io HIS 'TORT of the SOCIEtr. 



fally molded upon the cryftals of felt-fpar. Now, were it true, 

 that all granite is formed by fufion, the very contrary, it 

 would feem, ought always to take place, as fek-fpar is very ea- 

 fily melted, and quartz refills the greateft efforts of heat that 

 have hitherto been applied to it. 



This difficulty is obviated thus : It is well known, that when 

 quartz and felt-fpar are pounded and mixed together, the mix- 

 ture may, without difficulty, be melted and run into a kind of 

 glafs, the felt-fpar ferving as a flux to the quartz. The fame 

 fact may be dated in another way, by confidering the felt-fpar, 

 when melted, as a fluid in which, as in a menftruum, the 

 quartz is difTolved ; and in this view, we may expect, by ana- 

 logy, that phenomena, flmilar to thofe of the folution of fait in 

 water, mould take place. Now, it is certain, that when ex- 

 ceffive cold is applied to fait water, the water is frozen to the 

 excluiion of the fait, the ice obtained yielding freih water when 

 melted, and the fait, when the experiment is pufhed to the ut- 

 moft, feparating from it in the form of fand. Why Ihould not 

 the fame thing happen in the folution of quartz in liquid felt- 

 fpar, when the mafs is allowed to cool below the point of con- 

 gelation of the menftruum? The felt fpar may cryftalize fepa- 

 rately from the quartz, as we have feen pure ice formed fepa- 

 rately from the fait ; in both cafes, the congelation of the folvent 

 being fimultaneous to that of the diffolved fubflance. Hence 

 the cryftals may mutually interfere with each other's forms, 

 and we may as naturally expect to fee quartz molded on cryftals 

 of felt fpar as the reverfe. 



In anfwer to an objection which might be urged againft this 

 reafoning, viz. that the refult of the fufion of granite is a glafs in 

 which no cryftalization can be feen, an accidental experiment was 

 produced, which had happened at one of theLeith Glafs- houfes 

 a few weeks previous to the reading of this paper. A quantity of 

 common green glafs having been allowed, in a great mafs, to cool 

 gradually and very flowly, it was found to have loft all the proper- 

 3 ties 



