Abnormal Specimen of Antedon rosacea. 197 
a distance equal to the thickness of seventeen sections, and 
again communicates with the exterior through the funnel- 
shaped projection already described (figs. 2 and 5, fp.). That 
this canal was a modified ambulacral groove is shown by the 
epithelial cells which line it. They are precisely similar 
to those which line the ordinary ambulacral grooves; and 
further evidence in the same direction is afforded by the 
presence in its walls of numbers of the deeply staining proble- 
matical bodies which are invariably seen in sections of the 
ambulacral grooves of this species. Beneath the epithelium 
of the ambulacral grooves the nerve-band can be recognized 
without difficulty in most sections. The circular water- 
vessel (fig. 5, c.w.v.) and radial water-vessels are also present, 
and from the former a considerable number of water-tubes 
‘ (fig. 5, w.t.) depend into the body-cavity. Water-pores 
traverse the body-wall in all the sections and are abundant on 
the interambulacral area, marked with an asterisk in fig. 1 
(see also fig. 3, w.p.). -The skeletal and axial nervous 
systems present in the normal disk are entirely absent in the 
supernumerary one; so also is the central plexus. 
The interesting question now arises—What was the mode 
of origin of the supernumerary disk? In answer to it two 
hypotheses may, I think, be advanced :— 
1. ‘That the supernumerary disk originated as a bud from 
the normal disk. 
2. That it is the result of incomplete evisceration. 
In favour of the former hypothesis is the intercommunica- 
tion of the body-cavities of the two disks—a condition of 
things one would expect to find in a budding organism. 
Against it is the entire absence of arms, skeleton, and axial 
nervous system in the supernumerary disk. ‘The compara- 
tively large size attained by the supernumerary disk and the 
fact that the remaining systems of organs had attained their 
adult condition add importance to this objection. <A still 
weightier objection lies in the fact that, so far as I know, the 
formation of a bud has never been observed in any Kchino- 
derm. 
I am indebted to Prof. Marshall for the second hypothesis, 
and it appears to me to explain the facts most conclusively. 
Though Antedon rosacea has never been proved to evisce- 
rate spontaneously, eviscerated specimens frequently occur in 
dredgings; and the experiments of Prof. Marshall* and 
Mr. Dendy ¢ have shown that evisceration may be and often 
is followed by complete regeneration of the visceral mass. 
* “On the Nervous System of Antedon rosacea,” Quart. Journ. Mier. 
Sei. xxiv. (1884) pp. 507-548. 
+ “On the Regeneration of the Visceral Mass of Antedon rosacea,” 
