318 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



and Ptilocrinus (fig. 144, p. 207); they may be immensely elongated, as in Demo- 

 crinus (fig. 133, p. 203); they may be turned inward so that they come to lie more 

 or less parallel to the dorsoventral axis and fused into a solid conical or subcylin- 

 drical ring or plate, as hi Rhizocrinus, Bathycrinus , and Monachocrinus (fig. 134, p. 

 203, they may be turned outward so that they lie flat and form a platform upon 

 which the radials and the calyx rest, as in the pentacrinites ; or they may be entirely 

 metamorphosed so that they come to form an internal septum, as in the great 

 majority of the comatulids. 



In the progressive specialization and perfection of the phylogenetic line ter- 

 minating in the comatulids and the pentacrinites the chief factor involved is the 

 progressive reduction and strengthening of the calyx. First the subradial and 

 interradial plates dwindle and disappear, persisting longest in the posterior inter- 

 radius and beneath the right posterior ray; next the infrabasals become affected, 

 decreasing in size and often also in number, gradually leaning outward and con- 

 tinually decreasing the diameter of their circlet until they become quite negligible 

 as integral parts of the skeletal system, when they fuse with the proximale or 

 disappear altogether; after the infrabasals the basals become affected, in their 

 degeneration following much the same path as that previously taken by the infra- 

 basals; they decrease in size and often become reduced to three, at the same time 

 either gradually leaning outward so that they ultimately form a small platform 

 upon which the radials and the visceral mass rest and finally, through a curious 

 process of metamorphosis, passing around the dorsal nerves and reappearing as a 

 thin septum between the dorsal nervous mass and the visceral cavity, or gradually 

 leaning inward and fusing so that they form a truncated conical plate or ring which 

 is in effect nothing more than a first columnal. 



Among the recent comatulids the genera Atelecrinus (figs. 123, p. 192, 124, 

 125, p. 193, 414, p. 319, and 573, pi. 8, and Ato,pocrinus (fig. 227, p. 245) are the 

 only ones in which the basals persist as basals instead of becoming metamorphosed 

 into a rosette. In the species of Atelecrinus, excepting only in A. anomalies (fig. 

 414, p. 319), in which they are still very large, the basals have become arrested in 

 their specialization so that in the adults they are at approximately the same onto- 

 genetical stage as are those of Antedon at the tune of the beginning of the free exist- 

 ence (fig. 594, pi. 16), or as are those of the pentacrinites. As described by Car- 

 enter ' ' they are in complete contact laterally so as to form an unbroken ring about 

 the central opening of the calyx" which is "encroached upon by excessively delicate 

 processes that project inward from near the lateral margin of each basal." These 

 delicate processes may possibly represent the partially resorbed infrabasals. 



Carpenter notes that in the young Atelecrinus balanoides (fig. 573, pi. 8) the 

 basals externally "form a kind of belt of tolerably uniform height with the inter- 

 radial angles somewhat produced which everywhere separates the * * * 

 radials from the centrodorsal. " He notes further that "the extent of development 

 of the basals varies with the size of the individual, apparently diminishing with 

 age. * * * In the smallest specimen they are wide but low pentagons which 

 fall away very rapidly from their interradial apices to the points where they meet 

 one another beneath the radials. The middle of each basal rests on the top of one 



