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



believe that the metamorphosis from the heap of organically 

 deposited material, like the reef of the present day, can have 

 been subaqueous. Neither do I believe that the change has been 

 to any great extent effected by the interfiltration from the sea 

 of carbonate of lime in solution. The deposition of finely divided 

 carbonate of lime would, of course, greatly tend to fill up the 

 pores of any coral, but this is not an interfiltration. The cavities 

 in the hard limestone can be accounted for by the shrinkage that 

 would naturally occur in its metamorphosis ; I have also already 

 shown, in my description of the reef at Funafuti, that in the 

 growth of a living reef spaces are often enclosed. The pockets of 

 oxide of iron are due to the solution and abrasion of much of the 

 rock by the rain during its metamorphosis, when it would of 

 course be left undissolved ; the higher lying rock is much freer 

 from it than the lower, as should in such circumstances be the 

 case. The rock at Kambara, covered by the lava, is almost 

 intermediate between the soft limestone of Suva and this hard 

 limestone. The beach sand rock of Rotuma, a white, loosely 

 compacted limestone, is, as I have pointed out, used by the natives 

 for gravestones, becoming on exposure exceedingly hard and com- 

 pact, the various constituents being cemented together by calcite. 

 I believe that in these small slabs a metamorphosis is now taking 

 place precisely similar to that which has formed the dense highly 

 crystalline limestone of the Lau Group out of the loosely aggre- 

 gated masses, such as form a coral reef at the present day. 



I am greatly indebted to Dr Pollard, of the Caius College 

 Laboratory, for having kindly analysed two specimens of this hard 

 limestone. Both came from the inner side of the ridge of Namuka, 

 the one being a pure white rock without any large crystals, and 

 the other a very fine grained reddish rock. The white rock (two 

 determinations) gave 78 - 6 per cent, carbonate of lime and 21*5 

 carbonate of magnesium. The red rock gave 85'2 per cent, 

 carbonate of lime, 6"7 carbonate of magnesium, 4 - 2 oxides of iron 

 and aluminium, and 3'5 silica 1 . 



These analyses show that dolomitization has gone on to a 

 considerable extent in this hard limestone. Dana states 2 that 

 95 — 98 per cent, of the common reef-corals is formed of carbonate 

 of lime, the remainder being made up of 1J — 4 per cent, of organic 

 matter with about 1 per cent, of earthy ingredients. Dana also 

 notes 3 that coral sand and mud contain little or no magnesia 

 generally; he gives however an instance where 1*38 per cent, was 

 obtained. These analyses, so far as they go, afford no probability 



1 Analyses of two very white specimens from Kambara and Fulanga gave me 

 19*8 and 10 ■ 7 per cent, respectively of carbonate of magnesium. 



2 Coral and Coral Islands. J. D. Dana, 1890, p. 99. 



3 Loc. cit. p. 394. 



