274 THE INFLUENCE OF INANIMATE SUKROUNDINGS. 



As one plateau rose to the region of reef-building corals, these 

 channels would either become perpendicularly deeper or super- 

 ficially wider, according to the force and direction of the 

 currents acting upon them. Its tendency to grow upwards, 

 combined with the general upheaval, would bring the highest 

 part of the coral-stocks to the surface and so expose them to the 

 influence of the tides as well as that of periodical or incidental 

 rainfalls. The process will now be repeated which I have 

 already described as ascertained by observation on separate coral- 

 blocks; if channels have already been formed by submarine 

 currents in the reef which has come to the surface, these will 

 become the most natural and convenient conduits for the water 

 flung up on to the reef; the shallows lying between the outer reef 

 and the land, though of small extent at first, will increase, in the 

 first instance by the outward growth of the reef, and then, in a 

 much greater degree, by the destruction of the enclosed land ; 

 as, at the same time, the amount of water thrown over on to 

 this inner space must increase in proportion to the increased 

 surface, the channels must, in spite of continued upheaval, eat 

 away the rock to a greater depth, and more or less quickly 

 according to the nature of the rock itself. 



Now, if we draw the obvious conclusions from the two an- 

 tagonistic views above mentioned, all the arguments derived from 

 observation again range themselves on my side. If it were true 

 that thick reefs could be formed only during subsidence and ex- 

 clusively by its action, we should be justified in expecting to find 

 the same species of coral throughout the whole depth of the 

 mass, and only the base, with a maximum thickness of about 

 120 feet, could under the circumstances exhibit a certain variety 

 in the materials composing it. To my knowledge, however, no 

 observations exist which afford grounds for this conclusion. 

 Granting, on the contrary, that the reef was formed diiring 

 a period of elevation, the inference naturally follows that 

 the composition of such a raised reef must be heterogeneous, 

 since in the first instance only deep-sea creatures can have' 

 established themselves on it; the reef-corals, properly so called, 

 would in s\ich a case form only a thin layer above the deep-sea 

 corals. And this actually is the case, according to my personal 



