SILICIOUS AND HORNY SPONGES WILSON. 387 



Genus GELLIUS Gray (1867). 



Gellius Gray, 1867, p. 538.— Ridley and Dexdy, 18S7, p. 37. 

 Gelliinae in which the skeleton is typically a renieroid network 

 without fibers. But spicule tracts may also be present, or indeed may 

 replace the simple renieroid network. The latter in some species 

 becomes very irregular and halichondrioid. Spongin is scanty. 

 "When spicule tracts are present, they are not sharply defined fibers 

 as in Gelliodes. 



Gellius, in which Gray (1867) gave the skeleton as "regularly 

 netted, r has been defined by Ridley and Dendy, 1887, Topsent, 

 1894<?, and Lundbeck, 1902, as without distinct fibers or at any rate 

 without long fibers. But compact spicule tracts, including a little 

 spongin, which are not far from the fibers of many Gelliodes species, 

 are found in Gellius flagellifer Ridley and Dendy (1887), Gellius 

 perforatus Wilson (1904), and in some other forms. 



The megascleres in the genus are characteristically oxeas, but there 

 are species with strongyles (Lindgren, 1898). Lundbeck, 1902 (p. 

 71), sums up the combinations of microscleres occurring in the genus 

 as follows: Sigmas; sigmas and toxas; sigmas and rhaphides sepa- 

 rately or forming trichodragmas ; rhaphides (trichodragmas) ; toxas. 

 He would merge Rhaphisia Topsent (megascleres — oxeas; micros- 

 cleres — rhaphides or trichodragmas, or trichodragmas and toxas). 

 The number of species is growing so that some subdivision, even if 

 artificial, based on the microscleres, would be useful. 



The distinction between the group of species constituting Gellius 

 and that constituting Gelliodes is vaguer than is sometimes assumed 

 (Dendy thus, 1905, p. 137, defines Gelliodes as having fibers with 

 abundant spongin), and a brief discussion of the latter genus may 

 not be out of place here. 



Gelliodes Ridley (1884, p. 426) may be defined: Gelliinae in which 

 the skeleton includes, sometimes in addition to a finer network or a 

 more scattered skeleton, sharply defined fibers forming a reticulum, 

 which in some species is produced by the dendritic branching and 

 anastomosing of longitudinal fibers. Spongin may be abundant, in 

 which case the fiber is cored (Chalina-like) by the spicules, or scanty, 

 in which case the fiber is a mere column of spicules (Petrosia- or 

 Pachychalina-like) cemented together, but the column not covered, 

 by spongin. Microscleres — sigmas, or sigmas with toxas. 



Lundbeck, 1902 (p. 75), has materially helped in the progressive 

 definition of Ridley's genus. In the type species (Ridley, 1884, p. 

 426) and in some other species, those described by Lundbeck (1902) 

 and in G. petrosioides Dendy (1905, p. 138), for instance, spongin is 

 very scanty, and yet the fiber has a distinctness which is absent from 

 our concept of Gellius, toward which the genus shades over. Some 



