24 Proceedings of the Boyal Physical Society. 



that subsidence is adequate to provide platforms for coral- 

 reefs, but that these platforms could likewise be furnished by 

 the two other processes just referred to. Subsidence has 

 been invoked because no other solution of the problem seemed 

 admissible. But as another solution has been found, the 

 argument in favour of subsidence has no longer the same 

 force. The new solution, being based upon facts which are 

 everywhere observable in the coral regions, appears to me to 

 be more probable than the older one, which is only an infer- 

 ence resting on no positive proofs. 



2. The precipitous descent of the outer face of the reefs to 

 depths far below those at which corals can live, is another 

 difficulty which finds a ready explanation on the theory of 

 subsidence. If it were true, as is popularly assumed, that a 

 coral-reef presents towards the ocean a vast perpendicular 

 wall of limestone, entirely composed of solid coral, there could 

 be no escape from the conclusion that subsidence must have 

 occurred to permit of such an aggregation of coral-rock. We 

 learn, however, that much misconception exists on this sub- 

 ject. Some of the earlier accounts of the coral islands 

 speak of *' unfathomable " depths at a short distance sea- 

 wards from the reefs ; but more recent soundings afford no 

 confirmation of these statements. Instead of being the 

 summits of vast submarine pillars of limestone, atolls, as 

 weU as barrier-reefs, appear to be really planted on the 

 tops of submarine peaks and ridges. The outer face of the 

 reef itself is undoubtedly steep, in some places vertical. At 

 Tahiti, for example, as shown in Fig. 1, the living face of 

 coral may extend to a depth of 30 to 35 fathoms, beneath 

 which huge detached blocks of coral are piled up and cemented 

 together, forming a steep face which descends to about 150 

 fathoms at a distance of 180 fathoms from the outer edge of 

 the reef. The sea- bottom beyond that point is covered with 

 coral-sand, and slopes at 25° to 30°, after which the angle 

 lessens to 6°. By the abrading action of the breakers in 

 tearing off blocks of coral, and strewing them down in steep 

 talus-slopes, a platform is prepared on which the actually 

 growing part of the reef can build outwards. 



In Darwin's section of the Gambier Islands, the thickness 



