PORIFERA. I. 



71 



the length is the same as in the preceding species, and is between 0-028 — o-io7""". In the largest ones 

 the thickness reaches to o-oo2"""; but in the smaller it is somewliat less. 



Locality: Egedesminde (Bergendal), depth 15 fathoms, one specimen; the mouth of the Ameralik 

 fjord (Ingolf) one specimen; both specimens grow on Balanoid shells. 



Although I have only had so slight a material of these two species I have thought it best to 

 describe them, as their spiculation is interesting, and separates them very well from the other Gellius- 

 species. Of the hitherto known GeUms-s^pecies provided with toxa, G. angulahis Bow., arcoferus Vosm., 

 and fiabelliformis Ridley and Dendy, have also sigmata; only G. pumicetts Frstdt. (Kgl. Sv. Vetensk. 

 Akad. Handl. Bd. 21, no. 6, p. 29, T. II, fig. 9 a— d), and G. toxins Tops. (Rev. Suisse de Zool. IV, 1897,470) 

 have a spiculation consisting of oxea and toxa without sigmata; but in puiiiiceus oxea have a length 

 of c.g""", and toxa have another form; neither seems G. toxins, of whose spicules we, unfortunately, 

 have no figures, to agree with any of the species described here. 



Note. According to what is stated above, the genus Gcllius may have the following combi- 

 nations of spicules: oxea and sigmata (the most common case); oxea, sigmata and toxa; oxea, sigmata 

 and rhaphides, in bundles or scattered {inassa Cart, iiiacrosigiiia Tops., G. sp. (pyriformis) Ridley and 

 Dendy, Challeng. Report, Monaxonida, 46, and p/iillipensis'D&wdy, Proc. Roy. Soc. Victoria VII, 1895, 247); 

 oxea and rhaphides [Lacazci Tops., Arch, de Zool. exp. et gen. Ser. I, I, 1893), and finalh' o.xea and toxa. 

 I think it doubtful whether the genus Rhaphisia Tops, the spiculation of which consists of oxea and 

 rhaphides or trichodragmata (ai/oiiyuia Cart, laxa Tops.), or of oxea, trichodragmata and toxa (spissa 

 Tops.), can be kept up, as the character given by Tops en t, of the skeleton forming no regular net- 

 work, is scarcely sufficient to make good the establishing of the genus; and it must therefore, I think, 

 be referred to Gcllius. The bows figured by Topsent for Rhaphisia spissa (Mem. de la Soc. Zool. de Fr. 

 XI, 1898, 232, fig. 2 c) remind of the bows in the two species described above, but they are larger. 



7. G. fiagellifer Ridle}- and Dendy. 

 PI. II, Fig. 9, PI. XIV, Fig. I a— d. 



1886. Gcllius nagtilifer Ridley and Dend\-, Ann. Mag. Nat. Hist. Ser. 5, XVIII, m. 



1887. — — Ridley and Dendy, Challeng. Report, Monaxonida, XX, 42, PI. XIII, figs. 5 and 10. 



Ohlong-cushionshapcd. The surface somewhat shaggy. The dermal membrane thin, resting on 

 the subjacent reticulation of polyspictdar fibres. The skeleton consists of an irregular network of poly- 

 spicular fibres. Spicula: Megasclcra oxea o-^iy— 0-447'""' ; microsclera common sigmata 0-02— o- 07'"'", 

 flagellate sigmata o-oy8—o-og""". 



Of this species we have one specimen agreeing both in outer form, structure and spiculation very 

 well with the quoted descriptions. The form of the specimen is as an oblong cushion, and it appears 

 to have been attached. It has a length of 34""°, a breadth of 20'"'", and a thickness of ca. 13"^'". The 

 colour (in spirit) is \ellowish white. The surface is somewhat shagg)- from projecting spicules. The 

 dermal iiiciubrane is thin and transparent, and rests on the subjacent skeleton which forms, below the 

 skin, a coarse reticulation of polyspicular fibres. From the nodes spicules or bundles of spicules arise, 



