566 Annals of the South African Museum. 



In the Rietfontein cultures numerous counts gave from 102 to 230 

 per colony, with an average of 180, but occasionally counts as low as 

 60 to 70 were made, while in the Kimberley material the numbers 

 were very much greater in one large colony over 400 were counted. 



2. F. capensis. On the Cape Flats the average is about 120, but 

 as in F. Rousseletii the limits of variation are wide, from 60 to 160. 

 Occasionally a strain has been found which has a very low average 

 of oospores, i.e. 60 to 70, but usually the number is over 100. 



As with the other reproductive cells, external conditions as well as 

 inherent characteristics are operative, higher temperatures apparently 

 causing the production of large numbers of oospores. 



Development of the Reproductive Cells. 

 A. Gonidium. 



1. In both V. Rousseletii and V. capensis the gonidium is at first 

 distinguished from the vegetative cells by its larger size, somewhat 

 darker colour, two or more pyrenoids, and large nucleus and nucleolus. 

 Like the vegetative cells it has two long cilia and several contractile 

 vacuoles. In surface view it is more or less star-shaped, and is 

 surrounded by a ring of some 9 to 12 vegetative cells with which it is 

 connected by a number of stout protoplasmic strands (fig. 2, A, B, 

 Plate XXXVIII, G). After the birth of the colony the gonidium 

 continues to enlarge, and during this second phase of enlargement 

 the cilia are lost, the gonidium sinks inwards, and the shape alters 

 considerably. In side view it is very broad in proportion to its polar 

 diameter, with a massive basal chloroplast and large conspicuous 

 nucleus ; in surface view it appears round instead of star-shaped, 

 and the connecting strands are becoming drawn out (fig. 2, C). 



Since this enlargement takes place within the peripheral layer of 

 vegetative cells, the latter are not much displaced thereby, except 

 that as the gonidium sinks from its original place in the outer layer 

 it draws down with it the circle of cells with which it is in direct 

 protoplasmic connection (Plate XXXVIII, E). Thus the developing 

 daughter lies beneath a marked depression in the parent, this depres- 

 sion or pore being floored with the outer wall of the original gonidial 

 cell. Eventually this becomes the pore of liberation for the mature 

 daughter. The space occupied by this pore is not much larger than 

 that occupied by the gonidium in its later phase of enlargement. 

 The fact that the number of cells round the gonidium is about double 

 that round any one somatic cell is another argument for regarding 



