4 K. S. LASHLEY 



likewise, to Dr. A. R. Middleton for assistance in equating the 

 energies of the stimulus lights. 



TECHNIQUE 



Game Bantam cocks of somewhat impure breed have been 

 used in all the tests, differential reaction to the stimuli being 

 obtained by the Yerkes discrimination method with food, or 

 food and punishment, as the stimulus to learning. Concerning 

 the suitability of the fowls for experiments little more com- 

 plimentary can be said than that they fit the apparatus. They 

 show most of the peculiarities recorded by Yerkes for the dove 

 and by Tugman for the sparrow, plus sexual reactions which 

 sometimes make work impossible. 



The Yerkes-Watson light apparatus, with some slight mod- 

 ifications for ease of manipulation, provided the visual stimuli. 

 This apparatus gives two fairly pure monochromatic bands of 

 which the wave-length, energy, and saturation are under ac- 

 curate control (Yerkes and Watson, '11). These bands are 

 reflected from similar plaster surfaces, 1 by 10 cm., exposed in 

 the alleyways of the discrimination box. Three surfaces are 

 used, one light being exposed constantly upon the middle sur- 

 face, the other being shifted to the right or left in unison with 

 lateral movements of the experiment box, thus effecting the 

 interchange of the relative positions of the lights. This technique 

 introduces a source of error which must be controlled carefully. 



The primary distinction to be made in tests for color vision 

 is between intensity and wave-length but reactions to other 

 characters of the stimuli and experimental procedure must be 

 guarded against. With the stimulus lights under complete con- 

 trol the discrimination method offers five other groups of stimuli 

 which may lead to deceptive results. These are listed below 

 together with the method of eliminating them in the final trials, 

 in part as recommended by Yerkes and Watson (op. cit. p. 84 ff). 



