The Development of Phascolosoma. 129 



Besides the resemblance in external form and in the general 

 homolog}^ of the prototroch, a still more crucial point of resemblance 

 between Phascolosoma and the Chaetopods is the indication of transitory 

 metamerism, which I have observed in the rudiment of the nerve 

 cord and in the mesoblastic bands, in the trochophore of Ph. gouldii 

 immediately before metamorphosis. This is so clearly marked in the 

 nerve cord of this species that I feel sure that it is a constant 

 character. 



In view of tliis fact I am inclined to regard the three pairs of 

 lateral bristles, which Selenka found in Ph. elongatum (?) at Ville- 

 franche, as indications of metamerism, although they do not occur in 

 the three forms which I studied. A renewed investigation of 

 Selenka's variety may show more fundamental signs of metamerism 

 than is furnished by these purely ectodermal structures, which appear 

 to arise, if Selenka's account is correct, in a much simpler manner 

 than the setae of Chaetopods. 



These evidences of metamerism may be interpreted either as 

 vestiges of a completer metamerism that ma}^ have existed in the 

 ancestors of the Sipunculids or, on the other hand, as incipient 

 tendencies of a somewhat primitive organism. If the former sup- 

 position be correct, Sipunculids are highly degenerate Annelids that 

 are closely allied to Chaetopods. If the latter view be accepted, 

 they are either primitive Annelids, which stand even nearer the 

 Coelenterata than Polygordius, or they constitute a distinct phylum 

 of monomeric Trochozoa. 



The idea has been advanced by Heath (1899) that metamerism 

 in Annelids may have arisen, like the incipient segmentation in 

 IscJmochiton, in the dorsal ectoderm of the trunk, and thence have 

 extended laterally to the ventral side. This hypothesis, it is claimed, 

 would account for the unsegmented condition of the ventral nerve 

 cord in Polygordius. although there are certainly no signs of early 

 segmentation of the dorsal part of the somatic plate in this form, 

 nor in the Chaetopods. On the other hand, the earliest evidences 

 of metamerism in the ontogeny of the Chaetopods are the rudiments 

 of seta-sacs along the lateral lines of the ectoderm of the trunk, 

 and the simultaneous segmentation of the mesoblastic bands, as has 

 been shown in PsygmohrancJms (E. Meyer, 1901, p. 256) and in 

 Nereis (von Wistinghausen, 1891). Metamerism in the Annelids thus 

 makes its first appearance in lateral ingrowths of ectoderm, which 



Zool. Jahrb. XXIII. Abt. f. Anat. 9 



