﻿Huntington — Crystalline Structure of Iron Meteorites. 293 



chiefly characterized by the system of fine lines seen here par- 

 allel to c b. The lower of the two drawings was made from 

 the original specimen, without a knowledge of what the angles 

 were ; but on examination it was found that the lines, and 

 prism faces as well, could all be referred to a cube with twin 

 members on all the trigonal axes. 



The face of such a twin is represented by the diagram be- 

 tween the two figures of the crystal, and the upper figure was 

 actually drawn from the cube diagram by means of a parallel 

 ruler, and it will at once be seen from the close coincidence of 

 the two figures that the upper one must be correct, the lower 

 one only varying within the limit of error of a mechanical 

 drawing. 



On examining the Hauptmannsdorf iron, which has long 

 been known to break with a cubic fracture, a cube face under 

 the microscope appeared, as shown in fig. 7, some of the lines 

 being parallel to the cube edges and others forming diagonals 

 of the cube face. These diagonals might ^ 



be octahedral lines or dodecahedral lines, 

 or they might be lines of cleavage parallel 

 to the face of the other individual of a 

 twin, and as the face shown in section re- 

 placing the cube angle at a was the face of 

 a twin cube, the lines parallel to that edge 

 would probably be due to the same plane ; 

 and this was proved to be the case, as on an 

 adjacent face the same lines followed the Hauptmannsdorf 

 direction of twinning. These lines of twin- (Braunau.) 



ning are not represented in fig. 7, to prevent confusion. 



Thus the Hauptmannsdorf iron appears to be purely cubic 

 in structure, while the Tazewell appears to be purely octahe- 

 dral. However, as the octahedron was observed in the 

 Coahuila iron, and the dodecahedron appeared in the Butler 

 iron, it became a question whether the cube and dodecahedron 

 could not be found in the typical octahedral irons. A very 

 large number of irons were studied with this end in view, but 

 nothing could be determined with specimens where only one 

 face could be examined. It was necessary to have two known 

 crystal faces, and to be able to follow the plates over an edge. 

 For on a cube face the octahedral plates give rectangular inter- 

 sections, while on an octahedral face the cube plates would give 

 intersections parallel to the octahedral edges. 



Figure 8 shows of original size a section of the well-known 

 La Caille meteorite, cut parallel to an assumed cube face, the 

 direction being determined by external well-developed octahe- 

 dral faces. Here the rectangular intersections a c and c d 

 result from octahedral plates intersecting the cube face, but 



