1885.] NATURAL SCIENCES OF PHILADELPHIA. 287 



onstrating that a pliable vault may enclose another flexible 

 integument and contain the food grooves underneath, which was 

 seriously questioned by Carpenter (Chall. Rep., p. 182). He 

 evidently overlooked Crotalocrinus, for we doubt if he could 

 have taken the small covering plates (Icongr.,Pl. 17, fig. 3 a) for 

 the representatives of the large rigid plates of figs. 6 and 7 on PI. G, 

 or the irregular pieces around the oral pole to be summit plates. 



Crotalocrinus and Enallocrinus have close affinities with the 

 Ichthyocrinidae, not only in that both have a flexible skeleton, 

 but the}' frequently possess no interradials dorsally, and they all 

 have the same peculiar arm structure. In speaking of a pliant 

 vault we do not mean a surface " formed of connective tissue with 

 numerous interradial plates imbedded in it," as supposed by 

 Carpenter (Chall. Rep., p, 182), but a continuous integument of 

 plates connected by ligament in place of suture, sometimes with 

 imbricating plates. We postulated the prevalence of this structure 

 in the vault of the Ichthyocrinidae from the construction of the 

 dorsal plates 2 which could not be movable unless the ventral side 

 was pliant also. Our views are confirmed by the vault structure 

 of Crotalocrinus, and we think the disk ambulacra of Ichthyo- 

 crinus were arranged in a similar manner, and covered by a 

 similar vault. 



A very different perisome is found in the higher types of the 

 Cyathocrinidse, which is not subtegminal, but exposed upon the 

 surface of the interradial plates. This form is found only in 

 genera in which the ambulacral tubes rest upon the upper edges 

 of the interradials. It is not restricted alone to the later genera, 

 but occurs in several Silurian forms. Angelin has figured such 

 a disk in Cyathocrinus Isevis (Iconogr., PL 26, figs. 2 and 3), and 

 Gissocrinus punctuosus (ibid., PI. 29, fig. 75 d), but we think 

 the structure was not correctly understood. In all cases the five 

 interradial plates are completely covered by small perisomic 

 plates, of which those at the four regular sides are not pierced 

 with water pores, while those toward the ventral sac are generally 

 profusely perforated. In some cases we found the summit plates 

 in process of resorption. In Cyathocrinus iovensis (PI. 5, fig. 7), 

 the larger proximals appear in the form of eight irregular pieces, 

 their edges rounded off; while in Cyathocrinus multibrachiatus 

 (PI. 4, fig. 6) only fragments of the plates are scattered over the 

 perisome. 



