80 ROBERT M. YERKES AND JOHN B. WATSON 



" To determine whether or not an animal has color vision, 

 it is essential, in testing the response of the animal to two differ- 

 ently colored beams, that the energy carried by these two beams 

 be the same — i.e., that they differ in wave-length only.^ This 

 control is obtained by the use of a selenium cell whose color 

 sensibility curve is known. For the details of the method used 

 in determining such sensibility curves the reader is referred 

 to the Physical Review, vol. 28, p. 324. Suffice it to say here 

 that the sensibility curve is plotted between galvanometric 

 deflections caused by changes in resistance of the selenium 

 cell, and the wave-length of incident beam, the energy carried 

 by each bundle of monochromatic rays being constant. Given 

 the sensibility curve of a cell it is evident that we have the 

 means at hand of determining the condition under which two 

 bundles of monochromatic light carry the same energy. 



" Selenium cells obtained from different makers do not neces- 

 sarily have the same sensibility; hence it is necessary that each 

 cell be calibrated as above indicated. Furthermore it is required 

 to know in terms of C.G.S. units the energy carried by the light 

 ^used in the excitation of the cell. This is necessary in order 

 that the result obtained by any one investigator may be strictly 

 comparable with those of any other investigator. Although the 

 selenium cell connected in series with two or three dry batteries 

 and the galvanometer may be used when the amount of energy 

 to be measured is large, it is best to use the cell as a part of 

 the simple potentiometer, thus placing all possible degrees of 

 sensitiveness at the immediate disposal of the investigator. 

 The following plan is suggested for the standardization of 

 this work : 



"A central bureau is to be decided upon, where the calibration 

 of all instruments is to be carried out. The color sensibility 

 curve for all cells is to be carried out in the wave-length interval 

 450-800, and for the same absolute intensity of light. Lower 

 intensities than that at which the cell is calibrated are obtainable 

 through the use of a rotating sector. Due to the fact that the 

 selenium cells, galvanometer, batteries and resistances will not 



^ This is only one of the reasons for an energy control. We desire oftentimes 

 to present the two beams when they differ very markedly in energy, and for the 

 sake of reproducibility of conditions it is essential to know the energy of the beams 

 at all times. 



