242 THE INFLUENCE OF INANIMATE SUBEOUNDINQS. 



the canal as it now exists could never have served for it to pass, 

 as in the deepest spots the bottom is only a few feet below high- 

 water mark. Thus everything points to a very recent upheaval, 

 which however can certainly not have been merely local, but 

 must have acted on the whole group of islets which are now 

 known under the name of the largest, Kriangle. 



This may be admitted without giving up the opinion that 

 the whole atoll was formed during subsidence ; for in order to 

 maintain this view it is merely necessary to assume that the 

 elevation thus proved only began at a very recent period, per- 

 haps at the very time when the Spanish ship cut the channel 

 in the eastern reef. In opposition to this only one obj ction 

 can be raised, but it is one of considerable importance. The 

 great difference above pointed out between the external slope 

 of the eastern and western reefs can hardly be reconciled with 

 the hypothesis of an equable subsidence throughout the atoll ; 

 for, if such had occurred, -the slope would be equally steep on 

 both sides. In point of fact it is steep only on the west, and 

 on the east very gentle, although this is the windward side. 

 This difficulty, again, might be removed by an arbitrary as- 

 sumption that the eastern side of the atoll might have re- 

 mained stationary while only the western side was sinking. 

 This hypothesis also would be easy to upset ; but I will post- 

 pone the discussion of it to a more favourable opportunityi 



III. The bank of Kossol. — The small reef known by this 

 name lies to the south of Kidangle ; its shape is a well-defined 

 horseshoe ; the channel that divides them is, according to the 

 maps, from 50 to 60 fathoms deep, and, so far as I have been 

 able to detect, entirely free from corals. The channel between 

 Kossol and the noi-them point of the large island of Babel- 

 thuap is, on the contrary, very shallow, at least in comparison 

 with the southern one ; on the maps, it is true, a considerable 

 depth is marked, but I must positively contradict the accuracy 

 of these indications ; the water in it is everywhere of a paleljlue 

 colour, while in the channel between Kossol and Kriangle it is 

 quite dark, almost blue-black. And while in this channel coral- 

 blocks are nowhere to be found standing up from the bottom, 

 thev occur in considerable numbers in the channel south of 



