Q50 On CrystdllbgrdpfnJ i 



\Thich fjass by the axis and by the oblique diagonals, giveS 

 the result represented in fig. 10, where we see that the 

 molecules are tetrahedrous. Observation also proves that 

 the oblique quadrangular prism which is the nucleus of 

 pyroxene, has naUiral joints situated parallel to a plane 

 which would pass by the small diagonals of its bases ; from 

 this we may conclude that its molecules are triangular 

 prisms. 



I shall not insist further on these modes of division, 

 which will be explained at greater length when we come to 

 speak of the articles which represent them ; but I ought 

 not to pass over a result which serves to connect the cry- 

 stallization of substances, whose molecule is the tetrahe- 

 dron or triangular prism, with that of substances which 

 havQ^, as primitive forms, simple assemblages of elementary 

 parallelopipcdons. 



This connection consists in the tetrahedral or prismatic 

 molecules being alway^^ assorted in such a manner, in the 

 interior of the primitive form and of secondary crystals, 

 tiiat on taking them by small groups of twos, fours, sixes, 

 or eio"hts. they compose parallelopipedons, so that the 

 ramies subiiacted by the effect of decrements are nothing 

 else than sums- of these parallelopipedons. 



That we may better conceive how this takes place, let us 

 conceive for a moment that the small rhomboids which re- 

 present the molecules of carbonated lime are divisible into 

 tetrahedrons, as we have seen with respect to the rhom- 

 boids which belong to the tourmaline. This view does not 

 change the explanations which we have given of the difie- 

 rent secondary forms of which carbonated lime is suscep- 

 tible : i. e. in order to determine these forms by the help of 

 theory, we should always confine ourselves to the consi- 

 deration of decrements by one or more ranges of rhoua- 

 boidal molecules. 



What is onlv an hypothesis with regard to carbonated 

 lime, is changed into reality with the tourmaline. Although 

 the rhomboids produced at first by the mechanical division 

 of the crystals of this substance are ultimately resolved intp 

 tetrahedrons, the decrements which give the secondary 

 forms are produced by subtractions of these rhomboids si- 

 tnilar to the primitive form ; so that we may suppose, in 

 calculations relative to the determination of these forn)S, 

 that the tetrahedrons which represent the true molecules 

 are connected with each other in an invariable manner in 

 each rhomboid. 



Let us cite another example drawn from a very simple 



structure. 



