NATURAL SCIENCES OF PHILADELPHIA. 325 



have been, during the life of the animal, two distinct openings in the vault, as 

 appears to be the case in the specimen of G. planus, Miller, figured by Prof. 

 Phillips and Mr. Austin. But on examining the specimen of C. lowensis men- 

 tioned above, we find that it shows the base of the small lateral proboscis, with 

 the five principal vault-pieces alternating with the first radials (the one on 

 the anal side being larger than the others), and the same ambulacral furrows 

 extending inwards from the arm- bases, all exactly as in the C. malvaceus. But 

 here we find the central opening undoubtedly closed hj several vault pieces, 

 while the ambulacral furrows, extending inward from the arm-bases, pass 

 in under these central pieces, and are themselves occupied, or covered, by a 

 double series of alternating, very minute pieces, which probably also extend 

 on, all the way up the ambulacral furrows of the arms as marginal pieces. 



From our examinations of these two specimens, which are the only examples 

 of the genus we have seen, showing the vault pieces, and seem to be typical 

 forms of the genus in all other respects, we are strongly inclined to think the 

 specimen of G. planus, figured by Prof. Phillips and Mr. Austin, has had these 

 central vault pieces removed by some accident. The fact that these pieces in 

 the specimen examined bj' us, in Mr. Wachsmuth's collection, seem not to be 

 deeply implanted between the five larger surrounding pieces mentioned, but 

 rather rest, as it were, partly upon the narrow bevelled points of the inner ends 

 of the latter, between the ambulacral furrows, so as t(3 allow room for these 

 furrows to pass under, would render them less firm, and more liable to be re- 

 moved by any accident, and may possibly account for their absence in the 

 English specimen mentioned. 



In regard to the pieces covering the central part of the vault, and which, 

 from the way they are arranged for the ambulacral furrows to pass under them, 

 were apparently more liable to be removed than the others, we would remark 

 that they do not present the prominent appearance, and uniformity of size and 

 form, of the movable pieces composing what is often called the ovarian pyra- 

 mid in the Cystids, but certainly have all the appearances of true fixed vault 

 pieces, and scarcely project above the others surrounding them. Consequently 

 we cannot believe it at all probable that this genus had a central mouth, open- 

 ing directly through the vault ; though its ambulacral canals evidently con- 

 verged from the arm-bases to the middle of the vault, partly above the outer 

 vault pieces, and under those composing the middle of the vault. That these 

 furrows terminated at the entrance of the alimentary canal, under the middle 

 of the vault, as those of Comatula converge to the mouth, in the same central 

 position, is highly probable ; and, as will be seen further on, we are much in- 

 clined to believe that the minute organisms upon which we are led, from ana- 

 logy, to think these animals subsisted, were conveyed to the entrance of the 

 alimentary canal along the ambulacral furrows, without the agency of any 

 proper mouth, opening directly through the vault. Hence we think it proba- 

 ble that the small tube, usually called the proboscis, situated near the poste- 

 rior side of the ventral disc, rather corresponds to the tubular anal opening 

 similarly situated in Comatula iMediterranea. 



From our description of the vault of these species, it will be seen to present 

 considerable similarity to that of Crotalocrinus rugosus, excepting that in that 

 genus, owing to its great number of arms, the ambulacral furrows, or canals, 

 bifurcate several times between the middle of the vault and the arm-bases, 

 while in Crotalocrinus there is no lateral proboscis, nor, apparently, even any 

 visible opening, judging by the figures we have seen, though we suspect it may 

 have a small opening at the periphery of the veutrnl disc, on the posterior or 

 anal side. In the group of depressed Platycrini for which Troost proposed the 

 name Cupellsecrinus we observe a somewhat similar vault, at least in some of the 

 species; also in Coccocnnus. In such forms there would seem to be, as it were, 

 an intermediate gradation between the modern Crinoids and the prevailing 

 PalcEozoic types, as has been pointed out by Mr. Billings. 



4. Convoluted support of the digestive sack, in the Actinocrinidx. The presence 



1868,] 



