50 BULLETIN 82, UNITED STATES NATIONAL MUSEUM 



pinnules. He said further that this assumption of the extension of the gonads under 

 the interradii being the chief cause of the enlargement of the perisome makes it clear 

 also why in young comatulids we do not get a high disk in connection with the pinnule 

 gap, for at this stage the gonads are wholly rudimentary. 



Dr. Gisl6n's explanation of the pinnule gap is most interesting and ingenious and 

 certainly is worthy of serious consideration. But the author still maintains his 

 belief in his own interpretation of the conditions in the early pentacrinoids. There is, 

 of course, the possibility that both of us are in error in ascribing the pinnule gaps in 

 the recent crinoids to factors having their origin in phylogenetical antecedents. These 

 gaps may well be nothing more than a response to the conditions of environment with 

 no phylogenetical background at all. 



In the case of the developing young of the comatulids the prune requisite is to 

 secure food as early as possible. The larger the diameter of the circle combed by the 

 food-collecting organs, the larger is the potential food supply. So pinnules first 

 appear at the tips of the growing arms, where it is obvious they will be most effective. 

 With the commencement of concentration of food upon the disk, resulting from the 

 action of the pinnules at the arm tips, the protection of the disk becomes increasingly 

 important, so that it would be natural to expect the tactile pinnules at the edge of the 

 disk to put in an appearance. The need for the intermediate pinnules, which are 

 either genital or supplementary tactile pinnules, or both, does not arise until later, 

 so that their development is delayed in favor of that of the more important pinnules. 



The absence of a varying number of basal pinnules in such types as the Atele- 

 crinidae and Bathycrinidae is probably due to the fact that, as a result of the crowding 

 of the arm bases, there is no room for them. If present they would be without value 

 to the animal, and so they are not developed. Such an explanation is easy to com- 

 prehend if pinnules are regarded in the light of articulated appendages, but is less 

 easy of comprehension if they are considered as reduced arm ramifications. 



INTERRELATIONSHIPS OF THE COMATULIDS 



Gislen's rearrangement of the comatulid groups was given in detail in Part 3, 

 pages 60-64. His notes on the evolution and interrelationships of the comatulids 

 are of much interest, for his was the first serious attempt to harmonize in detail the 

 systematic background of the recent and the fossil forms. 



The evolution of certain comatulid characters. Gislen said that originally cirri 

 were arranged in five radial columns. It was not until the suppression of the inter- 

 nodals in the pentacrinite stem and the increased shortening of the nodals took place 

 that two alternating columns occurred, to conserve space, in each radius. Since 

 after the development of the centrodorsal the number of the cirri continued to increase 

 a third column appeared, and finally the arrangement became a series of closely 

 crowded irregular alternating rows. This developmental course may be traced 

 more or less distinctly in the ontogeny of certain comatulids. Centrodorsals with 

 closely crowded alternating rows of cirri are therefore the most highly specialized. 



Originally the cirrus sockets had a distinct sculpture, but this has disappeared 

 more or less completely among the younger forms. 



The cavity within the centrodorsal was originally, as all the fossil comatulids 

 show, from fairly small to very inconspicuous, never larger than in the corresponding 



