Selection of Food in Stentor Cceruleus (Ehr.) 8l 



While the method outlined above is the best possible for solving 

 many of the questions that are bound up v^ith that of choice of 

 food, there are nevertheless other points which cannot well be 

 cleared up in this way. For it is well known that S ten tors eat and 

 thrive upon such small organisms as bacteria, and it would be 

 impossible for several reasons to note what happens to each indi- 

 vidual bacterium as it is swept into the Stentor's pouch. For 

 getting at questions of this nature there was employed another 

 method which may be called an indirect method as compared with 

 the one described above. 



When it is desired to test the relative readiness with which very 

 small particles such as bacteria, Peranema, yeast cells, finely 

 ground carmine, etc., are taken up by the Stentor, it has been 

 found that the best method is to mix up quantities of them in the 

 desired proportion and then to introduce into this mixture some 

 normal S ten tors which have very little or no food in them. After 

 a stated time the Stentors are taken out and squeezed under a 

 cover glass in order to examine their contents. The most difficult 

 point is to maintain in the mixturue the original uniform distribu- 

 tion of the particles. This difficulty was mostly overcome by fre- 

 quent stirrings and by placing the dish in the dark to prevent reac- 

 tions to light, as will be described more fully later. By the use of 

 this method very important results have been reached which 

 could not have been obtained otherwise. Each of these two 

 methods acts as a check upon the other, and at the same time they 

 verify each other's results. 



THE NORMAL BEHAVIOR OF STENTOR 



To understand the experiments bearing on the choice of food 

 the reader should have a more or less clear idea of the normal 

 behavior of Stentor. This subject has been dealt with to some 

 extent by various observers, notably by Jennings, but in the course 

 of my investigations several points have been cleared up that w^ere 

 heretofore more or less imperfectly understood. Several new 

 features of behavior have also been discovered that play at cer- 

 tain stages essential roles in the choice of food, and it is necessary 



