290 BULLETIN OF THE UNITED STATES FISH COMMISSION. 



reason to suppose that the somatic mesoblast extends much beyond the 

 sides of the body at this stage, but it ends abruptly on either side be- 

 tween the epiblast and sphmchnopleiu-e the same as in Alosa, as shown 

 by transverse sections. Fig. G shows the optic vesicles o developing at 

 the head end of the neural or spinal cord, which is solid in this species at 

 this time, as in embryo bony fishes generally. The notochord is also 

 faintly indicated at this stage. The vesicle 1c, the nature of which is so 

 puzzling, shown at the under side of the tail in Figs. G, 8, 9, and 10, is well 

 developed ; it was originally described by Kupffer, and it has been sup- 

 posed to be primitively joined to the posterior end of the intestine, but 

 of this there is as yet no satisfactory ])roof. It disappears entirely at a 

 later stage of the evolution of Belone, and appears to play only a tran- 

 sient and comparatively inconsequential j)art in the process of develop- 

 ment. The usual lateral flattening of the anterior end of the neural or 

 spinal cord takes place, as shown in Fig. 9, in optic section through it 

 and the optic vesicle on either side. Further stages in the development 

 of the optic vesicles are shown in Figs. 7 and 10, in which the rudiments 

 of the auditory invaginations are also represented. In Fig. 10 the em- 

 bryonic body only is represented, as in Fig. 9; the mesoblast has been 

 segmented into a greater number of muscular somites, and the point of 

 closure of the blastoderm is shown at the tail, where it forms the caudal 

 plate already alluded to. 



DEVELOPMENT OF THE HEART AND BLOOD. 



As stated at the beginning of this paper, the genesis of the blood of 

 Belone is perhaps the most interesting ])art of its history. The heart 

 develops in the usual way in the segmentation cavity below the head; 

 at first an annular mass of cells, it soon becomes tubular, and is pro- 

 longed forwards until its venous end a extends to the front end of the 

 head. An arterial channel is at once developed from its hinder end 

 through the body of the embryo between the notochord and intestine, 

 and just below the tail at z it widens into a capacious vessel of very 

 uneven caliber and passes entirely around over the yelk between the 

 epiblast and the hypoblast to again empty. its contents into the anterior 

 venous end of the heart at a. The vitelline blood-system is at this stage, 

 seventy-two hours after impregnation, as simple as it can possibly be. 

 It is a mere channel which is as yet hardly provided with proi>er walls, 

 except in the region of the heart and body, running the whole length 

 of the body of the young fish and continued around the yelk back to 

 the heart. On either side of the body of the embryo a small vessel also 

 makes its way outwards from the aortic channel or vessel, but suddenly 

 returns again to empty its fluid contents into the heart at a. This ves- 

 sel, or rather the one on the right side, is shown at v" in Fig. 11. With 

 the progress of development the most noteworthy change which takes 

 place in the arrangement of the two vessels on either side of the body 

 is their rapid extension and growth outwards over the yelk, as shown 



