( 3/2 ) 



I next proceeded to the lavas filled with 

 pores and bubbles ; but here more caution 

 was neceflary. In one of them, two large 

 bubbles appeared perforated on one fide. 

 On pouring the water into the crucible, it 

 exploded with a noife equal to the report 

 of a fmall piftol, and forced out the lava in 

 fcattered fragments. I was then convinced 

 that this explofion had not been produced 

 by the water that had merely touched the. 

 furface of the lava, but by that which had 

 penetrated into the two open bubbles. I 

 had recourfe, therefore, to a proof which 

 could not but be decifive, which was, to re- 

 peat the experiment on the fame kind of 

 lava, but on fuch as contained no fracture 

 of any bubble, which may frequently be 

 obtained among fufed lavas of this fpecies. 

 When this lava was liquefied I let fall into 

 the crucible the fame quantity of water, 

 which produced only a fimple ebullition ; 

 and the fame happened on repeated trials, 

 which convinced me I was not miftaken in 

 my opinion. 



< I con- 



