INTESTINAL RESPIRATION IN ANNELIDS. 811 



or Chaetopterus. A peculiarity about these carmine experiments, however, deserves 

 notice. The experiment may fail altogether to demonstrate an entering current at the 

 anus, and this although a violent ascending ciliary action is going on all the time; — 

 compare the account of Tryj^ctnosyllis zebra ; the animal seems to try to get away from 

 the carmine. In other cases, though the carmine enters, it seems never to ascend very 

 far, and soon to be rejected (compare the accounts of Poly cirrus and Spirorhis). It 

 would appear that the posterior end of the intestine is not fitted to deal with solid 

 particles, and that if possible it avoids receiving them. 



(2) Antiperistaltic contractions of the intestine are even commoner than ascending 

 ciliary action. Of the nineteen families before enumerated, the Opheliida3, concerning 

 which I have no note, are to be excluded ; of the rest, it appears to be absent only in 

 the Chsetopteridfe, of which, however, very few specimens were examined, and probably 

 in the Aphroditidse, which have (at least Aphrodite aculeata has) special arrangements 

 for intestinal respiration ; for practical purposes it may also be reckoned as absent in 

 the Phyllodocidse. It occurs therefore in fifteen families out of eighteen. 



That a phenomenon which occurs so widely and constantly, in so large a group as 

 the Polychseta, has some essential significance, will be immediately admitted. That in 

 a number of forms, both of Oligochseta and Polychseta, the contractions are partly 

 circulatory in their eff"ect, driving forward the fluid in the perienteric plexus or sinus, 

 is obvious ; that they are even mainly circulatory in some (Enchytrseidse, Sabellidse, 

 Serpulidse, etc.), is no less true. These facts have already been insisted on, and have 

 an essential place in the speculations on the manner of evolution of the vascular system 

 which have been already put forth. 



But that this is the fundamental meaning of antiperistalsis appears to me to be an 

 error. The question is discussed at length on pp. 779-781 ante. The association of 

 antiperistalsis with ascending ciliary action ; the fact that the waves obviously cause 

 (in most cases) a considerable constriction of the alimentary tube, and are hence much 

 more violent than is necessary merely to propel the fluid in the intestinal vascular net- 

 work ; and that the function of peristalsis generally in the animal kingdom, including 

 the groups from which the Annelida may be supposed, directly or remotely, to be 

 derived, is to propel the contents of a tube along its lumen ; are among the arguments 

 there adduced in support of the view that the reversed peristalsis of the intestine had 

 originally, and in many cases has still, a respiratory or at least ingestive significance, 

 and that this is the essential meaning which underlies the phenomenon. I may here, 

 in addition, recall the periodical opening and closing of the anus (gulping action), and 

 especially the direct association of this with the starting of an antiperistaltic wave from 

 the anus {Leptoneris vaillanti), which in Polycirrus and Amphitrite may be fairly 

 described as a regular gulping and swallowing of " mouthfuls " of water. 



(3) Associated with this use of the anal end of the body are several other phenomena 

 which, like that, induce us to believe that this extremity has in many cases a greater 

 physiological importance than would naturally be attributed to it. Eye-spots occur by 



