Combinations of Arsenic "with Cobalt. 335 



in which case the atomic relation should be as 



7*74. : 18-06. 



Such a numerical relation appears at first sight, it is true, 

 somewhat unusual. The formula 



Co 5 ~1 

 Fe 5 y As 3 

 Cu 5 J 



would, perhaps, be more probable ; however, the atomic rela- 

 tion corresponding to it, viz. 



7*74 : 19-35 

 differs still more from the ascertained result. 



There sometimes occurs a third kind of crystallized arse- 

 nical cobalt in the smelting processes at Modum. It forms 

 dendrites of a kind quite similar to those of speiss-cobalt, or 

 such as are obtained as pure copper in the [aussaigerung] of 

 gun-metal*. This, therefore, would prove that its crystal- 

 line form would belong to the regular system. An analysis 

 which was made about seven years since gave approximately 

 the formula q q o -\ 



Fe 5 I As 

 Cu 5 J 

 or, more correctly, 



co 5 i rco 5 "i 



Fe 5 y S + 2 J Fe 5 y As 

 Cu 5 J LCu 5 J 



since beside 14- per cent, of arsenic, it contained 3 per cent, 

 of sulphur. The copper amounted to 7'86 per cent. How- 

 ever, the iron and cobalt were separated in a defective 

 manner, whence the former contained a considerable amount of 

 cobalt. The analysis needs therefore repetition ; although, 

 since iron and cobalt do not differ much in their atomic values, 

 scarcely any essential change in the formula might ensue. 



XLIX. Examination of a crystallized Nickel Ore. By 

 William Francis, A.L.S.f 



THE combination which is the subject of the present paper, 

 although consisting mainly of arsenic and nickel, yet has 

 to be distinguished from the kind of nickel ore analysed some 

 time since by Professor Wohler. Whilst the latter is crystal- 



* I obtained of Dr. Marchand, who is at present engaged in very inter- 

 esting experiments on alloys for fire arms, a piece of dendritic copper 

 bearing very great resemblance to the native. — Sch. 



f Communicated from Berlin by the Autiior. 



