350 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



The crinoid arms are, as will be later shown, i)rimarily paired interradial 

 structures which have become joined alonp: their radial edges, forming a radial 

 biserial api)enihige, the ossicles later slipi>ing in between each other so that an 

 elongate uniserial appendage results. The original arms were therefore i)rimarily 

 ten in numl)er, ancl were probably homologous with the auricles of the urchins 

 including the ossicles of the dental i)3'rami<ls, though turned outward from the 

 body of the animal instead of being wholly internal. Originally, before their 

 union into five, the arms probably bore no ventral ambulacral structures, and had 

 no function other than that of increasing the surface of the disk by increasing the 

 distance between the points of attachment. 



Now there are definite indications that not only the arms but also the radials 

 were originally ten in number, two on the distal edge of each basal, anil that each 

 of the interbasal radials as we now know them in the crinoids is composed of two 

 primitive radials, one from the distal border of each of the underlying basals. 



The dorsal nerve cords arise as stout interradial processes lying exactly over 

 the nerv.es leading to the cirri in the monocyclic forms. These two sets of nerves 

 thus bear exactly the same relation to each other that the dorsal and ventral nerves 

 do which ijmervate the legs and wings of insects, and are probably to be considered 

 as in a way analogous to these. Within the basals each of these primary nerve cords 

 divides into two secondary nerve cords, each of which enters an adjacent radial; 

 at the distal border of the radials the two cords from the two adjacent basals fuse 

 and form a single cord which is continued into the arms. Thus the arms are 

 innervated by a radial nerve cord which is formed by the ultimate union of the 

 two halves of interradial primar}'' nerve cords. 



Each primary trunk within the radials, as also just after its di\dsion within the 

 basals, indicates its primarily interradial origin by a conomissuro wliich joins the 

 derivatives from the original nerve trunk (fig. 64, p. 89). Each of the great doi-sal 

 nerves of the arms is made up of half of each of the two primary nerve trunks of the 

 basal <m either side of and below the radial at the base of the arms which have moved 

 together and have become fused into a single nerve. 



We thus have each of the primary nerve cords dividing and sending out two 

 diverging brandies (wliich happen to fuse with similar branches from the adjacent 

 primary cords in the recent forms) that are connected by two transverse cords, one 

 within the basals, the other within the radials. These transverse commissures I 

 consider to bo strictly comparable to similar commissures in the ventral nervous 

 system of primitive molluscs, phyllopod crustaceans, nemerteans and Peripatus, 

 and to show conclusively that the five radiathig units of which the nervous system 

 of a crinoid is made u]) are not the five radial nerves fnmi the radials outward, 

 but the interradial primary cords and their branches and connectives as far as the 

 point of union in the radials ; and from that point onward the axial cord of the arm 

 must be considered as being coni])osed of two halves, each belonging to the ad- 

 jacent interradial nerve cord, and tlierefore as being in reahty two halved inter- 

 radial nerves lying side by side in a radial position. The radial commissures (wliich 

 collectively form the so-called circular commissure) are therefore to be regarded as 



