COLOR VISION IN REPTILES 495 



could be trained to distinguish grays only if their shades were very differ- 

 ent; but they distinguished each of several colored papers from any of 

 seventeen grays. 



Twelve spectral lights were discriminated qualitatively, and the indi- 

 cations were that with longer training a great many more hue differences 

 might have been shown to occur for the species. The trained human 

 observer can distinguish about 160. The spectral limits for the turtle 

 were at least as low as 7,401m^ and as high as X760m\l — much the same 

 as for man and most other vertebrates. Hues were most easily told apart 

 when in the neighborhood of orange-red (A,634m|x), with weaker max- 

 ima of discriminability at the blue-green (A504m[l) and violet, and a 

 minimum in the blue. 



The most important hues for the turtle appeared to be orange, green, 

 and violet. Yellow and yellow-green, when not accurately discriminated, 

 were apparently most often seen as orange; but red was separated from 

 the general orange category and seemed to be more akin to violet for the 

 animal, which thus has a closed color circle. These peculiarities were 

 attributed, probably quite properly, to the restrictive filtering action of 

 the meager assortment of oil-droplet colors possessed by the turtle (see 

 Chapter 8, section D) . In contrast to the fishes, and in keeping with the 

 predominantly red and orange oil-droplets, the turtle showed an elevated 

 capacity for hue-discrimination in the long-wave region, where also the 

 position of the photopic brightness maximum was shifted from its human 

 (yellow-green) value well into the orange, toward the red. This incident- 

 ally is not in keeping with the properties of the turtle's 'zapfensubstanz' 

 (maximum absorption at A,560m[x) as given by Studnitz. The oil-droplets 

 also account for the observed low ability to discriminate hues in the 

 green and especially the blue region, but the apparent slight rise found 

 in the violet region seems paradoxical. 



Wagner, a year before Wojtusiak, and working in the same laboratory, 

 published the only study to date on lizards, apart from the peculiar 

 results of Schlieper on Lacerta vivipara mentioned above. His technique 

 was simple but effective. He found that Lacerta agilis was violently dis- 

 gusted by the taste of salt. Offering meal-worms pinned in front of discs 

 of colored paper on long handles, he obtained rapid training positive to 

 normal worms presented with one stimulus and negative to brine-soaked 

 ones offered with an alternative stimulus. When each discrimination was 

 finally established, as evidenced by twenty successive correct choices. 



