i88 CORALS A.YD CORAL LSLANDS. 



century, it is stated, among other notes upon their natural 

 history, that flocks of crows, no doubt the same species {Corvus 

 Amcricaniis) which now inhabits them, were in the habit every 

 evening of winging their flight from the main island toward 

 the north. This observation, which from its simplicity I should 

 the more readily believe to be a true statement, would clearly 

 prove the existence of land in that direction at no great dis- 

 tance ; for the habit of this bird to leave its roosting-place for 

 distant feeding-grounds during the day, to return at random, 

 is one of its well-known characteristics. 



" Taking these matters into consideration, I see everything 

 to support the supposition that the Burmudas once presented 

 a much more extensive aspect than they do at present ; and 

 certain additional evidences which I hope to bring forward 

 shortly in a collected form will, I conceive, tend to confirm 

 my impression that the restricted ten-aqueous area lying within 

 the limits of the outer barrier reef is merely the summit of one 

 of a range of islands which extended in somewhat semicircular 

 form for a distance of seventy or eighty miles, and which have 

 suffered submergence to a dep-th only to be correctly ascer- 

 tained by borings, which might be successfully accomplished 

 under the auspices of the Government at a trifling expense." 



