Evolution of Animals 449 



each with an external pore as in the last, to a segmentally 

 arranged and unpaired or paired system, in which many sets 

 of nephridial or excretory glomeruli are in intimate contact 

 with special branches of the vascular system, as in the highly 

 complex arrangement described by Montgomery for Sticho- 

 stemma Eilhardi (95: 1). It would be impossible with our 

 present knowledge to say whether we have here a graded ad- 

 vance from a simple pronephros of two tubules to a highly 

 complex pronephros composed of many sets of tubules, or 

 whether rather the nemerteans may not present us with all 

 stages in formation of a pro-, meso-, and metanephridial sys- 

 tem of still diffuse type in Tetrastemma. 



But, as indicated in a preceding paragraph, it seems ex- 

 tremely probable that the primary and specially the posterior 

 main excretory tubes of Rhabdocoela became closed externally, 

 in transition from that group to the Nemertinea, and so formed 

 a rudimentary vascular system. The anterior ends of the 

 tubes then, along with their branches, made special, usually 

 paired, antero-lateral external openings for themselves. In 

 the earlier stages of the process, as exemplified by most nemer- 

 teans, only two paired anterior openings were formed. But, 

 in types that must once have connected these with Sticho- 

 stemma, the primary nephridial tubes became broken up into 

 separate constituents, that gradually ranged themselves along 

 the length of the body. Some of these remained closed, but 

 most made external openings along the sides of the animal. 

 Internally, they gave origin to groups of flame cells of excretory 

 character, that ramified alongside the vascular tubes. 



Since the tubes are of unequal length in Stichostemma, alike 

 in themselves, and on the two sides of the body, and since 

 the external openings, real or potential, vary from 16 to 19 

 on each side, there is strongly suggested here an archinephric 

 system such as Lankester has postulated. Further the very 

 irregular lengths of the tubes, and the irregular disposition 

 of the openings, might indicate that, in some still more evolved 

 fresh-water form that once existed, the archinephridial system 

 consisted of at least 20 pairs of archinephridia, that opened 

 symmetrically by as many pores, and which internally formed 

 as many sets of coiled tubules. Still later, a common internal 

 archinephric duct may have joined the outer ends of the nephri- 

 dia, and so the external openings would become closed up. In 

 this connection Goodrich remarks (U6:85): "In the cyclo- 

 stomes there is evidence of the whole duct being really formed 

 by the fusion of segmental rudiments representing tubules, 



15 



