ON THE ELECTROLYTIC METHODS OF QUANTITATIVE ANALYSIS. 183 



In the electrolysis of manganese salts in presence of dilute mineral 

 acids, sodium pyrophosphate, or ammonium oxalate, the manganese is 

 separated at the anode as hydrated peroxide. This property is of little 

 value, however, for the determination of manganese itself, because it is 

 difficult to effect the complete separation of the metal as oxide, and special 

 conditions must be adopted to cause the precipitate to adhere to the 

 anode. On the other hand, the difference in the behaviour of iron and 

 manganese when subjected to electrolysis in ammonium oxalate solution 

 is attractive as a method for the separation of the two metals. 



The practical difhculty in effecting a separation on these lines is that 

 the precipitated manganic oxide always carries down some iron with it ; 

 this can only be overcome by adopting such conditions of electrolysis that 

 only one of the metals is separated, the other remaining in solution. By 

 means of a divided cell Engels states that manganese can be completely 

 separated as peroxide, using a sulphuric acid solution ; this involves the 

 subsequent determination of the iron. The chief work on the subject, 

 however, has aimed at the determination of the iron by deposition on the 

 cathode, obviously the more useful line of separation. 



According to Classen, the separation is possible if 8 to 10 gr. of 

 ammonium oxalate are added to the solution of the mixed salts and the 

 mixture electrolysed warm with a C.D.,(,n of about 1-0 ampere. This 

 proportion of oxalate is said to outlast the deposition of the iron ; the 

 manganic oxide does not separate until the mass of the oxalate has been 

 decomposed, and even with large proportions of manganese only very little 

 peroxide separates at the anode under these conditions. 



Neumann ' and Engels ^ both state that the separation is incomplete, and 

 our own experiments confirm this view. We have not found it possible 

 to completely deposit iron without a separation of manganese peroxide, 

 nor to separate the latter free from iron. The presence of even small 

 proportions of manganese has, moreover, quite a remarkable effect in 

 hastening the separation of iron as hydrate in the electrolysis of oxalate 

 solutions. In some early experiments on the determination of iron the 

 results were from 3 to 4 per cent, too low, and a separation of hydrate 

 always took place after about two hours ; on testing the precipitate it 

 was found to contain manganese, derived from the iron wire used in 

 making up the solution. In the subsequent work recorded above the 

 iron was always purified from manganese by precipitation as basic 

 acetate. 



The following experiments show the extent of the error when the 

 separation is conducted under the conditions most favourable for the 

 deposition of the iron. The mixed chlorides of the two metals were 

 neutralised with ammonium hydrate, 5 gr. of ammonium oxalate added, 

 and electrolysed as usual. In experiments 4, 5, and 6, a CD. ion of only 

 0*2 ampere was used ; but still the deposition of the iron was incomplete, 

 and in all cases manganese peroxide contaminated with iron separated 

 at the anode or remained suspended in the solution. A comparison of 

 the results tabulated on p. 184 shows that the error in the iron increase?! 

 with tlie proportion of manganese taken. 



' Theorie v. Praxis der anahjtischen Electrolyse, p. 194. 

 ^ Cliem. Rundschau, 1896, pp. 5 an:l 20, 



