E. C. Andrews — Coral Reefs in the Fiji Islands. 137 



feet above the sea, and the deep dissection by streams of the 

 uplifted peneplain, were the main activities which had deter- 

 mined the present topography of the mainland, nevertheless 

 the Great Barrier Reef of Australia, as it appears to-day, has 

 resulted from the growth, upwards and outwards, of coral-reef 

 organisms, upon a submerged portion of the coastal area. This 

 coastal submergence is of recent date. 



This is in accordance with the general principles of Barrier 

 Reef origins laid down by Darwin and Dana. Jukes also con- 

 sidered that Darwin's explanation of the Great Barrier Reef 

 of Queensland was the most satisfactory supplied to date.* 



Professor Alexander Agassiz, however, considered that the 

 action of subsidence had been called in unnecessarily by these 

 writers.f 



In his opinion the coastal portions of the recent Queensland 

 area had been cut by the waves to the condition of a submarine 

 platform, and that coral and other reef-building organisms had 

 grown upon " the terraces of erosion " to which the coastal 

 area had been reduced thus, and by the coalescence of coral 

 masses once isolated, the Great Barrier Reef had been formed. 



In the Fiji Report of 1899 by the writer, sufficient account 

 was not taken of coral-reef formation during submergence 

 whereas, at a later period, the writer was compelled to regard 

 submergence as essential in the formation of the Great Barrier 

 Reef of Queensland. On this account a revision of the Fiji 

 work was made from memory to ascertain whether submergence 

 might not have been operative there also. Not having had the 

 chance to reexamine the Fiji Islands since the report of 1899 

 was made, it has been deemed advisable to reopen the case for 

 submergence in Fiji, merely as a possible, or rather probable, 

 alternative to the published statement. 



The accompanying brief notes on several islands in the Fiji 

 Group may suffice to make the case intelligible (see accom- 

 panying map). 



Viti Levu, the main island of the Fiji Group, is about 1,500 

 square miles in area.;}; Strong streams draining the island rise 

 in high plateaus, thence they enter upon torrent tracks between 

 high ranges, while still lower down stream they flow through 

 deep wide valleys, finally entering the sea as long ramifying 

 salt water arms often dotted with islands and occupied by del- 



* Voyage of H. M. S. " Fly," vol. i, pp. 333-347. 



|A. Agassiz, " Great Barriei Reef of Australia," Bull. Mus. Conip. Zool. 

 Harvard College, xxviii, p. 110, 1898. 



\ For a general description of the structure of the Viti Levu two papers 

 by W. G. Woolnough might be consulted with advantage. 



The Continental Origin of Fiji, General biology, Proc. Linn. Soc, N. S. 

 W., vol. xxviii, p. 457, 1903. 



