256 PROCEEDINGS OP 1 II i: AMERICAN ACADEMY. 



can be determined. [gnition of ferric nitrate was therefore chosen as 

 the next method of obtaining ferric oxide. 



The material used in the following work was purified with great can'. 

 In the firs( place a solution of a so-called "chemically pore" chloride of 



iron was treated with an excess of sulphuretted hydrogen and the result- 

 ing sulphides and sulphur were removed l»y liltration. From the filtrate, 



after oxidation with nitric acid,, the iron was precipitated with an excess 

 of amnionic hydrate and the precipitate was thoroughly washed by de- 

 cantation. In this process traces of a r of metals which might 



have been present must have been removed. The precipitate was dis- 

 solved in sulphuric aeid. and by cautious addition of amnionic hydrate 

 basic ferric sulphate was thrown down : a precipitate which is much 

 more easily washed than ferric hydrate but which was redissolved in 

 sulphuric acid with considerable difficulty. In order to reduce to the 

 ferrous state the ferric sulphate thus formed, the solution was next sub- 

 jected to the action of a galvanic current of several amperes. The solu- 

 tion was contained in a large platinum dish which served as the negative 

 electrode, the positive electrode being a flat spiral of platinum wire. 

 Since the solution was verj concentrated and contained a considerable 

 excess of sulphuric acid, the greater part of the ferrous sulphate crystal- 

 lized out when the solution was allowed to cool. By alternately elec- 

 trolyzing and cooling, almost all of the iron was eventually obtained as 

 ferrous sulphate. 



The next process in the purification was to deposit the iron electrolyti- 

 cally from a solution of its oxalate. By means of this separation, the 

 aluminum and manganese could be eliminated, since the first, together 

 with alkaline and other impurities, is not deposited by electrolysis, while 

 the second is deposited upon the anode. The amnionic oxalate needed 

 was prepared with great care. Oxalic acid which had been repeatedly 

 crystallized with hydrochloric acid, and then until free from chlorine, was 

 Baturated with pure redistilled ammonia, and the resulting amnionic oxa- 

 late was subjected to several crystallizations. To a hot concentrated so- 

 lution of this amnionic oxalate, the ferrous sulphate was added as long as 

 no permanent precipitate resulted, and the solution was then electrolyzed 

 until the greater part of the iron had been deposited. A slight deposit 

 of hydrated manganese peroxide appeared upon the positive pole, which, 

 however, presented enough surface to prevent the deposit from becoming 

 detached. The film of iron was carefully washed and dissolved in pure 

 dilute nitric acid. When all of the ferrous sulphate had been thus con- 

 verted into ferric nitrate, the solution of the nitrate was filtered and 



