Not more than O'l mgr. of iron -was left in the solution as determined 

 colorimetrically with sulphocyanide ; the remainder must therefore have 

 been carried down by the manganese peroxide ; it was detected qualita- 

 tively in each case in the precipitate, but no quantitative estimations were 

 made. 



Direct experiments on the electrolysis of solutions of manganese 

 chloride, to which 5 gr. of ammonium oxalate were added, and in which 

 variations both of current and of electromotive force were tried, showed 

 that it is not possible to electrolyse such solutions, under conditions per- 

 mitting the deposition of iron, without the separation of manganese per- 

 oxide. The separation is eflPected the more rapidly the greater the propor- 

 tion of manganese present and the higher the current density and the elec- 

 tromotive force. With only G-Ql gr. of manganese in solution, an E.M.F. 

 of 3 volts, and C.D.ioo = 0"2 ampere, the precipitation of hydrate occurred 

 after four hours' electrolysis in the cold solution, and in eighteen hours, 

 the time required for an electrolytic determination of iron, with only 

 0-002 gr. of manganese, anE M.F. of 1-35 volts, and CD. , on = 0-1 ampere, 

 the hydrate also separated. With the view of delaying this separation of 

 the manganese a series of experiments were tried in which a small 

 quantity of hydroxylamine sulphate was added to the solution to be 

 electrolysed. It has been shown that this reagent acts favourably in 

 preventing the separation of stannic acid in the electrolysis of tin salts in 

 ammonium oxalate solution,^ and it might, therefore, have a similar 

 favourable effect in the case of manganese. 



To a small extent this is the case; the addition of 1 gr. of hydroxyl- 

 amine sulphate, under conditions similar to those recorded in Series II. 

 of our experiments, considerably delays the separation of the hydrated 

 peroxide. But the deposition of iron is also delayed, and attempts to 

 separate the two metals with this addition gave results similar to those 

 of Series VI. The separated peroxide contained iron, and the deposited 

 metal was from 3 to 16 per cent, too low ; in addition, the iron deposit 

 was uneven and showed a tendency to scale off. 



We therefore conclude that the quantitative separation of iron and 

 manganese in ammonium oxalate solution cannot be effected. Further, 

 the influence of small proportions of manganese ou the electrolytic deposi- 

 tion of iron, referred to above, is a factor that detracts very considerably 

 from the analytical value of the electrolytic method for the determina- 

 tion of the latter. 



On the other hand, very small proportions of manganese can be sepa* 



' Tliirfl npro''f> 1896. 



