56 BULUETIK 184, UNITED STATES NiATTONlAL MtPSEfUM 



range where atomic mobility is slight, the transformation will be 

 sluggish and incomplete. As a result there is much black plessite 

 and "spotted" taenite, consisting of a very fine and imperfectly un- 

 transformed aggregate of gamma and alpha particles. There is 

 also, with strong etching, more or less grayish darkening even of clear 

 taenite, due to traces of supersaturation, which produces a slight 

 admixture of a gamma-alpha aggregate. 



The time required for transformation increases very rapidly be- 

 low 500° as the nickel content rises, and experiments show that 

 even at that temperatm-e in alloys with more than 8 percent or 10 

 percent of nickel atomic mobility practically ceases. Therefore 

 when nickel exceeds that proportion, the gamma-alpha transforma- 

 tion cannot be completed artificially before the atoms "freeze in." 

 Below 8 or 10 percent nickel the gamma-alpha zone is traversed at 

 temperatures above 500° where there still is sufficient diffusion to 

 make equilibrium possible. 



There is no temperature, however, below which diffusion wholly 

 ceases, although the time required for a structural change may be 

 immeasurably great. As meteorites may have cooled through a 

 period of extreme and unknown duration, it is impossible to place 

 a temperature limit below which some change might not have taken 

 place. 



Fineness of octahedral structure. — The mechanism as described 

 is consistent with the fact that the fineness of the Widmanstatten 

 pattern increases with the nickel content. With a low nickel con- 

 tent transformation takes place at a relatively high temperatm-e 

 where diffusion is active. The kamacite therefore precipitates freely 

 in plates, which may reach a thickness of several centimeters, fol- 

 lowing the octahedral crystallographic planes of the gamma phase, 

 while the relatively small amount of gamma (taenite) migrates easily 

 to the planes of the plates where it segregates. With a high nickel 

 content transformation takes place in a lower range where diffusion 

 is more sluggish. In such case not only is taenite more abundant 

 but the diminished atomic mobility does not permit it to migrate so 

 freely; thus it segregates in lamellae more closely spaced. 



Thus the formation of thick plates (showing on a section as wide 

 bands), which took place easily with a lower nickel content and 

 higher temperature, becomes impossible with a higher content and 

 lower temperature. The progressive narrowing of the bands sub- 

 stantially parallels the rise in nickel content; but not exactly, as the 

 successive types of octahedrites overlap somewhat as respects com- 

 position and width of bands (fig. 8). With the same composition 

 greater rapidity in cooling would tend to produce a finer Widman- 

 statten structure. 



