SPONdES. 63 



species, and it is quite possible that a new genus may be necessary for the reception 

 of the Ceylon sponge, but this is a point which cannot be decided until we have more 

 information as to the European species. 



Hexadella in,<Uc-ci may, as I have already suggested, possibly be identical with 

 CARTER'S Halisarca, ritbitingens (5), which was described from dry material from the 

 Gulf of Manaar and referred to the genus Halisarca provisionally by its author. 

 Having examined one of Mr. CARTER'S original preparations of Halisarca rubitingens, 

 now in my possession, I have come to the conclusion that we cannot say for certain 

 what that organism was, and that it will be best to abandon the name altogether. 



R.N. 26, 40 (Gulf of Manaar), 254 (south of Adams Bridge, 5 fathoms). 



ORDER: TRIAXONIDA (HEXACTINELLIDA). 



Non-calcarea in which the skeleton is composed of siliceous spicules whose funda- 

 mental form is triaxonid and hexactinellid. 



Owing, doubtless, to the collecting operations having been confined to comparatively 

 shallow water, there are no representatives of the Hexactinellida in the collection. 

 There are, however, a few specimens of the so-called Hexaceratina which may be 

 found ultimately to belong here. As, however, I do not consider that their 

 systematic position is yet definitely settled, I have decided to deal with them later 

 on, in connection with the other horny sponges (see below, p. 200). 



ORDER: TETRAXONIDA. 



Non-calcarea in which the fundamental form of the spicule is tetraxonid and 

 tetractinellid. The spicules may, however, be more or less reduced, and also to 

 a greater or less extent replaced by spongin or even sand. 



It will be seen that I use the term Tetraxonida in a wider sense than that in 

 which it is employed, for example, by MINCHIN, in his admirable article in 

 LANKESTER'S ' Zoology.' It appears to me necessary to do this in order to bring out 

 the most fundamental character of the group, viz., the primitively tetraxonid and 

 tetractinellid character of the spicule, in contrast with the triaxonid and hexactinellid 

 character of the spicule in the Triaxonida.* The term Tetractinellida may be 

 reserved for those non-Lithistid Tetraxonida which retain to a greater or less extent 

 the primitive spicule-form, a form which doubtless characterised the ancestors of all 

 the Tetraxonida, but has been lost by reduction in many of the existing members of 

 the group. 



The Lithistida have evidently developed along special lines ot their own, and 



* The occurrence of triocls in the most primitive family of the Tetraxonida (viz., Flakinidte) suggest.- 

 a triaxonid (and triradiate) precursor of the tetraxonid and tetractinellid spicule. The triaxonid spicule 

 of the Plakinidse is, however, a fundamentally different form from that of the Triaxonida (Hexactinellida). 



