144 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY. 
to me to be that of Semon, who has suggested that they serve, 
when filled with fluid, as they normally appear to be, to protect 
the central nervous system against injuries incident upon violent 
contractions of the body-wall. The absence of the stainable 
coagulum in the canals, which are lined with a flat epithelium like 
that of the body-cavity, distinguishes them from the genuine haemal 
lacunae. The fact that the hyponeural canals are closed and 
isolated vessels indicates, at any rate, that they cannot be of value 
as circulatory organs. 
2. PERIPHERAL NERVES. 
: a. Nerves arising from the Nerve Ring. 
There are in Caudina fifteen nerves arising from the posterior 
part of the abaxial side of the nerve ring and ten buccopharyngeal 
nerves, which spring from the axial side of the nerve ring imme- 
diately behind the circular furrow in its anterior face. 
(1) The tentacular nerves resemble in their essential features 
those of other holothurians in which they have been described. 
Each nerve (Plate 4, fig. 48, 7. ta.) begins as a wide sheet, which is 
crescentic in cross section, the concavity being directed toward the 
axis of the tentacle. It is thickest at the base of the tentacle, 
gradually diminishing toward the tip of the tentacle in conse- 
quence of the distribution of branches from it to the epithelium. 
The histological conditions of the tentacular nerve resemble those 
of the nerve ring and the outer band of the radial nerve. An 
anterior prolongation of the epineural ring canal accompanies 
each tentacular nerve for a short distance on its axial side (Plate 3, 
fig. 42). 
From the tentacular nerve are distributed (a) solid branches of 
nerve fibrillae, containing ganglionic cells, which run directly to 
sensory buds (Plate 2, fig. 16), and (b) isolated fibers, which are 
connected with subepithelial ganglionic cells; these in turn are 
doubtless connected with the sensory epithelial cells. These 
various structures have already been described in connection with 
the epithelium of the integument. 
Though I have spent much time in endeavoring by the Golgi 
method to ascertain the manner in which the members of this 
group of neurons are connected, my efforts have been unsuccess- 
