MAYOR— THE REEFS OF TUTUILA, SAMOA. 233 



Vaughan (1914, Journal Washington Acad. Sci., Vol. 4, pp. 26- 

 34) showed that the platform upon the seaward edges of which the 

 Florida reef is now growing, extends northward into a region too 

 cold for coral growth. Moreover, the disconnected coral patches 

 which rim the seaward edges of the Great Bahama Bank are many 

 of them growing not at the extreme edge of the bank but at an 

 appreciable distance inward from its seaward margin. The hard- 

 rocky floor of this bank is covered with a layer of flocculent cal- 

 careous mud which when the water is agitated becomes churned into 

 a milky mass fatal to coral growth Thus coral heads can very rarely 

 attain a foothold excepting near the seaward edges of the bank 

 where pure ocean water in large measure replaces the silt laden 

 waters of the bank. 



In other words, the coral patches which rim the Bahama Bank 

 have merely grown in modern times near the seaward edges of a 

 submerged flat, the extraordinarily level character of which can 

 only be explained by assuming it to have been formed in conformity 

 with sea level. Only a water-level could be so flat. 



Daly's (1915, Pi'oc. Amer. Acad. Arts and Sci., Vol. 51, pp. 157- 

 251) opinion that the cooling of tropical seas in glacial epochs had 

 much to do with determining the relative abundance of corals, has 

 opened an interesting field for research, but according to Vaughan 

 (1919, U. S. National Museum Bulletin No. 109, p. 256), the West 

 Indian fossil reefs do not- support this idea, for corals grew exten- 

 sively in this region in Pleistocene times. 



W. M. Davis, in numerous papers,^ has called prominent atten- 

 tion to the following well-established facts : That under still-stand 

 conditions, if the land be not surrounded by reefs, the sea will cut 

 into the clifTs faster than the valleys can be excavated by subaerial 

 erosion, and thus the streams will cascade into the sea. Then if the 

 island subsides, or the sea level rises, and drowns the valleys, the 

 submarine slopes at the spur-ends will be steeper than the slopes of 

 the submerged sides of the drowned valleys. Also silt will be 

 largely pocketed at the stream mouths in the inner ends of the 

 drowned valleys, and will settle to the bottom before it reaches the 



1 A good resume is given in Davis, 1919, Trans. New Zealand Inst., Vol. 

 51, pp. 6-30. 



