120 Mr. H. J. Carter on the Genus RosscUa. 



a specific distinction, any more tlian tlie same complement of 

 spicules is always acconipaniod by the same form of sponge : 

 thus, two sponges may be almost uiulistinguishablc in their 

 general forms, and yet, after all, be totally ditVerent in thc^ 

 forms of their spicules respectively. Hence the necessity of 

 examining even/ specimen of sponge microscojncally before we 

 decide on its specific characters. 



As in 7?. antarctica, so here we have younger specimens of 

 B. phiUp2>e7isis (fig. 2) growing u]X)n the hair-like locks of 

 the older ones (fig. 1,/), but much larger in dimensions, 

 similar to those noticed in the concluding ])art of my descrip- 

 tion, which suggested to Dr. Gray the name of '■^PsetctUa glo- 

 hulosa'''' (/. c). But whether originating in ova or puUulation 

 I am, as above stated with reference to the minute ones on 

 R. antarctica, unable to determine. 



Lastly, it might be observed generally that although the 

 hooked extremities of the anchoring-spiculcs have been for 

 the most part torn off, there are many among them, especially 

 coming from the upper part of the sponge, which naturally 

 have never had any, but have always been fine-pointed. 



Rossella velata, Wyville Thomson (' The Depths of the Sea,' 



p. 418). 



General form ovoid, hollow ; ti-uncatc and open at the upper, 

 closed at the lower end. Aperture subcircular, slightly widen- 

 ing inwards from a thin margin to a cavity of much the same 

 shape as the sponge itself externally. Sessile or fixed by 

 anchoring-spicules. Colour brownish grey. External surface 

 uniformly net-like and monticular, resting on a widely cancel- 

 lated structure below, and covered by the latticework spicular 

 layer above, which is again surmounted by three forms of 

 appendicular spicules, whose relative positions and forms will 

 be described hereafter. Internal surface or that of the cavity, 

 the structure of the body or wall, and that of the latticework 

 layer the same as in both the foregoing species. Pores and 

 vents respectively the same in form and situation, ilonticules 

 of the surface round or boss-like, looking like so many stelliform 

 eminences regularly linked together by interradiating lines. 

 Spicules of three kinds, viz. appendicular, structural, and 

 flcsh-spicules. A. Appendicular of three forms, viz. 1, pointed 

 at each extremity, and 2, four-hooked at the free end (like 

 nos. 1 and 2 in Ji. antarctica respectively), associated, scantily 

 scattered over tlie upper part of the body in small groups issu- 

 ing from the summits of the boss-like eminences, becoming 

 more numerous towards the lower part, where they attain 



