THE METALLOGRAPHY OF METEORIC IRON 85 



The composition of Locust Grove is consistent with tiie foregoing 

 conckisions, as it contains 0.02 percent carbon and 0.18 percent 

 phosphorus. The inclusion evidently marks a spot of local enrich- 

 ment in which the percentages of either or both were higher than 

 the above averages for the mass. 



A second example of an iron-carbon structure (pi. 66) was found 

 in Chesterville — hke Locust Grove, a nickel-poor ataxite and with 

 approximately the same composition. At the center of a spot of 

 carbon enrichment is a rounded area that is substantially identical 

 with a white phosphoretic cast-iron, dendrites of a pearlitic nature 

 in a white groundmass of steadite. This is surrounded by an area 

 of dark pearlitic grains with white ferrite (kamacite) along their 

 boundaries, the equivalent of a steel with about 0.55 percent carbon. 

 Farther from the center this gives place to a lighter granular struc- 

 ture, the equivalent of about a 0.25 percent carbon steel, in which 

 the ferrite (kamacite) predominates. 



Tliis unique structure may have been due to the presence in the 

 original iron (before the heat alteration which turned it into an 

 ataxite) of a body of cohenite em-iched with phosphorus in the form 

 of sclu-eibersite inclusions, thus maldng possible the production by 

 fusion of an iron-carbon structure and also of a limited amount of 

 the eutectic steadite. 



Of three analyses of this iron by Sjostrom (in Cohen, 1908) two 

 shoAved 0.34 percent phosphorus and 0.02 percent carbon; one of the 

 three showed no phosphorus, and one no carbon. The variations 

 apparently were due to the presence or absence of cohenite or 

 sclireibersite in the samples analyzed. 



Origin of such structures. — It is to be noted that the only two irons 

 in which any kind of iron-carbon Structures were found are nickel- 

 poor ataxites, probably formed by the strong reheating of a hexa- 

 hedral iron which happened to contain inclusions of cohenite, fol- 

 lowed presumably by relatively rapid cooling. Thus the conditions 

 of production would in somic measure resemble those of artificial 

 irons, where the rate of cooling is too rapid for the segregation of 

 cementite. 



With annealing the pearlite in artificial iron breaks doAvn, the 

 cementite segregating in spheroids (pi. C), but with the utmost 

 attainable slowness of production, such particles are minute and 

 thickly dispersed. In normal meteoric irons, on the other hand, 

 the cooling was probably of secular slowness; and thus it was pos- 

 sible for the carbide to segregate completely in relatively large and 

 scattered cohenite crystals, and no pearlite remained, even though 

 the average content of combined carbon might have been such as 

 to produce some pearlite under artificial conditions. 



