THE DEVELOPMENT OF OPHIOTHRIX FKAGILIS. 593 



of cells in the Bipinnaria. The coelom on each side was three- 

 lobed, and the central division in all probability was produced 

 into tentacles, whilst locomotion was effected by means of a 

 ciliated band. Now, if we survey the existing animal 

 kingdom in order to find some type still living which will 

 approximately answer to this description we are led to the 

 group of the Ctenophora. Here we find a group of animals 

 with an anterior neuro-epithelial plate, swimming by means 

 of cilia. Connected with the endodermic stomach or archen- 

 teron, as we may term it, are right and left outgrowths. 

 From each of these a canal runs forward into the anterior 

 portion of the body to end near the brain (so-called excretory 

 canal) whilst another runs backwards to near the mouth 

 (paragastric canal). The middle division of the outgrowth 

 leads into a tentacular canal which also gives rise to canals 

 which open into the meridional canals lying beneath the 

 ciliated ribs. Now I am very far from suggesting that the 

 ancestor of Echinoderms resembled in details a living 

 Ctenophore, but if Ctenophora are the remnants of a great 

 group of pelagic animals with considerable variation in 

 structui'e, the Echinoderm ancestor may well have belonged 

 to that group. The Ctenophore represents a stage before 

 the coelom had been separated from the gut or the primi- 

 tive mouth divided into mouth and anus. It is to me a 

 matter of extreme interest to find that not only have the 

 researches of Lang and the discoveries of Willey and others 

 with regard toCtenoplana led to the conclusion thatTurbellaiia, 

 and therefore all the group of Platyhelminthes, are descended 

 from a Ctenophore-like ancestor, but that Woltereck (32), 

 working on the development of Polygordius, is led to the same 

 conclusion with regard to this animal also. This conclusion, 

 if justified, carries with it the consequence that the whole 

 of the Annelida and Mollusca, which are bound together 

 by the common possession of the trochophore larva, are also 

 derivable from this root. When such diverse lines of research 

 seem to converge upon one point, it really does seem as if we 

 were within measurable distance of gaining some idea of what 



