{ U9 ) 



It at laft overflowed the edges of the cruci-» 

 ble, forming fmall ftreams down its fides, 

 which, when they reached the plane on 

 which the crucible flood, gave origin to 

 fmall currents, if that plane was at all in- 

 clined. When I put more of the fame pro- 

 dudl into the crucible, the currents became 

 larger. If the plane was then taken from 

 the furnace, and the fmall currents thus pro- 

 duced examined, they were found full of 

 minute bubbles, as was likewife the matter 

 which remained in the crucible. This cu- 

 rious experiment I made with feveral glafles 

 and volcanic enamels, as alfo with a variety 

 of cellular lavas, and always with the fame 

 fuccefs. 



It is likewife probable that this elaftic 

 vapour, when collected in a large quantity, 

 if it finds under the earth any impenetrable 

 obftacle, produces local earthquakes, and fub- 

 terraneous thunders and roarings ; burfting 

 open the fides of the lava, and forcing out 

 the lava. We have an example of this, if 

 1 may fo fpeak, in miniature, in the two 



matrafles 



