562 Annals of the South African Museum. 



numbers than the latter. An area of varying extent about the 

 anterior pole is always free of reproductive cells of any kind. 



Number and Arrangement of Reproductive Cells. 



A. The Gonidium. 



The gonidia are usually confined to the posterior half of the colony. 

 In both species the number of gonidia which develop is most commonly 

 8, arranged alternately in two planes, one plane approximately 

 equatorial, the other half-way between equator and posterior pole. 

 There appears to be no rule governing the order of development, nor 

 is there any variation in size attributable to position in the colony 

 (Plate XIII, D, and Plate XX, A). The number is, however, by no 

 means constant ; it may vary in F. Rousseletii from extreme cases of 

 1 up to 16 (cf. Plate XVIII, A to D) or in V. capensis to 20, the highest 

 number as yet seen in one colony of this species (Plate XIV, A). 



Sometimes a number of gonidia which have not divided, or have 

 divided once or twice and then ceased development, can be seen in a 

 colony ; particularly if fewer than 8 have developed, the full number 

 is usually made up by such abortive gonidia (Plate XX, A). 



In F. capensis it is not unusual to find cells which are like the 

 majority of the somatic cells, but are very slightly larger and richer 

 in food reserves, often without visible protoplasmic connections. It 

 is possible that these represent incipient gonidia as postulated by 

 Klein in F. globator (1890, p. 18). They are, however, very different 

 in appearance from the undoubted abortive gonidia which in colonies 

 with few daughters are nearly always to be found in the positions 

 which daughters should normally occupy. More probably these 

 aberrant cells are somatic cells which have in some way been cut off 

 from their neighbours by the withdrawal of the connecting strands 

 and in which metabolic substances have accumulated. 



When more than eight daughters develop they tend to lie more or 

 less in the same two planes as the eight primary ones (Plate XVIII, A), 

 but are not confined to these planes. They are rather more regularly 

 disposed in F. Rousseletii than in I*, capensis. 



West's (1910, p. 102) description of the daughter colonies in F. 

 Rousseletii as " regularly eight " is misleading and founded on observa- 

 tion of a single collecting of material obviously at an early stage of an 

 asexual phase. In his account (1918, p. 1) of the second lot of material 

 from Ussangu West makes no further remark on the asexual colonies, 

 and his figures show only sexual colonies. 



