194 



Transactions. 



Fig. 3. — Perieodrilus ricardi. 

 Sketch of the appearance 

 presented by the blood - ca- 

 pillaries as they follow the 

 branches of the narrow canal. 



of Microchaeta (1), where the narrow duct, indeed, is made up of a network 

 of anastomosing canals ; but each branch is separate, and forms a single 

 loop round the ducts which lie between the narrow canal and the outer 

 side of the tubule. Sometimes, however, the branch divides into two before 

 it bends round ; but, as a rule, there is no connection between the indi- 

 vidual loops (c/. also Eisen's figure of A. marmoratus). 



Each such excretory loop is accompanied by a fine capillary blood- 

 vessel, and when seen in optical section the vessels form a perfectly regular 

 trelliswork radiating from the inner to the outer side of the fold (fig. 3), 

 and the appearance presented when the ne- 

 phridium is mounted in glycerine is remark- 

 able, for the blood-vessels shine out very 

 clearly and form a striking characteristic 

 of the nephridium. 



But for this regular trelliswork of blood- 

 vessels the looping of the excretory canal 

 might be easily overlooked — at any rate, in 

 preserved specimens — for the canals are so 

 transparent that they are seen with great 

 difficulty except in that part of a fold where 

 the ciliated regions occur ; but the canals 

 are here easily distinguishable, and the direc- 

 tion of the current may be deduced from 

 the direction of the cilia. A more minute 

 examination of the blood-vessels proves, however, that each capillary accom- 

 panies a loop from the narrow duct, and that both have a semicircular 

 course, passing outwards to the edge of the fold and back again to the duct 

 round the intervening tubes. 



These capillary blood-vessels are exceedingly fine, and their actual con- 

 nection with the main blood-supply cannot usually be seen. It was observed, 

 however, in a few cases, and these were sufficient to set at rest any doubts 

 of a possible misinterpretation of the appearance presented. 



It is worth noting the behaviour of similar outgrowths from the 

 narrow canal in the various genera which possess a branching system of 

 tubules. 



In Microchaeta rappi (1), the nephridium of which, excluding the well- 

 developed bladder, is not unlike that of P. ricardi, there occurs in each 

 looping fold of the rosette an anastomosing system of ducts which ramify 

 over and around the wider canals of the fold, in this si)ecies there is no 

 narrow canal sending outwards regular processes from its lumen ; the 

 branching has become quite irregular, and the duct itself has become lost 

 in its branches. 



In Moniligaster grandis (7) the narrow canal is not branched in such a 

 complicated manner ; it is still distinguishable in most regions, but in the 

 distal portion of the glandular lobe transverse connections appear between 

 the two parallel regions of the narrow tube, though in some parts the out- 

 growths of ductules are irregular. 



In ArgilopJiilus, as described by Eisen (8), we find a condition which 

 resembles more nearly that of P. ricardi. Here, however, the branching is 

 limited to a definite portion of the narrow tube — viz., along the whole 

 length of the anterior fold — in accordance with the perfectly definite form 

 of the nephridium, while in Perieodrilus it appears to occur in many 

 regions and in any. 



