( 382 ) 



be incapable of producing the explofions 

 continually obferved at Stromboli. I omit 

 feveral other objedlions, as 1 think thefe 

 two are fufficient to demonftrate that this 

 fecond hypothefis, likewife, cannot be main- 

 tained. 



Shall I venture to propofe a third, which) 

 in my opinion, is not contemptible, though 

 I ofFer it merely as conjectural ? May not 

 the phenomena of this volcano be attributed 

 to the adtion of oxygenous gas ? On the 

 one fide, it is well known that the fulphatc 

 of iron and the fulphate of alumine afford 

 this gas in abundance, when ad:ed on by a 

 ilrong fire ; and, on the other, it is equally 

 certain that thefe two falts are copioully 

 produced in volcanos. They may, there- 

 fore, in the profound recefies of Stromboli, 

 furnifh a rich and inexhauftlble mine of 

 oxygenous gas, which, mixing with the 

 lava, and afcending through it by its levity, 

 may colled:, in a confiderable quantity, in 

 the narrow neck of the crater, and producing 

 inflations and bubbles in the lava, force 



its 



