STUDIES ON POLYCHAT LARVA. 623 
and under these circumstances this may be the direct cause 
of the erection of the spines. 
That the appearance of these provisional setz in the 
trochophore is primitive seems hardly likely in view of the 
evidence, furnished by Polygordius, of the acquisition by the 
ancestors of the Annelida of a vermiform body before the 
appearance of any parapodial structures ; it is more probable 
that they first appeared in the larva of some ancestor of the 
Spioniformia. At first, we may suppose, there was a tendency 
in these forms towards the precocious development of the 
normal sete of the adult; and this was favoured by natural 
selection for purposes of defence or buoyancy in the water. 
In the presence of these sete throughout the order Spioni- 
formia'—with the exception of the Chetopteride and 
Claparéde and Mecznikow’s ‘ Unbestimmte Spionidenlarven ’ 
(see footnote, p. 622)—-we have then some definite embryo- 
logical evidence, independent of adult characteristics, that 
this order originated from some definite ancestral stock dis- 
tinct from that of the Nereidiformia. Claparéde, it is true, 
includes the larva of Odontosyllis gibba in his list of the 
Metacheetz (1863, p. 87), but I think it may reasonably be 
doubted whether the first-formed sete of this larva can be 
definitely grouped as provisionaland distinctly different from 
their successors. 
There is little evidence, however, to show that in the 
provisional setz of the Spioniformia we have a better represen- 
tation to-day than in their permanent sete of the sete 
* The larve ascribed by Fewkes to the genus Aricidea (1885, pp. 
174-176, pl. ii, figs. 4-6, and pl. vi, figs. 1 and 10), possess provisional 
sete like those found in larve of the Spioniformia. I have been unable 
to consult any account of the adult characteristics of this genus, but 
the name suggests a close relationship with Aricia, a genus included 
by Benham (* Cambridge Natural History,’ vol. ii, p. 321) in the order 
Nereidiformia. Fuchs (‘Die Topographie des Blutgefasssystems der 
Chatopoden,’ Jena, 1907, p. 12), on the other hand, refers the Ariciide 
to the Spioniformia. Fewkes’ larve of Aricidea are only referred to 
that genus provisionally. It would be of great interest to know defi- 
nitely whether free-swimming larve of the family Ariciide do or do 
not bear provisional sete. 
