282 CORAL A^ND ATOLLS 



casual visitor can hope to examine with any thoroughness — 

 and it is easy to draw the most diverse conclusions from the 

 inspection of restricted areas ; but with time and famiharity 

 with more of the islands, it becomes evident that there is 

 a compensation in the process : but it is more than a mere 

 compensation, for the repair tends over the whole atoll to 

 outweigh the destruction. 



Although much importance has been attached to the 

 various indications of land gain and loss of the atoll, I do not 

 think that most of the processes of building and wasting that 

 can be seen daily in progress give any real clue to the actual 

 earth movements to which the land maybe subjected. What 

 we really see when we watch a sandy beach piled up, or a 

 strip of shore denuded and the palms uprooted by the waves, 

 is the fluctuating gain and loss of the islands' floating capital. 

 A bank or a spit is made and unmade by a shift of wind or 

 current, it does not indicate a rise or fall of island-level. 



We have seen besides that great alterations maybe brought 

 about by actual temporary change of ocean-level, for, during 

 periods of several days at a time, the surface of wide tracts of 

 ocean may be raised or lowered locally by the banking-up 

 effects of winds and currents, and there is nothing to dis- 

 tinguish this temporary alteration of water-level from an actual 

 depression or elevation of the land. I do not think that 

 this curious phenomenon of the local variation of ocean-level 

 is sufficiently considered ; for it is a very real and a very 

 potent fact ; it is only when your extreme elevation above the 

 sea happens to be a few trivial feet that its reality strikes you. 

 For days at a stretch during storms to the westward, the level 

 of the ocean in the neighbourhood of the atoll will be most 

 distinctly raised ; and any denudation that may occur during 

 this time will present all the appearances of alteration of land- 

 level. For actual indications of land movements, other things 

 than the shifting sand-banks must be looked for, and evident 

 indices of elevation are furnished in the raised platforms of 

 breccia which may be found in certain portions of the barrier 

 ring. The breccia of the shore platforms is entirely the 



