78 



PORIFERA. I. 



cules and especial!}' by the form of the oxea, which is quite different from that in G. plrxa\ for the 

 fact that the sizes of all three forms of spicules are above those in G. plexa, can scarcely by itself be 

 reo-arded as a sure mark of separation between the two species. 



Locality: Station 113, 69° 31' Lat. N., 7'^ 06' Long. W., depth 1309 fathoms (temperature — i°-oC.), 

 one fragment. 



These two Ge///odc's-s])ecies differ from the hitherto known siDecies b)- being possessed of toxa. 

 Thus they show an approaching to the genus Toxochali^ia^ but in this genus no sigmata should be 

 found. Ridley and Dendy (Challeng. Report, Monaxonida, XX, 47) say that Gelh'odes is only differ- 

 ent from Toxochaliiia by the presence of sigmata in stead of toxa, and they continue: It is perhaps 

 doubtful whether the last character (sigma instead of toxa) is of generic value, and whether Toxo- 

 chaliiia and Gelh'odes should not be merged in one genus, but as no species is yet known, whose spi- 

 cular complement comprises both toxa and sigmata, they may at present be kept apart . Now the 

 two species described here show just this combination of toxa and sigmata, and consequently the genus 

 Toxochaliiia^ I suppose, must be merged in Gelliodes. I shall, hovewer, call attention to the fact that 

 in the diagnosis of Toxachalina the fibres are said to be - rectangularly arranged , while, in the two 

 species described here, they are irregularly branched, what seems upon the whole to be predominant 

 in the Gelliodes-s,-^^c\Q.'s>. 



Oceanapia Norman. 



The sponge consisting of a round body from zuhich more or fewer, branched or itnbrandied 

 fistnlcE arise, lejhich arc closed at the ends. Outermost the body is surrounded by a hard, rindlike layer. 

 Spongin present. Spicula: Mega sclera diactinal, o.xea ; microsclera sigmata. 



As mentioned under Phlceodictyon, I place Oceanapia inider the Gelliince, as I regard it to be 

 closely allied to Gelliodes. The character that separates it from this genus, would even seem chiefly 

 to be the form only; biit nevertheless the genus may, at all events for the present, be kept ujd. 



At present perhaps only the one species is known, as the only one of the three Oceanapia- 

 species described by Dendy that is possessed of sigmata (Proc. of the Roy. Soc. Victoria VII, 1895, 248), 

 O. mollis, according to its skeletal structure, is more likely to be a Gel Hits, what Dendy himself so 

 far directs the attention to, as he sa}'s that it is very closel}- related to Gellius. Of the two species 

 established by Topsent (Rev. Suisse de Zool. IV, 1897, 4^75 P^- XIX, fig. 13, PL XXI, fig. 29; 469, 

 PI. XIX, fig. 17), amboenensis ^\\A fragilis, the external form is not known, and both may perhaps be 

 referred to Gelliodes. Of the Rhizochalina fbulata established by Schmidt (Spong. Meerbus. v. 

 Mexico, 76), nothing can be said on account of the incompleteness of the description'). 



I. O. robusta Bow. 



PI. XV, Fig. I a — c. Figs. 2 — 4. 



1866. Isodictya robusta Bowerbank, Mon. Brit. Spong. II, 304, 20. 



1866. Dcsmacidon feffreysii Bowerbank, Mon. Brit. Spong. II, 347, 2. 



') On the other hand, as will be seen hereafter, there is a possibility that Ridley and Dendy's O. robtista is not 

 identical -nath the species of Bowerbank. The footnote of these authors under Rliizoclialiiia fisttilosa might Ukewise make 

 it possible that here we had still another species of Oceanapia. 



