28 FREDERICK S. BREED 



Curve II-A, fig. i, shows the course of reaction 2, and Curve 

 III-A of reaction 3, for the chicks of Group A. The errors of 

 swallowing tend to vanish more rapidly and more completely 

 than those of seizing. The data on reaction i are too meager 

 to justify consideration here. 



Curve IV-B, fig. 2, is the curve of development of reaction 

 4 in Group B, chicks nos. 9 to 13, inclusive. These chicks were 

 hatched just eight days after those of Group A. The two groups 

 were segregated until chicks nos. 9 to 13 were beginning their 

 seventh day, and then were allowed to mingle. The pecking 

 tests were made throughout in the regular way, both groups being 

 treated as nearly alike as possible. Curves IV- A and IV-B, fig. 

 2 , are matched for corresponding days in the lives of the chicks — 

 that is, for corresponding ages, and not for the dates on which 

 the records were made. Although it' was not part of the plan, 

 when the experiment was conceived, to study social influence, 

 the higher curve for Group B indicates that possibly association 

 with the glder chicks had the effect of facilitating improve- 

 ment in the younger ones, and suggests a method of measuring 

 this influence. A glance at Curve IV- A, fig. 2, shows that the 

 efficiency of the chicks of Group A on their fifteenth day, when 

 those of Group B joined them, was numerically 16 on a scale of 

 20, and that the efficiency of Group B at the same time, the 

 seventh day for the latter, was 14.6. But the difference in the 

 curves may be due to unsuspected irregularities in method, or 

 to variability in the chicks. It seemed hazardous, therefore, to 

 rest such an important conclusion on one test, and a repetition 

 of the experiment was planned and executed. 



Curves IV-A and IV-B indicate, then, that there is a very 

 rapid improvement in the integration of the components of the 

 complete reaction within the first three days, followed by a 

 slower but fairly steady improvement after that for some time. 

 It is interesting to note just how nearly the development approx- 

 imates perfection, — 18.4 on a scale of 20 representing the degree 

 of approximation at the highest point. 



Curves III-A and III-B, fig. 2, trace the course of reaction 

 3 for Groups A and B respectively. 



The study of the pecking reaction was resumed almost a year 

 later with chicks nos. 57 to 65, inclusive, which we shall style 

 Group C. These chicks were hatched Oct. 25, 1908. With a 



