222 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1917. 



to North America, and by Pliocene time the corals of distinctive 

 Indo-Pacific facies had become extinct on the Atlantic side, so that 

 the Pliocene coral fauna of Florida is purely Atlantic in its affinities. 

 After the differentiation of the Atlantic from the Indo-Pacific fauna 

 it seems that there was a short connection somewhere that permitted 

 the Atlantic fauna to extend on the Pacific side of America up to 

 the head of the Gulf of California. 



THEORIES OF THE FORMATION OF CORAL REEFS. 



Three kinds of coral reefs are generally recognized, viz : (1) fring- 

 ing or shore reefs which occur along the shore; (2) barrier reefs 

 which occur at variable distances offshore and have lagoons from 

 1 or 2 to as much as 30 or even 40 fathoms in depth between them 

 and the shore line; (3) atolls, which are ringlike and inclose lagoons 

 above whose surface no land masses of importance protrude. 



As the literature on coral reefs is 30 enormous that a detailed re- 

 view of it in this paper is impossible, coral reef theories will be here 

 classified into three general categories, with a subordinate division 

 of the third. 



1. The first theory is that of Darwin and Dana. According to 

 these authors corals first form a fringing reef along the shore of the 

 gently sloping bottom of a subsiding land area; the reef grows up- 

 ward at such a rate that its top remains near the surface of the 

 water and through retreat of the shore it is converted into a barrier. 

 Continued subsidence, where the inclosed land area is an island, may 

 result in the production of an atoll circumscribing a lagoon without 

 any land mass projecting above the water level. But the Darwinian 

 hypothesis involves more than mere subsidence and the conversion 

 of a fringing into a barrier reef, for it also attempts to account for 

 extensive submarine platforms by assuming that they have been built 

 upon sloping basements through agencies dependent on the presence 

 of reefs. 



The accompanying two figures (p. 223) are reproductions of Dar- 

 win's original illustrations; wdiile the third (p. 224) is J. B. Jukes's 

 diagrammatic cross section of the Great Barrier Eeef of Australia. 



2. The next general theory of coral reef formation was originated 

 by Carl Semper,^ who, in 18o3, after studies in the Pelew Islands 

 and noticing evidence of uplift there, announced the opinion that 

 atolls could be formed in areas of stability, or even uplift, by the 

 solution of the interior of limestone masses, and that erosion by 

 currents and wave cutting could develop channels behind fringing 

 reefs, and in that way transform a fringing into a barrier reef. 



^ Semper, Carl, Rcisebcricht : Zeitschr. fiir wiss. Zool., vol. 13, pp. 503-5G9, 1863. 



