SpicultS.(l,) Short, stoul, fusiform, slightly curved oxea (Plate IX., fig. :t, a, 6, c), 

 usually sharply and fairly gradually pointed at each cud (often Incoming strongylote 

 or stylote); measuring about 0'25 millim. by 0'017 millim., but varying a good deal 

 in thickness. 



(2.) Sigmata (Plate IX., fig. 3, d) ; slender, C-shaped or contort, about 0'022 millim. 

 from bend to bend ; very numerous. 



Were it not for the presence of the sigmata, this species would be a typical Petrosia. 

 It appears to be very nearly related to TOPSENT'S Gelliodes fayalenxis* from the 

 Straits of Pico-Fayal, but is distinguished by the absence of the large oscula, and, 

 perhaps, by other characters. 



R.N. 146 (deep water off Galle and onwards up West Coast of Ceylon). 



Gelliodes petrosioides, var. fibrosa, nov. 



I propose this name, at any rate provisionally, for a single small, much-damaged 

 specimen of irregular shape and cavernous structure, with large exhalaiit canals and 

 vents (?) and very soft, fibrous texture ; agreeing very closely with the type of the 

 species in spiculatiou and in the structure of the main skeleton fibres, but with the 

 fibres better defined and the meshes of the reticulation mostly very wide and not 

 filled up by scattered spicules. The soft texture of the specimen, which is in 

 striking contrast with the hardness and density of the type, is due simply to this 

 greater laxity in the skeleton arrangement. 



Except for the presence of the sigmata, this variety closely resembles a Pachy- 

 chalina. It may ultimately, when better specimens are forthcoming, have to be 

 considered as a distinct species. 



R.N. 272 (deep water off Galle and onwards up West Coast of Ceylon). 



Toxochalina, RIDLEY. 



Gelliinse with much spongin, more or less completely enveloping or even replacing the 



megascleres, and forming distinct fibres. The microscleres are toxa. 

 This genus differs from Gelliodes only in the replacement of the sigmata by toxa, 

 and in view of the occurrence of both these forms in the same species of Gell'ms 

 (vide under Gelliux angulatus, var. canaliculata), it may be questioned whether the 

 two should be kept distinct. They are interesting as forming an obvious starting 

 point in the evolution of the great sub-family Chalininaj. 



Toxochalina robusta, RIDLEY. 



1884, Toxochalina robusta, RIDLEY (16). 



There are a number of specimens of this sponge in the collection, agreeing closely, 

 both as regards external form and skeletal characters, with RIDLEY'S description of 

 the type from Port Jackson. It is perhaps noteworthy, however, that the megascleres 



* TOI-SKXT (48), p. TS. 

 I 2 



