Darwin's Theory of Coral Beefs. Ill 



of barrier reefs, as represented on large-scale charts, are not as 

 a rule characterized by these systematically associated features. 

 The markedly irregular shore line of Kandavu in the Fiji group, 

 with its many headlands and bays, cannot have been produced 

 by marine action either before or after the barrier reef was 

 formed ; and the absence of extensive deltas in the bays dis- 

 credits the idea of a long still-stand of the island during the 

 cutting of the assumed platform around the headlands in the way 

 postulated in this theory. The platform would be most effect- 

 ively cut during a period of slow subsidence, provided no coral 

 reefs grew up in the way of the waves ; but the essence of 

 this theory lies in the exclusion of subsidence. In the 

 typical barrier reef of Bora Bora in the Society group, the 

 shore line of the central dissected volcanic island is of very 

 irregular pattern, such as marine erosion cannot produce ; the 

 sprawling ridges descend gently to their extremities in the 

 lagoon, and are not cut off in cliffs ; hence this barrier reef can- 

 not be regarded as a veneer on a wave-cut platform. More- 

 over, the embayments between the sprawling ridges are little 

 filled with delta plains ; hence the island cannot be supposed 

 to have stood still during the long and slow development of 

 the barrier reef by outward growth. The value of wave-cut 

 platforms therefore seems to be limited to narrow examples on 

 which fringing reefs can first establish themselves. 



Furthermore, if a veneering barrier reef were uplifted and 

 exposed to normal erosion, it would take the form of a more 

 or less dissected terrace ; new-cut ravines would disclose its 

 volcanic structure, the upper surface, strewn with coral sand 

 and silt, would transect the rock structure, and its outer edges 

 might retain patches of the veneering reef. Whether terraces 

 of this kind occur or not may be left to observant explorers to 

 determine. 



Murray's theory of outward growth and solution, whereby 

 fringing reefs are converted into barrier reefs during a pro- 

 longed still-stand of a volcanic island, is also a manifest possi- 

 bility ; but it involves several consequences, easily deducible 

 from the theory but not usually stated with it. For example, 

 if the central island be several miles in diameter and a thousand 

 or more feet in height, its streams will wash down abundant 

 detritus upon the fringing reef. By the time the reef has 

 grown outwards far enough to be called a barrier, the stream- 

 borne detritus will have formed deltas fronting the mouth of 

 each valley ; and with farther outgrowth of the reef the deltas 

 will become laterally confluent, so as to form a low alluvial 

 plain around the original shore line of the island ; and this 

 original shore line should not exhibit sinuosities of the kind 

 that are produced by subsidence. 



