20 



Ephydatia and always with the same result. But now the contra- 

 experiment (Table 2) ! When I proceeded with a colourless sponge 

 in the same way as described for a green one, I found that, 

 after exposure to Hght, the movements were not resumed by the 

 bacteria. Accordingly, the preparation of the colourless sponge 

 proved to contain n o green corpiiscles at all , neither isolated 

 nor in amoebocytes. I repeated this experiment also several 

 times. 



I think to have given in these experiments the decisive proof, 

 that the green corpiiscles of the SpongilUdae produce 0^ in light, 

 hut not so in darkness. 



2. Now we have to show the photosynthesis (production of carbo- 

 hydrate or oil) of the green corpuscles in light '). In the first 

 place I should mention that in fact most corpuscles show one — 

 sometimes more — drops of oil, but never any carbohydrate 

 (pag. 25). To prove their production, I proceeded as follows (Table 3): 

 Preparations were made from a light-green Spongilla (grown in 

 twilight and afterwards kept for some time in darkness) in a 

 darkroom by candle-light. One half of them was immediately put 

 into complete darkness, while in the other ones the percent of the 

 isolated green corpuscles containing an oildrop was examined 

 (Table 3 c). This proved to be 42° j^. These preparations were 

 then exposed to tempered day-light. Now we see in the table 

 that in two days the percent rosé from 42 to 79, but that 

 in the preparations in darkness it remained 35. Then being ex- 

 posed to day-light for two days, the latter too showed a rising, 

 namely from 35°/o to 78°/^; while at the end a percent of 

 + 90 was reached in all preparations. I could also observe that 

 the oildrops of the corpuscles, which had been exposed to the 

 light for a longer time (so in culture n". 396 a — d) were larger. 

 So it is clear^ that the green corpuscles of the SpongilUdae 

 photosynthesise oil in light^ hut not so in darkness. I have re- 



1) It sccms Ihat Brandt (1- c.) has observed this for Spongilla; he at least men- 

 lions that the isolated corpuscles remaiued alive for weeks, „according to their pro- 

 ducing assimilates". But he doesn't give more about it; and where he should have 

 treated the phenomenon more rxtensivcly, he does not mention it at all. 



