CORAL FORMATIONS. 625 



began to disappear, the channel would become a lake, with a few peaks 

 above its surface ; then, a single peak of the old land might be all that 

 was left ; and finally this would disappear, and the coral reef come 

 forth an atoll, with its lagoon complete. 



Referring again to the figure : if, in the subsidence, the horizontal 

 line 2 become the sea-level, the former fringing reef f is now at b, a 

 barrier reef, and f is at V, and ch, ch', ch" are sections of parts of 

 the broad channel or area of water within ; over one of the peaks, P, 

 of the sinking island, there is an islet of coral, i : when the subsidence 

 has made the horizontal line 3 the sea-level, the former land has wholly 

 disappeared, leaving the barrier-reef, t, t', alone at the surface, around 

 a lagoon, III, with an islet, u, over the peak T, which was the last 

 point to disappear. 



These steps are well illustrated at the Feejees. The island Goro 

 (Fig. 966) has a fringing reef; Angau (Fig. 967), a barrier; Explor- 

 ing Isles (Fig. 968), a very dis- 

 tant barrier, with a few islets ; 

 Numuku (Fig. 969), a lake with 

 a single rock. The disappear- 

 ance of this last rock would make 



the island a true atoll. W- ' J 3 **' 



Whenever the subsidence ceases, ,<dPi«P% iJcta::: ^ 



the waves build up the land above 1^-^sF 



the reaeh of the tides ; seeds take Islands of the Fee J ee & rou P : Fi «- 



967, Angau ; 968, Exploring Isles ; 



root ; and the reef becomes cov- muku. 

 ered with foliage. 



The atoll Menchikoff (Fig. 962) was evidently formed, as explained 

 by Darwin, about a high island, consisting of two distinct ridges or 

 clusters of summits, like Maui and Oahu in the Hawaian group. 



If the subsidence be still continued, after the formation of the atoll, 

 the coral island will gradually diminish its diameter, until finally it 

 may be reduced to a mere sand-bank, or become submerged in tha 

 depths of the ocean. 



The rate of subsidence required to produce these results cannot exceed the rate of 

 upward increase of the reef-ground. On page 591, some estimates are given with re- 

 gard to the exceeding slowness of the movement. 1 



As coral debris is distributed, by the waves and currents, according to the same laws 

 that govern the deposition of silt on sea-coasts, it does not necessarily follow that the 

 existence of a reef in the form of a barrier is evidence of subsidence in that region. 

 On page 670, the existence of sand-barriers of similar position is shown to be a coni- 



1 For further information on the subject of Corals and Coral Islands, the reader ma}'- 

 refer to the author's Exploring Expedition Report on Zoophytes, 740 pp., 4to, and 61 

 plates in folio, 1846, and to his recent work on Corals and Coral Islands. 398 pp., 8vo, 

 1872; also to Darwin on the Structure and Distribution of Coral Reefs, 214 pp., 8vo, 

 with maps and illustrations, London, 1842. 

 40 



