Watson, Color Fision in Monkeys. 17 



small differences in tlie visual characteristics of the boxes did not 

 serve as secondary criteria to which the animal might have reacted. 



In presenting the lights, care was taken in the successive tests to 

 vary the height to which the screen was raised. In order to keep 

 this factor variable, I purposely left off a stop which would allow 

 the screen to go a certain height and no higher. The screen could 

 be raised from 0"-7", the full height of the band. The animals 

 would, e. g., on one test have to react with the two bands only I/2" 

 high, while on the next the bands would be 7" high. This procedure 

 tended to eliminate the possibility of their reacting to slight dif- 

 ferences in form. A few of the tests where the vertical and hori- 

 zontal forms of the two bands were changed independently are in- 

 cluded in the report. 



In the early part of the work, only six trials per day were given. 

 Gradually, as the animal became more accustomed to his work, 

 fifteen to twenty were given. The animals were always eager for 

 the grapes, and so long as they were unafraid, the tests could be 

 repeated again and again until the experimenter became fatigued. 



The positive color was presented very irregularly as regards left 

 and right positions. In the early stages of the association, the posi- 

 tion of the two bands was interchanged regularly. After the first 

 three or four days, however, irregularity was introduced. The 

 positive color would be presented first on the right on a given day, 

 while on the succeeding day it would be presented first on the left. 

 In order to give an idea of the variations introduced in the order 

 of the presentation of the stimulus, I offer the following notes 

 taken from J's reactions to red-green: 



April 23. 



Red on left to begin : 



Red and green alternated for the first three trials. 

 Red given three times on right. 

 Red given three times on left. 



April 24. 



Red on right to begin : 



Red and o-i-een alternated for three trials. 



