18 



E. ARNESEN. 



[REP. OF THE "MICHAEL SARS" NORTH 



often verging upon styli, 0.444 mm. long and 0.006-0.011 

 mm. tick. As I have had Topsent's preparations for 

 comparison I am sure of the identity. 



Geographical distribution: This species has been 

 recorded from Exmouth, England (Bowerbank, op. eft), 

 off Roscoff (Topsent, op. cit.) and from the Mediterranean 

 Coast of France (38, p. 123). 



The specimens from the "Michael Sars" were taken 

 between Gran Canaria and Cape Bojador (lat. 26° 6' N, 

 long. 14° 33' W), depth 39 metres; deposit shingel; 

 bottom temperatur 15.63° C. 



Stylotella topsent! sp. n. 



PI. Ill fig. 4 & pi. V, fig. 4. 



St. 37. One specimen. 



Sponge erect, ramouse, with a few long flexible 

 branches dichotomising in whiplike ends (fig. 4). The 

 transverse section of the branches is oblong or triangular. 

 Oscula, about 2 mm. in diameter, are serially arranged 

 along the narrow sides of the branches. Dermal mem- 

 brane distinct and pellucid (only patches preserved). 

 Consistency rather tough. Colour i spirit light brownish. 



Skeleton consists of a somewhat irregular network 

 of densely packed parallel styli imbedded in strong spongin 

 fibres. There is no pronounced difference between the 

 fibres, but those radiating towards the surface are somewhat 

 stronger than the other ones. The meshes of the network 

 and the dermal membrane are filled with dense masses 

 of scattered spicules. 



Spiculation. The spicula of the fibres are styli 

 varying in length from 0.230 to 0.296 mm. by 0.195 mm. 

 (in the thickest part). They are either straight, cylindrical, 

 with long and sharply pointed ends or, usually, somewhat 

 fusiform and slightly bent; the other end is simply rounded 

 off. The scattered spicules in the meshes and in the 

 dermal membrane consists partly of styli of the same kind 

 as in the fibres, but there are also some very slender, 

 irregularly curved ones, the form of which is difficult to 

 define, but usually they have blunt ends. 



The above described form does not agree in any 

 respect with any one of the Atlantic species described by 

 Bowerbank and Topsent (St. Jullieni 38, p. 137), 

 but in form and proportions of spicules it is like one of 

 the four Australian species of Lendenfeldt — (St. digitata 

 Lndf. 14, p. 185) — which has a digitate branching form 

 and straight or slightly curved spicules, 0.250 in length 

 and 0.004 mm. in thickness. 



Locality. Off Cape Bojador; depth 39 metres; 

 shingel; bottom temperature 15.63° C. 



Cladorhiza gelida Lundbeck. (?) 

 Vide litter: 17, p. 83, 42, p. 5. 



St. 102. Several fragments. 



At station 102 a few, mostly denuded fragments of 

 Cladorhiza were obtained, which externally exhibits most 

 resemblance to Lundbeck's figure of CI. gelida Lndb. 

 (PI. Ill fig. 1, op. cit.), but shows certain peculiarities. 



Thus besides the filiform appendages tapering towards 

 the point in the characteristic way of CI. gelida, there are 

 also filiform appendages, ending in a globular or club- 

 like swelling (5 cm. long). Further the spicules are about 

 double the size of those indicated by Lundbeck and 

 Topsent. Several fragments have been examined from 

 different parts of the sponge, with the same result, though 

 in the branchlets a few spicules approximating in size 

 to those in the type have also been observed. 



All the types of micro-spicules indicated by Topsent 

 (PI. II fig. 4, 17) have been found except the small sig- 

 mates. Thus, the anisancorce unguiferae were present in 

 great abundance, generally varying from 0.051—0.055 mm. 

 (thickness 0.007 mm.); in the branchlets a few measured 

 only 0.037. Large sigtnata, 0.159 — 0.259 mm. (greatest 

 thicknes 0.259 mm.) were likewise abundant in all parts, 

 while only a few ancistres, of about the same size, have 

 been observed. Sigmansistres, 0.051 — 0.092 mm., were 

 numerous especially in the branchlets, where the large 

 sigmata were less frequent. 



This difference in the size of the spicules in my speci- 

 mens, as compared with those in CI. gelida Lndb., has 

 made me hesitate to refer them to that species. But 

 perhaps CI. gelida is like CI. longipinna R. and D. (21, 

 p. 92) in having spicules of different sizes (anisancorae 

 varying from 0.034 — 0.060 mm.) in different parts of the 

 sponge — the larger ones belonging to the lower surface. 

 At least two of my specimens are fragments broken off 

 very near to the base of the sponge, all the rest being 

 only detached branches — so that it may be so. I there- 

 fore think it right to refer them provisionally to CI. gelida 

 Lndbk. 



Geographical distribution: Cladorhiza gelida 

 Lundbeck is known as an inhabitant of the cold area 

 having been recorded from: — 



