edge overlaps a little the central cellular lava. This structure is so
symmetrical round the entire circumference, that one is forced to suppose
that the bomb burst during its rotatory course, before being quite
solidified, and that the lip and edges were thus slightly modified and
turned inwards. It may be remarked that the superficial ridges are in
planes, at right angles to an axis, transverse to the longer axis of the
flattened oval: to explain this circumstance, we may suppose that when the
bomb burst, the axis of rotation changed.

AERIFORM EXPLOSIONS.

The flanks of Green Mountain and the surrounding country are covered by a
great mass, some hundred feet in thickness, of loose fragments. The lower
beds generally consist of fine-grained, slightly consolidated tuffs (Some
of this peperino, or tuff, is sufficiently hard not to be broken by the
greatest force of the fingers.), and the upper beds of great loose
fragments, with alternating finer beds. (On the northern side of the Green
Mountain a thin seam, about an inch in thickness, of compact oxide of iron,
extends over a considerable area; it lies conformably in the lower part of
the stratified mass of ashes and fragments. This substance is of a reddish-
brown colour, with an almost metallic lustre; it is not magnetic, but
becomes so after having been heated under the blowpipe, by which it is
blackened and partly fused. This seam of compact stone, by intercepting the
little rain-water which falls on the island, gives rise to a small dripping
spring, first discovered by Dampier. It is the only fresh water on the
island, so that the possibility of its being inhabited has entirely
depended on the occurrence of this ferruginous layer.) One white ribbon-
like layer of decomposed, pumiceous breccia, was curiously bent into deep
unbroken curves, beneath each of the large fragments in the superincumbent
stratum. From the relative position of these beds, I presume that a narrow-