ftY COMPRESSION. ^QJ 



taelong to the kind called aniygdaloid. In greater depths still, 

 the pressure would be proportionally increased, till sulphur, 

 and even water, might be constrained ; and the carbonate of 

 lime would continue undecomposed in the highest heats. 



If, while the lava was in a liquid state, during the eruption Submarine 

 or previous to it, a new rent {d e. Fig. 41.), formed in the ^^"P*'^"*' 

 solid country below the volcano, was met by our stream (at c?J, 

 it is obvious that the lava would flow into the aperture with 

 great rapidity, and fill it to the minutest extremity, there 

 being no air to impede the progress of the liquid. In this man- 

 ner, a stream of lava might be led from below to approach the 

 bottom of the sea (fjj, and to come in contact with a bed of 

 loose shells (gg), lyi"g on that bottom, but covered with 

 beds of clay, interstratified, as usually occurs, with beds of 

 sand, and other beds of shells. The first effect of heat would 

 be to drive off the moisture of the lowest shell-bed, in a state 

 of vapo-'.ir, which, rising till it got beyond the reach of the heat, 

 would be condensed into water, producing a slight motion 

 of ebullition, like that of a vessel of water, when it begins to 

 boil, and when it is said to simmer. The beds of clay and 

 sand might thus undergo some heaving and partial derange- 

 ment, but would still possess the power of stopping, or of very 

 much impeding, the descent of water from the sea above; so 

 that the water which had been driven from the shells at the 

 bottom, would not return to them, or would return but slowly ; 

 and they would be exposed dry to the action of heat.* 



In this case, one of two things would inevitably happen. Ei- andformatlen 

 ther the carbonic acid of the shells would be driven off by the ^^ ''"lestone. 

 heat, producing an incondensable elastic fluid, which, heaving 

 tip or penetrating the superincumbent beds, would force its 

 way to the surface of the sea, and produce a submarine erup- 

 tion, as has happened at Santorini and elsewhere ; or the vo- 

 latility of the carbonic acid would be repressed by the weight 

 of the superincumbent water (k kj, and the shell-bed, being 

 softened or fused by the action of heat, would be converted 

 into a stratum of limestone. 



* This situation of things, is similar to what happeos when 

 small-coal is moistened, in order to make it cake. The dust, 

 drenched with water, is laid upon the fire, and remains long wet, 

 wlijle the heat below suffers little or no abatement. 



The 



