64 BULUETEN 184, UNITED STATES NATIOiNiAIi MUSETJM 



Though Horse Creek resembles an octahedrite, its planes are not 

 octahedral, as the phosphide is tetragonal and because of the relation 

 of its space lattice with that of iron it would precipitate on other than 

 octahedral planes. This was true of a Widmanstatten structure pro- 

 duced by Meal, Barrett, and Jarabek (1934) from a solution of 3.2 

 percent phosphorus in alpha iron, though the exact crystal structure 

 of the precipitate was not determined. That the planes in Horse 

 Creek are not octahedral is apparent from plate 60, where deviations 

 from the octahedral planes are apparent. 



Transitional octahedrites. — When the octahedral structure is nor- 

 mally developed it is easy to assign an iron to its proper class by 

 simple measurement of the bands. The transitional low-nickel 

 and high-nickel types, however, give rise to differences of opinion. 



The distinction between coarsest octahedrites and granular hexa- 

 hedrites in some cases is doubtful, there being large areas with di- 

 versely oriented sheen in which some see traces of an octahedral 

 pattern and others do not. Ainsworth, for example, with 6.9 percent 

 Ni-Co, has always been classified as a coarsest octahedrite; yet the 

 author's specimen, a full-sized slice, has both the general appear- 

 ance and structure of a granular hexahedrite, with no suggestion of 

 an octahedral pattern. 



El Burro, a newly reported Mexican iron with 6.36 percent Ni-Co, 

 has a fairly recognizable octahedral pattern with bands one or two 

 centimeters wide. In Santa Luzia de Goyaz (Ni-Co 6.23 percent) 

 only local traces of an octahedral pattern are discernible, most of 

 the mass being made up of very large and irregular grains with 

 rosette inclusions of schreibersite similar to Ainsworth. In both of 

 these irons the Ni-Co content is no greater than that of some 

 hexahedrites. 



Mount Joy (Ni-Co 6.40 percent), which Linck and Brezina called 

 a hexahedrite and Berwerth an octahedrite, the author considers a 

 typical granular hexahedrite, its Ni-Co content being little above 

 that of a number of irons in that group. The same is true of Central 

 Missouri, the only available analysis of which gives the very low, 

 perhaps too low, percentage of 4.80 Ni-Co; its large grains, reaching 

 dimensions of inches, seem altogether irregular. This iron is re- 

 corded in museum catalogues as a coarsest octahedrite, though 

 Weinschenck apparently regarded it as a hexahedrite. 



New Baltimore (Ni-Co 6.74 percent) presents an anomalous 

 feature. It is made up of large grains, like Central Missouri and 

 others, which have a characteristic microstructure, with no visible 

 taenite. One large grain, however, is filled with a very distinct 



