176 Davis — Dana's Confirmation of 



on so good a foundation until the truncation is complete. The 

 ordinary relation of fringing and barrier reefs to their central 

 island suffices to show that the work of marine truncation 

 would be arrested by the growth of reefs, just as soon as the 

 abraded platform became broad enough to afford a foundation 

 for coral growth a moderate distance away from the outwash 

 of fresh water and its detritus ; and when such a reef is once 

 established, the further truncation of the island by wave work 

 is practically stopped ; the waves of most lagoons are too weak 

 to be effective agents in cutting away the land. Moreover the 

 Alexa bank described by Wharton is as fairly explained by 

 regarding it as a slightly submerged atoll as by taking it for a 

 truncated volcanic island. Hence the theory of truncation, 

 though easily conceivable as to its inorganic elements and 

 received with approval by several writers, involves organic 

 elements which are not admissible. It must be rejected as a 

 means of explaining atolls. 



The formation of veneering barrier reefs on the outer mar- 

 gin of submarine platforms cut by the sea around still-standing 

 islands, an old idea* recently given prominence by Agassiz, 

 seems entirely possible, provided that the coral larvae can be 

 prevented from establishing themselves on the platform soon 

 after its cutting is begun ; but the explanation of barrier reefs 

 outside of broad lagoons in this way would involve the same 

 difficulty that is fatal to Wharton's theory. However, if bar- 

 rier reefs have sometimes been formed in this manner, certain 

 special features should be found in close and constant association 

 with them on the still-standing central island, i^ or after a plat- 

 form a mile or so in width has been worn by the waves, the 

 central island should rise from the cut-back shore line in a wall 

 of steep cliffs, as Darwin clearly stated, f and the general out- 

 line of the cliff-wall should be simplified as compared with the 

 original outline; for marine erosion, after a brief introductory 

 period of increasing irregularity of shore line, tends to decrease 

 the sinuosities of the initial form. If any initial bays existed, 

 due to irregular island-building by volcanic eruption, they 

 should fill with delta plains while the intervening headlands 

 are cut back. The wider the platform becomes, the simpler 

 should be the island outline and the more continuous the cliff 

 wall; but the cliffs should be interrupted here and there by 

 valleys, as long as the area of the island is large enough to 

 maintain streams. After the formation of a barrier reef, deltas 

 should be rapidly built forward from each of these valleys in 

 the quiet and shallow waters of the enclosed lagoon. 



Now it may be confidently asserted that the central islands 



*See footnote in Darwin's " Coral Reefs," 1842, 49. 

 f Coral Eeefs, 1842, 49. 



