406 W. M. DAVIS 
points of the island are rather strongly clift, and depths of 15 or 20 
fathoms are shown close to the cliff base. Mayer has recently 
announced the occurrence of cliff-base benches 8 feet above present 
sea-level, and of a fringing reef that surmounts a submerged plat- 
form about 20 fathoms in depth; but he does not explicitly corre- 
late the origin of the valleys, now drowned in embayments, and the 
headland cliffs, some of which are 500 feet high.‘ The Marquesas 
Islands, farther east in the Pacific and much nearer the equator than 
the many atolls in the neighboring Paumotu group, have no reefs, 
though fragments of coral are found in the bay-head beaches; 
their headlands are strongly clift, and the depths of from 12 to 
20 fathoms are found a short distance offshore. Both of these 
examples strongly suggest that unprotected modern volcanic islands 
may have their spur ends clift by the ocean while valleys are eroded 
on their slopes; both suggest also that, as Dana suggested for the . 
Marquesas and as Darwin pointed out in more general form, rapid 
subsidence may have drowned former reefs; if so, the subsidence 
must have been of so recent a date that no new reefs have yet been 
built up. These two island groups deserve abundant soundings on 
their banks and critical physiographic study of their shore lines. 
If we assume for the moment that none of the islands here 
mentioned have subsided, several inferences may be drawn from the 
foregoing facts. ‘The first is that the central depths of the extra- 
tropical banks, generally 20 fathoms or less, are decidedly smaller 
than the central depths of the large intertropical banks, which 
usually measure 40 fathoms or more; in both cases the banks may 
now be more or less aggraded, but no changes of depth thus brought 
about permit us to regard the present dissimilar depths as accordant 
and hence as confirmatory of abrasion at the same level by the 
lowered glacial ocean. Another inference is that of the two groups 
of banks it may well be the extra-tropical that are best ascribed to 
abrasion by the lowered ocean, because a lowering of sea-level by 
20 or 25 fathoms, which would suffice to cut these banks, is in 
the opinion of some glacialists a more probable measure of the 
glacial lowering of the ocean than 4o or more fathoms, which is 
demanded for the cutting of the intertropical banks. 
t A. G. Mayer, “‘Coral Reefs of Tutuila .. .. ,”’ Proc. Nat. Acad. Sct., III (1917), 
522-26. 
