2-2 •ENDEAVOUK" SCIENTIFIC KESULTS. 



Loc. — Near Kangaroo Island, South Australia, 17 Ims. 

 ("Endea\()ur"). 



(b). IV. uxypliiia, \ ar. ? (Plate xxxiw, fig. 3). 



The different appearance presented by the second specimen 

 is partly due to its intense purplish colour, and perhaps partly 

 also to its dry macerated condition. The colour difference is 

 of little consequence since it probably results from the presence 

 of a symbrotic alga similar to that which affects certain other 

 Australian species, notably Clialina polychotoma, Carter, 

 Aretiochalina niirabilis (Lendf.), Whitelegge,i and Echino- 

 clathria ramosa, sp. no v. 



In shape (PI. xxxiv., fig. 3) it is not materially different 

 from the typical specimen. It consists of a stipitate flabelli- 

 form plate (170 mm. in height and about the same in breadth, 

 and 6 to 8 mm. thick) to which are added, mainly on one side, 

 a few lamelliform outgrowths, joined to it along vertical lines.. 



The washed-out condition of the specimen is most unsatis- 

 factory from many points of view for the purpose of full and 

 accurate description, but in this case it possesses the ad- 

 vantage of permitting the rather characteristic gross structure 

 of the skeleton to be readily perceived. I have not been able 

 to convince myself that the type specimen, if macerated, would 

 show a similar structure. The surface is everywhere irregu- 

 larly covered with shallow, roughly polygonal or rounded 

 honeycombcell-like pits on an average 2 mm. in diameter, and 

 more or less distinctly arranged in longitudinal series running 

 from the base to the margin of the plate. VX'here the sponge 

 is thin, the "cells" may completely perforate it; when the 



1 From an examination of a section which I have prepared of a frag- 

 ment of a British Museum specimen labelled " Arenochalina mirabilis 

 Lendenfeld, Torres Straits," and a comparison of its skeletal structure 

 with that figured by Lendenfeld (Zool. Jahrb., 1887. taf. xxvii., fig. 28). 

 I feel sure that this specimen is truly representative of Lendenfeld's 

 species. I find, also, that Whitelegge's Arenochalina inirabilis, from Port 

 Jackson, is closely allied to, if not identical with, the same species. 

 Lendenfeld's description is accordingly wrong in stating that the megas- 

 cleres are oxea; they are slightly fusiform subtylostyli with relatively 

 large axial canal. Whether the typical A. tnirabilis possesses chelae or 

 not I am unable to say, since, in the fragment referred to, owing to its 

 washed-out condition, interstitial spicules are entirely absent. However, 

 in the Port Jackson sponge there occur scattered anisochelae palmatae of 

 simple form, and, since the close relationship of this sponge to Lenden- 

 feld's is beyond doubt, one can therefore say that the genus Arenochalina 

 possesses the spiculation of Mycale, and that it will probably form one of 

 the sub-genera into which the latter genus will no doubt ultimately 

 be subdivided. The external resemblance of Arenochalina mirabilis to 

 Spongelia elegans (c./., Whitelegge -Rec. Austr. Mus., iv., pi. x., fig. 7, and 

 Lendenfeld— Monogr. Horny Sponges, pi. xxxix., fig. 2) is so striking that 

 it almost casts doubt on Lendenfeld's record of the occurrence of the 

 latter species in the same area as the former, viz., at Broken Bay, New 

 South Wales. 



