THE METAMORPHOSIS OF ECHINODERMS. 71 
entirely obliterated, since in a young Luidia in which the disc 
measures about 5 mm. in diameter, I find no trace of any 
anus at all. 
This Luidia presents certain other features which merit a 
brief description. The dorsal sac is very conspicuous on the 
aboral side of the water-pore, and a considerable portion of the 
dorsal organ projects into it; it is quite distinct from the 
ampulla, as well as from the rest of the celom, and is evidently 
identical with the space which Ludwig found (18, p. 159) 
containing the aboral termination of the dorsal organ. Imme- 
diately below this sac, at the level of the original dorsal 
mesentery (persisting at this stage transverse to the axis of 
the cesophagus and stomach), the genital cord starts from close 
by the dorsal organ (though I cannot from my own observa- 
tions assert the connection of the two structures) ; but it is 
still very short, and cannot be followed far round the disc. 
For the most part the original mesentery is fragmentary (at 
any rate in my sections), but enough of it remains to indicate 
that it follows the growth of the hepatic ceca into the arms— 
these ceca being derived from the ten stomachic pouches seen 
in fig. 24 (compare figs. 25 and 26). This extension, however, 
of the right body-cavity into the arms is probably to be re- 
garded as secondary, as I shall explain later. There is much, 
in my opinion, to point to the conclusion that the more primi- 
tive line of division of the two cavities is marked in the adult 
by the ring of the genital cord; but the proof of this is 
indirect, and cannot be entered upon here. 
2. Brachiolaria. 
I have selected Brachiolaria as the representative of the 
second type of Asterid development because it is the form 
which I have had most opportunity of studying; but the same 
general plan of development is found in many true Bipin- 
nariz, as well as in the larva of Asterina, which is, as it seems, 
only a modified Brachiolaria (1, p. 62; 15, p. 154), 
The general external features of Brachiolaria have been 
admirably figured by Agassiz (1) ; while the internal anatomy 
