i6o EFFECTS of HEAT 



At the place of delivery, a quantity of gafeous matter is pro- 

 pelled violently upwards, and, along with it, fome liquid lava ', 

 which lalt, falling back again in a fpongy ftate, produces one 

 of thofe conical hills which we fee in great number on the vaft 

 fides of Mount iEtna, each indicating the difcharge of a parti- 

 cular eruption. At the fame time, a jet of flame and fmoke iffues 

 from the main crater, proving the internal communication be- 

 tween it and the lava ', this difcharge from the fummit gene- 

 rally continuing, in a greater or a lefs degree, during the in- 

 tervals between eruptions. (Fig. 41. reprefents an ideal fection 

 of Mount iEtna ; a b is the direct channel, and b c is a lateral 

 branch). 



Let us now attend to the ftate of the lava within the moun- 

 tain, during the courfe of the eruption \ and let us fuppofe, that 

 a fragment of limeflone, torn from fome flratum below, has 

 been included in the fluid lava, and carried up with it. By the 

 laws of hydroftatics, as each portion of this fluid fuftains pref- 

 fure in proportion to its perpendicular diftance below the point 

 of difcharge, that preflure muft increafe with the depth. The 

 fpecific gravity of folid and compact lava is nearly 2.8 ; and its 

 weight, when in a liquid ftate, is probably little different. 

 The table fhews, that the carbonic acid of limeftone cannot be 

 conftrained in heat by a preflure lefs than that of 1708 feet of 

 fea, which correfponds nearly to 600 feet of liquid lava. As 

 foon, then, as our calcareous mafs rofe to within 600 feet of the 

 furface, its carbonic acid would quit the lime, and, afluming a 

 gafeous form, would add to the eruptive effervefcence. And 

 this change would commonly begin in much greater depths, in 

 confequence of the bubbles of carbonic acid, and other fubftan- 

 ces in a gafeous form, which, rifing with the lava, and through 

 it, would greatly diminifh the weight of the column, and would 

 render its preflure on any particular fpot extremely variable. 

 With all thefe irregularities, however, and interruptions, the 



preflure 



