INTESTINAL RESPIRATION IN ANNELIDS. 771 



plexus, on both intestine and oesophagus, are the most obvious ; but the system com- 

 prises numerous and definite longitudinal and oblique vessels also. 



Wo thus have anatomical evidi-nee that a very considerable proportion of the blood- 

 vessels in both lower and highei- Oligochseta are derived from the alimentary ple.xus, 

 which, situated in the gut-wall, is the most primitive portion of the system ; the 

 vessels have arisen by a progressive differentiation of the channels in the plexus, 

 which at first, as still in ^Eolosonia, consisted of a system of irregular lacunar spaces. 



Of the remaining portion of the vascular system, I propose to consider in detail 

 only the lateral loops or commissures, a certnin Jiumber or all of which are, like the 

 dorsal vessel, contractile, and share in the function of propelling the blood. 



I'he Lateral Commissures. 



The direct evidence for the origin of the lateral loops from the alimentary network 

 is less than in the case of the vessels previously considered, and the earlier stages of 

 their development are not clearly documented for us. The argument must be largely 

 one of analogy. 



The contractility of the dorsal vessel is originally the same thing as the contractility 

 of the gut-wall ; and unless we are to assume an altogether different origin for the 

 contractility of the commissures, these also must be imagined as having their origin 

 in the gut-wall and as having later separated from it. 



Anatomically, it is perhaps worth while to recall such features as the parallel 

 channels running transversely between dorsal and subintestinal vessels in the crop of 

 Chietogaster, and the well-marked transverse vessels in the intestine and oesophagus 

 of Phcretima and other earthworms. These examples show that not only longitudinal, 

 but transverse channels also, may be formed in the gut-wall. From such channels 

 the commissures may be supposed to have arisen, sepai-ating themselves from the 

 alimentary wall in a manner like that of the dorsal and ventral vessels. As to why 

 such channels only separate from and become independent of the alimentary canal to 

 the number of one pair per segment, the answer is probably that this can only happen 

 where there is some support for them, along which they can travel, i.e. can only happen 

 at the positions of the septa. 



Is a Small or a Large Niimher of Commissures the Primitive Condition ? — In the 

 Tubificidae and in the higher families of the Oligochseta there is, speaking generally, 

 a pair of lateral loops to each segment of the body ; in ^Eolosoma there are none, in 

 the Naididio and Enchytrteidaj a few only, situated in the anterior region. Which 

 condition are we to look on as primitive ? In other words, are we to consider the small 

 number of loops in the Naididse and Enchytrseidae as the remnant of an originally 

 complete series, and is yEolosoma the last stage in the path of degeneration ? 



Let us consider first the case for degeneration. It has, in fact, been urged by 

 Beddard (3, pp. 72, 167, 170) that yEohsonia and the Naididse, though simple forms, 

 are secondarily simple, and not primitively so. 



