INTESTINAL RESPIRATION IN ANNELIDS. 747 



and a tubular " buccal cavity" intervenes between mouth and pharynx. In Chsetogaster, 

 the large circular mouth reaches to the anterior end of the body (a prostomium being 

 practically non-existent), and opens immediately into the barrel-shaped suctorial 

 pharynx. 



Thirdly, the absence of dorsal setse, and reduction of the number of ventral setal 

 bundles. The dorsal setse, which in the other Naididse are frequently much elongated, 

 being as long as or sometimes much longer than the diameter of the body, would be 

 detrimental to the success of an internal parasite by very considerably restricting its 

 movements. The ventral setse also are absent over a considerable region in the anterior 

 part of the animal (segments iii.-v.). In C. orientalis the setse of segment ii., which 

 are longer than the rest, are directed anteriorly, not perpendicularly to the body- wall ; 

 the ventral setse in this species are "somewhat small compared with the size of the 

 animal." Such conditions, as more readily allowing of movement within the body of 

 the host, would be of advantage to an internal parasite. 



Fourthly, the incompleteness of the dissepiments. This is illustrated by the fact 

 that no sperm- or egg-sacs are formed, and that sperm-morulas ripen in the general 

 body-cavity, passing freely from one part of the body to another, even to the posterior 

 end of the last animal of a chain (48). This may be paralleled by the absence or in- 

 complete development of the dissepiments in boring Polychsetes, and by their incomplete 

 development at the anterior end in earthworms. This character is useful, inasmuch as 

 it allows the anterior end of the animal to be made firm and resistant by being dis- 

 tended with coelomic fluid, and it may, in the present instance, be correlated with 

 previous boring habits of parasitic ancestors. 



The commencing disappearance of the vascular system in the anterior part of the 

 body in (7. crystallinus (Vejdovsky, 54), and the entire absence of external annulation 

 in the genus, are also perhaps worthy of mention in this connection. 



There are thus a number of features, peculiar to Chsetogaster alone among the 

 genera of the Naididse, which seem to point to the existence of a parasitic mode of life 

 in the ancestors of living forms, or which at least are consonant with such a supposition. 

 If we accept this view, it is easy to understand the absence of antiperistalsis and 

 ascending ciliary action in the intestine in this genus ; these functions, existing origin- 

 ally in this as in other genera of Naididse, were given up on the assumption of the 

 parasitic habit ; and consequently are no longer found, though the genus has returned 

 almost completely to a free-living existence. 



It is possible that ChMogaster does not stand alone in having returned to a free life, 

 and a parallel is perhaps to be ' found among the Nematodes. I quote the views of 

 Hatschek and Hubrecht, as given by Kauther (39a) : — " Es ist fraglich [he is quoting 

 Hatschek] ob die parasitischen oder die freilebenden Nematoden dem Ausgangspunkt 

 der Gruppe naher stehen. Man mochte der ersteren Vermuthung den Vorzug geben 

 und annehmen, dass viele Eigentiimlichkeiten dieser aberranten Thiergruppe durch den 

 Parasitismus erworben worden sind, doch spricht auch manches fur die zweite Theorie. 



