440 M, Berzelius on the [Junk, 



nickel, or cobalt, when the acid is concentrated. This solu- 

 tion of the sub-arseniate was decomposed by caustic potash, 

 leaving 0*82 part of sub-arseniate, which I have mentioned 

 above. The potash had, therefore, taken up 0*16 part of 

 arsenic acid. The 0*82 part of subarseniate, produced by 

 the potash, contained 0*764 of oxide of iron, and conse- 

 quently the 0*98 part are equivalent to 0*1 IG of metallic 

 arsenic, and 0'529 of iron. 



f. The ammoniacal liquid, mixed with the caustic potash, 

 furnished 3-44 parts of oxide of nickel, equivalent to 2*7 

 parts of the metal. The liquid, separated from the oxide of 

 nickel, was evaporated until the ammonia was completely vo- 

 latilized, without depositing any thing. It was afterwards 

 mixed with the alkaline liquid, resultino- from the precipitation 

 of the oxide of nickel, and the subarseniate of iron. This mix- 

 ture now contained arsenic acid. It was neutralized by means 

 of muriatic acid. 



d. To separate the arsenic acid, I used M. Berthier^s me- 

 thod. I dissolved ten parts of common iron wire in muri- 

 atic acid, and poured the solution into the preceding, and pre- 

 cipitated with an excess of ammonia. The precipitate, washed 

 and heated to redness, weighed 22*4 parts. But ten parts of 

 metallic iron, which in general contain 0*0005 of its weight 

 of carbon, give 14*35 parts of peroxide. Subtracting 14*35 

 from 22*40, gives 8*05 parts for arsenic acid, which, added to 

 the 0*116 part obtained in 6, gives a result of 8*166 parts of 

 arsenic acid, equivalent to 5*332 of metallic arsenic. 



The analysis therefore gives 



Arsenic 53*32 



Sulphur 14*40 



Nickel 27*00 



Iron , 5*29 



100*01 



The difference between the two analyses is inconsiderable, and 

 shows that the analytic methods agree very nearly : especially if 

 we consider that the sulphur, determined by the same method in 

 the two experiments, varies in proportion, and appears to indi- 

 cate a difference of composition in the ores. Still as the arsenic 

 •acid could not be entirely separated from the peroxide of iron, 

 the result is always subject to uncertainties, with respect to the 

 relative quantities of the iron and arsenic. 



(C.) Analysis by Means of Nitromuriatic ylcid and Acetate of 



Lead. 

 This analysis, as well as the following, were made on pieces 

 taken from a specimen coming also from Loos ; but as I have 

 said above, forming another variety of the nickel ore. 



