No. 2.] THE HAVYPOPHYSES OF AMIA. 503 
ment of the adhesive organ and the neuropore. Its tissue is 
continuous with that of the neuropore and remote from the 
stomodzeum. 
2. It loses its connection with the ectoblast and the neuro- 
pore tissue and comes to lie between the infundibulum and the 
dorsal wall of the alimentary canal. 
3. In changing its position it never unites with the ento- 
» blast. 
4. Its cells contain smaller yolk granules than those found 
in entoblast cells and may consequently be readily distin- 
guished from cells of entoblastic origin. This is true of the 
earliest stages. . 
5. In 22 mm. larve it shows a division into a number of 
elongated vesicles with well marked cavities. 
6. By the penetration into it of nervous tissue (neuroglia, 
Haller, 1896) from the infundibulum, it becomes divided into 
an anterior and a posterior portion. 
7. It is richly supplied with blood-vessels. 
DISCUSSION OF RESULTS. 
From the foregoing account it appears that the hypophysis 
of Amia shows in its development a striking likeness in some 
features to the hypophysis of Acipenser as described by v. 
Kupffer (1893). 
In both cases the hypophysis does not originate from the - 
ectoblast of the stomodzeum or near it, but from the ectoblast 
at the anterior end of the neural tube in connection with the 
anterior neuropore (mediane Riechplatte and lobus olfactorius 
impar of v. Kupffer). In both cases there is an adhesive 
organ intervening between the fundament of the hypophysis 
and the stomodzum. In both cases the hypophysis detaches 
itself from the ectoblast and neuropore and takes up a posi- 
tion posterior and dorsal to the stomodzum. In both cases its 
further history is not essentially different from that of other 
vertebrates (Haller, 1896). 
Acipenser differs from Amia in that its hypophysis is, ac- 
cording to v. Kupffer, tubular in an early stage. This hypophy- 
