2 ME. L. OWEN OK THE [vol. lxxix, 



Flint, Starbuck, Browse, Lacepede, Laysan, Cornwallis, and 

 Clippefton. 



Althougb from time to time various theories have been put 

 forward as to the origin of these phosphate deposits, there can be 

 no reasonable doubt that they are all due, primarily, to bird- 

 droppings. 



With the one exception of Clipperton Island, the deposits 

 consist entirely of more or less leached guano overlying a coral- 

 floor which is, to a greater or less extent, changed by metasomatism 

 into tricalcium phosphate. The deposits mentioned as being of 

 commercial importance have been leached so completely that the 

 soluble phosphates have entirely disappeared, some of the phosphate 

 on Ocean Island, for example, containing over 90 per cent, of 

 tricalcium phosphate (Ca 3 P 2 8 ). On Maiden Island, where the 

 birds still congregate, all three forms of calcium phosphate 

 {CaH 4 P 2 O s , CaHP0 4 , and Ca 3 P 2 8 ) are found; while on Clipperton 

 Island l a mass of trachyte, about 60 feet high, rises through the 

 coral in the south-eastern part of the island. The original structure 

 of the trachyte is still discernible ; but the alkaline silicates have 

 been replaced by phosphates of calcium, aluminium, and iron. 

 The main portion of the deposit on this island is, however, of a 

 type similar to that found on the other islands. 



On most of the islands the underlying coral shows signs of 

 partial dolomitization (43 to 45 per cent, of magnesium carbonate), 

 the magnesia being chiefly derived from the guano. 



It is not my purpose to attempt to consider the numerous and 

 complicated chemical reactions which may certainly, probably, or 

 possibly occur when the complex solution, derived from the guano 

 by the action of meteoric waters, interacts with the underlying 

 rock. ■ Here it will be sufficient to consider the solution as con- 

 taining, in its first phase, calcium phosphate, ammonium phosphate, 

 and free phosphoric acid ; and in its second phase as containing 

 tricalcium phosphate as a major constituent. 



The products formed by the interaction of this solution of 

 guano on the underlying coral mass are similar in all the deposits ; 

 but the ultimate results of the reactions are so completely exhibited 

 in the Ocean Island deposit that I propose to deal with them when 

 making detailed reference to that island. 



II. BlBLIOGBAPIIY. 



The literature having reference to the phosphate deposits of the 

 South Seas is very scanty, and few publications exist which deal 

 with anything more than a superficial examination of the deposits. 

 The list given below may be taken as practically complete : — 



(1) ' The Guano & other Phosphatic Deposits occurring on Maiden 

 Island.' W. A. Dixon, Journ. Eoy. Soc. N.S.W. vol. xi (1877) p. 176. 



This is, so far as I am aware, the first serious attempt to observe 

 1 W. J. Wharton, Q. J. G. S. vol. liv (1898) p. 228. 



