BULLETIN OF THE UNITED STATES FISH COMMISSION. 147 



one side. This space, the segmentation cainfjj, is fiUed tcith fiuid and 

 groics tcith the groicth of the germinal disl-, as the latter becomes converted 

 into the hlastoderm^ and does not disappear until some time after the embryo 

 has left the egg as a young fish, after remaining as a sjMce around the 

 yelk-sack as long as a vestige of the latter remains, as may be seen in 

 the young of Cybium, Gadus, Elecate, Syngnathns, and Alosa." 



My observations have been conducted without hardening reagents, 

 since it has been found that such methods abstract the water from the 

 embryo, and cause the segmentation cavity to collapse and be obliter- 

 ated, so that the only way in which the writer has been able to follow 

 the history of this space was to study it in the living transparent eggs, 

 whi(;h may be got into various positions so as to show all the x)hases of 

 its development in the diiferent stages of the evolution of the embryo. 

 I believe it will be found to be present in the blastoderms of the ova of 

 almost all teleostean fishes. 



"Should this prove to be the fact [quoting from the same source] the 

 teleostean egg will be as distinctly defined, in respect to the sum of the 

 developmental characters which it presents, from the developing ova of 

 other vertebrates, as the adult teleost is from the remaining classes of 

 the subkingdom to which it belongs." 



Later, as is shown in Fig. 7, or after seven hours, the blastoderm 

 has grown so as to inclose nearly one -half of the vitelline globe or 

 yelk, the rim is very distinct, and when viewed from above as a trans- 

 parent object, the segmentation cavity is visible as a somewhat creseent- 

 shai)ed area more transi^arent than the embryo or the rim. The embryo 

 bounds the concave side of the crescent and lies in immediate contact 

 with the yelk, except over a small space just under the fore part of the 

 head, which is found to be continuous with the segmentation cavity 

 beneath the latter ; this space will be found to be very significant, and 

 is the cavity in which the heart develops. In taking another look at 

 Fig. 7, it will be noticed that the blastoderm is a hemispherical cap, 

 and that on the left hand from its center to the edge of its rim there is 

 a thicker i^ortion shown ; this is the enibryo mackerel seen from the 

 side with its head end lying in the middle of the disk and its tail at the 

 edge. To the right hand and below a clear space is shown ; this in 

 like manner is the segmentation cavity seen from the side, and to the 

 right of it the blastodermic wal's are seen to be double, consisting in- 

 ternally of the hypoblast and externally of the epiblast, with a space, 

 sg, between them ; this is the segmentation cavity in optic section, 

 which is seen to extend a little way under the head of the embryo at 

 cr.s', to form the cavity in which the heart will be formed. To the left of 

 crs the keel or carina, cr, of the embryo dips down into the vitellus ; the 

 carina is simply the fore part of the medullary canal, which for the 

 most part becomes the great median nervous or spinal cord of the young 

 fish ; in all embryo teleost fishes it is much flattened laterally in its 

 fore part, and in consequence it dips down far into the yelk as a flat 



