446 THE EVOLUTION OF THE METAZOA 



of the Oligomeria, a development which in all probability 

 never actually took place. It could not be considered as a 

 case of neoteny if this larval form evolved into a new type after 

 its descent to the sea bottom, thus after a metamorphosis. 

 The Chordonia, if they evolved in this way, would have a 

 planktonic larva similar to the so-called dipleura. Something 

 of this kind, however, is unknow^n to us. The interest- 

 ing proposition made by Garstang to derive the central 

 nervous system and the endostyle of the Chordonia from an 

 immersion of ciliated bands of the Echinodermata larvae must 

 be considered as an entirely erroneous concept, as a purely 

 artificial construction. Where is here the origin of a secondary 

 metamerization, of the notochord, of gill slits, etc.? Such a 

 suggestion could perhaps be taken into consideration if there 

 were no longer any such animals as the Enteropneusta. 

 The adult Enteropneusta are comparatively less specialized than 

 the planktonic larvae of the Ambulacralia. It is therefore easier 

 to derive the Chordonia from the Enteropneusta than from 

 the planktonic larvae of the Echinodermata. Such a derivation 

 is also, therefore, much more probable. The suggestion that 

 the hypothetical immersion of ciliated bands could be the 

 starting point for the evolution of new organs is somewhat 

 similar to Krumbach's attempt to derive the protonephridia 

 from an immersion of the swimming plates of the Cteno- 

 phora. 



Even less probable than a derivation of the Chordonia from 

 the planktonic larvae of the Echinodermata seems to be a 

 derivation of the Chordonia from some very ancient, armour- 

 ed, adult forms of the Echinodermata. Such a suggestion 

 was made by Gieslen (1930) on the basis of a comparison 

 between the carpoidal echinoderm Cothurnocystis eli^ae and the 

 Ostracodermata in which he took into consideration above all 

 their armour. Such a derivation does not seem to be much 

 better than is a derivation from 'Lifnulus. These are clearly 

 cases of parallelisms or similarities which are not based 

 on any closer relationships. Berrill (1955:5) has correctly 



