1907.] AND SUPPOSED SPECIES IN CORALS. 551 



conditions of life in tliese barriei' pools are peculiar ; the pools are 

 filled with sand, for the fragments which are triturated by their 

 journey to and fro over the baiTier are deposited in them, they 

 contain the minute green filaments of the boring algte, and at 

 mid-day low tides they become heated by the sun to 93*^ Fahr. 

 and more. Theii- coral fauna is practically always the same, 

 Madrepora flourishes in its most highly branching forms, Pocillo- 

 pora always has a good foothold, and the other species are in 

 plenty, but hardly in a state of luxuriance. Beche-de-mer in 

 hundreds live in these pools, and Crustacea, polychfetes, mollusca, 

 eels, and the myriad brilliant fish make up the conspicuous fauna. 

 Of the many rough- water colonies that were transplanted experi- 

 mentally into this environment, not one remained alive at the end 

 of 50 days, and most were dead within a month. The first sign 

 by which a colony shows that its environment is not suitable 

 is by becoming highly pigmented ; both rough-water forms are, 

 when flourishing, very pale corals, being usually of a light bufi" 

 colour, but within a fortnight the transplanted colonies had 

 become of a dark yellow-brown, aiid in Madrepora there was a 

 more than usual tendency to lateral branch formation. In 20 

 days most colonies had some portion dead, and the dead parts 

 became rapidly the site of gi-owing alga? ; in 30 days nearly all the 

 numerous transplanted colonies were dead or dying, and by 50 

 days no portion of any colony remained alive. It was silt that 

 had determined the death of them all ; the stunted, flattened 

 types of Madrepora and Pocillopora are both corals of a rough- 

 water habitat, they are used to clear water in which there is little 

 or no suspended matter, and not a single colony was able to with- 

 stand the slow but certain sedimentation of the bai-rier pool ; 

 when the silt had once fairly determined their death, the fine 

 boring algee completed the ruin. If aiiy colony of branching- 

 coral be i-emoved entire from the lagoon, it will be found that the 

 lower portion is invariably dead ; but this death is in most cases 

 not a natural one resulting from the senility of the zooids, biit 

 merely an index of the amount of silting up of the lagoon that 

 has taken place since the establishing of the colony. Sand is ever 

 being washed into the lagoon through the numei^ous gaps in the 

 island ring, and most decidedly the tendency all over the lagoon 

 is a gradual filling up, by the deposition of finely triturated frag- 

 ments : the floor of the lagoon is fairly steadily rising, and those 

 colonies of corals growing in its bed are for ever being encroached 

 upon by the gradual rising of the sand level. The deposited sand 

 most certainly kills the zooids with which it comes in contact, and 

 the result is that the lower portion of every lagoon colony is 

 killed. 



Silt then, in this atoll, is the most potent factor in causing coral 

 death, and next in importance to the silt comes the seaweed. 



There is a green alga that, at some seasons of the year more 

 than at others, comes to the bai-rier pools in great quantities : it 

 is a growth of fine afreen threads and its efiect on coral-growth 



