VISUAL PERCEPTION OF THE CHICK 89 



or difficultness of discrimination, but upon the basis of an 

 optimum dependent upon the sensitiveness of the animal. 

 Experimental data can determine this optimal punishment 

 within the limits of the distribution of sensitiveness in any 

 type of animal. One should expect this optimum to vary not 

 only with the different types of animals, but also with different 

 individuals and at various stages of individual development 

 and progress. 



In general, the chick tends to acquire a fairly definite succes- 

 sion of acts in arriving at the appropriate conclusion of the 

 series. In this series ther^ may be a choice which involves 

 entrance, for instance, into the left compartment. If the left 

 compartment happens to be +, the trial at the left exit may 

 culminate the series. If the left compartment, on the contrary, 

 is — , the series of acts must be continued until the right exit 

 is tried. A certain definite "set" in serial acts tends to occur, 

 making certain errors a natural consequence. 



In terms of neural organization, an integration occurs of 

 which a portion must be disintegrated and another portion 

 must be retained and further stamped in. That portion of the 

 "set" w^hich corresponds to the error must be eliminated and. 

 If possible, without destroying the appropriate portion of the 

 "set." Thus, the way to eliminate the error portion of the 

 series is to shake up the "set" at the place corresponding to 

 the error, but without shaking up the appropriate portions of 

 the "set." A shock, then, just severe enough to have only 

 an immediate effect is the desirable degree of punishment. 

 Its effect should not spread beyond the error portion of the 

 neural organization, and yet it should be efficient at that 

 particular point. An excessive shock may uproot the entire 

 "set" and destroy all progress that previous training has 

 fostered. The optimal punishment, then, should be just 

 enough to disintegrate the undeslred act (or acts) of the series 

 but not enough to extend beyond that one element in the series 

 to other valuable acts. 



My method of training chicks of the early groups was merely 

 to adopt what appeared to be a satisfactory shock (tickle) 

 and accept the chance results it brought. With the three 

 chicks discussed in this paper, particular attention was given 

 to the effect of the shock. It appeared that the heavier chicks, 

 24 and 25, reacted more strenuously than the lightest. No. 27. 



