LIMNOCODIUM (CRASPEDACUSTES) SOWERBIV. 365 
sules is indicated in a rare phenomenon, observed bj my 
assistant (Mr. Bourne), namely, the fusion of neighbouring 
tubular capsules, an instance of which is shown in PL XXX, 
fig. 1. 
Deformities and abnormalities of many kinds are not unfre- 
quent in the capsules, the marginal bodies themselves, and the 
tentacles. The vacuolated cells of the abumbral velar surface 
a])pearto stand out from the stratum in which they originally lay, 
and sometimes to form separate vesiculi, which may or may not 
fuse with the growing capsule of the refringent body. 
The adumbral ectoderm is complete and continuous beneath 
the tubular capsules, its circular muscular fibrils being traceable 
in every part beneath (that is, on the adumbral surface of) the 
tubular capsules (see PI. XXX, fig. 2 muse). 
The development of the velar canals or tubular capsules of 
the marginal bodies is, then, shown to be due to the fusion of 
vacuolated cells of the abumbral ectoderm of the velum. 
It remains to be shown what is the first origin of the capsule 
and of the refringent body itself. 
In PI. XXXI, figs. 10 to 20, I have represented a number of 
stages in the development of the refringent body, and it is perhaps 
as well at once to say that the refringent body is nothing more 
nor less than a modified solid tentacle. It consists in the fully 
formed state of a number of cortical cells enclosing four, six, or 
eight large refringent cells (PL XXXI, figs. 10, 11) . The cortical 
cells correspond to the ectoderm of a tentacle, and the medul- 
lary refringent cells to the endodermal axis of a tentacle, the 
whole group of cells being tightly pressed together so as to form 
a spherical refringent body. The cortical as well as the medul- 
lary cells of the free hemisphere of the body are highly refringent. 
There is not in the refringent body of Limnocodium any 
separate concretion^ any otolith in the strict sense of the term. 
The whole structure is purely cellular, and consists entirely of 
nucleated cells which can be isolated and recognised as distinct 
cell elements by means of reagents. 
In this respect the refringent body of the new Medusa appears 
to differ from any previously described organ of the kind. 
Certainly it differs entirely from any marginal body either of 
Leptomedusae or of Trachyline Medusae described by the 
llertwigs in their recent work ^ Nervensystem der Medusen.’ 
There seems very small justification for regarding the refringent 
body of Limnocodium as an auditory organ at all. It consists 
simply and solely of nucleated cells, the substance of which is of 
a higlily refringent nature. Refringent bulbs would be a 
sufficiently appropriate name for these bodies. The cells which 
correspond to the tentacle ectoderm, and which I term cortical 
