at 0° and formanilide, carbonic oxide, and carbon diox- 

 ide are formed, which makes it likely that oxalic anh} T dride, proba- 

 bly formed as an intermediate product, decomposes spontaneously. 



* Ber. d. chem. Ges., XVI. 145. 



t Ann. Chem. (Liebig), LXXXVII. 149. 



