28 Proceedings of the Royal Physical Society. 



the origin of coral-reefs are accepted. If, as Darwin sup- 

 posed, the coral-islands of the Pacific and Indian Oceans 

 represent the last peaks of submerged continents, it is in- 

 credible that continental rocks should not be found among 

 them. The oceanic islands (except, of course, those composed 

 of coral-rock) are of volcanic origin, and show none of the 

 granites, schists, and other rocks which might have been 

 looked for on such elevated summits. They have been piled 

 up by the accumulation of lavas and tuffs discharged from 

 the earth's interior, and, where they occur, point to uj)heaval 

 rather than to subsidence. Again, as Mr Murray has shown, 

 the inorganic deposits of the ocean-floor are composed of 

 volcanic debris, with a singular absence of the minerals that 

 constitute the usual crystalline rocks of our continents.^ 



No satisfactory proofs of a general subsidence have been 

 obtained from the region of coral-reefs except from the 

 structure of the reefs themselves, and this is an inference 

 only, which is now disputed. From the nature of the case, 

 indeed, traces of subsidence can hardly be expected. A few 

 examples have been cited, such as the occurrence of trunks 

 of cedar trees in a layer of red soil in Bermuda, lying between 

 the calcareous deposits at a depth of 42 feet below low-water 

 mark. This indicates a recent subsidence of that tract, but 

 it may be merely local, and may be due to the sinking down 

 of the roof of one of the caverns with which the limestone is 

 so abundantly honeycombed. Occasionally, along the margins 

 of lagoons, trees are found at the water- edge, in a position 

 suggestive of subsidence. But the removal of the calcareous 

 rock by solution or wave action might equally account for 

 their condition. 



Of elevation in the region of atolls and barrier-reefs there 

 is almost everywhere more or less distinct evidence. Pro- 

 fessor Dana has collected the facts which prove that recent 

 elevatory movements of unequal and local extent have 

 occurred in all parts of the ocean.^ 



Upheaval has taken place even in areas where barrier- 



1 Proc. Roy. Soc, Edinb., 1876-77, p. 247 ; Murray and Reuard, Brit. Assoc. 

 Report, 1879, p. 340. 



2 Corals and Coral Islands, 2d edit., p. 284. 



