210 DYNAMICAL GEOLOGY. [Book III. 
Idaho overlie denuded masses of earlier trachytic lavas, and similar 
proofs of a long succession of intermittent and widely-separated vol- 
canic outbursts can be traced northwards into the Yellowstone Valley. 
When a volcanic vent is opened it might be supposed always to find 
its way to the surface along some line of fissure, valley or deep de- 
pression. No doubt many, if not most, modern as well as ancient 
vents, especially those of large size, have done so. It is a curious 
fact, however, that in innumerable instances minor vents have 
appeared where there was no line of dislocation to aid them. This 
has been well shown by a study of the ancient volcanic rocks of the 
Old Red Sandstone, Carboniferous and Permian formations of Scot- 
land... It has likewise been most impressively demonstrated by 
the way in which the minor basalt cones and craters of Utah have 
broken out near the edges or even from the face of cliffs rather than 
at the bottom. Captain Dutton remarks that among the high 
plateaux of Utah, where there are hundreds of basaltic craters, the 
least common place for them is at the base of a cliff, and that, though 
they occur near faults, it is almost always on the lifted, rarely upon 
the depressed side.” Ona small scale a similar avoidance of the valley 
bottom is shown on the Rhine and Moselle, where eruptions have taken 
place close to the edge of the plateau through which these rivers 
wind. Why outbreaks should have occurred in this way is a question 
not easily answered. It suggests that the existing depressions and 
heights of the earth’s surface may sometimes be insignificant features, 
compared with the depth of the sources of volcanoes and the force 
employed in volcanic eruption. 
Conditions of Eruption.—Leaving for the present the general 
question of the cause of volcanic action, it may be here remarked 
that the conditions determining any particular eruption are still 
unknown. An attempt has been made to show that the explosions of 
a volcano are to some extent regulated by the conditions of atmos-— 
pheric pressure over the area at the time. In the case of a voleanic 
funnel like Stromboli, where, as Scrope pointed out, the expansive 
subterranean force within, and the repressive effect of atmospheric 
pressure without, just, balance each other, any serious disturbance of 
that pressure might be expected to make itself evident by a change 
in the condition of the volcano. Accordingly, it has long been 
remarked by the fishermen of the Lipari Islands that in stormy 
weather there is at Stromboli a more copious discharge of steam and 
stones than in fine weather. They make use of the cone as a weather- 
glass, the increase of its activity indicating a falling, and the diminu- 
tion a rising barometer. In like manner, Etna, according to S. von 
Waltershausen is most active in the winter months. When we 
remember the connexion now indubitably established between a more 
copious discharge of fire-damp in mines and a lowering of atmospheric | 
pressure, we may be prepared to find a similar influence affecting the 
! Trans, Roy. Soc. Edin. xxix. p. 437. 
* “High Plateaux of Utah,” Geol. and Geog. Survey of Territories, 1880, p. 62. 

