84 ROBERT M. YERKES AND JOHN B. WATSON 



subject, the reader may be referred to the work of Ives, which 

 is to appear shortly in the Physical Review. 



Spectrometers. — There are many simple instruments upon 

 the market which will give the wave-length of the bands used 

 with sufficient accuracy for the present work. A pocket form 

 of instrument supplied with a scale is the most convenient. 

 The relation of the scale divisions to the distinct lines in the 

 solar spectrum is then determined, and the "curve" of the 

 instrument draw'n on millimeter cross section paper. 



The Browning pocket spectroscope is supplied by Max Kohl 

 for 82 marks. Hans Heele also furnishes a pocket spectroscope 

 with scale for the same price (purchase with Amici prism of 

 high dispersion). Either of these instruments may be imported 

 through the Arthur H. Thomas Company, Philadelphia. 



We suggested on p. 63 that the scale of the double slit be 

 calibrated directly in terms of wave-lengths. Once this has 

 been accurately done there is no further need of a spectrometer. 

 It is w^ell, however, to have one at hand for use in making 

 occasional tests. 



g. Experiment box. — On account of the great differences in 

 the structure, form, and instinctive capacity of animals w^hich 

 are likely to be used as subjects in these tests, no fixed type 

 of experiment box can be recommended for use with the color 

 apparatus. That shown in fig. 4 suggests a simple type of con- 

 struction. 



One of the chief difficulties in the way of using spectral light 

 in dark-room work is the peculiarity w^hich the shorter wave- 

 lengths exhibit in illuminating the interior of the black experi- 

 ment box. Even the darkest of our black paints reflects the 

 green rays sufficiently to show the outlines of the floor, the par- 

 tition, and the walls of the experiment box. The red rays, 

 on the other hand, are completely absorbed by some black 

 paints. This means that under the ordinary conditions of our 

 tests a secondary criterion is afforded the animal. On the basis 

 of the criterion, it soon learns to avoid the more highly illumin- 

 ated compartment, if its food is given with the color of 

 longer wave-length, or to seek it if its food is given with the 

 color of the shorter wave-length. Two animals, a white rat 

 and a rabbit, under experimental observation soon learned to 



