VISUAL PERCEPTION OF THE CHICK 47 



was increased to o 15 + , chick 16, the only individual of the 

 group which had failed to make a perfect record in one of the 

 two preceding tests, is the only one in this series to make a per- 

 fect record. 



Comparing, now, these records of chicks 15-18 with the rec- 

 ords of chicks 6 and 7 of the first group, the haste is further 

 justified. Chicks 6 and 7, after continuing in the easy training 

 until the reactions were well perfected, subsequently fail to 

 show up superior to chicks 15-18. Furthermore, the records of 

 chicks 20 and 21, belonging to group 3, are no less clear cut for 

 the variables © 12 + and o 15 + than the records of chicks 6 

 and 7. 



The second answer to this possible criticism is more theoretical 

 than the first, but the two are closely allied. A review of the 

 actual results indicates that it was not essential to insist upon 

 100 per cent correctness. In the second place, mine is a study 

 of detail vision in the strict sense; earlier studies, though pre- 

 suming to present data on visual details, have probably dealt 

 with visual complexes. For this reason, my results cannot be 

 justly estimated by means of the same criteria that have been 

 adopted in connection with earlier work. In studies of detail 

 vision, animals are compelled to react appropriately to a single 

 visual factor. Now, when the remaining factors happen to be 

 so grouped that the complex becomes especially attractive, — ■ 

 and well controlled tests will contain such combinations, — is it 

 surprising that the attractive complex gains the attention of the 

 animal for the time being and causes the animal to ignore the 

 more insignificant single visual detail? In control tests, one is 

 always demanding, in the face of other attractions, responses 

 to this more or less insignificant factor. Due allowance should 

 be made for momentary lapses of the attention of the animal. 



Furthermore, as the threshold of the particular visual factor 

 is approached, the single detail becomes less and less significant. 

 It is reasonable to suspect, then, that even before the percepti- 

 bility of the discriminate detail wholly vanishes, the animal 

 will be tempted to try out novel reactions. On the analogy of 

 human behavior, it seems reasonable that the animal actually 

 ignores at times the discrimination problem that has been set. 

 This supposition is especially plausible when we consider that 

 the vividness of punishment may decline as the discrimination 

 habit approaches perfection. Making allowance, then, for 



