CORAL REEFS AND SUBMARINE BANKS 205 
succeeded subsidence” in the Friendly or Tonga group, and of 
“subsidence having probably succeeded recent elevation” in the 
Harvey or Cook group (140). Finally Darwin expressed the general 
opinion: ‘It has already been shown (and it is, perhaps, the most 
interesting conclusion in this volume) that the movements [of 
subsidence] must either have been uniform and exceedingly slow, 
or have been effected by small steps, separated from each other by 
long intervals of time” (145). 
Structural features of reefs formed during subsidence.—The 
formation of coral reefs along recently uplifted coasts or around 
young volcanoes is improbable, because the loose detritus which 
prevails on the shore belts of such coasts is unfavorable to coral 
Fic. 1.—Structure of coral reef formed around a volcanic island by subsidence 
growth. In the absence of reefs, coasts of these kinds will, while 
their slopes are dissected by streams, be attacked and cut back 
by the sea, except where the deltas of large rivers are built forward; 
as long as no change of level takes place, beach detritus will be 
spread along the shore and reef growth will be prevented. But if 
subsidence occurs, the dissected coast will be embayed; detritus 
will then be held in the embayments and reef growth may begin 
on the rocky headlands or on bare ledges off shore... The longer 
the subsidence continues at a moderate rate the thicker the reefs 
will become. ‘The essential structural features of reefs thus formed 
around a volcanic island are represented in Fig. 1, in which UVU 
is the initial cone, formed by eruption, with sea-level at S:, and 
UWU is the dissected and clift cone at the time when subsidence 
begins; ZT represents the ring of detritus swept off shore. Re- 
union, a clift and almost reefless island in the Indian Ocean, repre- 
sents this stage. 
tSee “‘Clift Islands in the Coral Seas,” Proc. Nat. Acad. Sci., II (1916), 284-88. 
