CORAL FORMATIONS. 277 



one of the encircling order, and this finally, by the disappear- 

 ance through the agency of the same movement of the central 

 land, into a lagoon island. In the former manner a reef skirting 

 a shore would be changed into a barrier extending parallel to, 

 but at same distance from, the mainland. 



Mr. Darwin then showed that there existed every inter- 

 mediate form between a simple well -characterised encircling 

 reef, and a lagoon island ; that New Caledonia supplied a link 

 between encircling and barrier reefs ; that the different reefs 

 produced by the same order of movement were always in juxta- 

 position, of which the Australian barrier associated with 

 encircled islets and true lagoons, affords a good example. He 

 then proceeded to show that within the lagoon of Keeling 

 Island, proofs of subsidence might be deduced from many 

 falling trees and a ruined storehouse ; these movements appear- 

 ing to take place at the period of bad earthquakes, which 

 likewise affect Sumatra, 600 miles distant. It was thence 

 inferred as probable, that as Sumatra rises (of which proofs are 

 well known to exist), the other end of. the level sinks down ; 

 Keeling Island thus acting as an index of the movement of the 

 bottom of the Indian Ocean. Again at Vanikoro, where the 

 structure indicates, according to the theory, recent subsidence, 

 violent earthquakes are known lately to have occurred. 



The author then removed an apparent objection to the 

 theory, namely, that subsidence would form a disc of coral, but 

 not a cup-shaped mass or lagoon, by showing that corals which 

 grow in tranquil water are very different from those on the out- 

 side, and less effective ; and that as the basin became shallower 

 they are subject to various causes of injury. The lagoon never- 

 theless is constantly filling up to the height of lowest water 

 spring tides (the utmost possible limit of living coral), and in 

 that state it long remains, for no means exist to complete the 

 work. Mr. Darwin then proceeded to the main object of the 

 paper, in showing that as continental elevations act over wide 

 areas, so might we suppose continental subsidences would do, 

 and in conformity to these views, that the Pacific and Indian seas 

 could be divided into symmetrical areas of the two kinds ; the 



