COLOR VISION IN FISHES 479 



not mention the species by name. And, where he finds that the fishes 

 scatter when the spectrum thrown upon them is progressively narrowed 

 down to red alone, it might only mean that the animals were avoiding 

 the red because it was such a strong stimulus to them and they were left 

 with no other light in which to congregate. Hess believed, on the other 

 hand, that in reduced illuminations blue lights became too strong for the 

 fish owing to the retraction of the retinal pigment, and in this way ex- 

 plained Bauer's results with red-blue pairs, wherein the fish would enter 

 the red light only when the intensities of both were lowered. For Bauer, 

 this meant that since the fish had become dark-adapted it could no longer 

 discriminate hues, and consequently had no redness to avoid; but to 

 Hess it indicated that the blue light had become unpleasantly bright, and 

 was seen achromatically just as in light-adaptation, but with far greater 

 brilliance because of the removal of the shielding pigment from its 

 pathway. 



It is interesting to note that in fishes the migration of the pigment 

 itself, and of the cones, has recently been shown to take place maximally 

 in yellow light when physical intensities are equated. Red and blue evoke 

 the least movement of the elements, with orange, yellow-green, and blue- 

 green intermediate in effectiveness. This checks with Bauer's demon- 

 stration of a subjective brightness-maximum in the yellow, and also with 

 the peak of absorption of the photosensitive substance which von Stud- 

 nitz claims to have extracted from fish cones. 



Von Frisch emphasized two techniques, one of which involved train- 

 ing the fish to respond positively to one of two stimuli regardless of what 

 the other one might be, and the other of which made use of the response 

 of the skin pigment cells to colored backgrounds, which since the work 

 of Pouchet in 1876 had been known to be mediated through the eyes 

 (see next section). Frisch's training technique was essentially the one 

 which he had applied so successfully to the bee, in another controversy 

 with Hess which lies outside the scope of this treatment. After the fish 

 had been trained to come to a certain colored tube for food, regardless 

 of where that tube might be in a series of gray tubes in the aquarium, 

 he omitted the food (thus controlling olfactory and gustatory cues) and 

 found that the fishes — 'EUritze' (Phoxinus Icsvis) — always went to the 

 colored tube though six gray tubes at a time, out of a total of 50 gray 

 shades ranging from white to black, were presented along with the color. 

 Since there was no shade of gray which the fishes mistook for the train- 

 ing color, Frisch concluded that they could discriminate hues. 



