POLYCHHTA FROM THE N.E. PACIFIC. 967. 
upon the differences in size of the animals. Speaking generally, 
however, these two species are far more definitely distinguishable 
than any pair of species in Phyllochetopterus or in Cheetopterus. 
No less interesting than the diversity in form of the two 
species is their curious distribution. While W/. taylori is so far 
only known to oceur in the coastal waters of the N.E. Pacific, 
M. minuta has already proved to have a much more extended 
range. The type-specimens come from the subtropical regions 
of the North Atlantic, but I was surprised to find a Jeso- 
chetopterus of common occurrence in Torres Straits, which is, 
without doubt, identical with JZ. minuta. It will prove, I 
venture to predict, a widely spread Indo-Pacific form, and its 
absence from previous descriptions only illustrates the difficulty 
of obtaining a representative idea of a Polychet fauna from 
general collections. Though J. minuta may be found in the 
Tndian Ocean, it is less likely to turn up in the Red Sea and the 
Mediterranean, where so much attention has been given to the 
obseurer forms of Polychet worms, and the distribution will 
probably remain discontinuous in type. 
The Position of Mesocheetopterus in the Family. 
It can hardly be doubted that a close relationship exists 
between Chetopterus and Mesochetopterus. The structural 
differences between them correspond closely with the different 
kinds of tubes which they occupy. Chetopterus possesses a much 
wider tube, in which it fits very loosely, and it is for this reason 
that the excessive and bizarre modifications of the median region 
have been produced. Adhesive organs are needed to maintain 
the position of the worm in the tube, and these are formed by 
the fusion of the notopodia giving a cup-shaped sucker. Ciliary 
action alone would be too feeble to produce an efficient circulatory 
current in so wide a space, and hence the fans of the 14th—16th 
segments exist. The notopodia of the anterior and posterior 
regions are concerned in the movement of the animal up and 
down in its tube. They must be long enough to touch the walls 
and so attain to much greater dimensions than in Mesocheto- 
pterus. The greater or lesser length of the parapodia causes, as 
we have seen above, some difference in the arrangement of the 
setie. 
Chetopterus possesses, too, a complicated method of feeding, 
which is responsible for further ditferentiation in its external 
structure. This method has been lately described by Enders in 
detail (4). The long aliform notopodia of the 12th segment and 
the dorsal cup of the 13th segment alike aid in separating food 
from the respiratory current, and compacting it into masses 
which are swept forward in the ciliary groove to the mouth. 
The middle region, then, fulfils a double function in promoting 
the circulation of water in the tube and collecting food, different 
segments bemg specialised for each task, In consequence of this 
