74 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.41. 



The next step is marked by the acquisition of cirri borne by the 

 proximale. This, according to definition, would constitute the for- 

 mation of the first true centro-dorsal. The extreme irregularity' as 

 regards the appearance of these cirri and their variable number point 

 to the conclusion that they are not modified verticils of cirri origi- 

 nally borne by a nodal. In such forms, indeed, I think we have to 

 look among stalked Crinoidea without verticils of cirri upon the 

 column for the ancestors of the free types. The cirri in such cases I 

 conceive to have been evolved in direct response to the needs of an 

 eleutherozoic habit. At the same time I do not believe that such 

 cirri could have been evolved were it not for the fact that within 

 this genetic line there was a tendency toward the acquisition of such 

 organs. These cirri appearing first in the free-swimming stages per- 

 haps are progressively found earlier and earlier in the ontogeny of the 

 animals until, no doubt, cirri were borne at an early stalked stage in 

 the development of the organ. 



Subsequent to the acquisition of cirri the changes in the gross 

 structure of the crinoid skeleton are of a less fundamental nature. 

 Detachment from the column takes place at progressively earlier 

 ontogenetic stages, as does the appearance of the cirri. There is, 

 moreover, a constant tendency toward an increase in the number of 

 cirri. The infrabasals continually decrease in size and importance, 

 and a reduction in the size of the basals is to be noted. The basals, 

 moreover, unite to form a ''rosette" and do not appear externally. 

 Naturally there are also changes m the musculature of the arms 

 and in the type of union between the ossicles. The degree of special- 

 ization is in direct proportion to the length of time that has elapsed 

 since the assumption of an eleutherozoic existence. 



PJiylogenetic steps in evolution of centro-dorsal. — The steps in the 

 evolution of the centro-dorsal as briefly outlined above find expres- 

 sion in well-known fossil genera. These genera may not represent 

 actual links in an evolutionary chain, but they have the character 

 that such ancestral forms undoubtedly possessed. Indeed, it is by 

 no means certain that these genera did not give rise to successful 

 comatulid lines. If they themselves did not, forms not widely unre- 

 lated may be held to constitute the necessary ancestral types. 



One may well, I think, start with such a form as the Cretaceous 

 Mesocrinus, which has been described and figured by Carpenter (1881). 

 This genus apparently has a fairly well-developed proximale, which 

 may or may not be fused with the infrabasals. It is inserted well 

 up in the cup formed by the basal circlet. The proximale is small, 

 however, and bears no cirri. The columnals in the upper portion 

 of the stem are circular in section, but the distal ossicles have oval 

 articular faces. The column evidently pertains to the Bourgue- 

 ticrinidse. The stem in its distal portion is comparatively heavy and 



