GEOLOGICAL CONCLUSIONS. 391 
is possible that such barriers may often have existed in ancient 
time, and have disappeared through subsequent denudation of 
the surface. But may not the difference between the great even 
layers of the continental formations and those of a coral island 
have proceeded from the difference in the depth of the seas? 
Over the great shallow continental seas where the limestones 
were in progress, the waves may have generally been feeble, 
and therefore there may have been a less tendency to form nar- 
row barriers and deep intervening channels. 
The marsh condition of a drying-up lagoon with its forming 
limestones has been compared above with that under which an- 
cient unfossiliferous limestones were made. The narrow limits 
of the former make the comparison unsatisfactory ; for, in the 
coral island, coarsely fossiliferous beds are all the while form- 
ing about the exterior of the island, but a few miles at the 
most from the lagoon-marsh ; while the ancient limestones re- 
tain their unfossiliferous character often through many thou- 
sands of square miles. Still, the above mentioned difference be- 
tween the continental sea and the existing deep oceans may 
perhaps account for the diversity of results. 
VIII. CONSOLIDATION OF CORAL ROCKS. 
All true coral-reef rocks are examples of the consolidation 
of material mainly of coral origin—either mud, sands, frag- 
ments, or standing corals, the last with mud or sands inter- 
mixed—by (1), an under-water process; (2), at the ordinary 
temperature ; and they exemplify the mode in which all other 
submarine limestones of organic origin have been consolidated. 
The process appears to depend on the presence (proved by 
chemical analysis) of carbonic acid in the sea waters that bathe 
and penetrate the sands. This carbonic acid is derived from 
three sources: from (1), the rains which wash it down from 
