202 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



rigid column, anil not accomplished after the calcareous deposition has been com- 

 pleted and the rigid character attained. 



No recent crinoids are known in which the first Imo of development obtains; 

 but it is well illustrated by the fossil genus Ednocnnus. No crinoids are known 

 in which the column is composed simply of two columnals, as would be the case in 

 the first stage in the second line of development. But suppose we carry- this line 

 further; we have a cruioid attached bj- a column iii which an articulation has been 

 developed in the center; such an articulation would of necessity develop a fulcral 

 ridge running across the joint faces and embracing the central canal, admittmg of 

 motion in a single plane, coinciding with that in which the original stimulus deter- 

 mining the fracture was received. Stem growth would continue; but as new 

 deposition occurs only just under the calyx, only the proximal columnal would 

 increase in length. Soon the proximal columnal would become so long as to become 

 brittle, as did the original stem, and fracture would again occur midway between 

 the first articulation and the calyx. Now, this fracture would almost certainly 

 differ from the original fracture in bemg formed at right angles to it, for any force 

 exerted in the same plane as that which caused the original fracture would be taken 

 up by the articulation which has formed; but, owing to the definite direction of, 

 and the close union along, the fulcral ridge, any force coming parallel to the fulcral 

 ridge — that is, at right angles to the original force — would meet with resistance, as 

 for a force exerted in this direction the original articulation would be practically 

 nonexistent, and a second fracture would occur in the weakest spot; namely, half 

 way between the original articulation and the calyx, developing into a second 

 articulation m which the fulcral ridge would run at right angles to the direction 

 taken by that of the first. A still further mcrease in stem length would mean a 

 progressive increase in the number of articulations, each of which would, in the 

 direction taken by its fulcral ridge, alternate ^vith those on either side; antl thus 

 would eventually be formed the primitive pol3^columnar crinoid stem, a stem 

 exactly comparable to the stem of Rhizocrinus (figs. 135, 137, p. 205), Bathycrinus 

 (fig. 527, pi. 2), and the young of the comatulids (figs. 407, p. 317, 532, 533, pi. 3). 



iUthough the origin of the polycolumnar crinoid stem appears undoubted!}' to 

 to have been from a single original calyx plate, a centrale corresponding to the 

 centrale in Marsupites (fig. 565, pi. 7) or in TJintacrinus (fig. 572, pi. 7) and to the 

 central plate of certain echmoids, it does not necessarily follow that the redupli- 

 cation of the columnals was the result of a series of actual morphological fractures 

 as just described. 



This is the most obvious explanation, and tlie one wliicli may be most readily 

 grasped; at the same time, through explaining the development of the alternating 

 fulcral ridges, it indicates with a reasonable degree of accuracy the method by which 

 the rapidly deveU)j)ing columns of the later fossil and of the recent types, as opposed 

 to the slowly developing columns of the palaeozoic forms, have come into existence. 



The primitive type of column, occurring in the palseozoic species almost 

 exclusivel}-, but persisting in the recent Plicatocrinidae, is cliaracterized by short 

 cylindrical columnals which have the articular faces marked with radiating 

 ridges. The explanation of the origin of this type of column is somewhat 



