954 The American Naturalist. [November, 
(radicle or caulicle) several inches long. These floit like corks 
on the water, and the little plant, which now resembles a cigar 
loaded at one end, is ready to strike root wherever it touches soil. 
The young tree grows apace, and further shows that it has come 
to stay by sending down roots from the branches, which serve 
as little guy ropes to anchor it firmly in the sand, This new 
land is in a state of constant ebb and flow, until its sand bars and 
dunes have been firmly cemented into coral limestone. The sea 
and the rain eat away the soft rock, carving it into fantastic forms. 
A soil however will gradually accumulate in little pockets at the 
surface, where the seeds of plants brought thither by birds, by 
wind or wave, immediately germinate, and cover the already old 
yet new island with a mantle of green. 
The whole subject of the formation of coral islands is now 
being vigorously discussed. 
Before Darwin's day it was generally believed that coral islands 
were incrustations on the top of lofty sub-marine mountains. But 
when, 50 years ago, Darwin made his celebrated "Beagle" voy- 
age (1832-1836), and afterwards published his account (second 
only in fame to his later theory of the origin of species), of the 
origin of these wonder-islands in mid-ocean, the older view was 
at once discarded. The key to his explanation was subsidence, 
the sinking of the ocean bottom. What were once table lands 
and mountains rising out of the Pacific and Indian Oceans, are 
now only sunken peaks crowned with coral limestone. The 
coral animals themselves impose, as we have seen, peculiar con- 
ditions. They require pure and warm sea-water and a bountiful 
supply of oxygen, and die if subjected to cold currents, to sedi- 
ments, and if carried to a greater depth than 100 to 120 feet. 
With these conditions the problem seems simple enough. The 
land, a volcano we will say, is very slowly sinking in a tropical 
sea. The coral polyps will attach themselves to its shores, will 
grow within the zone of their bathymetrical life limits, and will 
gradually build up a fringing reef As the mountain sinks this 
reef grows out from the land, since the outer corals exposed to 
the wash of the waves from the open sea, are in a better environ- 
ment than those next the shore, and hence grow the fastest 
