MR GARDINER ON ATOLLS. I I9 



The crescentic or ring shape of islets on the reef is explained by 

 sand, carried into the lagoon by the currents, being deposited at 

 their extremities. 



A very suggestive paper by Mr J. J. ListerS in 1891, pointed 

 out the occurrence of both high volcanic and limestone islands in 

 the Tonga Group. The elevation is considered to have been " in 

 progress for a long period," and to have been of a recurring nature 

 with intervals of rest. 



In 1894 appeared a long account by Professor Alexander 

 Agassiz of the Bahamas and the elevated reefs of Cuba-. The paper 

 concludes with a chapter on recent views of the formation of coral 

 reefs in reference to the West Indies. Attention is especially drawn 

 to the formation of marine limestones at great depths by accretion. 



Prof. Sollas' " Report on the Coral Reef at Funafuti^," in which 

 he comes to the conclusion that the facts collected by himself 

 support the subsidence theory, was closely followed by a short 

 paper from Admiral Wharton* to impress more fully the importance 

 of the cutting down action of the sea on volcanic islands and the 

 formation of flat banks at depths, within which the reef-building 

 organisms are known to live. 



An important paper by Prof. Agassiz on " The Great Barrier 

 Reef of AustraliaV' in which he points out the absolute impossi- 

 bility of explaining its manifold conditions on the subsidence 

 theory, suggested to me the propriety of re-examining my speci- 

 mens and notes, collected during a six months' residence at Funafuti 

 and Rotuma, and four months in Fiji, which I spent simply in 

 travelling round the islands of that Group, on which also a short 

 paper has since been published by Prof. Agassiz®. 



In a previous paper' I pointed out that a determinable amount 

 of oxygen is given off by corals of the genera Pocillopora, Prion- 

 astraea, Madrepora^ and Euphyllia. Since then I have examined 

 a large number of polyps from different specimens of these genera, 

 and find that they are all crowded with zooxanthellae, either 

 imbedded in the endoderm or free in the gastro-vascular cavity. 

 In no case have I been able to find any organisms, which might be 

 supposed to have served the polyps for food. It is noticeable too 

 that Fowler^ and Bourne® in their accounts of the structure of 

 different corals almost invariably mention the occurrence of these 

 algae. Some specimens of the Pocillopora and Madrepora examined 



^ "Notes on the Geology of the Tonga Islands," Qnar. your. Geo. Soc, vol. XLVii, 

 p. 590 (1891). 



^ Bull. Afus. Comp. Zool., Harvard, vol. xxvi, No. i, p. i (1894). 

 ■' Nature, vol. LV, p. 373 (1897). 



* Nature, vol. LV, p. 390 (1897). 



* Bull. Mus. Comp. Zool., Harvard, vol. xxvni. No. 4, p. 98 (1898). 

 ' Ann. and Mag. Nat. Hist., ser. 7, vol. I, p. 231 (1898). 



^ " The Coral Reefs of Funafuti, Rotuma and Fiji, together with some Notes on the 

 Structure and Formation of Coral Reefs in general," Proc. Camb. Phil. Hoc, vol. IX, 

 p. 417 (i8y8). 



» Q. y. M. S., vols. XXV— XXX (1885—90). 



