202 
GEOLOGY: W. M. DAVIS 
Andaman islands in the Bay of Bengal are elaborately embayed, and a sub- 
marine platform several miles in breadth and from 20 to 40 fathoms in depth 
adjoins them; a bank of similar depth, measuring 35 by 10 miles lies not 
far away to the east; yet these islands have no barrier reefs and only very 
narrow and discontinuous fringing reefs; their submergence must be very 
recent. If the postglacial rise of ocean level were the cause of all these sub- 
mergences, reefs should everywhere be of about the same volume: as a matter 
of fact they vary in volume enormously. The fringing reefs of southwest 
Palawan, of Yap in the western Pacific, and of Rodriquez in the southern 
Indian ocean are two or three miles wide; various atoll and barrier reefs are 
half a mile or a mile wide. Others are much narrower, and still others are 
discontinuous, or so imerfectly developed as not to reach sea level: finally, 
some submarine banks are flat, without any reef rim. These irregular values 
speak for variable subsidence of the reef foundations, not for a uniform rise of 
ocean level. 
The occurrence of subsidence in the Australasian region at a more rapid 
rate than that of coral upgrowth has an interesting bearing on the origin of 
the numerous submarine banks, of various depths down to 40, 50, or 60 
fathoms, in the China sea. It is evident that if islands suffer a movement of 
subsidence rapid enough to drown their barrier reefs and thus to develop 
fringing reefs of a new generation, the same subsidence would completely sub- 
merge neighboring atolls. Furthermore, if the rapid subsidence of a group of 
islands were of so recent a date that the resulting fringing reefs are still nar- 
row, the drowned reefs of the submerged atolls might still remain below sea 
level, even if the amount of submergence had not been great enough to kill 
all their corals. 
Now in view of the proximity of the China sea to the Philippine Islands, it 
seems reasonable to suppose that its floor has shared some of the many move- 
ments by which the islands of the archipelago have been disturbed; hence 
the submarine banks of that deep basin are best explained, following Darwin's 
theory, as drowned atolls not yet rebuilt. The date as well as the rate and 
the amount of a subsidence is therefore of importance in determining whether 
the atolls that it drowns shall still be submerged. Certainly recent submer- 
gence, presumably due to subsidence, has affected the Macclesfield and certain 
other large submarine banks of the China sea, for corals are growing on the 
rims of many of them, but have not yet built the rims up to the sea surface. 
On the other hand in spite of the proximity of the unstable Philippines, the 
Glacial-control theory explains the Macclesfield bank, which is taken to be 
a typical example of its kind, as the remains of a huge volcanic cone, origi- 
nally as large as Hawaii, which stood still long enough to be worn down to 
small relief in preglacial time, and which, still fixed, was truncated by abra- 
sion at a lower level while the ocean was depressed about 40 fathoms during 
the Glacial period. So long enduring a stability seems improbable enough 
even for the central Pacific, and much more improbable for a next-door 
neighbor of the Philippines. 
