4« theory of Cryjlallizatiott. 



the paradox arifing from this diverfity of appearance is ex- 

 plained by the double ufe of the rhomboidal form, which 

 ferves here to difguife itfelf, and which conceals fixed and 

 conftant chara&erifing marks under a variable outfide. 



If we take a cryftal of another nature, fuch as a cube of 

 fluor fpar, the nucleus will have a different form. This, 

 in the prefent cafe, will be an odtaedron, to which we (hall 

 arrive by taking off the eight folid angles of the cube. Heavy 

 fpar will produce, for nucleus, a right prifm, with rhomboidal 

 bales ; feld fpar an oblique-angled parallelopipedon, but not 

 rhomboidal * ; the apatite or beryl a right hexaedral prifm ; 

 the adamantine fpar a rhomboid, a little fharpened •, blend 

 a dodecaedron, with rhomboidal planes ; iron of the ifland 

 of Elba, a cube, &c. and each of thefe forms will be conftant 

 in regard to the whole fpecies, fo that its angles will undergo 

 no variation of importance ; and if an attempt be made to 

 divk'e the cryftal in any other direction, we fhall not be able 

 to find a joining. Nothing will be obtained but undefined 

 fragments : in a word, it will be broken rather than divided. 



Thefe folids, infcribed, each, in all the folids of the fame 

 fpecies, ought to be confidered as the real primitive forms 

 on which all the other forms depend. All minerals, I con- 

 fefs, are not fufceptible of being divided mechanically, but 

 there arc a greater number than I at firft thought ; and with 

 regard to thofe cryftals which have hitherto refilled my ef- 

 forts to find their natural joinings, I have remarked that their 

 furface, ftriated in a certain direction, or the relation of their 

 different forms among thofe which belong to the fame fub- 

 ftance, often prefent indications of their ftructure ; and that, 

 reafoning from their analogy with other divifible cryftals, 

 we may determine that ftruc~rure at leaft with great pro- 

 bability. 



I call fccondar y Jbnps all thofe that differ from the primi- 

 tive form. We fhall fee, hereafter, that the number of thefe 



f Mem. de l'Acad. des Sciences, An 1784, p. 237. 



forms 



