34 



FREDERICK S. BREED 



given their tests at the same period in the morning, and of course 

 were fed together in the same litter, both groups being com- 

 pelled to scratch for their food. Yet the similarity of the curves 

 for the first eight days, when modification progresses most rap- 

 idly, is conspicuous, the curve for the younger brood running 

 slightly lower. These young chicks, in spite of the examples of 

 more accurate pecking furnished them by the behavior of the 

 chicks of Group E, began less accurately than their elders, 

 remained behind them by about the same margin during the 

 critical period of development, and hardly equalled them while 

 the experiment continued. 



Tria/s 

 50 



40 

 3S 



20 

 15 

 10 

 5 



Days 1 



12 U 14 is 16 17 18 19 20 21 22 23 24 25 



Figure 5 — A comparison of the curves of development of the pecking instinct to 

 show the possible effect of social influence. Distances along axis of abscissae 

 represent days of age ; distances along axis of ordinates, the number of a given 

 type of reaction in a daily series of fifty reactions. Curves I-E and I-F show 

 the rate of decrease in the number of reaction 1 for Groups E and F, respect- 

 ively; Curves II-E and II-F, the rate of decrease in the number of reaction 2 

 for the same groups. 



A feature worthy of notice is the fact that development of 

 reaction 4 in both groups seemed to halt on the third day, with 

 very rapid improvement preceding and following. A search for 

 the cause of this retardation of development in the chain of 

 actions is interesting. A detailed examination of the relations 

 of Curves I, II, III, and IV for both groups brings out the fol- 

 lowing points: 



ist day, the course of development is unknown; 

 2nd day, striking and seizing improve rapidly, while swallow- 

 ing (in the restricted sense) declines considerably in effectiveness ; 



