308 Annals of the South African Museum. 



investigated is by no means suited to the existence of such. It is, 

 nevertheless, highly probable that such forms will eventually be 

 found in the moister and well-wooded parts of the Union, or more 

 especially in the more tropical country north of the same. Such 

 forms, if obtained, should in all probability possess a special interest 

 as indicated by one of us (E. J. G.) in the Proc. Linn. Soc. of 

 N.S.W., 1910. 



In this paper we restrict attention to representatives of the 

 Glossiphoniidse and Gnathobdellidae. 



The types of the species described are preserved in the South 

 African Museum. 



I. GLOSSIPHONIID^. 



We know of no previous record of any representatives of this 

 group south of German East Africa. 



The occurrence of Placobdclla sp. calls for no special remarks, 

 since at the most it cannot have any more than mere specific 

 importance, and as ws have only a single and poorly preserved 

 specimen for examination we can do little more than record the 

 genus for these parts, and note any characters which we deem 

 reliable in a preserved specimen. The absence of the genus from 

 South Africa would give more ground for surprise than does its 

 occurrence. It is perhaps not out of place to remark here that, 

 although like so many other fresh water groups of animals, the 

 fresh water Hirudinea enjoy a wide distribution (most parts of the 

 earth possessing characteristic species), yet, at least in the main 

 fresh water group, Glossiphoniidae, this distribution cannot be 

 rashly attributed to the same ready means of dispersal as serve 

 to explain the distribution of other fresh water groups. This is 

 supported not only by the occurrence of definite characteristic 

 species in most large areas, but also by the facts that the eggs 

 are carried in most Glossiphoniidae attached to the ventral surface, 

 and experiments indicate that they cannot withstand the effects of 

 sea water or desiccation. This may be taken to indicate that their 

 dispersal has been carried out over short distances and that their 

 present universal distribution must signify vast changes in geography 

 in the past. In other words their distribution seems to prove that 

 the family is an ancient one. This is strengthened by the simple 

 nature of the somite, the comparatively well-preserved ccelorne, &c. 



For the second representative of the group described in this paper 

 we institute a new generic name, Marsupiobdella. 





