114 PROCEEDINGS OF THE NATIONAL MUSEUM. yol. 41. 



in all probability the opposite process has obtained. At times, no 

 doubt, arm branches have become reduced in size forming ramules 

 and even ultimately what might be styled pinnules. However, I 

 do not think that this has been the normal process. 



As has been elsewhere noted, certain fossil Crinoidea may be held 

 to have led an epi-planktonic existence, being attached to floating 

 wood. Edriocrinus may be added to the list of epi-planktonic 

 types. The species involved is probably E. dispansus, a fairly young 

 specimen of which has been found attached to the float of a ScypJio- 

 crinus. The Edriocrinus was cemented to the lower surface of the 

 bulb, and inasmuch as these floats are almost invariably found in the 

 rocks with the lower surface down, it seems probable that attach- 

 ment was had while the ScypJiocrinus was alive and freely floating. 



Cotylederma, Eudesicrinus , and Holopus. — For comparison with 

 Edriocrinus figures are here given of the curious genera Cotylederma, 

 Eudesicrinus, and Holopus (PL 11, figs. 3-5, 6-8, 10). It will be 

 noted in all of these forms that the crinoid is cemented to the bottom 

 by a solid calcerous disk. No infrabasals or basals are to be observed. 

 The origin of these forms, as in the case of Edriocrinus, is more or less 

 obscure. Likewise the composition of the base by which the crinoid 

 is cemented is a matter of conjecture. It is only I tliink by tracing 

 the possible derivation of these crinoids that we may get a clue as to 

 the composition of the peduncle. It is not essential in tracing the 

 evolution of these forms that we establish the exact genetic lines. 

 Rather we should attempt to establish the evolution of the structural 

 features that might evolve such a type of base. 



Origin of pedunculate type of structure. — The evolution of these 

 forms may perhaps be held to be somewhat as follows: Having a 

 stalked ancestor with a well differentiated proximale, detachment 

 becomes eft'ective immediately beneath this plate. For a time a 

 free-swimming or possibly crawling existence is maintained. Sub- 

 sequently reattachment takes place. Tliis is accomplished by the 

 cementation of the proximale to some object on the sea bottom. If 

 such be held to be the origin of the peduncle we may consider it to 

 have essentially the composition of a centro-dorsal, barring the lack 

 of cirri. 



The reason for refixation in these forms is not altogether clear. 

 As is elsewhere noted there is considerable evidence to prove that 

 alternate periods of freedom and fixation frequently obtain among 

 the Pelmatozoa. In these genera tliis condition may be expressed 

 in a single period of freedom foUowed by permanent fixation through- 

 out the subsequent life of the organism. The reason for the lack 

 of maintenance of an eleutherozoic habit may not certainly be stated 

 although there are several possible causes. It may be that the 

 structure of the arms in these genera is not such as to enable them to 



