£g 2 Kfl.C ROSCOFIC AL Es SAYS', 



2. That in fimple or little-compounded bodies, this tendency 

 of integrant parts is more obvious and fenfible than in others more- 

 compounded : hence the former are much more difpofed to 

 cryftallize. 



3. .That. although we do not know the figure of the- primitive 

 integrant molecules of any body, we cannot doubt but that thofe 

 of every, different body have a conftantly uniform and peculiar : 

 figure. 



4. That thefe integrant parts cannot have an equal tendency to 

 unite indifcriminately by any of their fides, but by fome prefer- 

 ably to others, excepting all the fides of an integrant part of a-* 

 body be equal and fimilar;.and probably the fides, by which, 

 they tend to unite, are thofe by which they can touch raoft ex- 

 tenfively and immediately. 



The mod general phenomena of cryftallization may be con- 

 ceived in the following manner : 



Let a body be fuppofed to have it's integrant parts feparated ; 

 from each other by fome fluid ; if a part of this fluid be taken 

 away, thefe integrant parts will approach together; and as the 

 quantity of intervening fluid dimini flies, they will at' laft touch 

 and unite. They may alfo unite when they come fo near to each 

 other, that their mutual tendency fliall be capable of Overcoming 

 the diflance betwixt them. If, befides, they have time and 

 liberty to unite with each other by the fides molt difpofed to this 

 union, they will form mafles of a figure conftantly uniform and ; 

 fimilar. For the fame reafon, when the interpofed fluid is haftily, 



« taken:/; 



