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



and fronted by a superb barrier reef several miles offshore. Cliffs do not 

 seem to have been cut on the southwestern side of the higher-standing 

 land-masS;, G, J, during the penultimate cycle of erosion and abrasion, 

 blocks 4, 5, that was occupied by the cutting of great cliffs, K, on the 

 northeastern side and closed by the general subsidence which introduced 

 the present cycle, block 6 : hence the penultimate cycle was probably in- 

 troduced by a flexure of a preexistent island, EF, which caused elevation 

 along the northeast side of the island, H, and depression along the south- 

 west side, G; thus there would have been a reef -free shoreline of emer- 

 gence along the northeast coast, where cliffs, K, would be cut as long as 

 the island stood still or subsided slowly; and a reef -fronted shoreline of 

 submergence along the southwest coast, where deltas would advance in 

 the bays, while the offshore reefs would grow up and broaden; until the 

 later and more general subsidence of the whole island caused reef up- 

 growth to the present sealevel, much less continuous on the northeast than 

 on the southwest, but in general extending all around its long oval cir- 

 cuit, a crosswise segment of which is shown in block 6. 



Irregular deformation, in which subsidence ultimately dominates, 

 seems indispensable in explaining these various features. The northeast- 

 ern side of the island, M, finds its best explanation as a recently sub- 

 merged coast of long previous, H, and long enduring, K, emergence ; the 

 recent submergence must be due to subsidence, because it is of greater 

 measure than can be reasonably explained by a rise of ocean level, and it 

 Avas this subsidence that gave opportunit}^ for the imperfect u]3growth of 

 the present barrier reef. The earlier stages of this scheme are manifestly 

 hypothetical in a high degree, but they have at least the merit of taking 

 account of all that is known of the geology of the island, and of proceed- 

 ing by reasonable steps from its former subcontinental area, block 1, of 

 unknown extent, AB, through such a series of longer or shorter erosional 

 cycles as will account not only for the present shoreline and reefs, but 

 also for the special features of the uplands and the lowlands that are 

 seen on the two sides of the actual island. The physiographic evidence 

 in support of the successive deformations shown in figure 16 can not be 

 fully given here, but the coherence of the scheme thus graphically repre- 

 sented certainly speaks in its favor and warrants the association of New 

 Caledonia with Tahiti in confirmation of the hypothetical views expressed 

 above as to the probable development of cliffs around many emerged 

 islands during an early reef-free stage of their development. 



The upshot of all this chapter is that coasts of recent emergence and 

 coasts of more remote emergence, as well as coasts (except small islands) 

 of remote submergence, do not offer conditions favorable for reef growth; 



