64 RICHTHOFEN —NATURAL SYSTEM 
course, depend in a great measure upon the width they would severally occupy within 
the channels, and those filling its wider parts must for a longer time remain susceptible 
of renewed expansion by accidental circumstances. Another cause which would have 
the same general effect, is the localization of the ingress of water. Some of its chan- 
nels would, to all probability, become obstructed, and the rate at which new ones 
would be opened in their place would probably diminish in nearly equal ratio with the 
total amount of the manifestations of energy connected with the phenomenon of ejection 
in its different stages. As it appears that the supply of water, which may be either 
constant or intermittent, is, next to an elevated temperature, the chief condition for 
entertaining voleanic action, it may be inferred that the cause mentioned would contrib- 
ute greatly towards the isolation of certain portions within the main fissure, by helping 
to keep the matter within them in a liquid state. If this second cause was coincident 
with the first, by the restriction of the ingress of water to a place where the fissure 
expanded, then both circumstances would combine to prolong the state of liquidity at 
that point. The connection between the bottom and the surface may have been 
kept open in a certain part of the mouth of the fissure, while solidification was proceed- 
ing over the rest of it. An isolated vent would then gradually be formed, and nar- 
rowed down to the size of a voleanie orifice. Obstructions of the outflow, by periodical 
consolidation, would become more frequent, and thus would proceed a slow change of 
the mode of action of massive eruptions to that which is peculiar to volcanoes. 
This is probably the simplest manner in which volcanoes can originate. It will 
apply particularly to a number of those which have undergone no change in regard to 
the character of their lava, and the lofty cones of which rise over mountain ranges 
consisting of the same material with their own lava and cinders, though owing their 
origin to massive eruptions. We must now consider a third cause which would aid in 
promoting volcanic action, and probably come very often into play. It is indicated by 
the frequent occurrence of series of voleanoes extending in lines parallel to the axis of 
the main outbursts. Their only possible cause is the formation of fractures, parallel to 
the main fissure, and branching off from it. In order, however, that these could be 
formed, solidification must have proceeded downward in the main fissure, without any 
communication with the liquid portion in depth having been kept open. This process 
would necessarily imply a temporary cessation of the process of extrusion. That this 
could take place may be the more readily understood, if it is taken into consideration, 
that the liquid masses filling the fissure in its whole extent must contract considerably 
by constant loss of heat, and that any additional expansion produced by the promotion 
of aqueous fusion at certain places had first to equalize this loss of volume, before 
it could manifest itself in a rise of the whole mass. This twofold action, which is 
probably one of the main causes of the intermittent character of volcanic ac- 
tivity, must produce an alternating motion of matter within the fissure, and 
there would be given ample opportunity, during a period of its subsidence, for 
the consolidation of the upper portion to a great distance down from the surface. 
If a period then followed in which, by dislocations of some kind, a change took 
place in the conditions subterranean, and expansion began again to prevail within the 
fissure, the new supply of force would manifest itself at the upper limit of the liquid 
(102) 
