268 THE INFLDENCE OF INANIMATE SUEEOUNDINQS, 



an old colony of Porites. Its surface, in the first instance quite 

 level, will be gradually hollowed out by various co-operating 

 influences, and so at last a raised margin, only cut through by a 

 few channels, will surround a central hollow. A precisely 

 similar result may be produced in a reef undergoing slow up- 

 heaval. Suppose, for example, that Kriangle had a tolerably 

 level surface, like the shallows to the south of Ngaur, or like 

 Kossol, before it was raised to the average level of the sea : from 

 the first moment when it was exposed to aerial influences a 

 process of destruction must have begun on the surface of the 

 reef, perfectly analogous to that which takes efiect on large 

 isolated coral blocks. Animals and plants first establish them- 

 selves on the surface of the reef, which is certainly highly 

 favourable to them; they excavate and penetrate the solid 

 limestone of the coral in every direction while the rain falling 

 on the face of it kills the polyps themselves ; the rain and the 

 sea-water flung over the margin of the reef must remain there 

 if they can find no outlet. In the first instance the water on the 

 summit of such a rising reef may often remain standing, but 

 channels must soon be worn through the constantly growing 

 and rising margin, or submarine drainage may easily arise, since 

 boring animals and plants are able gradually to destroy even 

 the hardest limestone. That there are channels of this kind 

 in Kriangle may be inferred from the fact that at low ebb- 

 tides the water in the lagoon stands at no higher a level 

 than it does outside the atoll ; but this could not occur if the 

 water thrown over into the lagoon at high tide did not easily 

 find an outlet throiigh the reef itself, for surface channels, 

 which might serve the same purpose, are wholly wanting. The 

 extreme porousness of the soil of Kriangle below the sea-level is 

 also proved by another fact. The inhabitants have sunk deep 

 wells, and a very large tank, about ten feet deep, for bathing pur- 

 poses ; the water in these is usually fresh, as they fill dui-ing 

 the rainy season. But when, after persistent drought in the 

 dry season, the level of the water sinks considerably in the 

 basin, the bottom of it is frequently brackish, and that without 

 storms having arisen to cause the sea- water outside to wash over 

 the low island and into the tank or the wells. This shows that 



