1 1 6 PL ANT- ANIMALS [CH. 



which were repeated many times, confirmed the 

 infection-hypothesis they were not sufficiently uni- 

 form to establish it absolutely. Time after time, the 

 minute larvse showed, when examined microscopic- 

 ally, a general absence both of green cells and colour- 

 less precursors of green cells ; but, time after time, 

 also, an occasional animal reared under these ap- 

 parently sterile conditions was found to contain 

 green cells. Though the occurrence of green cells 

 among animals hatched in filtered sea-water was 

 infrequent and sporadic, yet there such animals were 

 and, to make matters worse, the longer the larvae 

 were kept under observation, the larger was the 

 number of specimens which contained green cells. 



Evidently, either the infection-hypothesis was 

 wrong or there was some defect in the experiment. 

 A careful examination of the conditions of the ex- 

 periment revealed ultimately a source of error and 

 opened up the possibility of a new origin for the 

 green cells. The mucilaginous capsules, enclosing 

 groups of eggs, were discovered to be infested with 

 all sorts of minute organisms. On the capsules, and 

 in them, were many different kinds of microscopic 

 animals and plants diatoms, infusoria and forms of 

 life unlike any to be seen elsewhere. Since the egg- 

 capsule is formed by a secretion of the skin, and since 

 the skin of the animal is covered with slime, it was at 

 once clear that repeated washing in filtered sea-water 



