486 ANNUAL REPORT SMITHSONIAN INSTITUTION, 196 2 



whose sketches are here used. You will note that the centrodorsal 

 even seems to have a vestige of a stalk, though Miss Clark could not 

 be sure what the structure may represent. Perhaps an asterozoan 

 may yet survive with a stalked pentacrinoid stage. The bottom row 

 of illustrations shows how infrabasals develop in crinoids and aster- 

 oids at early stages of development. Thus, exactly comparable vari- 

 ations in the structure and modification of the calyx occur in young 

 crinoids and in yomig asterozoans at equivalent stages, the radials, 

 basals, and infrabasals being evidently subject to the same laws. 

 Young stages of hundreds of species of asterozoans are now known, 

 and every one without exception has been found to have a calyx. 

 The conclusion is surely inescapable that the skeleton in crinoids 

 and asterozoans is built to the same fundamental plan, namely a 

 cup-shaped central body, dominated by meridional gradients, from 

 which radial gradients emerge and tend to become pinnate. 



Figure 14, H, compares the pentacrinoid stage of a crinoid with the 

 equivalent stage in an asterozoan (fig. 14, I). In the crinoid, the 

 main radial gradient passes from the radial plate of the calyx (num- 

 bered 2) directly into the first brachial ossicle (numbered 3), and so 

 outward through the successive brachial ossicles. Thus the arm 

 plates rest upon the radial plate of the calyx without any interrup- 

 tion. Consequently, the gut is necessarily restricted to the central 

 region of the body. In the asterozoan, on the other hand, the first 

 arm ossicle (that is, the ambulacral ossicles, here numbered 3) is dis- 

 located adorally, so that all the brachial ossicles come to lie in a nearly 

 horizontal series, separated by a gap from the radial plate of the 

 calyx. During development the two first arm skeletal elements actu- 

 ally migrate into the jaw, to which they contribute, while growth 

 of the disk and insertion of secondary dorsal plates widens the gap 

 between the original radial plate and the first arm ossicle. Thus it 

 is that the gut can send a caecum outward into the arm, carrying with 

 it as it goes an envelope of perivisceral coelom. Thus it is by no 

 means difficult to understand how the asterozoan arm could arise from 

 the pelmatozoan arm, for in the very life span of asterozoans still 

 living we can observe the action of the processes we postulate. That 

 the earlier embryologists were unable to interpret the data in this 

 way was because they had studied only genera of families which 

 we now can prove to have been late derivatives, loaded with secondary 

 dorsal and ventral plates, completely concealing the course of the 

 original radial gradients. Thus secondary plates, such as asteroid 

 carinals and ophiuroid dorsals, and terminal plates of the arms, were 

 mistakenly supposed to continue the series started by the radial plates 

 of the calyx. In fact, no carinals and no dorsals ever develop in the 

 groups of asteroids and ophiuroids which have now been isolated as 

 the archaic ones, and none whatever occur in somasteroids. Terminal 



