26-L W. M. Davis — Shaler Memorial Study of Coral Reefs. 



This conclusion becomes all the more reasonable when the 

 intimate association of barrier reefs and atolls in the Fiji, New 

 Hebrides, and Society groups and elsewhere is noted. The 

 conclusion may seem unreasonable to some students of the 

 problem in the case of large groups of atolls which include no 

 barrier reefs ; but, as far as I have been able to discover the 

 ground of that seeming unreasonableness, it rests chiefly on a 

 hypothetical objection to the occurrence of subsidence in 

 oceanic areas — an objection which I believe to be based more 

 on our ignorance than on our knowledge of the earth's struc- 

 ture and behavior. The objection is greatly weakened when it 

 is seen that the submergence of barrier-reef groups by regional 

 subsidence is highly probable, and when it is recognized that 

 if their submergence is due to a rise of ocean level, the uplift 

 of the ocean bottom thus demanded in some other region is 

 much greater than the sinking of the ocean bottom in the coral- 

 reef region demanded by the theory of subsidence. 



It is instructive to note that two different schools of geology 

 to-day teach, explicitly or implicitly, directly opposite theories 

 on this recondite problem. One school insists that highlands, 

 such as those of eastern Australia or of northern Arizona, owe 

 their relief, not to radial uplift of their own areas, but to the 

 radial subsidence of the neighboring areas ; yet, inasmuch as 

 both these highlands include peneplains of normal erosion, they 

 must once have been lowlands continuous by gentle grades 

 with neighboring low areas and with the sea-surface ; and if 

 the neighboring low areas have since then subsided without 

 being submerged, then all the oceans and all the ocean bottoms 

 — and all the continental areas that did not then gain an 

 increased altitude above sea level by retaining their former 

 distance from the earth's center — must have sunk at the same 

 time and by essentially the same amount ; thus an implicit 

 consequence of this theory — the theory of still-standing high- 

 lands, as it may be called — is that the radial sinking of the 

 ocean bottom must be the sum of all the continental subsi- 

 dences of different dates, whereby peneplains have been given 

 the appearance of uplifted plateaus. Another school teaches 

 that the ocean bottoms are relatively fixed, or at least very 

 slow to decrease their distance from the earth's center, and 

 explains local high-standing peneplains by local uplift, without 

 notable disturbance of the rest of the world ; this might be 

 called the theory of local responsibility. 



The theory of still-standing highlands evidently demands a 

 rate and an amount of terrestrial shrinking far in excess of 

 that usually ascribed to the earth ; and if it be true, the earth 

 can hardly yet have reached the advanced stage of evolution 

 usually attributed to it, when, as Leibnitz long ago said, "con- 



