138 ME. A. E. FINCKH. 



patches of the eastern lagoon platform, already mentioned in Sub-section (1), indicate 

 that, when growing on a surface level enough to allow of an accumulation of loose 

 material, the Halimeda builds up on its own debris. One of these patches, when broken 

 into, shows the actually living joints, which are indicated by their green colour, to 

 occupy only the upper 4 to 6 inches of the mass ; below this the mass is made up of the 

 dead, but still connected, joints of the plant, which pass downwards gradually from 

 the coarse into the fine Halimeda sand of the bottom of the lagoon. This process of 

 heaping up of ffcdimeda sand is undoubtedly going on throughout the lagoon and on 

 the ocean slopes. On the wall bounding the lagoon platform of the eastern rim of 

 the atoll it was observed that the Halimeda forms an almost continuous living fringe. 



(3) The Corah and Hydrocorallines. — Of these corals, the mode of occurrence of 

 which is of importance in the formation of the reef, we have as in the previous 

 chapter, (a) Heliopora cceridea ; (h) The Millepores ; (c) The Poriles family ; 

 {d) Madrepora loripes, and [e) PociUopora. 



(a) The Heliopora is of primary importance. 



At the present time it is mainly this coral which is laying the foundations for 

 further additions to the land in the shape of shoals and platforms in the lagoon. Its 

 shoal-like clusters have already been mentioned in Sub-section (l) as occurring on the 

 western rim of the atoll. The highest parts of those shoals, or clusters of coral, which 

 are still from 2 to 3 fathoms below the surface of low- water spring tide, consist entirely 

 of this coral. Once, however, risen to within a few feet of low-water level spring tide, 

 their summits become overgrown by tlie lichenous Lithothamnion, The interstices 

 between the branches of the coral become filled in, so that finally a continuous and 

 compact rock results, with a surface a comparatively smooth plane, which eventually 

 reaches tlie level of low- water spring tide. Tlie rock consists of the stems of the coral 

 with which the shoal began its existence, of foraminiferal sand, of the remains of the 

 finely branching Lithothamnion and of Halimeda, all bound togetlier liy the agency of 

 the lichenous Lithothamnion. By this process shoals arising on the edge of the reef 

 platform are added to it as solid rock. 



The remarkable uniformity of level of the tops of the reticulating brandies of the 

 Heliopora, as illustrated by tlie dead remains in the Mangrove Swamp at Fimafuti 

 (see Plate 18) and, but less clearly, on all the lagoon platforms, Is due to the surface of 

 the water forming a sharp plane of limitation to their upward growth. In deeper 

 water where there is no limit to its growth in height the coral expands irregularly 

 upwards as well as outwards. 



A somewhat singular occurrence o^ Heliopora ccBrulea, growing 2 feet above low- water 

 spring tide, was observed in a sliallow basin on the reef at the north end of the Islet 

 of Amatuku, Both Heliopora and Porites were seen to be growing there in a healthy 

 condition. At high tide the outline of the basin is entirely obliterated, but as the tide 

 retreats a distinct rim becomes visible, hemming in a sheet of water, the level of which 

 only falls a certain depth with the tide and remains stationary during the rest of the 



