622 F. H. GRAVELY. 
In the Spioniformia all the segments are commonly 
developed in series, as in Ophryotrocha amongst the 
Nereidiformia, there being no distinct ‘‘ primary ” segments ; 
and where any such segments can be recognised they are 
very few in number (e.g. three, in Polydora ciliata—see 
Leschke, 1903, Pl. VI, figs. 3-5). Moreover the appendages 
do not develop any adult characteristics until a large number 
of segments have been formed, being preceded by tufts of 
long and often deeply-toothed provisional set,! those situ- 
ated on the first segment being longer than any of the others, 
and often appearing before the commencement of segmenta- 
tion in species with a free-swimming trochophore stage (e. g. 
Claparéde’s Polydora [= Leucodora] larva, 1868, PI. 
VII, figs. 4,5; and Fewke’s Spio larva, 1885, PI. II, fig. 3). 
There is at present no evidence to show that these larval 
sete are in any way the result of the correlation between 
adult and larval characters ; they appear rather to be purely 
larval organs developed for purposes of defence or buoyancy. 
Their defensive function has been urged by Fewkes (1885, p. 
168), because immediately upon irritation, either by a fixing 
fluid or an encounter with some obstacle to progression, their 
usual position flat against the side of the body is changed for 
an erect one, when the larva bristles with spimes in a manner 
that might well defy any delicate pelagic organism. In spite 
of its bristles, many of which, though somewhat slender, con- 
siderably exceed the length of its body, several of the Port 
Krin Spio larvee (see Gravely, 1909) have, however, been seen 
in the intestine of the peculiarly delicate-looking larva of 
Magelona; and in view of this it may be suggested that the 
longer and more slender forms of provisional sete assist in 
floating, as do the spines of diatoms, for any stimulus which 
causes the erection of the spines also causes the larva to cease 
from its regular movements of progression and so to sink, 
' Claparéde and Meeznikow’s ‘ Unbestimmte Spionidenlarven ’ (1869, 
pp. 13-15, pl. xiii, figs. 1—1 F’) does not develop these {provisional set, 
but I know of no other case of their absence in a free-swimming 
Spionid larva. 
