bY E F. HALLMANi>r. 533 



ing spicules. The branches are tiexibleand tough, with an outer 

 layer of softer consistence; this layer has disappeared from the 

 stalk, which is dense and tough tliroughout, and has a smooth 

 and even surface. The colour in spirits is greyish-yellow. 



Skdet07i. — The formation of the skeleton differs from that of 

 .S". dendioides, described above, mainly in two respects; and 

 these differences are to some extent consequent upon the much 

 greater length (up to 1-5 mm.) of the megascleres in the present 

 species, and upon the relative narrowness of the external layer 

 intervenino; between the central axis and the surface. In the 

 central axis, there ai'e not to be distinguished, as in S. dendroides, 

 definite longitudinal fibres joined by transverse ones in more or 

 less ladder-like fashion, but the megascleres are disposed rather 

 loosely in ill-defined tracts which cross one another at acute 

 ano-les, thus tjivino- rise to a somewhat lattice-like arrangement. 

 And, secondly, the extra-axial skeleton(Pl. xxxv., fig. 2) is entirely 

 without transverse fibres, and consists simpl}' of numerous, short, 

 pauciserial lines of spicules running outwards to the surface in a 

 direction nearly perpendicular thereto, — these spicules being 

 arranged more or less penicillately and united by, at most, an 

 infinitesimal amount of spongin, and the terminal ones project- 

 injf far bevond the surface. 



The axial skeleton clianges considerably in character with age, 

 owing to gradual increase in the amount of spongin developed in 

 connection tlierewith, and presents a very different appearance 

 in the older and more basal parts fi'om that which it exhibits in 

 the. uppermost portions of the branches. In the latter region, 

 for a considerable distance (several centimetres at least) from 

 the extremities of the branches, the amount of spongin present 

 is so slight that its existence is apparent only in sections from 

 which the fleshy tissue has been removed by maceration(Pl.xxxv., 

 tig 1). In this portion of the skeleton also, the spongin appears 

 difiused, and is without definite outlines. Proceeding towards 

 the base of the sponge, the spongin gradually becomes more and 

 more concentrated upon the sides of the lattice-like meshwork 

 formed by the megascleres, which is thus converted into a reticu- 

 lation of spiculo-spongin fibre. The elongated, narrow meshes of 



