68 HENRY BURY. 
stage, from which it will be seen that this separation occurs 
in the same interradius as the water-tube. Already the 
pouches of the hydroccel have increased in number to twenty- 
five (they are much more conspicuous in sections than in ex- 
ternal views), but there is no pause in this condition, fresh 
pouches being rapidly and continuously formed between the 
terminal one and the next adjoiningit. In this way the hydro- 
coel spreads over the surface of the larva until its five main bran- 
ches come in contact with and fuse with the five excrescences 
(see figs. 20 and 22) which contain the terminal plates ; but it 
is to be noted that this is effected without that rotation of the 
two series of organs noticed by Ludwig in Asterina (15) ; the 
anterior dorsal pouch of the hydroccel becomes enveloped by 
the anterior terminal plate, and the pouch nearest the anus by 
the corresponding terminal excrescence. 
Fig. 21 will also serve to illustrate the relation of the anterior 
and posterior body-cavities on the left side in later stages. It 
will be seen that the “ ventral horn ”’ of the posterior cavity 
has there grown so far round that it forms a long thin mesentery “/ 
with the ventral wall of the anterior cavity. The dorsal wall 
of the latter is marked, as we have seen, by the line of the 
water-tube ; and along this line, at the close of larval life, a 
secondary separation of the two cavities is effected by the forma- 
tion of a septum. It will at once suggest itself that in the 
remnant of the anterior cavity thus enclosed we have the 
equivalent of the ‘‘ axial sinus”’ of the adult (see 5, p. 37), and 
a study of its subsequent history shows that this is the case. 
It is also important to notice that this sinus is bounded on 
both sides by the left body-cavity. This is important for my 
theory, and though confirmed by MacBride (19, p. 433) is 
opposed to the statements of some other observers; Ludwig 
(15, p. 87) states that in Asterina the ventral wall of the sinus 
is part of the original mesentery separating the right and left 
cavities, while Semon (82, p. 37) regards the whole mesentery 
containing the water-tube as a part of the dorsal mesentery 
of Synapta. 
No portion of the posterior body-cavity ever intervenes 
