18 REACTION OF THE INTERIOR OF THE EARTH 



the chain of the Andes separated by intervals of almost a 

 century. To this law I long since called attention ; 

 and where exceptions occur, they may perhaps be explained 

 by the channels of communication between the volcanic 

 seat of action and the crater of eruption not being in all 

 cases alike permanently free; since this channel may be 

 temporarily obstructed in some volcanoes of moderate eleva- 

 tion, so that eruptions may become more rare, without 

 any immediate prospect of their absolute extinction. 



These considerations, respecting the relation of the height 

 of volcanoes to the frequency of their eruptions, naturally 

 conduct us to an examination of the causes which determine 

 the place at which the lava issues from the mountain. In 

 many volcanoes eruptions from the crater are extremely rare; 

 they more often take place (as was remarked in the case of 

 Etna in the sixteenth century by the celebrated historian 

 Bembo ( 216 ), when a youth), from lateral openings formed in 

 those parts of the walls of the upheaved mountain, which, 

 from their nature and shape, may offer the least resistance. 

 Sometimes "cones of eruption" rise over these lateral 

 fissures ; and in this case the larger cones, erroneously 

 denominated "new volcanoes/' are ranged in rows, in- 

 dicating the line of fissure, which is speedily reclosed; 

 while the smaller cones, which are shaped like bells 

 or bee-hives, form numerous crowded groups, which cover 

 large spaces of ground. To the latter class belong the 

 "hornitos de Jorullo" ( 21 ?) and the cones of eruption of 

 Vesuvius in October 1822, those of the volcano of Awatscha 

 according to Postels, and of the lava field described by 

 Ennan, near the Baidar mountains, in the peninsula of 

 Kamtschatka. 



