476 Mr Gardiner, The Coral Reefs of [Mar. 7, 



more outwards, and anastomose more frequently with one another. 

 The whole structure then, provided that there are no movements 

 of elevation or subsidence, should be an open spongework, full 

 of chambers and passages, some of which are partially, or com- 

 pletely, blocked by sand and boulders. 



The consolidation of sand and fragments by the infiltration 

 of sea water is not, I believe, of any importance. That molecules 

 of a salt can out of a saturated solution of that salt attract 

 more molecules admits of no doubt, but there is no evidence to 

 show that this is going on with regard to lime salts to any con- 

 siderable extent in the sea ; the pores of the dead corals in 

 reef- flats and in the raised reefs of Viti Levu do not show any 

 marked filling up, which can be ascribed to such a cause. Sand 

 is constantly to be found in small holes in the reef completely 

 walled in by the growths of nullipores and corals ; it exists also 

 in greater or less quantity in nearly all the pools on the reef, 

 in all boat channels and on all inner reefs, but I could not 

 succeed in finding it consolidated together in any way or to the 

 rock below it under low tide level. The sand in the soft limestone 

 quarry at Suva, although it has evidently been upheaved for a 

 considerable time, is still badly consolidated, and the thin layer 

 between this rock and the soapstone is almost loose. The Report 

 on the Coral Reef at Funafuti by Prof. Sollas mentions that loose 

 sand and boulders were met with as deep as the bore went. At a 

 depth of 72 feet in the bore, on the outer sides to windward of the 

 atoll " the bottom of the hole was surrounded by quicksand, con- 

 taining boulders of coral, and as fast as the sand was got out, so 

 fast it flowed in and faster." The consolidation of the fragments 

 here has certainly not been effected, and I am inclined to doubt 

 whether infiltration does not have an entirely opposite effect by 

 means of the carbonic acid gas in the water. 



The hard parts of a reef, above low tide, are made up of frag- 

 ments of coral, sand and nullipores, the consolidation of which 

 seems to be due to the deposition of carbonate of lime on the 

 evaporation of the water at low tide. Loose fragments after a 

 time get a bed of fine mud, which, as more fragments are piled up, 

 gradually surrounds and buries them. On the reef-flat there is 

 less coral and less sand, and much of the rock is formed by a 

 structureless and very compact limestone, which also forms the 

 whole rim. Where corals are found in this rock, they are not as a 

 rule as much eaten into or worn, as they are in the parts above 

 the low tide level ; they do not seem like fragments at all, but as 

 skeletons of corals in the position of growth. This appearance leads 

 me to believe that the rock here has grown up to the surface exactly 

 in the position in which it is found, and is not made up of masses, 

 thrown up from outside the reef by the seas and cemented together. 



