Algodoniie and some artificial copper-arsenides. 441 



ing concentr. nitric acid dissolves the mineral without leaving 

 a residue, forming first a green, then a murky blue solution 

 (owing to the interference between nickel-nitrate green and 

 cobalt-nitrate red). HC1 does not act upon the mineral. The 

 quantitative analysis was made by several methods : 



(a) Nitric solution made ammoniacal, alcohol added and 

 magnesia mixture. This is not a good method, because part of 

 the nickel and cobalt enter into the magnesium-ammonium 

 arsenate. 



(h) Nitric solution made ammoniacal, diluted properly and 

 IPS passed into hot solution to saturation. Filtrate evapo- 

 rated to dryness, residue oxidized with HNO 3 and magnesium- 

 ammonium arsenate precipitate. Separation of copper from 

 nickel and cobalt by H 2 S ; separation of nickel and cobalt by 

 nitrite method. 



(c) Powder fused with nitrate and carbonate of sodium. 

 This gives very good results for arsenic ; but for the fine pul- 

 verulent condition of the oxides these latter must be carefully 

 handled to avoid loss. 



I find that the magnes. amm. arsenate can be heated on the 

 asbestos pad of a Gooch crucible without loss into pyro-arse- 

 nate. From four analyses of perfect material I obtain 



Cu 61-67 



Ni . 7-03 



CoO 2-20 



Fe .. trace 



As 28-85 



99-75 



hence the atomic ratio 



Cu 0-9803 



Ni 0-1200 



Co 0-0373 



1-1376 2-958 (Cu, Ni, Co) 3 



As 0-3847 1-000 As 



This is an exact ratio of 3:1, the ratio of domeykite. I 

 look upon this as a case of isomorphic replacement. We may 

 expect to find all gradations of replacement within the ratio of 

 3 : 1. Synthetic experiments lead me to consider this ratio as 

 representing an especially strong one, as of bonds well satisfied ; 

 because I have been enabled to obtain it artificially in well 

 defined crystals, of which more hereafter. The molecule 

 Cu 2 As is the strongest of all, the other ratios seem to partake, 

 more of the nature of alloys, of unsatisfied bonds. It may be 

 urged that a mechanical mixture of the copper arsenide with 



