SHALLOW-WATER FAUNA OF THE TROPICS. 71 



sand which fills up the interstices between the 

 living Corals, or gets washed over into the 

 lagoon, or falls as a talus over the sea-ward slope 

 of the reefs. In the latter case lumps of Coral 

 torn off from the reef become embedded in it and 

 form with it a bank which gives support to more 

 living Polyps on the sea-ward slope. Conse- 

 quently, in the course of many years, a reef 

 which was at one time only fifty yards from the 

 beach may have extended to a distance of a 

 hundred yards or more, growing, as it were, on 

 the skeletons or shells of the Corals that have 

 died. There can be little doubt that Coral-reefs 

 do grow sea- wards in this manner in some places, 

 but they may also be either beaten back or kept 

 for a long time perfectly stationary if the tides 

 are too strong or too slack. 



What the precise conditions are which favour 

 the growth sea-wards of Coral-reefs has not yet 

 been systematically investigated ; but we may 

 suppose that if the tides are too strong the 

 sand has no opportunity of settling between the 

 Coral blocks and forming a solid limestone rock, 

 and if they are too slow the Coral-polyps do 

 not get sufficient nourishment to allow them 

 to build fast enough to counteract the solvent 

 action of the water. 



An interesting point connected with the Coral- 

 reefs is the manner in which they are formed at 

 first. A volcanic upheaval gives rise to a new 

 island which, in the course of time, is surrounded 

 by a Fringing reef. How does this reef begin ? 



The answer to this question has been recently 

 given by the observation of the formation of new 



