W. M. Davis — Shaler Memorial Study of Coral Reefs. 265 



sistentior emergeret status rerum." The theory of locally 

 uplifted highlands is, on the other hand, more accordant with 

 generally accepted views regarding the well-established stabil- 

 ity of the terrestrial mass. Hence of the two theories the 

 latter seems much the better grounded ; yet not so rigidly 

 grounded as to exclude the possibility of regional ocean-bottom 

 subsidence in the conversion of fringing reefs into barrier reefs 

 or of barrier reefs into atolls, for this is only another case of 

 local responsibility. However this may be, the important thing 

 to note here is that choice between the two theories must be 

 made, not by observation of the visible highlands of the earth's 

 surface, for all the features of the highlands are as well ex- 

 plained by one theory as by the other ; choice must be made by 

 the discussion of inferences regarding the possible behavior 

 of the earth's invisible interior ; and in this respect the discus- 

 sion of the theories of these two opposed schools of geology 

 resembles the discussion of the theories of coral reefs. 



Nevertheless, possibility of ocean-bottom subsidence is not 

 a necessity. In the case of atoll groups in which no barrier- 

 reef islands are found, it is easily conceivable that the reefs 

 may have been formed in other ways than by upgrowth from a 

 sinking foundation ; indeed, even if nine out of every ten atolls 

 were proved by means of abundant, large-core borings, to have 

 been built upwards during the subsidence of their foundations, 

 some other origin for the un bored atolls might still be con- 

 ceivable to the agile-minded inquirer ; but conceivability is not 

 necessarily credibility. Hence the problem of atolls cannot be 

 regarded as absolutely solved : nor can it ever be absolutely 

 solved until we make vast additions to our present knowledge. 

 Indeed, however far observation of any coral reef is carried, it 

 cannot go beyond the facts of existing structures ; while the 

 origin of reefs, whether barriers or atolls, involves "unobserva- 

 able processes of the invisible past as well as unobservable 

 conditions of the unattainable interior of the earth ; and the 

 absolute determination of these unobservable elements of the 

 problem is far beyond our reach. The origin of coral reefs 

 must therefore, like the displacement of faulted structures, the 

 uplift of high-standing peneplains, and the metamorphism of 

 sedimentary strata, always remain a matter of inference, how- 

 ever far the observational study of existing reefs is carried. 

 The most that can be done will be to make our inferences rea- 

 sonably safe ; but if in the meantime a highly probable solu- 

 tion for the general problem of coral reefs, including atolls as 

 well as barrier reefs, is wanted — that is, a solution of the kind 

 that is accepted as a " conclusion " in many other geological 

 problems — such a solution is, in my opinion, provided by Dar- 

 win's broad theory of upgrowing reefs on foundations that sub- 

 side with the subsiding ocean bottom, the submergence thus 



