38 li. KIRKl'ATUIt'K. 



A single round oscule at the summit. Pores nnind or oval, scattered, 20 m 

 in diameter. Dermal layer not detachable; sub-dermal spaces 190 m in depth. 

 Flagellated cluimbers about 30 /* in diameter. 



Skeleton composed of rather thick main fibres about 1 spicules broad viewed 

 en face, ahout 190 m or more in thickness, witli a few loose spicules between the 

 main fibres, scattered, Init with a tendency to l)e parallel to the surface, and 

 forming distinct horizontal bundles just below the surface. Spongin is present, but 

 not easy to detect. 



Spicules. IMegasclercs. Oxeas G7G x 2G m, slightly curved at centre, terminating 

 gradually in fine points. 



Microscleres. Isocheles 19-5 |U long, 6-5 m broad in lateral view, with slender 

 convex shaft without perceptible alae ; with broad saddle-shaped teeth with their distal 

 edges not far from each other, but not touching or overlapping. 



Four small specimens of this sponge were obtained. The largest is 4 cm. long 

 and 1 • 4 cm. in diameter. The surface has a finely hispid aspect in two specimens, but 

 is finely conulated in the third, and smooth and worn down in the fourth. The variety 

 differs from the type (15. p. 110) in having much larger megascleres, th(jse of the type 

 being only 350 x 18 /x. The isocheles of the type, again, are longer, viz., 28 m- After 

 a careful examination of the isocheles of the Kerguelen specimen in a good light and 

 with the spicules slowly rotating in balsam, I found that the median sub-dental 

 continuation of the shaft did not bend backwards from the tooth as in Desmaculon 

 {Ilumoeodlctya) palmata ; the appearance of bending back (sec 15. PI. xxiii., fig. 36) being 

 due to the lateral bend of the palmate tooth ; accordingly, even if Homoeodictya be a 

 true genus or sub-genus, D. kerguelenensis would not belong to it. 



Three specimens were dredged near Winter Quarters, from No. 12 Hole, 25-30 

 fms. ; off Hut Point, 100 fms. ; and off Coulnian Island, 100 fms. 



Desmacidon kerguelenensis mr. cactoides. 

 (Plate XIX., fig. 2, and Plate XXHL, fig. 2a-bM 



Sponge pyi-iform, laterally compressed ; surface prickly, with an Acanthella-like 

 aspect. Consistence rather firm. Colour in spirit, pale yellow. 



Skeleton formed of rather stout main fibres proceeding to the surface, where they 

 con.spicuously push up the dermal membrane, with scattered spicules in between the 

 main fibres. 



Spicules. Megascleres. Slightly curved strongyles, but sometimes with rather 

 blunt pointed ends (/.''. oxeas), 7G8 x 31 "25 /a. 



Microscleres. Isochelae palmatae 17 m long and G*25m broad with the wide 

 saddle-shaped palmate teeth almost in contact or even overlapping. 



The single specimen is attached to a worm tube by a contracted stalk-like base ; 

 the length is 5*5 cm., and the greatest breadth 3' 5 cm. It was a matter of doubt 

 whether to regard this form as a distinct spe(;ies ; but it will be obvious from the 



