BULLETIN OF THE UNITED STATES FISH COMMISSION. 181 



the yolk and the intestine in some embryo sharks. The general form 

 of the yelk vesicle, as may be seen in the figure, is ovoidal, slightly 

 flattened on its upper side, with a depression or furrow traversing the 

 flattened portion lengthwise. It is entirely surrounded, on all sides, by 

 a spaee tilled with a serous fluid. At the anterior end of the yelk vesi 

 ele this space, in the stage of development here described, is most ca- 

 pacious, and comprises all that cavity marked by c between the posterior 

 pericardial membrane j)>n and the yelk envelope yhy. This space c 1 

 have identified with the segmentation cavity, for reasons which it will 

 not be necessary to present iu detail in this place. If the heart does 

 not actually develop within this cavity, its immediate connection with 

 this space is an incontestable fact. Practically the heart develops within 

 it, as we have elsewhere described. In the cod (Gadus) the mesoblast 

 from which the heart is developed lies upon this space, and as develop- 

 ment proceeds each step of the heart's evolution maybe watched most sat- 

 isfactorily. At most, the only separation between the pericardial space p 

 and the heart is effected by the development of the posterior pericardial 

 membrane pm, which is usually of extreme tenuity in the stage of de- 

 velopment here described. In fact, I am not sure that the membrane 

 pm may not be perforate, for the reason that blood corpuscles are almost 

 always found iu the pericardial space at about this stage of development. 

 In Tylosurus I am quite sure that the pericardial cavity is not shut off 

 from connection with the homologue of c, because iu that genus it is 

 crowded, in some stages, with blood corpuscles, which vibrate in unison 

 with the pulsations of the heart in the fluid in which both are immersed. 



In the shad, as in other species, the membrane pm is continuous with 

 the splanchnopleural or peritoneal layer. As the venous end of the 

 heart, just above p, pulsates, the membrane pm is pulled back and forth 

 by the action of pulsation. Moreover, the membrane pm is continuous 

 with and joined to the venous end of the heart, just above p, and in 

 front of c. Iu fact, the heart opens freely into the cavity c. Free com- 

 munication is thus established between the cavity of the heart itself 

 and the segmentation cavity, or the serous space which surrounds the 

 yelk vesicle. 



We are now ready to comprehend in a measure the manner in which 

 the material of the yelk is generally broken up into small spherules or 

 corpuscles and sucked up out of the space e by the heart, and carried 

 into the body of the embryo to be appropriated in the processes of further 

 development. On the surface of the structureless membrane yhy care- 

 ful observation will reveal the fact that minute spherical prominences 

 are developed. If one will be content to observe patiently for a 

 couple of hours, these bodies will finally be seen to free themselves 

 from all further connection with yhy and to drop freely into the 

 surrounding fluid. These corpuscles are quite colorless and present 

 the irregular globular appearance of the white blood cells found in 

 human blocd. Owingto their continual vibration iu the serous medium 



