182 Mr. D. Forbes' s Researches in 



silver, or when tested with nitroprusside of sodium. "When cal- 

 cined od charcoal and then treated with borate, it shows the re- 

 actions of nickel and iron, — the former reaction so strong, how- 

 ever, as generally to obscure the reactions both of the iron and 

 cobalt which it contains. The iron-reaction is evident, however, 

 by treating the raw mineral for only a very short time with borate 

 in the oxidating-flame. The cobalt-reaction is easily obtained 

 by fusing a portion of the calcined mineral on charcoal with bo- 

 rate in the reducing-flame of the blowpipe and adding a small 

 globule of gold, which, by uniting with the reduced nickel, re- 

 moves it from the glass and allows the cobalt-colour to show 

 itself distinctly. 



The qualitative examination showed the mineral to contain 

 arsenic, sulphur, nickel, and iron in large amount, along with a 

 smaller quantity of cobalt and manganese, as also some silica and 

 insoluble rock-matter from which a little magnesia had been re- 

 moved by the action of the acid. Traces only of copper were 

 found present. 



The quantitative analysis was conducted as follows : — ■ 



I. A weighed portion of the mineral in fine powder was well 

 mixed with six times its weight of a mixture of equal parts of 

 pure nitrate of potash and carbonate of soda, and fused in a 

 porcelain crucible; the resulting mass was boiled with hot water 

 repeatedly and filtered. This filtrate was acidified with hydro- 

 chloric acid, and then neutralized with ammonia in order to see 

 whether any substance precipitable by ammonia was contained in 

 the solution ; but as this was found not to be the case, the whole 

 was again acidified by hydrochloric acid, and a solution (in nitric 

 acid) of 12*46 grains of iron wire of known purity added to it, 

 after which the whole was precipitated by ammonia and deter- 

 mined as usual : the weight of the precipitate, less the amount of 

 sesquioxide of iron from the quantity of iron added, gave the 

 amount of arsenic acid equivalent to the arsenic in the mineral . 



From the filtrate, after being rendered slightly acid with hy- 

 drochloric acid, the sulphur was estimated by precipitation as 

 sulphate of baryta after adding chloride of barium in excess. 

 The iron was determined in the oxides remaining upon the filter 

 after having treated the fused residue with water; this was dis- 

 solved in hydrochloric acid, treated in the cold with carbonate of 

 barytes, and the precipitate containing the iron was dissolved in 

 sulphuric acid, filtered from the sulphate of barytes, and the ses- 

 quioxide of iron in the solution precipitated by ammonia and de- 

 termined as usual. An unfortunate accident prevented the other 

 substances being determined in the filtrate from the iron preci- 

 pitate. In order, therefore, to determine the nickel, cobalt, man- 

 ganese, and magnesia, another portion was taken and dissolved 



