( 377 ) 



pn die fpot, I was fufficiendy convinced 

 that the fluid contained in the tumors 

 could not be derived from the evaporation 

 ©f the lava, fince this evaporation (which, 

 from the fad:s adduced, cannot be denied) 

 was, indeed, fufEcient to burft them, but 

 not to produce the ejedions continually " 

 pbfervable. Nor can v^'e attribute them to 

 evaporated v^^ater, as this would not have 

 deferred its detonation when included in 

 thofe bubbles till it had reached the fuiface 

 of the lava ; but the explolion and ejed:ioii 

 would have taken place at the moment the 

 water had come in contad: with the burning 

 lava, and, confequently, at a much greater 

 depth, Befides, had that multitude of ca- 

 pacious tumors been filled with water which 

 had padded into the ftate of vapour, it muft 

 be maniieft that every eruption would be 

 much m.ore violent and loud. "We muft 

 therefore conclude that the fluid in quef- 

 tion is a permanent gas, and probably one 

 of thofe which are difcover?ble in volcanos 

 when, in a ftate of tranquillity, they are ac- 



ceflTible, 



