GEOLOGY: W. M. DAVIS 
473 
ing Isles here seems to have results from the repetition of one sector in 
six successive positions: the reef is really of irregular pattern and the 
islands are not centrally placed within it. A consideration of all the 
pertinent facts shows that, if the strong changes of level demanded by 
Vanua Mbalavu and Avea have extended to the exterior atolls, as is 
eminently probable, then the true atoll and the almost-atoUs have 
been formed, during a recent subsidence of a few hundred feet, by up- 
growth on the worn-down remnants of uplifted atolls, which had like- 
wise been formed by upgrowth, presumably on a volcanic foundation, 
during an earlier time of greater subsidence. 
The changes of level demanded by Vanua Mbalavu and Avea may 
be inferred from the structure of Avea, two miles long and 600 feet 
high, as shown in part in figure 2. It should be announced that this 
figure has been drawn from a rough outline made from a passing 
steamer; confidence is nevertheless felt in its essential correctness, for 
I fully agree with Gardiner and Agassiz that volcanic slopes and lime- 
stone cliffs in the Fijis are easily recognized at a distance. The under- 
FIG. 2. 
lying volcanic rocks here, as in Vanua Mbalavu, have the forms of 
rounded hills, subdued by deep erosion before they were covered with 
limestones. The eroded slopes dip under water, as at x\ hence the 
island formerly stood at least x feet higher than now. The subdued 
hills and slopes are partly covered by limestones, now rising to a height 
of 600 feet; the limestones have suffered much erosion, which has 
greatly diminished their area, and has somewhat reduced their height, 
as at y. Hence the island after being eroded, must have sunk x -f 600 
+ y feet while receiving its unconformable limestone cover. The lime- 
stones are now dissected and their shore line is so well embayed as to 
necessitate a recent submergence of more than 100 feet; and it must have 
been during this submergence that the present great barrier reef has 
grown up. It may be noted in passing that the postglacial rise of 
ocean level may have had to do with the recent submergence, but for 
reasons elsewhere stated I believe that true subsidence has also been 
concerned in it. 
The six sectors, G to M, of figure 1 graphically represent the succes- 
