THE OLIGOCH^TA OF NATAL AND ZULULAND. 4-i9 



3 a. At each setal segment there are more than eight setse ; they 



form nearly complete circles round the body. Prostate 

 pores at the eighteenth segment, spermathecal pores 

 four pairs at the intersegmental fiirrows 5-6 to 8-9. 



Pheretima heterochseta {Mich.} 

 Sb. Every setal segment has in general eight setse. . . 4 



4 a. Setai of the hinder end are much enlarged and are not 



placed in eight longitudinal rows. 



Pontoscolex corethrurus (Fr. Miill.) 



4 b. Setge of the hinder end not, or not much, enlarged, in 



eight regular longitudinal rows. ... 5 



5 u. Six or seven segments of the pre-clitellar part of the body, 



viz. the fourth to the ninth or tenth segment, divided 

 by a ringlet furrow, quite as sharp as the intersegmental 

 furrows, into two ringlets ; the front one bears the 

 lateral nephridial pores and the setse, if they are 

 present.' ..... t^ 



5 b. All segments are simple or only divided by more or less 



slight ringlet furrows, which are fainter than the inter- 

 segmental furrows, and are not restricted to a small 

 number of segments. . . . .14 



6 fl. Setse all ventral ; median dorsal distance of seta; equals 



about two thirds of the whole circumference of the 

 body. Copulatory walls on segments 19-21, and they 

 touch each other in the median venti-al line. 



Microcha;tus sulcatus (Kinb.) 

 a The median ventral distance between the inner- 

 most ventral seta? diminishes distinctly towards 

 the copulatory cushions. . . f. typicus 



/3 The median ventral distance between the innermost 

 ventral setse does not distinctly diminish towards 

 the copulatory cushions. var. howickianus n. var. 



' This biannulation is not always easy to recognise, as often the setse, 

 which nuiy Ije comj)letely absent, and the nephridial pores are difficult 

 to detect. But the biannulation may be distinguished by the fact that 

 on front of the clitellum a certain segment-like part (really the hind 

 ringlet of the liindermost biannulated segment) is somewhat shorter 

 than the segment behind, while the next segment-like part (really the 

 front ring of the hindermost biannulated segment) is distinctly longer 

 than the posterior normal segments. Wlien enumerating the segments 

 of Microchsetus species it must be borne in mind that often the first 

 and second segments are retracted into the mouth, and are consequently 

 invisi])le. 



