614 BP. Hy UGRAVELY: 
and 5, G.). Thus in the Polydoridz nototrochs are present 
on every segment after the second, but I have been unable to 
determine with certainty the presence of any ciliaatall between 
the lateral rows of long cilia; gastrotrochs, on the other 
hand, are confined to segments 3, 5, 7, 10, 13, 15, 17; ater 
and although complete, the cilia across the middle of the 
body are very small and inconspicuous. In the Spionidee 
nototrochs appear to be entirely absent in many species, but 
gastrotrochs occur on every segment, broken into short 
lengths as shown in fig. 5, and always with some long cilia at 
each end, 
Another peculiarity of the ciliation of the larvee of the 
Spionide and Polydoridz would appear to be the presence of 
a slight gap in the course of the telotroch in the mid-dorsal 
line. This gap is very small and often difficult to determine 
with certainty, but by careful watching of the cilia I have 
been able to find it in every species in which I have paid 
special attention to the point. 
Some GENERAL CONSIDERATIONS ON THE CLASSIFICATION OF 
Potycua#t LARvV2z. 
Polychet larvee were originally arranged by Miiller and 
others in the groups Monotroche, Telotroche, Meso- 
troche, Polytroche, and Atroche, according to the 
ciliated bands borne by them. Claparéde (1863, pp. 84-88) 
has pointed out some of the difficulties which render this 
classification unsatisfactory, and has proposed to divide these 
larve primarily into the Metacheta, with provisional sete, 
and Perrenichete, without them, and to subdivide these 
classes according to the nature of the ciliated bands—the 
former into Gastrotrochew, Nototroche, and Amphi- 
troche, and the latter into Cephalotroche, Polytroche, 
and Atroche. 
Agassiz (1867, p. 252) has briefly pointed out the value of 
the class Metacheetee, and the almost certainly provisional 
nature of the rest of this classification, but does not make any 
