INTESTINAL RESPIRATION IN ANNELIDS. 813 



physiologically from the alimentary system. Siphonostoma and Terebella may also 

 be compared. 



In some ways, however, the course of evolution has led in the Serpulimorpha to a 

 more specialised state of things than that seen in the Enchytrajidfe. In the Enchy- 

 trseidse the "sinus"' is probably a close network [cf. note on p. 778) ; in the Serpuli- 

 morpha it appears to be really a continuous sinus ; this is, according to the views put 

 forward in the present paper (p. 778), a secondary condition, which has been 

 developed out of that of a network by the fusion of adjacent channels. Secondly, the 

 antiperistalsis is mainly circulatory in its effect ; it usually constricts the lumen of 

 the gut but little, and can therefore have only a slight influence in moving the fluid 

 in the interior of the intestinal tube. This is to be both compared and contrasted 

 with the results of evolution in the Naididse and Tubificidge ; as in these latter the 

 antiperistalsis becomes limited in its effects to the dorsal vessel, the main circulatory 

 channel in the intestinal region, and the cavity of the gut is but little affected, 

 so here also the effects of the antiperistalsis are confined to the sinus, the main 

 component of the circulatory system in this region, and the lumen of the gut is 

 affected only slightly. 



(6) The phenomena observed {vide ante) in one specimen of Nereis pelagica are 

 interesting as indicating the occasional persistence of a primitive condition of things in 

 a group where the vascular system has progressed to a high evolutionary level. In 

 this specimen antiperistalsis began at the anus, travelled forwards for a few segments, 

 and was then continued on without interruption as a contraction of the dorsal vessel. 

 The same wave was at the hinder end a general antiperistalsis of the intestinal wall, 

 further forwards a contraction of the dorsal vessel, — a temporary relapse to an earlier 

 physiological condition or stage of evolution (cf. yEolosoma and the Enchytrseidse), 

 and indicative of the original unity of the two phenomena. 



(c) Eusiillis tuhifex is also noteworthy. In the Syllidae, according to Fuchs (19), 

 the dorsal vessel is unconnected with the intestinal canal, although it is sunk within 

 a fold of the gut- wall ; though there is thus apparently a close spatial relation, complete 

 anatomical differentiation has taken place, and the dorsal vessel is not merely a special 

 channel in the intestinal plexus. On referring to the foregoing observations, it will 

 be seen that in the Syllidse no relation was found to exist, as a rule, between antiperi- 

 staltic and vascular contractions — indeed, antiperistaltic contractions were not often 

 visible ; the physiological condition corresponds with the anatomical. In a specimen 

 of Eusyllis tuhifex, however, the dorsal vessel contracted whenever the intestine 

 contracted ; the physiological unity was in some degree preserved — a condition which 

 again points back to a stage where intestinal and vascular contractions were not only 

 coincident in time, but undifferentiated parts of the same phenomenon. 



{d) Finally, there is a large number of families in which the condition is that of the 

 higher aquatic Oligochaeta. In Autolytus (Syllidse), Castalia (Hesionidge), the Eunicidse, 

 Ariciidse, Spionidse, and Arenicolidee, both anatomical and physiological differentiation 



