^ Proceedings of the Royal Physical Society. 



that, in his opinion, the theory of subsidence could not be 

 applied in the explanation of the Florida reefs ; that, on the 

 contrary, the southern end of Florida is built up on successive 

 contentric barrier-reefs, whicli have been gradually connected 

 and cemented into continuous dry land by the accumulation 

 and consolidation of mud-flats between them, and that this 

 process is still going on, and must eventually convert the 

 present keys and reefs from Cape Florida to the Tortugas 

 into similar land.^ 



In 1863, Professor Carl Semper published the results of 

 his researches among the Pelew Islands. He found himself 

 unable, by the theory of subsidence, to account for the 

 phenomena there presented, and threw doubts upon the 

 general applicability of that theory. He pointed out that, 

 while the southern islands, probably once atolls, consist of 

 coral rock upraised to from 400 to 500 feet above the sea, 

 and are flanked by living coast reefs, true living atolls exist 

 at the northern end of the group. He contended that there 

 is absolutely no evidence of subsidence ; that the association 

 of all the different kinds of reef within so circumscribed an 

 area seems entirely to disprove the notion of subsidence ; and 

 that, at least in this group of islands, Darwin's theory cannot 

 be applied. In some suggestive observations on their pro- 

 bable origin, he remarks that the reefs depend mainly, for 

 their form, upon the nature of the bottom on which they 

 begin. Atolls spring up on submarine banks. A species of 

 Porites takes root in little colonies, varying from the size of 

 the fist to masses six or eight feet in diameter. In time the 

 central portions of these growing colonies die, while the outer 

 parts flourish and gradually build up a ring of coral. This 

 ring, which maybe circular or, elongated in form, is some- 

 times continuous, but more commonly is traversed by one or 

 more channels. The interior portions are scoured out and 

 deepened by the tidal currents ; or, if the form of the bottom 

 and other conditions be suitable, a great many individual 

 masses of coral gradually grow into a more or less continuous 



^ Bull. Mus. Comp. Zool., vol. i. See also J. Le Coiite, Proc. Amer. 

 Assoc, X., 1856, part ii., p. 103; and E. B. Hunt, Sillimun's Journal, xxxv., 

 1863, p. 388. 



