52 Troost on the Pyroxene. 



base forms, with the edge of the prism, corresponding with 

 the two obtuse angles, angles of 106° 6' and 73° 54'. In the 

 primitive nucleus of the Amphibole, which is also an obtuse 

 rhomboidal prism, the inclinations of the planes of the prism 

 are to each other 124° 34' and 55° 26', a difference so great 

 as to be easily recognized. In the nucleus of the Pyroxene, 

 the smaller diagonal of the base is to the length of one of 

 the edges of the prism, nearly as 18 to 5; in the Amphibole, 

 on the contrary, it is nearly as 4 to 1. The nucleus of the 

 Pyroxene is divisible, in the direction of the two diagonals 

 of the base, into four oblique triangular prisms. 



The varieties of Pyroxene, which have been considered 

 as so many different species, have given to Plauy the same 

 primitive nucleus.* 



Pyroxene is generally black or green, or offers interme- 

 diate shades between these two colours; it is also whitish 

 green or greenish white, sometimes grey, rarely white; its 

 longitudinal fracture is lamellar, and this fracture is more 

 or less apparent, according to the varieties; its transverse 

 fracture is granular, uneven, or conchoidal. It is seldom 

 transparent, mostly translucent on the edges, or opaque. 

 When transparent, it is endowed with the double refracting 

 power. It is hard enough to scratch glass. Forms a dark 

 green, or greyish green powder. In general, its sp. gr. varies 

 from 3.223 to 3.373. It melts before the blow pipe with 

 difficulty, into a brownish or whitish glass. 



The varieties of Pyroxene have given, by analysis, near- 

 ly one half silex; lime and magnesia, in a quantity equal 



* This departed sage, in whom I deplore the loss of a true friend, has 

 made investigations of these different varieties the subject of many me- 

 moirs which were published in the Annates and Memoirs of the Museum 

 of Natural History; where those who wish to study the manner in which 

 the laws of crystallization are explained, and the results that are to be 

 deduced, may satisfy themselves- 



