214 



If one holds that the ciliated funnels function for carrying ex- 

 cretory waste products to the exterior then the venous connection 

 is utterly unexplainable. 



Later than this, the nephrostomes in the Amniota atrophy alto- 

 gether and the kidney becomes transformed into an excretory 

 organ (Fig. 1 0) . 



In the lowest Chordata as the Hemichorda, there are simple pores 

 or short ciliated tubes, leading from the coelom to the exterior -^'^ (pro- 

 boscis pores, collar pores) an advance upon the diffuse and generalised 

 ways of exit in the Echinoderma, though the ontogeny of the latter 

 points to a definite series of pores in the bilateral ancestors of the 

 group (Dvirham!. 



This is succeeded in the Cephalochorda by metamerie nephrosto- 

 mial funnels opening freely into the coelom and the branchial chamber 

 throughout life^^ 



The modifications of these funnels through the Cyclostoma, Am- 

 phibia, etc., bear out to some extent the history indicated by the 

 Amphibian ontogeny ^i. 



The morpholoi^y of the nephridial funnels is further complicated 

 by the fact that from the condition where some of the nephrostomes 

 function for discharge both of gonocytes and ek-phorocytes (a stage 

 represented in ontogeny of Amphioxus by the period where the peri- 

 gonadial coelom communicated freely with the pronephric funnel, 

 and its mesonephric homologue in Scyllium [Boveri]) to that in 

 which some funnels function solely for the exit of gonocytes whilst 

 others having lost their egestive function atrophy, the tubules only 

 remaining. 



In the development of any pronephric (or for that matter meso- 

 nephric) tubule the funnel and tubule are first formed before the 

 cells lining the walls take on an excretory function or become closely 

 connected by glomeruli to the vascular system, so that it is not unrea- 



39 W. Bateson, Q. J.M.S. Vol. XXIV— XXVI. — G. H. Fowler, Festschr. 

 fur Rud. Leuckart. 1892. — S. F. Harmer, » Cliallenger« Zoology XX. Appendix. 



40 T. Boveri, Zool. Jahr. Abth. für Morph. V. 



41 It is obvious that the above derivation from coelomic pores metamerically 

 repeated precludes the likelihood of the nephridia being derived from branched 

 flame-cell excretory organs. In this case the branching nephridia of Capitellidae 

 (Eisig, loc. cit. p. 181, of rontohdella (A. G. Bourne, Q. J. M. S. 1884), and of 

 Perichaeta {¥. E. Beddard, Q. J. M. S. XXVIII) must be assumed to be seconda- 

 rily acquired, in a precisely similar manner to the reduplication of nephr. funnels 

 in the Vertebrate phylum. Flame-cells and their tubes belong to the stage in phylo- 

 geny in which egestion (monocytic) is diffuse, hence are entirely excretory, are not 

 segmented and drain the primary body-cavity. Compare: Gegenbaur, Comp. 

 Anat. and R. S. Bergh, Kosmos. 1885. 



