186 BULLETIN OF THE UNITED STATES FISH COMMISSION. 



a structure which disappears so entirely as not to leave behind any 

 organ which may be with certainty traced to it as its source of forma- 

 tion, I am loath to regard as oue of the primary embry6nic layers. The 

 yelk is entirely included within the abdominal cavity in hshes as soon 

 as the blastoderm has closed over it. In this regard it widely differs 

 from the chick, a point to be borne in mind in this discussion. The serous 

 space around the yelk in the shad represents the body cavity. Look- 

 ing again, during the present writing, at sections made from embryos 

 shortly after the inclusion of the yelk by the blastoderm, I am convinced 

 more forcibly than ever of the correctness of the view herein maintained. 

 I cannot persuade myself, even while examining this early stage, that 

 the yet thin and incipient yelk wall is continuous, or likely to have, been 

 with any of the embryonic layers, except during the very earliest stages 

 of development and before the differentiation of the layers. The yelk 

 hypoblast, therefore, has only a physiological and mechanical function to 

 perform, which ends with the final and complete absorption of the yelk 

 out of the serous abdominal cavity. 



Immediately after the heart is formed, as soon as it begins to pulsate, 

 and long before hatching, it seems to open directly into the serous cav- 

 ity already described. In this condition and even much later it seems 

 to the observer almost like an independent being within the embryo, 

 sucking up the yelk; an appearance which, at this time, is of course 

 illusive, as the breaking down of the yelk by the help of germination 

 and the circulation probably does not begin until about the time the 

 embryo is free from the egg. Previous to tbat time the appropriation 

 of the yelk material probably goes on by intussusception, as already 

 mentioned. The communication of the heart with the serous cavity 

 surroundiugtheyelk, as stated before, is direct, but as soon as the Ouvi- 

 erian ducts are developed, its venous end is almost entirely fed by them 

 from the cardinal veins, the serous cavity in front of the yelk only com- 

 municating imperfectly with the heart. In four or live days the hulk 

 of the yelk is absorbed, some remnants of it sometimes remaining for 

 a long time afterwards, or up to tin' tenth day or even later. The rate 

 of j T elk absorption is profoundly influenced by temperature, which is no 

 more than was to \te expected. 



The diminution in the bulk of the yelk is accompanied by a gradual 

 collapse of the outer sac, the diminution of the capacity of one seem- 

 ing t» keep pace with that of the other. This is the case with the shad, 

 and in fact with most embryo hshes. The most notable exception to 

 this rule being the very remarkable phenomena first observed by the 

 writer in the embryos of Cybium and Parephippus, where the collapse of 

 the yelk mass in its vesicle, as absorption goes on. is not followed by 

 an immediate and equivalent diminution of the capacity of the external 

 sac. It follows from this state of affairs, in these species, that the se- 

 rous cavity around the yelk becomes remarkably enlarged. The ques- 

 tion, then, also arises, how does the extra water lind its way into this 



