262 DR M/INTOSH ON SOME POINTS IN THE STRUCTURE OF TUBIFEX. 



In the eighth segment the perivisceral branch is considerably enlarged, and, 

 curving outwards and downwards, enters the ventral as a large trunk, only 

 slightly less than at its commencement. The dorsal half of the vessel pulsates ; 

 the ventral does not. It would thus seem that the vessel, probably where fixed 

 to the body- wall, ceased to pulsate. This arrangement constitutes the so-called 

 " hearts" of these annelids, and in this species both contracted simultaneously. 

 The perivisceral branches of the seventh and ninth segments acted similarly, 

 though in a less conspicuous manner. The perivisceral vessel of the latter 

 (ninth) segment is often noticed to give off large branches at the body- wall. 



In specimens whose generative organs are much developed, the periviscerals 

 of the ninth and tenth segments are of considerable size, but neither approach 

 the periviscerals of the eleventh, which are enormously dilated, indeed, nearly 

 as large as the main trunk itself. This enlarged perivisceral sends branches 

 over the succeeding segments, sometimes as far backwards as the twenty-third 

 in ripe animals (Plate X. fig. 6, e, e). The periviscerals of the twelfth segment 

 under the same circumstances are also dilated, and those of the thirteenth 

 more so, three large branches being directed forwards. The arrangement of 

 these trunks would seem to countenance my view of the circulation in the 

 ordinary condition, viz., that the periviscerals as a rule do not proceed as con- 

 tinuously cylindrical trunks into the ventral, but that they communicate by 

 their branches on the body-wall. 



Continuing forwards, the main trunks (dorsal and ventral, Plate IX. fig. 17) 

 have the same arrangements in the fifth and sixth segments as previously de- 

 scribed, the only noteworthy change being an occasional abnormality in the 

 origin of the intestinals — one coming off before the other, and thus affording a 

 better view of their distribution. It is to be observed, also, that the ventral 

 trunk has in this region faint contractions, which render the vessel pale ; it then 

 fills again. The periviscerals of the fourth segment are somewhat smaller than 

 usual. At a point corresponding to the posterior border of the third segment, 

 the dorsal gives off two large trunks, doubtless the homologues of the perivis- 

 cerals, though they generally slant obliquely forwards and outwards rather 

 than transversely, and divide into many small branches towards the margin of 

 the body. The dorsal then pursues its course straight forwards, again gives 

 off two branches behind the mouth — the branches subdividing and inosculating 

 with others in front and behind, and finally terminates by forking in the snout 

 in front of the mouth. The two divisions thus formed split, after bending 

 backwards, into numerous twigs, which unite with the capillaries that go to 

 form the feeders of the main ventral vessel. The latter originates, by the 

 junction of the two great branches constituted by the feeders just mentioned, 

 towards the posterior border of the third segment. In many positions under 

 pressure, the whole anterior part of the animal is one vast series of vascular 



