136 Proceedings. 



formation under different circumstances, and in larger 



quantities, appears interesting. 



In the rebuilding of a flue to a furnace in which arsenical 



nickel-cobalt ores had been smelted, numerous globules of 



a metal were found intermingled with half-fused brick. An 



analysis gave : — ' 



Calculated 

 Found. for NioAs. 



Nickel 6773%] 



Co ,. i-26„ 



Cu I"20„ 



Fe 0-23,, 



Arsenic 28-14 „ ^ 



S 0-45 „ J 



•70-42% 70-2^ 



.28-59% 29-8% 



99-01 



The pieces were dull on the surface, due probably to oxida- 

 tion with a corresponding loss of arsenic. 



A specific gravity determination at 20° C. gave 8-668. 



Each globule seems to have contracted considerably on 

 cooling, and has a crystalline structure. Having been left 

 in a damp atmosphere, there formed on the surface small 

 green particles which were insoluble in water but soluble in 

 acids and ammonia. This, I thought, might be the green 

 arsenate formed by direct oxidation, but in the few milli- 

 grammes obtained after several months exposure I did not 

 succeed in detecting arsenic. 



As to its origin, the ore which had been smelted, and 

 the metal it yielded, contained cobalt and nickel in the pro- 

 portion of one to two. I cannot discover that anything 

 similar has before been noticed after working ores containing 

 much nickel, but it is certain that nothing has ever passed 

 over to the flues from cobalt ores, and this might be inferred 

 from the different behaviour of the two metals in the 

 present case. 



