CH. I] RANGE OF GEOSCOLICTD-ffi. 63 



in this country — in which the spermathecae are oval or 

 roundish pouches without any caeca attached to them. 

 But the Geoscolicidae agree with the majority of earth- 

 worms to differ from the Lumbricidee in that the gizzard 

 is as it were strung upon the oesophagus, instead of lying, 

 as it does in Lumbricus, at the junction of oesophagus and 

 intestine. The nephridia are always paired ; but the 

 setae differ from those of the Lumbricidae in being nearly 

 always ornamented with raised ridges or sculpturing of 

 some sort; moreover there is frequently a difference 

 between those which occur on the segments of the 

 clitellum and those which are found elsewhere upon the 

 body. The Geoscolicidae of Africa belong to the genera 

 Microchceta, Siphonogaster, Ilyogemia and Callidrilus. I 

 have already referred to Kynotus of Madagascar which 

 is allied to these but still is different from any of them. 

 The family extends in the Old World into the Malay 

 region where it is represented by the two allied forms, 

 perhaps hardly generically separable, Glyphidrilus and 

 Annadrilus. It is represented in Europe by Hormogaster, 

 whose affinities are uncertain. Putting aside Siphono- 

 gaster, which is a remarkable and isolated type with two 

 extraordinary processes of the ventral body wall which may 

 be of the nature of penes, and Ilyogerda which may be, as 

 has been suggested by Dr Michaelsen, really a repre- 

 sentative of another family which has got to resemble the 

 Geoscolicidae by convergence, all these Ethiopian and 

 Oriental genera are alike in having a great number of 

 small spermathecaB in several segments of the body and in 

 having for the most part a glandulo-muscular structure at 



