480 ADAPTATIONS TO PHOTIC QUALITY 



In this same way Frisch also determined the discriminability of colors 

 from each other, and found that the fishes confused red with yellow, but 

 not blue with green or either blue or green with either red or yellow. 

 Purple was also sometimes confused with red and yellow, suggesting a 

 closed color circle which was firmly estabUshed years later for Phoxinus 

 by other investigators employing spectral lights. In 1923, Burkamp 

 offered Phoxinus as many as 23 pigmentary grays simultaneously with 

 a training color, whose position in the mosaic of grays was irregularly 

 varied, and found that the color was never confused with any gray — 

 an abundant substantiation of Frisch's earlier work. 



Frisch essentially repeated Zolotnitzky's experiments, which Hess had 

 also repeated (with altered technique and complete failure) . After train- 

 ing Elhitze on yellow meat, Frisch substituted bits of yellow paper on 

 gray backgrounds of the same texture, including a shade of gray which 

 matched the yellow for his own dark-adapted eye. On each gray back- 

 ground he also fastened bits of other gray papers both lighter and darker 

 than the yellow. Trained fishes snapped mostly at the yellow bits, un- 

 trained ones equally at all three. Hess repeated this experiment also, 

 again altering the technique, and got negative results. His fishes trained 

 positive to yellow would afterwards snap at blue objects as often as at 

 yellow ones; but it has been pointed out that he had not kept blue objects 

 out of the situation during the training, and had made no effort to 

 prevent any 'blue = food' association at the same time that he was build- 

 ing up the 'yellow = food' one. Frisch also turned the weapon of retinal 

 migration back upon Hess, when he eventually demonstrated that the 

 particular intensity in which Phoxinus ceases to discern the chroma red — 

 is dark-adapted, in other words — is one in which the photomechanical 

 changes will run to completion of the dark-adapted pattern (see p. 149). 



The response-to-background technique also yielded positive results on 

 Phoxinus in Frisch's hands; and since he first popularized the method it 

 has since, on less drab species, yielded even more striking findings than 

 his own. Phoxinus ordinarily responds to a yellow (or red) background 

 by becoming yellowish. It has no other capacity for color change; but it 

 responds to light and dark backgrounds by corresponding lightenings 

 and darkenings of the skin. Frisch took advantage of the fact that the 

 change in tone takes place in a few seconds, while several hours are 

 required for the change in hue to be accomplished. From a finely graded 

 series of black-gray-white papers, he was able to select a gray to which the 

 fish made the same brightness-response as to a particular yellow paper. 



