178 
the tail, but are not usually recognisable in the first few 
segments, which are very small, and probably only 
recently formed. Typical giant cells may be seen in some 
of the middle and posterior caudal segments, but careful 
search has failed to reveal them in a_ considerable 
number of the tail segments of the two worms examined. 
In the mid-dorsal region of the cord there are one, two 
or three giant fibres seen in section (figs. 55, 36). 
Anteriorly there is only one, in the middle region of the 
body either two or three, and in the tail usually one. At 
first sight they appear to be tubes with distinct walls and 
homogeneous contents, but by suitably staining, longi- 
tudinal sections the contents are found to be distinetly 
fibrillar, but the fibrils are so fine that they are scarcely 
recognisable in transverse sections. The wall of the tube 
is composed of nucleated cells. These giant fibres arise 
from the giant cells in the cord. This is not so easily 
demonstrated in Arenzcola marina as in A. grubii, in which 
the cells and fibres are larger (fig. 53). In the latter 
species each giant cell gives off a stout process, which runs 
in a somewhat sinuous course through the fibrous part of 
one side of the cord towards the dorsal surface. Shortly 
after leaving the cell the process gives off one or two 
branches, which pass into and ramify in the fibrous 
part of the cord. The process then passes dorsally 
and enters either the median or, more usually, one of the 
lateral giant fibres. The single median fibre present in 
the anterior region of the cord arises from the first giant 
cell situated just behind the union of the cesophageal con- 
nectives. In those portions of the cord where two or three 
giant fibres are present there are connections between 
them, generally in each cheetigerous annulus. Branches 
from one or other of the giant fibres may occasionally be 
traced into the spinal nerves. 
