107 A short notice of Earthquakes, [JanxtakY 



the presence of combustible matter ; various otber experiments were 

 followed by the same inferences upon another occasion, when flame 

 and steam issued from the crater, there was no indication of carbon. 

 I have however mentioned sufficient to prove, that the causes ancient- 

 ly assigned for volcanic fires, are proved by Sir H. Davy to be fallaci- 

 ous, the most current of these was the combustion of mineral coal, 

 this he says is most inadequate, for however large the strata of pit 

 coal might be, its combustion under the surface could never pro-* 

 duce intense heat, the production of carbonic acid gas, would im- 

 pede the process and if this cause existed, carbonaceous matter 

 would be found in lava. In England, there is an instance of strata 

 of pit coal having been long burning, but the only result was, baked 

 clay not in any way resembling iava, the action of sulphur upon 

 iron has also been assigned as a reason, but were this the 

 case sulphate of iron would be the chief product of the volca- 

 nics, " Sir H. Davy therefore assigns these phenomena to chemical 

 causes, the products of volcanoes giving an idea of the substances 

 primarily active ; these are found to contain mixtures of the earths 

 and alkalis in an oxydated and fused state from intense ignition, 

 water and saline substances are also found, such as may be furnish- 

 ed by the sea and air but altered in such a manner, as may be 

 expected from the formation of fixed oxidigied matter. He moreo- 

 ver states, as a reason, why the combustion is not increased, when 

 the volcanic productions pass into the atmosphere, that volcanic 

 fires take place in immense subterranean cavities, and that the ac- 

 cess of the air to the acting substances occurs long before they reach 

 the day, lava being the refuse of combustion below, Lyall and 

 several other practical naturalists support these opinions, observing 

 that earthquakes are generally attended and preceded by heavy 

 torrents of rain inundating contiguous regions, water being supposed 

 to percolate and descend by fissures to those cavities accounting in 

 some degree for the prevalence of earthquakes and valcanoes in Is- 

 lands, and in countries, bordering on the sea. Before the eruption of 

 the hot springs in Iceland, the ground is always agitated, and if the 

 vapours which are condensed in these small cavities are capable of 

 agitating the ground to any extent, what may be expected when they 

 are confined below in extensive subterranean cavities. The intensity 

 of the heat in these regions, and the density of the air from 

 the immense pressure from above can only be imagined 

 when it is known that limestone melts without giving off its 

 carbonic acid under a pressure of 1700 feet of sea which only cor- 



