322 THE TANGANYIKA PROBLEM. 



Spongilla loricata and most of the Lubomirskia, while in 

 Spongilla tanganyikce they are straight. However, there is 

 among the Lubomirskinse a variety of Lubomirskince inter- 

 media, described by Dybowski as var. a, in which the 

 spicules are spiny and almost straight. The spines are 

 evenly distributed, and in many cases the ends of the 

 spicules present the amphistrongylote character. Another 

 feature of Lubomirskia intermedia agreeing with Spongilla 

 tanganyikce is that the microscleres are smooth and almost 

 equal to the megascleres in length. In Lubomirskia bracili- 

 fera and Lubomirska papyracea the spicules are Amphi- 

 strongylote, though in the former the spines are arrayed at 

 the ends of the spicules, in contrast with those of Spongilla 

 tanganyikce, but to a certain extent agreeing with those of 

 Spongilla bokmii, while in the latter the spines are evenly 

 distributed over the shaft of the spicule, in contrast with 

 those of Spongilla bokmii, but similar to those of Spongilla 

 tanganyikce. 



From these points of comparison it seems that Spongilla 

 tanganyikce, as well as Spongilla bokmii, must be closely 

 related to the Lubomirskinse. Had it not been for the 

 presence of the gemmule in the small piece of Spongilla 

 tanganyikce, I should certainly have placed it among the 

 Lubomirskinae. On the other hand, were the gemmules to 

 be found in any species of the Lubomirskinse, the Tanganyika 

 form would have to be removed from that sub-family as at 

 present defined. Consequently, I venture to suggest that 

 the sub-family Lubomirskinae should be abolished and the 

 species contained in it placed under the Spongillinse, which 

 then could be arranged into a number of genera according 

 to the character of their megascleres. 



