in the fufion of gold in a mafs of filver, for if the Hlver ht 

 very {gradually cooled the gold will feparate from it, as Komberg 

 obferved, Mem- Par. 17 13. 



Now llie affinity of filex to the alkali, being loofened by a 



flight diminution of heat, the affinity of argil to the filex to 



which it united only as to a compound in the given temperature, 



is alfo neceflarily loofened. That in the dry way argil unites 



to filex in temperatures below 150°, only in confequence of the 



previous union of the filex to the alkali, is clearly deduced 



from this fad, that if the alkali be abfent the union will not 



take place in temperatures below 150°, whereas it takes place 



by Sir James's own experiment at temperatures below 100° when 



the alkali is prefent, for he found the whins fufible at 55". 



It is true the whins contain lime alfo, but though the prefence of 



a certain proportion of lime contributes materially to the fu- 



fibility of filex and argil, yet it would be inefFedual in degrees 



of heat below 120S if an alkali were not prefent to affift it, 



as I know by experience. 



The prefence of argil contributes alfo to the diminution of 

 the afiinity of the alkali to the filiceous ingredient, as the alkali 

 feems to have nearly as ftrong an affinity (fome think ftronger) 

 to argil as to filex ; hence it is that all Analifts fince Berg- 

 man's time employ an alkali to loofcn the intimate union of 

 filex and argil in precious ftones. 



• Thefe 



