DESTRUCTION OF REEFS ON STATIONARY COASTS 541 



vancing headlands and outlying islands alternating with reentrant embay- 

 ments branching into many coves. The headlands are soon swept clear 

 of detritus, and thereupon they become bordered by outgrowing fringing 

 reefs and are thus protected from further wave attack. The bayheads 

 on the contrary are invaded by delta deposits and, if submergence cease, 

 any fringing reefs at first formed in the bays are in time overwhelmed. 

 The longer such a coast stands still, and the higher and larger and rainier 

 the drainage area of its rivers, the farther forward will the deltas ad- 

 vance ; the detritus may indeed be eventually drifted so abundantly along 

 the prograding shore as to smother the headland reefs: thereupon the 

 waves will cut away the dead coral rock, and the delta fronts also if the 

 rivers are not too strong, and in time the waves will attack the headlands 

 again. It thus seems eminently possible that the coast of a considerable 

 land area that v^as bordered with reefs for a time after submergence took 

 place, will later become reef -free and so remain as long as the land stands 

 still and the shoreline suffers uninterrupted abrasive retrogression. 



It is manifest, however, that this untoward result may be prevented if 

 the submergence of the coast be intermittently continued at frequent in- 

 tervals; for if the bays are lengthened and deepened and the headlands 

 are shortened by renewed submergence before detritus overwhelms the 

 previously established reefs, the reefs will grow upward and more or less 

 outward, and will thus be transformed into discontinuous barrier reefs; 

 at the same time new fringing reefs will grow on the shortened headlands, 

 and the intermediate lagoon will be aggraded with deposits inwashed 

 from the outer reef and outwashed from the land, as well as by organic 

 deposits formed in the lagoon itself. It also appears probable that small 

 islands, once partly submerged, will retain their reefs indefinitely in a 

 succeeding stationary period, because the detritus that they shed will not 

 suffice to fill the lagoon and smother the corals on the reef face. This 

 scheme represents, I believe, the ordinary condition of reef growth, for 

 most reefs occur on mountainous coasts of long-continued submergence, 

 as has been shown in the preceding chapter; in other words, reefs are 

 ordinarily formed as Darwin explained. In case renewed submergence 

 should be more rapid than reef upgrowth, the previously formed reefs 

 will be drowned and fringing reefs of a new generation will be formed 

 along the new shoreline; and reefs of this kind are also, as has already 

 been shown, explained by a special phase of the theory of subsidence 

 which Darwin explicitly announced. 



If our observational acquaintance with the coasts of the world covered 

 several geological periods, we might reasonably expect to find actual ex- 

 amples of reef-fronted and reef-free coasts in all stages of developmenf 



