104 Memoirs of the Indian Museum. More 
But there are difficulties in supposing that the micronephridial condition, itself 
derived from the meganephridial, can retrace its history; and it seems better there- 
fore to derive Erythraeodrilus from a situation on the line of descent anterior to the 
point of appearance of micronephridia, v.e. from some place between Notiodrilus and 
Octochaetus. 
Howascolex (Michaelsen, 7, c/. also remarks of the same author, 12 and 13) here 
suggests itself as a possible point of departure I have placed it, in imitation of 
Michaelsen (8), at the side of the basal stem in the preceding scheme. Howascolex 
already presents the mixed mega- and micronephridial condition found in Erythrae- 
odrilus, and the latter could be derived from it by (a) the microscolecine reduction, which 
as we have seen takes place in the younger branch represented by Eutyphoeus, and 
(0) an increase in the number of setae, such as also occurs in the Hoplochaetella branch. 
If geographical considerations (Howascolex has only been found in Madagascar) stand 
in the way of this, then the only remaining possibility would be a derivation from 
the original Acanthodriline itself,—which scarcely make things easier. 
Assuming provisionally therefore that Erythraeodrilus is a descendant of Howa- 
scolex, there remains the question whether it is to be reckoned to the Acanthodrilinae 
or Octochaetinae (there is at present no known indigenous Acanthodriline in India). 
Howascolex is placed (with some reserve} by Michaelsen in the Acanthodrilinae, 
and is considered as possibly but by no means certainly in the line of descent from 
Notiodrilus to Octochaetus, i.e. the ancestral line of the Octochaetinae (8, 12, 13). 
If Howascolex is judged not to be in the direct line, its descendant Erythrae- 
odrilus obviously cannot be included in the Octochaetinae. Even if Howascolex is a 
direct ancestor, Erythraeodrilus cannot be so included unless Howascolex is included 
also, for this would offend by making the Octochaetinae diphyletic (cf. table inf. If 
the Octochaetinae are to begin below the line ab, then to reckon Erythraeodrilus to 
them would deprive them of genetic unity). So long as Howascolex is placed in the 
Acanthodrilinae therefore, Evythraeodrilus must also go to the Acanthodrilinae, or 
must constitute a separate subfamily. 
The very wide difference between Erythraeodrilus and the rest of the Acanthodri- 
linae, and the inconvenience of instituting a subfamily for a single species, render 
both these solutions of the problem undesirable. I think the best way of dealing 
with it is to consider Howascolex as already, in virtue of its micronephridia, an Octo- 
chaetine, and as the direct progenitor of Octochaetus. There will then be no diff- 
culty in placing Erythraeodrilus as an Octochaetine also, —as a twig which arises 
independently from the base of the subfamily. 
Imay add that morphologically there is no reason why Howascolex should not be 
in the direct line between Notiodrilus and Octochaetus ; the only essential difference 
between Howascolex and Octochaetus is the coexistence of meganephridia with the 
micronephridia in Howascolex,—a difference which has not always been considered as 
necessitating even a generic separation. I presume that the reason why Michaelsen 
hesitates to consider Octochaetus as the direct and immediate descendant of Howa- 
scolex is that thelatter is found in Madagascar, the former in India and New Zealand. 
