256 MR. C. F. ROUSSELET ON THE POLYZOA [Mar. 5, 



slightly expanded poi'tions at intei'vals, fi'om which a long erect 

 tube rises ; side branches arise at the base of the zocecia. 

 Polypides Avith circular lophoiohore of 8 tentacles. The whole 

 zoarium and zocecia embedded in Spongilla tcmganyikce, on the 

 surface of which the polypides protrude. 



The freshwater sponges brought back by Dr. Cunnington 

 from Lake Tanganyika include specimens of Sj^ongilla tcmganyikce 

 Evans, in which was found embedded this gymiroliematous polyzoon 

 Avhich cannot be identified with any known species. Its affinities 

 seem to lie between the genei-a Gylmclrcecium, Potts ieUa, and 

 Victorella ; but its habitat in fi^eshwatei' and the number of 

 tentacles possessed by the polypides, have decided me to place it 

 in the last-named genus. 



Potts's Paluclicella erecta (11), afterwards renamed PottsieUcc 

 erectco by Prof. Kraepelin (6), is the only other freshwater species 

 which is known to penetrate encrusting sponges, but its circular 

 lophophore has 20 tentacles. 



The meandering cylindrical stolons, glassy white when cleared 

 from adhering sponge-fragments, form an entangled mass difficult 

 to follow, and are entirely embedded in the substance of the 

 sponge. Adjoining and crossing stolons and tubes adhere, but 

 can be separated by tearing apart. No septa have been observed 

 in the stolons. 



The upright tubes, together with the slight expansions of the 

 stolon, form the cell or zooecia, and lateral branches, one on each 

 side, wei-e observed springing from the expanded stolon, but these 

 side branches are not always present. 



The tubes are long, erect, single, slightly widened at the base^ 

 of glassy transparency, except the small portion above the surface, 

 which is rendered more or less opaque by fine granules, and 

 emeru-e on the surface of the sponge where the polypides expand 

 their small circular lophophore of 8 tentacles. The length of the 

 tubes in this species is no doubt due to the necessity of reaching 

 to the surface of the sponge. In VictoreUcc pavicla of Kent (1, 5) 

 the cell is formed by a very distinct expansion of the stolon and 

 the tube arising therefrom is of moderate length. Length of 

 erect tubes up to 1-7 mm. (j^ in.). 



It has been a matter of considerable difficulty to dig the 

 complete stolon with the tubes out of the substance of the sponge 

 without breaking up these delicate structures, and the operation 

 has been only partially successful. 



Hc^b. Growing within and through specimens of S'pongillcc 

 tcmganyikce, which were foiuid encrusting rocks in shallow water 

 at Ohamkaluki, and also on shells dredged from about 20 fathoms 

 near Mshale. 



Bibliography. 



1. BousFiELD, E. C. — The Victorella pavicla of Saville Kent. 

 Ann. Mag- Nat. Hist. ser. 5, vol. xvi. 1885, pp. 401-407, 

 1. plate. 



