'theory of Cryflallifatlort. zyj 



tondary cryftals by their re-entering and falient angles, thus 

 afford a fatisfa&ory folution of the difficulty a little before 

 mentioned; which is, that the fragments obtained bydivifion, 

 the external facets of which form part of the faces of the fe- 

 condary cryftal, are not like thofe drawn from the interior 

 part. For this diverfity, which is only apparent, arrfes from 

 the facets in qutftion being compofed of a multitude of 

 fmall planes^ really inclined to one another, but which, on 

 account of their fmailnefs, prefent the appearance of one 

 plane ; fo that, if the divifion could reach its utmoft bounds, 

 all thefe fragments would be refolved into moleculae, fimilar 

 to each other,' and to thofe fituated towards the centre. 



The fecundity of the laws on which the variations of cryf- 

 talline forms depend, is not confined to the producing of a 

 multitude of very different forms with the fame moleculac. 

 It often happens alfo, that molecula; of different figures ar- 

 range themfelves in fuch a manner as to give rife to like po- 

 lyedra in different kinds of minerals. Thus the dodecaedron 

 with rhombufes for its planes, which we obtained by com- 

 bining cubic moleculae, exrfts in the granite with a ftru&ure 

 compofed of fmall tetraedra having ifofceles triangular faces, 

 as I fhall prove hereafter ; and I have found it in fparry 

 fluor, where there is alfo an affemblage of tetraedra, but re- 

 gular ; that is to fay, the faces of which are equilateral tri- 

 angles. Nay more : it is poffible that fimilar moleculae may 

 produce the fame cryftalline form by different laws of de- 

 crement*. In fhort, calculation has conducted me to ano- 

 ther refult, which appeared to me £1111 moire remarkable, 

 which is, that, in confequence of a fimple law of decrement, 

 there may exift a cryftal which externally has a perfect re- 

 femblance to the nucleus, that is to fay,' to a folid that doe* 

 not arife from any law of decrement f. 



* Mem. de V Acad, an 1788, p. 17 & *<>• 

 f Ibid. p. 23. 



U 3 Varitut 



