158 ^OBSERVATIONS OK ft. -SALT, &C. 



cesbfe to wonder at the almoft daily formation of petrified wood, 

 (in which, though cryftalllzation does not actually take place, 

 a very perfect arrangement is indicated, by the intimate union 

 of the filiceous particles,) or of hydrophanous femi-opals in 

 the decompofed ferpentine of Muffinet, near Turin, or of 

 chalcedony containing drops of water, in the decompofed ba- 

 falt of Vicenza. 

 Since the cry- I have endeavoured to fhow, that. in the cryftallizations re- 



ftalhzingmaynot fu 1 1 i n g from igneous fufion, it is not only pofiible but probable, 

 order of fufibili- that the mofi infufible fubftances might not be the firfi to 

 ty, refractory cryfiallize; and this appears to involve important confequences, 

 be irnpreffeTby f° r '* P art ty removes one of the greatefl difficulties that embar- 

 fuch as are more raffes the igneous theory, by explaining the poffibility of re- 

 fractory fubftances generated by fire being impreffed by the 

 forms of more fufible ones. It feems, however, that the fame 

 order of arrangement would prevail in fubftances that were 

 ftsfpended in a fluid medium, as the degrees of attraction would 

 be the fame. In either cafe, the fir ft ilep by which the arrange-, 

 ment of an apparently homogenous mafs commenced, would 

 probably be the accumulation of particular molecules into little 

 globules. Such feems to have happened in variolites, and other 

 rocks which contain fpherical concretions of a different nature 

 Inftances and from their bails. Still farther advanced is the arrangement of 

 porphyries : the molecules of one fpecies have affumed a regular 

 cryftalline form ; and fomelimes two or even more varieties o 

 cryftals are formed, which remain unmixed in the unarranged 

 bafis. If the remaining molecules of that bafis are fufceptible 

 of eryfiallization, it may be fairly concluded, that an extenfion 

 of the procefs of arrangement would convert the porphyry into 

 granite, or at leafl into one of the compound aggregates of 

 cryftals which conftitute the numerous tribes of granites, grun- 

 fteins, and fienites ; and it feems equally probable that this 

 might be accomplished, whether the molecules were indebted 

 to a fuitable temperature, or to an aqueous medium, for the 

 requifite facility of movement. 



The formation of granite and other rocks, muft however be 

 porpHyry^"" referred to the ultimate perfection of eryfiallization, by which 

 all the molecules have been permitted to arrange. Thofe gra- 

 nites called porphyritic, in which large. cryftals of feldfpar are 

 imbedded in a bafis compounded of the ordinary ingredients of 

 granite in fmall grains, are apparently generated from a men- 



ftiuum 



remaiks. 



Granites and 





