OP ARTS AND SCIENCES. 



495 



Fig. 10. Cranberr}' Plains. 



faces of a very perfect octahedron, drawn of original size, from the 

 Cranberry Plains (Poplar Camp) iron. It will be seen by this sketch 

 that the octaliedral outline has been sharply formed ; but while 



many of the Widmanstattian plates are 

 parallel to this outline, there are others 

 which are markedly curved. These curved 

 plates must have originally formed through 

 the liquid mass as true planes, like their 

 neighbors, and have been bent in the sub- 

 fequent solidifying of the remaining mate- 

 rial. For, if they had been distorted by 

 an exterior force, the regularity of the octahedron would have been 

 at the same time destroyed. 



This phenomenon of curved plates has been made a basis of sub- 

 division in the classification of iron meteorites. But it is evident from 

 the figure that the bending is not a constant or essential character, 

 as the plates are not all curved, but only appear so at particular parts 

 of the mass, and some specimens show no trace of such a feature. In 

 a word, such curved plates are simply accidents of the crystallization. 

 Very similar bending is common in various minerals, as in gypsum, 

 and very conspicuously in the beautiful crystals of stibnite which 

 within the few last years have been brought from Japan. 



A similar remark might be made in regard to the swellings formed 

 in the kamacite plates around inclusions of troilite, Schreibersite, 

 and the like, which have been so minutely described by Reichenbach, 

 and named by him " Wulsteisen." * These again are accidents of 

 crystallization, which have their counterpart in other crystals, and 

 are most beautifully shown on the plates of mica from Chandler's 

 Hollow, Delaware, where the depositions of magnetic oxide of iron 

 on the planes of the crystalline growth of the mica produce effects 

 which imitate in a most striking manner the Widmanstattian figures. 



We give in Fig. 11 as perfect a representation of one of these mica 

 plates as could be obtained by the photographic process used in illus- 

 trating this paper, and it will be seen that the resemblance is very 

 close. The analogy here is far more than superficial, and shows, as 

 we conceive, the mode of action by which the Widmanstattian figures 

 were produced. This will be evident if it is borne in mind that the 

 figures on the mica plate are also sections of planes of crystalline 

 growth on which the particles of oxide of iron eliminated during the 



* Poggendorff's Annalcn, 1801, Bd. cxiv. j). 477. 



