742 



LIGHT AND LIFE 



'Rhodcpsm. lodopsin, ond speciral sensitivity of 



rods ana cones: 

 • -frog 

 O snof^e 

 e cot 

 O guinea pig 



400 



700 



SCO 600 



Wove length - m/j 

 Fig. 5. The absorption spectra of chicken rhodopsin (\,„a^ 502 ni/x) and iodopsin 

 (Xmax ■'>fi2 mix), compared with the rod and cone sensitivities of various animals. 

 The lines show the absorption spectra of the \isual pigments, the points electro- 

 physiological measurements of spectral sensitivity (quantized). Rod data: frog 

 (Granit, 1947, p. 292); cat (Donner and Granit, 1949); guinea pig (Granit, 1942b). 

 Cone measurements: frog (Granit, 1942a); snake (Granit, 1943a); cat ((iranit, 

 1943b). (From VVald, Brown, and Smith, r.2). 



visual pigments. By the same token the Purkinje phenomenon — the 

 shiit ol spectral sensitivity toward the red in going from dim to 

 bright light, Irom rod to cone vision — involves only the shilt of 

 vision from depending upon the absorption spectrum of a rod visual 

 pigment in dim light, to depending upon the spectriun of a cone 

 visual pigment in brigiit light. 



Similar comparisons between the spectral sensitivity of human rod 

 vision and the absorption spectrum of a human rod suspension (50) ; 

 and between the spectral sensitivity of an invertebrate eye, that of 

 Limulus, as measured by Graham and Hartline (16), with the ab- 

 sorption spectrum of its visual pigment (29) , have proved equally 

 satisfactory. 1 think it is safe to conclude that when the data are 

 adequate, the spectral sensitivity of vision corresponds exactly with 

 the absorption spectrum of the associated visual pigment. 



This brings us to the second thesis, that the visual threshold, and 

 hence its reciprocal the visual sensitivity, rises and falls during light 



