EMBRYONIC DEVELOPMENT OF THE TELEOSTEANS. 55 



around this cavity tends to confirm this view, as Kupffer has 

 pointed out. I am anxious to insist, moreover, upon the fact 

 that the name of " segmentation cavity " given to this cavity is 

 not at all appropriate to it. 



The segmentation cavity, as known in Amphioxus, the Cyclos- 

 tomes, the Batrachians, and a host of invertebrates, is invariably 

 situated between the ectoderm and the endoderm. It is limited 

 on one side by the concavity of the ectoderm, on the other by 

 that part of the primitive vesicle which will, after its invagination 

 is accomplished, constitute the endoderm. If we once admit that 

 the intermediate layer of Osseous Tishes is the homologue of the 

 endoderm of the other Metazoa, it is clear that the cavity which 

 develops between the blastodisc and the intermediate layer alone 

 deserves this name. This cavity, which existed in my eggs, and 

 which the majority of embryologists have observed in TeleosteaJis, 

 is generally designated by the name of germinal cavity (Keim- 

 hohle). It is indispensable to modii'y our terminology in accord- 

 ance with the preceding observations. I propose, then, to 

 designate under the name of coAnUf^ofi^Laiekoullet the cavity 

 pointed out by this author in the midst of the blastodisc in the 

 Perch and the Pike, by Van Bambeke in the Eoach, and by 

 Kupffer in Gohius niger. Balfour (20) has discovered a homo- 

 logous cavity in the Elasmobranchs. 



It is necessary, on the other hand, to give the name of cavity 

 of von Baer, segmentation caviti/, or blastocal, to the space 

 which appears at the conclusion of egg-cleavage, between the 

 blastodisc (ectoderm) and the intermediate layer (endoderm). 

 It is this cavity which was described in the Trout by Strieker, 

 by Ileineck, by Weil, by CEUacher, by Klein, by His, and by 

 Gotte, and which has been called sometimes Furchungshbhle, 

 sometimes Keimholde. In common with CEUacher I consider 

 this cavity as the homologue of the germinal cavity of the chick. 

 It is for this reason that 1 think the name of Keimhohle may be 

 retained equally with the other names for this cavity. The 

 existence of this cavity appears doubtful in the Eoach and in 

 the genera Gasterosteus and Spinachia, if we may judge by the 

 observations of Van Bambeke and Kupffer. 



IV. A final question which I wish to enquire into relates to 

 the ultimate destination of the two primordial layers of the 

 embryo of osseous fishes — the blastoderm on the one hand, and 

 the intermediate layer on the other. At the phase represented 

 in figure 6 of my plate the embryo is composed, if we except 

 the enveloping lamella, of three cellular layers well defined in 

 the marginal welt. The external one, limited by the enveloping 

 lamella, is evidently derived from the blastodisc ; the internal 



