22 TROFESSOR W. J. SOLLAS. 



northern limb of the swamp. Down these the water may be seen descending as 

 the tide falls, and that they communicate freely with the outer sea is shown by the 

 fact that the body of a native child, drowned in the swamp, was afterwards found 

 floating in the waves outside the reef. If these holes, as would appear to be the 

 case, are tlie chief or the only means by which the sea can enter, it is possible that 

 the high-water level in the swamp may stand a little below that of the open ocean. 



The clumps of Porlfcs are not, as we at first imagined, continued downwards into 

 solid reef rock ; on the contrary, they are isolated masses, which may be readily dis- 

 lodged and overturned with the aid of a crowbar ; between them, and filling the 

 interstices of the Heliopora up to the general level of the floor of the swamp, is a 

 fine silt abounding in foraminifera and dead molluscan shells. 



The breccia which bounds a great part of the swamp extends in places to form 

 its surface, lying in a thin encrusting sheet upon the Heliopora reef It would thus 

 appear that the reef is older than the breccia, and as this is but a part of the general 

 hard crust which extends from ocean to lagoon, the reef is probably the oldest super- 

 ficial feature of the atoll. It w^ould seem to have concluded its growth before the 

 present ocean-ridge and lagoon mound were in existence, and then to have become 

 covered up with coral debris, which after consolidation was broken up to form the 

 little beach now bordering the swamp on the landward side. 



The flat summits of the Porifes clumps are a suggestive feature, indicating an 

 arrest of upward growth against a jdaln surface, such as would be aflbrded by the 

 level of low water. Corals with similarly flat summits may be seen at present living 

 in the lagoon, where it is evident that on reaching the low-water level they have 

 ceased to grow upwards, but have extended laterally and again upwards to low -water 

 level till a flat tabular form has been acquired (fig. 15). 



<©© 



-uu .- 1 ^D057r-i — Low water level 



Plan Elevation. 



Fig. 15. — Diagram to illustrate the flat siimmits of corals which have grown up to the low-water level of 



the lagoon. 



If it be admitted that the surface of the Porites clumps marks an ancient level of 

 low water, then it obviously becomes a problem of extreme interest to compare this 

 level with that of the existing sea. Accordingly, with the assistance of Mr. Payne, 

 the boatswain, and a crew of seamen, I made two sets of measurements across the 

 dead coral reef and the swamp from the lagoon beach to the ocean. These gave 

 concordant results, and showed that the summits of the Pontes clumps now stand 

 1 foot 4 inches above mean tidal level, i.e., 4 feet 6 inches above low water at spring 

 tides or 3 feet 9 inches above low water at neap tides. I therefore concluded that a 

 change of sea-level in a negative direction to the extent of about 4 feet had occurred 

 over the site of the Mangrove Swamp since the growth of its ancient reef It is 



