STUDIES ON POLYCHAT LARVZ. 613 
the original mouth of the trochophore opens, not direct to the 
exterior, but into a “ vestibule’’ closed in by the lips. It is 
interesting to note that in Spionid D, in which the anterior 
parts of the lips are prolonged and specially modified, the 
part of them situated posterior to the prototroch is much 
smaller than in either of the more typical forms, Spionid A 
and Polydora A, with the result that the posterior margin 
of their ciliated areas extends from its lateral termination 
inwards and forwards, instead of inwards and backwards, 
as in the other two forms described, the whole of the lips 
appearing to be drawn forwards in every possible way. 
The presence of the vestibule in front of the mouth in 
larvee of the Spionidz and Polydoridz causes of necessity a 
ventral gap in the prototroch; and, correlated with this 
perhaps, we find—apparently with equal constancy—an ex- 
tensive dorsal break in its continuity. When the prototroch 
is thus confined to the sides of the head, its efficiency as an 
organ of locomotion must be seriously diminished, even when, 
as in some of these larve, its length is increased by lateral 
extension on to the bases of the tentacles, and this no doubt 
accounts, in some measure at least, for the importance of the 
telotroch in these larve, its cilia being at least as long as 
those of the prototroch, and often longer—a great contrast 
to their usual insignificance and frequent absence in the 
Nereidiformia, in which the prototroch appears to be always 
complete. It should be noted in this connection, however, 
that the most elongated of the Nereidiform larve that I 
have examined—that of Nephthys—also bears a more power- 
ful telotroch than prototroch. 
In Spioniform larve with the prototroch confined to the 
sides of the head too, interparatrochs are invariably present, 
and have frequently, if not always, become very definitely 
specialised, the greater part of their strength being concen- 
trated near the sides of the body, where the cilia are very 
much longer than they are across the middle; they are also 
frequently confined to either the dorsal or the ventral surface, 
and even here their continuity is often broken (see figs. 3 
