420 COLLECTIONS FII«>M M I'.LANESIA. 



matic position, which Carter (I.e.) is inclined to regard as among 

 the Renieridse, I notice that RhizoeJialina oleracea, at any rate, has 

 a true Chalinid fibre ; but Bowerbank's and Carter's species never 

 approach this condition more closely than by producing a few 

 scattered compact fibres, wholly composed of spicules, like those of 

 some Pachychalince : but the greater part of the organization is 

 Etenierid, and it appears to approach Schmidtia, Balsamo-Crevelli. 

 On the other hand, Oceanajpia, which seems to be nearly allied, has 

 the bihamate spicule. Taking this fact in conjunction with the 

 horny fibre of Rh. oleracea, it seems to me best to place the two 

 genera in the Desmacidinidaa. 



54. Rhizochalina fistiilosa, Bowerbank, var. infradensata, now 



? Alcyonium putridosum, Lamarck, Mem. Mus. Hist. Nat. i. p. 108. 

 Desmacidonfistulosa, Bowerbank, P. Z. S. l>7-'5, p. 19, pi. iv. tigs. 7, 8. 



Two more or less imperfect specimens in spirit, the largest about 

 25 millim. (1 inch) across the body, and some detached dry tubes. 

 An arrangement here found, which I have not seen described in 

 this species, is that of a dense layer of the skeleton-spicules, pacl 

 side by side, at right angles to and about -25 to -8 millim. below the 

 surface of the sponge— below, that is to say, the superficial Isodictval, 

 or rather Halichondrioid (in Bowerbank's sense) network containing 

 the subcortical crypts and other cavities. This layer recalls the verti- 

 cal layer of small subspinulates of Rinalda uberrima, Schmidt, only 

 that it is not, as there, placed at the surface. It is represented in 

 the type specimen of the species by a layer in which the spicules are 

 set obliquely to the surface at various angles. As both the present 

 specimens present this peculiarity, I think it well to establish for 

 them a distinct variet}'. 



One of the specimens exhibits the small crateriform eminences 

 figured on the outside of Bowerbank's specimen, but the other does 

 not ; hence they probably have no systematic, and but little physio- 

 logical importance. The spicules agree closely in proportions with 

 those of the type. 



Arafura Sea, Arafura Sea, 

 Type specimen. Spec. no. 1. Spec. no. 2. 



Acerate spicule . -27 by -01 1 to -0127 -25 by -Oil -25 by -012 mm. 



Hah. Arafura Sea, X.W. coast of Australia, 32-36 fms. ; bottom 

 mud, sand, and shells. 



Distribution. Fremantle, W. Australia {Bowerbank) (the typical 

 form ). 



The dermal membrane of one of the specimens contains a large 

 number of smooth acerate spicules of about half the length and 

 breadth of the proper spicules : they do not occur below the mem- 

 brane, nor, apparently, in the other specimen. A similar circum- 

 stance occurs in R. singapormsis described below; in that case a 

 number of short blunt cylindrical spicules occur of the normal, or 

 almost the normal thickness, but only one half to two thirds the 

 length of the adult form. 



