{Z Theory of Crfta'dization. 



cryftals, and arrives at refults which exactly reprefent thofe 



cf nature \ and this is the only end 1 propofed. 



When the nucleus is a parallelopipedon, that is to fay, a 

 foKd with fix parallel faces, two to two, like the cube, the 

 rhomboid, &c and this folid does not admit of any other 

 dlivifions than thofe made in the direction of the faces, it \% 

 evident that the moleculse refulting from fubdivifion, both 

 of the nucleus and of the enveloping matter, are fimilar to 

 the nucleus. In other cafes the form of the molecula? differs 

 from that of the nucleus. There are alfo cryftals which, by 

 help of the mechanical divifion, yield particles of different 

 figures, combined together throughout the whole extent of 

 thefe cryftals. I fhall explain hereafter my conjectures oil 

 the manner of refolving the difficulty prefented by thefe 

 kinds of mixed ftructures ; and it will be feen that this difli-> 

 euJty does not affect the theory at bottom. 



II. Laws of Diminution. 



j. Diminution at the edges. 



The primitive form and that of the integral moleculae 

 Being determined, after the diffeclion of the cryftals, it was 

 neeeffary to difcover the laws according to which thefe 

 roofoculs were combined to produce, around the primitive 

 form, thofe fpecies of envelopes^ terminated fo regularly, and 

 from which refulted polyedia fo different from each other - % 

 though originally of the fume fubftance. But fuch is the 

 mechanifm of the ftru&ure fubject to thefe laws, that all the. 

 parts of the fecondary cryftal, fuperadded to the nucleus,. 

 are formed of laminae, decreasing regularly by the fubtraction 

 ©f one or more ranges of integral molecula? ; fo that 

 theory determines the number of thefe ranges, and, by a; 

 meeeflary confequence, the exact form of the fecondary 

 cryftal. 



To give an idea of thefe laws, I fliall firft make choice of a 

 T«nry fimple a:?d very elementary example. Let us conceive 



