ys of the island find
a long descending passage to the foot of the precipice, and emerge into
light on the edge of the grassy talus half-way down the hill. It
reminded me of the tunnel in the rock through which Imlac opened up a
way of escape to Rasselas from the happy valley,--the "subterranean
passage," begun "where the summit hung over the middle part," and that
"issued out behind the prominence."

From the commencement of the range of cliffs, on half-way to the
shieling, I found the shore so thickly covered up by masses of trap, the
debris of the precipices above, that I could scarce determine the nature
of the bottom on which they rested. I now, however, reached a part of
the beach where the Oolitic beds are laid bare in thin party-colored
strata, and at once found something to engage me. Organisms in vast
abundance, chiefly shells and fragmentary portions of fishes, lie
closely packed in their folds. One limestone bed, occurring in a dark
shale, seems almost entirely composed of a species of small oyster; and
some two or three other thin beds, of what appears to be either a
species of small Mytilus or Avicula, mixed up with a few shells
resembling large Paludina, and a few more of the gaper family, so
closely resembling existing species, that John Stewart and Alister at
once challenged them as _smurslin_, the Hebridean name for a well-known
shell in these parts,--the _Mya truncata_. The remains of
fishes,--chiefly Ganoid scales and the teeth of Placoids,--lie scattered
among the shells in amazing abundance. On the surface of a single
fragment, about nine inches by five, which I detached from one of the
beds, and which now lies before me, I reckon no fewer than twenty-five
teeth, and twenty-two on the area of another. They are of very various
forms,--some of them squat and round, like ill-formed small
shot,--others spiky and sharp, not unlike flooring nails,--some straight
as needles, some bent like the beak of a hawk,--some, like the palatal
teeth of the Acrodus of the Lias, resemble small leeches; some, bearing
a series of points ranged on a common base, like masts on the hull of a
vessel, the tallest in the centre, belong to the genus Hybodus. There is
a palpable approximation in the teeth of the leech-like form to the
teeth with the numerous points. Some of the specimens show the same
plicated structure common to both; and on some of the leech backs, if I
may so speak, there are protuberant knobs, that indicate