230 THE INFLUENCE OF INANIMATE SURROUNDINGS. 



passage from atolls to barrier-reefs, and then to true fringing 

 reefs, till, at the southernmost end of the group, all reefs have 

 disappeared. 



The fact that all the varieties of reef are met with in combina- 

 tion in a district which, in accordance with Darwin's theory, 

 we were accustomed to regard as a district of subsidence, and 

 in which, consequently, only atolls and barrier-reefs ought to 

 occur, is fiankly admitted, in the second edition of Darwin's 

 \vell-known work, to be a very grave difficulty. We might 

 evade this difficulty, as Darwin does that is, by assuming 

 that even islands which are in process of subsidence, and which 

 therefore ought properly to be enclosed in true barrier-reefs, 

 might sometimes form true fringing reefs, particularly if the fall 

 of the coast were so steep that even during a period of slow sub- 

 sidence the reef remained close to the shore, and thus preserved 

 the form of a fringing reef. This assumption, however, stands 

 in sharp contradiction to the facts long since observed and 

 announced by myself, and into which I must presently go some- 

 what more closely. We may also endeavour to overcome this 

 difficulty in another way for instance, by supposing that within 

 the Pelew Group itself there exists an axis of motion independent 

 of the general change of level of the bottom of the Pacific Ocean. 

 This point might be sought perhaps somewhat to tho north of 

 Pelelew, or in that island itself, to the north of which the 

 whole group would be constantly sinking more and more 

 rapidly, while to the south it was rising in proportion. This 

 would, in fact, apparently account for the circumstance that 

 Angaur has hardly any reefs, that Pelelew exhibits fringing 

 reefs, and plainly though feebly developed barrier-reefs. It 

 would also explain why the reefs to the north of this fulcrum 

 become deeper and deeper in the deep sea, till at last, at the 

 extreme north, only atolls occur, or reefs of the atoll character. 

 Now I purposely avoid laying any stress on the fact that 

 such an assumption is in the highest degree improbable, as 

 that there should be such a fulcrum or point of rest localised in 

 so small and isolated a district as these islands constitute in 

 the Pacific Ocean, with upheaval proceeding to the south of it, 

 and subsidence to the north. But, even granting this as pos- 



