12 ME. L. OWEN ON THE [vol. lxxix, 



deposition of the guano and, contemporaneously, to a practically 

 uninterrupted elevation. The deposition of guano, at least to any 

 •considerable extent, appears to have ceased before the final gentle 

 tilting of the island occurred. 



The solutions from the guano would, first of all, react vigorously 

 with the coral and coral-sand directly beneath, the reaction 

 resulting in a structureless mass of impure tricalcium phosphate. 

 Following on the first violent action Avould be a slower replacement 

 of calcium carbonate, in the underlying material, by tricalcium 

 phosphate. Where the action occurred in connexion with the 

 oolitic calcium carbonate so common on coral reefs, it would give 

 vise to oolitic phosphate-particles, and form those t} r pes of phos- 

 phate which imitate various detrital and concretionary limestones. 

 The coral not eaten away by the guano solutions would be changed 

 by metasomatism into the phosphate pinnacles observed in the 

 workings. Finally, the solutions would consist of an equilibrium 

 mixture of tricalcium phosphate, calcium carbonate, and carbon 

 dioxide bereft of the power to attack coral. By the loss of its 

 carbon dioxide this solution would yield the substance to which 

 reference has been made under the name of translucent 

 phosphate. 



The form of the coral pinnacles underlying the phosphate, and 

 that of the concretionary limestone (now changed to phosphate) 

 occurring between the pinnacles, suggest that the island suffered 

 at least one submergence before the deposition of the guano, in 

 any quantity, commenced. This supposition is strengthened by 

 the fact that newer corals (now phosphatized) are occasionally 

 found attached to the dolomitized pinnacles. 



The island shows no evidence of having suffered any extensive 

 submergence since the deposition of the guano. If submergence 

 had occurred, there would certainly have been left traces of coral 

 beaches overlying the phosphate deposit. No such beaches- have 

 t)een observed. It may be argued that such beaches were con- 

 verted into phosphate hj solutions leached from guano subsequently 

 deposited. This may have happened, but I consider it very im- 

 probable, as I have shown that the variation in the percentage of 

 tricalcium phosphate, within the deposit proper, rarely exceeds 

 1 from top to bottom at any point. If fossil beaches existed 

 within the body of the deposit, there can be little doubt that they 

 would be indicated by a local change in the phosphate content. 



Since the cessation of the dejDosition of guano, the island shows 

 evidence of having been tilted from the north-north-west towards 

 the south-south-east, as recorded by the dips of the platforms of 

 coral and by the isophosphatic planes. 



Many hundreds of analyses have been reviewed in arriving at 

 these conclusions. These show that in bulk samples, the per- 

 centage of tricalcium phosphate varies between 80 and 90. One 

 representative analysis of the phosphate, when dried at 100° C, is 

 recorded here : — 



