ART. 21 OXIDATION OP METEOEIC IRONS SHANNON 7 



oxide and again titrated for total iron. The insoluble was filtered 

 out, ignited and weighed, after which citric acid was added to the 

 solution and the nickel determined with dimethylglyoxime. Water 

 was determined as loss on ignition of a separate portion and cor- 

 rected for oxidation of the ferrous iron. The results were as follows : 



Insoluble 0. 36 



Fe203 79. 50 



FeO 3. 68 



NiO 6. 44 



H2O 8. 19 



Undetermined 1. 83 



100. 00 



Mineralogically combined, following the same assumptions as used 

 in the interpretation of Tassin's analyses above, the figures give the 

 following : 



Limonite (2Fe203.3H20) 56.73 



Magnetite (FeO. Fe^Oa) 11.86 



Trevorite (NiO.FesOs) - 20.03 



Fe203 (excess) 8.36 



Insoluble (SiOs etc.) .36 



97.34 



In the absence of appreciable amounts of schreibersite and of any 

 residual unoxidized metallic iron, the magnetism exhibited by the 

 material must be attributed to the oxides themselves. The 8.36 per 

 cent of free Fe^Og indicated above is very probably not present as 

 such, but is in the limonite, the variability of that mineral in water 

 content making water a very unsatisfactory index to the amount of 

 hj-droxide present. The ferrous oxide, equivalent to only 11.86 per 

 cent of ordinary magnetite, is not sufficient to account for the degree 

 of magnetic susceptibility. If the nickel oxide be combined as 

 ferrite, it adds 20.03 per cent of trevorite, giving a total of 31.89 

 per cent of magnetic ferrites. The excess of FcsOg may or may not be 

 the ferromagnetic form of this compound, but its amount is small 

 enough to be disregarded, especially as it is quite probably limonite, 

 as stated above. If, however, the existence of magnetic nickel ferrite 

 is not admitted, it is necessary to assume the presence of ferromag- 

 netic FcoOg to account for the properties of the material. 



There is considerable evidence in support of the assumption of the 

 existence of nickel ferrite in this material. J. Lawrence Smith ^ as 

 long ago as 1875 pointed out certain magnetic peculiarities of ferric 



''J. Lawrence Smith. Singular anomaly of the sesquioxide of iron as prepared from' 

 meteoric iron. Amer. Chemist, vol. 5, 1875, pp. 356-358 ; Chem. News, vol. 31, 1875, pp. 

 210-212; Compt. Rend., vol. 53, 1875, pp. 301-304. 



