314 THE TANGANYIKA PROBLEM. 



The spicular systems of the third class are seldom found in any other position than 

 in the fibres. 



As a rule, the spicules are arranged in the fibres with their axes parallel to one 

 another, and in the deeper parts of the sponge the connecting spicules are rather 

 numerous, and more strongly developed than in the more superficial parts. The 

 connecting spicules are usually the most strongly developed spicules in the whole 

 sponge as regards size, differing, however, only in thickness from the smooth, curved 

 amphitornota which constitute the fibres (Fig. 2 — a. ). Speaking generally, the largest 

 spicules of the first class, together with a few of the second and all the third, form 

 the fibres and the connecting links between them, while the smaller spicules belong- 

 ing to the first, and nearly all those belonging to the second class, are scattered 

 about irregularly in the meshes between the fibres. The smallest spicules of all 

 seem to be absolutely independent of the skeletal meshwork, and this is the strongest 

 argument that can be adduced in favour of the view that they are microscleres and 

 not young megascleres. 



The arrangement of the skeleton at large is reticulate. The most general feature 

 of the general conformation of the fibres is the way they pass from the surface of 

 fixation of the sponge to the dermal membrane which they support. Along their 

 course from one surface to the other they present a wavy appearance, often dividing 

 and again reuniting, approaching the dermal membrane nearly always at right 

 angles, and in many cases expanding into a brush-like structure which supports that 

 membrane (Fig. 2 — c. ). In some of the largest lobes of the sponge the fibres nearest 

 the centre pursue a straight course, while those furthest from that position curve 

 outward, so as to form supports to the dermal membrane which covers the flanks 

 of these mound-like elevations. Owing to this arrangement a longitudinal radial 

 section of one of these lobes presents an almost fan-like appearance, as regards the 

 skeletal fibres. 



(C) The Spongin. — All the skeletal fibres of this sponge are enclosed in a distinct 

 sheath of spongin, which is greatly thickened at the points where the connecting 

 spicules occur, these being either partially or completely surrounded by it (Fig. 2 — b. ). 

 Not only are the fibres and the connecting spicules enclosed in a sheath of spongin, 

 but the surface of the sponge is covered by a thin layer or cuticle of the same sub- 

 stance, which dips down between the cells of the dermal membrane and communi- 

 cates with that which envelopes the fibres (Fig. 2 — a.). 



(3) The Canal System. — Owing to the fact that the material which had been 

 preserved for histological study of the sponge had been shaken considerably in 

 moving from place to place, a great number of cells had, apparently, become loose, 

 and were found lying in the spaces of the canal system. In consequence it was 

 impossible to make a complete and thorough study of that system, though individual 

 cells were in many places nicely preserved ; nor is Spongilla, for other reasons, a 

 favourable object for the study of the canals in the Monaxonida. 



The canal system in Spongilla moorei belongs to the type usually described as the 

 third. The dermal pores, which are situated on the flanks of the mound-like 

 elevations of the surface and in the intermediate depressions, are small, and open 

 into the subdermal cavity, which is lined by flattened epithelium, and considerably 

 reduced by the passing through of the skeletal fibres, which are enclosed in a sheath 

 of spongin, which is covered by cells of the epithelial layer. 



