67 



culaily around the stem and the branches, thin fascicles of the same sort of spicula occur in 

 the same manner; so that their basis frequently rests on several fascicles diverging towards 

 the main axis (see fig. 18) while the process otherwise (tig. 19) only contains one single thin 

 fascicle, which is continued centrally up to the very point. In the enlarged end of the bran- 

 ches these central spicula are very irregularly arranged; although they appear usuallv to 

 exhibit a radial arrangement; their points are seen protruding (fig. 21) on all sides through 

 the outer skin. 



On examining more closely these central spicula, we find them, wherever they form 

 the solid axis, of tolerably similar appearance and size (fig. 2.5), about O.GO Mm. long, perfectly 

 straight, thickest in the middle, and evenly tapering towards both extremities, of which one 

 is simply rounded (fig. 28) and the other shortly pointed (fig. 27); in the interior there may 

 be observed a distinct but extremely fine axial canal. Only in the extremities of the branches 

 there are found, among the normal spicula, some few considerably shorter (scarcely half as 

 long) which are not perceptibly enlarged in the middle, and the proximal extremity of which 

 often shews an irregular curve towards one of the sides, (fig. 211). 



The organic substance forms everywhere a tolerably thick stratum, which surrounds 

 the spicular axis (see fig. 18) and in which there are found imbedded some few needle-shaped 

 spicula apparently detached from the axis. Otherwise it makes a uniform parencliym, wherein 

 absolutely no canals or cavities can be discerned. Outwardly this parenchym is covered with 

 a more solid dermal layer, in which are found imbedded numerous spicula of a figure quite 

 different from that of the central spicula above noticed. Three different sorts of them are 

 to be distinguished, all belonging to the gi'oup named by Bowerbank .,retcntive spicula". The 

 largest of these (fig. 29 & 30) the length of which scarcelv exceeds 0.10 Mm. are simply 

 two-hooked (bihamate) and are rather scantily distributed round the stem and the branches 

 with their processes (see fig. 18), but more plentifully in the enlarged ends of the branches 

 (see fig. 22). The second form of superficial spicula (fig. .S3) which I have found only in the 

 enlarged ends of the branches, but here in rather considerable number (see fig. 22), are like 

 the former bihamate, but much thinner and smaller (only 0.04 Mm. long) and decidedly distin- 

 guished by the incurved hooks not lying in the same plane but inclining to different sides, 

 often so that they lie nearly opposite, giving to the spiculum the form of a more or less 

 distinct S. They thus belong to what Bowerbank calls contort bihamate spicula. The S'."" and 

 last form (fig. 20. 31. & 32) belongs to the inequi-anchorate spicula; and these are found 

 in quite extraordinary numbers everywhere in the outer skin, particularly at the ends of 

 the secondary processes, very closely packed together (see fig. 19). They are remarkably 

 small (only 0.02 Mm. long) so that their exact figure cannot be represented, unless by ai)pli- 

 cation of the highest magnifying power. They appear (see fig. 32) to consist of a somewhat 

 curved stem enlarged at both the bent ends, but much more at one end than at the other. 

 Both ends seem (fig. 32 a) to terminate in 3 sharp hook-Uke, somewhat flattened, acutely 



pointed teeth; but at the one end these teeth are so small that they can only be discerned 



9* 



