Fig. 3. 



1823.] Mr. Levy on a new Mineral. 133 



formula will give the value of A, and thus the dimensions of the 



primitive form will be completely determined. There will then 



be no difficulty in calculating from the measured incidences of 



the other faces, either on m, or some other known plane, the 



indices of the decrements from 



which they are derived. Thus I 



find that the primitive form may 



be supposed to be an oblique 



rhombic prism, fig. 3, in which 



the incidence of m onmis 1 15° 45', 



that of m on p 91° 25', and one 



side of the base is to the lateral ^ 



edge nearly in the ratio of 15 to j 



16. The letters placed on the 



faces of the crystal, represented 



in fig. 4, indicate respectively the 



decrements which produce them, 



and their incidences on p and ware 



as follows : 



Incidence of h x on p 91 



e 1 on p . 

 e* on p . 



o 2 on p . 



i 



a on m 



128 

 148 

 138 



157 



138 



b l on m 



3 



Z»" on m 127 



b- on m 1 19 



41' 

 45 

 19 

 3 



6 

 55 



6 

 13 



30" 















30 

 10 



30 

 



I was unable to measure the faces marked i in the drawing : 

 they result from an intermediary decrement. The intersections 

 of these faces with the two faces m and o 2 appear to be parallel. 

 Therefore, by supposing the face o- the result of a decrement by 

 one row instead of two, the faces i would be produced by simple 

 decrements on the lateral angles o of the prism. In this hypo- 

 thesis, the ratio of h to d would be half what I have taken it, 

 and the indices of all the modifications observed, except 

 that of h' should also be divided by 2. The oblique rhom- 

 bic prism, the dimensions of which have just been stated, does 

 not possess the property which, according to Haiiy, belongs to 

 all primitive forms of this kind, and which consists m their hav- 

 ing such dimensions that the diagonal joining the solid angle o r 

 fig. 3, with its opposite, is perpendicular to the edge h. But m 

 the present case, the incidence of/j on m being nearly 90 , it is 

 very obvious that if the ratio between h and one side of the base 

 of the prism were determined by such a condition, it would be 

 considerably less than I have made it ; and consequently that 

 even on the supposition that the observed modifications could be 



