500 W. M. DAVIS SUBSIDENCE OF REEF-ENCIRCLED ISLANDS 



reason to suppose that volcanic islands in their originally completed form 

 "iisnally possessed a weak ash cloak over a resistant lava core ; and in the 

 absence of snch a cloak, the strong slopes now visible in dissected volcanic 

 islands above sealevel may be reasonably supposed to extend with but 

 small decrease of declivity to a considerable depth below sealevel. 



It is also possible that a sloping platform might be produced by the 

 offshore deposition of waste from a reef -free island, but in this case the 

 island would be strongly clift by the unrestrained waves; and the cliffs 

 would still be partly visible after moderate submergence. The best way 

 of making a platform in the coral seas is in association with reef up- 

 growth around subsiding islands, or with reef outgrowth around still- 

 standing islands, as Darwin supposed. 



The reasonableness of the supposition that supermarine slopes may be 

 prolonged below sealevel is increased when the depth of the sea-bottom, C, 

 figure 4, outside of a barrier reef, E, is taken into account; it then ap- 

 pears more clearly than before that the. reef foundation is best indicated 

 by prolonging the visible slope, AB, of the island spurs down to the sea- 

 bottom, C, outside of the reef. The total reef is thus seen to form a huge 

 terrace-like mass on the submarine fianks of the volcanic cone; and the 

 thickness of the terrace is thereby shown to be much greater than the 

 depth of the lagoon; indeed, judging by the slopes that prevail in dis- 

 sected volcanic islands, the thickness of the reef mass may usually be 

 taken at a fifth or a quarter of the lagoon breadth: a barrier reef that 

 stands a mile away from its encircled island may well have a thickness of 

 1,000 or 1,300 feet. Subsidence is manifestly necessary to produce reef 

 upgrowth of so great a measure. But it is eminently possible that lagoons 

 several miles in breadth, containing eccentrically placed islands, may be 

 underlaid by an uneven surface of hills and valleys, maturely worn down 

 on a group of volcanoes of unequal heights : here the slope of a surviving 

 island can not be safely prolonged far from its shoreline. 



In the case of barrier reefs along continental borders, the large area of 

 land back of the shoreline will usually afford opportunity for better de- 

 termination of changes of level than is provided by comparatively small 

 volcanic islands: thus the Great Barrier reef of Australia is best inter- 

 preted by taking account of the physiographic development of the interior 

 highlands as interpreted by Andrews (1903), from which it appears that 

 the reef foundation probably subsided while the highlands rose, thus indi- 

 cating, as I have lately shown (1917, c), that the coast and the adjoining 

 sea-floor have been flexed, the coast up and the sea-floor down, and that 

 the reef, growing up from the down-flexed area, has a great thickness. I 

 must therefore dissent from the conclusion reached by Yaughan, that 



