Qn Earthquakes — their causes and effects. 21 3 



marine volcanoes, which show themselves by throwing up 

 large quantities of water into the air, and are in a state of 

 action in concert with those at a considerable distance. A 

 continuance of those eruptions frequently produces islands, 

 some of which again disappear suddenly, as though they 

 were the apex of a huge crater falling into its own bowels. 

 Of this description was the island of Sabrina, among the 

 Azores. In June 1811, it made its appearance where the 

 sea was sixty fathoms deep, and was nine hundred toises in 

 diameter. Many of the West India Islands experienced an 

 earthquake at the same time, at a distance of eight hundred 

 leagues. 



It may be objected to this theory that rocks do not burn ! 

 In answer I slate, that when rocks are decomposed, their 

 metallic bases, calcium, silicium, aluminum, magnesium, Sic, 

 are highly incandescent, and prove a never failing source of 

 matter for combustion. 



If there should be an accumulation of gaseous vapours in 

 one of those channels, and it should be unable to pass in 

 consequence of some obstruction, to a liberating volcano, it 

 forces another opening, and causes, by passing through a 

 distant channel, an earthquake at a considerable distance 

 from whence it originated. Something of this nature took 

 place when Riobamba was destroyed. The volcano of 

 Pasto, which had continued to eject smoke, ceased at once ; 

 at that moment the city was destroyed at the distance of 

 thirty-tive leagues south. 



Earthquakes frequently occur without any symptoms 

 of connexion with volcanoes. These may be accounted 

 for by supposing that in consequence of a slight influx 

 of the sea to one of those channels, the water, meeting 

 with an ignited mass, is suddenly converted into elas- 

 tic vapour, and expands itself throughout its openings, 

 causing a concussion or trembling of the superincumbent 

 strata. Dolomieu, in his examination of Calabria, says, " La 

 force motrice paroit avoir reside sous la Calabre ellc 

 meme." He afterwards informs us that the motion appears 

 to have advanced progressively along the chain of the Appe- 

 nines, from the south to the north. The action of an earth- 

 quake was almost instantly communicated from Chili to 

 Guayaquil, a distance of six hundred leagues. If Hum- 

 boldt's idea, that the Cordilleras form a vast volcanic wall, 

 be correct, the pulsation must have been communicated 



