m of the crater; a sort of gateway, by
which to enter and quit the cavity. This passage had no doubt been
formed by the exit of lava, which centuries ago had doubtless broken
through at this point, and contributed to form the visible reef beyond.
The height of this hole was some twenty feet, having an arch above it,
and its width may have been thirty. When Mark got to it, which he did by
descending the wall of the crater, not without risk to his neck, he
found the surface of the crater very even and unbroken, with the
exception of its having a slight descent from its eastern to its western
side; or from the side opposite to the outlet, or gateway, to the
gateway itself. This inclination Mark fancied was owing to the
circumstance that the water of the ocean had formerly entered at the
hole, in uncommonly high tides and tempests, and washed the ashes which
had once formed the bottom of the crater, towards the remote parts of
the plain. These ashes had been converted by time into a soft, or
friable rock, composing a stone that is called tufa. If there had ever
been a cone in the crater, as was probably the case, it had totally
disappeared under the action of time and the wear of the seasons. Rock,
however, the bed of the crater could scarcely be yet considered, though
it had a crust which bore the weight of a man very readily, in nearly
every part of it. Once or twice Mark broke through, as one would fall
through rotten ice, when he found his shoes covered with a light dust
that much resembled ashes. In other places he broke this crust on
purpose, always finding beneath it a considerable depth of ashes,
mingled with some shells, and a few small stones.

That the water sometimes flowed into this crater was evident by a
considerable deposit of salt, which marked the limits of the latest of
these floods. This salt had probably prevented vegetation. The water,
however, never could have entered from the sea, had not the lava which
originally made the outlet left a sort of channel that was lower than
the surface of the outer rocks. It might be nearer to the real character
of the phenomenon were we to say, that the lava which had broken through
the barrier at this point, and tumbled into the sea, had not quite
filled the channel which it rather found than formed, when it ceased to
flow. Cooling in that form, an irregular crevice was left, through which
the element no doubt still occasionally entered, when the adjacent ocean
got a