1 16 M. Humboldt on the iVatets of the Rio Vinagre 



in 1631 ; fabulous traditions, which are perpetuated by an in- 

 scription at Portici, 



In the volcanos of the Andes which exceed the limit of per- 

 petual snow, the causes of inundations are very different from 

 those which we have just indicated. As the eruptions of 

 these colossal summits take place only after long intervals 

 (every thirty or forty years, or still more rarely), banks of snow 

 of an enormous thickness accumulate on the sides ol the moun- 

 tains. These snows do not melt at the time of the explosion 

 only, but sometimes several days before. Thus in February 

 1803, during my stay at Guayaquil, the inhabitants of the 

 province of Quito were frightened at the appearance of the 

 cone of Cotopaxi, which lost a great part of its snows in 

 a single night, and showed plainly the black colour of its 

 burnt rocks. Whatever idea may be formed of the power 

 of the volcanic forces, and of the intensity of the subterraneous 

 fires in the Andes, it cannot be admitted that the thick sides 

 of a cone could be uniformly warmed, and transmit the heat 

 with such rapidity (by the conductibility of their mass) to the 

 outside. The sudden melting of the snows, when, in the Cor- 

 dilleras, it precedes the eruptions, is probably owing only to 

 an infinity of little fumarolcs which disengage hot vapours 

 through the fissured rock of the cone. These vapours, accord- 

 ing to what I have had opportunity of observing in the cra- 

 ters of Vesuvius, the Peak of Tenerifte, and the volcano of 

 Jorullo in Mexico, are most frequently pure water, which 

 does not act at all on the most sensible re-agents ; at other 

 times they contain muriatic acid. It is remarked that the 

 same crevice gives, at very near epochs, distilled (pure) water 

 and very acid waters. The artificial spring which M. Gim- 

 bernat has had the ingenious itlea of forming at the sum- 

 mit of Vesuvius, by the condensation of the vapours in a glass 

 tube, has sometimes shown these variations : they prove either 

 the change of chemical action in the interior of the volcano, 

 or the accidental opening of some new communications. In 

 the Andes of Quito, as in Iceland, and in the eruptions of 

 iEtna of March 23, 1536, and March 6, 175.5, the sudden 

 melting of the banks of snow produced great devastations*. 



At other times, by slow infiltrations the snow waters are ac- 

 cumulated in the lateral cavities of the volcano; shocks of 

 violent earthquakes, which do not always coincide with the 

 epoch of the fiery eruptions, open these cavities ; and waters 

 long kept in, which support little fish of the genus Pimelodes, 

 carry with them pidverized trachytes, piunice-stones, tufas, 



• Ferrara, Campi Ftvgrn,\H\{), p. 1 6.^.— Idem, Dcsn-k. ddP Etna, 1818, 

 p. 89, 116-120. 



and 



