106 MR. P. H. CAEPENTEE ON THE GENUS ACTINOMETRA. 



whole length of their sides, so as entirely to cut off the radial pentagon from the top 

 stem-segment, just as in S. Jcegeri. 



There can therefore he little douht that the hasals of Pentacrinus are homologous with 

 those of Solanocrinus, and therefore analogous to the compound hasals of Actinometra, 

 which, as we have seen, are not entirely developed out of the embryonic basal plates. 

 It would seem, in fact, as if in Pentacrinus and Solanocrinus the embryonic basal 

 plates became directly transformed into the basals of the adult ; while in Comatula 

 they undergo metamorphosis into the central rosette by the absorption of the greater 

 portion of their dorsal or primary tissue, and the development of a secondary ossification 

 on the ventral side of the original plates. 



In Ant. rosacea the metamorphosis is much more complete than in most Antedons 

 and in Actinometra, in which new skeletal elements are developed by a more or less 

 complete tertiary ossification in masses of connective tissue, that correspond precisely in 

 position, and to a certain extent also in shape, with the basals of Solanocrinus and Penta- 

 crinus. The latter being most probably direct products of the growth of the embryonic 

 basals are therefore strictly homologous only with the rosette of Actinometra, although 

 analogous in position to the whole circlet of compound basals in this genus, viz. to 

 the rosette and basal star taken together. 



(§ 72) It is interesting to observe the different position of the basals with regard to 

 the chambered organ in Comatula and in the various species of the stalked Crinoids. 



In Comatula this organ is situated in the cavity of the centrodorsal piece (PL VIII. 

 fig. 3) which is on the dorsal side, not only of the radial pentagon, but also of the rosette 

 or metamorphosed basals; and the nervous cords proceeding from its fibrillar envelope 

 to enter the central canals of the first radials come off from its ventral angles. 



The large centrodorsal piece of Comatula is developed by the growth of the top stem- 

 segment of the Pentacrinoid larva. In Pentacrinus, which remains pedunculate 

 throughout life, the top stem-segment is the youngest and smallest. Its central 

 cavity is far too small to contain the quinquelocular organ forming the upper end of the 

 central axis of the stem, which contains five longitudinal chambers expanding slightly 

 at every nodal segment, where each of them gives off a single cirrhus-vessel 1 . 



There is no special increase in the diameter of these chambers in the top stem-seg- 

 ments, and they do not expand into the large chambers of the quinquelocular organ 

 until near the level of the ventral surface of the basal circlet which surrounds the 

 dorsal half of the chambered organ. The ventral portion of the latter is contained in the 

 lower part of the central funnel-shaped space enclosed within the radial pentagon, where 

 it is surrounded by a very dense calcareous network, through which the axial prolongation 

 containing the superior continuations of the five chambers of the quinquelocular organ 

 ascends, on its way to enter the visceral mass, just as in Comatula. In consequence of 

 this relatively higher position of the chambered organ in Pentacrinus than in Comatula, 

 the nervous cords which enter the central canals of the first radials come off from its 

 dorsal angles, and not from the ventral ones as in Comatula. 



Iu Comatula, therefore, the walls and floor of the cavity enclosing the chambered organ 



1 Pentacrinus and Bhizocrinus, loe. eit. pp. 43-40. 



