16 BULLETIN 52, UNITED STATES NATIONAL MUSEUM. 



In the following pages theie will be found much speculation in regard to the 

 hypothetical ancestor of the crinoids and of the echinoderrns, based upon a study 

 of each of the various systems which, when taken together, make up the crinoid 

 or echinodcrm whole, and a figure of the hypothetical ancestor will be found 

 embodying all the data acquired from this study. It is well, perhaps, to emphasize 

 the fact that no claim is made that such a creature ever existed; we see in all the 

 echinoderrns to-day most perplexing combinations of primitive and highly special- 

 ized characters, associated in all sorts of different ways, and this leads us naturally, 

 as I have already stated, to the assumption that there was no definite intergrade 

 between the echinoderrns and the barnacles, but that the former sprang from the 

 latter (or, more strictly speaking, from the same phylogenetic line which can be 

 traced by easy stages to the latter) by a broad saltation in which the assumption 

 of the free habit and the correlated assumption of the pentaradiate symmetry 

 combined to render the existence of intermediate types impossible, while at the 

 same time it caused the formation by the echinoderrns, at the very moment of their 

 origin, of two widely diverse stocks, the heteroradiate, including the Pelmatozoa, 

 the Echinoidea, and the Holothuroidea, and the astroradiate, including the Asteroidea 

 and the Ophiuroidea, between which there are, and can be, no intergrades. 



The comatulids must therefore be considered as a biologically extremely com- 

 plex and mixed group in which each organ and structure occurs in a single series 

 all the way from a primitive to a highly specialized type, but in which the various 

 degrees of specialization of each organ or structure, in other words, the progressive 

 steps in the series, as not in any way correlated with species or with genera, or with 

 the comparable degrees of specialization of any other organ or structure. 



Thus it is at once evident that there is a most extraordinary uniformity 

 throughout all the comatulid families and genera, and that each is potentially on 

 essentially the same phylogenetic plane as are all of the others. 



The comatulids as a group are exactly parallel and comparable to the penta- 

 crinites as a group; they are descended from the same ancestral stock and represent 

 exactly the same phylogenetic stage, but during their development they have 

 diverged from their phylogenetic mean in exactly the opposite direction. The 

 pentacrinites have departed widely from their prototypes by enormously increasing 

 the length of the column and at the same tune indefinitely reduplicating the cirrifer- 

 ous proximale, a departure which has to a considerable degree lessened the mobility 

 of the crown, this being hi part compensated by a corresponding increase in the length 

 of the arms; while the comatulids have departed just as widely by compressing 

 what is virtually the entire column of the pentacrinites within the compass of the 

 single proximate or nodal from which numerous cirri are extruded, fixation by these 

 cirri reducing the possibility of motion by the crown to a minimum so that under 

 ordinary conditions the animals are almost as firmly attached as is Eolopus. 



As the greater part of the enormously elongated stem of the pentacrinites lies 

 on the sea floor and therefore becomes neutral in its relation to the mechanics of the 

 animals, these forms do not exhibit any very radical departure from a more gener- 

 alized type, such differences as they show being chiefly the result of the very large 

 size of the crown and arms correlated with a reduction in size of the calyx; nor do 



