THE METALLOGRlAIPHY OF METEO'RIC IRON 83 



as that of the Cape iron would be stable above that line and might 

 be preserved to ordinary temperatures, even if cooling were very 

 slow, because of the rigidity of the mass. 



The phosphorus content of the Cape iron is slight from the metal- 

 lographic standpoint. Other high-nickel ataxites with similar struc- 

 tures contain as little, or even less (e. g.. Deep Springs 0.06 percent, 

 Iquique 0.05 or 0.07 percent, Illinois Gulch 0.07 percent, Guffey 

 0.02 percent), and sodium picrate etching reveals no phosphide in 

 them. 



References. — The relationships of phosphorus with iron, nickel, and 

 sulphur, and the resulting structures in artificial and also in meteoric 

 irons, have been treated exhaustively in the contributions of Vogel 

 and his collaborators. Those who wish to pursue the subject further 

 are referred to the publications of Vogel (1927, 1932, 1938), Vogel 

 and Tonn (1930), Vogel and DeVries (1931), and Vogel and Bauer 

 (1931), the titles of which will be found in Literature Cited. 



XV. CARBON AND SULPHUR IN METEORIC IRON 



Martensite and troostite. — Quantitatively carbon is a very minor 

 constituent in meteoric irons, appearing only in fractional percent- 

 ages and in most analyses not at aU. WTien present it is practically 

 always in the form of inclusions, either of graphite or of the carbide 

 cohenite, which were discussed in Chapter IV. Except for two 

 small areas of exceptional character, the author has observed no 

 instance of iron-carbon structures analogous to those produced 

 artificially. 



Some writers, however, have alluded to supposed iron-carbon 

 structures in meteoric irons, and Vogel (1927) reported and illus- 

 trated an example of what he termed martensite or troostite in Casas 

 Grandes. That iron contains 0.177 percent of carbon, of which 

 0.145 percent is in combined form. The latter percentage would 

 represent 2.175 percent of carbide, which in theory might make 

 possible a local development of martensite or troostite as reported 

 by Vogel. The author, however, regards the structure referred to 

 not as an iron-carbon but as an iron-nickel structm-e. 



The presence of carbon would have the same general retarding 

 effect upon the gamma-alpha transformation in natural nickel-iron 

 alloys as it has in artificial ones. As may be seen by the iron-nickel 

 diagram, however, the presence of nickel greatly lowers that point; 

 with 10 percent nickel it is depressed to 400°, and with about 25 

 percent to room temperatm'e. Martensite is produced by rapid 

 cooling from the A13 point (Chapter VII). Thus its preservation 

 would theoretically be possible in areas of carbon enrichment in 

 octahedrites with 7 percent or more of nickel. 



