632 



OLIGOCHAETA 



necessary to separate the three genera into a subfamily or to unite them into a single 

 genus. The irregular disposition of the setae is not, however, a differential character 

 of first-rate importance ; for in some specimens of Pontoscolex, which do not appear 

 from their size to be especially young, the setae are 'paired. This fact, first noted 

 by Fkitz MiJLLEE, has been also observed by myself. Moreover, in the genus 

 Tnchochaeta one species has and one has not this irregular disposition of the setae. 

 As there are no other points which distinguish them from the other Geoscolicidae, 

 it does not appear to be necessary to include them, as I formerly did, in a special 

 subfamily. The four genera, Geoscolex, Aniens, Bhinodrilus, and Urobertus, agree 

 in having paired setae which are usually ornamented; the last three of these are 

 united by Michaelsen (10) into a single genus. I explain later my reasons for 

 disagi'eeing with this fusion of three genera ; but the views of Michaelsen serve 

 to emphasize the very close relationship which exists between all three of them. 



The anatomy of the different genera which are contained in this subfamily will 

 be found treated of under the several genera. In this place the mutual relationships 

 of these genera may be discussed. 



This discussion will be facilitated by a study of the accompanying table, which 

 shows the principal differences and resemblances: — 



