l66 OXIDATION PRODUCTS OF DIANISIDINE. 



micro-crystalline — presumably the basic salts as in the case of 

 the HCl and HNO3 salts. Tap water liberates the isomerised 

 chestnut base CssHggO^N^, from which boiling water extracts 

 the dianisidine which accompanies it. 



A specimen dried in vacuo gave the following" analyses : 

 0.2605 used 0.0398 Sn for reduction = 15.25%. 

 0.3210 gave 0.1125 Ag' — 26.3% HBr. On correcting for trace 

 of potassimn bromide ash, these figures become Sn = 15.4, 

 HBr = 25.6. 



CagHggO^N^Br, requires Sn =r 18.4, HBr ^ 25.0 



The reduction product was again pure dianisidine. 



On sciaking a weighed quantity in cone, ammonia and finally 

 washing out with dilute soda, and finally with warm water, it 

 yielded 71.7% of crude chestnut base, and the washings con- 

 tained a trace of dianisidine. 



Thus 71.7 base + 26.3 HBr ^98.0 is accounted for, the 

 amount of base (mixture of C28H28OJN4 and dianisidine) agree- 

 ing" "vvith above formulation. On percolation with cold alcohol 

 about 48 % of the original was left undissolved: (m.p. 

 190-210° ), being the crude polymerised diii"nine Co,iH280^Nj. 



3. Basic Nitrate of the quinhydroiv:^. — This is much less 

 soluble than the above salts, consequently an almost quantitative 

 yield is obtained when dianisidine in excess of 5% HNO. is 

 treated with ferric nitrate. It can be completely freed from iron 

 by washing, and forms a black-violet paste, soluble in acids to a 

 deep blood-red solution : a trace gives an intense garnet colora- 

 tion in concentrated sulphuric acid. Owing to this intense colour 

 the substance could be used as its own indicator in titrating it 

 with stannous chloride, after dissolving in warm dilute hydro- 

 chloric acid. 0.1370 dried in vacuo used 10.7CC stannous chloride 

 dec = 0.00274 Sn) to disappearance of red colour. Sn for re- 

 duction -:: 21.4% of substance. Titration of a specimen dried in 

 the steam-bath left a ]M-oportion of violet azo-dye unreduced and 

 gave Sn = 15.3%: so that some degree of isomerisation occurs 

 on heating. 



Q^Hg^O^N^ (NO.,) requires Sn = 21.7% : it thus seems to 

 be a basic nitrate analogous to the basic hydrochloride above 

 described. 



4. Qiiinhydrouic sulphate. — From action of ferric chloride 

 on dianisidine in very dilute H.^SO^ at 30°. The almost black 

 precipitate obtained on standing is the sulphate, and i^" quite 

 insoluble in dilute sulphuric acid, though on washing it com- 

 mences to dissolve to a blood-red solution : inky-violet paste 

 almost black on drying in vacuo, which treatment alters it to some 

 extent, as the dried substance gives a purple shade in cone. HgSO^ 

 due to the polymer (see belr.w). On reduction with stannous 

 chloride the liquor was brown, and was found to give a black 

 precipitate with iodine (so deep as to mark the starch endpoint) 

 which was used as the endpoint of the back-titration. 0.2400 

 used 0.0456 Sn for reduction, and filtrate worked up witli Barium 



