250 BULLETIN 103, UNITED STATES NATIONAL MUSEUM. 



shore. Such a floor if formed by agencies associated with the pres- 

 ence of the reef will not be composed of rock demonstrably older 

 than the reef, and will not exhibit geologic phenomena that in age 

 clearly antedate the reef; but if it can be shown that the rock of the 

 floor is older than the reef and that the floor has had a geologic his- 

 tory antecedent to the formation of the reef, it is demonstrated that 

 the reef is merely growing on the surface of a flat whose formation 

 is entirely independent of the reef development. 



(d) The growth rate of corals, which furnishes one of the checks 

 to be applied to the Glacial-control hypothesis of the formation of 

 living reefs, is further considered on pages 253, 254. 



SOLUBILITY OF CALCIUM CARBONATE IN SEA WATER. 



As the formation of lagoon channels behind barrier reefs and of 

 atoll lagoons by the solvent action of carbon dioxide (C0 2 ) dissolved 

 in sea water is a part of the coral-reef hypotheses of Semper, Murray, 

 A. Aggassiz, and Gardiner, if lagoons and lagoon channels have been 

 formed in the way indicated, in the Tropics the surface waters of 

 the ocean should contain an excess of carbon dioxide (C0 2 ) and 

 should exercise a demonstrable solvent effect on calcium carbonate 

 (CaC0 3 ). If it should be found that there is no excess of carbon 

 dioxide (C0 2 ) in such water and that the water is saturated with 

 reference to calcium carbonate (CaC0 3 ), the hypothesis of the forma- 

 tion of lagoons and lagoon channels in the manner postulated by 

 Murray and others must be definitely abandoned. 



In 1913, Mr. E,. B. Dole undertook at Tortugas, Florida, certain 

 examinations that were intended to solve this problem, if possible. 

 In 1914 I summarized in the following words the results I had ob- 

 tained from a study of the bottom samples along the Florida reef 

 tract, those of Drew on dentrifying bacteria, and those of Dole on 

 the chemistry of the waters. 1 



There are two rival hypotheses for the formation of atolls : One of these attributes 

 them to the submarine solution of the interior of a mass of limestone, the other ac- 

 counts for them by constructional agencies. In order thoroughly to test the solution 

 hypothesis the results of four lines of investigation were brought to bear upon it, 

 and all are accordant. (1) All the bays, sounds, and lagoons within the Florida reef 

 and key region are filling with sediment; (2) Drew's investigations of dentrifying 

 bacteria show that chemical precipitation of calcium carbonate is taking place in 

 the lagoons; (3) the chemical examination by R. B. Dole of samples of sea water 

 flowing into and out of Tortugas lagoon, collected twice daily for a lunar period, show 

 that although both carbonate and bicarbonate radicles are in solution uncombined 

 carbon dioxide is not present, and that the water possesses no capacity for further 

 solution of calcium carbonate by virtue of its content of free carbon dioxide; (4) the 

 determinations by Dole of the salinity of the water within the Tortugas lagoon and 

 at the southern end of Biscayne Bay show a higher concentration than that in the 

 open sea water on the outside, indicating that tidal inflow and outflow are not suffi- 

 cient completely to mix the water in the lagoons with the water of the surrounding 



i Wash. Acad. Sci. Journ., vol. 4, pp. 27-28, Jan. 19, 1914. 



