20 'Journal of Comparative Neurology and Psychology. 



the color with the food, a total break in the process, and a resort, on 

 their part, to the use of a sensory process genetically lower than the 

 visual, namely, the kinsesthetic. Such behavior is suggestive of the 

 relatively unimportant role which color vision plavs in the life of 

 til is animal. 



3b. The onset of the position error in such a crucial place in J's 

 series (see p. 26) is especially unfortunate. His percentage of 

 correct choices increased normally during the formation of the asso- 

 ciation. This percentage remained high for the several days, during 

 which the various changes were being made in the negative color, 

 then all at once we find the habit disintegrating. On the whole, in 

 ■this case, it seems best to reserve judgment as to whether the animal 

 was reacting to blue on the basis of its possibly greater intensity, 

 and to await further tests upon him. 



4. If 0716 were to draio the general conclusion that the wave 

 length of a given monochromatic light stimulus is, or might he. 

 under suitable conditions, a factor in the adjustment of the animal 

 to that stimidus, one apparently would find abundant support for 

 the position in the above tables. The writer for the present, however, 

 prefers to allow the experimental data to stand as such without 

 drawing any conclusions from them. The reason for this position 

 becomes apparent when we consider the following points : 



Did the changes which were made in the relative intensities of 

 any two bands really reverse the intensity relation for the animal 

 (a glance at the table of constants will show the enormous changes 

 which were made) ? The answer to this question must come from 

 experiments. In order to answer it we would need to know for each 

 species of animal, 1st, the relative stimulating eifect upon it of the 

 different parts of the spectrum, that is, we would need to have a 

 curve for the animal corresponding to the curves which have been 

 constructed for similar reasons for the human eye in its normal and 

 abnormal states. The first step in constructing such a curve might 

 come through obtaining the animal's reaction thresholds (stimulus 

 limen, R. L.) for the separate spectral bands, e. g., the red, yellow, 

 green and blue ; 2d, beginning with these values (expressed in 

 photometric and in radiometric terms) as the lowest points in the 



