— S — 



HO ntfcntion to ilio Cocos 1ap;oon, allho'' it was the only one of 

 llie class, wliicli he had an opportunity ot' obscrving, bccause 

 masses of growiiig coral are to be seeii in it, larger and harder, 

 than any that he could liave seen along the exterior reef, ex- 

 eept tlie reef itself , or, as for distin ctiou it may be named, the 

 breaker wall, and as to that wall, it is mainly composed of 

 small coral and shells firraly cemeuted together; the exterior 

 on which the billows fall into breakers, being covered with a 

 slielly glaze (much like that on coarse earthenware) and wliich 

 1'orms its grand security against the demolishing actidn of tbe 

 breakers. At tlie Cocos he could fartlier have observed, that along 

 the exterior of the islets, frora W. S. W. by the Southward to 

 E. N. E-, a considerable extent of the upper platform on which 

 the soil of tlie isles is laid, has been underinioed (not overflo- 

 wed but undermined) and as broken up, by the action of the 

 surge from the breakers, and the debris, partly thrown up along 

 and upon the outer margins of the islets and partly transporled 

 hy the surge and tide into the lagoon. The denuded space, 

 n<jw intcrvcning betveen the foot of the beacli and the brea- 

 ker wall, is generally below ebb tide level, and on many places 

 lias several feet depth of water, lying on a sandy bottora, so pre- 

 senting the same sort of appearances, that are elsewhere obser- 

 vable between the sliores and fringing reefs, subjected to the 

 action of breakers. We nced not to be informed, that these 

 sliallow spaces present very different aspects and quality of 

 bottoms, from those of the submarine valleys orravines, between 

 Ihe submarine prolongations of supermarine ridges, or ranges 

 forming the shorcs, or between the exterior walls of submarine 

 craters and the shores of the upheaved central islands. 



VIT. vH.iving thus specified the several kinds of reefs, which differ in 

 »lheir forins and relalive positiori, vvilh regard to the neiglibouring land , 

 » bilt which are siinilar in all olher respect, it will I think , be allowed , 

 5) ihat tio cxplrinatjon can be satisfactory which does not include the whole 

 »series. The theory which I woidfl oif'er , is sirnply ihis, that as the land 

 vwilh the allached reeis subsides very gradually Irom ihe action of sub- 

 » lerraueaii uauses , the coral building polypi soon laise again their solid 

 » indibcs to iLc kvel of the water, but, uol SB wilii llie liud , eacU iucli 



