r,2s O LI GOGH A ETA 



All these genera agree to differ from the next subfamily in these characters: 



(1) Spermathecae, one to four pairs, placed in neighbourhood of gizzard. 



(2) No copulatory papillae. 



These two characters are nearly absolutely distinctive; there are, in addition to 

 these, a few other characters, which, without being found in all the genera belonging 

 to this subfamily, are not found in any of the genera belonging to the second 

 subfamily of the Geoscolicidae. 



Thus the sperm-sacs are frequently a single pair of sacs which are of very 

 considerable length ; in Trichochaeta, for example, they occupy as many as twenty 

 segments ; the irregular distribution of the setae, often met with in this subfamily, 

 does not occur in the Microchaetinae. The statement that there are no copulatory 

 papillae requires some explanation ; MiCHAELSEN (10) has described in Anteus 

 papiliifer numerous papillae on the segments of the clitellum and on a few 

 segments in front of the clitelluin ; these are formed round modified sexual setae. 

 In the second subfamily of the Geoscolicinae, for instance in the species Kynotus 

 mi( haelsenii, the sexual setae are provided with glandular sacs, having thick muscular 

 walls, and recalling the spermiducal glands of other Oligochaeta. 



It is doubtful, however, how far this distinction can be retained, having due 

 regard to the structure of MICHAELSEN'S Anteus callichaetus ; in this species 

 MICHAELSEN (10) describes the presence in segments xviii-xx of five pairs of 

 ' unregelmassig kugelige Taschen, deren Wandung ein driisig-zottiges Aussehen hat,' 

 a description which might apply to glands such as those which occur in the genus 

 Perichaeta in connexion with the genital papillae, as well as to pouches like those of 

 Kynotus and JMicrochaeta. And though, as I have pointed out in discussing the 

 spermiducal glands of the Oligochaeta (p. 113), it is quite possible that the glands in 

 Perichueta are referable to the same category as those of Microchueta, &c., there are 

 obvious differences of detail. 



A good deal of stress has been laid upon the position of the nephridiopores, 

 although the fact that they alternate in position from segment to segment in more 

 than one genus tends to throw doubt upon the usefulness of this character ; it may 

 be useful to see how far this character can be made use of in dividing up the 

 genera of the present family. We have not complete information. 



It will be noticed from the accompanying table that all the New-World genera, 

 with the exceptions of Geoscolex and Tykonus, agree in possessing nephridia which 

 opt n in front of one or other of the five lateral setae ; now it is interesting, if not 

 significant, that both these genera present some feature of affinity to the Old-World 

 Gecscolicidae : Geosrolex agrees with Siplionogaxter in having no spermathecae, and 



