DEVELOPMENT, STEUOTCJRE, ETC., OF ACTINOTliOCHA. 489 



opposite result, i. e. that of sfcraighteuing out these de- 

 pressions. 



A second group of discrepancies is apparently due to 

 specific variation. Amongst these we may instance Ikeda's 

 failure (as in the case of Roule) to find two pleurochords. 

 In the North Sea larvae the anterior wall of the stomach 

 pushes forward as a diverticulum below the oesophagus, 

 which opens into the antero-dorsal corner of the stomach. 

 If it simplifies matters at all to regard this diverticulum as 

 " ventral " instead of anterior, there seems no objection to 

 doing so. As development proceeds, this diverticulum 

 throws out two lateral pleurochords as I have described. In 

 the Japanese larvae this stage is never reached, but the 

 diverticulum remains single with only one row of vacuoles ; 

 hence we must be prepared for other organs in like manner 

 being abbreviated or even undeveloped. For example, the 

 nephridia in these larvas, as figured by Ikeda, are very small, 

 very simple, and very embryonic as compared with the large 

 branching tubules in the North Sea type. This considera- 

 tion may easily account for the failure of Ikeda to find the 

 internal openings of these organs in any of his larvas. It is 

 possible that the absence of proboscis pores, of a post-oral 

 nerve-band, and of dorsal and ventral nerve-fibres n^ay be 

 due to a similar cause. In future work upon the central 

 plexus of Actinotrocha I shall have a further occ;isiou to 

 refer to these matters. 



A third source of seeming discrepancy is due to an un- 

 intentional misunderstanding, on the author's part, of my 

 statements. 



I had to notice the existence of small processes of the 

 trunk coelom which lie in the perianal sinus, and in doing so 

 was obliged to hold in view the possibility of their being the 

 first indications of the adult nephridia. Further investiga- 

 tion has convinced me that they have little morphological 

 significance, but both Roule and Ikeda have imputed to me 

 the definite statement that the trunk coelom has definite 

 nephridia, apparently only for the purpose of contradicting 



