784 Professor j. stephenson on 



the intestine ; (3) respiration through the general intestinal epithelium, in the way 

 described for the aquatic Oligochseta. 



(l) Respiration hy means of the anterior 'part of the alimentary tract [especially 

 by oesophageal diverticula). 



QuATREFAGES (37) stated so long ago as 1850 that in certain species of Polychseta 

 the digestive tube appears to aid the skin in carrying out the function of respiration. 

 A small Syllid was seen frequently to swallow large gulps of water, which it subsequently 

 returned by the mouth a little at a time ; and the same, but in a less pronounced 

 degree, was seen to happen in Dujardinia. On one occasion the intestine of a living 

 specimen of Hesione pantherina was observed to be partly distended with gas, which 

 was from time to time expelled either by the mouth or by the anus. The same author 

 subsequently (38, p. 70, vol. i. ; p. 92, vol. ii.) refers again to his previous observations, 

 and adds that the quantity of air in the intestine of the Hesione was such as to cause 

 the animal to float ; he considers it probable that the introduction of air into the 

 alimentary canal was an act voluntarily performed for the purpose of respiration. 



EisiG (16) refers to this observation of Quatrefages on H. pantherina, and to the 

 fact that Claparede had found bubbles of gas in the alimentary canal of H. sicula. 

 Specimens of this latter on being killed often expel bubbles of gas from mouth or anus ; 

 the species is, moreover, often found floating at the surface, the body swollen owing to 

 the distension of two air-containing diverticula of the alimentary canal. These gas- 

 bladders open into the stomach (the region of the alimentary canal between pharynx 

 and intestine), and extend forwards along the sides of the pharynx ; they are very 

 extensible, and are contractile both spontaneously and in response to stimuli. The 

 intestine of this species is also capable of energetic contractions ; the application of any 

 stimulus is followed by a more or less extensive peristalsis, which, if the intestine is 

 filled with gas, may force this gas with great force into the bladders or into the 

 oesophagus. The bladders are not vascular in H. sicula, but in Tyrrhena claparedii, 

 a related form, they are vascularised in a manner similar to the intestine. 



Similar vesicles, the T-shaped glands of other authors, are found in the Syllids. 

 In Syllis aurantiaca they are hammer- or T-shaped, and distensible ; whether they are 

 vascular or not is not stated. The greater number of Syllids have only slightly 

 developed or no air-bladders. 



Discussing the physiological value of these structures, Eisig states that neither 

 food nor secreted matters are ever found in them, but only a varying amount of clear 

 fluid, or gas, both of which can be pressed out of the stomach into the bladders, or 

 vice versd. The fluid is sea-water, and its purpose is respiratory. Though long and 

 carefully watched, the animals were never seen to take in air ; hence the gas must be 

 secreted by the animal, probably in the intestine ; and it is probably oxygen, which 

 becomes mixed with carbon dioxide and nitrogen through the respiratory processes. 

 A parallel is drawn with the swim-bladders of fishes. 



The swim-bladders of Annelids are therefore to be considered as appendages of the 



