ix] GEOGRAPHICAL DISTRIBUTION 135 



occurrence in other parts of the tropics is perhaps not 

 yet explained satisfactorily. Also we may record as 

 characteristic of this Ethiopian region a few peculiar 

 genera such as Nannodrilus and Gordiodrilus. Alma 

 being a partly aquatic genus is perhaps less distinctive 

 and as a matter of fact it strays into the Palaearctic 

 region, being found in the lower waters of the Nile. 

 It will be observed that with this exception the limits 

 of the Ethiopian region according to earthworms 

 agrees with that delimitation afforded by a con- 

 sideration of other groups since it stops short at the 

 Sahara, leaving northern Africa to be referred to the 

 Palaearctic region. At the same time we have an 

 analogy with South America as concerns the southern 

 extremity of the African continent; here we meet 

 with Notiodrilus and allied Acanthodrilinae just as in 

 Patagonia and as also in that quarter of the world 

 these forms just stray into the Ethiopian region 

 above specimens of Notiodrilus being met with in 

 Madagascar as well as in tropical Africa. This bit 

 of Africa as it appears to me must also be cut off 

 from the Ethiopian region and included in an Ant- 

 arctic region. Madagascar offers a further problem. 

 Are we to include this in Ethiopia or speak of a 

 Malagasy region? Apart from a few forms which 

 are at least possibly to be looked upon as accidental 

 immigrants, such as members of the genera Pheretima 

 and Gordiodrilus, the fauna of Madagascar consists 



