he supposed, in a state of security upon a
small island, but that, before the farmer had
returned to his own house, the fire appeared
to break out from that very island,
and he had the misfortune to be the sad
spectator of the ruin of himself and family.
This account, indeed, was, as far as I was
able to ascertain, perfectly correct, but nevertheless
it does not at all prove that the
accident was caused rather by the earth itself
being on fire than by a fire-stream: for, at
the very moment that the farmer had collected
his sheep upon this spot, the lava was
rushing along with the greatest imaginable
rapidity, and took quite a different course
from that which was at first expected; proceeding
towards a neighboring river and
along its channel, till it arrived at the
island, which it burned together with the
sheep.
The Icelandic annals relate a long series
of such eruptions, continued through whole
centuries; but we do not find anv account
distinctly describing the nature of the lava-
streams which formerly over-ran whole districts.
The damage sustained is simply
noticed, but upon the subject of the progress
of the fire, authors are entirely silent. It is
therefore quite natural that the late fire
should appear particularly frightful to the
spectators, and that they should be led to
suppose that the fire broke forth from the
entire and uninjured crust of the earth, at
considerable distances from the fire-stream
itself. My own opinion is that such appearances
always exist when the stream of lava,
for the first time, pours down from the
mountains upon a fertile tract of land, or
upon a soil that is loose and free from obstructions.
We may readily imagine what
an immense weight must fall upon the earth,
when the lava rushes down from the high to
the low lands, and we may in like manner
conceive it possible for the lava to burn and
force itself a passage to a considerable depth
below the surface of the earth. The lava
itself being a fluid, driven forward by every
new accession of matter, it can, without
doubt, proceed in its course as well below
as above the surface of the ground, and even
in some instances with greater rapidity.
Above it must work its way over all inequalities,
and where the stream of lava has