278 
GEOLOGY: W. M. DAVIS 
had been introduced by warping to a reef -free and retrograded coast in an old 
stage of the cycle is a most natural consequence of a long stationary period in 
the history of an island in the coral seas. 
Another warping is then inferred, chiefly because the change from the sub- 
continental land of block 1 to a narrow island adjoined by deep seas on both 
sides, is not likely to have been accomplished in a single period of deformation. 
This warping must be supposed to have affected the island unsymmetrically, 
as in block 4, probably drowning any previously formed barrier-reefs along the 
southwestern coast, and re-embaying the shore line there, where a new barrier 
reef, G, would be developed, but this last point is not essential; at the same time 
the northeastern coast appears to have been uplifted, so that a coastal plain 
of marine sediments, such as may have accumulated in the shallow sea, F, was 
there added, as at H. The reason for the last inference is that the elevated 
peneplain areas along the northeastern side of the island were cut back in 
cliffs by the sea in the early stage of the cycle introduced by this elevation; 
and the simplest way of accounting for this is to suppose that the elevation here 
laid bare a narrow coastal plain, covered with loose sediments, on the shore line 
of which reef-building corals could not establish themselves, and on which the 
waves could therefore work unimpeded. No other supposition can so reason- 
ably account for the abrasion of cliffs along one side of an island in the coral 
seas during the early stages of a subrecent cycle of erosion. 
Block 4 is then gradually transformed into block 5, in which the weak-rock 
areas of the southwestern coast are again worn down to moderate relief, and 
the reef -enclosed lagoon is largely filled with delta plains, as at /; and in which 
the uplifted peneplains of stronger rocks along the northeastern coast are dis- 
sected by narrow valleys and cut back in high cliffs, as at K. A recent sub- 
mergence has converted block 5 into block 6, drowning the previously developed 
delta plains of the southwestern coast, where the reef has grown higher and the 
sea has advanced farther than before on the lowland border, thus leaving the 
broad lagoon, L, of today between the young shore line of submergence and 
the barrier reef; the same recent submergence has partly drowned the cliffs 
of the northeastern coast, so that their valleys are now beautifully embayed, 
and a barrier reef has grown up from the sea bottom in front of them, as at M. 
It is chiefly upon the highland peneplains back of the cliffs of this coast, and 
upon similar highland areas which occur along the northwestern half of the 
other coast, that metaliferous laterites occur. 
Abundant variations on the earlier stages of the foregoing scheme may be 
proposed. The changes here outlined are probably much simpler than the 
changes that have actually taken place, and some of the changes here indi- 
cated are very uncertain. For example the cutting back of the embayed coast, 
D, in block 2, to the low cliffs, F, of block 3, is by no means assured; but the 
unsymmetrical warping by which block 3 was transformed into block 4 seems 
to be essential as a means of reasonably providing for the development of sea 
cliffs in a relatively early stage of the cycle of erosion on one side of the island 
