March 26, 1915] 



SCIENCE 



457 



almost neutralized by the lowering of the sea- 

 level during the oncoming of a glacial epoch, 

 and at such a time coral reefs would broaden 

 and lagoons would become shallow; but with 

 the passing of a glacial epoch the return of 

 ice-sheet water to the ocean would accelerate 

 the submergence due to subsidence, and at 

 such a time coral reefs might be more or less 

 completely drowned: thus the discontinuity 

 of certain reefs on so-called " platforms " may 

 be explained. 



All the phenomena which testify to the f orr 

 mation of coral reefs on subsiding foundations 

 can be equally well explained by the assump- 

 tion of a rise of the ocean surface around or 

 over fixed foundations : but a rise of the ocean 

 surface in any coral-reef region demands a 

 rise of the whole ocean surface; and if the 

 coral-reef foundations are to stand still, a rise 

 of the whole ocean surface can be explained 

 only as the diminished result of a greater rise 

 of the ocean floor in some non-coral-reef region. 

 The conditions involved in this alternative for 

 the simple theory of local subsidence are so 

 extravagantly improbable that, as soon as they 

 are explicitly defined, they must be rejected. 



Wo absolute demonstration of the origin of 

 coral reefs, or, for that matter, of any other 

 geological structure, is possible : the most that 

 can be hoped for is a highly probable conclu- 

 sion. The conclusions announced above in 

 favor of Darwin's theory are believed to have 

 about the same order of probability as that 

 usually accepted as " proof " in geological 

 discussions. 



A number of local conclusions may be briefly 

 announced as follows : 



The elevated reef along the south coast of 

 Oahu, Hawaii, was formed during or after a 

 sub-recent period of subsidence, for its lime- 

 stones enter well-defined valleys that must have 

 been eroded when the island stood higher than 

 now, before the reef-limestones were deposited 

 in them. 



The Fiji group has suffered various move- 

 ments of subsidence and elevation by which 

 its many islands were affected in unlike ways. 

 Elevation has taken place at different times in 

 different islands, for some of the elevated reefs 



are elaborately dissected, others are very little 

 dissected, and still others remain at sea-level. 

 The embayments due to the latest subsidence 

 on the larger islands, Viti Levu and Vanua 

 Levu, are now largely filled with delta plains. 

 All the reefs, those now elevated as well as 

 those at sea-level, appear to have been formed 

 during periods of subsidence, the evidence af- 

 forded by the elevated reefs of Vanua Mbalavu, 

 Mango and Thithia, being especially significant 

 on this point. The medium-sized island of 

 Taviuni has few visible reefs, because its 

 flanks and shores are flooded by sheets of 

 recent lava. The small island of Wakaya 

 seems to be a tilted block of lava beds, not a 

 dissected volcano. 



The extensive barrier reef of New Caledonia 

 has grown up during a recent subsidence by 

 which that long and maturely dissected island 

 has been much reduced in size and elaborately 

 embayed; but unlike most encircled islands this 

 one was strongly cliffed around its southeastern 

 end and along much of its northeastern side 

 before the recent subsidence took place. 



The two southeastern members of the 

 Loyalty group. Mare and Lifu, are former 

 atolls, evenly unlifted about 300 feet: Mare 

 shows a small hill of volcanic rock in the center 

 of its limestone plateau or elevated lagoon 

 floor. Uvea, the northwestern of the three 

 Loyalty Islands, is a slightly tilted atoll; its 

 eastern side shows an uplifted reef in crescentic 

 form, 100 or more feet high at the middle of 

 its crescent, and slowly descending to sea- 

 level at its horns ; the tilted lagoon floor slowly 

 deepens westward and is enclosed by discon- 

 nected, upbuilt reef-islands. 



The New Hebrides show signs of uplift in 

 their elevated reefs, and of depression in their 

 embayments. There is some evidence that 

 certain uplifted fringing reefs on the island of 

 Efate, near the center of the group, were 

 formed during pauses in a subsidence that 

 preceded their uplift, and not during pauses in 

 their uplift as inferred by Mawson. The nar- 

 rowness of the lagoons enclosed by the barrier 

 reefs that encircle certain strongly embayed 

 islands in this group may be explained by sup- 

 posing alternations of slow and rapid subsi- 



