drostatic pressure to
which the flanks of the mountain are exposed, when the column of lava
has risen to a great height. Or if, before it has reached the top, there
should happen to be any stoppage in the main duct, the upward pressure
of the ascending column of gas and lava may burst a lateral opening.

In the case however of Mount Loa, in the Sandwich Islands, there appears
to be a singular want of connection or sympathy between the eruptions of
the central and the great lateral vent. The great volcanic cone alluded
to rises to the height of 13,760 feet above the level of the sea, having
a crater at its summit, from which powerful streams of lava have flowed
in recent times, and having another still larger crater, called Kilauea,
on its southeastern slope, about 4000 feet above the sea. This lateral
cavity resembles a huge quarry cut in the mountain's side, being about
1000 feet deep when in its ordinary state. It is seven miles and a half
in circuit, and scattered over its bottom, at different levels, are
lakes and pools of lava, always in a state of ebullition. The liquid in
one of these will sometimes sink 100 or 150 feet, while it is
overflowing in another at a higher elevation, there being, it should
seem, no communication between them. In like manner, lava overflows in
the summit crater of Mount Loa, nearly 14,000 feet high, while the great
lateral cauldron just alluded to (of Kilauea) continues as tranquil as
usual, affording no relief to any part of the gases or melted matter
which are forcing their way upwards in the centre of the mountain.
"How," asks Mr. Dana, "if there were any subterranean channel connecting
the two great vents, could this want of sympathy exist? How, according
to the laws of hydrostatic pressure, can a column of fluid stand 10,000
feet higher in one leg of the siphon than in the other?" The eruptions,
he observes, are not paroxysmal; on the contrary, the lava rises slowly
and gradually to the summit of the lofty cone, and then escapes there
without any commotion manifesting itself in Kilauea, a gulf always open
on the flanks of the same mountain. One conclusion, he says, is certain,
namely, _that volcanoes are no safety valves_ as they have been called;
for here two independent and apparently isolated centres of volcanic
activity, only sixteen miles distant from each other, are sustained in
one and the same cone.[774]

Without pretending to solve this enigma, I cannot refrain from
remar