16 



ing numerous brown particles, 2. Generally not the whole (green 

 or colourless) sponge had such a dirty tint — at least not Spon- 

 gilla. As one knows, this sponge grows on stones or wood as a 

 thin crust, from which long (10 — 20 c.M.) finger-shaped branches 

 proceed ; contrary to Ephydatia, which generally forms flat 

 cushions with but little elevations. Of course, the growth of Spon- 

 gilla chiefly takes place at the end of the branches, and is rather 

 quick. But then it is obvious that the tissue of such a region, 

 where the cells are rapidly dividing, can not possibly have the 

 same degree of dirtiness (by particles from the water) as an older 

 not growing region. And besides, in young tissue there are no 

 flagellated chambers ; so there the possibility of taking particles 

 from the water is, of course, considerably reduced. This proved 

 to be the case. In sunimer, whon Spongilla is growing quickly, 

 it generally shows brightly (not dirty) coloured tops, '/a — 'A ^•^• 

 long; which, however, in specimina from cleaner water contrasted 

 much less distinctly with the older tissue. 3. When such a sponge 

 (with bright coloured tops on dirty-brownish branches) was put 

 into an aquarium filled with water from the conduit, all the 

 diff'erence in colour between the tops and the older tissue had 

 disappeared within a few days; also this last tissue had got a 

 bright colour, whether it was green, creamy-white, or an inter- 

 mediate. ïliis tissue, originally loaded with brown particles, then 

 proved under immersion to contain almost none. 4. Something 

 the like can be said of Ephydatia. 



So the tivo cJdef forms of Spongilla lacusfris and Ephydatia 

 fluviatilis are a grass-green (Fig. 1) and a colourless (creamy-white) 

 one (Fig. 2). 



I am going to treat the green form first. 



The sponge owes its green colour to numerous little green cor- 

 puscles preserit in its tissues, especicdly in amoeboid cells (Eig. 69), 

 which for shortness' sake I shall indicate with the common name 

 of amoebocytes. (As for the different cell-forms in the fresh-water 

 sponges, see Weltner (68) 1907). These amoebocytes crowded with 

 fJic f/rmtt corpuscles form the greater mqjority of the sponge cells, 



