148 CORAL FORMATIONS. 
older limestone strata. Other portions of the rock, of less extent, are 
made of standing corals with the intervals filled in by reef-debris, and 
the whole cemented solid. The former variety here mentioned prevails 
about outer reefs exposed to the open seas; the latter among the zmner 
patches growing in quiet waters. The first kind is common about 
outer reefs, as large areas in the coral plantation are mere sand. It is 
still more abundant, forming the bottom among the inner patches, or in 
the lagoons, where the finer detritus is washed by the sea. A glance 
at the chart of the Feejees and Kingsmills will show how large a 
portion of the body of the reef increases from these fine accumulations. 
The exterior of a coral island, for a few hundred yards, excepting 
some islets within, is the only part which is the proper growth of the 
living reef. Within the exterior reef the coral structure may consist 
almost wholly of the compact homogeneous white limestone we have 
described. The elevated island of Metia was for a long time after 
elevation exposed to the ravages of the sea, before the present shore- 
reefs accumulated to give it protection. Proofs of degradation along 
the coast have been referred to. ‘There is much reason, therefore, for 
believing that the Metia now existing exposes on its eastern and 
southern sides at least (where particularly examined by us,) the interior 
of the original structure; and this view is supported by the compact 
character of the rock, (p. 67.) 
These reef-rocks receive also large contributions of sand or frag- 
ments from shells, which unite with the coral debris. 
II. Coral reefs, though they may stretch along a coast for scores of 
miles, are seldom a single mile in width at the surface : and if elevated 
above the sea, they would stand as broad ramparts separated by pas- 
sages mostly 20 to 200 feet deep, and often of great width. The sub- 
stratum, however, is continuous coral-rock ; and if these more elevated 
parts were removed by any process, after an elevation, they would 
leave an area of coral limestone often as extensive as the whole reef- 
grounds. ‘This is at once seen from the preceding figures. In an 
island like Dean’s, one of the Paumotus, these reef-grounds are 1000 
square miles in extent. It is true that the reefs at the surface gradually 
- widen if the land is undergoing no subsidence. But when situated on 
a sloping bank, as usual, this widening, as already illustrated, gradu- 
ally renders the bank steeper, and the rate of increase in width is rapidly 
diminished. And if the bank were not sloping, there is still reason to 
believe that the patches would not attain a great width at the surface of 
the sea, owing to the currents sweeping over them, occasioned partly 
