532 KEVISIOX OF THE AXINELLID^. ii., 



membrane thin, without contained megascleres. Skeleton witli 

 a central axis, in which the megascleres for the most part are so 

 disposed as to produce a lattice-like pattern, and in which 

 (except in the older portions of the sponge) spongin is only 

 scantily developed. Extra axial skeleton consisting of numerous, 

 short, pauciserial lines of (relatively very long) spicules, radiat- 

 ing from the central axis to the surface, — the spicules composing 

 which are more or less divergently directed, and are not united 

 by visible spongin. Megascleres : slightly curved styli, 320 to 

 1525//, in length by 18/x in maximal stoutness. Microscleres : 

 sigmataof two sizes, respectively 18/x and 50/x in maximal length, 

 the larger ones in part occurring in dragmata; and fusiform 

 trichites, 22 to iSfj. in length, occurring both in dragmata and 

 scattered singly. 



Zoc — Great Australian Bight (exact localitj' unknown). 

 External characters. — The single specimen (PI. xxxiii., fig. 4)^ — 

 280 mm. in total height — consists of about half-a-dozen more 

 elongated or main branches, 130 to nearly 200 mm. in length, — 

 one of which is a direct continuation upwards of the stalk and 

 gives off the others at different levels,— and of a score or so 

 shorter branches, ranging from 5 to over 100 mm in length, 

 which arise from llie former at distant intervals, and nearly 

 always proceed off from them at ver}' wide angles, often almost 

 or quite perpendicularly. The mode of branching, therefore, is 

 not dichotomous (as it usually is in the case of ramose sponges) 

 but irregular. The l)ranches are at most 55 mm. in diameter 

 proximally, and diminish in stoutness to slightly less than 2-5mm. 

 at their extremities. Tlie stalk has a length of 55mm. measured 

 from its base to the origin of the first branch, and terminates 

 below in a tuft of branched rootlets. The species is very similar, 

 in general habit, to Kaspailia tenuis Ridley and Dendy(^33). 



The specimen, although in alcohol, is not in a very good state 

 of preservation, the superficial layer being much damaged and 

 the dermal membrane almost completely destroyed through 

 maceration. Whether there are oscula or not, is accordingly 

 not evident; but, if present, they must be rather small and in- 

 conspicuous. The surface is everywhere hispid with far project- 



