Volvox in South Africa. 599 



vacuoles, the periods of which do not coincide, nor exactly alternate. 

 With the oil-immersion lens they can be seen very clearly. Im- 

 mediately above the position of the eyespot and vacuoles the sperm 

 narrows suddenly to form the long, narrow hyaline beak, at the apex 

 of which the two long cilia are inserted. They are directed back- 

 wards, and this fact probably gave rise in F. globator to the idea that 

 the cilia are sometimes inserted laterally near the eyespot (Klein, 

 1890, p. 33, and Table III, fig. 23). The spiral twist often seen in the 

 sperms is very apt to make it appear as though the cilia were inserted 

 at the base of the beak (cf. the two central sperms in fig. G, Plate XLV), 

 but in life in the African species it is quite evident that this is not 

 the case ; the insertion of the cilia is apical, or so nearly apical that 

 it is impossible to detect the difference. 



Except for the slight variation in the posterior end either rounded 

 or drawn out into a short point and an equally slight one in the 

 shape of the whole, the sperms are very uniform in size and form. 

 Often the body has a slight spiral twist, apparently consequent on 

 the rapid movement. The beak, sometimes equal to one-quarter 

 the total length of the sperm, is an extraordinarily mobile structure, 

 and seems to function in some way in the location and penetration of 

 the oosphere. 



Living sperms were watched inside female colonies, where they tend 

 to collect round the egg-containing vesicles ; probably owing to the 

 greater viscosity of the mucilaginous fluid filling the cavity of the 

 coenobium, the cilary action so characteristic in water was seldom 

 seen, the movement of the sperms being of an amoeboid type. The 

 beak is constantly bending and turning, often apparently with a boring 

 movement. They were often seen close to the vesicle surrounding an 

 egg, the beak apparently trying to bore into the wall, while the cilia 

 floated out motionless behind (fig. 8, B, C ; Plate XLVI, A). 



Volvox capensis. 



In the main the above account of spermatogenesis in V. Rousseletii 

 applies also to F. capensis, but there are several points where the 

 process in the latter differs markedly from that in the former. 



In the first place, the number of male initial cells is always very 

 much smaller and the sexual colonies are always monoecious and 

 protandrous. As a rule, the 8 primary antheridia develop more or 

 less simultaneously, others often developing much later, but usually 

 all are liberated before the oospheres are ready for fertilisation. As 



