OF THE CAUSES OF THE PHENOMENA OF VOLCANOES &C. 77 
June 1813 — first the West India Islands, afterwards the vallies of 
the Ohio and the Mississippi, and at last the opposite coast of 
Venezuela. Thirty days after the complete destruction of the town 
of Caraccas, the eruption of the volcano on the island of St. Vin- 
cent took place. At the same moment when this explosion hap- 
pened, on the 30th of April, a subterranean noise was heard 
throughout a country of nearly fifty thousand square miles in ex- 
tent." 
OF THE CAUSES OF THE PHENOMENA OF VOLCANOES 
AND EARTHQUAKES. 
41. We come now to the most difficult part of this subject, the 
task of accounting for the phenomena of volcanoes and earth- 
quakes. How is that heat which fuses the rocks generated in the 
bowels of the earth, and how does it operate in shaking the solid 
globe? Werner held that the seat of volcanic fires is the 
coal formation ; a secondary stratum. But the volcanoes of 
South America, have their seat beneath the primitive rocks. 
Those of Auvergne, in France, have forced their way through 
a bed of granite. The showers of red hot stones that are 
projected from the craters of burning mountains during an erup- 
tion, and the cloud highly charged with electricity that over- 
hangs them, render it impossible to approach them when in their 
highest state of activity, for the purpose of studying attentively 
and accurately the changes that are going on. Nor does the mat- 
ter that is thrown out afford us much information. Sulphur either 
free or in some of its combinations appears to be a constant pro- 
duct of all volcanoes. In the crater of Vesuvius there is the smell 
of burnt bitumen but not in that of Stromboli. Muriatic acid may 
also be detected, though not in general in any considerable quan- 
tity, in the vapours escaping during an eruption. Silica consti- 
tutes about one half of the substance of lava ; the rest is alumine, 
magnesia, lime, and iron. 
But the first three substances, sulphur, bitumen, and muriatic 
acid are present in too small quantity to admit of our attributing 
to them the tremendous convulsions that accompany intense vol- 
canic action. Carbonic acid is disengaged from the fissures and 
saturates the water of the springs of a volcanic district, nor does 
this effect cease for many ages after the fire is extinct. In what 
quantity it may be given out from the crater during an eruption, 
has not been ascertained. When Sir Humphrey Davy, ascertain- 
ed that the bases of the alkalis and earths are metallic, and that they 
take fire when brought into contact with water, the attention of 
philosophers was directed to them, as probably the agents by 
which the phenomena of volcanoes are produced : but their ex- 
treme levity is one objection to them, nor does the composition 
of lava accord with this idea. Silicon does not, like potassium 
and sodium, decompose water by the abstraction of its oxygen 
