CORAL AND CORAL REEFS. 15 



whereas with barrier-reefs and atolls there is a great difficulty 

 on this head. In barrier-reefs, from the improbability of the 

 rock of the coast, or of banks of sediment, extending in every 

 instance so far seaward within the required depth in which the 

 polyparia can work ; and in atolls, from the immensity of the 

 spaces over which they are interspersed (they spread, in a rough 

 line, 4500 miles in length), and the apparent necessity for 

 believing that they are all supported on mountain summits, 

 which although rising very near to the surface level of the sea, 

 in no one instance go above it. 



Dr. Darwin considers this a most improbable supposition, 

 and holds that there is but one alternative the prolonged sub- 

 sidence of the foundations on which the atolls were primarily 

 based, together with the upward growth of the reef-constructing 

 corals. On this supposition, every difficulty vanishes. Fring- 

 ing-reefs are thus converted into barrier -reefs ; and barrier- 

 reefs, when encircling islands, are thus converted into atolls, 

 the moment the last pinnacle of land sinks beneath the surface 

 of the sea. 



By this hypothesis alone can be explained the existence of 

 breaches opposite valleys, to which I have already alluded. 

 Little direct proof of subsidence can, however, be found in the 

 case of atolls and barrier-reefs, whereas the presence of up- 

 raised marine bodies on the fringed coasts show that these 

 have been elevated. The recent finding of fossil coral at a 

 considerable height in the island of Viti Levu would, in my 

 opinion, indicate that the lands encircled by the barrier-reefs 

 were elevated in like manner. This coral was found by a friend 

 of mine in 1876, during the campaign against the cannibals. 



The same authority goes on to say : ' We thus see vast areas 

 rising with volcanic matter every now and then, and bursting 

 forth through the vents or fissures with which they are tra- 



