336 Mr. W. Francis's Examination of a crystallized Nickel Ore, 



lized in octahedrons, and is deposited from the smalt fur- 

 naces ; the former is crystallized in large laminae, very similar 

 to the second cobalt ore described in the preceding Article, 

 and is evidently the product of a further process of purifica- 

 tion : nor was it obtained directly from a smalt-work, but 

 from the German silver manufactory of Mr. Henniger in 

 Berlin. This process of purification seems to resemble that 

 which is practised at Modum with the cobalt ores. 



The analysis was performed in the laboratory of Professor 

 Henry Rose; and the course pursued quite similar to that 

 employed for combinations of arsenic. The nickel and iron, 

 however, were not separated by succinate of ammonia, but 

 determined according to Dr. Scheerer's method. The result 

 of the analysis was 



Sulphur 1-01 



Arsenic 34*07 



Nickel 52-58 



Cobalt 3-28 



Iron 10-06 101-00. 



The peroxide of iron obtained was again dissolved in hy- 

 drochloric acid, and precipitated by ammonia in excess. Sul- 

 phuret of ammonium produced no perceptible alteration in 

 the filtered ammoniacal fluid. From the obtained oxide of 

 nickel, however, 0*009 grammes of nickeliferous peroxide of 

 iron were separated by solution and the addition of ammonia : 

 which quantity had not been separated by the above method. 

 In this separation, therefore, Dr. Scheerer's method gave 

 a better result than that with succinate of ammonia in the 

 analysis of the cobalt ore described in the preceding Article. 



If it be admitted that the slight quantity of sulphur takes 

 the place of a part of the arsenic, the atomic weight of the ar- 

 senic is to that of the other metals as 

 7*74 : 18-10, 

 therefore, exactly as in the combinations detailed in the pre- 

 ceding Article. The formula for this nickel ore would ac- 

 cordingly likewise be 



Co 7 Us 8 



Fe 7 J 

 In this case the atomicproportion should be 



7*74 : 18-06, 

 coinciding, therefore, very nearly with that given by the ana- 

 lysis. \\\ on the other hand, the formula 



Ni 5 l 



Co 5 ^As 2 



Fe 5 



