riety
of coral formations in different parts of the globe, was induced to
reject the opinion that their shape represented the form of the original
bottom. Instead of admitting that the ring of dead coral rested on a
circular or oval ridge of rock, or that the lagoon corresponded to a
preexisting cavity, he advanced a new opinion, which must, at first
sight, seem paradoxical in the extreme; namely, that the lagoon is
precisely in the place once occupied by the highest part of a
mountainous island, or, in other cases, by the top of a shoal.

The following is a brief sketch of the facts and arguments in favor of
this new view:--Besides those rings of dry coral which enclose lagoons,
there are others having a similar form and structure which encircle
lofty islands. Of the latter kind is Vanikoro, (see map, fig. 39, p.
351,) celebrated on account of the shipwreck of La Peyrouse, where the
coral reef runs at the distance of two or three miles from the shore,
the channel between it and the land having a general depth of between
200 and 300 feet. This channel, therefore, is analogous to a lagoon, but
with an island standing in the middle like a picture in its frame. In
like manner in Tahiti we see a mountainous land, with everywhere round
its margin a lake or zone of smooth salt water, separated from the ocean
by an encircling reef of coral, on which a line of breakers is always
foaming. So also New Caledonia, a long narrow island east of New
Holland, in which the rocks are granitic, is surrounded by a reef which
runs for a length of 400 miles. This reef encompasses not only the
island itself, but a ridge of rocks which are prolonged in the same
direction beneath the sea. No one, therefore, will contend for a moment
that in this case the corals are based upon the rim of a volcanic
crater, in the middle of which stands a mountain or island of granite.

The great barrier reef, already mentioned as running parallel to the
north-east coast of Australia for nearly 1000 miles, is another most
remarkable example of a long strip of coral running parallel to a coast.
Its distance from the mainland varies from twenty to seventy miles, and
the depth of the great arm of the sea thus enclosed is usually between
ten and twenty fathoms, but towards one end from forty to sixty. This
great reef would extend much farther, according to Mr. Jukes, if the
growth of coral were not prevented off the shores of New Guinea by a
muddy bottom, caused by