146 BULLETIN OF THE UNITED STATES FISH COMMISSION. 



ity. At the end of tliree hours the germinal disk, as shown in Fig. 4 

 has undergone profound changes ; the cells are no longer arranged in a 

 single stratum, but in several, superimposed upon each other, which has 

 been the result of the segmentation of the cells of the morula stage of 

 Fig. 3 in a plane parallel to that of the great diameter of the disk. 

 At the same time there has been a wreath of flat cells segmented off from 

 the edge of the disk, which would be considered by some as originating 

 separatelj' or independently of the disk, an opinion from which I dissent 

 for reasons which I have already stated in dealing with the origin of 

 the disk from a primitively homogeneous, external germinal layer of 

 Ijrotoplasm. 



THE BLASTODERM AND SEGMENTATION CAVITY. 



In Figs. 5 and 6, a half hour later, we see that the fate of the mar- 

 ginal cells has been to form the rim of the incipient blastoderm, which 

 is beginning to spread out, become thinner, and lose its distinctive feat- 

 ures as a biscuit-shaped germinal disk. The inner edge of the rim of 

 cells Just alluded to limits the margin of the segmentation cavity of 

 the mackerel egg, and I can see no reason why this space should not be 

 considered homologous with the cavity which bears the same name in the 

 blastoderm of the embryo elasmobranch and chick, in both of which it 

 is probably of greater extent than it has been hitherto suspected. The 

 roof of the cavity is at first two or three or even more cells deep, but 

 as soon as the rudiment of the embryo fish is defined at the edge of the 

 blastoderm its roof soon after is found to be composed of but a single 

 layer of cells, corresponding to the epiblast or skiu layer, while its floor 

 is the intermediary layer of Van Bambeke, and corresponds to the hy- 

 poblast, mucous or deepest layer from which the intestine, the blood for 

 the most part and perhaps the notochord is derived as development 

 advances. With slight verbal alterations I will here quote from what 

 the writer has published elsewhere* : " The disk begins to spread over 

 the vitellus or yelk, and soon acquires the form of a watch glass, with 

 its concave side lying next the surface of the yelk. Coincident with 

 the lateral expansion of the disk, now become the blastoderm, a thick- 

 ening api^ears at one poiut in its margin, which is the first sign of the 

 apx^earance of the embryo fish. With its farther exi^ansion the embryo 

 is developed more from the margin of the disk towards it center ; in this 

 way it happens that the axis of the embryo lies in one of the radii of 

 the blastoderm, its head directed towards the center, its tail lying at the 

 margin" and continuous with the rim, which is soon two or three cells 

 thick, and extends all the way round the edge of the blastoderm like a 

 ring. " Before the embryo is fairly formed a space appears in the blas- 

 toderm, limited by the thickened rim of the latter, and the embryo at 



* Structure and ovarian incubation of the Top-minnow (Zygouectes), " Forest and 

 Stream," August 18, 1881. 



