[37] EMBRYOGRAPHY OF OSSEOUS FISHES. 491 



days later tlian in the stage represented in Fig. 14, or on the seventh 

 day of development. The disk spreads somewhat and becomes decidedly 

 concave on its inner face, at the same time a cavity appears which 

 occupies an eccentric position at one side of and within tlie blastoderm, 

 as shown in section in Fig. 15. This cavity appears to be the result of 

 further cleavage and is filled with a serous fluid; the cells which inclose 

 it frequently jut into the cavity somewhat irregularly. In Fig. 16, the 

 space occupied by the segmentation cavity is shown by the lighter 

 area sg somewhat crescentic in shape and bounded by a thicker rim of 

 cells around its outer margin and the thinner portion of the embryonic 

 disk above. The portion of the blastoderm from which the head of the 

 future embryo will be developed is shown just below A, as a rounded 

 promontory of cells projecting into and forming the concave margin of 

 the segmentation cavity.v This promontory from A to B is composed 

 of a number of layers of cells and represents the embryonic disk or 

 shield of authors, in which the first trace of the axis of the body of the 

 embryo becomes apparent. An outline, Fig. 17, more magnified and 

 somewhat older, shows comparatively little change in the form of the 

 blastoderm and segmentation cavity. 



The origin of the segmentation cavity as well as the character of its 

 walls has engaged my attention considerably. 



Klein * represents it as originating by the elevation of the blastoderm 

 at one side, so that it is freed from contact with the parablast layer 

 lying just below it. In this way a space, filled with fluid, is developed. 

 As far as my own observations enable me to reach a conclusion it appears 

 that the above view of its origin is probably the correct one. The 

 imperfect floor of the cavity is afterwards apparently develoi^ed by an 

 ingrowth and budding of scattered cells at its edges and bottom, jirob- 

 ably from the yelk membrane (parablast); this floor disappears during 

 a later stage. It is singular that no investigators have recognized the 

 homology of this cavity with the segmentation cavity in the eggs of 

 Elasmobranchs and Amphibia, as pointed out by Balfour, with whose 

 conclusions in this regard I was wholly in accord long before the ap- 

 pearance of the second volume of his monumental work on embryology 

 in 1881. With regard to the details of its development, however, I 

 ditfer with this authority ; and of his statement thatit disappears " shortly 

 after the appearance of the medullary plate" I can only say that I Jiave 

 accumulated a very large amount of evidence in proof of the contrary. 

 Inasmuch as the whole of the evidence on this point is now in mj' ijos- 

 session in the form of sketches from living ova as well as sections, it 

 may be well to give a summary of my views regarding this i)oint with 

 references to previous investigators. H. Rathke is the first to have 

 described the growth of the blastoderm over the yelk and its complete 

 inclosure of the latter as observed by him in 1832 in the development 

 of Zoarces. 



• Quar. Jour. Mic. Sci., No. LXII, 187G, pp. 113-131, plate VI. 



