88 ROBERT M. YERKES AND JOHN B. WATSON 



It is our conviction that the methods which are herein recom- 

 mended will enable us to discover, by the comparative method^ 

 facts which are of fundamental importance for visual theory. 



We desire to offer the following recommendations concerning 

 methods of studying the various aspects of vision which have 

 been considered in this report. 



Light perception. — That the standardized light apparatus, as 

 herein described or in improved form, be employed for accurate 

 determination of the capacity of organisms to respond to light. 

 In connection wdth the use of this apparatus w^e deem it im- 

 portant that tungsten lamps be used as sources of stimuli and 

 as photometric and radiometric standards; that the light box 

 be provided with movable screens which shall minimize the 

 reflection within the light box; that the standard plates be 

 accurately cut and kept in perfect condition; that differences 

 in the visibility of portions of the two symmetrical halves of 

 the experiment box, resulting from different intensities of the 

 two stimuli or from differences in the surfaces of the parts 

 of the box be excluded, as far as possible, by means of such 

 devices as are suggested on page 85 ; that experiments be con- 

 ducted in ordinary illumination whenever possible; that the 

 experimenter to the extent of his ability strive to render the 

 apparatus automatic in its operation so that he shall be able 

 to avoid disturbing the reacting animal; that intensive pains- 

 taking observations with one or a very few individuals be sought 

 in preference to extensive and superficially made observations 

 on a large number of individuals ; that a thorough investigation 

 of the light vision of an animal preface and prepare the way for 

 any investigation of its color vision ; that the intensity of lamps 

 used as sources be accurately determined" to begin with and 

 at stated intervals by means of reliable standards and measuring 

 devices, such for example as the standardized tungsten lamp 

 and the Lummer-Brodhun contrast photometer; that radio- 

 metric measurements be made with a radiomicrometer or a 

 selenium cell; that stimulus areas be photometered by com- 

 paring them (balancing them against) a precisely similar standard 

 area, 5 or 6 cm. circle, one meter distant, which in turn is illu- 

 minated by the standard source (tungsten). This standard 

 source is to be mounted upon a movable carriage. It is further 

 recommended that in arranging the apparatus for making 



