Sponges from Funafuti. 3. r >9 



Polyjibrospongia Sweeti, sp. n. (PI. XV. fig. 2, a-c.) 



Sponge forming an irregular incrusting mass, 4 centim. in 

 length, 2*5 centim. in width, and 1 centim. in thickness, 

 nearly encircling the stem of a Gorgon ia ; with long mem- 

 branous yellow oscular tnbes about 2 centim. in height, 

 occasionally branched. Dermal membrane smooth, perga- 

 mentaceous, yellowish white. 



Skeleton. — Ectosomal : composed of stout primary fibres 

 30 to 60 fi in thickness, partly cored with foreign bodies, 

 the meshes being filled in by thinner clear fibres and crowded 

 with foreign bodies ; the oscular tubes consisting of a con- 

 tinuation of the dermal layer, but witli the cored fibres forming 

 a regular network with square meshes. 



Choanosomal skeleton composed of an oval-meshed network 

 of loose bundles of slightly branched anastomosing fibrils, the 

 diameter of the bundles being 210 ft and of the fibrils 9 /jl ; 

 also of sparsely scattered slightly branched fibres 50 fi thick, 

 and passing from the base to the surface, composed of foreign 

 bodies cemented by spongin. 



Filaments absent. 



This remarkable species is nearly related to Polyfibro- 

 spongia jlabellifera, Bowerbank, a thin flabellate species from 

 New Guinea, but differs in shape and in having the oscular 

 tubes. The type specimen of Bowerbank's species is in the 

 Dresden Museum, but the Bowerbauk collection contains two 

 slides prepared from the type, and quite sufficient to show the 

 close similarity in the skeletal structure of the two species. 



Lendenfeld puts Bowerbank's species in a subgenus Poly- 

 jibrospongia under the genus Hircinia. After a careful search 

 I have been unable to find filaments in either of the above 

 two species, though these mysterious objecrs are abundant in 

 a specimen from Port Jackson named by Lendenfeld Hircinia 

 (Polyjibrospongia) gigantea. 



To include H. gigantea in the same genus with Polyjibro- 

 spongia Sweetiandflabelltferawould be tantamount to regarding 

 the filaments as unessential elements from the systematic point 

 of view. I have not at present a definite opinion on this difficult 

 controversial question of the systematic value of the filaments, 

 a question on which equally eminent spongologists take 

 diametrically opposite views. 



For the present I shall retain Bowerbank's genus with 

 the two species Sweeti and jlabellifera, placing it next to 

 Stelospongus, and shall leave H. gigantea and fasciculata 

 under Hircinia. The new species is named after Mr. G. 

 Sweet, whose munificent donations rendered possible the 

 success of the Australian Boring Expedition. 



Locality. W. of Tutanga, 86 fath., Funafuti Atoll. 



