38 THE VOYAGE OF H.M.S. CHALLENGER. 



Skeleton regularly netted. Spicules of two kinds: — (l) Fusiform. (2) Bihamate, simple, 

 iind contorted." The increase of our knowledge, due to recent discoveries, has necessitated 

 several not very serious alterations in the original diagnosis. Thus, it has been found 

 advisable to omit the words " sponge massive," for we now know at least one species, 

 Gellius calyx, nobis, which has a beautifully symmetrical and elegant form. Again, the 

 genus has had to be enlarged to admit certain forms with toxa, or with long, straight, 

 hair-like microsclera, although at present only a few such are known, viz., Gellius 

 angulatus, Bowerbank, sp., Gellius arcoferus, Vosmaer, Gellius Jlabelliformis, nobis, 

 Gellius pyriformis, nobis ; and it is possible that these and kindred forms may 

 subsequently require a separate genus or subgenus to include them. 



The smooth oxeote character of the megasclera appears to be a very good and 

 constant point, but it is important to notice that the ends of the oxea may become 

 completely rounded oflF, as in Gellius pyriformis and Gellius carduus, the spicules then 

 becoming almost cylindrical (strongylote), but showing their real nature by the marked 

 manner in which they still taper to each rounded extremity, and also by their close 

 correspondence in shape and curvature with those of allied forms. Intermediate stages 

 in this process of rounding off are also sometimes met with. 



Vosmaer, in the Notes from the Leyden Museum {loc. cit.), does not accept the 

 genus Gellius at all, but has in its place Desmacodes, Schmidt, ^n his later work on the 

 Porifera in Bronn's Klassen und Ordnungen des Thierreichs, he, however, substitutes 

 Gellius for Desmacodes, but gives the genus a much wider scope than we are inclined to 

 admit ; his diagnosis runs as follows : — " Stabnadelu glatt oder gedornt ; Spongin wenig 

 entwickelt. Keine Anker. Statt dessen Haken und Bogen. Atlantischer und Arctischer 

 Ocean (Mittelmeer V). Untief bis 1 8 Fad. " ^ The Challenger dredgings extend the range 

 of depth at which Gellius occurs to 600 fathoms, and show it to be as well established in 

 the deep-sea as in shallow water. 



Gellius varius, Bowerbank, sp. (PI. VIII. fig. 4). 



1875. Ealiclwndria varia, Bowerbank, Proc. Zool. Soc. Lond., p. 292. 



1875. Isodictya virgata, Bowerbank, Proc. Zool. Soc. Lend., p. 294. 



1880. ? Desmacodes fihulatus, Vosmaer, ^)ars, Notes Leyden Mus., vol. ii. p. 105. 



1884. Gellius varius, Ridley, Zool. CoU. H.M.S. "Alert," Brit. Mus., p. 424. 



With this species we identify several good-sized fragments from Station 208. They 

 are all more or less cylindrical, and the longest is a branch about 80 mm. long by about 

 6 mm. in diameter, slightly forked at the extremity. Other pieces are thicker, much 

 more distinctly branched, and the branches show a strong tendency to anastomose. The 

 measurements of the spicules are as follows : — Oxea about 0-22 by 0-013 to 0-0145 mm. 



> Vide Klass. u. Ordnung. des Thierreichs, Porifera, p. 349. 



