CORAL REEFS 97 



large. There are many objections to this premiss, and 

 there is no evidence that such a general subsidence is 

 going on as would be required by the theory — i.e., from 

 Africa across the East Indies to Hawaii and the 

 Paumotuas. Indeed, it seems probable that the 

 method of formation by subsidence, if it occurs at all, 

 is very local, but the theory is of such interest that 

 in every place visited evidence of subsidence should 

 be sought. The charts should be examined for sunken 

 valleys, etc., and any additional soundings made out- 

 side the reefs will be of interest. A difficult point in 

 this theory is the explanation of the causes which 

 prevent the lagoons of atolls and barrier reefs from 

 ever exceeding 50 fathoms in depth. The question 

 — how far, if at all, material is swept over the 

 reef to fill in the lagoon — is worthy of investi- 

 gation, and in this connection the examination of 

 the mode of formation of the lagoon flat should be 

 interesting. 



The subsidence view of the successive formation of 

 barrier reefs and atolls is strongly opposed by all those 

 who have in the last thirty or forty years devoted 

 themselves to the examination of extensive series of 

 reefs, mention in particular being made of the late 

 Prof. A. Agassiz, Sir John Murray, and Prof. Semper, 

 who, in addition to having made explorations themselves, 

 have had the benefit of the experience of numerous 

 other expeditions. It has in the first place been shown 

 that fringing reefs might owe much of their breadth 

 to the abrasion of the land round which they form a 

 belt. Tall cliffs and masses of rock overhanging at their 

 base are often found bounding the shoreward edges of 

 the reefs. The land in such a position is evidently 

 being rapidly removed between tide-marks, and as it 

 recedes it leaves at low-tide level a flat platform, 



7 



