228 PERCY SLADEN TRUST EXPEDITION 



opinion as yet. With regard to the Oeratosa I adhere to the views I have already 

 expressed [1905]. Only the " Pseudoceratosa " have any claim to be included in the 

 Tetraxonida, and whether these can be constituted into a distinct sub-order is extremely 

 doubtful. Some of them appear to me to be certainly Chalinine in origin, while others 

 are very possibly Ectyonine. 



Sub-order 1. Homosclerophoea Dendy [1905]. 



Tetraxonida in which microscleres and megascleres have not yet become sharply 

 differentiated from one another and no trisenes are as yet developed. 



I cannot agree with Hentschel in including [1909] the Oscarellidse in this sub-order, 

 which seems to me to be a distinctly retrograde step. I adhere to the opinion which 

 I expressed in 1905, that Oscarella must be placed in a separate order, Myxospongida, 

 which represents the common ancestors of all the siliceous sponges, both Triaxonida and 

 Tetraxonida, and also of the Euceratosa. 



Family Plakinidse. 



With the characters of the sub-order. 



This family was proposed by Schulze in 1880 for the reception of the three genera 

 Plakina, Plakortis and Plakinastrella. References to its history between 1880 and 1900 

 are given by Lendenfeld [1903, p. 118]. In addition to the three original genera 

 Lendenfeld includes in the family Corticium and Thrombus. In 1905 I removed 

 Plakinastrella from the Plakinidai on account of the presence of short-shafted trisenes, 

 and placed it in the Pachastrellidse, an arrangement to which I must adhere. 



Genus Dercitopsis Dendy [1905]. 



Plakinidse with calthrops, oxea and sometimes triods, but no candelabra. AU 

 spicules smooth. 



When I proposed this genus in 1905 I unfortunately overlooked the existence of 

 two species which must certainly be taken into account in discussing its aflSnities, viz. 

 Plakinastrella clathrata, described by Kirkpatrick [1900 b] from Funafuti, and P. oxeata, 

 described by Topsent [1904 a] from the Azores. More recently Lendenfeld [1906] has 

 described a species under the name Plakinastrella mammillaris, from the west coast 

 of Australia, which must also be considered in the same connection. That all these three 

 species are closely related to my Dercitopsis ceylonica there can be no doubt, but I am 

 not disposed to agree with Lendenfeld [1906] that Dercitopsis ceylonica should be 

 associated with them in the genus Plakinastrella. On the contrarv I think that all 

 three should be removed from Plakinastrella and placed in Dercitopsis, and that for 

 the following reasons. 



The type species of Plakinastrella is P. cop>iosa, described and figured by Schulze 

 [1880]. That species possesses well-differentiated, shoi-t -shafted trisenes, definitely 

 orientated beneath the surface of the sponge, and, as already stated, it was mainly for 

 that reason that in 1905 I placed it in the Pachastrellidse. Not one of the species placed 



