STRUCTURE AND DEVELOPMENT OF LOXOSOMA. 309 
area which Hatschek (17) finds to be invariably present behind 
the velum in Trochospheres. 
The shell-gland of Mollusca is commonly situated more 
posteriorly than the sucker of Polyzoa; it is not impossible 
that if it were once used for purposes of temporary attachment, 
it would be most conveniently situated in the centre of the 
aboral surface, in order to allow the cilia connected with the 
mouth to assume the position most favorable for nutritive 
purposes. The shell-gland might thus travel somewhat forward 
as a result of its new function, the cilia concerned in nutrition 
and locomotion alike becoming confined simultaneously to the 
ventral surface. 
Hatschek (17) has pointed out that in Polygordius the 
ciliary apparatus (velum and oral cilia) does not originate as 
a closed ring ; it appears first on the ventral side, and subse- 
quently completes itself dorsally. This mode of development 
is also characteristic of some Gastropod Trochospheres, as 
shown by Bobretsky (48) . Thus in the phylogenetic history 
of the Entoprocta, it is possible that when the sucker was 
situated more posteriorly, the velar ring was at first completed 
on its anterior side ; that as the sucker travelled forwards, it 
reached a point where it interrupted the velum dorsally ; and 
that finally, when it had arrived at a subcentral position, the 
posterior union of the two halves of the velum took place 
behind it, thereby increasing the efficiency of the nutritive 
cilia, when the animal was fixed. It is worth noting that in 
Dentalium, Kowalevsky (loc. cit.) describes the shell-gland 
as originating during the gastrula stage, whilst in Loxosoma 
the sucker similarly develops exceedingly early, and is func- 
tional whilst the brain, for instance, is still quite rudimentary. 
Although this early appearance of the sucker in Loxosoma 
may be merely due to the fact that it is an important embryonic 
organ, the period of its development, as well as its striking 
similarity to the young shell-gland of Mollusca, renders 
possible the view of its homology with the latter. The 
constancy of the occurrence of the shell-gland in larval Mol- 
lusca and of the sucker in embryos of Poly zo a (Ento- 
