rivers charged with sediment which flow from the
southern coast of that great island.[1125]

Two classes of reefs, therefore, have now been considered; first, the
atoll, and, secondly, the encircling and barrier reef, all agreeing
perfectly in structure, and the sole difference lying in the absence in
the case of the atoll of all land, and in the others the presence of
land bounded either by an encircling or a barrier reef. But there is
still a third class of reefs, called by Mr. Darwin "fringing reefs,"
which approach much nearer the land than those of the encircling and
barrier class, and which indeed so nearly touched the coast as to leave
nothing in the intervening space resembling a lagoon. "That these reefs
are not attached quite close to the shore appears to be the result of
two causes; first, that the water immediately adjoining the beach is
rendered turbid by the surf, and therefore injurious to all zoophytes;
and, secondly, that the larger and efficient kinds only flourish on the
outer edge amidst the breakers of the open sea."[1126]

[Illustration: Fig. 118.

Supposed section of an island with an encircling reef of coral.

A, The island.

_b_, _c_, Highest points of the encircling reef between which and the
coast is seen a space occupied by still water.]

It will at once be conceded that there is so much analogy between the
form and position of the strip of coral in the atoll, and in the
encircling and barrier reef, that no explanation can be satisfactory
which does not include the whole. If we turn in the first place to the
encircling and barrier reefs, and endeavor to explain how the zoophytes
could have found a bottom on which to begin to build, we are met at once
with a great difficulty. It is a general fact, long since remarked by
Dampier, that high land and deep seas go together. In other words, steep
mountains coming down abruptly to the sea-shore are generally continued
with the same slope beneath the water. But where the reef, as at _b_ and
_c_ (fig. 118), is distant several miles from a steep coast, a line
drawn perpendicularly downwards from its outer edges _b c_ to the
fundamental rock _d e_, must descend to a depth exceeding by several
thousand feet the limits at which the efficient stone-building corals
can exist, for we have seen that they cease to grow in water which is
more than 120 feet deep. That the original root immediately beneath the
points _b c_ is actually as far from the surfac