THE TANGANYIKA PROBLEM. 321 



(B) The Spicules. — It is generally stated that the skeletal 

 spicules of the several species of the Spongillidae have no 

 characters of higher than specific value. It is difficult to 

 make out from the literature of the family how far such a 

 statement is justified. However, the spicules of Spongilla 

 tanganyika possess such characters that it is almost im- 

 possible to believe that they have not a wider application. 

 This sponge, considered from the point of view of the 

 skeleton, seems to present a certain amount of affinity with 

 a few species of the Spongillinae on the one hand and 

 of the Lubomirskinae on the other. 



The megascleres of the greater number of species arrayed 

 under the sub-family Spongillinae are sharp-pointed, that is, 

 they are either amphioxea or amphitornota. There are, 

 however, a few species which possess spicules with rounded 

 ends, that is, amphistrongyla. The species in question are 

 Spongilla nitens (Carter), Spongilla bohmii (Hilgendorf) 

 and Spongilla loricata (Weltner), to which may be added 

 Spongilla tanganyika:, now described for the first time. 

 Spongilla tanganyika?, therefore, seems to be more closely 

 related to these species, so far as the characters of the 

 skeleton are concerned, than to any other species of the 

 Spongillinae. Of the three specie named above it appears 

 to present closer affinity with Spongilla bohmii than with 

 either of the other two, for in Spongilla nitens and in Spon- 

 gilla loricata the amphistrongyla are smooth, while in both 

 Spongilla bohmii and Spongilla tanganyika- they are spiny. 

 In the former the spines are more thickly set at the end, 

 which is a special feature of the megascleres of some species 

 of the Lubomirskinae, and which may point to a certain 

 amount of affinity in that direction, while in the latter they 

 are evenly distributed over the whole spicule. In Spongilla 

 bohmii the megascleres are curved as in Spongilla nitens, 



21 



