28 



least 3700 at the end of July. Also stages of divisiou occui' /Vee, 

 quite identical to those of the corpuscles of the sponges (Table 10). 

 The diameter of the single corpuscles may sometiines be somewhat 

 smaller than that of those in the sponges, viz. 1.4 — 3/y. ; the oval 

 shaped body may be lengthened. One may find the same forms, 

 however, in old cultures of the isolated corpuscles of sponges 

 too. Finally I will mention the fact that, free in nature, the cor- 

 puscles often stick together in masses of 10 — 120 specimina. I will 

 return to this subject later on. 



4. Wheri the sponge dies ifs green corpuscles survive. 



5. It proved p)0ssihle to me to durahly transmute colourless 

 Spongillidae into the green form, hy infecting them with green 

 corpuscles isolated from a green sponge. Lateron I will treat the 

 method, in which this infection is brought about, more extensively 

 (Table 7). Now I will mention only, that for that purpose the 

 colourless sponges were placed for 3 — 72 hours into a diluted 

 suspension of isolated green corpuscles in water ; after that 

 they were transported, either colourless or already light-green, 

 into an aquarium filled with water from the conduit only and 

 placed into day-light. The green colouring then proved, in the 

 course of some weeks, not decreased at all ; on the contrary, it was 

 strongly increased ; many sponges had obtained rather a normally 

 green colour in a month's time (see Table 8 W 103, 207, 246, 

 248, 254, 257, 258, 326, 327), and their amoebocytes proved to 

 be normally laden with normal green chlorophyll corpuscles (pag. 

 16 — 17). At the same time such colourless sponges, but without 

 having been in a suspension of chlorophyll corpuscles, were — as 

 a contra-experiment — also exposed to daylight in water from 

 the conduit. These sponges got a faint green tint (colourless 

 sponges always grow green in daylight; I will treat this after- 

 wards), but in intensity their colour remained far behind that of 

 the infected specimina (Table 8 n° 102, 206, 259, 328); proof, 

 that the laftor really owed their normal green colour to the 

 infection with chlorophyll corpuscles, foliowed by the rapid mul- 

 tiplication of these corpuscles. 



From these experime^üs we mag deduce^ that the green chlorophyll 



