W. M. Davis — Shaler Memorial Study of Coral Beefs. 269 



as to the stability of the ocean bottom, which permit the inference 

 that the older oceanic volcanoes must have suffered subaerial 

 denudation long enough for them to approach or reach pene- 

 planation (309) — that is, that they stood still long enough to 

 be worn down almost to sea level. 



If the question of insular stability is approached rather from 

 the observational than from the theoretical side, and if the 

 interval of time under consideration is expressed not in terms 

 of standard geological chronology, such as " since the Tertiary," 

 but in terms of the period needed to erode the embayed valleys 

 of barrier-reef islands — this period being, according to Daly's 

 exposition of the glacial-control theory, less than post-Tertiary 

 time, inasmuch as the embayed valleys were, under the glacial- 

 control hypothesis, eroded only during epochs of glaciation — 

 a very different conclusion is reached. In the Fiji group 

 there are several elevated reefs, 600 or 800 feet above sea level, 

 as yet very little dissected ; hence they must have been exposed 

 to subaerial erosion for a much shorter period than that 

 required to erode the embayed valleys in volcanic rocks. Two 

 of the Loyalty islands are former atolls, now elevated about 

 300 feet, but their dissection is hardly begun. On Efate in 

 the New Hebrides, weak " soapstones " overlaid by fringing 

 reefs at various altitudes up to 800 feet show no change since 

 emergence except narrow valleys cut by streams of rapid fall. 

 In view of these large measures of uplift — 50, 100 or 130 

 fathoms — in very recent time, it seems unreasonable to allow 

 subsidence no more than " a fathom or two " in a longer 

 period. Hence if the ocean surface were lowered by a score of 

 fathoms or more during the last glacial epoch, it does not 

 seem altogether unreasonable to suppose that certain parts of 

 the ocean bottom, with the islands standing on it, should have 

 risen or sunk by some such amount in the same time. The 

 combination of subsidence with changes in the level of the 

 ocean surface therefore seems legitimate. 



Summary of Hesults.—'Yh.e general result of my voyage 

 already announced above, as well as several special results, 

 may now be concisely stated : — 



The origin of coral reefs cannot be determined by a study 

 of the visible features of sea-level reefs, for they can be 

 explained by any one of the eight or nine theories that have 

 been proposed to account for them, provided that the pos- 

 tulated conditions and processes of the theories are accepted. 

 Apart from research by deep borings, the true theory can be 

 detected only by the study of associated problems, such as the 

 form of the central islands within barrier reefs, or the struc- 

 ture of uplifted reefs. 



The origin of coral reefs, like many other geological problems, 

 involves the discussion of invisible structures and processes to 



