4-1 12 COLLECTIONS FROJl MELANESIA. 



has, however, done good service in pointing out an essential differ- 

 ence between the two groups into which he divides the Chalino- 

 psidinse, viz. in the way in which their echinating spicules are 

 attached. It must, however, be remarked that Echi nodi dy ion, niihi, 

 as now understood, approaches Axinella decidedly in this point. 



The presence of spined echinating spicules is not (see Raspailia, 

 infra) distinctive of the family as here constituted. 



The absence or slightly pronounced tendency to difference in size 

 and form between the corresponding spicules of allied species, when 

 the outward form of the sponge differs unmistakably, is a most 

 characteristic feature of this family, and is especially well exhibited 

 in the genera Echinonerna, Claikria, Echinodictyum, and Raspailia 

 (s. str.), whereas in Axinellida? the relative thickness of the spicules 

 usually gives good characters (sec Aeanthella, sp., p. 403, where the 

 external form differs little from A. obtusion). 



OPHLITISPONGIA. 



Ophlitaspongia, Bowerbank, Mon. Brit. Spong. ii. p. 14. 



Vosmaer (Family Desmacid. p. 107) places 0. seriata, Bowerbank, 

 the typical species of this genus, under Desmacodes, and says ( /. c. 

 p. 155) of 0. -papilla, id., which seems to me not to be specifically 

 distinct from it, that it is probably a Clathria, but that no ancho- 

 rate spicules have been described in it ; I have examined the 

 original slides (Bowerbankian) without finding anchors. The sponge 

 which I am about to describe from the present collection agrees 

 with these British OpMitispongiai (but not with the foreign ones) of 

 Bowerbank in their fibre and spiculation, except that the fibre is 

 cored by a cylindrical spicule which is wanting in 0. seriata and 

 papilla. I am inclined to believe that we have here a natural 

 genus, differing from Clathria mainly in the absence of anchorate 

 spicules. I do not see how these species can be placed under 

 Desmacodes, when they have echinating spicules, but no skeleton 

 acerates (except tricurvates) and no bihamates. 



to. 



Ophlitispongia australiensis. (Plate XL1I. figs. <•, c'.) 



Habit of Clathria fron&ifera ; spiculation of C. coralloides and 

 allied species. Massive ; structure cellular, i.e. interior and surface 

 broken up into angular cells by walls of tough denser sponge- 

 substance, projecting at surface in low ridges and slight points ; 

 between them are extended thin membranous expansions. Texture 

 in dry state firm, tough, subelastic ; colour pale dirty brown. 



Main skeleton — meshes rounded, narrow ; primary fibres stout, 

 amber-yellow, cored with about one third their thickness of spicules, 

 proceeding straight to surface, diameter about -1 to *14 mill i m. j 

 secondary fibres abundant, irregular in direction, amber-yellow, 



