of Edinburgh , Session 1884-85. 
7 
with any tubular structure whatever. It disappears as the intestine 
is formed; but the authors have not recognised the fact that the 
periblast aids in forming the floor of the intestine, as I shall show 
below. They believe that a lumen exists in the part of the intes- 
tine formed in the region of the vesicle, from the time of the latter’s 
disappearance onwards, It is very surprising that, having thus 
grasped clearly the relations of the vesicle, Agassiz and Whitman 
have not seen its meaning ; they say that they are not ready to 
accept Balfour’s interpretation, but the views of Kupffer and 
Henneguy are still more unsatisfactory. I hope to show that 
Henneguy’s is the obvious and only morphological interpretation 
possible of the structure in question. 
I may he allowed to explain that I arrived at the view I hold in- 
dependently, not having seen any account of Henneguy’s results 
till my opinion had been formed ; at present I have only been able 
to see an abstract of Henneguy’s paper, and do not know if his 
foundation of fact was more or less firm than my own. 
Kupffer’s vesicle, then, as seen in a successful series of transverse 
sections through a herring ovum at the proper stage, is a hemi- 
spherical cavity, hounded above by the lowest layer of the blasto- 
derm, and laterally and interiorly by the periblast. The latter is 
simply the outer layer of the yolk containing nuclei, but not 
divided into cells. In the herring at this stage the hypoblast cells 
are not distinctly differentiated from the mesoblast. The notochord 
is well formed, not yet vacuolated, and, as usual, not sharply marked 
off from the hypoblast. The neurochord is present as a thick cord 
of cells derived from the epi blast, and containing no canal (see 
Plate I. fig. 1), 
It is clear, then, that Kupffer’s vesicle has the same relation to 
the embryonic layers as the invagination cavity in Elasmobranchs, 
Amphibia, and Cyclostomi (Petromyzon). The vesicle has no dif- 
ferentiated cell- walls of its own ; it is simply a depression in the 
periblast. The differences between its relations and those of the 
cavities with which I am comparing it are — first, its small extent, 
and second, the want of an opening to the exterior. As to the 
extent, the vesicle is a rudiment, a remnant of a larger cavity ; as to 
its opening, it is on a par with all the other cavities in the Teleostean 
embryo. The neural canal, the cavity of the otocyst, the cavity of 
