[ 7 J 



which the former contains a large proportion of the latter, as 

 effential to its compofition, and is fufible only by reafon of its 

 compound nature, but if the quantity of the quartzy ingredient 

 be increafed the whole becomes infufible, as I have experienced- 

 .whereas if the proportion of fait in water be increafed, ftill the 

 water will be congealable if confiderably cooled ; moreover quartz 

 frequently bears the iropreffion of ftones more fiifible than 

 itfelf, which could not happen in any poflible fuppofition, if all 

 had been in a ftate of fufion. 



Again, Sir James obferved, that a quantity of green glafs, 

 which had been allowed to cool flowly, was found to have loft 

 all its vitreous properties, being opake, white and refraflory ; 

 but being again melted by a blow-pipe and fuddenly cooled, 

 it refumed its former properties and became glafs ; hence he infers 

 that if the glafs produced by the fufion of granite had been 

 allowed to cool with fufficient flownefs, it might have cryftallized, 

 producing a granite fimilar to the original, p; 1 1 . 



The obfervation on glafs here mentioned is perfedly juft and 

 has been often repeated ; but the analogy betwixt this cafe and 

 the formation of granite from a complete fufion of its ingre- 

 dients is far from being accurate. .Glafs confifts of a fimple 

 earth, namely, the filiceous united' to an alkali. To form this 

 union it is necelTary that the integrant- affinity of the filiceous 

 particles to each other fhould yield to the chymical affinity 



which 



