1907.] 



AXD SUPPOSED SPECIES IX CORALS. 



531 



this method of growth ; and though the fully developed colony is 

 so totally different to the calm-water form, still the process is 

 easily seen in the making by means of artificially inflicted injury. 

 Besides the alteration of the general appearance of the vegetative 

 growth, the rough-water environment causes other changes in the 

 coral ; for it is diagnostic of a rough-water coral that its structure 

 is compact and dense, and its corallites tend to be flush with the 

 general surface of the growth. The question of the raising of 

 the coi-allites will have to be discvissed again Avhen the action of 

 sedimentation is gone into ; it must be stated here, that though 

 the levelling of the corallites in I'ough-water types is doubtless 

 partly mechanical, it is also due to the fact that in rough water 

 sediment does not tend to be deposited. 



Text-fig. 154. 



Rough-water type of highly branched Madrepore ; from an actual specimen 

 of Madrej>ora pulchra taken from the barrier. 



In marked contrast to the I'ough- water types are those forms of 

 a species that happen to have become fixed in an environment 

 where the water is more or less calm ; and here, as every grade of 

 environment is to be found, eveiy grade of modification of the 

 colony is represented. 



Corals grow in great luxuriance on the wave -stirred outer 

 slopes of the atoll, but this is a site by no means to be confounded 

 with the surf-line of tlie barrier. They grov/ also in the numerous 

 pools of the barrier flats, in the inlets to the lagoon, and in the 



