METHODS OF STUDYING VISION IN ANIMALS 15 



the light vision of an animal involves the presentation, in a 

 dark-room, of two visual areas which differ only in intensity 

 of illumination. These visual areas are obtained by illuminating 

 semi-opaque glass plates with a standard source which is placed 

 on the opposite side from the animal. The chief advantages of 

 this general method of presenting lighl stimuli, in contrast with 

 the reflection methods which are usually employed by psychol- 

 ogists and often by physiologists, are three: (i) The visual 

 stimulus 'can be controlled accurately with respect to intensity ; 



(2) it may be photometered and radiometered with accuracy; 



(3) other visual factors than intensity of light may be excluded 

 or kept constant; (4) other factors than the visual may be ex- 

 cluded or controlled more certainly and satisfactorily than in 

 the case of most methods. 



While recognizing the value of the commonly employed 

 methods as means of obtaining certain kinds of information about 

 the light perceiving capacities of an organism, we wish to urge 

 the use of more accurately controllable and measurable stimuli 

 in all quantitative investigations. 



II. METHODS OF INVESTIGATING VISUAL SIZE, FORM, AND 

 DISTANCE PERCEPTION 



I . Size perception 



Students of vision have investigated the perception of size, 

 with few exceptions, by means of crude methods ; and the results 

 of such quantitative studies as have been made by excellent 

 methods can not be compared because of fundamental differ- 

 ences in experimental procedure. It has been our purpose to 

 devise a method which shall be adaptable; which shall yield 

 accurate quantitative results; and which shall render possible 

 and profitable the direct comparison of results obtained with 

 different animals. 



Size perception, or the ability of an organism to detect dif- 

 ferences in size, has been studied chiefly by " the method of 

 the discrimination of objects " (boxes, blocks, balls, glasses, 

 etc.) on the basis of difference in size. Thus, for example, an 

 animal is tested by being required to select the smaller of two 

 boxes in order to obtain the reward of food. The method has 

 many advantages — like all other reasonably well planned quali- 



