NEURAL ACTIVITY IN THE RETINA 



707 



tivity distributions of individual receptors and b) their 

 representation in the organized message dehvered by 

 the ganglion cells through the optic nerve. The latter 

 delivers the information which the striate area has to 

 interpret and so problem b is as important as prob- 

 lem a. The act of interpretation itself is at the moment 

 beyond the reach of electrophysiological approach. 

 Solution of the first problem requires more reliable 

 microelectrode records of individual receptor po- 

 tentials than is found in any paper hitherto presented. 

 Our most definite quantitative data are still the ones 

 obtained from individual optic nerve fibers of animals 

 (66, 68, 6g, 73). It is necessary to understand how a 

 spectral distribution is defined in order to comprehend 

 the color problem. A simplified presentation of this 

 question can be given in the following way. 



Assume that a color-sensitive substance absorbs light 

 along a spectral distribution curve represented in 

 every wavelength by S\. Dependent upon the lamp 

 used and upon other properties of the spectrum (e.g. 

 slit width, diffraction) each of these wavelengths tested 

 represents an amount of energy E\. Finally, again, 

 each of the wavelengths tested elicits an effect Lx. It 

 is immaterial now if this effect is considered in terms 

 of a receptor potential, a spike frequency or as per- 

 ceived brightness. This efTect L>, will be proportional 

 to the sensitivity Sx and the amount of energy Ex so 

 that Lx = Ex-Sx. In order to measure Sx which is the 

 function we want to study and which clearly is .Sx 

 = Lx/Ex we must first of all measure Ex of the spec- 

 trum used (which should be of a high degree of 

 purity). The next step is to set up the biological ex- 

 periment so that the physiological effect Lx is kept 

 constant in every wavelength. Then, with Lx and 

 Ex known, the equation can be solved, i.e. Sx can be 

 calculated. It is proportional to i/Ex. For c|u;)ntita- 

 tivc work it is therefore not enough to keep the energy 

 E of the spectrum constant and measure the physio- 

 logical effect L, even though such results may have 

 indicative value and can be approximately corrected if 

 the relation between E and L initially has been meas- 

 ured over the working range for each wavelength. 

 Very serious errors can also be introduced by filters 

 of which even the best have narrow color bands over 

 one or two log units only. Therefore spectra should 

 be used for quantitati\e work. A good method is to 

 mea.sure i/Ex for a constant response (Lx) such as 

 the threshold. This was the method employed in the 

 experiments on the discharge from indi\idunl nerve 

 fibers. 



These results, for which a large number of different 

 species of animals were used, some with pure cone 



450 



500 



0-550 



mO 700 



FIG. 14. Photopic dominator curves of the frog ( ) and 



the snake Tropidonotus natrix (• •). Equal quantum in- 

 tensity spectrum. Sensitivity plotted against wavelength. 

 [From Granit (66).] 



retinae, others with mixed retinae (after light- 

 adaptation), led to the dominator-modulator con- 

 cept. The optic nerve fibers were found to deliver two 

 types of curves, broad-band dominators and narrow- 

 band modulators. Figure 14 shows photopic domina- 

 tors of the snake cone eye and the light-adapted frog 

 eye which are of interest because the photochemical 

 systems of these eyes seem to he very similar to our 

 own. Actually their photopic dominators agree very 

 \\ell with the average photopic distribution of sensi- 

 tixity of the human eye. When the mixed eye is 

 dark-adapted, the same fiber that previously gave a 

 photopic doiuinator now gives a scotopic one with 



b 



maximum around 5000 A which agrees with the 

 sensitivity distriljution of visual purple or rhodopsin. 

 We recall that Polyak (122) had shown that both 

 rods and cones converge towards the saiue ganglion 

 cell. Thus the dominators are the carriers of the 

 Purkinje shift of retinal sensitivity with state of adap- 

 tation. In man the point of maximum shifts from 

 5560 to 51 00 A, just as in frogs and cats. The photo- 

 chemical aspects will be discus.sed elsewhere in this 

 volume (Wald, Chapter XXVIII), but it deserves to 

 be pointed out that dominators in various systems 

 have been synthesized by Wald out of vitamin A 

 aldehydes and rod and cone proteins with the aid 

 of various enzymes and that these synthesized prod- 

 ucts have absorption spectra in good agreement with 

 the experimental results obtained from optic nerve 

 fibers [see also the summaries by Granit (73, 75)]. 

 Examples of modulators from different animals are 



