276 



Mr. Levy on anew Mineral Substance* [April, 



Fig. 1. 



Fig. 2. 



represented by fig. 2 ; in some of thern, however, the plane m 

 and its opposite are wanting, so that the prism is then six-sided 

 instead of being eight-sided. These crystals cleave easily with 

 brilliant surfaces parallel to the planes p and t, fig. 2. There is 

 also an indication of cleavage parallel to the plane m. All the 

 natural planes, as well as those obtained by cleavage, are suffi- 

 ciently brilliant to allow the use of the reflective goniometer for 

 the measurement of their incidences. From the measurements 

 I have taken, and the cleavages already mentioned, I am induced 

 to take for the primitive form of this substance a doubly oblique 

 prism, fK r . 1, in which the incidences of/) on m and t are respect- 

 ively 92 o0 34 / and 88°, that of m on t 112° 30', and in which the 

 three edges d, f, h, which meet at the solid angle o, are to each 

 other nearly as the numbers 13, 20, 11. The incidences of p on 

 m and t are nearly supplement of each other, the only difference 

 being 34', hence the primitive form differs but little from an oblique 

 rhombic prism ; for if these two angles were exactly supplement 

 of each other, then the incidences of p on m, and on the face 

 behind parallel to /, would be equal, and consequently the pri- 

 mitive would, at least so far, have the character of an oblique 

 rhombic prism. There is another incidence which might, 

 without a proper attention, lead to the same conclusion respect- 

 ing the nature of the primitive. It is the incidence of the planes 

 h l and g", fig. 2, which is equal to 89° 20' ; that is to say, very near 

 a right angle. Now if the planes g 2 and A 1 were considered as 

 the diagonal planes of a rhombic prism, they should be perpen- 

 dicular to each other. These indications of an oblique rhombic 

 prism, as the primitive form of this substance, are, however, car- 

 ried no further, and are entirely destroyed by the want of the 

 symmetry which should accompany them. The faces A 1 , g\ if 

 ihe diagonal planes of a rhombic prism, should be equally inclined 

 upon the two lateral planes which they meet ; and here we find 

 that the incidences of A 1 , with the planes m and t, as well as 

 those of g- with m, and the plane parallel to t, differ widely from 

 each other. The occurrence of the plane </' without being 

 accompanied by a plane replacing the edge of intersection of 

 plane p with the plane parallel to t, is also incompatible with an 

 oblique rhombic prism. No doubt can remain, therefore, as to 



