86 SMITHSONIAN INSTITUTION 



causing inhibition of its lateral growth and its peculiar asymmetrical develop- 

 ment (PI. B, fig. 8a). The path of this migration is indicated by the dotted 

 line in figure 6 of Plate C, and the course of the plate is also that of the anal 

 end of the intestine to which it is attached. If on the figure of the primitive 

 Protaxocrinus (PI. XL VI, fig. 8b) a line is traced from the radianal to the end 

 of the tube, it will be found to follow the same course. 



The course of the radianal in the successive stages shown by this series 

 of comatulid larvae is closely paralleled by its position at corresponding stages 

 in the phylogeny of the Paleozoic crinoids, especially the order Flexibilia. 

 Figures I and 2, Plate B, with the radianal at the base of the potential right 

 posterior ray, are analogous to the stage of the Ordovician and Silurian 

 Protaxocrinus, Ichthyocrinus, Clidochirus, etc. ; figures 3 and 4, obliquely be- 

 low and to the left of the radial, to that of the Silurian Lecanocrinns and 

 Gnorimocrinus ; figure 5, in the notch between the right sloping face of the 

 posterior basal and the radial, to that of the Carboniferous Forbesiocrinus and 

 Taxocrinus; figure 7, midway between the two posterior rays and sym- 

 metrically truncating the basal, to that of the so-called anal x in the Devonian 

 and Carboniferous Dactylocrinus and Mespilocrinus; while specimens in the 

 stage of figure 5, Plate C, with all anal structures eliminated from the dorsal 

 cup, represent that of the Carboniferous Nipterocrinus and Wachsinuthicrinus. 

 And throughout the series of larval changes there is to be observed the same 

 tendency of the anal structures to bend to the right which, as I have before 

 pointed out, prevails among the genera of the Flexibilia. Even in the tegmen 

 of the adult there is not perfect symmetry. 



I have called the migrating plate the " radianal," following the course of 

 Mr. Clark already quoted, 1 which was the result of conference between us, 

 because it seems to be perfectly homologous both in position and function with 

 the plate of that name in the Flexibilia and many of the dicyclic Inadunata of 

 the Paleozoic crinoids, at least up to the stage in which it assumes the function 

 of the symmetrically located first plate of the median anal series. We know 

 that in our comatulid larvae this plate is not replaced or changed in any re- 

 spect except by its migration and ultimate resorption, because it is fixed to the 

 gut, and is lifted with it by the upward growth of the tegmen from its point of 

 origin to its latest position. But in the phylogeny of the fossil forms the per- 

 manent fixation of interradial plates in some of them introduced another ele- 

 ment, which would have to be accounted for in the posterior interradius as well 

 as in the regular areas, and an intermingling of characters with the radianal 

 may have resulted. 



It is thus clearly demonstrated that these great morphological changes are 

 due mainly to the growth of the gut and its mode of discharge, phases of which 



1 Jour. Washington Acad. Sci., 1912, p. 310, 311; Monograph of the Existing Crinoids, 1915, p. 332. 



