CHAETOPODA 157 
“siphon,” that is, a tube which opens at both ends into the 
alimentary canal. The siphon never contains food, and its 
function is probably respiratory. A similar 
structure runs from one part of the alimen- 
tary canal to another in the Echiuridae, 
and in Echinoids. The CAPITELLIDAE and 
some other families are without any blood 
system. In other worms the principal vessels 
are similar to those deseribed in Avenicola ; 
the red blood of some forms is due to 
haemoglobin dissolved in the plasma, in 
others the blood is green or almost colourless. 
The typical arrangement of the nephridia, 
one pair in each segment, is often interfered 
with. They usually fail in the anterior 
segments when there is a large pharynx, and 
in the tubicolous forms their number is 
usually much reduced: eg. eight pairs in Pe eee ae 
Terebella and six pairs in Arenicola. The hae ce 
genus Capitella are remarkable for having 
several pairs of nephridia in each segment, the number increas- 
ing in the posterior end up to six or seven pairs in this genus. 
The nephridia are themselves subject to much variation ; 
one nephridium may have several funnels, and may be con- 
nected by a tube with another, and sometimes the organ 
breaks up into small tubules. The whole arrangement recalls 
the excretory system of some of the earthworms described 
above. 
The ventral nerve cord of some of the tubicolous Poly- 
chaeta has its right and left half divaricated, and connected 
by numerous transverse commissures. ‘This is well shown in 
Serpula, and in a less degree in Sabella. Eyes are very generally 
present, and are usually confined to the prostomium.  Poly- 
ophthalmus, however, has a number of lateral eyes, a pair to 
each somite; whilst Branchiomma bears them on its branchial 
filaments. Otocysts, such as those of Avenicola, are rare. 
The Polychaets, with some exceptions, are dioecious. The 
generative organs are usually developed in relation to a blood- 
vessel, which no doubt serves to nourish them; in the Seden- 
