1 822.] Method of analyzing the Ores of Nickel. 215 



mixed with carburet of iron, which separates from the precipi- 

 tating iron in proportion as it dissolves ; on the other hand, the 

 reduced metal can scarcely be dried without suffering consider- 

 able oxidation. If the oxide of copper be precipitated by a car- 

 bonated alkali, this alkali, when added in excess, always dissolves 

 a small quantity of carbonate of copper. This may indeed be 

 obtained, if the solution be evaporated to dryness, and the resi- 

 duum be made red-hot ; the carbonate of copper is then decom- 

 posed, and water separates the alkaline subcarbonate from it; 

 but then the oxide of copper covers the sides of the crucible, and 

 adheres very firmly to it. The crucible must first be weighed 

 alone, and then with the oxide of copper, which must be dis- 

 solved by an acid. No one of these methods is convenient. I 

 have found that the method of separating oxide of copper from 

 its ammoniacal solution by caustic potash gives a much more 

 correct result than the foregoing processes. The separation i& 

 not, however, perfect; for the ammoniacal liquor, when filtered^ 

 becomes brown on the addition of hydrosulphuret of ammonia, 

 and in a few days flocculi are deposited, but so inconsiderable 

 in quantity that I could not weigh them with certainty. I also 

 endeavoured to precipitate the copper from its solutions by means 

 of sulphuretted hydrogen, and to weigh the dried bisulphuret ; 

 but it always gave me at least three or four per cent, too much 

 weight for the oxide of copper employed, for the bisulphuret of 

 copper becomes acid during exsiccation, as occurs with the 

 similar sulphurets of rhodium and of platina. When distilled in 

 a small apparatus to expel the excess of sulphur, the sulphuric 

 acid, and moisture, the remaining protosulphuret of copper gives 

 the quantity of copper with more exactitude. 



But let us return to the mixture of oxide of copper with oxide 

 of nickel. What I have already said of the analogy of these twa 

 oxides proves that the oxide of nickel, when precipitated by 

 caustic potash from a solution which contains copper, must con- 

 tain some of this metal, a part of which, however, still remains 

 dissolved in the ammonia, unless a great, excess of potash be 

 added. It is, however, very easy to separate the copper from 

 the nickel by sulphuretted hydrogen, which precipitates the 

 former from its solution in an acid without acting upon the 

 latter. 



Oxides of Nickel and of Zinc. — The oxide of zinc dissolved by 

 ammonia is also precipitated by an addition of caustic potash ; 

 but it precipitates more slowly than the oxide of nickel, and 

 requires more potash. Its presence is discovered in oxide of 

 nickel by reducing the latter by means of soda in the flame of 

 the blowpipe. If it contains zinc, the charcoal is covered with 

 a white incrustation of oxide of zinc ; but to effect this, it is- 

 necessary to use a strong heat. In an analysis of a metallic 

 mixture which I performed a long time since, I endeavoured to 



