it now is, or nearly
at the same point, and there must, therefore, have been a tendency, from
the beginning, to a conical or dome-shaped arrangement in the ejected
materials. Secondly, were we to admit a great number of separate points
of eruption, scattered over a plain or platform, there must have been a
great number of cones thrown up over these different vents; and these
hills, some of which would probably be as lofty as those now seen on the
flanks of Etna, or from 300 to 750 feet in height, would break the
continuity of the sheets of lava, while they would become gradually
enveloped by them. The ejected materials, moreover, would slope at a
high angle on the sides of these cones, and where they fell on the
surrounding plain, would form strata thicker near the base of each cone
than at a distance.

What then are the facts, it will be asked, to account for which this
hypothesis of original horizontality, followed by a single and sudden
effort of upheaval, which gave to the beds their present slope, has been
invented? M. de Beaumont observes, that in the boundary precipices of
the Val del Bove, sheets of lava and intercalated beds of cinders, mixed
with pulverulent and fragmentary matter evidently cast out during
eruptions, are sometimes inclined at steep angles, varying from 15
degrees. to 27 degrees. It is impossible, he says, that the lavas could
have flowed originally on planes so steeply inclined, for streams which
descend a slope even of 10 degrees form narrow stripes, and never
acquire such a compact texture. Their thickness, moreover, always
inconsiderable, varies with every variation of steepness, in the
declivity down which they flow; whereas, in several parts of the Val del
Bove, the sheets of lava are continuous for great distances, in spite of
their steep inclination, and are often compact, and perfectly parallel
one to the other, even where there are more than 100 beds of
interpolated fragmentary matter.

The intersecting dikes also terminate upwards in many instances, at
different elevations, and blend (or, as M. de Beaumont terms it,
articulate) with sheets of lava, which they meet at right angles. It is
therefore assumed that such dikes were the feeders of the streams of
lava with which they unite, and they are supposed to prove that the
platform, on the surface of which the melted matter was poured out, was
at first so flat, that the fluid mass spread freely and equally in every
direction, and no