234 



ANNUAL EEPOET SMITHSONIAN INSTITUTION, 1917. 



canic island Saba, and the gteep shore along the north side of St. 

 Croix. The presence of a flat seems necessary to initiate vigorous 

 coral growth. 



Only a few paragraphs will be devoted to atolls, of which there 

 are two kinds. Those of the first kind are ring-shaped segments 

 of long reefs that rise above shallow platforms, such as the atolls of 

 the Great Barrier Eeef of Australia and the Tortugas atoll of 

 Florida. These are shaped by currents that are mostly wind-in- 

 duced. The convex sides of such atolls are toward the wind and the 

 open sides are to the leeward. The accompanying diagram, copied 

 from Hedley and Griffith Taylor, illustrates the principles of their 

 formation. That there never was any central land area in such atolls 



is obvious. The 

 other kind of 

 atolls is those 

 whose rims more 

 or less completely'' 

 margin the flat 

 summit areas of 

 submarine moun- 

 tains or plateaus 

 that almost reach 

 the surface of the 

 sea. This kind 

 of atolls was the 

 subject of special 

 . study by Admi- 



FiG. 15. — Diagram to show how a linear reef lying across ral VYnarton, Or 



THE WIND IS FORMED INTO A HORSESHOE. AFTER IlEDLEY AND tllCBritisll NaVy, 

 GRIFFITH TAYLOR. ^^,j^^ ^^-^^^^ ^^^ 



the uniformity of the depth of the lagoon floors, and stated, as 

 Chamisso years previously had done, that the margining reefs are 

 only more or less continuous. He also laid special stress on the 

 fact that the flat floors of the lagoons did not accord with Darwin's 

 hypothesis, according to which they should be concave, more or 

 less bowl-shaped, and expressed the opinion that the summits had 

 been leveled by marine erosion previous to the formation of the 

 atoll rims. It appears to me that the most plausible explanation of 

 atolls is that they have formed on flat summit areas during moderate 

 submergence. 



In reply to a criticism of my interpretation of the relations of 

 offshore reefs to the platforms above which they stand because I have 

 not attempted to explain the origin of the platforms,^ I may say that 



^ Wharton, W. J. L., Foundations of coral atolls : Nature, vol. 65, pp. 390-393, 1897. 

 = Davis, W. M., The origin of coral reefs : Nat. Acad. Sci. Proc, vol. 1, pp. 14G-152. 

 March, 1915. 



