METHODS OF STUDYING VISION IN ANIMALS 27 



one another in a definite, constant, measured way with respect 

 to some one characteristic, and only one. It should, further, 

 enable him to exclude all secondary and irrelevant sensory cri- 

 teria by which the discriminative reaction of the animal might 

 be influenced. 



Account must be taken, in any accurate visual tests, of the 

 animal's degree of visual adaptation. This must be controlled. 

 It is necessary, therefore, to control carefully, and to measure 



(a) the illumination of the experiment room between tests, and 



(b) the intervals of exposure to the influence of stimuli and to 

 general illumination, or the lack thereof. No deflnite procedure 

 can be recommended because the method must be determined 

 by the characteristics of the animal under observation. 



The stimulus adapter must be adjusted, between experiments, 

 noiselessly, and the experimenter must keep himself beyond the 

 range of vision of the animal as much as possible. 



In order, then, that reasonably accurate and strictly com- 

 parable data concerning the light, size, form, and distance per- 

 ception of animals may be gathered in dift^rent laboratories 

 we recommend : 



(i) That the simple discrimination method be employed by 

 presenting to the animal simultaneously tw^o photic stimuli, the 

 one of which calls for a positive response and the other for a 

 negative response.' 



(2) That the following motives, singly or in combination, be 

 considered, and if possible used as conditions favoring the estab- 

 lishment of the proper method of reacting: (a) Escape from 

 the experiment box to a larger, more comfortable, and more 

 natural (home-like) cage; (b) return to companions; (c) the 

 .obtaining of food by choice of a certain stimulus; (d) the avoid- 

 ance of punishment — weak electric shock or other suitable 

 form — for inappropriate reactions. 



In our opinion it is desirable, for quantitative investigations, 

 that conditions for the establishment of a habit be chosen with 

 careful consideration of their constancy and controllability. 



^ For full description of the discrimination method and various modifications 

 thereof the reader is referred to the following publications: 

 Yerkes, R. M. The dancing mouse. New York, 1907. 

 Watson, J. B. Some e.xperiments bearing upon color vision in monkeys. Jour. 



Comp. Neurol, and P.'^ychol., 1909, vol. 19, pp. 1-28. 

 Casteel, D. B. The discriminative ability of the painted turtle. Jour. Animal 

 Behavior, 1911, vol. 1, pp. 1-28. 



