24 K. S. LASHLEY 



rate investigation, that there is a marked change in the reac- 

 tion value of the spectrum between 500 and 530 up with no 

 marked change between 500 and 460 pp, i. e., the region corre- 

 sponding to the blue and violet for man begins at a longer wave- 

 length for the chick. 



To summarize this part of the work, the accuracy of the 

 discrimination of the fowls is affected by changes in the position 

 of the stimulus lights through the following intervals and it 

 is suggested that the wave-lengths lying between each pair of 

 these intervals, which are probably shorter than the tests indicate 

 have a fairly uniform reaction value. 



630 — 600 mm. 



590 — 580 " 

 { 565 — 550 " 

 j 565 — 535 " 

 j 520 — 500 " 

 "1540 — 470 " 



THE RELATIVE REACTION VALUE OF INTENSITY 

 AND WAVE-LENGTH 



In field studies of animals where the conditions of illumination 

 are not under control it is difficult to determine whether the 

 animal, even if known to have color vision, is reacting to color 

 or to brightness. The distinction is of importance in many 

 biological problems where the necessity for natural conditions 

 makes an actual test of the efficient stimulus impossible. The 

 only clue to it in such cases is given by the relative ease with 

 which habits of reaction to the two attributes of the light stim- 

 ulus are formed. The data upon the rate of learning in my 

 experiments gives some indication of the relative efficiency of 

 wave-length and intensity in light stimuli. 



a. Birds already trained in brightness discrimination, when 

 trained with red and green of different brightness values for 

 them (red 650, green 520 pp. at equal energies) changed quickly 

 to reaction upon the basis of wave-length, without any alteration 

 in the brightness values of the lights which might have destroyed 

 the habit of reacting to brightness. After training with colors 

 the association with wave-length was dominant even when the 

 brightness values of the stimuli were so greatly different as to 

 call out the instinctive brightness preference in untrained birdss 



b. The conditions under which training for brightness and 

 color discrimination have been carried out are not strictly 



