460 
BULLETIN OF THE UNITED STATES FISH COMMISSION. 
causes which brought about a retardation of the functional intestine would not neces- 
sarily affect the functionless postanal section. It thus happens that the postanal 
vesicle in teleosts appears as Kupffer’s vesicle before any lumen is formed in the 
intestine. In Serranus , at least, it lies at the end of a postanal gut. In the trout, 
according to Henneguy, and in Cymatogaster it is at least in part incorporated in the 
intestine. Kupffer’s vesicle is, moreover, the only part of the intestine raised from the 
yolk before a floor is acquired. [In Serranus (Wilson, 1890) this raised portion is not con- 
fined to Kupffer’s vesicle.] Kupffer’s vesicle, in Cymatogaster at least, is more than 
the postanal vesicle of elasmobranchs. The archenteron, the postanal vesicle, and 
neurenteric canal all seem represented by it. 
In Cymatogaster it is seen that a part of this vesicle is for a time the direct con- 
tinuation of the alimentary tube and that during this time a narrow slit (neurenteric 
canal) extends upward from its anterior half. This upward extension is formed in 
Ctenolabrus , according to Agassiz and Whitman, at the closing of the blastopore. But 
the condition described by them I have never been able to see in any of the pelagic 
eggs examined by me. 
If we consider a part of the vesicle the homologue of the postanal vesicle of elasrao- 
branchs it remains to be shown why in Cymatogaster it forms part of the permanent 
intestine. The cause is not far to seek. The embryo, in the first place, is shortened on 
account of the small yolk at the periphery of which it is formed, the tail being repre- 
sented by a large knob of undifferentiated cells. On the other hand, the alimentary 
canal is precociously developed, owing to viviparity, and the whole of the hypoblastic 
area is utilized in forming the permanent alimentary tract. 
Agassiz and Whitman (1884) traced Kupffer’s vesicle in several species of pelagic 
eggs. In the formation they found what Kingsley and Conn had already well described. 
It u arises by the fusion or confluence of a cluster of granules. * * * In Ctenolabrus 
the granules appear soon after the embryonic ring passes the equator, when the length 
of the embryo is about four-fifths of the diameter of the ovum. Its maximum diameter 
when fully formed is seldom more than 0-03 mm. During its formation, till it reaches its 
maximum size, it lies beneath the chorda and the entodermic stratum and has no sort 
of relation with any tubular structure whatever. * * Yentrally and laterally it is 
bounded by periblast material. * * * It grows smaller after the closure of the blasto- 
pore, and during this period in a number of species it rises from the periblast into the 
entoderm, where it vanishes.” Behind this they have found a variable number of 
secondary caudal vesicles. 
Henneguy (1889) describes some of the distinguishing phases of Kupffer’s vesicle 
in the trout (Salmo fario.) (He first described it as early as 1880.) The first indications 
of a modification in the region of the future vesicle were noticed very early and the 
vesicle itself was quite large when but two or three protovertebrae had been formed. 
The cells in this region are larger than the others and are undergoing division. There 
are but few of these cylindrical cells, and they are in contact with the periblast, and 
later one sees’ “une invagination se produise dans l’embryon pour former la vesicule.” 
This mass of cells is the first indication of Kupffer’s vesicle. Its growth must be 
quite rapid, for it is 0T1 mm. long and 0-09 wide when but two or three protovertebrae 
are formed, and occupies u la place de la corde dorsale,” i. e., it is entirely surrounded by 
hypoblast. It lies just in front of the caudal swelling “au point ou commence a se 
differencier le mesoderm.” He points out that it differs in its position in the entoderm 
