This section is from "Scientific American Supplement Volumes 275, 286, 288, 299, 303, 312, 315, 324, 344 and 358". Also available from Amazon: Scientific American Reference Book.
If a solution of ferric oxide and manganese ammonium oxalate is submitted to electrolysis, without the previous addition of ammonium oxalate, the characteristic color of permanganic acid immediately makes its appearance, and the peroxide gradually precipitates itself on the positive, while the iron is deposited on the negative electrode. When the examination is made in the above manner, it is impossible to separate the two metals, for the peroxide will bring down with it a considerable quantity of ferric hydrate. The separation of the two metals is only possible when the precipitation of the manganese peroxide is prevented, until the greater portion of the iron has been deposited. This result may be attained by adding sodium phosphate, or, better still, by the addition of ammonium oxalate in great excess. In both cases the characteristic coloration from permanganic acid is developed by the action of the current at the positive pole; this, however, disappears in the direction of the negative electrode. After the greater portion of the ammonium oxalate has been converted into carbonate, the coloration and necessarily the formation of manganese peroxide begins.
Ammonium oxalate is added to the solution, and heat applied; then three or four grammes more of ammonium oxalate are dissolved in the liquid, which is then immediately submitted to electrolysis. When the amount of manganese is small, the separation of the two elements takes place very rapidly, and the results are accurate. If the amount of manganese is more than double that of iron, the separation of the latter will take a much longer time. Then, in order to effect a complete separation of the two elements, it is necessary to redissolve the deposited manganese in oxalic acid (the acid is added, without interrupting the current, until the liquid becomes red), and the current is allowed to continue its action.
It was found desirable, in effecting this separation, not to employ too strong a current (two Bunsen elements will suffice), and only to increase the strength of the current when it is necessary, in consequence of a large amount of manganese being present, to redissolve the peroxide.
When the process is completed, it is not advisable to allow the current to act any longer, for otherwise some of the peroxide may adhere firmly to the iron, and the latter (after previously having poured off the liquid) must be redissolved in oxalic acid, that is to say, the electrolysis must be repeated. As has been already mentioned in the determination of manganese as peroxide, its precipitation from ammonium oxalate is not complete. The solution which contains the greater portion of manganese, suspended as peroxide, must first, therefore, be boiled to decompose the ammonium carbonate; the remainder of the ammonium oxalate is neutralized with nitric acid, and the manganese converted into the sulphide by ammonium sulphide. The manganese sulphide is then ignited in a current of hydrogen, and weighed as such.