250 Proceedings of the Royal Physical Society. 



was tlie result of the division of an originally single pair of 

 nephridia to each segment. Now, however, numerous genera, 

 including most of those with the largest number of species, have 

 been shown by myself,^ by Benham,'-^ and by Spencer,^ to possess 

 an excretory system of the same kind. These genera include 

 representatives of three out of six of Rosa's families. As both 

 conditions may occur in the same genera (for example Acantho- 

 drilus, Cri/ptodrilus, Perichceta [sensu lato]), it seems clear that one 

 of the two conditions has been several times independently 

 produced.^ Thus, after all, it may perhaps be said that Eisig's 

 objections are so far not removed, as there is a simple mul- 

 tiplication of instances. As a question of mere probability it 

 seems to me easier to suppose a reduction than a multiplica- 

 tion of nephridia in a segment, especially as there is at the 

 same time in some genera (in Perichceta and Megascolides at 

 any rate) a connection between the nephridia not only of ^ the 

 same segment, but also from segment to segment. In these forms, 

 moreover, there is no regularity in the position of the external 

 pores or the coelomic funnels ; they cannot with any approxi- 

 mation to the truth be called " segmental organs." On a 2^'riori 

 grounds, therefore, the existence of dysmetameric organs in so 

 regularly metameric an animal as an Annelid suggest an inheritance 

 rather than a modification within the group. Another argument 

 for considering the dysmetameric condition as the more primitive 

 is afforded by the genera Megascolides and Acanthodrilus. In 

 the former genus Spencer^ has described nephridia opening by 

 numerous ducts into the pharynx ; in A. ')7iultipo7-us 1 have 

 myself found ^ that the hinder region of the intestine is furnished 

 with numerous diverticula, which become continuous with tubes 

 indistinguishable from the ordinary nephridia. Now it is more 



^ Preliminary Note on the Nephridia of Perichceta — Proc. Roy. Soc, vol. 

 xliii., p. 309. The Nephridia of Earthworms — Nature, vol. xxxviii,, p. 221 

 (1887). On the Presence of Numerous Nephridia, etc. — Q. J. M. S., vol. 

 xxviii., p. 397. On certain Points in the Structure of Urochceta, etc. — 

 Ibid., vol. xxix., p. 235. On the Structure of Three New Species of Earth- 

 worms, etc. — Ibid., vol. xxix., p. 101. 



- Studies in Earthworms, No. I. — Q. J. M. S., vol. xxvi., p. 213. 



^ The Nephridia of Earthworms — Nature, vol. xxxviii., p. 221 ; The 

 Anatomy oi Mer/ascolides australis — Trans. Roy. Soc. Vict., vol. i. , No. 1. 



^ Tile following have or maj' have " diffuse " nephridia : — Perichceta (and its 

 subdivisions), Cryptodrilus, Megascolides, Digaster, Didymog aster, Dichogastcr, 

 Acanihodrihis, Trigaster, Typhceus, Deodrilus, Deinodrilus. They include 

 one-half of the known species. There are 19 genera in which the nephridia 

 are always paired. 



^ Log. cit., pi. iii., fig. 10. 



^ On the possible Origin of the Malpighian Tubules in Arthropods — Ann. 

 and Mag. Nat. Hist., 1889, p. 290. 



