Freshwater Medusa of Lake Tanjanyika. 273 



a flattened epithelium (fig. 7; s in figs. 4 and 5). These 

 egg-shaped bodies are composed of numerous cells and appa- 

 rently attached by a thin stalk or thread to that side of the 

 capsule which is nearest the ring-canal. The basal cells of 

 the marginal bodies have a granular and opaque appearance, 

 being coloured dark brown bj the action of the osmic acid. 

 The apical cells, however, are quite clear and refrinj,-ent 

 (Hg. 8). These extraordinary bodies at once recall the corre- 

 sponding structures in Limnocodium described by Professor 

 ]iay Lankester*, to which they bear a remarkably close 

 resemblance. At the same time they differ in structure from 

 all other sense-organs hitherto described in Medusae, the chief 

 difference between the marginal bodies of Limaocnida (as 

 I shall call the Tanganyika Medusa) and those of Limno- 

 codium consisting in the absence of any tubular extension of 

 the caj)sules into the adumbral ectoderm layer of the velum 

 as is the case in Limnocodium. Another less important 

 point of difl'erence lies in the marked contrast between 

 the grannlar basal cells and the refringent apical cells 

 in the marginal bodies of Limnocnida. However, in spite 

 of these differences -there is a very great similarity of 

 structure in the organs of both, and probably the marginal 

 bodies of Limnocnida will be shown to be endodermal in 

 origin, just as the " refringent bulbs " of Limnocodium have 

 been proved to be. 



Among the individuals of Limnocnida examined some 

 were found to have the outer wall of the m.anubrium quite 

 smooth, while others have the proximal half of the manu- 

 brium covered with small swellings visible to the naked eye. 

 When snbjected to microscopic examination the individuals 

 with the smooth manubriums turned out to be males and 

 females with the external wall of the manubrium covered with 

 spermatozoa or ova in all stages of development. 



The other set of individuals with the small swellings proved 

 to be examples showing stages of bud-formation. Buds may 

 originate anywhere on the basal half of the manubrium, so 

 that in some specimens several rows of buds occur, completely 

 encircling that organ. 



As far as the method of bud-formation has been observed 

 at present, it appears that the young buds first arise as out- 

 growths on the external wall of the manubrium. Besides 

 these, a certain number of older buds with rudimentary 

 tentacles occur, projecting into the stomach and present- 

 ing the appearance of having actually grown through the 



* "On Limnocodium Sotverhii, E. Ray Laiikester," Quart. Jouni. Micr. 

 Sci. vol. XX,, 1880. 



