DEVELOPMEyt OF TRACHINUS TIPEKA. 285 



are eleven or twelve protoverfcebrse. In the side view it is seen 

 as a thickening beneath the embryo (see fig. 20). I found it 

 difficult to get a clear view of the changes taking place in this 

 patch in the living egg, and can add nothing from this point of 

 view to its development. Many of the cells are, however, so far 

 as I can make out, budded off from the mesoblastic roof oE the 

 cardiac cavity, just as Ryder describes, and then grow down and 

 come in contact with the floor of the cavity. Messrs. Kingsley 

 and Conn observed the same process of development in the 

 Gunner. The heart is solid at first, but a lumen afterwards 

 develops when it has reached the floor of the pericardiac space. 

 The heart is only a simple hollow cellular tube at the time it 

 commences pulsating, and has its broad venous end closely 

 applied to the vitellus. The first pulsations, faint and somewhat 

 intermittent, were observed at the stage of about twenty proto- 

 vertebrse, and about the same time the first spontaneous move-, 

 ments of the embryo have been noticed. The venous end of the 

 heart is somewhat funnel- shaped, and remains applied to the 

 vitellus up to the time of hatching. No blood-corpuscles nor 

 circulation is visible up to three or four days after hatching. The 

 observations of Ryder and Kingsley and Conn seem to agree with 

 my own on this point — no vascular system has been found up to 

 a considerable time after batching. Ryder found that in the 

 Spanish Mackerel the aorta only begins to develop 16-20 hours 

 after hatching, the whole development up to hatching only 

 occupying 24 hours. I therefore conclude that I have not been 

 able to keep my embryos long enough to follow the development 

 of the vascular system. This seems in strange contrast with 

 observations on non-pelagic eggs, which usually show a very 

 marked circulation, both in the embryo and around the vitellus, 

 a considerable time before batching. 



Intestine. — On the closure of the blastopore, and when Kupffer's 

 vesicle is at its maximum development, a thin layer of granules 

 extends from the vesicle to the blastopore, which is either the 

 homologue of the postanal gut or of the neurenteric canal. If 

 Balfour is correct in identifying Kupffer's vesicle with postanal 

 vesicle of Elasmobranchs, this layer of granules should represent 

 the first formation of the neurenteric canal, and the anus should 

 be formed at a point anterior to the vesicle. The mcsenteron is, 

 however, developed from the granular band forwarrls, and after 



LIXN. JOURN^. — Z00L0(3;T, VOL. XVIII. 20 



