192 Proceedings of the Royal Physical Society 
the retraction changes have progressed to near the stage shown in 
6, pl. xxi. fig. 14 for Asterias rubens. The median brachium is almost 
opposite arm rudiment II, but the aboral brachiolarian notch, or space 
between arm rudiments I and II, is unduly great, probably because retraction 
was rendered imperfect by absence of fixation. The hydroccele pouches (see 
explanation of Fig. 1) are near the stage shown for A. rubens in 6, pl. xxi. fig. 13, 
that is, they have as yet increased little in size and show no trace of out- 
pouchings for the first pair of sucker feet. There is also an unduly wide 
interval (oral brachiolarian notch) between hydroccele pouches I and V, the 
first and last respectively of the series (see 6, p. 243). 
The specimen was not sufficiently well preserved to allow satisfactory 
histological examination, and in particular nothing could be made out regard- 
ing the skeleton, but on cutting sections it appeared that the larval mouth 
and anus were closed, that the wall of the alimentary canal was greatly 
swollen, and that in general the arrangement of the enteroccelic cavities was 
comparable with that found in the A. rubens brachiolaria at a corresponding 
stage. Practically all the typical larval ciliated processes, namely, preoral, 
dorsal median, anterior dorsal, posterior dorsal, postero-lateral, and post-oral 
were still recognisable on one side or the other, and they appeared to be 
cylindrical, not flattened, at the extremities. The ciliated bands, however, 
could not be made out. The brachiolaria is remarkable for its size, measuring 
nearly 2 mm. across the disc, that is, more than twice as much as that of 
A. rubens at a corresponding stage. 
At first the absence of mouth and anus, the general shape of the larva, its 
large size, and the possibility that it might come from a great depth, raised 
the question in my mind whether this was not a pelagic brachiolaria with 
abbreviated development, originating like the Solaster larva from a large 
yolky egg. Against this, it is to be noted that the characters of the brachia 
and sucker, the form and degree of development of the hydroccele pouches, 
the width of the oral and aboral brachiolarian notches, the presence of 
remains of the various ciliated processes, the closure of the mouth and anus, 
and the thickened condition of the hypoblast lining, all make up just such a 
complex of characters as we expect to find in the early metamorphosis of 
feeding brachiolariw like those of A. rwbens and A. vulgaris. On the whole, 
therefore, it seems in every way probable that our brachiolaria has been a 
feeding free-swimming one. Should the contrary by any chance prove true, 
we would have to infer that the yolky character of the eggs permitting direct 
development in this instance is an ever so much more recent acquirement 
than the similar condition in the eggs of Asterina (16), Oribrella (19), and 
Solaster (6). 
