IN THE FORMATION OF CORAL REEFS. 347 



species. The spicule determinations were consequently restricted to these 

 more abundant forms. In selecting material for the determination, colonies 

 of various sizes were taken, in order to have the final averages cover as wide 

 a series of the different ages of the colonies as was possible without using the 

 larger specimens of any species. The results of the determination on five^ 

 specimens of the several species are shown in table i. 



As would be expected, those forms having the thickest layer of coenen- 

 chyma about the horny axis showed the highest percentage of spicules. 

 Those also in which the ccenenchyma is most dense, e. g., Eunecia rousseaui, 

 have a much higher spicule content than others in which the ccenenchyma 

 is soft — tor example, Pseudoplexaura crassa, in which the entodermal canals 

 and the polyps are relatively very large. This last-mentioned difi^erence 

 becomes much more evident when the dry weight of the colonies is taken as 

 the basis for computing the percentage of spicules. In Eunecia rousseaui 

 the proportion between the fresh weight and the dry weight was as lOo to 85, 

 while in Pseudoplexaura crassa it was as 100 to 55. The same characteristic 

 was especially noticeable in Briareum, in which (although the central axis 

 consists entirely of interlaced spicules) the spicule content was smaller than 

 in several other forms and, indeed, less than the average for the entire 12 

 species on which determinations were made. That the size of the spicules 

 is also an important factor is shown by the fact that while all members of the 

 genus Gorgonia have a relatively thin coenenchyma, the presence of very 

 small, densely packed spicules brings the spicule content well up to that 

 shown by other forms. 



The data contained in table i were obtained primarily to give a basis 

 from which the amount of spicules in any mass of Ciorgonians could be com- 

 puted without the necessity of making separate determinations on such large 

 amounts of material as would necessarily be handled in an extended survey 

 of the reefs. In order, however, to have a more reliable check for these 

 computations, a series of determinations of the spicule content of the Gor- 

 gonians were made from a number of squares, each with sides a yard in 

 length. The results of these determinations are shown in table 2. 



In securing the material for these determinations, a square frame made 

 from iron pipe, with an area of one square yard, was thrown to the bottom 

 without any previous observation of the number of Gorgonians there present. 

 After their removal, the colonies were sorted, those of any one species being 

 kept together, and determination was made of the total amount of spicules 

 for each kind. For these observations the only selection made was that in 

 practically all instances the specimen was gathered on portions of the reefs 

 where the water was sufficiently shallow to permit one to wade. 



Squares i to 6 were on the shallow reef west of Loggerhead Key and 

 were scattered along a line some 2 miles in length. Squares 7 to 14 were 

 taken along a line about 0.5 mile in length on the east side of Loggerhead 

 Key. Square 15 was on a reef about 3.5 miles east of Loggerhead Key, 

 where the Gorgonian fauna was very sparse, but was included to make the 



'Only three determinations were carried out on Gorgonia acerosa, for the reasons previously mentioned. 



