80 SMITHSONIAN INSTITUTION 



several of the elements already discussed, for a separate chapter in which its 

 relations to the former matters may also be explained. 



There are in all 200 specimens of the larvae, which were found clinging to 

 the cirri of several young individuals of Comactinia (Actinometra) meridion- 

 alis, a widely distributed littoral species. They show almost every stage, from 

 that of a nearly closed cup of basals and orals with faint incipient radials and 

 a long, tapering stem, through development of anal, radial, and interradial 

 structures, arms, pinnules and cirri, to those of diminution and ultimate re- 

 sorption of orals and interradials and loss of stem. A series of specimens has 

 been selected for figuring to represent the various phases of this development, 

 to which I have added for comparison an adult specimen from another locality 

 (Plates B and C). The minute " pentacrinoids," from .31 to 1.85 mm. in 

 diameter of calyx, were attached to the cirri of the host by the usual flat, 

 encrusting plate (Pis. B, fig. 6; C, fig. 7), similar to that of the fossil Calpi- 

 ocrinus (PI. VII, fig. la) and Taxocrinus (PI. L, fig. 1). Some of them have 

 already been figured by Mr. Clark in his Monograph of the Existing 

 Crinoids — see on p. 317, figures 408 and 412; plate 4, figure 548. 



The outstanding difference in the development of this type from that of 

 Antedon lies in the position, size, and much greater sweep of migration of the 

 radianal, and in the prominence of the anal tube — both evidencing the great 

 importance of the movements of the intestine in producing changes in the form 

 and composition of the calyx, as indicated in my paper of 1906, already quoted. 

 And in these characters the analogy between the larval stages of this living 

 crinoid and the successive phylogenetic modifications of the group Flexibilia is 

 most remarkable. It will be remembered that in W. B. Carpenter's account of 

 the growth of the calyx in Antedon rosaceus the so-called anal plate was first 

 observed to appear " between two of the radials, and at the same level with 

 them." That is the position shown in Sir Wyville Thomson's figure * of its 

 first appearance as observed by him ; and in none of the researches upon species 

 of that genus has this plate been noted as beginning in any different position. 

 In the present material, however, we shall see that it appears in a far more 

 primitive position. 



Our series of Comactinia larvae begins (PL B, fig. 1) with a calyx com- 

 posed almost exclusively of basals and orals, the latter closed and standing 

 quite erect; and a stem with ossicles lengthening distally, on which cirri have 

 not begun to appear. The plates at this stage consist as usual of a thin cal- 

 careous reticulation formed by spicules imbedded in undifferentiated sarcode; 

 later this reticulation progressively consolidates, thickens, and becomes a less 

 and less porous skeletal substance. Very small patches of stereom, which are 

 the beginnings of radials consisting of a few spicules, have appeared in the 



1 Philosophical Transactions Roy. Soc. London, 1865, p. 529, pi. 27, fig. 1. 



