aggregation of
smaller cavities--all being filled with steam and gases--together with
dust and volatile products which are ejected when the cavity opens up,
and its contents escape at the upper surface of the lava stream in
virtue of the continuation of the twistings and convolutions due to the
stream motion itself, and to the unbalanced hydrostatic pressures acting
upon the parietes of the bubble. Very large single bubbles of like
character rise in the fluid lava within craters in vigorous action, and
often so regularly that their recurrence causes a sort of rhythmical
rush and roar in the column of steam, etc., issuing above the mouth.
This was evident in the discharges issuing in 1857 from the highly
instructive minor _bocca_, then existing, examined by me, and referred
to ("Report, Naples, Earthquakes," etc. Vol. II., pp. 313, 314), as
presenting at the time great facilities for determining pyrometrically
the temperature of the lava within, and of the dry superheated steam
issuing with a rhythmic roar from it. M. Le Coq ("Epoques Geologiques
d'Auvergne," Tome IV.) has recorded some examples of the formation and
opening-out of large bubble-like cavities in lava already ejected.
Perhaps that able and laborious vulcanologist, whose death a few months
ago science still deplores, attributes too much importance as well as
magnitude to them, when attributing the formation of what he has
denominated "craters of explosion," to the mechanism of the rise and
bursting of such bubbles upon a gigantic scale. Such blowings forth,
sudden or prolonged, from particular spots of lava streams, _en route_,
undoubtedly may also have their origin in damp places, or water or
air-filled cavities in or beneath the bed over which the lava rolls,
which, getting gradually heated, generate steam, or air or gases under
tension by expansion, etc., which thus at length blow through the liquid
or pasty lava flowing above, and which in bursting through delivers much
dust also, and so simulates a little eruptive crater. Examples of this,
upon a great and convincing scale, can be pointed to in the Val di
Calanna and elsewhere on Etna.

[3] (P. 96). There are strong grounds for the gravest doubts that there
exists any real connection of a physical character between Volcanic
Eruptions, and Earthquakes more or less _approximately_ coincident only,
in time of occurrence; the respective sites being widely apart, and the
less the probability as the interven