392 COMECTIONS FEOM MELANESIA. ' 



granules, lying in a diffusely stained subtranaparenfc matrix of the 

 same colour, but paler. A transparent membrane, consisting of an 

 almost colourless matrix, containing few purple granules, appears to 

 invest the fibre. 



HaK Thursday Island, Torres Straits, 4-5 fms. ; bottom sand, or 

 sand and rock. 



Distribution. " Indian Ocean " {Pallas). 



Ohs. In many particulars this species recalls Aplysina purpurea of 

 Carter, but appears to differ fundamentally in the distinctness, large 

 size, and non-multiplicity of the fibres ; -whereas in that species the 

 axes of the conuli and the skeleton generally consist of aggregated 

 masses of fine fibrils. If Hyatt's species is really like lanthella 

 homei, with which he compares it, it cannot be this sponge, as it 

 would be of flattened growth ; but he appears to be uncertain on 

 the point. 



If one of the dermal cones, with the surrounding membranes, is 

 treated with a strong solution of caustic potash, a dark brownish- 

 yellow colouring-matter is dissolved out, thereby differing from 

 that of lanthella, which is said to be violet under similar circum- 

 stances (Gray, Proc. Zool. Soc. 1869, p. 50); nothing of the tissues 

 is left but a branched fibre or two and some flocculent matter ; 

 therefore the only truly fibrous structures here are the terminal 

 twigs of the skeleton. 



The wall of the main skeleton-fibre of this species is much thicker 

 than in most Aplydnce, and its axial substance is not granular. 



20. lanthella flabelliformis. 



Spongia flabelliformis, Pallas^ Elench. Zooph. p. 880. 

 lanthella flabelliformis,. Gray, P. Z. S. 1869, p. 50. 



A specimen in spirit, somewhat imperfect, and not showing any 

 of those exfoliations of the lateral surfaces which specimens com- 

 monly exhibit. 



Hah. Port Darwin, 7-12 fms. ; bottom sand. 



Distribution. Indian Ocean (PaZZas) ; " Australia" (ffraj/). 



CHALINIDJE. 



In this family must now be included some forms with minute 

 fiesh-spicules. Homoeodictya, Ehlers, is the earliest discovered case 

 of this combination ; and I am able to add another, in a new genus, 

 Toxochalina, which possesses fine tricurvates scattered in the sarcode. 

 I have relegated Bhizochalina to the Desmacidinidse for reasons given 

 below (family Desmacidinidse). I have also given reasons for a 

 belief that Platyehalina, Ehlers, is a Ceratose sponge (see Euspongia, 

 supri). 



The fact cannot bo ignored that the genus Toxochalina, charac- 

 terized below, runs counter to the plan of classification hitherto 



