ORAL AND APICAL SYSTEMS OF THE ECHINODERMS. 377 



no trace of anything representing a stem, not even its upper- 

 most segment which indicates its presence in Comatula. 

 The problematical Crinoid Cyathidium from the chalk of 

 Fctsoe seems to have resembled Holopus in this respect, and 

 to have had a spreading base of attachment immediately 

 beneath the cup. In these two cases we have, I believe, 

 the nearest approach in the Crinoids to the embryonic con- 

 dition of the apical system in the other Echinoderms. 



I do not think that the extremely irregular shape of the 

 spreading base of attachment of Holopus, Apiocrinus, and 

 other Crinoids is any serious objection to the view advanced 

 above. In the Pentacrinoid larva of Comatula the ter- 

 minal plate has at first a definite circular form (figs, viii 

 and IX, 1), but gradually increases both in diameter and 

 thickness, absorbing into itself (as it were) nearly the whole 

 of the organic substance of the basal disk, while its margin 

 usually becomes more or less deeply divided into lobes. 

 We do not know the condition of the original base of 

 the stem in Pentacrinus, but in Apiocrinus the terminal 

 plate and much of the lower part of the stem are sur- 

 rounded by a thick extraneous deposit of calcareous matter, 

 forming an irregular cone which would give a very firm 

 support to the animal. I imagine the irregularly ex- 

 panded character of the base of attachment of Holopus to 

 be due to a similar secondary deposit, which has obscured 

 its originally symmetrical shape, in the same way as the 

 partial resorption of the subanal plate in the Echinids 

 causes its primitive pentagonal symmetry to become less 

 marked. In neither case is it fair to regard the condition of 

 the part in the adult animal as an argument against any 

 views of its homology which may be derived from embryo- 

 logical considerations. 



In order to make clear the reasons which have led me 

 to the view above suggested, I must go back to a very early 

 period in Echinoderm development, namely, to the Gastrula 

 stage. If Gotte's figures of the Gastrula of Comatula be 

 compared with those of other Echinoderm Gastrulse as given 

 by other observers, a very striking difference is apparent 

 between them. The Gastrula of Aster acanthion, according 

 to Agassiz,^ is somewhat "pear-shaped, with rounded ex- 

 tremities, having at one end an opening leading into a 

 pouch which extends half the length of the cylinder " (fig. 

 X, a). The Toxopneustes Gastrula closely resembles that of 

 Aster acaiithion, except that its transverse axis is longer in 

 proportion to the longitudinal one, i.e., to the one occupied 

 * ' North American Starfishes,' loc. cit., p, 9. 



