118 L. W. COLE 



white box, since, in the preference tests, the dancers selected 

 the black one in more than one-half the trials/ 



As described above (p. 1 12), the preliminary series were followed 

 by tw^o series of ten trials each, called "preference series," and 

 designated in Table II by the letters A and B. On the day follow- 

 ing the completion of the "preference series," the training series 

 was begun and they were continued until the chick had made 

 twenty consecutive choices of the darker screen. The order 

 of change of illumination of the two screens appears in table II. 

 The letter 1 indicates that the screen at the left was the darker 

 one, the letter r, that the one at the right was the darker. 



Since the preference series were preceded by the twent}^ pre- 

 liminary trials, in which the chick escaped from the experiment 

 box by going alternately through the right and left passage- 

 ways, the preference, so-called, was interfered with by the par- 

 tially formed habit. Untrained chicks chose the brighter screen 

 uniformly. 



During the training series, if a chick chose the lighter passage- 

 way, it received an electric shock, whereupon it usually retreated 

 from the wires, the door of the darker passageway was opened 

 and through that it escaped to the hover box. Under this stimulus 

 the chicks quickly learned to choose the darker screen under 

 conditions of easy and medium discrimination. A few chicks 

 were unable, even after many trials, to learn to choose the darker 

 screen under the difficult condition of discrimination. 



Results of the Experiments. The results of the experiments 

 appear in table III. This table gives the three conditions of dis- 

 crimination, easy, medium, and difficult, the relative strengths 

 of the stimuli, the numbers by which the individual chicks were 

 designated, and, opposite each of these, the number of trials 

 which preceded twenty consecutive connect choices, or the number 

 of trials "up to the point at which errors ceased." 



In order to spare the reader an annoying repetition of the 

 phrases, "easy, medium, and difficult conditions of discrimin- 

 ation," I shall sometimes refer to them, respectively, as great, 

 medium, and slight differences of illumination or brightness of 

 the two glass screens. 



It is evident from table III that under the condition of easy 

 discrimination the rate of learning is more rapid the stronger 



" Yerkes and Dodson, loc. cit., p. 462. 



