536 W. M. DAVIS^ SUBSIDENCE OP REEF-ENCIRCLED ISLANDS 



from the valleys and detached from the cliffs must have been swept off- 

 shore by the cliff-cutting waves. If no subsidence had taken place, the 

 islands would be today not reef -encircled but cliff -rimmed. Even if the 

 changes of ocean level during the Glacial period be considered, high cliffs 

 must have been cut around the still-standing islands in preglacial time; 

 and the erosional and abrasional changes accomplished during the Glacial 

 epochs of lower sealevel could not have transformed the cliff-rimmed and 

 non-embayed islands of preglacial time into the non-clift and broadly 

 embayed islands of today. 



The only reasonable explanation that I have been able to invent for the 

 disappearance of the great volume of detritus shed from a deeply dis- 

 sected, reef-encircled island involves the aid of wave abrasion during a 

 considerable reef -free period of erosion before subsidence began or when 

 it proceeded slowly, and the absence of effective abrasion during a follow- 

 ing period of greater or more rapid subsidence of the island accompanied 

 by reef growth, after erosion was well advanced. Under such conditions, 

 most of the detritus would be swept offshore and deposited, while subsi- 

 dence was slow or absent, on the submarine flanks of the island before it 

 was encircled by reefs ; after subsidence became rapid or great enough to 

 permit reef establishment and upgrowth, the coarser part of the detritus 

 still to be discharged would be laid down in embayment deltas, and some 

 of the finer part would be swept out of the lagoon through the passes in 

 the reef, which were probably broader and more numerous in the earlier 

 stages of reef growth tlian they are today. 



THE SUBMERGED CLIFFS OF REEF-ENCIRCLED ISLANDS 



If the processes thus sketched have really taken place on such islands 

 as Huaheine and Borabora, they must have once had a rock platform, 

 DE, figure 4, and strongly clift spur ends, EF, around their shores; if 

 the platform gained a width of a mile, the spur-end cliffs must have had 

 a height of 1,000 or 1,300 feet. But Huaheine and Borabora have taper- 

 ing spurs, at the ends of which any little cliffs that occur, as at B, have 

 been cut by lagoon-wave abrasion at present sealevel, for they rise at the 

 back of narrow^, low-tide rock platforms. Hence the greater cliffs, that 

 were presumably cut before reef growth began, have disappeared by sub- 

 mergence, and submergence of so great an amount as is here implied can 

 be accounted for only by subsidence. 



It is not necessary, however, that the island should have stood still and 

 that the inferred platform and cliffs should have been cut back to a great 

 breadth, DE, and height, EF, before any subsidence took place; it suf- 



