CTENOPHORA. 49 



type'). Regarding the nervous system it does not appear that it is in any wa}- more primitive than 

 in the Polyclads. But, above all, it seems rather impossible to reconcile the peculiar character of the 

 entoderm in the Acoela with their derivation from a radiate type, the Coelenterates being decidedly 

 not characterized by the lacking of a gastral cavity, v. Graff maintains that "die Thatsachen der ver- 

 gleichenden Anatomic und der Entwickelungsgeschichte bieten . . . gar keine Handhabe dafiir die Acolie 

 als Riickbildungserscheinung oder als Folge einer secundaren Erwerbung zu bezeichnen" (p. 1932), and 

 it is maintained that during the development no trace of a gastral cavity appears. The researches of 

 Pereyaslawzewa^) are against this statement; comp. the section of a "gastrula" of Aphanostoma diver- 

 sicolor 0rst., represented in Taf. IV, Fig. 32, of Graff's work, where a very distinct cavity is shown 

 between the large entodermal cells. According to v. Graff (p. 1970) these researches are, however, not 

 very trustworthy. But even if there is really no trace of a gastral cavity in the embryonal development 

 of the Acoela, general morphology decidedly supports the view that the acoelous condition is a 

 secondary character, as also the derivation from the Coelenterates maintained by v. Graff seems irre- 

 concilable with the view that the acoelous condition is a primitive character. I may also recall the 

 fact that in the Polyclads the entoderm generally forms no gastral cavity from the beginning 

 (sterrogastrula). 



Upon the whole I must say that I cannot find any real support for v. Graff's theory. I cannot 

 doubt the correctness of Lang's view that the Polyclads are the most primitive of the Turbellarians. 



Having disposed of these two main questions, we may proceed to discuss the special objections 

 raised against the theory of the derivation of the Polyclads from the Ctenophora. 



The homologies of the axes of the Poly clad and Cteuophoran body as represented by Lang 

 (comp. p. 38— 39) are maintainedby Will ey to be quite erroneous; he thinks, that he has demonstrated con- 

 clusively, that the long axis of the Polyclad body corresponds to the transverse (tentacular) axis of 

 Ctenophores, not to the sagittal axis of the latter, as Lang maintains. Accordingly the nuchal ten- 

 tacles of Polyclads are not homologous with the tentacles of Ctenophores, — upon which assumption 

 Lang's theory "rests in the first instance", his interpretation of the axial relations being "framed in 

 accordance with this assumption" 3) — but with the sensory tentacles of Ctenoplaiia, which undoubt- 

 edly represent the polar fields. The tentacles of the Ctenophores, on the other hand, "belong to the 

 same category of structures as the proboscis of Nemertines and of certain Rhabdocoele Planarians" 

 (Willey Op. cit. p. 335). As argument for this interpretation of the axial relations Willey adduces 

 the peculiar Thoe paradoxa of Chun, in which, at first, only one "directive" tentacle is developed, 

 the other appearing later. In Ctenoplana the tentacular axis is certainly equipolar, both tentacles being 

 equally developed, so that we cannot from them conclude which side is the anterior; "but if we con- 

 sider about which axis the paired structures are situated, we are simply forced to acknowledge that 

 the plane of the tentacles corresponds to the sagittal plane — in other words, that the tentacle axis 

 of Ctenoplana and Ctenophora corresponds to the longitudinal axis of Bilateralia". (Willey, Op. ciL 



') Also Hatschek (Lehrbuch d. Zool. p. 319 — 320) assumes a wandering of the mouth and pharynx from the lower 

 end of the body to the middle of the ventral side. 



2| S. Pereyaslawzewa. Monographie des Turbellaries de la mer noire. Schr. d. neuruss. Naturf. Ges. Odessa. XVII. 1892. 



3) This is scarcety a fair statement. Langs theory rests on an elaborate comparison of the anatomj- and develop- 

 ment of Polyclads and Ctenophores, and the tentacles are not even used "in the first instance" for determining the axial 

 homologies. (Comp. Lang's Monograph, p. 646). 



The Ingolf-Expedition. V, 2. 7 



