252 BULLETIN 103, UNITED STATES NATIOISrAL MUSEUM. 



give qualitatively the relations of currents to land forms, and com- 

 pletely confirm the more qualitative generalizations of Hedley and 

 Griffith Taylor, which in brief are the axis of elongation of linear reefs 

 is parallel to the direction of the dominant current while the bow of 

 a crescentic reef is directed toward the direction whence the dominant 

 current comes. These relations of reef form to current direction are 

 most striking where the reefs rise above comparatively shallow plat- 

 forms, as along the Great Barrier reef of Australia and along the 

 Florida Keys. In atolls that more or less encircle the flat tops of 

 submarine peaks, although currents are undeniably important in 

 shaping sections of the reefs, they are not of so great importance as 

 reefs that rise above shallow, long, Avide platforms. 



CRITERIA FOR DETERMINING THE EFFECT OF GLACIATION AND DEGLACIATION 

 ON THE DEVELOPMENT OF LIVING REEFS. 



Daly's elaborate paper on the Glacial-control theory of coral reefs 

 has been cited on page 245. If the Glacial-control theory is true the 

 following conditions should prevail: (a) There should be evidence of 

 geologically Recent submergence of most of the shore-lines of the 

 earth; (b) the average amount of submergence should be equal to the 

 amount of lowering of the ocean-level during Pleistocene glaciation; 

 (<?) the position of the strand line during Pleistocene glaciation should 

 be indicated by scarps separating flats, and the amount of sub- 

 mergence indicated by their present position below sea level should 

 agree with the amount of raising ocean level due to deglaciation; 



(d) rate of growth corals should be such that since the disappearance 

 of the continental ice sheets coral reefs could grow to a thickness 

 equal to the amount sea level was taised as a result of the deglaciation; 



(e) living barrier coral reefs and atoll reefs should be superposed on 

 antecedent basement flats or platforms. It should here be stated 

 that the fact that there has been local differential crustal movements 

 does not at aU invalidate the importance of the Glacial-control theory 

 in its application to the explanation of the modern coral-reef develop- 

 ment. 



Of the criteria stated in the foregoing list only the amount of 

 vertical change in the position of sea level because of glaciation and 

 deglaciation, the length of time since the disappearance of the great 

 continental glaciers, and the rate of growth of corals need discussion 

 at tliis place. After their consideration some attention will be given 

 to other criteria of less determined value. 



AMOUNT OF VERTICAL DISPLACEMENT OF STRAND UNE BY GLACIATION AND DEGLACIATION. 



It is entirely obvious that the withdrawal of water from the ocean 

 to form the Pleistocene continental glaciers would lower sea level, 

 and that the return of the waters so locked up to the ocean upon the 

 melting of the continental glaciers would raise sea level back to where 



