ON THE GENUS LUCERNARIA. 273 
radial canals which communicate with each other at the 
border of the bell, up to which the lines of connection do 
not quite reach, and thus leave, as it were, an annular passage. 
In L. octoradiata these lines of connection are very regularly 
arranged. They always extend from the middle of one of 
the sides to the oral tube, in the direction of the longitudinal 
ridges of the stem, and the openings which represent the 
annular vessel are but small; in L. campanulata, on the con- 
trary, where the lines of connection can scarcely be detected 
from the outside, though their existence can be proved by 
the introduction of a probe, they are often not absolutely radial 
in direction, and the annular vessel is of considerable and irre- 
gular width. When the arms of the bell are united into four 
groups, a similar line of connection always divides one of 
these groups or arms of the first order into two secondary 
arms, as I have already noticed. Frey and Leuckart describe 
eight such pouch-like radial canals in L. quadricornis, but 
Milne-Edwards has already remarked that this must be a 
mistake, as in that species, as well as in all the others hitherto 
examined, only four partition-walls and radial canals exist. 
In the gastro-vascular system I have constantly found a 
clear granular fluid, which is kept in motion by the cilia, 
and the cavity is very much crowded in places by the before- 
described nematophorous follicles, the muscles, and the 
generative organs. 
8. Muscular system.—The very distinctly marked muscular 
system in Lucernaria is easily recognised ; it is composed of 
bundles of fibres running in definite directions, beyond which 
I have remarked no further structure. 
In the stem of L. octoradiata (fig. 18) there are found in the 
above-described longitudinal ridges (/), but lying free in the 
gelatinous substance, four cylindrical or flat bundles of muscular 
fibres (m), which arise below from the pedal disc, and suddenly 
terminate at the apices of the four points of the natatory sac. 
It is to these muscles that is due the great contractility of 
the stem in this species. In L. campanulata these muscles 
are entirely wanting; and in accordance with this the stem 
exhibits very little or no contractile power, so that sections in 
any direction can as easily be made in it as in the stalk of a 
plant. . 
Two sets of muscular fibres can be distinguished in the 
bell, a radial and a circular, but these belong, as in all Me- 
dusee, to the natatory sac alone. 
The radial muscular bands (m’) are eight in number, one 
running in the centre of each arm. At the apex of each 
VOL. III.—NEW SER. U 
