I02 Asa Arthur Schaeffer 



show that 56 per cent of the Euglense were eaten and only 10 per 

 cent of the Trachelomonas. But a better and truer idea of the 

 accuracy of selection is obtained by a study of the results as set 

 forth in the table. As in the sixth and the seventh experiments, 

 the beginning of this experiment differs from the latter part in 

 thatthe variety of ingested particles is greater in the beginning, and 

 also in that the proportion of food ingested as compared with 

 the amount fed, is greater in the beginning. There seems also to 

 be a marked change in the basis upon which discrimination is 

 effected; it evidently became more restrictive towards the end. 



What is the cause of the change in the degree of restrictiveness in 

 selection in these experiments? Is the change due to differences 

 in the food or to an alteration in the S ten tor itself.? 



Evidently different food particles possess different strengths or 

 powers of giving the stimulus that causes Stentor to ingest them. 

 The stimulus from Phacus triqueter is stronger than that from 

 Trachelomonas hispida; and that from Euglena viridis is still 

 stronger than that from Phacus triqueter. This is shown by the 

 fact that at certain times (when nearly replete), Stentor takes 

 only Euglenae (or Phacus if Euglenae are not present), from a 

 mixed stream of organisms. There probably are slight differ- 

 ences between the strengths of the stimuli from different individ- 

 uals of the same species, but these differences are evidently less 

 than those between different species, for when nearly satiated, 

 Stentor takes onlv the particles of one species. 



Yet we find in the experiments described above that specimens 

 of the same organism (as Euglena) are at first eaten without any 

 exception, while later less than half of them are eaten. The 

 change in the proportions eaten as the experiment progresses are 

 well seen in the following tabulation of the results of the eighth 

 experiment. The experiment is divided into successive groups 

 of about ten particles to each group. 



