312 Annals of the South African Museum. 



Somites v-xxvi are then triannulate, somite xxvii biannulate, but 

 nothing can be said in regard to somites i-iv, except that the 

 annulation is very indistinct or absent. It is interesting to note 

 that somite v, which is triannulate ventrally, shows a much more 

 indistinct furrow between annuli 1, 2 than between 2, 3 ; further it 

 is biannulate dorsally and the anterior annulus is decidedly broader 

 than the posterior. These facts bear out the idea that the order of 

 origin of the annuli is, as in the majority of Glossiphoniidse. 



1, 2, 3 (annuli) 



cab (order of origin). 



In a number of the individuals examined the brood pouch aperture 

 was a very prominent structure. At first it was supposed that it was 

 an abnormally developed genital aperture, and it so happened that 

 one specimen had a minute structure protruding through the 

 aperture. The fact that no genital aparture other than this could be 

 made out on an external examination supported this idea, and sug- 

 gested that the protruding structure was a penis. The posterior 

 position of such a genital aperture, and the presence of penis in an 

 undoubted Glossiphonid, were so unique that it was found necessary 

 at this stage to section an individual. It was then found that in 

 several individuals which were sectioned, the pore led into a large 

 pouch which extended through six somites, and in this were found, 

 in different individuals, young in all stages of development. One 

 such individual contained segmentation stages only, and although 

 the position of the pore or slit is very definitely indicated, no 

 communication to the exterior was found. Several other specimens 

 examined externally showed no definite aperture, although the posi- 

 tion of such if it did exist was clearly indicated. It seems then that 

 the absence of such in some specimens is due not to its being 

 obscured by contraction, but by the fact that the aperture exists only 

 when the contained young are well advanced. The supposed 

 "penis" mentioned above was found to be a young individual 

 protruding through the aperture. It has been stated already that 

 the female ducts no longer open directly to the exterior, but into the 

 brood pouch ; so that the ova would pass directly into this chamber. 

 Hypodermic impregnation which we know takes place in some 

 Glossiphoniidae, and probably in all, as previously suggested by one 

 of us (E. J. G.), would seem to be necessary here owing to the 

 absence of a female pore opening to the exterior. This is strongly 

 supported by the remarkable tenuity of the body wall in the brood 



