DISTRIBUTION OF SUBMARINE BANKS 527 



with a number of minute reefs which I have described in Fiji and which 

 are now brought to the surface again by elevation (1916, h). 



The Sulu Sea, inclosed by the southern Philippines and Borneo, con- 

 tains near its middle the Sultana bank, 18 miles long, with very little 

 marginal reef ; near by are the Kagayanes reefs, which imperfectly inclose 

 a bank of similar length; farther south, where the occurrence of broader 

 fringing reefs indicates a less rapid subsidence, the sea is occupied by 

 smaller, atoll-like surface reefs of irregular development. The China Sea 

 is much more remarkable for the number of its submarine banks, of which 

 the Macclesfield, above mentioned as lying in the center of this sea, is the 

 largest. Westward from the Macclesfield bank, a third or half way to the 

 coast of Tonquin China, a number of smaller banks are partly rimmed 

 with reefs, as if subsidence were of smaller measure or slower in that 

 direction. Farther south is a remarkable group of banks — the Tizard, 

 Eifieman, Prince of Wales, Prince Consort, and Vanguard — with imper- 

 fect reef rims or with no rims, and with varying depths, as will be more 

 fully shown below. 



It may here be noted that Niermeyer (1911) and Wichman (1912) 

 have called attention to the relative rarity of atolls in the East Indies 

 and to the small size of those that occur. Both these authors, although 

 differing in other points, agree that the reefs which are found there can 

 not be explained according to Darwin's theory: but the discussions are 

 so incomplete — the evidence of embayments and unconformities being 

 overlooked — that the conclusion seems untrustworthy. Sluiter's expla- 

 nation of the scarcity of reefs in the shallow Java Sea, 10 or 20 fathoms 

 in depth, as a consequence of the inability of corals to establish them- 

 selves on the muddy bottom (1889), is better supported. Whether atolls 

 were more numerous in the southern archipelagoes when the non-embayed 

 islands stood lower will be determined by future exploration; Roti, near 

 Timor, must, according to Brouwer (1914), have been an atoll when its 

 highest limestones were formed, and certain other limestone islands, less 

 satisfactorily described, appear to be of the same origin. 



It seems impossible to make the peculiarly grouped facts of the Philip- 

 pines and the China Sea accord with the sharply defined requirements of 

 the Glacial-control theory, though they accord remarkably well with the 

 more elastic requirements of the theory of subsidence. One theory de- 

 mands that atoll lagoons and submarine banks shall have depths of about 

 40 fathoms or less ; that islands shall, as a rule, be bordered by platforms 

 of similar depth, fronted with barrier reefs and backed with partly sub- 

 merged cliffs; that the cliffs shall not be interrupted by embayments of 



