{ 28 ) 



ture of the Vefuvian garnets, (o far as it is vU 

 treous and lamellar, is fimilar to that of thofe 

 of Lipari j but that, when expofed to the fire, 

 a difference is found between thefe two 

 fiones, the one eafily melting in the furnace, 

 and the other proving refradtory. 



Finding, therefore, that thefe four fpecies 

 of garnets were infufible in the furnace, 

 though continued in it for feveral days, I 

 had recourfe to oxygenous gas (or dephlo- 

 gifticated air}, by the adion of which they 

 all melted, though flov ly. When the ma- 

 trix lava flowed like common glafs, the 

 fmali pieces of garnet within it remained 

 unchanged ; but at length fufed, though 

 without incorporating with the lava, fo a3 

 to form a homogeneous whole. 



Thofe chemifts and naturalifts, who, be- 

 fore me, have made experiments with fire 

 on the Vefuvian garnets, have defcribed re- 

 fuks fimilar to thofe I obferved. Bergmann 

 fays thefe garnets melt with the blow- 

 pipe alone, but a vehement fire is necef- 



fary. 



