46 HAROLD C. BINGHAM 



which emphasize habit formation my results do not always ap- 

 pear clear cut and definite. Frequently the positive choices 

 that have been allowed to pass as though satisfactory do not 

 exceed 75 per cent of the total, whereas earlier work seems to 

 have been more regularly favored with ultimate perfection. 

 Does this not indicate that my results are untrustworthy? 



In general, there are two answers to this possible criticism. 

 I shall refer first to the least important, namely, that my task 

 was not one in habit formation. Perfection of a habit was only 

 incidental to the primary task of sensory acuity. Experience 

 with the first group of chicks indicated that time was wasted 

 by holding rigidly to the perfection requirement. The uncer- 

 tainty of the chick's health, further, made it advisable to hasten 

 towards the threshold of discrimination without waiting in all 

 cases for perfect habits. 



Chicks 6 and 7, representing the first group, were held to the 

 preliminary training, o 28 H g 7 +, until two successive per- 

 fect series were obtained. Then the variable was changed to 

 o 12 +. From the first, there was evidence of discrimination 

 of this new variable but only a single perfect series appeared. 

 After six or seven series, there was no evidence of improvement 

 over the beginning of the ©12+ discriminations. This sug- 

 gested that time could have been saved by increasing the variable 

 as soon as there was satisfactory evidence of discrimination, 

 even though positive reactions had not reached 100 per cent. 



With the second group, including chicks 15-18, therefore, 

 this suggestion was tried out. The series numbered 1 in table 

 3, which was the 14th for this group of chicks, represents the 

 final series of the preliminary training. In that series chick 17 

 is the only one to react perfectly. But the reactions of 15, 16, 

 and 18 were so closely approaching perfection that there was 

 unmistakable evidence of discrimination. Moreover, it was 

 more convenient to present to the other chicks the same stimuli 

 which were presented to chick 17. Without further training, 

 then, all chicks of the group were confronted with the new 

 variable, o 12 -f. 



That such haste was justified is indicated in the single o 28 + 

 — o 12 + series for each individual of this second group. 

 Chicks 15, 17, and 18 responded perfectly while the positive 

 responses of No. 16 are 8 out of 10. Again, when the variable 



