462 COLLECTIONS FROM MELANESIA. 



frond, though only curved somewhat to one side in the plane of 

 expansion, was in life decumbent, so that the terminal branches 

 were then in contact with the sea-bottom. The species differs from 

 B. australiensis in the branching and anastomosing habit and the 

 flattened knife-edged branches, and in having the longitudinally 

 arranged spicules not confined to the axis, but extending to the 

 cortex. From B. syringella, Schmidt, it also differs in its growth 

 (though Schmidt mentions that the branches of 11. syringella some- 

 times unite) and in the absence of heads to the acuate spicules. 



AXINELLID^E. 



Axinellida, Carter, Ann. $ Mag. N. H. 1875, xvi. p. 133. 



This family differs from the Ectyonidoe in the much greater 

 importance of size of spicule as a factor of specific distinction. The 

 relations of the two families, however, require readjustment on 

 more satisfactory bases than at present. 



90. Axinella echidnaea. (Plate XLIII. fig. «.) 

 ? Spongia echidnaea, Lamarck, Ann. Mus. Hist. Nat. xx. p. 4-48. 



It seems likely that this will prove to be Lamarck's species. 

 That author refers (I. c.) to Seba (Thesaurus, iii.pl. xcix. fig. 7) in 

 illustration of his sponge. This figure has a strong resemblance to 

 the present species, but does not show the same tendency to lateral 

 junction between the branches, and has most of the latter somewhat 

 enlarged at the tips, whereas in these specimens they usually, though 

 not invariably, are either of about the same diameter throughout 

 or else taper to points. The dark reddish-brown colour of these 

 specimens and the peculiar echination of their surface by angular 

 wedge- or knife-shape processes about 2 to 4 millim. high, projecting 

 outwards and somewhat upwards, are decidedly indicated in the 

 figure. In texture the specimens are tough, elastic, harsh to the 

 touch in the dry state ; the surface-processes are flexible, almost 

 soft, in spirit. In structure it is a true Axinella, and thus does 

 not support Lamarck's surmise that it might be identical with 

 Sponr/ia muricata of Esper (Pallas, sp.), which is Tricentrium muri- 

 catum of Ehlers. The main skeleton exhibits the usual longitudi- 

 nally elongated meshes of loose spiculo-fibre, which in the stem is 

 composed in part of a transparent and almost colourless horny uniting 

 material, which seems to bo wanting in the surface-tufts ; distance 

 between longitudinal lines of axis -07 to - 1 millim. Surface covered 

 with a fuscous-brown subopaque pigment, which penetrates to a 

 slight distance below. Sarcode transparent, almost colourless, very 

 pale reddish brown. Spicules : — (1) Smooth, slightly curved acerate, 

 tapering gradually to sharp points, or more or less blunted si 

 one or both ends ; size -3 by -0095 to -44 by -0127 millini. : these 



