et below they die. But now suppose that the land sinks very
gradually indeed. Let it subside by slow degrees, until the mountain
peak, which we have in the middle of it, alone projects beyond the sea
level. The fringing reef would be carried down also; but we suppose that
the sinking is so slow that the coral polypes are able to grow up as
fast as the land is carried down; consequently they will add layer upon
layer until they form a deep cup, because the inner part of the reef
grows much more slowly than the outer part. Thus you have the reef
forming a bed thicker upon the flanks of the island; but the edge of the
reef will be very much further out from the land, and the lagoon will be
many times deeper; in short, your fringing reef will be converted into
an encircling reef. And if, instead of this being an island, it were a
great continent like Australia, then you will have the phenomenon of a
barrier reef which I have described. The barrier reef of Australia
was originally a fringing reef; the land has gone slowly down; the
consequence is the lagoon has deepened until its depth is now 25 fathoms
and the corals have grown up at the outer edge until you have that
prodigious accumulation which forms the barrier reef at present. Now let
this process go on further still; let us take the land a further step
down, so as to submerge even the peak. The coral, still growing up, will
cover the surface of the land, and you will have an atoll reef; that is
to say, a more or less circular or oval ring of coral rock with a lagoon
in the middle. Thus you see that every peculiarity and phenomenon
of these different forms of coral reef was explained at once by the
simplest of all possible suppositions, namely, by supposing that the
land has gone down at a rate not greater than that at which the coral
polypes have grown up. You explain a Fringing Reef as a reef which is
formed round land comparatively stationary; an Encircling Reef as one
which is formed round land going down; and an Atoll as a reef formed
upon land gone down; and the thing is so simple that a child may
understand it when it is once explained.

But this would by no means satisfy the conditions of a scientific
hypothesis. No man who is cautious would dream of trusting to an
explanation of this kind simply because it explained one particular set
of facts. Before you can possibly be safe in dealing with Nature--who is
very properly made of the feminine gender, on account o