The Classification and DistrihiUion of Earthworms. 240 



It appears to me, in fact, that the key to the classification of the 

 group is to be found in the modifications of the excretory system. 



It is obvious that the way in whicli any group sliould be 

 classified is that which will indicate its course of develop- 

 ment. Clearly, therefore, characters should be chosen which 

 have a relation to lower forms from which the group to be 

 classified has been evolved. Characters peculiar to the 

 group, however much or appropriately they may vary, can 

 only be regarded as of secondary importance. Where, how- 

 ever, it is a question of indicating the affinity of particular 

 species and genera, then characters peculiar to the group are 

 available. Hence it may be perfectly reasonable to sketch the 

 main outlines of a scheme of classification by the modifica- 

 tions of only a single character; and perfectly unreasonable to 

 do so by the use of even a large number of other characters. 



It is a common mistake to think that several characters 

 are necessarily better than one. 



Now it appears to me that structures like the clitellum, 

 the setae, the gizzard, and so forth, are so distinctively 

 " Oligochtetous," that it is dangerous to commence the broad out- 

 lines of a classification by using them as diagnostic characters. 

 It seems to me quite conceivable that these characters and 

 others like them may have changed about so greatly during 

 the course of the evolution of the group as to have several 

 times (independently) produced the same result. I do not 

 think, for example, that the Lumbricidse and Geoscolecidas are 

 necessarily related on account of the absence in both of atria 

 and penial setae, and in the saddle-shaped clitellum. Such a 

 modification may have occurred more than once. 



The nephridia, however, are not distinctively Oligochaetous 

 structures even in the actual form which they assume in 

 that group. 



As long as one species of Acanthodrilus {A. multijmrus)'^ was the 

 only form known with numerous nephridia per segment, it was 

 perfectly legitimate for Eisig to refuse ^ to admit this arrangement 

 as the archaic one. It might readily be supposed, as the Naples 

 zoologist supposed, that the multiplication and interconnection 



^ Beddard, Preliminary Note on the Nephridia of a new Species of Earth- 

 worm — Proe. Roy. Soc, June 1885. 

 2 Die Capitelliden in Fauna und Flora des Golfes von Neapel. 



