off by waves filling up the space between the reef and the rock, the whole became an annular coral reef in closing a lagoon. The island and reef would thus present in a plan the following features successively.
In fig. 1. the rock is skirted by coral; in fig. 2. the rock is separated from the coral which widely encircles it by a channel of (deep) water; in fig. 3. the rock is not seen, but forms the base of the lagoon, and is covered by fragmented corals.
If, instead of a sinking, we next imagine a gradual rising or a stationary situation of some island, on which a circle of coral was fixed, the additional growth of this substance would be always on the outside, and the land would never be separated from a widely encircling reef by a channel of deep water.
In Mr. Darwin's views, the presence of a lagoon coral island is an evidence of depression of the solid land there; and, on the contrary, marginal coral reefs often supply evidence of rising, in addition to that furnished by shelly beaches at high levels. Thus may
