230 DB. W. B, BENHAM OK [NoV. 15, 



number of ch^etse remain normal, though they are differently 

 arranged. 



Then there is Dinodrihbs, with 12 chfetse and micronephridia. 

 But both these genera have the prostate pores and male pores 

 arranged as in the archaic form. 



It seems to me that Michaelsen is in error in separating these 

 two genera from other Acanthodriline forms and associating 

 them in a separate subfamily, the Octochsetinse, with Etdyphoeus 

 and Hoplochcetella ; for, apart from the micronephric condition, 

 there is really little to distinguish Octochcetus from Notiodrilus ; 

 moreover their presence in New Zealand indicates their close 

 association therewith. The step from the Acanthodriline seiies 

 to the Megascolecine is a small one, and apjDears easily conceivable 

 from analogy with the origin of Rhododrilus : that is, one prostate 

 pore has shifted so as to open close to and in common with the 

 male duct in the 18th segment, for several species of Cryptodriline 

 worms have prostates similar to those of Rhododrilus. 



Ehobodbilus edulis, sp. n. (Text-figs. 48-54.) 



Two individuals of " Tarao," a worm eaten by the Maoris, were 

 received from Mr. Best preserved in formol. 



Dhnensions. Length 275 and 285 mm. respectively, with a 

 diameter of '13 mm. just behind the clitellum. The worm is 

 cylindrical, tapering only very slightly posteriorly, and then 

 suddenly decreasing as the anus is approached, so that at segment 

 ?i-8 the diameter is 10 mm., and at n-2 it is still 8 mm. The 

 total number of segments in the larger individual is about 300, 

 but as the hindmost segments are very closely contracted and 

 small, and triannulate, it is not easy to count them with absolute 

 correctness — nor does this matter. There is no doubt that the 

 living worm was, when extended, at least 15 inches and possibly 

 more. Mr. Best in a letter says : " These Tarao are small, I have 

 seen them 18 inches in length." 



The segments 2-5 are biannulate, 6-12 are quadriannulate, 

 with the chsetse in the third annulus, while those of the clitellar 

 and postclitellar regions are triannulate. 



The colour of the " Tarao " is (in formol, which does not cause 

 the colour to change much) red ; the anterior end apjoears pale 

 greyish, much paler than the rest of the body, which is pale 

 reddish, with a tinge of purple on the upper half of the body ; 

 the clitellum is buff. 



The prostomium, like the first segment, is a good deal furrowed ; 

 it is epilobic, being dovetailed into the peristomium for about 

 half the length of the latter, and it ends in a transverse groove. 



The chcetce are eight in number, rather widely spaced ; the 

 lateral spaces (be) on each side are nearly equal in the mid-body 

 and posterior region, while the dorsal space (dd) is about twice 

 the ventral (aa) and three times the lateral. 



As is frequently the case in allied genera, the ventral couple 

 (a, h) approach one another in the region of the genital pores ; 



