igoS.] EXCRETORY ORGANS OF METAZOA. 565 



except medially, these opening into a network of intracellular ducts 

 placed outside of the peritoneum, and the latter discharge at the 

 surface by irregularly arranged canals (Spencer, 1889). In 

 Mahbenus each micronephridium has its own excretory duct 

 (Bourne, 1894). The network of fine canals may be continuous 

 from segment to segment, as in Perichcrta, or only the micro- 

 nephridia of one and the same segment may be so connected as 

 exemplified by Deinodrilus, Acanthodrilus, and Dichogastcr (Bed- 

 dard, i888Z^). 



Both of these kinds of nephridia may occur in the same animal 

 and even in the same segment, as in Megascolidcs; and in this genus 

 there is a pair of ventral longitudinal canals continuous from seg- 

 ment to segment into which both open (Spencer). In Dichogastcr 

 the posterior segments contain both kinds (Beddard, 1888&). 



Development of the Meganephridia. — With great hesitation I 

 attempt to give a brief review of this subject, that has proved the 

 Austerlitz of many a theory. Kowalevsky (1871) was the first 

 to demonstrate the mesoblastic origin of these organs in Ena.vcs 

 {Rhynchelmis) and Liimhricus. Vejdovsky and Bergh have fur- 

 nished more observations on the subject than any other writers. 

 Vejdovsky's results (1885, 1892a, 1900) on Rhynchclmis, Stylaria 

 and Tubifex are as follows : Each nephridium arises from three 

 separated anlages : ( i ) A large preseptal funnel cell, giving rise to 

 the nephrostome; (2) a cord of small cells budded off behind the 

 former, producing the secretor}- loop; and (3) an ectoblastic in- 

 vagination that joins wuth the latter and forms the distal canal 

 and the excretory vesicle. Bergh's studies (1888, 1890, 1899) on 

 Lumhrictis, Criodrihis and Rhynchclmis differ from those of 

 Vejdovsky mainly in deriving each nephridium from a single meso- 

 blastic anlage instead of from three parts; in his mind the organ is 

 essentially an embryonic unit. Wilson (1889) concluded for 

 Allolohophora that the funnel arises from a large mesoblast cell, 

 and the loop from a postseptal mass of cells that is continuous with 

 the ectoblastic nephridial cell cords, though he admits the loop may 

 nevertheless be mesoblastic. And Lehman (1887, Allolobophora) 

 derived the nephridium from a large preseptal cell. 



These researches agree in finding that the nephridia arise seg- 



