52 



systems. Such communicating systems of canals often run parallel to the surface in the neigh- 

 bourhood of the latter as we saw in 1945 (PI. VI, fig. 1). This seems especially to be the case 

 in the incurrent system. Therefore I believe the network of canals shown in fig. 1 on PI. XIV 

 to be incurrent. As stated before, we find in such specimens of trop. tuberosa central canals, 

 surrounded by transparent tissue, which canals I believe to be equivalent to the incurrent 

 reservoirs of 1365 etc. Here, however, the canals are widely open, whereas in g2a-l (trop. 

 tuberosa) they are constricted. 



4. The soft parenchyma. 



In a former paper (1908 a, p. 40) I drew attention to a sort of lymphoid or adenoid 

 connective tissue, composed of a reticulum of flat cells with membraneous processes. I stated 

 this tissue in Poterion and Cliona to surround the main canals and suggested that it probably 

 is widely spread in Sponges. Sections through specimens of tropus tuberosa (g2a-l) exhibit 

 exactly the same transparent tissue (PI. IV, fig. 3). Careful observation of sections of various 

 thickness, stained in various ways and taken from different specimens, have tought me, that 

 almost the whole parenchyma is of this kind. It has, however, a very different aspect in different 

 places and different specimens. In fig. 6 on PI. VI I have marked one of the canals with an x. 

 In fig. 7 this place is magnified a hundred times; one clearly sees now that the parenchyma 

 between larger and smaller canals is composed of a reticulum, the meshes of which have different 

 sizes. Just as we find gradual transitions from the largest canals to the smallest visible with 

 low power, so we find that there are likewise transitions in lumen between such small canals 

 and the meshes. If we now look with higher power at a certain canal or large mesh, e. g. 

 the place marked y in fig. 7, we find again smaller meshes and we clearly see that they are 

 formed by membraneous expansions of cells (fig. 8). Such cells generally possess ellipsoid 

 nuclei and ramifying processes which are not threads but membranes. Between these cells 

 all sorts of other cells may occur and I believe also connective tissue fibres. The size of 

 the membrane cells is variable ; hence the aspect of the tissue will be different. In fact, the 

 difference depends on the size of the cells and accordingly on the size of the meshes, 

 but also on whether many or few other cells occur. It will also make a great difference 

 whether there are mastichorions in the neighbourhood or not ancl finally it will depend on the 

 quantity of spicules. However, as far as I can see, the structure of the whole parenchyma is 

 fundamentally the same. 



Now we have already mentioned, that various contractile elements occur. Obviously the 

 aspect of the lymphoid tissue will also depend on the state of contraction of such elements, 

 say the muscle cells. This is plainly seen in the tissue surrounding larger canals, especially in 

 those, which I called reservoirs. In specimens like 948 b or 1365 the lumen of the central canal 

 is wide ; the tissue surrounding it so far as it is practically without spicules forms but a narrow 

 band, if seen in section (PI. VII, fig. 1); the concentrically placed muscle cells are fully expanded, 

 and the tissue itself is consequently contracted. The meshes between the flat membraneous cells 

 show their long axes to lie concentrically (PI. VII, fig. 4). We find quite another aspect if we 



