INTESTINAL RESPIRATION IN ANNELIDS. 755 



nected with tln' alimentary wallas far as the oesophagus; it is again attached to the 

 dorsal surface of the posterior part of the pharynx, becoming free once more before it 

 divides. It is continued as two lateral vessels, one on each side of the buccal cavity, 

 which unite ventrally in the ventral vessel. 



A pair of contractile lateral commissural vessels are present in the oesophageal region. 



'i'he ventral vessel is separated from the alimentary canal, with wliicli, however, it 

 is connected by a number of wide communications ; these in the intestinal region, and 

 presumably in the other regions also, pass from it to the mid-ventral sinus of the 

 alimentary wall. 



It will be remembered that in CIiMogaster orientalis and C inmjahensis the 

 alimentary canal exhibits no antiperistaltic movements, and there is accordingly no 

 question here of any relation between this phenomenon and the vascidar contractility. 

 I have nevertheless described the anatomy of the circulatory apparatus in the same way 

 as has been done for previous forms, because the anatomical condition seems to form a 

 link in a natural series. 



There is, again, a well-marked gut- plexus, specialised here anteriorly into a regular 

 series of lateral channels ; but a dorsal vessel is here distinct throughout the length of 

 the animal ; a separate ventral vessel is present, independent of the gut- wall ; and, in 

 addition, there is a pair of contractile lateral commissures. The blood, as in the 

 ^olosomatidaj and many Euchytra^ids, is colourless. 



The subsequent argument may here be so far anticipated as to say that I propose 

 to consider the vascular system in the Chtetopoda as a specialisation of and a develop- 

 ment from a primitive lacunar network or plexus in the gut-wall ; and the contractility 

 of the vascular system, which throughout the group is manifested pre-eminently in a 

 postero-anterior peristalsis of the dorsal vessel, as a derivative of the contractility of the 

 alimentary canal ; which latter is also, in the aquatic and presumably more primitive 

 forms, in the greater portion of its extent a postero-anterior peristalsis. 



I have previously given reasons for supposing that the genus Chietogastcr originally 

 posse.ssed a postero-anterior peristalsis of the gut, like its congeners ; and that this 

 disappeared during an endoparasitic stage in the history of the ancestors of present-day 

 forms. I have here to remark, that though the general antiperistalsis of the gut-wall 

 may be supposed to have disappeared during the parasitic stage, as being useless for 

 respiratory purposes, it would nevertheless be perpetuated as far as regarded that 

 longitudinal strip of the alimentary tube which contained the now ditterentiated dorsal 

 vessel, in order to subserve its second function, the circulation of the blood ; and hence 

 we now have what is, in these aquatic forms, the anomaly of a contractile dorsal vessel 

 with a quiescent gut-wall. 



Other Naidida'. 



The Naidida) in general show a very interesting transition stage in the relations of 

 the contractions of the vascular system to those of the alimentary canal. In the families 



