288 PROCEEDINGS OF THE ACADEMY OF [1885. 



The disk ambulacra were probably differently constructed from 

 those of the arms (PI. 4, figs. 6 ami 7). The specimens indicate 

 that the plates of the former were suturallj r connected, while 

 those :ilong the arms were movable. All Cyathocrinidae, so far 



as observed, have side-pieces which support Saumpliittchen ; and 

 these rest upon two series of subambulacral (?) plates, which form 

 the floor of a tube as in the Actinocrinidse. 



The '' ventral sac " of the Pistulata was always regarded by us 

 as functionally and structurally distinct from the "anal tube" or 

 " proboscis " of the Camarata. We held the former to be an 

 essential part of the body, and perisomic in its origin; the latter 

 as a mere prolongation of the azygous interradius, and con- 

 structed of abactinal plates. 



To understand the two structures, we must bear in mind that 

 in the growing Actinocrinoid the capacity of the calyx adequately 

 increased with the growth of the body, and hence was at any 

 time capable of holding the visceral mass. In the Fistulata, 

 however, in which all brachials remain permanently free, and the 

 calyx is not enlarged in proportion to the visceral mass, the pos- 

 terior side of the disk forced its wa3' out through the anal open- 

 in--, and formed the so-called ventral sac, which has always a 

 narrow neck along the base. According to our interpretation the 

 ventral sac is an enormously developed interpalmar area supported 

 by the anal plate, and as such reminds us of the asymmetrical 

 disk in the recent genus Aetinonometra, in which the anus is 

 central and the mouth marginal. 



In most of the Fistulata, the ventral sac is perforated with 

 round or slit-like openings, transverse^ arranged, which enter 

 the outer mar-ins of two adjoining pieces, but never penetrate 

 the inner portions of the plates like the water pores of the 

 Neocrinoidea. The openings either extend over the whole surface 

 of the sac; or are arranged in Longitudinal rows — porous plates 

 alternating with solid ones; — or the terminal end is composed of 

 large solid pieces, frequently spiniferous ; or as in the Carboni- 

 ferous species of Cyatkocrinus the entire tube is composed of solid 

 hexagonal plate-, and the porous or anambulacral plates are 

 restricted to the small area usually occupied by the smaller 

 proximals. In the Poteriocrinidse, the anambulacral plates 

 extend over the greater part of the ventral sac. but in the 

 Catillocrinidse and Calceocrinkhe they are limited to one side of 



