PACIFIC AND BEERINGS STRAIT. 257 



also Henderson's Island, afford good examples of c ",ti P 



VIII. 



Feb. 



what I have been describing. To enable the reader 

 more readily to comprehend the nature of these ^26. 

 singular formations, I subjoin a sketch and a section 

 of a coral Island, with the slope of the sides of seve- 

 ral of them* laid down according to the soundings 

 and the depths at which attempts were made to 

 reach the bottom. 



All these islands are situated within the trade- 

 wind, with the exception of Oeno, which is only on 

 the verge of it, and follow one general rule in hav- 

 ing their windward sides higher and more perfect 

 than the others, and not unfrequently well wooded, 

 while the opposite ones are only half-drowned reefs, 

 or are wholly under water. At Gambier and Ma- 

 tilda Islands this inequality was very conspicuous, 

 the weather side of both being wooded, and of the 

 former, inhabited, while the other sides were from 

 twenty to thirty feet under water, where they 

 might be perceived equally narrow, and well de- 

 fined. It is on the leeward side also that the en- 

 trances into the lagoons generally occur, though 

 they are sometimes situated in a side that runs in 

 the direction of the wind, as at Bow Island ; but I 

 do not know of any one being to windward. The 

 fact, if it be found to be general with regard to 

 other coral islands, is curious, and is not fully ac- 

 counted for by the continued operation of the trade- 

 wind upon its side, as the coincidence would sug- 

 gest. -After the reef has arrived at the surface of 

 the sea, it is easy to conceive what would be the 

 effect of the trade- wind ; but it does not seem pos- 

 sible that its influence could be felt so far under 

 water as some of the reefs are situated. 



vol. i. s 



