oG-i TV. M. DAT1S SUBSIDENCE OF EEEF-EXCIECLED ISLANDS 



margin of a full-width platform. KO, the slope would be worn during the 

 slow submergence into the curve. QTSK. and the next abraded platform 

 would have small ^idth, near K, and a vague margin toward S. 



It is of course eminently possible that unduly narrow platforms, like 

 JH, and worn-off slopes, like ST, may have formerly been abraded at 

 times of small or slow submergence : but the many examples of fuU- width 

 platforms, BA, that I saw on various islands in the Fiji group, on both 

 sides of Xew Caledonia, on small islands in the lagoon of the Great 

 Barrier reef of Australia, and on some of the Society Islands, suggest that 

 submerged platforms, DE, are full-width like the visible platforms, AB. 

 and that a submergence of 10, 20, or 30 feet has usually been accomplished 

 in a relatively short-time interval between longer stationary periods. This 

 appears to be better accounted for by intermittent subsidence than by 

 climatic fluctuations of ocean level. In any case the present relation of 

 ocean surface and island level appears usually to have been assumed by a 

 relatively rapid advance of the sea on the land, sufficient to drown any 

 earlier bluff, DC, corresponding to the present bluff, AL. 



It is therefore probable that such a change of level would submerge any 

 preexistent barrier reef (not drawn in figure 17) and transform it for a 

 time into a ^^sunken barrier," similar to the sunken barriers of Xew 

 Guinea; and that while it was submerged the waves of the lagoon were 

 strengthened by so much of the exterior waves as could cross the barrier. 

 The greater part of the spur-end platform and bluffs may have been then 

 cut before protecting friaging reefs were formed on the new shoreline. 

 Since the barrier reef has again grown up to sealevel and the fringing 

 reefs have broadened in front of the spur ends, wave-work on the plat- 

 forms and bluffs must have been weakened. The lagoon waves are, how- 

 ever, stiQ strong enough to lift coral blocks onto the margin of the fring- 

 LDg reef and to wash the platform fairly clean, leaving only patches of 

 gravel on its surface and occasional large rock blocks near the cliff base. 



The Origix of Atolls 



rs'direct etidexce from barrier reefs 



This article has been given the title, ''The Subsidence of Eeef-encircled 

 Islands,'* because it is devoted chiefly to the behavior, as to elevation or 

 subsidence, of islands that are bordered by fringing reefs or inclosed 

 within barrier reefs. As far as such islands and the reefs around them 

 are concerned, the various lines of evidence given on the preceding pages 

 appear to me to warrant the acceptance of Darwin's tlieorj' of intermittent 

 subsidence as superior to all its competitors : and in so far as the theorv is 



