238 1)R. W. B. BENHAM ON [Nov. 15, 



each sac is more or less ovoid, somewhat bent at the commence- 

 ment of the duct (? perhaps artificially by contraction of the 

 body). 



The dvict is short, but wide, and receives a very long, tubular 

 diverticulum, which is spirally coiled at its lower end, where it 

 enters the common duct close to the body-wall. The lower end 

 of the diverticulum is muscular, though of the same diameter as 

 the rest. 



Loc. Ruatahuua, Urewera Country, North Island, N. Z. A 

 single individual. 



Remarks. — It will be seen that this species differs from R. edulis 

 in details with regard to the arrangement of chfetse and copulatory 

 tubercles, in form of penial chsetfe and spermatheca, as well as in 

 size, form, and colour of the worm itself. 



The Genus Rhododrilus. 



It is clear that in many features R. edidis and R. besti agree with 

 Microscolex ; nevertheless there are certain points of difference, 

 which, in view of geographical distribution, may be of more 

 importance than the resemblances. 



For example, most noticeable, but perhaps of no great sys- 

 tematic importance, is the great size of these two species, and 

 especially of R. edidis (285 mm.), as compared with the small 

 worms included in the genus Microscolex, ranging as they do 

 from a length of 30 mm. to that of 58 mm. 



j3ut the point upon which I would lay sti'ess is as to the form 

 and extent of the prostate. 



In all the species of Microscolex this gland is of comparatively 

 small size, and limited to segment 17, or extends, in 

 M. heinpeli and M. novce-zecdandice, into the next segment; 

 Avhereas in the two species jvist described these glands ai-e of 

 very considerable extent, passing through 5 or even 8 segments. 



It is also worth noting, in view of the fact that Beddard is 

 inchned to lay some importance on the point in some genera, 

 that in all the species of Microscolex the last heart is in segment 

 12, whereas in my two species it is in segment 13. 



Michaelsen points out, too, that in the genus Microscolex the 

 single pair of spermathecse open in the fui-row 8/9 : this is the 

 case with one of my species, whereas the pore in the other is 

 at 7/8. 



The gizzard, too, in Microscolex, is " absent or rudimentary," 

 which is not true of our edible worms ; and, finally, the clitellum 

 of the latter is not "complete" or "girdle-like," but saddle- 

 shaped. 



These small points of difference necessitate one of three courses : 

 either we must amend the diagnosis of Microscolex, or we must 

 create a new genus, or we may place our species in the closely 

 allied genus Rhododrilus. The last plan, it will be seen, is the 

 one I have followed, though even this procedure necessitates the 



