ON DINOPHILUS GIGAS, 117 



the head cavity of Criodrilus and of many Polychoets is, at 

 an early stage, ^ exactly in the condition which is permanent in 

 Dinophilus ; it is a cavity, not bounded by any definite 

 "coelomic" epithelium, but traversed by mesodermic fibres, 

 which form a plexus running through it. 



From these considerations it may plausibly be argued that 

 we have in Dinophilus a form representing in its main features 

 a stage in the evolution of Chsetopods which is in the existing 

 members of that group repeated only in the larval condition — 

 a form in which the only archiannelid character which is not 

 developed is the epithelial and segmented character of the 

 body cavity. 



That the epithelial character of the body cavity may be 

 acquired within the limits of a group, Saccocirrus, as already 

 pointed out, seems to prove ; while the acquisition of segmen- 

 tation is well seen in the various species of Dinophilus itself. 

 Thus, in D. vorticoides^ we find the whole body unseg- 

 mented, with a uniform covering of cilia; in D. apatris^ we 

 have an external segmentation which is not shared by the ex- 

 cretory system; while in D. gyrociliatus we find the ne- 

 phridia composed of '' simple, intracellular, segmental organs, 

 terminating in flame cells ;"^ and lastly, in D. metame- 

 roides we have the appearance of a commencing segmenta- 

 tation of the body cavity.^ 



But the anatomy of Dinophilus seems to show that from its 

 near connection with the Trochozoon ^ it is related to other 

 forms besides Chsetopods. The pharynx seems especially to 

 show this. Comparing the longitudinal section (fig. 11) with 

 a similar section through the pharynx of Histriobdella 

 (fig. 13) we see that the pharyngeal apparatus is obviously 



1 Cf. Hatschek, " Stud. iib. Entw. d. Anneliden," ' Arb. a. d. Zool. Inst. 

 Wien,' 1878, and others. 



^ E. van Beneden, ' Bull. Acad. Roy. Belg.,' Tome xviii. 



^ Korschelt, loc. cit. 



•* Ed. Meyer, quoted by Lang, loc. cit. 



* Hallez, loc. cit. 



* I use this term to imply simply the type, whatever that may have been, 

 which is now ontogenetically represented by the trochosplieres. 



