496 EDWIN S. GOODRICH. 



of coelomic follicles and their funnels. This may lead to a 

 corresponding differentiation of structure ; in the first set the 

 peritoneal funnel becomes the most important part, in the 

 second the nephridial portion (such specialisation has been 

 well described in many forms by Eisig, Meyer [80], Trautzsch 

 [98], Cunningham [26], Marion and Bobratzky [78], &c.). 



The fusion between the nephridium and peritoneal funnel 

 does not occur in all Polychsetes ; fortunately, we appear still 

 to have all the intermediate stages between this condition and 

 that of the Oligochsetes. 



Meyer (80) has shown that in Nereis the nephridia of the 

 first five segments have the typical pronephridial structure 

 with a flame end-cell, and that in the posterior segments 

 this end-cell (judging from his figures) opens into the 

 coelom (true nephrostome ; compare Rhynchelmis).^ I have 

 also described in the Lycoridea (41) a ciliated region of the 

 coelomic epithelium which I believed to be the peritoneal 

 funnel. It is, however, to Eisig that we owe the description of 

 what appear to be intermediate stages. In Dasybranchus and 

 Tremomastus we have conditions in which the peritoneal funnels 

 (Genitalschlauche) are separate from the nephridia and open 

 independently (fig. 14), and in which the two organs are con- 

 nected but still open separately (fig. 13). Perhaps in certain 

 segments of these forms and of Capitella we have the more usual 

 Polychgete arrangement, in which the peritoneal funnel no 

 longer acquires an independent opening (31). 



Arthropoda. 



It will be best to begin our review of this group with a brief 

 recapitulation of the development of Peripatus, which has been 

 so excellently described by Mr. Sedgwick (94), and Dr. von 

 Kennel (64). Soon after the metameric somites have been 

 hollowed out to form the coelomic follicles, the upper half of 

 each coelomic cavity becomes nipped off" from the lower half. 

 From the wall of each of these lower coelomic sacs a peritoneal 



1 Such would appear to be tl>e coiiditiou in Protodrilus, where the nephri- 

 dial funnels figured by Hatschek (52) are small, and provided with a 

 flagellum. 



