DEVELOPMENT OF THE REEF 241 



below the bathymetrical limits of the reef-building corals 

 (Stanley Gardiner, op. cit. p. 175). 



Mr. Stanley Gardiner has said that, in determining the 

 cause of the 30 -fathom limit, "the real factor is the light, 

 which to reach the commensal algse has to penetrate the tissues 

 of the polyps as well as the water" (op. cit. p. 177). Now, 

 without wishing to controvert the assertion that corals " feed 

 mainly by their commensal algse," I would point out the fact 

 that light is certainly not essential for their feeding. It has 

 been my invariable experience with the corals in this atoll 

 that they are actually absolutely nocturnal in their activities. 

 It must be noticed by every one who visits a reef that, during 

 the hours of daylight, the zooids of the corals are not ex- 

 panded ; they are in a passive phase, from which a further 

 stage of retraction may result on stimulus — but they are 

 practically never expanded. It is most instructive when 

 examining a reef to see great soft masses of Alcyonaria with 

 myriads of zooids expanded and waving in the current, and 

 alongside it a rock colony of Pontes in which not a single 

 member shows activity. 



It is also one of the most striking things to examine the 

 reef corals by night, and see the extenb of alteration their 

 appearance undergoes, when the full expansion of their zooids 

 occurs. Mr. J. E. Duerden observes : * " As a general rule the 

 polyps are not expanded to their full degree during the day, 

 either on the reef or in the laboratory ; but the process begins 

 immediately after sunset, and full expansion is maintained for 

 the greater part of the night. In the morning the polyps are 

 again found retracted." These observations were made in 

 Jamaica, and so the habit is not one that is due to locality, and 

 I think we may therefore regard it as the fact that the reef- 

 building corals are active and that they feed in the dark. I 

 agree with Mr. Duerden in the observation that artificial light 

 appears to have no effect on zooids already expanded in the 

 dark. Again, the corals are certainly not entirely dependent 

 on their commensal algse for their food-supply, for any one may 



* Memoirs National Acad. Sci., vol. viii., 7th Memoir, p. 419 (1902). 



