172 
CORAL ISLANDS* 
Fig. 39. 
a a, the habitable part of the island, b, the lagoon. 
islands are owing to their being the crusts of sub- 
marine volcanoes, Mr. Lyell states that every island 
yet examined in the wide region termed Eastern 
Oceanica, consists either of volcanic rocks or coral 
limestones ; and that in some of them, as in Gam- 
bier's Group, rocks of porous lava actually rise up 
in the centre. That many of these islands have 
been raised from the sea by volcanic action, there 
can be no doubt ; for on the summit of the highest 
mountain in Tahiti, an island composed almost en- 
tirely of volcanic rocks, there is a distinct stratum 
of fossil c,oral, resembling that of modern reefs. In 
addition, we may state that MM. Quoy and Gaim- 
ard describe the shores of Coupang and Timon as 
formed of coral beds twenty-five or thirty feet in 
thickness, and that above these repose vertical beds 
of slate, traversed by quartz ; also, that in the Isle 
of France, a coral bed ten feet thick occurs be- 
tween two lava-currents. 
One circumstance in relation to these coral isl- 
ands is worth remarking, and that is the deep, nar- 
row passage which almost invariably leads from the 
sea to the lagoon in the centre. This is kept open 
by the water rushing in during high tide and rush- 
ing out again at low tide, and with such force 
as to prevent the coral animals from raising their 
structure. In the same manner, the deepest chan- 
nel of our harbour (Gedney's Channel) is kept open 
by the strong ebb tide which sweeps out the sand, 
which would otherwise soon block it up, and de- 
posites it in the ocean. But if we follow up the 
Hudson to the Overslaugh, near Albany, where the 
tide is scarcely felt, we find the channel constantly 
blocking up with sand, and obstructing navigation, 
