268 journal of Comparative Neurology and Psychology. 



ing the amount of difference in the illumination of the boxes with 

 ease and accuracy. These are only a few of the objections to white, 

 grey, and black cardboards or papers that experience enables me 

 to raise. Naturally I shall neither use them nor recommend their 

 use hereafter in investigations of the visual powers of animals. 

 Later, in connection with a report on '^Methods of studying vision 

 in animals" which is to be made by the committee on standardiza- 

 tion of tests of the American Psychological Association, I shall pro- 

 pose a substitute method. 



The investigation has shown, I believe, the great importance 

 of choosing conditions of experimentation Avhich may be readily 

 and accurately measured and controlled, and of determining, as a 

 preliminary to any experimental study of habit-formation, the value 

 for the individual animal of the several important factors in the 

 experimental situation. It has further shown that we should Avork 

 with individuals, and with relatively simple and perfectly analyz- 

 able situations ; and that no treatment of our results is so likely 

 to hide their real significance as the averaging of groups of ob- 

 servations for different individuals. Evidently the best prepara- 

 tion for an exj^eriment is a thoroughgoing study of the character- 

 istics of each animal to be used in the investigation ; and the best 

 result which an experiment can yield is evidence of the relation 

 of experimentally controlled conditions to the particular traits of 

 an individual animal. Averages are important, but we should 

 not sacrifice individual facts for the purpose of presenting them. 



The primary aim of the investigation, it will be remembered, was 

 the discovery of the relation of plasticity to age and sex. The 

 data prove that dancers at the age of one month acquire the white- 

 black habit more rapidly than do older individuals, and that the 

 females, on the average, acquire the habit more rapidly than do 

 the males. Of gi-eat importance is the fact (represented by the 

 curves of Fig. 3) that whereas the females make more mistakes 

 of choice at first than do the males they very soon begin to choose 

 with a higher degree of accuracy, and ultimately acquire a perfect 

 habit with considerably fewer training tests than the males. 



That these age and sex differences in the form of the habit- 



