described by Mr Dane in. 45 



indicating either that the shores are stationary, or that they 

 are now rising. 



Mr Darwin went painfully over every work in which any 

 account of coral reefs was to be found, and marked by colours 

 on a map to which of the three classes they belonged— of 

 " fringing reefs," " barrier reefs,"" or " atolls." On classifying 

 them in this way, the following general facts arrested his at- 

 tention : — 



1. They are not mingled indiscriminately, but generally those 

 of each class appear in groups, spread over a considerable area. 



2. Where they are mingled, the barrier reefs and atolls, 

 both of which indicate subsidence, are found together. 



3. On the other hand, fringing reefs and coral beds on terra 

 Jlrma, indicating that the land is either stationary or uprising, 

 are generally found together. 



4. Active volcanoes, the agents of elevation, are numerous 

 in the stationary or uprising groups, and, except in a very few 

 cases, are absent from the subsiding groups. 



Mr Darwin was thus led to conclude that the ocean contains 

 areas of elevation and areas of subsidence ; in other words, that 

 in some parts its bottom is sinking, and burying ancient lands 

 under the waves ; while in others, it is rising, and unveiling to 

 us the germs of future islands and continents. Let us pursue 

 this idea into a few details. 



The Maldive and Lacadive Atolls and Great Chagos Bank, 

 probably mai-k the former existence of on island extending 

 1500 miles from north to south, or equal in length to Britain, 

 France, and Spain united. 



In the Caroline Archipelago, northward of New Britain, we 

 have perhaps the ti*aces of a second island of similar size, of 

 which two or three small portions are still above water ; in the 

 Marshal], and Gilbert, and Ellice groups, traces of a third ; 

 in the Society Isles and Low Archipelago, a few remnants of a 

 fourth ; and in the Fidgi Islands, remnants of a fifth. According 

 to the theory also. New Caledonia and the north-east coast of 

 Australia have subsided, and may still be subiding. 



On the other hand, Sumatra, Java, Sumba, Timor, with Gil- 

 olo, the Philippines, Formosa, and Loo Clioo, which abound in 

 active volcanoes, and perhaps also Borneo and Celebes, belong 

 to the category of uprising lands. If we suppose that the cle- 



