512 ZOOLOGY SECT. 



(p. 33n) and the Histriobdellea (p. 33 J). The general occurrence of 

 the trochophore larva may be taken as pointing to descent from an 

 unsegmented ancestor having resemblances to the trochophore, and 

 a form like Dinophilus would afford us an intermediate link between 

 such a hypothetical ancestor and Polygordius or Protodrilus. 



The position of the Sipunculoidea in the Annulata is, as already 

 noticed, a matter of doubt ; if we dissociate them from the Echiurida 

 there is little to connect them positively with the other members 

 of the phylum. But, on the whole perhaps, the evidence in favour 

 of regarding them as allied to the Echiurida, and through them with 

 the Cheetopoda, is sufficiently strong. The segmentation has 

 become more completely aborted in the Sipunculoidea, and appa- 

 rently, in further adaptation to a sedentary life in fissures and 

 burrows, the anus has been displaced forwards as in several groups 

 of animals that are permanently fixed. 



Affinities between the Phoronida (p. 348) and the Sipunculoidea 

 have often been supposed to exist, and by some zoologists it has been 

 proposed to unite the two groups in one class. It seems probable, 

 however, that the very manifest resemblances which undoubtedly 

 exist do not indicate a near relationship, but are the result of con- 

 verging modifications of originally widely different stocks. The 

 most striking of these points of resemblance are two (1) the 

 approximation of the anus towards the oral aperture, and (2) the 

 presence of the tentacular circlet. But a study of the development 

 shows that these common features arise in totally different ways 

 in the two cases. The forward position of the anus in the Sipun- 

 culida is brought about by a gradual displacement resulting from 

 the growth of the aboral region of the body ; and the invagination 

 and evagination by which the corresponding result is attained in 

 Phoronis do not occur. Again, while in Phoronis the tentacles of 

 the adult may be looked upon as formed by the development of 

 processes along the line occupied by the post-oral circlet of cilia, in 

 the Sipunculida the tentacular lobes have nothing to do with the 

 post-oral circlet, but are formed by the growth of a series of lobes 

 from the margin of the mouth itself. The larva of the Sipunculida 

 again is, as already pointed out, very nearly related to the larva of 

 the Chsetopoda, and is a typical trochophore ; while the Actino- 

 trocha larva of Phoronis diverges somewhat widely from that type. 



In adult structure, particularly in the absence of parapodia and 

 setae and the reduction of the ccelome, the Hirudinea differ 

 markedly from the Cheetopoda ; but a study of their earlier 

 developmental stages shows unmistakably their close connection 

 with the latter group, more particularly with the Oligochseta ; and 

 the existence of an undoubted Leech (Acanthobdella) with setse 

 and with a well-developed ccelome traversed by mesenteries helps 

 still further to bridge over the gap between the two classes. 



