NARRATIVE OF THE EXPEDITION IN 1896. 23 



scarcely likely that this change has been merely local and confined to this spot ; had 

 its amount proved to be slight, explanations based on possible alterations in the direc- 

 tion of waves or currents, or on a possible difference between low-tide level in the 

 sea and the lagoon might have been offered for consideration ; as it is we seem driven 

 to the admission that a general elevation of the atoll or fall of the sea-level must 

 have occurred since the Heliopora reef was in active growth. 



The level summits of the Porites clumps have been regarded as marking a once- 

 existing level of low tide ; this is a minimum limit, for they may not have reached so 

 high ; it is very unlikely they grew higher. Abruptly succeeding them are the breccia 

 beds, evidently formed under very different conditions, such as might be brought 

 about by a change of sea level. These beds are very difficult to explain, but the 

 evidence of the Taiisale reef seems to preclude an appeal to depression, and we are led 

 to consider the effect of slight elevation of the atoll ; a large elevation being equally 

 precluded by the evidence of the Taiisale reef At the close of the Taiisale period 

 we imagine that the atoll was wholly submerged, and no ocean ridge nor lagoon 

 mound was in existence. An elevation by which the mean-tide level became the low- 

 tide level would expose the outer portion of the reef to the destructive action of the 

 breakers, which, tearing off corals and coral fragments, would drive them across the 

 inner portion of the reef into the lagoon, where they would accumulate in a low bank, 

 and this would increase by continual accretion seawards till the waves had piled up a 

 bank of breccia broad enough to present sufficient resistance to their further advance 

 over the reef The final stage might even have been the growth of an ocean ridge. 



The formation of the breccia beds would thus depend on a nice balance between 

 erosion and deposition, and might have occurred at a time when the summits of the 

 Porites clumps corresponded with mean-tide level ; a further slight elevation would 

 then appear to have followed, raising the atoll about 1 or 2 feet ; the breccia beds 

 thus became exposed to marine erosion, but the resulting fragments, partly as a result 

 of its increment in height, could no longer be everywhere driven across the whole 

 breadth of the reef, but accumulated, along with material contributed by growing 

 corals, to form the ocean ridge. The pinnacles and pinnacle ridges of the outer glacis 

 are remnants of the breccia beds, and their form and disposition are to be explained 

 as the result of marine erosion acting upon beds having a seaward dip, as already 

 suggested in discussing the islet of Pava. They are now the chief irregularities on 

 a slope of maximum stability, worn by the waves during a stationary period into a 

 curve best fitted to resist their onslaught. This and the existence of the living 

 breakwater afforded by the nullipore rim will explain the surprising smallness of the 

 amount of destruction* at present accomplished on the reef by the great waves of the 

 Pacific Ocean. 



* This was a frequent source of surprise. During a windy night one would often be kept awake by 

 the incessant thunder of the surf upon the outer beach, and it was difficult to resist the impression that 

 the island was being washed away. On visiting the reef in the morning one expected to find the beach 



