THE TANGANYIKA PROBLEM. 



305 



this process. Instead of the evaginations of the manubrial 

 wall of the parent leading into single buds, these evagina- 

 tions branch, so that a short string of buds are, as it were, 

 attached to a hollow pipe, and before the manubriums of 

 these buds have broken through, the string of vigorously- 

 pulsating individuals becomes detached, and for a time 

 floats and swims about, presenting a singular analogy to a 

 Siphonophoran (cf. Fig. 1). Eventually the individual 

 buds become constricted off from one another, and the 



Fig. 4. — Young asexual bud of 

 the Tanganyika medusa X 10, 

 showing the base of the gastric 

 cavity still lined with endo- 

 derm. 



final stages of their evolution seems to be the same as 

 that witnessed in the case of the sessile buds. 



Returning now to the consideration of the life-history of 

 Limnocnida, it has been seen that at the end of the wet 

 season, that is in March, there are a few adult medusae in 

 the lake which are only to be found with great difficulty. 

 Such individuals begin about this time to bud, asexually, in 

 the manner I have just described, and the buds thus formed, 

 themselves produce buds, this process being repeated 

 through many generations, so that in June and July the lake 



20 



