20 T ran sact ions. 



I 

 waves washed the cliffs and the rock platform in front of them bare, and 

 this gave the corals a firm foundation on which to attach themselves. 



If this interpretation be correct, the cliffs and reefs of Tahiti are not 

 only beyond explanation by the glacial-control theory, but the island is a 

 strong witness against the theory. It testifies that a reef-free island may 

 be strongly cliffed in a time that suffices only for the erosion of steep-sided 

 valleys : hence other islands, like Murea, Raiatea, and Huaheine, in the 

 Society Group, in which the submerged valleys are much less steep-sided 

 than those of Tahiti, ought to have been more strongly cliffed than Tahiti 

 if they were reef-free while their now-submerged valleys were in process of 

 erosion. The fact that they are not cliffed shows that they must have been 

 protected by living reefs, and thus discredits the assumption that reef-corals 

 were killed during the Glacial period. The possible less resistance of the 

 lavas on Murea and the other islands than on Tahiti does not affect the 

 argument, for if the Murean valleys are wider than the Tahitian valleys 

 because the rocks of Murea are weaker than those of Tahiti, then for 

 the same reason the spurs of Murea ought to be cut back in cliffs of 

 greater height than those of Tahiti. 



The reason for giving a special account of Tahiti is that, among the 

 many reef-encircled volcanic islands of the Pacific, it is unique in being 

 cut nearly all around its circuit by strong cliff's the bases of which are 

 now belov\' sea-level. Similarly, as noted above, Reunion is unique among 

 islands in the coral seas in being cut all around by cliff's the bases of which 

 are at the sea-level of to-day and are now undergoing attack by the sea in 

 the absence of protecting reefs. Two intermediate stages are represented 

 by the Marquesas and Tutuila (Samoa), which have submerged cliffs but 

 are not surrounded by barrier reefs. Of these two stages, Tutuila is the 

 later, because it has well-developed fringing reefs, while the Marquesas are 

 reef-free. Search for other islands of the Reunion, Tahiti, and intermediate 

 types is evidently desirable, for it is manifestly unsafe to generalize on a 

 few examples. Yet, inasmuch as these few examples confirm each other, 

 one is tenapted to ask whether they do not show the typical stages of early 

 development through which many deeply-dissected, reef-encircled volcanic 

 islands have long ago passed — that is, whether many deeply-dissected, reef- 

 encircled volcanic islands would not show reef-buried cliffs and platforms 

 on their submarine slopes if they could be examined. 



In a group of phenomena which ofl'er few examples of early stages and 

 many examples of later stages of development it certainly seems reason- 

 able to regard the examples of later stages as having passed through the 

 stages represented by the early examples, particularly when the early 

 exainples present the very features which a deliberate anal5'sis of the 

 problem leads one to regard as essential preliminaries to the features of 

 the more advanced examples. This interpretation appeals strongly to me, 

 because, instead of empirically entering the problem of reef-encircled 

 islands at a middle stage of progress, the attempt is made to trace out all 

 the stages of the problem from beginning to end. 



On the other hand, many students of coral reefs may regard it as 

 fanciful, not to say fantastic, to say that the cliff's which are still in process 

 of abrasion on Reunion, and wliich are partly submerged and fairly well 

 protected from wave-attack on Tahiti, are probably the counterparts of 

 similar cliffs now completely submerged on the reef-buried lower slopes of 

 many other volcanic islands. But another aspect of the problem deserves 

 consideration before a decision on this question should be declared. 



