EVOLUTION OF THE ECHINODERMS — FELL 481 



ophiuroids and the auricularia larva of starfishes cannot have any 

 broad phylogenetic significance, and by the same token the close re- 

 semblance of the pluteus larva of ophiuroids to the pluteus larva of 

 echinoids cannot mean any relationship between echinoids and 

 ophiuroids. Not the slightest trace of any fossil intermediate between 

 ophiuroids and echinoids has ever been found, whereas we have found 

 abundant and overwhelming evidence of the direct relationship of 

 ophiuroids and asteroids through common pinnate somasteroid an- 

 cestors, whose fossil remains we now possess in considerable number 

 and variety. 



A further point now emerges, which goes some way toward solv- 

 ing the impasse mentioned earlier, by which it appeared that indirect 

 development must be older than direct development, since some forms 

 with direct development have vestigial larvae. Although we still 

 know nothing of the embryology of Ophiocanops^ many features of 

 its anatomy declare its affinity to the Ophiomyxidae, a group of 

 ophiuroids in which absolutely direct development occurs, without 

 any trace of a larva at all. On the other hand, those genera of Ophi- 

 uroidea which have vestigial larvae have now been shown to fall in 

 families of relatively late derivation, from gi^oups which have pe- 

 lagic larvae. They are groups with numerous secondary features in 

 the skelton, far removed from the archaic forms with somasteroid- 

 like features. Thus it is now extremely probable that there are two 

 quite distinct types of direct development in ophiuroids, one ancient, 

 with no vestige of a larva, the other secondary and showing both by 

 the vestigial larva and by the characters of the skeleton that it is of late 

 origin. I now suspect that the pluteus larva will eventually be proved 

 to be a feature evolved by ophiuroids after the separation of ophiu- 

 roids from the somasteroid line, and that the pluteus of echinoids 

 is an entirely independent development of that group. It is most 

 urgent to ascertain the nature of the development in Ocliiocanops^ 

 as also of course in Platasterias. 



An analogous situation exists in regard to pedicellariae, which 

 can be proved to have originated in the Asteroidea as a late develop- 

 ment, subsequent to the differentiation of the Luidiidae. They occur 

 in no other subclass of Asterozoa. It is thus obvious that the super- 

 ficially similar pedicellariae of echinoids are an entirely independent 

 development, most of which has occurred subsequent to the differenti- 

 ation of the Cidaroida, which are the oldest surviving echinoids. 



ORIGIN OF ASTEROZOA 



We can further elucidate stages in the evolution within the Aster- 

 ozoa by interpolation between fixed points on our frame of reference. 

 Wliat happens if we now attempt to extrapolate backward, in search 

 of the ancestry of the somasteroids themselves ? 



