388 CORALS AND CORAL ISLANDS. 
This is the only method by which any thick stratum of lime. 
stone could be made out of a single set of species, for all 
such species have a narrow range in depth ; and the only way, 
from any succession of species, if those species are alike in 
range of depth. | 
In the case of existing coral reefs, there is yet no evidence 
that the species of the lower beds differ from those of the top. 
There is also no evidence, in any part of any ocean, that there 
is a set of cold-water corals fitted to commence a reef in deep 
water and build it up to such a level that another set of 
species may take it and carry it up higher; the facts thus far 
gathered are all opposed to such an idea. Should it be here- 
after proved that the corals of the inferior beds differ in 
species from those now existing, it will probably be found that 
the predecessors of those now living were also shallow-water 
species; so that the subsidence in any case was necessary. 
_V. DEEP-SEA LIMESTONES SELDOM IF EVER MADE FROM CORAL 
ISLAND OR REEF DEBRIS. 
This point has been discussed on pages 143, 211. The facts 
show that the sediment or débris from a shore is almost wholly 
thrown back by the waves against the land where it originated, 
or over its submerged part in the shallow waters, and that it — 
is not transported away to make deep sea formations. 
The facts have also a wider bearing, for they teach that 
lands separated by a range of deep ocean cannot supply one 
another with material for rocks. The existence of an Atlan- 
tic ocean continent—an Atlantis—has sometimes been assumed 
in order to make it a source of the mud, sand and gravel, out 
of which the thick sedimentary formations of the Appalachian 
region of North America were made. But if this Atlantis 
were a reality, there would still have been needed, in addition 

