1890.] WORMS OF THE GENUS PERICH^TA. 55 



If Rosa's definition of the two genera be slightly altered, it will 

 be possible to arrange most of the species of " Perichceta'^ as 

 follows : — 



Megascolex. Line of setae interrupted; clitellum occupying 



more than three segments. 

 Perich-«;ta. Line of setae continuous ; clitellum consisting of 



three segments only. 



There are, however, other species which present more important 

 differences among themselves than those above mentioned. Prof. 

 Bourne (11) calls attention to the fact that there are anatomical 

 differences, to «hich considerable weight must be attached, between 

 several of the foims described by him. 



Perichceta novce zelaiidice (Beddard, 7) differs from all other 

 species in the following combination of characters : — No dorsal pores ; 

 nephridia paired ; atria tubular. 



P. bakeri and P. barronensis of Fletcher agree in their nephridia 

 and atria, but have dorsal pores. 



A number of species described by Fletcher (17), viz. P. atte- 

 nuata, P. enonnis, and P. coxii, agree to differ from others in the 

 very remarkable fact that the setae in the anterior segments are eight 

 in number to each segment, increasing in the posterior segments to 

 20-30. 



In both these cases the different geographical area occupied by 

 the species is, perhaps, a further argument in favour of separating 

 them. 



Then Perichceta stuarti has four pairs of atria, as in Acanthodrilus ; 

 they open on to the same segments as in that genus, and appear to 

 be similar in structure ; tlie diverticula of the spermatheca also 

 appear, from Bourne's description (11. p. 667), to be more like those 

 of Acanthodrilus than of most Perichceta. P. ceylonica (Beddard, 

 9) has two distinct pairs of atria, but they open both of them on to 

 the same segment, the 18th; at present I am not inclined either 

 definitely to unite this form, generically, with P. stuarti, or defi- 

 nitely to separate it ; further information as to its structure is first 

 needed. 



If these various types be accepted by zoologists as of generic value, 

 it must still be admitted that they cut off very small corners from the 

 mass of species (about sixty in number) of which the family con- 

 sists. At present, however, it does not seem ensy to make any further 

 alterations, and I am not satisfied that Megascolex has the same 

 value, as a generic type, that Aporochceta has. 



The following is a definition of the family Perichsetidae and its 

 various genera as advocated in this paper. 



Fam. Perich JETiD^, Claus. 



Earthworms with a continuous circle of numerous setae round all 

 the segments of the body (with the occasional exception of a few of 

 the most anterior) ; clitellum commencing with the thirteenth or 



