•_>S1 FREDERICK S. BREED 



stand near the edge of the electric wires and crane its neck toward 

 the exit. It is easy to see how a slight noise on the proper side 

 or a slight movement of the proper shutter might have precip- 

 itated a correct reaction when no discrimination of the im- 

 portant stimuli had taken place. The method was open to 

 improvement at this point. So far as possible, chances for the 

 involuntary interference of the experimenter should be elim- 

 inated. The writer was at work on an automatic device to 

 meet this difficulty at the time these experiments were brought 

 to a close. 



The tests were given in series of ten trials each. In each 

 series there was an equal distribution of the two stimuli on the 

 right and left sides of the apparatus. These changes of position 

 were handily produced by shifting the sliding frame, which 

 contained like mounts at either end and the one to which these 

 were opposed in the central opening. It is readily seen how a 

 shift of the frame altered the relative positions of the two stimuli 

 as well as changed the source lights for each. 



The order of procedure was preference tests, training tests, 

 and control tests. The first consisted of one or more series in 

 which the chosen stimuli were alternated right and left and no 

 electric shock was used. In the color work certain "preliminary" 

 tests were made which w T ere really of the nature of preference 

 tests on various combinations of stimuli. In the training tests 

 the right and left shifts were equally balanced but varied in 

 order. In the control tests the shock was dispensed with. 



REACTIONS TO COLORS 



In the previous color tests, where the reflection method was 

 used and black and blue were opposing colors, chicks not only 

 formed the habit of rejecting blue but continued this rejecting 

 reaction both when white was substituted for black and when a 

 different tint of blue was substituted for that used in the original 

 training. It seemed very probable that difference in color was 

 the basis of discrimination. The experiment reported below was 

 an attempt to test the color discrimination of the chicks by the 

 transmission method. The colors chosen for this experiment 

 were red and blue. During the course of the preliminary tests 

 the following setting of the screens was adopted for the preference 

 tests and the later training : Two red screens, one at either end 



