MERRILL AND TASSIN] THE CANYON DIABLO METEORITES 



209 



the surface, and that this moisture is brought back by capillarity and 

 evaporated rather than drained off at lower levels, as in more humid 

 regions. An iron thus buried, even though rich in chloride and sul- 

 phide, would therefore endure for a long period. 



Part II. 



By WIRT TASSIN 



A. The Shale-ball Iron 



Several sections of the "shale balls" having iron centers were pol- 

 ished and etched. These etched surfaces were apparently so differ- 

 ent in appearance from the ordinary Canyon Diablo irons that had 

 their locality not been known they would not have been regarded as 

 parts of the same fall. Sections of the Canyon Diablo meteorite, 

 which had been in the possession of the Museum since 1894, were 

 found which showed areas having the usual very coarse octahedral 

 structure, with broad kamacite plates, together with small areas hav- 

 ing a structure closely comparable with those of the shale-ball irons 

 (pi. xx). Further, the analysis of the shale-ball iron compared with 

 those of the well-known Canyon Diablo agrees as closely as could 

 be expected in a heterogeneous mass subject to segregation. Thus: 



Analysis "A" was made by Moissan, 1 "B" by Booth, Garrett, and 

 Blair, 2 "C" and "D" by myself. The analyses "A" and "B" were 

 made prior to the discovery of the shale-ball irons, and, like "C," 

 which was made on an iron collected by Mr. G. K. Gilbert in 1892, 

 are of Canyon Diablo irons of the ordinary type. "D" is that of a 

 shale-ball iron. It will be noted that "A," "B," and "C" are distin- 



1 Compt. Rend., 1904, cxxxix, 776. 



' Proc. Phila. Acad. Sci., 1905, lvii, 875. 



