CORAL ISLANDS. 199 



of the Jardin des Plantes, the edge of the reef is produced by 

 Madrepora corymbosa, M. pocillifera, and two species of Astrea, 

 which pursue their operations at the depth of from eight to fifteen 

 fathoms. At the base is a bank of Scriatopora, from fifteen to twenty 

 fathoms in height. At the bottom, the sand is covered with Seriatopora. 

 At twenty fathoms we also meet with fragments of Madrepora. Be- 

 tween twenty and forty fathoms the bottom is sandy, and the sounding- 

 rod brings up great fragments of Caryophyllia. According to MM. 

 Quoy and Gaimard, the species oi Astrea, which, as these naturaUsts 

 consider, constitute the greater part of the reefs, cannot live beyond 

 four or five fathoms deep. Millepora alcicornis extends from the 

 surface to the depth of twelve fathoms ; the Madrepores and Seriato- 

 pores down to twenty fathoms. Considerable masses of Meandrina 

 have been observed at sixteen fathoms ; and a Caryophyllia has been 

 brought up from eighty fathoms in thirty-three degrees south latitude. 

 Among the polyps which do not form solid reefs, Mr. Darwin 

 mentions Gorgonia at i6o, Corallines at loo, Millepora at from thirty 

 to forty-five, Sertularians at forty, and Tubulipora at ninety-five 

 fathoms. 



According to Dana, none of the species of the genera which form 

 reefs — ndcmQly, Madrepora, Millepora, Porihs, Astrea, axid Meandrina — 

 can live at a greater depth than eighteen fathoms. It is only near 

 the surface of the water that the zoantharia which produce polypidoms 

 and form coral banks put forth their powers ; the points most exposed 

 to the beating of the waves is that which is most favourable to their 

 growth ; it is there that the finest specimens of the genera Astrea, 

 Forites, and Millepora most abound. 



The proportionate increase of the structures, according to Mr. 

 Darwin, depends at once upon the species which construct the reef-: 

 and upon various accessory circumstances. The ordinary rate of in- 

 crease of the madrepores, according to Dana, is about an inch and a 

 half annually ; and, as their branches are much scattered, this will not 

 exceed half an inch in thickness of the whole surface covered by the 

 madrepore. Again, in consequence of their porosity, this quantity 

 wU be reduced to three-eighths of an inch of compact matter. The 

 sands, too, filling up the destroyed part of the polyp are washed out 

 by the currents in the great depths where there are no living corals, 

 and the surface occupied by them is reduced to a sixth of the whole 

 coralline region, which reduces the preceding three-eighths to one- 

 sixth. The shells and other organic debris will probably represent a 

 fourth of the total produce in relation to corals. In this manner, 

 taking everything into account, the mean increase of a reef cannot 



