14 



Scientific Proceedings (94). 



Fourteen hundred c.c. of dry aniline oil are cooled with an ice 

 mixture to o° C. or lower, and the cooled arsenic acid is added 

 slowly with vigorous stirring. The mixture soon becomes thick 

 and then granular, after which it is finely ground and thoroughly 

 mixed. This powder has roughly the composition of (C 6 H 5 NH 2 )3 

 -(H 3 As0 4 )2. 



Two hundred grams of this powder are heated in an Erlen- 

 meyer flask, by means of a "Crisco" bath to 160 0 C, when the 

 powder begins to melt. The substance is stirred continually 

 and when it has all melted, a reflux condenser is then attached; 

 for one and a half hour it is heated from i6o°-i70°, and then one 

 hour from 180 0 to 183 0 C. After allowing the mixture to cool 

 somewhat 225 c.c. of 6N sodium hydroxide and 225 c.c. of water 

 are added, which causes the substance to dissolve and separate 

 out the aniline oil left uncombined. 



When the mixture is cool, the aqueous layer is drawn off with 

 the aid of a separatory funnel, and after shaking with 15 to 20 

 grams of infusorial earth or kaolin, it is filtered with suction. 



To the clear filtrate is added 100 c.c. of 6 N hydrochloric acid. 



By means of 25 c.c. portions and further additions of 0.5 c.c, 

 1. 00 c.c, 2.00 c.c, etc, of 6 N hydrochloric acid, one finds out if 

 further addition of acid to the whole filtrate will give a larger 

 yield or not. When after waiting a few minutes for the crystal- 

 lization to take place, and when the best medium has been deter- 

 mined on the small aliquot portion, an equivalent amount of acid 

 is then added to the whole filtrate. Usually the crystals fill the 

 entire solution so that it has the appearance of being solid. 



After standing for an hour or longer the precipitate, which is 

 white and crystalline, is filtered with suction. It is then washed 

 by suspending the precipitate in 200 c.c. of water and filtering 

 with suction. When all the wash liquid has drained from the pre- 

 cipitate, it is dried by fanning, or in any other suitable way. The 

 yield is about 30 per cent, of the theory, whereas 10 per cent, 

 seems to be the average heretofore. 



The products obtained by the above directions are almost 

 pure enough for further use, but if necessary can be reprecipitated 

 with alkali and acid, or by dissolving it in boiling water and allow- 

 ing the solution to cool. 



