116 PROCEEDINGS OF THE NATIONAL UUSEUM. vol.38. 



Were there no such things as fossil crinoids, cadi of these groups 

 would have the standing of an order; and for the purposes of this 

 paper they may be referred to as (1) the Holopida, (2) the Ptilo- 

 crinida, and (3) the Comatulida. 



The stem of Holopus is of the simplest possible construction, 

 being practically but a simple thickening of the primitive central 

 plate. Our Holopida therefore are a surviving early offshoot from 

 the main line of descent, representing a developmental stage of 

 great antiquity, before the inception of articulations in the crinoidal 



base. 



I have elsewhere shown that the theoretically most primitive type 

 of articulated crinoidal columnar is that found in Rhizocrinuft , and. 

 especially, Bathycrinus; these two genera, therefore, stand at the 

 foot of the Ptilocrinida, though not on the same plane, for the second 

 is much more specialized than the first. The stem of Phrynocrinus 

 is a curious and unique adaptation of the primitive articulated col- 

 umnar to a great increase in size and the necessity of supporting a 

 greatly increased weight; the individual columnars of Phrynocrirtus 

 are morphologically the same as those of Bathycrinus, hut modified 

 by being greatly shortened, the diminution in length being correlated 

 with a broadening of the terminal ellipses. The family Phryno- 

 crinidse therefore marks the furthest development in the line of the 

 simplest possible adaptation to increasing stress of a column of the 

 type found in the Rhizocrinidae. In very large specimens of Bathy- 

 crinus we notice that the primitive fulcral ridge on the articular 

 faces of the columnars shows a tendency to form a pair of triangular 

 -i i iid ures with their apices at the central canal, and these triangular, 

 or more properly fan-like, structures are marked more or less disl inct 1\ 

 with radiating lines. This indicates a second line of adaptation to 

 increased stress, which eventually results in the formation of circular 

 art icular surfaces uniformly covered with radial ing lines. Such articu- 

 lar surfaces are common to Ptilocrinus, Calamocrinus , Gephyrocrinus, 

 and Hyocrinus, and therefore we should group these four genera 

 together (a course already in pari indicated by Doctor Bather) 

 making of t hem t he superfamily I Iyocrinoida which, most ohvioush 

 on the basis of the arrangement of the brachials, falls into two 

 families, Ptilocrinidae and Hyocrinidse. 



Tin' great order of t he Comatulida, and the family Pentacrinit idae, 

 to which the yasi majority of the recent forms belong, represent a 

 higher state of development and phylogenetic (as opposed to generic 

 -■I- familj I specialization than the Ptilocrinida: for we find that the 

 primitively uniform stem of t he latter in this group is modified by the 

 development, at regular intervals, of cirriferous nodes; and the basals, 

 which throughout the echinoderms form part of the body wall and in 

 the 1 teteroradiata lie in t he same plane as the radials, forming in con- 



