210 FOUNDERS OF OCEANOGRAPHY 



It seems that A. Tylor, T. Belt, and others, had to some 

 extent anticipated Daly in attributing the origin of existing 

 coral reefs to a change in the ocean level consequent on 

 deglaciation ; but Daly has discussed the matter much 

 more fully than his predecessors in all its bearings, and has 

 brought forward many new facts in support of his views. 



The glacial-control theory is fundamentally opposed to 

 the Darwin-Dana theory, but is not inconsistent with 

 Murray's theory, from which it differs in details, such as 

 the method of formation of the platforms, but not in general 

 principle. Daly doubts whether archipelagos of atolls and 

 barrier reefs ever existed before the glacial period, though 

 possibly rare atolls may have been developed locally where 

 a limited subsidence affected the floor of the Tertiary ocean. 



In conclusion, it may be remarked that every serious 

 investigator of coral reefs seems to have added something 

 of importance, and that each of them, according to our 

 present views, seems to be right on some points and wrong 

 on others. It must be remembered that it is unlikely that 

 one theory will explain all the details of all reefs, which may 

 lie thousands of miles apart, and may have been formed under 

 very different conditions. 



Darwin and Dana showed how an atoll might be formed 

 on an area of subsidence, but their theory does not apply 

 to most atolls and barrier reefs that have been carefully 

 examined. 



Semper and A. Agassiz were correct in their criticisms of 

 Darwin's theory in the case of the reefs they had investigated, 

 and showed that atolls might be present where there was 

 no subsidence. 



Murray was right in his views as to the formation of 

 submarine platforms, and the possibility of these being 

 bmlt up to the required level, and also as to the process 

 by which a coral patch would naturally assume the atoll 

 form, but he was probably wrong as to the formation of 

 lagoons by solution. 



