244 THE INFLUENCE OF INANIMATE SUEEOTJN DINGS. 



Kriangle and Kossol. Now, if the two channels had been 

 formed by subsidence, as is required by Darwin's theory, it 

 would be quite incomprehensible why, in the channel south of 

 Kossol only, corals should have grown in blocks of various 

 sizes, from the bottom nearly up to the surface, while in the 

 northern channel they are totally absent. For though reef 

 corals certainly do not seem usually to build up from a depth 

 of sixty fathoms, since, on the hypothesis of subsidence, the 

 southern channel must at some time have been high and dry, 

 even the bottom of the north channel must have lain high 

 enough to allow of the establishment of reef-corals. But then 

 it is not clear why they should not have continued to grow up 

 to the present time, and hence, on the theory of subsidence, we 

 might expect to find exactly the opposite of what in fact has 

 happened. But all these diiEculties vanish as soon as we sup- 

 pose that both these channels, and of course the reef also, have 

 been formed during a period of upheaval. In that case the 

 Kriangle channel, which was originally the deeper, would not 

 yet be raised high enough to allow of the establishment on any 

 large scale of reef corals, while the bottom of the channel be- 

 tween Kossol and Babelthuap much sooner attained the re- 

 quisite level ; and it is ia accordance with this idea that we 

 find in this channel numerous blocks of various sizes. To this 

 point I must return in a later section. 



IV. The northern reef of Babelthuap. — This is a. true 

 barrier-reef running round the island of Babelthuap, which 

 rises to a tolerably conspicuous height, and is almost entirely 

 composed of quite recent eruptive rock. The northern portion 

 of the island is extremely narrow, and beyond the northern 

 point lie three or four islets in the channel within the reef. 

 According to the maps hitherto published this would be in- 

 correct ; even in the latest by Friedrichsen, founded on data 

 supplied by a naturalist named Kubary, its breadth at the line 

 passing through Aibukit (see Map I.) is made at least three 

 times as gi-eat as in mine, constructed on triangular measure- 

 ments. I found that at this point the width from reef to reef — 

 from east to west — was at most 5 to 6 geographical miles, while 

 Friedrichsen gives it as about 20. The island enclosed by the 



