562 



SCIENCE 



[N. S. Vol. LIII. No. 1382 



"lofty cliffs"; yet the very fact that most other 

 barrier reef islands are not "bounded by lofty 

 cliffs" but "slope down to the lagoon channel" 

 shows that they have not been "deeply eaten 

 into"; or df they have been then the resulting 

 cliffs have been completely submerged by later 

 subsidence. 



His general scheme of upgrowing reefs on 

 subsiding foundations therefore takes in with- 

 out any difficulty the special case of an island 

 around which a platform had been abraded. 



Good reasons may be given for believing 

 that the peculiar case of completely submerged 

 platform-back cliffs, just alluded to, is a very 

 probable one; for wave-cut platforms and 

 cliffs presumably occur as normal features in 

 an early, pre-reef stage of young volcanic 

 islands; and their rarity to-day is best ex- 

 plained by the strong subsidence of the islands 

 since the platforms were cut; but the discus- 

 sion of this question would lead away from 

 the matter here under consideration. 



Another passage from Darwin's book, di- 

 rectly following the one above quoted about 

 the possibility of reefs growing on the coast 

 of an island that has been deeply eaten into 

 by the sea, is pertinent here, as it explicitly 

 considers the growth of a reef upon a plat- 

 form margin and the depth of the resulting 

 lagoon : 



On this view,3 moreover, the cause of a reef 

 springing up at such a great distance from the 

 [non-subsiding] land, leaving a deep and Ibroad 

 moat within, remains altogether unexplained. 



Or otherwise phrased: If a reef sprang 

 up from the outer margin of a broad platform, 

 cut by waves around a still-standing island, 

 the enclosed lagoon could not be so broad 

 and deep as barrier-reef lagoons usually 

 are, unless subsidence had occurred along with 

 reef growth. The quoted statement is not so 

 clear as Darwin's writing generally is, but the 

 modified phrasing here suggested is believed 

 to represent his fuller meaning; it is certainly 



' 3 A footnote in Darwin 's book at this point 

 reads: "The Rev. D. Tyerman and Mr. Bennett 

 . . . have briefly suggested this explanation of 

 the origin of the encircling reefs of the Society 

 islands. ' ' 



consistent with the context In any case, Dar- 

 win clearly knew that a platform could be 

 abraded around a volcanic island and that 

 such a platform must be backed by cliffs; 

 and he further believed that, if a reef grew 

 up on the margin of the platform, the lagoon 

 thus enclosed would not have the depth of 

 most barrier-reef lagoons; but that if the 

 abraded island subsided and the reef grew 

 higher, the depth that is usually found in 

 barrier-reef lagoons would thereupon be pro- 

 duced. According to the present understand- 

 ing of the coral-reef problem, it is precisely 

 the occurrence of such subsidence that puts a 

 stop to further abrasion by making reef -growth 

 on a platform margin possible; but Darwin 

 did not detect this point, nor did he see that 

 the opportunity for abrasion of platforms 

 around volcanic islands in the coral seas is best 

 provided, as above mentioned, when the islands 

 are young and high, with simple, non-embayed 

 margins, so that a large amount of detritus 

 shall be washed down from their steep slopes 

 to the shore, where its accumulation in beaches 

 inhibits coral growth and permits abrasion. 

 Indeed, this explanation of the condition un- 

 der which the abrasion of a platform may 

 occur is not mentioned even in Chamberlin's 

 summary, though its omission there may be 

 due rather to the conciseness of the summary 

 than to a rejection of the explanation. The 

 explanation has, however, a considerable theo- 

 retical importance in giving reasonable con- 

 sideration to an early pre-reef stage of island 

 development that has been generally over- 

 looked ; * and it was in view of this explana- 

 tion that the common occurrence of completely 

 submerged platform-back cliffs was above sug- 

 gested as probable in barrier-reef islands; but 

 the platforms associated with these submerged 

 cliffs need not have been nearly so broad as 

 the submerged platform of Tutuila. 



It may be added that the opportunity for 

 platform and cliff cutting on Tutuila can not 

 be advisedly ascribed to the inhibition of 

 coral growth by the lowered temperature of 

 the lowered Glacial ocean, as is postulated 



4 ' ' Clif t Islands in the Coral Seas, ' ' Proc. Nat. 

 Acad Sci., II., 1916, 283-288. 



