Mr. II. J. Carter o?i the Genus Rossella. 117 



arm 34 by l-6000th of an inch in its greatest diameters, that 

 of the inflation and rays about 7j-G000ths of an inch long 

 (N.B. Tliis is the characteristic rosette of the species). Size 

 of entire specimen 1| inch long by 10-12ths of an inch broad, 

 and 7i-12ths of an inch tliick ; aperture about 7i-12ths of 

 an inch long by 2-12ths wide; margin thick, round; depth of 

 cavity 1^ inch; thickness of wall about 3-12tlis of an inch. 



Hub. Ocean- bed. 



Loc. Antarctic Sea in 300 fathoms, and lat. 74^° S. 



Obs. Tlie hexactinellid character of the spicules of this 

 sponge, together with the free termination of the " anchoring- 

 spicule " in four stout spines or hooks recurved and opposite 

 (PI. X. fig. 3), characterizes the genus, viz. that of ^^ Rossella]^^ 

 while the erect beard of spicules round the aperture (fig. 4, c), 

 and the peculiar form of the flesh-spicule (no. 10, fig. 6), which 

 is by far the most abundant, determine tlie species, viz. R. aat- 

 arctica. Tiiere is no /V6e<fc-like flesh-spicule that I have yet 

 seen wherein the arms are so distinctly, altliough so sparsely, 

 spined, and the rays so parallel, so little divergent at their 

 extremities, and so little inflated or capitate. (Altogether, 

 the slightly inflated end of the arm, and tlie microspined rays 

 which it supports, are a miniature form of the head of the 

 scopiform spicule of Aplirocallistes beatrix^ 'Annals,' 1873, 

 vol. xii. pi. XV. fig. 2.) It is not improbable that there are 

 other forms of the rosette flesh-spicule present in this species 

 besides those described ; but if so, I have not seen them, 

 and if there are any, they are of no consequence in a specific 

 point of view after no. 10. 



R. antarctica further differs from the two following species, 

 so far as my observations extend, in not possessing the other 

 flesh-spicules or forms of rosette which are common to both 

 R. velata and R. philippensis ; while it agrees with R. velata 

 in the more or less developed state of all the arms of the sex- 

 radiate spicule of the latticework layer on the surface, thus 

 differing from R. philijypensis, in which for the most part the 

 four horizontal arms alone are present. 



I have described the monticular and latticework layer of 

 the surface in a much more perfect state than it exists in the 

 specimens of R. antarctica to wliich I have alluded, where, 

 from rough usage at some time or other, as in some of the 

 specimens of R. lihilippensisj a great part of the latticework 

 layer has been abraded, thus rendering the cribellate and 

 monticular surface below more evident ; but still enough of 

 the former remains here and there to show what the specimen 

 was in its entirety. 



From tlie presence of several minute specimens of this 



