1893.] NATURAL SCIENCES OF PHILADELPHIA. 375 



spot with peculiar markings within. The straight lines which cross 

 each other are cleavages and have, in all probability, a structural 

 meaning. If we focus into the deep pits of the specimen, crystalline 

 bodies appear, of which a few have been drawn, (figure 3). These 

 solid crystals, sometimes show a pale yellow color and may be 

 troilite ; some, however, have a deep blue tint, similar to the thin 

 coat of oxide produced on heating iron or steel much below 

 redness. 



The microscopic investigation, which could be made only by 

 reflected light, revealed some additional points. The mass is so very 

 cellular and the cells, if the term be allowed, so close together as to 

 require magnifying power in studying them. The quasi cells are 

 filled out with the other material, thus constituting the whole mass. 

 It is not difficult to recognize the faint brass (or yellow) colored 

 troilite and the jet-black magnetite, the latter sometimes filling cells 

 having a distinct crystalline contour as shown in figure 4. 



The nickel is probably not evenly divided through the whole of 

 the metallic mass, as part of the metal is readily affected by the 

 nitric acid, while the rest is not. The tempered steel meteorite is, 

 therefore, not homogeneous but highly compound. 



Results were obtained by the qualitative chemical analysis show- 

 ing conclusively the distinct difference between the granules and the 

 separated dark powder. The former contains a sulphuret, probably 

 troilite ; the latter contains no sulphuret but, instead, a sulphate. 



Iron, nickel, sulphur, traces of carbon, chlorine, phosphorus and 

 chromium were found; also a silicate in which were recognized lime 

 and magnesia. Copper and cobalt were searched for, but in vain. 

 This leads me to direct attention to the Disco Island terrestial iron, 

 in which, according to Professor A. E. Nordenskiold and J. L. 

 Smith, are found copper, cobalt, phosphorus and comparatively 

 large quantities of carbon, differences too great to be overlooked 

 in comparing analytical work. 



Composition of the granules : 



Iron .... 86'66 per cent. 



Nickel .... 2 - 32 per cent. 



Sulphur .... 0*19 per cent. 



x Silicate . . . 4*41 per cent. 



Oxygen . . . 6 - 42 per cent. 

 If the sulphur is combined with iron to form troilite and the oxygen 

 to form magnetite we have : 



