on Basalt , See. 
*95 
invites them to crystallize. In all crystallization from compound 
fluids, the order in which the several bodies crystallize must be 
determined by their relative quantities and attractions. It is 
perfectly obvious, that no molecules can form a crystal in a 
heat sufficient for its fusion ; but it by no means ensues, that it 
will be formed as soon as the molecules are cooled to the point 
where the crystalline polarity overcomes the disintegrating 
power of heat; for they may remain suspended in a fluid 
formed by more fusible bodies, provided this fluid be sufficiently 
abundant to keep them from contact with each other, for the 
crystalline polarity appears to exert itself only at extremely 
small distances. In a mass composed of substances in a state of 
fluidity, with refractory molecules suspended among them, it is 
pretty clear, from the preceding paragraph, that the most abun- 
dant ingredient will be the first to crystallize. But the removal 
of a portion of the suspending fluid must bring the refractory 
molecules nearer together, and perhaps so near that the crys- 
talline polarity may overcome the attraction of the fluid for 
them; they will therefore crystallize next, and will be followed 
by the remaining ingredients, in the order their attractions 
dictate. 
As the crystals last formed must necessarily be impressed, at 
the parts in contact, by the peculiar forms of those which have 
been first generated, it also follows, if the preceding reasoning 
be just, that the infusible crystals may be found impressed by 
the more fusible substance, which crystallized first ; and the re- 
maining ingredients of the mixture, which were subsequently 
arranged, may be moulded on the refractory crystals ; and thus, 
in the same specimen may exist, a refractory substance gene- 
rated by fire, impressed by more fusible bodies, and impressing 
MDCCCIV. O q 
