c. The girdle divides two similar parts which are spherical in outline 

 and are ornamented with symmetrically disposed verrucae. Such an 



arrangement results in the spicule which we will call the. . . . Double Head. 



d. The girdle divides two similar parts which are spherical in outline 



and ornamented with radiating points. These are Double Stars. 



e. The spicule is rod-like in general form, with a girdle dividing two 

 parts which are gradually enlarged toward the outer ends, have 

 comparatively straight sides, rounded distal corners and are covered 



with fine and densely aggregated verrucse. These will be called . Double Bars. 

 /. The spicule is rod-like, without the girdle Bars. 



2 nd . Unsymmetrical forms. 



a. The girdle separates two unlike parts, one a club and the other a 



star. Such forms may be called '. . .. . Club-Stars. 



b. The girdle separates two unlike parts, one a club and the other 



a head. Forms of this kind may be called Club-Heads. 



c. The girdle separates two unlike parts, one a star and the other a 



head. These will be called Star-Heads. 



d. The girdle separates two unlike parts, one a club and the other a 



spindle. This rather rare form may be called Club-Spindle. 



e. The girdle separates two unlike parts, one a star and the other a 



spindle. These may be designed as Star-Spindles. 



Besides the above, which may be called girdled forms, there are often crosses produced 

 by a double head being longitudinally divided by an impressed vertical zone. 



While it is true that a given species, or even a single specimen, may show several of 

 these forms and numerous intergradations, it is also true that certain forms predominate in a 

 given genus, and it is these dominant types of spicules, and not the others, that are available 

 for generic diagnosis ] . 



While the spicules are the most important features for generic distinctions, we may also 

 avail ourselves of any other character which seems to be possessed by a group of allied 

 species, e. g. mode of branching, shape of calyces, character of axis, etc., in attacking the 

 problem before us. 



Artifical key to the genera of the GORGONELLID^;. 



Main branches arising from a forking of the main stem. Colony flabellate, 

 widely diverging, the main branches bearing a series of simple branchlets 

 on upper side only. Branchlets vertical and parallel, and resembling the 

 teeth of a comb Ctenocella 



1 See a discussion on this point, as well as a statement of the position of the author on the matter of the retention of established 

 generic names, in the Report on the Muriceidas of the Siboga Expedition, NUTTING, 1910, p. 5. 



