VOLCANOES. 229 



by the elevation of the volcano, which varies from the incon- 

 siderable height of a hill (as the volcano of Cosima, one of the 

 Japanese Kurile islands) to that of a cone above 19,000 feet 

 in height. It has appeared to nie that relations of height have 

 a great influence on the occurrence of eruptions, w^hich are 

 more frequent in low than in elevated volcanoes. I might in- 

 stance the series presented by the following mountains : Strom- 

 boli, 23 1 8 feet ; Guacamayo, in the province of Quixos, from 

 which detonations are heard almost daily (I have myself often 

 heard them at Chillo, near Quito, a distance of eighty-eight 

 miles); Vesuvius, 3876 feet; .^tna, 10,871 feet; the Peak 

 of Teneriffe, 12,175 feet ; and Cotopaxi, 19,069 feet. If the 

 focus of these volcanoes be at an equal depth below the sur- 

 face, a greater force must be required where the fused masses 

 have to be raised to an elevation six or eight times greater 

 than that of the lower eminences. While the volcano Strom- 

 boli (Strong)4e) has been incessantly active since the Homeric 

 ages, and has served as a beacon-light to guide the mariner in 

 the Tyrrhenian Sea, loftier volcanoes have been characterized 

 by long intervals of quiet. Thus we see that a whole century 

 often intervenes between the eruptions of most of the colossi 

 which crown the summits of the Cordilleras of the Andes. 

 Where we meet with exceptions to this law, to which I long 

 since drew attention, they must depend upon the circumstance 

 that the connections between the volcanic foci and the crater 

 of eruption can not be considered as equally permanent in the 

 case of all volcanoes. The channel of communication may be 

 closed for a time in the case of the lower ones, so that they 

 less frequently come to a state of eruption, although they do 

 not, on that account, approach more nearly to their final ex- 

 tinction. 



These relations between the absolute height and the fre 

 quency of volcanic eruptions, as far as they are externally per 

 ceptible, are intimately connected vAih. the consideration of 

 the local conditions under which lava currents are erupted. 

 Eruptions from the crater are very unusual in many mount- 

 ains, generally occurring from lateral fissures (as was observed 

 in the case of -^tna, in the sixteenth century, by the cele- 

 brated historian Bembo, when a youth*), wherever the sides 



* Petri Bembi Opuscula {^tna Dialogus), Basil, 1556, p. 63 : " Quic- 

 quid in ^tnae matris utero coalescit, uunquam exit ex cratere superiore, 

 quod vel eo iuscoudere gravis materia nou queat, vel, quia inferius alia 

 Hpiramenta sunt, non fit opus. Despumant flammis urgentibus ignei riv' 

 pigro fluxu totas delambentes plagas, et in lapidem indurescunt." 



