34 ROBERT M. YERKES AND JOHN B. WATSON 



light is not so satisfactorily applied thus as in a dark-room, and 

 quantitative experiments are impossible) ; impossibility of 

 obtaining readily any desired quality of stimulus. 



Unlike the reflection method, the transmission method may 

 be used fairly satisfactorily for quantitative work, for the stim- 

 ulus may be accurately measured and it may be applied, in a 

 dark-room, under simple controllable conditions. 



Since animals in nature view colors mostly as colored objects 

 seen in white light, it seems only fair that their color vision 

 should be investigated under conditions similar to these, as 

 well as under such conditions as are furnished by the clear 

 cut and controllable methods of transmission and dispersion. 

 The quantitative tests described in this report are open to the 

 objection that the darkened rocm, in which they should be con- 

 ducted, is an unnatural environment for most animals. In 

 view of this objection, we believe that for rough qualitative 

 tests of vision the reflection and the transmission methods 

 may be employed to advantage in daylight or sunlight. But 

 even in the case of qualitative experiments in sunlight, we 

 recommend the transmission method over the reflection method 

 because it enables the experimenter, first, to get a great variety 

 of qualities (hues) and intensities; second, because it enables 

 him to measure both the wave-length and intensity; and third, 

 because it enables him, within narrow limits, to vary intensity 

 independently of quality. 



We shall now point out the special characteristics of a number 

 of the substances which exhibit selective absorption and trans- 

 mission. 



Glasses. — Colored glasses are readily obtainable, but most 

 of them, when used singly, are practically worthless as filters. 

 There are two kinds in use : that in which the pigment is mixed 

 with the glass, known as pot glass; and the variety which has 

 a thin coating of pigment on one or both surfaces, known as 

 flashed glass. The former in good quality, is more expensive 

 than the latter. To these we might add the spectro-optical 

 glass, a special kind of pot glass 



Colored glasses may be described as ray filters in relatively 

 stable and permanent form, convenient for handling. Unfor- 

 tunately many colored glasses fade more or less rapidly in strong 



