284 PROCEEDINGS OF THE ACADEMY OP [1819. 



perhaps reproduction. Anal opening rarely observed, evidently- 

 lateral not posterior and low down. 



3. In having the calyx constructed of only three rings of plates 

 alternating with each other. Proximal or underbasal plates some- 

 times imperfectly developed or even wanting ; the plates of this 

 order rarely anchylosed. No interradials, but anal plates gener- 

 ally found within the calyx. 



These are the dominant characters upon which we propose to 

 establish the Cyathocrinidae as a family. 



An attempt to subdivide this family by means of differences in 

 the construction of the anal area and presence or absence of pin- 

 nulre into two groups, represented by Poteriocrinus and Cyatho- 

 crinas, proved unsuccessful. It is true the two typical genera 

 are well distinguished by these characters, and most genera from 

 {he Carboniferous might readily be separated in this wajr ; but it 

 would be very difficult with the Silurian genera, in which the 

 anal area is imperfectly developed. Such a division might pos- 

 sibly be established if we had before us the living types instead 

 of fossils, and it is very likely that the earlier forms, which include 

 HeAerocrinus, Iocrinus, Dendrocrinus, Carabocrinus, Hybocri- 

 nus, and Anomalvcrinus, in which the proximal plates or under- 

 basals are as 3'et very small and scarcely visible, might also be 

 arranged in a division by themselves. So, too, Woodocrinua, 

 Zeacrinus, Hydreionocrinus, and Coeliocrinus are well charac- 

 terized by their enormous balloon-shaped ventral sac; and in like 

 manner Ewpachycrinus, Erisocrimis, and Stemmatocrinus, b} T the 

 dwarfing of the same organ to its minimum size, represent a 

 transition in the direction of the Encrinidae. All these groups 

 apparently have some claim to be classed independently, but in 

 doing so we should rather increase than diminish the difficulties 

 before us, and gain nothing in the end. The genera are so closely 

 linked together, and shade so easily from one to another, and 

 from one group into another, that a satisfactory separation is 

 next to impossible. But, in order to facilitate the comparative 

 study of the genera, we have arranged them sj'stematically in 

 such a manner that allied types ai*e brought together in the order 

 in which they pass into one another, and we are confident that 

 this arrangement will materially assist in the recognition of the 



genera. 



The following is the list: 



