21 



poatod this experiment several times (Table 3 a. b.j; always witli 

 the same result. 



It appearod to me, that the oildrops are formed much more 

 qiückly, when the green corpuscles are exposed to light after 

 having been kept — as a preparation — ■ for some weeks in 

 darkness: in that case the percent rosé in half an hour from 

 38 to 70. This short exposure to light seems to have extended 

 lts influence also to later on ; at least in darkness the percent 

 rosé up to 88, This is not inconceivablo. The oildrops, once 

 having been produced , seem to be used but very slowly by the 

 corpuscles — even in darkness — (Table 3 a, n". 163 ; 3 b n*^. 210). 

 These two cases are by no means the only ones ; I have sometimes 

 been able to state that most green corpuscles of a culture, after 

 a period of 6 months in darkness, still contained a drop of oil. 



Finally I have been able to directly observe the forming of 

 an oildrop in a green corpuscle, which at first contained none 

 at all. The colour grain had not been kept in darkness; so the 

 forming did not proceed so quickly: the first appearance was 

 made after 3 hours of exposure to light, a real oildrop only to 

 be seen after 21 hours. 



So we have proved the identity of the green colour mg -matter of 

 the Sponglllidae to the chlorophyll of the plants hg physiological 

 argnments. We may now speak of the chlorophyll, the chloro- 

 phyll corpuscles of these sponges. 



III. TlIE NATURE AND STRUCTURE OF THE GREEN CHLOROPHYLL 



CORPUSCLES. 



As I stated in the Introduction, the results of Brandt (8, 9) 

 concerning the nature of the chlorophyll corpuscles of the fresh- 

 water sponges are quite different from those obtained by Kay 

 Lankester (35) and Geddes (21, 22). Brandt concluded that 

 these corpuscles were algae, Lankester and Gteddes however, 

 that they consisted of chlorophyll formed by the sponge itself 

 as an inherent part of its cells. 



