7o8 



HANDBOOK OF PHYSIOLOGY 



NEl'R0PH^■S10L0Cn• 



100 



% 



80 

 60 

 40 

 20 



450 



500 



550 



600 



650 



r 



FIG. 15. Modulator curves. Doli, rat; broken line, guinea pig; 



continuous line, frog; O O, snake. Equal quantum intensity 



spectrum. Sensitivity plotted against wavelength. [From 

 Granit (66).] 



given in figure 15. The modulator at 5000 A was 

 obtained after light-adaptation from the rat, an ani- 

 mal with rod eyes. In the dark-adapted state it had 

 been a dominator of the ordinary rhodopsin type. It 

 is by no means rare to find, in practically pure rod 

 eyes such narrow curves with ma.ximum around 5000 

 A, when visual purple activity has been suppressed 

 by light-adaptation. The other modulators very 

 clearly occupy three regions of predilection which 

 have recurred since in many other measurements. 

 The narrowest modulators ever seen were found by 

 Donner (52) in the pigeon cone retina where they 

 also occurred in three regions of predilection and 

 shifted slightly towards wavelengths which are long 

 compared with those of frogs. Donner suggested that 

 this shift was due to colored oil globules. Modulators 

 have also been obtained by selective adaptation to 

 different wavelengths as well as by electrical polari- 

 zation. 



A much debated question is whether these modu- 

 lators represent the more or less pure absorption 

 curves of photochemical substances or are products 

 of neural interaction based, for instance, on a mini- 

 mum of three broad-band curves. The simplest basis 

 for such interaction would obviously be overlap of 

 liroad-ljand curves, such as those of Dartnall C35)> 

 the pathways of one set of cones synaptically sup- 

 pressing or e.xciting those of the neighbor cones with 

 which they overlap in spectral sensitivity. As to such 

 interaction, it is true that it has been shown to e.xist 

 (43, 71), especiallv b\' polarization methods (70), but 



this does not necessarily constitute proof that it 

 actually did occur under the circumstances of thresh- 

 old experiments of the type used to establish the 

 concepts. On the other hand, retinal photochemistry, 

 though highly developed in many interesting experi- 

 ments by several workers [see Wald's summary in 

 Chapter XXVIII of this volume; also Granit (73, 

 75)], has not yet reached the point when it would be 

 possible to state that narrow-band photochemical 

 substances do not occur in living retinae. The most 

 that could be said is that broad-band curves seem to 

 be easier to demonstrate. Further work will no doubt 

 solve this problem. 



Much work has lately been devoted to the study of 

 the action of intermittent or flickering light. In 

 general, rod eyes have been found to fuse flicker at 

 lower values than cone eyes. In a mixed eye in the 

 dark adapted state rod sensitivity is high, for example 

 in the frog retina with roughly equal numbers of rods 

 and cones. Fusion frequency of the ERG to a light 

 of some 2000 lux in this animal will then be around 

 7 to 10 flashes per sec. But if this light is allowed to 

 shine for a while so as to light-adapt the eye, the 

 fusion frequency will soon rise to values around 20 

 flashes per sec. Now why could not the faster cones 

 also participate in tlie dark-adapted state and raise 

 the fusion frequency to their higher rate? Why was 

 light-adaptation necessary, particularly if the ERG 

 is a pure receptor affair and not influenced by inter- 

 action? Perhaps the reply is that interaction does 

 occur so that highly sensitized rods suppress the 

 cones, as was suggested by Granit & Riddell (80) 

 when they made this experiment. They also demon- 

 strated that the flickering wavelets change character 

 as light-adaptation proceeds and, besides, are dif- 

 ferent in rod and cone eyes. The same changes can 

 be seen in the ERG of man (41). 



Figure 16 is from experiments with ERG in guinea 

 pigs, cats and pigeons (45, 49) and shows a graph of 

 fusion frequency against light intensity in double 

 logarithmic plotting. Clearly there are two branches 

 of the curve in cats and guinea pigs. Much evidence, 

 presented in Granit's summary (73), goes to show 

 that the lower branch is a scotopic and the steeper 

 portion a photopic function. The less the number of 

 cones (their being far fewer in guinea pigs than in 

 cat.s), the higher the intensity at the kink of the curve. 

 The pigeon with cone dominance in the ERG has 

 no low-intensity branch but the curve rises steeply to- 

 wards values as high as around 1 30 per .sec. 



Records from the large retinal ganglion cells of 

 cats have shown (54) that fusion frequenc\' is pro- 



