THE AFTER HISTORY OF THE ATOLI. 265 



barrier in thousands, and its attention is wholly confined to the 

 coral rocks of the atoJl, for no other feeding- ground exists 

 for over five hundred miles. I have shot large numbers 

 of Kakatua Ijou, and have examined their intestinal con- 



FiG. 63. 



Photograph ov a Boulder of Alga-coveeed Dead Coral 



Rock, 



To show the bites of the fish of the genus Scarus. The black line 

 marks the edge of the alga covering not bitten away by the fish. 



tents. Every fish that I have killed has been filled with coral 

 sand of its own producing : the average weight of washed 

 and dried sand found in them is two ounces. Considering 

 the activity of these fish and their vast numbers, the addition 

 that they make to the sand-supply of the atoll must be very 

 considerable. Both Darwin and Forbes have described these 

 Scari as eating the living coral, but there is no doubt that 

 the observation is incorrect, for they never by any chance 

 touch a living colony. 



The Holothurians are also a source of some sand-pro- 

 duction, but I do not think that their contribution is a very 



