MONOGEAPH OF THE EXISTING CKINOIDS. 377 



served, and it therefore becomes essential, if we would arrive at a true knowledge 

 of the systematic position and interrelationslui)3 of these fossil species, to devote 

 particular attention to the same sti-uctures in the recent forms. 



The surface of each radial typically shows five small rounded openings leading 

 into the interior; one of these, referred to previously as the central canal (figs. 9-11, 

 p. 65, 431, 432, p. 349, and 439, 440, p. 351), is on the articular face just above the 

 middle of the transverse ridge; there is one on either side near the dorsal inner 

 angle (fig. 600, pi. 17), and there is a pair (sometimes united into a single one) 

 at or near the inner margin of the dorsal surface (figs. 443, p. 351, and 600, pi. 17). 

 These openings serve for the passage of the chief cords of the dorsal nervous 

 system. 



In the comatulids these cords lie usually just within the inner surface of the 

 radials, or they may even be on the surface so that they are not covered, except 

 in part, by calcareous deposit. In the Pentacrinitidse, however, they lie well 

 within the calcareous substance of the plate so that their course within the segment, 

 wliich is the same in the pentacrinites as in the comatulids, may be much more 

 readily made out. 



Each radial receives a branch from the two adjacent interradial nerve cords 

 which arise from the central capsule (figs. 63, 64, p. 89) ; these two branches enter 

 through the two apertures at or near the inner margin of the dorsal surface (fig. 

 600, pi. 17); within the radial they gradually converge, meeting and fusing just 

 \vithin the opening of the central canal on the articular face. From this point of 

 union of the two derivatives of the prunary interradial nerve cords a branch is 

 given off laterally to either side which passes through the apertures near the dorsal 

 inner angle and continues through the adjacent radial to the correspondmg posi- 

 tion within it. These connectives thus form a circular commissure all around the 

 calyx, as will be further explained when the nervous system is considered. 



In many species, particularly among the Comasteridas, Charitometridae, Tha- 

 lassometridse and Zenometrinae, deep subradial clefts are found extending inward 

 between the dorsal surface of the radials and the ventral surface of the ccntro- 

 dorsal (figs. 166-169, p. 229, 194, p. 237, 203, p. 239, and 208-216, p. 241). These 

 clefts are narrow and slitUke externally, but are more spacious interiorly. They 

 are bounded laterally by the basal rays and the ridges in which tlicso rays lie, 

 and inwardly by a wall formed by the close apposition of the small jioavily calci- 

 fied bars which form the tliickened edges of the inner part of the doreal faces of 

 the radials and the iimer part of the ventral surface of the centrodoreal. There 

 is thus no connection whatever between the subradial clefts and the body cavity 

 of the animal, nor are the five clefts at all connected vnth each other. They are 

 in all respects, as stated by P. H. Carpenter, similar to the so-called interarticular 

 pores seen in the stems of the pentacrinites (fig. 127, p. 197, in upper part of stem). 



The amount of concealment of the radials by the centrodorsal is, of course, 

 in direct proportion to the comparative size of that structure. In most s]>ecies 

 the radials extend to the ventral rim of the centrodorsal, or slightly beyond it. 

 When the centrodorsal is reduced in size more of the surface of the radials is shown, 



70146°— Bull. 82—15 25 



