VISUAL PERCEPTION OF THE CHICK 83 



A significant change in this learning process occurred when 

 the chick, moving in the direction of the stimulus, turned 

 aside and approached the exit with that peculiar side-to-side- 

 head-bobbing-and-neck-stretching behavior that is character- 

 istic of a curious old hen. Pushing close to the exit, the chick 

 would squat lower and lower with repeated upward bobs of 

 the head and stretching of the neck as though in anticipation 

 of the upward movement of the exit door. As the door 

 opened, admitting to the relatively dark experiment box a 

 cheerful flood of light, the chick would commonly express 

 its satisfaction by means of the familiar hovering twitter. 

 Where food was a rewarding motive, the hovering twitter was 

 often replaced by a clearly distinguishable feeding twitter, 

 and frequently the expressed satisfaction was a combination 

 of the two types of vocalization. 



From the detailed records of behavior in each individual 

 test,3 table 15 has been computed as a measurement of the 

 learning process in relation to the electric compartments. The 

 table presents quantitatively the changes in the behavior of the 

 chicks with reference to the + responses, the outside corners, 

 the stimuli, and the inside corners. The first column merely 

 represents the series of ten tests to which any of the quanti- 

 tative measurements belong. The last column indicates the 

 date on which the series was made. The table further presents 

 for comparison the records of chicks 24, 25, and 27. The four 

 columns arranged under each individual chick are arranged to 

 show in any series (1) how many correct ( + ) choices were 

 made, (2) how many times the outside corners were approached, 

 (3) how many times the stimuli were tried as a place of possible 

 exit, and (4) how many times the inside corners were similarly 

 tried. For my method of collecting these data, the reader is 

 referred to Chapter IV, pages 37 and 38. 



1. If the chick profits by preceding experiences in this maze 

 task, tendencies toward certain definite numerical limits should 

 appear in each of the four columns presented in table 15. The 

 number of correct choices, if there be no preference for either 

 stimulus, should start at approximately five. Advance in the 

 series should ultimately bring an increase in this number of 

 correct choices which approaches ten as a limit. 



^ For method of recording this behavior see page 36. 



