Narrative of the surveying voyages of his majesty's ships Adventure and Beagle (vol.3)

566 CORAL FORMATIONS. April, 1836. best characterized of these, namely, the Maldive islands, ex- tend in length for 480 miles, with an average breadth of sixty. These atolls agree in most respects with the lagoons of the Pacific; they differ, however, in several of them being crowded together — such little groups being separated from other groups by profoundly deep channels. Now if we look in a chart, at the prolongation of the reef towards the northern end of New Caledonia, and then complete the work of sub- sidence, so as to continue producing the same results ; we should have the original reef broken up into many patches each of which, from the vigorous growth of coral on the out- side, would have a constant tendency to assume a rounded form. Every accidental break in the continuity of the first line would determine a fresh circle. In the case, therefore, of the Low or Dangerous Archipelago in the Pacific, I beheve that the lagoon islands were moulded round the flanks of so many distinct islands ; but in the Maldives, that one single mountainous island, bordered by reefs, and very nearly of the same actual figure and dimensions with New Caledonia, for- merly occupied that part of the ocean. Lastly, to the extreme westward, the coast of Africa is closely skirted by coral reefs, and according to facts stated in Captain Owen's voyage, has probably been uplifted within a recent period. The same remark applies to the northern part of Madagascar, and, judging from the reefs likewise at the Seychelles, situated on the submarine prolongation of that great island. Between these two, N.N.E. and S.S.W. lines of elevation, some lagoon and widely-encircled islands indicate a band of subsidence. When we consider the absence both of M'idely-encircling reefs and lagoon islands in the several archipelagoes and wide areas, where there are proofs of elevations ; and on the other hand the converse case of the absence of such proof where reefs of those classes do occur; together with the juxta- position of the different kinds produced by movements of the same order, and the symmetry of the whole, I think it will be difficult (even independently of the explanation it