194 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



there are no oblong proximal brachials at the base of the arm, all the brachials 

 beyond the second being triangular. The arms in Atelecrinus are unusually long, 

 and apparently they assume from the first the attributes of true comatulid arms, 

 at least the triangular brachials and the regular distribution of the syzygies, not 

 passing through a stage in which they are comparable to the arms of the Flexibilia 

 Impinnata. The arms of Atopocnnus have never been observed, but, except for 

 the absence of the IBr series, they do not appear to differ from the arms of 

 Atelecrinus. This acceleration of the development of the arms in Atelecrinus is 

 correlated with a retardation of the development of the pinnules before the twelfth 

 brachial, which are always suppressed, and with a retardation of the meta- 

 morphosis of the basals, which remain visible externally. In Atopocrinus it is 

 correlated with the insertion of a supernumerary brachial between the first 

 brachial and first syzygial pairs and the absence of the IBr series. 



In Thaumatocrinus (part 1, fig. 113, p. 181) and in Pentametrocrinus (part 

 1, figs. 119, p. 185; 120, p. 187; and 121, p. 189) the arms are greatly elongated, as 

 in Atelecrinus. This excess of arm growth is correlated with the absence of the 

 IBr series and with the insertion of a supernumerary brachial between the first 

 and second brachial pairs as in Atopocrinus, and with the frequent absence of P^, 

 suggesting the condition found in Atelecrinus. 



In all of these genera the intersyzygial interval is unusually long, and ail 

 of the syzygies are evenly spaced, agreeing with the high state of specialization 

 of the arms in other ways. 



Pinnules. — In the form and number of the component segments and in the type 

 of articulation between them the oral pinnules of such types as Analcidometra 

 and Himerometra, and to a slightly lesser A&gvQQ Anted on, show a striking simi- 

 larity to the arms of certain of the Flexibilia Impinnata, while at the same time 

 they are equally similar to the arms of the pentacrinoid before the appearance of 

 the first pinnules. The correspondence is so complete that we are probably safe 

 in considering them as strictly comparable structures (compare figs. 270 and 271, 

 p. 207, and 1207, 1208, and 1210. pi. 34: also figs. 273, p. 207, and 1214, pi. 34). 



The oral pinnules therefore are arm branches which have failed to assume the 

 adult character, but instead have remained in the condition of the main arm at 

 the time of their first appearance; that is to say, they are Flexibilia arms borne 

 upon a comatulid arm trunk. 



We have already seen that in 20-armed species the outer arms arising from 

 each IIBr series bear more specialized pinnules than the inner, which indicates 

 that they are phylogenetically more developed; in other words, of later phylo- 

 genetic origin. In most types, though the outer arms are more developed than 

 the inner, they are developed simultaneously with the latter; but in Comatula 

 etheridgei (part 1, fig. 78, p. 131) they actually appear much later and never 

 reach more than a fraction of their length. The first oral pinnule (PJ develops 

 upon the ossicle which in species with the division series of two ossicles each 

 becomes the axillary of the IIBr series and which in Comatula etheridgei bears 

 the short arm exteriorly. In 20-armed species this ossicle is, together with the 

 larval arm beyond it, thrown off and a new element is regenerated bearing two 



