in the Pacific, ^c. 105 



have been filled up by fragments broken from the outer mar- 

 gin of the plateau sind carried landward by the surf, together 

 with the drift and wash from the beach of loose coral sand 

 and shingle. Others we may suppose have been obliterated 

 by the natural increase of the coral, till, from both causes, 

 only one remains here and there. Still there are sufficient to 

 show us that the formation of these channels cannot with 

 propriety be attributed to causes that would operate but in a 

 certain direction, such as we must regard the tide, which can 

 scarcely be supposed to seek an egress against the continual 

 rushing in of a powerful surf. 



Much light would no doubt be cast upon this and many 

 other peculiarities in the coral formations we have been con- 

 sidering, by an examination and comparison of the phenomena 

 presented by those of the Antilles, the Caribbean Sea, Euid 

 along our southern coasts. In fact, I believe that without 

 such comparison, it is impossible for a person to arrive at a 

 full knowledge of the structure of these rocks. Those es- 

 pecially, bordering the South coasts of Cuba and Florida, 

 deserve particular attention, on account of the numerous ca- 

 nals intersecting the extensive reefs, the varied nature of their 

 lagoons, and the opposite character of their neighboring lands, 

 which, broken and mountainous in the one case, are in the 

 other low, sandy, and level. I trust that at some future peri- 

 od it will be in my power to examine these interesting forma- 

 tions, with more time and opportunity for their study at com- 

 mand, than were enjoyed in my brief and restricted visit to 

 those in the great coral region of the Pacific. 



(To be coutinued.) 



14 



