46 Ohservafions upon th& 



really have, that we may not found systems upon nilstakea 

 or exaggerations. 



M. Fleuriau de Bellcvue does not admit that sea water is 

 absolutely necessary for producing volcanoes ; and he quotes 

 as^ainst this opinion, which seemed at first very alluring, a 

 volcanic eruption related by Messrs. Humboldt and Bonpland, 

 which took place, in 1759, " in a plam of Mexico, yiir/y 

 leagues from the sea in a straight line ; an eruption which ele- 

 vated in one night a volcano 1494 feet high, surrounded by 

 more than two thousand mouths, which still smoke." 



If burning volcanoes could be manifested any where other- 

 wise than under the influence of the waters of the ocean, we 

 should not need to confine ourselves to a single example] a 

 great number in that case would certainly be produced ; and 

 if this were the case, the opinion I have hazarded would ne- 

 ver have come into my mind. But after having directed my 

 attention to the great fact, that there is no burning volcano 

 whatever in inland districts ; that no expanse can be pro- 

 duced of fresh water, however vast it may be, having ever 

 produced any volcano ; that all of them are near the sea, or 

 encompassed by it ; and having observed that the smoke of 

 volcanoes deposits marine acid in abundance ; I drew this 

 unavoidable conclusion from all these circumstances, that 

 sea water is absolutely necessary, from the salts it holds in 

 solution, for producing the fermentations which raise and 

 keep up volcanoes. 



This conclusion has been since confirmed by the eruptions 

 of water from the volcanoes of Iceland, which deposit ma- 

 rine salt in great abundance ; and lately also by an observa- 

 tion of Messrs. Humboldt and De Buch, who were witnesses 

 of the eruption of Vesuvius in the month of August 1805, 

 and saw the sides of a crevice of its crater covered with a 

 crust of muriate of soda two or three inches in thickness. 



From which it results that the fact quoted by M. Fleuriau 

 de Bellevue proves nothing, if not that there are subterra- 

 nean galleries which extend 40 leagues from the sea, and that 

 in this instance its waters have penetrated that length. If 

 all the circumstances accompanying this fact were well known, 



it 



