340 ALEXANDER AGASSIZ 



of each island stretches a coral reef. After seeing the 

 coral reefs growing on the denuded rims of these islands, 

 Agassiz was inclined to revert to the old opinion that 

 some of the lagoons of atolls represent the remains of 

 extinct craters. He found nothing unreasonable in the 

 suggestion that many of the small, fairly round atolls, 

 and others rising from great depths and isolated, are 

 the denuded rims of such craters as Thombia, or, if 

 larger, Totoya, upon which corals have obtained a 

 footing. 



The existence of some deep lagoons has been cited, 

 by the supporters of the theory of subsidence, as a proof 

 of its truth. Agassiz pointed out that if the theory were 

 true, all large lagoons should be deep. Lagoons of con- 

 siderably greater depth than that at which corals can 

 thrive he believes may be explained as the remains of 

 old craters into which the sea has broken during the 

 washing away of their walls. In support of this theory 

 he cites the fact that Haleakala in the Hawaiian Islands, 

 Aso San in Japan, and several volcanoes in Java, have 

 craters of a diameter fully equal to a number of the 

 Fiji atolls. 



Great bluffs are a characteristic feature of the lime- 

 stone islands of Fiji. Some of these rise to a height of 

 a thousand feet, and attest the elevation that has taken 

 place in that region. The faces of these bluffs are evi- 

 dently coralliferous. Agassiz found it was almost impos- 

 sible to collect corals from the exposed surfaces of these 

 cliffs with the appliances at hand, as the limestones had 

 become so hard that a hammer produced no impression 

 on them, and the corals were so well embedded that they 

 could not be cut out. 



From such examination as he was able to make, he 



