766 
ROBERT PAYNE BIGELOW 
the HertiwgS; so far as it goes. I have noticed additional feat- 
ures which, however, may be pecuHar to the species studied. 
On the upper surface of the hood immediately above the base 
of the rhopalium there is a funnel-shaped dorsal sensory groove, 
fig. 13, s.g. It differs from the groove in Chrysaora in being much 
smaller and in sinking into the mesogloea not more than half so 
far. Hesse ('95) confirms Elmer's statement that a similar groove 
is to be found in P. noctiluca, but found no sensory epithelium 
in it. 
While, as stated by the Hertwigs, the general surface of the 
niche is covered by the common ectoderm of the body, I find on 
the rhopalial ridge, which is short and low, fig. 15, a peculiar 
epithelium which distally passes into the ectoderm of the rho- 
palium and proximally extends along the wall of the niche towards 
the muscle band. My material is not sufficient for me to make 
out very clearly the structure of this layer. It is twice the height 
of the ordinary epithelium. The outer part of the layer is made 
up of the thicker parts of the cells, which stain deeply, while the 
inner portion seems to be composed of processes running from 
the cells to the surface of the mesogloea. There seems to be a 
loose network of nerve fibers intermingled with these processes. 
At any rate, my preparations show very clearly a number of large 
ganglion cells scattered through this layer just below the deeply 
stained part, fig. 16, g. When I speak of these cells as large, I mean 
that they are many times larger than the ordinary epithelium cells. 
They are provided with comparatively large nuclei and are appar- 
ently bipolar, the stout processes running parallel to the rho- 
palium. Cells of this kind similarly situated have been found by 
Hesse ('95) in Rhizostoma, and, as he says, probably constitute 
a nerve center in each sensory niche. 
In the thick layer of nerve fibers on the basal portion of the 
rhopalium there is a cluster of nuclei like those in the sensory 
epithelium above them, fig. 14, n. These probably belong to very 
small ganglion cells. At each side of the ridge at the base of the 
rhopalium the layer of nerve fibers comes into contact with the 
endodermal lamella, and the latter, which otherwise has the same 
relative position as in Chrysaora, may be traced in this species 
