W. II. Melville — Jose-phinite, a new Nickel- Iron. 513 



calculated from the total sulphur, 0'22 per cent. Hydrogen sul- 

 phide was liberated in considerable quantity when the pow- 

 dered pebble was treated with diluted chlorhydric acid. Mil- 

 lerite does not act thus. 



Copper. — The copper is not a constituent of the pyrrhotite, 

 but belongs to the nickel-iron. Copper minerals were not 

 detected under magnifying power. 



Arsenic. — Arsenious oxide was condensed in the cool parts 

 of a tube through which air was allowed to pass, as also in all 

 the tubes used for the determination of volatile matter. No 

 sulphide of arsenic dissolved by digesting in the cold with 

 ammonium carbonate, nor could niecolite (nickel arsenide) be 

 detected. Arsenic forms a part of the metallic portion of the 

 pebbles. 



Chlorine. — The chlorine was determined in a nitric acid 

 solution of the original powder, a blank experiment with the 

 use of the same quantity of reagents being run parallel with 

 that. In an aqueous solution were found iron, nickel, magne- 

 sium, sodium and chlorine, and this fact together with the 

 small available quantity of chloride caused the difficulty in 

 determining to which element chlorine was combined. The 

 presence of iron and nickel, since pyrrhotite carries nickel, can 

 be accounted for in the aqueous solution by the easy oxidiza- 

 bility of the sulphuret with free access of air. Green ferrous 

 chloride could not be detected in the powder, nor any soluble 

 salt possessing that color. The permanence of the pebbles in 

 air is also a presumption in favor of the absence of lawrencite. 

 Nickel chloride has not before been observed, but may pos- 

 sibly exist in these specimens. It is more probable that the 

 chlorine is united with sodium O04 per cent requiring 0*026 

 per cent ]STa or 0-035 per cent JSTa 2 0. The magnesium is in 

 the form of carbonate 



Water and Volatile Matter. — The powdered substance in 

 a platinum boat was heated in a glass tube in a stream of 

 dry carbonic anhydride, and the water, both hygroscopic mois- 

 ture and water of constitution, was weighed in a chloride of 

 calcium tube. Another portion was heated in hydrogen gas, 

 and hence the oxygen in the oxides was known, although this 

 amount might have included some, if not all, of the oxygen 

 combined with iron in the ferric state in the silicates. After 

 burning in a current of dry air and then reducing in hydrogen 

 the total loss was ascertained from the difference in weight of 

 the contents of the boat before and after the operation. In 

 all cases except when burnt in air a brownish cloud possessing 

 an empyreumatic odor was driven off without leaving a brown 

 black sublimate in the tube. It was a volatile organic sub- 

 stance whose nature was not discovered. The pulverized peb- 



