[Ill] EMBRYOGRAPHY OF OSSEOUS FISHES. 5G5 



into tlie caudal plate cp, Fig. B, which is wholly converted into the cau- 

 dal mass cm of Fig. E. To urge the history of the closure of the blas- 

 toderm in Ehismobranchs as evidence against the process as it occurs 

 in Teleostei is unfair, because a considerable part of the blastodermic 

 rim in the embryos of cartilaginous fishes has evidently nothing to do 

 with the development of the body of the embryo, but closes after the 

 latter has been elevated above the yelk upon a stalk. 



An actual crater-lilie depression yb, Fig. B, with a tine canal ending 

 upon the yelk, is seen at the time of closure of the blastoderm in 

 Teleostei. Viewed from above, as in the dotted outline in B, yh is seen 

 to have radiating wrinkles extending out over the concrescing rim r. 

 This yelk blastopore so formed is not homologous with the similar pore 

 or cleft in the frog's ovum, because it cannot be shown to have anything 

 more than an indirect connection with the intestine in the embryo 

 fish. This is illustrated in Fig. C, showing an embryo of Tylosurus 

 in outline, unrolled from the yelk, with the oval blastodermic rim r 

 attached. The latter is concrescing; its contents are flowing together, 

 as it were, where its two limbs join the hind end of the body at hi. The 

 opening y& is the yelk blastopore, and is distinct from the place where 

 the actual concrescence is occurring viz, at hi. In fact, while there is 

 no actual opening into the neural or neurenteric canal at hi, neither ex- 

 isting at this stage of development, only the extreme anterior part of 

 the yelk blastopore coincides ultimately with that of the blastopore 

 of the frog, but onlj^ momentarily, in that the lumen of both intestine 

 and neurula, contrary to wlmt obtains with the frog's embryo, are want- 

 ing at this stage. The sides of the primitive groove coalesce so quickly 

 in Teleostei — if, indeed, it can be said that there is such a gToove in those 

 forms as we know it in the frog — that evidence of its presence even is 

 evanescent, or it is at most very feebly developed. In the frog the 

 margins of the groove are free, and the neural furrow in the medullary 

 plate is most conspicuous even at the time the blastopore closes, but the 

 latter marks approximately the position of the permanent anus. In 

 Teleostei there is no sucli relation between the yelk blastopore and the 

 vent, which arises in them on the ventral side of the caudal end of the 

 embryo, as shown in Fig. K. An exact comparison of the parts of the 

 frog's ovum with those of the osseous fish is not possible, asZieglerhas 

 endeavored to show ; his failure to make out the true state of the case 

 was his want of the ai)preciation of the true nature of the yelk. In fact, 

 the continuity of the e])il)last with the mesenteron, as in the embryo 

 frog, is broken in the embryo fisli by the time the yelk blastopore has 

 closed and the caudal plate has formed. 



The frog's ovum undergoes total but unequal segmentation; the fish 

 ovum, on the other hand, undeigocs equal segmentation of its germ; 

 partial segmentation, as regards its whole mass, while the final seg- 

 mentation—gemmation — of the yelk substance as leucocytes into the 

 segmentation cavity or blood vessels is carried on after the embryo is far 



