Freshwater Polyzoon from S. Africa. 



267 



processes split longitudinally in the plane in which the stato- 

 blast is flattened some time before tl>e statoblast itself splits 

 into its constituent valves (fig. 3). By counting the number 

 of hooks on a given length of a process before it has split, and 

 afterwards, it is found that there were twice as many before 

 as after on a given length and that the distances between them 

 are half as great. It tlius appears that the hooks themselves 

 do not split, but are arranged in a double row on each side, 

 which, when splitting occurs, becomes divided into a single 

 row on each side of each half of the pi-ocess. 



Fig. 3. 



f 



Three immature statoblasts of L. capensis. 



Some immature statoblasts were sent by another collector 

 to Dr. Rousselet, who, on hearing that I was describing the 

 species, very kindly lent his preparation of them to me. 

 Among the immature statoblasts are some which have the 

 size and form of the central capsule. The terminal process is 

 added last of all, during the completion of the development of 

 the annulus (fig. 3). 



Annandale (' Records of the Indian Museum,' 1907, vol. i. 

 part ii. pp. 145-149) describes a new variety of Lophopus 



