of Native Iron found in Brasil. 425 



configurations as are usual in the fracture of octohedral crystals, 

 and are found in many simple native metals. 



The magnetic qualities of the fragments, fortunately, enable 

 us to appreciate rightly those of the entire mass from which 

 thev have been detached ; for tiiough the mass, when tried upon 

 the' spot bv Mr. Mornay, gave indications of having distmct 

 N. and S. poles, it is prettv clear that tiiese were only so by in- 

 duction, in consequence of 'position with respect to the magnetic 

 meridian. For though the fragments are not in the least at- 

 tractive as magnets, and have in themselves no polarity, they 

 are precisely like anv other pieces of the best soft iron, and as- 

 sume polarity instantlv, according to the position in which they 

 are held with respect to the magnetic axis of the earth. Wheu 

 a long fragment is held in a vertical position, its lower extremity 

 being then within 20^ of the dip of the N. magnetic pole, be- 

 comes N. and repels the N. pole of a magnetic needle suspended 

 horizontally. But this power is instantly reversed l)y bemg 

 suddenly inVerted. So that the apparent contradiction betvyeen 

 the observed polarity of the mass, and the seeming want of it m 

 the fragments, is thus completely removed. 



Althoush Mr. Mornav reasonably expected that this iron 

 would not differ from the many others now on record that have 

 been found in various parts of the world, and from his experi- 

 ments was led to infer the presence of nickel, it appeared de- 

 sirable to ascertain this point with more precision than he had 

 been enabled to do, and to determine also in what proportion 

 this peculiar ingredient of meteoric bodies might be found to 



prevail. i , r 



I believe the means which I am accustomed to employ tor 

 detecting the presence of nickel in native iron to be new, and 

 may deserve to be described, on account of the very small (pian- 

 tity of the iron required for this mode of examination. 



Having filed from my specimen as much as 1 judged sufficient 

 for my purpose, (whicli need not exceed , l^ «f ^ gi'^dn,) 1 dis- 

 solved it in a drop of nitric acid, and then evaporated the solu- 

 tion to drvness. A drop or two of pure ammonia was then 

 added to the dried residuum, and gently warmed upon it in or- 

 der to dissolve any nickel that might be present. The trans- 

 parent i)art of the fluid was then led by the end of a rod of glass 

 to a small distance from tiie remaining oxide of iron, and the 

 addition of triple prussiate of potash immediately detected the 

 presence of nickel by the appearance of a milky cloud, which 

 was not discernible by the same means from a similar quantity 

 of common wrought iron tried at the same time. 



For the determination of the quantity of nickel I employed a 



different 



