44 



BULUETIN 184, UNITED STATES NlATIOOSilAIi MXPSEUM 



5 should drop vertically from a point corresponding with about 6 

 percent nickel. It is approximately so drawn in the early diagram 

 of Osmond and Cartaud (fig. 4), and it was accepted by Guertler 

 and other older authorities. It also appears, at a somewhat higher 

 percentage, in the recent diagram of Bradley and Goldschmidt 

 (fig. 7), which is presently discussed. 



The effect of cobalt on the transformation of iron is virtually 

 identical with that of nickel, so the iron-nickel system may be as- 

 sumed for practical purposes to be an iron-nickel-cobalt system. 

 Therefore the cobalt content of a meteoric iron should always be 

 considered in connection with its transformations, and an iron con- 

 taining 14 percent of nickel and 1 percent of cobalt may be judged 

 in the same way as if it contained 15 percent of nickel. 



Bradley and Goldschmidt 's diagram. — Yet another iron-nickel dia- 

 gram should be noted — that offered by Bradley and Goldschmidt 

 (1939) as the result of annealing experiments and of X-ray powder 

 photograms of natural and artificial iron-nickel alloys. The diagram 

 is shown in figure 7. 



15 20 25 



PCRCCNT NICKEL 



Figure 7. — Bradley and Goldschmidt's iron-nickel diagram. 



Affirming the conclusion of Owen and Williams (1938) that octa- 

 hedral structures in meteorites represent a kamacite-taenite equilib- 

 rium between 350° and 580°, they assert that a further transformation 

 occurs along a line drawn at 350°, which coincides with the eutectoid 

 line in the earlier diagrams of Osmond and Cartaud, Bendicks, and 

 the Hansons. 



The very narrow gamma-alpha range between percent and about 

 9 percent nickel in their diagram is open to question as applied to 



