528 REVISION OF THE AXINELLID^, ii., 



anastornosing. The mode of branching is dichotomous, and 

 successive dichotomies are visually in the some plane, but owing 

 to irregularities, partly resulting through mutual interference, 

 the branches come to be disposed in various planes : it is very 

 probable, however, that specimens occur in which the branching- 

 is confined entirely to the one plane. The division of the stem 

 to form the first two branches takes place 25 mm. above the 

 base, each of those again dichotomising at about the same dis- 

 tance above their origin, and each of the resultant four branches 

 also at about the same distance above theirs; the subsequent 

 divisions for the most part occur at increasingly longer intervals, 

 some of the terminal branches having an uninterrupted length 

 of 70 mm. 



The specimen is imperfect, consisting only of the dried 

 skeleton, — in which condition it appears to have been also when 

 first described. Nothing can be said, therefore, in regard to the 

 dermal features; but evidently the outer surface was even, with- 

 out conuli or elevations of any kind. Oscula are not indicated. 

 The skeletonised sponge being held between the eye and the 

 light, the skeleton is plainly perceived to consist, in each branch, 

 (i.) of a sharply circumscribed C3'lindrical core, of diameter 

 generally less than one-fourth and (except in the lowermost 

 parts of the sponge, up to about as far as the third dichotomy) 

 not greater than one-half the diameter of the branch, and (ii.) of 

 an outer region formed of slender radiating fibres, which are 

 inclined to the longitudinal direction of the branch at an angle 

 varying from 60° to nearly 90°, and present collectively an ap- 

 pearance somewhat resembling that of fur. The colour is a 

 faintly creamy-tinted pale grey or dirty white, its paleness being 

 due to the extremely small amount of spongin entering into the 

 composition of the radial fibres. In the original description, 

 the consistency is described as " tough, resilient, and compres- 

 sible," but this is not strictly correct : the axial region is fairly 

 tough and slightly compressible (and the branches consequently 

 are flexible), but the extra-axial layer is soft, and on compression 

 remains partially crushed. 



Details of skeletal str%icture (PI. xxxiv., fig. 2).— Except towards 



