139 



Cyathocrinites, which I hesitated to point out before I had proved the non- 

 existence of the column in Marsupites. If we consider the lower pentangular 

 plate in this genus (which is similar in position to that in Comatula) to serve 

 in both as a keystone to their vaulted cup, or as a plate performing in part the 

 office of a first columnar joint, (since the other plates adhere to its edges) and 

 thus apply the anatomical nomenclature used when describing the Cyatho- 

 crinites, we shall find the five plates forming the pelvis in that genus truly 

 conformable to the five costals in Marsupites, differing only in their upper 

 edges being bent, to form as its were a concave rim, and the lower surface 

 being excavated to give articulation to the column. The five intercostal plates 

 in Marsupites correspond in figure and marking with the five costals in 

 Cyathocrinites, and the scapulae, integument, and commencement of the arms 

 exhibit a conformity conspicuous at first sight. It is truly pleasing by 

 this change of anatomical momenclature, to point out in a more striking 

 manner the method in which nature effects the transition from one being to 

 another. In the Comatula, the absence of the column showed the first link 

 from the Crinoidea articulata to the Stelleridae, in the Marsupite a similar link 

 is offered from the Crinoidea inarticulata to that family. 



