314 
SIDNEY F. HARMER. 
adult, it is almost inconceivable that it should not attach itself 
by the end which bears the sucker. On the theory that the 
process is a mere metamorphosis, the fact that the larva should 
fix by its oral face, and that it should be thereby obliged to 
develope a new vestibule, mouth, and anus at its primitively 
aboral end, is an occurrence of the most extraordinary nature. 
The metamorphosis of a more or less bilaterally a symmetrical 
Echinoderm larva to the radiately- arranged adult form is not 
a parallel case. The sexually mature Pedicellina has pre- 
cisely the same general arrangement of its organs as its larva, 
whose most direct way to become adult would consist in the 
development of a stalk and budding stolon at the aboral end, 
with the possible atrophy of the dorsal organ. The whole life 
cycle of the Entoprocta appears to me to consist in a division 
of labour, leading to an alternation of generations ; the indi- 
viduals produced from the egg having given up the sexual 
function, but having retained many of the ancestral features 
which favour a free pelagic existence; the individuals produced 
by budding having, on the contrary, retained the generative 
function, but having acquired a sessile habit. The organisa- 
tion remains, however, fundamentally the same in larvae and 
adults alike, and both kinds of individuals are capable of the 
production of buds, one of the most characteristic features of 
the Polyzoa. Barrois’ account probably shows that the larval 
Pedicellina produces a single bud instead of two, as in 
Loxosoma. PI. II, fig. 14 (loc. cit.), represents a stage some 
time after the fixation. In this individual can be distinguished 
a disc of attachment, and connected with it a cylindrical body, 
at whose apex is a young vestibule with its tentacles. The 
proximal portion probably represents part of the larva, which 
is fixed by its oral face ; the larval organs have degenerated, 
but the bud is already well developed. The nature of this 
process is here less obvious than in Loxosoma owing to the 
absence of any constriction separating the bud from the larva. 
Of all the Ectopr octa, with the details of whose development 
we are acquainted, Cyphonautes, the larva of Membrani- 
pora, appears most easily comparable to that of the Ento- 
