Baron Humboldt on Volcanoes. 291 



tion, which occasionally rise high above the edges of the crater, 

 and give the volcano its characteristic physiognomy for years ; 

 but, on the occurrence of fresh eruptions, they sink suddenly 

 down, and disappear. The openings of these cones of erup- 

 tion, which rise from the floor of the crater, must not, as is too 

 frequently done, be confounded with the crater itself, which 

 encircles them. When the crater is inaccessible, from its vast 

 depth, and the perpendicular inward slope of its sides, as in 

 the case of Rucu-Pichincha (14,946 feet high), one can still 

 look down from the edges upon the summits of the monticules 

 which rise within the cauldron-like crater, partially filled with 

 sulphureous vapour. A more wonderful or grander natural 

 prospect I have never enjoyed. In the interval between two 

 eruptions, the crater of a volcano may exhibit no luminous 

 phenomenon, but merely open fissures and jets of watery 

 vapour ; or hillocks of ashes, that can be approached without 

 danger, are found upon its scarcely heated bottom. These 

 often gratify the wandering geognost, without making him 

 run any risk, by casting out glowing masses, which fall on the 

 edges of the cone of scoriae, their appearance being regularly 

 announced by slight, and entirely local, shocks — earthquakes 

 on a sanall scale. Lava occasionally flows from open fissures, 

 or snjall fiery gorges, into the crater itself without bursting 

 through its walls, or OA'erflowing its edges. But if it does 

 break through, the molten spring generally flows smoothly, 

 and in such a determinate direction, that the great cauldron- 

 like valley, called the crater, can still be visited during the 

 period of the eruption. Without a particular description of 

 the form, and also of the normal structure of burning moun- 

 tains, phenomena cannot be rightly comprehended which have 

 been distorted by fantastical descriptions, and the various signi- 

 fications attached to the words Crater, Cone of eruption, and 

 Volcano ; or rather, to the indefinite and indeterminate use of 

 these words. The edges of the crater sometimes shew them- 

 selves much less liable to change than might be expected. A 

 comparison of De Saussure's measurements with my own, yields 

 the remarkable result, in connection with Vesuvius at least, 

 that the north-west edge of the volcano, the Rocca del Palo, 

 may be regarded as having remained for forty-nine years 



