CLARK: NERVOUS SYSTEM OF CRINOIDS 69 
must look upon the ambulacral system as primarily double, com- 
posed of two similar interradial outgrowths which have moved 
outward from their primitive interradial position and become 
united to each other by their outer borders. Thus the ambulacra 
arose as paired interradial structures, the two components of 
each pair gradually moving away from each other and fusing with 
those of the adjacent pairs. In the gap thus left various inter- 
radial or interambulacral plates have been developed. 
It therefore follows that the five fundamental divisions of the 
echinoderm, the five units of the pentamerous symmetry, are not 
the five ambulacral series bordered with one-half of the inter- 
ambulacral series on either side, but the five interambulacral or 
interradial areas, plus one-half of each of the adjacent ambulacra. 
In this way the fundamentally double structure of the ambulacra 
is easily explained. 
The ventral ambulacral structures must be considered in the 
same light as the ventral ambulacral nerves; they are merely 
extensions of the ring systems about the mouth which have grown 
outward over the supports formed by the double dorsal ambulac- 
ral system. 
Thus while the double ambulacral skeleton and dorsal nervous 
system of the crinoids are primarily interradial structures, the 
single processes from the various circumoral systems are pri- 
marily radial, arising from an enormous development of five radial 
buds or outgrowths on each of the circumoral structures. The 
pentamerous symmetry of the crinoids, therefore, is primarily 
interradial in regard to the dorsal structures, but radial in regard 
to the ventral. 
It might be argued that, as the radial plate (or corresponding 
structure) is always single, the dorsal ambulacral structure must 
be primarily single also. The echinoderm skeleton was originally 
merely a spicular covering developed in the mesoderm of the body 
wall, comparable to that seen in certain holothurians (tho usu- 
ally in this group the individual spicules have become curiously 
specialized) and in general comparable to the skeletons of sponges, 
red coral, tubipora, etc.; mechanical considerations caused the 
spicules to become localized in definite areas so that they have 
