1907.] AND SUPPOSED SPECIES IN CORALS. 553 



oflfensive when the growth is exposed to the air, and this strange 

 odour is no sign of death, foi- a stinking coral when replaced in 

 water or when re-covered by the rising tide, flourishes again. At 

 such a low tide the bai-rier-flats present a picture of bushes and 

 boulders of living coral all freely exposed to the sun, all dry, all 

 smelling very offensively, and yet the returning tide finds them 

 all living as actively as when it left them. A coral may be taken 

 from the bottom of the lagoon, may spend the best part of a hot 

 day, high and dry, in the bottom of a boat, and yet, when it 

 is replaced in the water, all its zooids will expand, and it will 

 resume all its vigour. Exposure to the sun and air between tide 

 limits plays but little part as a causative factor in the question of 

 coral death. Of course no coi-al could grow beyond the normal 

 high-tide level, but the remarkably level appearance of the barrier- 

 flats is not the result of the action of sun and air on the coral 

 colonies, so much as of the levelling effects of the weaves, and the 

 moving fragments that they wash to and fro. It is the grinding 

 action of the surface waters at their level of maximum activity 

 that determines the limiting level of upward coral-growth, far 

 more than the death of the apices from the effects of sun and air. 

 The waves that sweep over the flats, and carry shorewards the 

 fragments that they have broken from the seawai'd margin of the 

 barrier, are for ever keeping the coral colonies within the limit 

 of upward growth ; but their action is not altogether detrimental 

 to the corals, for though, where island beaches are formed, many 

 fragments of living coral are cast ashoie only to perish, still many 

 more, where no such dry land exists, are safely lodged in a new 

 resting place. Broken fragments are swept across the flats, they 

 lodge in pools, they become sti-anded under the lee of boulders, or 

 are washed into the lagoon ; and each of these fragments, if not 

 too badly damaged, will form the nucleus of a new colony in a, 

 suitable habitat. If a large colony of a Madrepore be broken up 

 in the rock-pool where it flourishes, the great majoi-ity of its 

 fragments will continue to grow and branch out into new colonies ; 

 and if some of these fragments are swept onwards by the waves, 

 they will form pioneers for the species when lodged in a suitable 

 environment. 



Freshening of the water from the excessive tropical rains has 

 been said to cause wdiolesale death among the lagoon corals ; and 

 in high islands were rivei's flow into the sea, the fresh water is 

 well known to be a great cause of the absence of coral-growth. 

 Before Darwin's visit to the atoll it is said tliat an abnormal 

 rainfall killed many corals, and again in 1866 the fresh water is 

 said by the Governoi- to have stood for a height of seveiul inches 

 on the surface of the lagoon, so heavy and continuous was the 

 rain. Again, in May 1896, the rains were abnormal, and the 

 freshening of the water destroyed the lagoon algas and fish ; and 

 this in such quantities that when Mr. Arthur Keyser visited the 

 islands in July, the dead fish wei'e still being cleared from the 

 lagoon. There is no doubt that the rain would have to be long 



