682 



HANDBOOK OF PHYSIOLOGY 



NEUROPHYSIOLOGY I 



the peripheral processes one can measure. Yet it is 

 important to learn how far one can come with these 

 if only to know that one must seek elsewhere for 

 what remains. 



Absorption Spectra and Spectral Sensitivity: 

 Purkinje Phenomenon 



The rise and fall of \isual sensitivity throughout 

 the spectrum is governed in the first instance by the 

 capacity of the visual pigments to absorb light of 

 various wavelengths, i.e. by their absorption spectra. 

 When properly corrected, the spectral sensitivity 

 should correspond closely with the absorption spectra 

 of the visual pigments. 



For such a comparison, the spectral sensitivity 

 must be corrected for distortions caused by colored 

 ocular structures, in the human eye principally the 

 yellow lens and macula lutea and similar structures 

 in the eyes of other animals. The spectral sensitivity 

 also should be quantized. What is measured generally 

 is the relative energy at each wavelength needed to 

 evoke a constant response. The reciprocal of this is 

 the relative sensitivity, and this divided by the wave- 

 length is the sensitivity in terms of relative numbers 

 of incident quanta. This is the form in which spectral 

 sensitivity data can best be employed for the present 

 purpose. 



The spectra of the visual pigments should be stated 

 in terms of percentage ab.sorption rather than ex- 

 tinction (cf 59). The point of this distinction is that 

 all extinction curves are simple multiples of one 

 another, whereas a percentage absorption curve has a 

 unique shape depending upon the actual value of the 

 absorption. However, extinction and percentage 

 absorption are almost exactly proportional to each 

 other up to 10 per cent absorption and depart only 

 slightly from proportionality up to aljout 20 per 

 cent absorption. All known cones and most rods seem 

 to have absorptions below this value. Extinction 

 therefore runs parallel with absorption for all cones 

 and for all but the more densely pigmented rods. 

 In the figures which follow, the absorption spectra 

 of the visual pigments have been plotted in terms 

 of relative extinction since the percentage absorption 

 usually is not known. This introduces appreciable 

 distortion only in comparison with frog rod \ision 



(cf. fig. 14)- 



Figure 13 shows the comparison between the 

 absorption spectra of chicken rhodopsin and iodopsin, 

 and the spectral sensitivity of rod and cone vision in 



the pigeon. It would, of course, be preferable to com- 

 pare the spectral sensitivity of the chicken, but 

 in the absence of accurate data measurements on the 

 closely related pigeon have been used. They were 

 obtained by inserting microelectrodes into the retina, 

 following removal of the lens and cornea (11, 19). 

 The pigeons were either dark-adapted i to 2 hours 

 following the operation, or were light-adapted. At 

 each wa\elength, measurements were made of 

 the energy needed to evoke a constant electrical 

 response. The reciprocal of the energy, the sensi- 

 tivity, was quantized by dividing by the wavelength. 



The scotopic sensitivity agrees very well with the 

 absorption spectrum of rhodopsin. The photopic 

 sensitivity however is displaced about 20 m/i toward 

 the red from the spectrum of iodopsin. This displace- 

 ment must be caused in large part by the brightly- 

 colored oil globules which lie in the cones of chickens 

 and pigeons in the position of color filters (79, 80). 

 The displacement seems larger than the color filters 

 of the chicken retina should cause and may mean 

 that many of the electrophysiological measurements 

 happened to fall within the 'red field' of the pigeon 

 retina, the dorsotemporal quadrant in which deep 

 red oil globules predominate. 



The shift of spectral sensitivity toward the red as 

 one goes from scotopic to photopic conditions, from 

 rod to cone vision, is the well-known Purkinje 

 phenomenon. Except for the distortion just alluded 

 to, this is accurately mimicked in solution by the 

 absorption spectra of rhodopsin and iodopsin. 



This comparison gains special force when made 

 with retinas which do not possess obviously colored 

 filtering pigments. In figure 14 the absorption 

 spectra of chicken rhodopsin and iodopsin are 

 compared with the spectral sensitivities of rod and 

 cone vision in the frog, snake, guinea pig and cat, 

 measured with electrical procedures by Granit and 

 co-workers. The scotopic data agree very well with 

 the absorption spectrum of rhodopsin. The photopic 

 sensitivities agree so well with the absorption spec- 

 trum of iodopsin that it seems probable that this is 

 the major pigment of cone vision in the frog, snake 

 and cat. 



Figure 14 shows that when colored ocular struc- 

 tures do not intervene, the Purkinje phenomenon 

 emerges quantitatively from the absorption spectra of 

 rhodopsin and iodopsin. In essence it in\olves 

 nothing more than the transfer of vision from de- 

 pendence on the absorption spectrum of rhodopsin 

 in dim light to that of iodopsin in bright light. 



