304 CORALS AND CORAL ISLANDS. 
been made by the caving in of caverns with which the reef- 
rock is supposed to have been honey-combed, but does not 
apply the theory to all other atolls. The view is opposed 
by Mr. Heilprin, after a more recent study of the islands. 
There is nothing to sustain it among the Pacific atolls 
that were visited by the author. Ordinary atolls have no 
caves ; for caves in reef-limestones are made only after their 
elevation. The constant effort and action of the sea upon 
submerged reefs is to fill up all cavities and consolidate all 
sands. In the drift-sand ridges or hills, however, they may 
be formed without an elevation of the island; but these do 
é 
not exist in the region of the lagoon-basin. « ., //..: 
; 
sie , % t t 
As a further reply to the arguments adverse to Darwin’s 
theory from the West Indian seas, the following facts are 
here added. 
The arguments from the Florida and West Bahama reefs 
in favor of no subsidence have more weight than for most 
coral-reef regions. They appear to indicate that any subsi- 
dence once in progress long since ceased; but they are far 
from proving that the reefs of the seas have been formed 
without help from subsidence. There is evidence that a 
great subsidence occurred during the coral-reef era, affording 
all that the Darwinian theory demands. 
In a valuable paper by Mr. Alexander Agassiz, published 
in 1879 in the Bulletin of the Museum of Comparative Zodl- 
ogy,” the author points out that the South American conti- 
nent, in comparatively recent geological times, had connection 
with the West India islands through two lines: (1) one along 
a belt from the Mosquito Coast to Jamaica, Porto Rico, and 
1 Heilprin, Bermuda Islands, p. 44. 1889. 
2 An abstract of the paper is contained in the American Journal of Science, 
1880, xviii. 230. 



