British Mineralogy. 175 



to the borate, and, if the treatment is prolonged, ultimately 

 leaves a globule of nearly pure disulphide of nickel (Ni 2 S), 

 which gives the reactions for nickel when further treated with 

 borate in the oxidatirig-flame; any copper present in the mineral 

 remains in this disulphide. The mineral, if calcined on charcoal 

 in the oxidating-flame at a heat below its fusing-point, reacts for 

 nickel and iron when treated with borate. The cobalt was found 

 to be present in so small quantity in the specimen analyzed, that 

 its reactions were quite overpowered by those of the nickel and 

 iron, and it could not be detected until a special search for it 

 was made by Plattner's arsenical process. 



The qualitative analysis showed the presence of sulphur, iron, 

 nickel, traces of copper and cobalt, along with a little rock-mat- 

 ter, which consisted chiefly of quartz, but which contained some 

 magnesia in the form of a silicate easily decomposed by acids. 



The quantitative analysis was conducted as follows : — A 

 weighed quantity was dissolved in nitrohydrochloric acid, which 

 left a small amount of insoluble matter, along with some free 

 sulphur, which was estimated by the loss found on igniting the 

 insoluble matter. The solution (having previously been evapo- 

 rated to drive off any large excess of acid) was now precipitated 

 by chloride of barium, and the resulting sulphate of barytes de- 

 termined. The sulphur calculated from the weight of this sul- 

 phate of barytes, in addition to the amount of free sulphur pre- 

 viously found., gave the total amount of sulphur present in the 

 mineral. 



The nickel and iron in the filtrate were separated by carbonate 

 of barytes, and the iron in the precipitate determined, as usual, as 

 sesquioxide. The solution containing the nickel, along with a 

 little magnesia, was first precipitated by sulphuric acid in order 

 to remove the excess of the baryta, which was filtered off, and 

 then the nickel was precipitated by colourless sulphide of am- 

 monium with the usual precautions. This sulphide was, after 

 incineration, dissolved in nitrohydrochloric acid, and the oxide 

 precipitated by caustic potash boiling. Upon examination, only 

 a trace of cobalt was found present in this oxide. In the solu- 

 tion from which the sulphide of nickel had been separated, the 

 magnesia, after removal of the excess of baryta by sulphuric acid, 

 was thrown down as phosphate of ammonia and magnesia, and 

 determined as pyrophosphate of magnesia as usual. 



20*07 grs. mineral afforded 0*52 gr. free sulphur, 0*05 gr. in- 

 soluble matter, 51*53 grs. sulphate of barytes, 14*30 grs. sesqui- 

 oxide of iron, 2*85 grs. protoxide of nickel, and 0*54 gr. pyro- 

 phosphate of magnesia ; and from these results the percentage 

 composition is calculated to be as follows : — 



