22 THE BERMUDA ISLANDS. 



even then the solidity and compactness of the growth would 

 render the process of undermining and disruption a slow one. 

 This is also true of the reefs on the south side of the island, the 

 crests of which are serpuloid, while the lee-slopes are fairly 

 covered with large Mseandrinas. We but rarely came across a 

 loose block of stone on the beach, and where such was found 

 it could be generally, if not always, identified as the disrupted 

 part of the cliffs upon which the fury of the surf was expended. 

 The examination of the lime-sand of the inner waters only 

 exceptionally showed recognizable coral fragments, although it 

 was very largely composed of the debris of the more friable 

 millepore. Indeed, it might be said that the sand is properly 

 a shell and millepore sand, rather than one of coral formation, 

 and this is also true in a measure of the cliff-sand of the main 

 body of the islands. 



We passed the better part of three hours on the reef, but by 

 the end of this time the water was gaining upon the spot 

 rapidly. In a few short hours the reef would again be entirely 

 covered, save the three gray pinnacles which constitute the lone 

 North Rock. 



