90 Asa Arthur Schaefjer 



In the above experiment there were eaten 12 Phacus and i grain 

 of sulphur, while 13 grains of sulphur and 3 Phacus were rejected. 

 It is evident therefore that in this case there is some sort of dis- 

 crimination between Phacus and sulphur. 



Experiment 2. Discrimination hetxveen Starch Grains and Phacus 



In another experiment designed for the same purpose but in 

 which iodine-stained potato starch grains and Phacus triqueter 

 were used, even more sharply defined results were obtained. The 

 technique was similar to that of the preceding experiment, except 

 that two pipettes were used, one for starch and the other for 

 Phacus. The starch was stained with iodine to facilitate obser- 

 vation. Previous to feeding, the starch was washed very thor- 

 oughly to remove all the superfluous iodine. The results are as 

 shown in Table II. 



This experiment shows conclusively that S ten tor can and does 

 discriminate between two kinds of particles differing as much as 

 Phacus and starch grains do from each other. The possibihty 

 of coincidence is entirely ruled out of court in that the stream of 

 particles was changed at least seven times from starch to Phacus 

 and from Phacus to starch. In the whole experiment there are 

 only four "mistakes" at the most, including the rejection of a 

 swarmspore twice and of a Coscinodiscus. But later experiments 

 will show that the rejection of the swarmspore and Coscinodiscus 

 were probably not mistakes, so that there occurred in this test 

 only one mistake, that of ingesting particle numbered 69 — a 

 starch grain. The size of the starch grains was variable, being 

 from one-eighth to four times the size of a Phacus specimen. 

 This shows that in this experiment, as in the preceding, size was 

 not the determining factor in the selection. Another point worth 

 noticing is that there are no loops in the paths of the particles 

 swallowed, and that there are very few particles rejected without 

 loops. The 45 loops of particle 32 are probably due to the fact that 

 the Stentorwas at that time lying on its side with pouch upper- 

 most, thus making the removal of the particle more difficult than 

 usual. When the animal turned over the particle was gotten rid of. 

 The usual number of loops does not exceed 10 in any one case. 



