LEARNING IN PARAMECIUM 



69 



times do show, however, in spite of this circumstance, a marked 

 decrease. Again, the averages fall roughly into groups: 1-2, 3-7, 

 8-15. The group-averages stand: 84. ±4.7, 45.8 ±9.3, 33.7 ±4.4. 

 The irregularity in the mean variation is plainly due to wide 

 individual differences of activity. 



Thus, G, H, N, and X were either consistently or spasmodi- 

 cally sluggish. The slow and uncertain changes in their beha- 

 vior may be the direct result of their state, or they may be due 

 to the long blank intervals of " unlearning " between successive 

 reversals. Comparison with D, E, and (long initial times) 

 suggests the former interpretation.^ 



It was, now, evident to the observers that the apparent increase 

 of facility in turning might be due to some change in the medium 

 (such as increase in CO2 or decrease in oxygen) brought about 

 by the paramecia themselves. Until this possibility had been 



^ The tests were in all cases carried beyond the fifteenth full turn. The highest 

 number of turns in this set was 123 (subject E). After the fifteenth, however, 

 the subjects either continued to turn with little or no delay or else became inactive. 

 For a further slight increase in facility, see Tables III and IV below. 



