tly the same pressure upon the
sides of the containing cavity as if it remained in the form of vapor.

If a tube, whether of glass or other materials, filled with condensed
gas, have its temperature slightly raised, it will often burst; for a
slight increment of heat causes the elasticity of the gas to increase in
a very high ratio. We have only to suppose certain rocks, permeated by
these liquid gases (as porous strata are sometimes filled with water),
to have their temperature raised some hundred degrees, and we obtain a
power capable of lifting superincumbent masses of almost any conceivable
thickness; while, if the depth at which the gas is confined be great,
there is no reason to suppose that any other appearances would be
witnessed by the inhabitants of the surface than vibratory movements and
rents, from which no vapor might escape. In making their way through
fissures a very few miles only in length, or in forcing a passage
through soft yielding strata, the vapors may be cooled and absorbed by
water. For water has a strong affinity to several of the gases, and will
absorb large quantities, with a very slight increase of volume. In this
manner, the heat or the volume of springs may be augmented, and their
mineral properties made to vary.

_Connection between the state of the atmosphere and earthquakes._--The
inhabitants of Stromboli, who are mostly fishermen, are said to make use
of that volcano as a weather-glass, the eruptions being comparatively
feeble when the sky is serene, but increasing in turbulence during
tempestuous weather, so that in winter the island often seems to shake
from its foundations. Mr. P. Scrope, after calling attention to these
and other analogous facts, first started the idea (as long ago as the
year 1825) that the diminished pressure of the atmosphere, the
concomitant of stormy weather, may modify the intensity of the volcanic
action. He suggests that where liquid lava communicates with the
surface, as in the crater of Stromboli, it may rise or fall in the vent
on the same principle as mercury in a barometer; because the ebullition
or expansive power of the steam contained in the lava would be checked
by every increase, and augmented by every diminution of weight. In like
manner, if a bed of liquid lava be confined at an immense depth below
the surface, its expansive force may be counteracted partly by the
weight of the incumbent rocks, and also in part by atmospheric pressure
acting c