INTESTINAL RESPIRATION IN ANNELIDS. 773 



of a chain ever has a complete series of loops ; and that of the loops which are actually 

 formed, a certain number at least do not disappear, but persist, in an adult worm, in 

 the middle of the animal's body; this is shown by Bourne's fig. ii., since the loops of 

 the segments numbered xviii., xix., xx., xxi., in that figure belonged, before the onset of 

 asexual reproduction, to segments xviii.-xxi. of the parent animal, and were not pro- 

 duced in a budding zone. 



An opponent might therefore reply that if the above interpretation of Beddard is 

 correct (in supposing that he bases his position on the appearance of a complete series 

 of loops, most of which subsequently disappear), it would seem that his argument needs 

 modification ; that it is doubtful how far the processes of asexual reproduction are 

 significant in a consideration of phylogeny ; and lastly, that while the interest and 

 importance of the observations are admitted, it will be advisable to wait for a complete 

 account of the history of the loops before drawing conclusions, and before extending 

 them to the Naididiv; in general and especially to the Enchytroeidag and /Eolosomatidae. 



(4) However the case may stand with regard to Uncinais, we have in Bmnchio- 

 drilus a genus of Naididai in which a complete series of lateral loops exists in the adult 

 condition. It may be argued that here at least there is exemplified for us the primi- 

 tive condition, in comparison with which the remaining genera are regressive. 



The question then is. Is a single form, Branchiodrilus, to be looked on as progres- 

 sive as compared with the rest of the Naididse ; or are these to be looked on as degenerate 

 in comparison with Branchiodrilus ? 



Branchiodrilus is a specialised form ; the loops are throughout a considerable 

 portion of the body continued into the gills ; and it is thus the presence of an extended 

 series of lateral loops which gives the possibility of the development of the gills, which 

 they supply with blood. 



It is open to us to suppose, therefore (a) that a progressive variation, viz. the 

 production of an extended series of loops, has been seized and taken advantage of for 

 the development of gills ; {b) or, considering the complete series of loops in Brcmchio- 

 drilus to be a primitive feature, that the development of gills is the reason of the 

 retention, in this isolated case, of the lateral loops which have disappeared in the 

 other genera. 



The first alternative would appear to be preferable, since the loops extend through- 

 out the body, including the posterior region, whereas gills are absent from the hinder 

 part of the animal. 



The argument is not conclusive either way ; but it appears to tell rather against 

 than for the assumption of a complete series of loops as a primitive condition. 



The opposite side of the question, i.e. that a small number of loops situated in the 

 anterior part of the body represents the more primitive condition, the complete series 

 being a later development, may now be considered. 



(1) In the first place, there is the fact that the circulatory system as a whole does 

 broadly, as has been shown, exhibit a gradual advance from the condition in yEolosoma 



