48 



HAROLD C. BINGHAM 



these interruptions, which are certainly not indicative of inabil- 

 ity to perceive the visual detail in question, wholly perfect 

 responses when the threshold is being approached cannot rea- 

 sonably be expected. 



By considering, in addition to the correct records, the pecu- 

 liarities of behavior which cannot be presented in tabular form,^ 

 I am convinced that the largest variable which the chick can 

 distinguish, under the conditions of this study, when presented 

 with a standard o 28 + , lies somewhat above o 15 +. The 

 quantitative measurements on the basis of correct choices sup- 

 ports this conclusion. Every subject, excepting 7 and 20, suc- 

 ceeded in making at least 80 per cent of the choices correct in a 

 series when the ©15+ variable was used. Even in the re- 



TABLE 4 

 Average time for choosing 



spouses of 7 and 20, an efficiency of 70 per cent was indicated. 

 But more convincing, perhaps, is the fact that two chicks, 3 

 and 17, reached 90 per cent and two, 16 and 21, made perfect 

 series. 



The time used by the chicks in choosing is a factor which 

 might throw supplementary light on this threshold question. 

 Having kept the time for every choice, and having casually 

 observed changes in time for choosing in the beginning, inter- 

 mediate, and final stages of discrimination, it is at least inter- 

 esting to note whether the actual time records confirm this ob- 

 servation. In table 4 appears the average time in seconds 

 consumed by each of eight subjects for choosing under the 

 three different discrimination settings. An average of the 



^ See sample test sheet, page 36, for method of recording these peculiarities of 

 behavior. 



