208 Mr. J. T. Cunningham on some 



the true segmentation-cavity. Bambeke and other investi- 

 gators have failed to recognize the homology of the segmen- 

 tation-cavity in Teleostei, Elasmobranchii, Amphibia, &c." 



In his account of segmentation in the general portion of 

 his work Balfour says that simple segmentation leads to 

 the formation of a hollow vesicle or blastosphere enclosing 

 a central cavity, which is called the segmentation-cavity or 

 cavity of von Baer. In German works this cavity is called 

 sometimes " Furchungshohle," sometimes " Keirahohle ; " 

 and some English embryologists speak of it as the subger- 

 minal cavity or germinal cavity. 



In my paper " On the Relations of the Yolk to the Gastrula 

 in Teleosteans " (2) I described the history and relations of the 

 segmentation- cavity as they are seen in the living pelagic 

 and transparent ova of the cod, haddock, and whiting. I 

 showed there that the segmentation-cavity does not become 

 visible as a distinct space between the centre of the blasto- 

 derm and the protoplasmic envelope of the yolk (periblast) 

 until the commencement of the invagination or the appearance 

 of the hypoblastic ring. I showed that the cavity never 

 exists beneath the germinal ring nor beneath the embryonic 

 or dorsal rudiment. I showed that in the process of the 

 envelopment of the yolk by the blastoderm the whole of the 

 geiminal ring becomes used up in the formation and increase 

 of the dorsal rudiment, and that the central part of the blasto- 

 derm with the segmentation-cavity beneath it comes to form 

 the ventral portion or yolk-sac of the embryo and larva. 



Before this paper of mine was written Agassiz and Whitman, 

 in a paper "On the Development of some Pelagic Fish-Eggs'* 

 (6) had demonstrated very clearly and conclusively the exact 

 nature of the processes which take place during segmentation 

 in the pelagic ovum. Hoffmann (14) had previously asserted, 

 and supported the assertion with beautiful ideal figures having 

 no relation to reality, that the first nuclear division took place 

 hoiizontally and produced two new nuclei, one vertically 

 above the other, the upper giving rise afterwards to the cells 

 of the blastoderm, the multiplication of the lower forming 

 the nuclei of the unsegmented periblast. Agassiz and Whit- 

 man showed that up to the four-cell stage there is no distinc- 

 tion between periblast and blastoderm, the cells being 

 continuous with one another below and externally with the 

 protoplasmic pellicle which envelops the yolk. But at the 

 sixteen-cell stage they showed that the four central cells have 

 separated from a thin layer of protoplasm below which covers 

 the yolk, and are thus definitely limited and defined on all 

 sides, while the twelve marginal cells remain continuous with 



