-MERRILL AND TASSXN] THE CANYON DIABLO METEORITES 205 



unaltered iron shows a structure quite unlike that of the typical iron, 

 as described (see pi. xx), a difference so striking that one unac- 

 quainted with the conditions under which they were found would 

 certainly be justified in pronouncing them independent falls. This 

 difference, which is further accentuated in composition, as shown in 

 Mr. Tassin's paper (p. 209), is very evident in the illustrations 

 (compare pis. xix and xx). 



As noted by Mr. Tassin and as is readily evident on even a casual 

 inspection, the shale iron differs from the other irons in the rela- 

 tively large proportion of iron chlorides and phosphides and in 

 the lack of the broad kamacite plates so conspicuous in pi. xx. So 

 abundant are the first mentioned that a polished surface soon tar- 

 nishes, the chloride, as is usual in chloride-rich irons, exuding in the 

 form of greenish drops which quickly oxidize, coating it with a layer 

 of "rust." Fig. 2 of pi. xviii and fig. 1 of pi. xix show polished 

 surfaces with oxidized coatings produced by an exposure of but a 

 few days in the atmosphere of this office. The common type of the 

 Canyon Diablo iron (pi. xx), on the other hand, is very stable, with- 

 standing the warm, damp atmosphere of Washington, for a pro- 

 longed period without serious oxidation. 



The writer believes that the discovery of these shale balls explains 

 the origin of the widely disseminated iron shale which is found on 

 the plain and which has been the subject of discussion by other 

 writers, particularly Derby, 1 Barringer, 2 and Farrington. 3 



The explanation now given is not new or, at least, not wholly so, 

 but is emphasized here and apparently substantiated by the finding 

 of the shale actually in contact with an unaltered nucleus and in 

 process of formation. The shale, according to this view, originates 

 through the oxidation of a peculiarly susceptible variety of the iron 

 and is not to be explained, as does Dr. Farrington, on the basis that 

 such were derived from buried irons, which would receive a larger 

 amount of water from being covered with soil and rock fragments, 

 and hence would oxidize more rapidly. That, however, the pro- 

 tective covering was instrumental in producing this particular type 

 of shale is probable, since the gradually oxidizing material would be 

 held in place rather than fall away to powder, as would be the case 

 if the decomposition took place on the surface. 



The presence of the shale scattered so widely over the plain, 

 usually in fragments rather than in the form of balls with metallic 

 nuclei, the writer would explain as follows : 



1 Am. Jour. Sci., vol. 49, 1895, p. 102. 



2 Proc. Acad. Nat. Sci. Phila., December, 1905, p. 879- 



3 Am. Jour. Sci. (4), 17, 1907, p. 300. 



