590 RADIATION BIOLOGY 



and hence is particularly capable of responding to a flickering stimulus. 

 It shows a renewed development of PII as the light goes off, as shown 

 by a large rf-wave in the electroretinogram. In some eyes (for example, 

 in the frog and owl) the retina is of the E-type in the dark-adapted con- 

 dition but changes to the I-type with hght adaptation (Granit, 1947). 

 Marked differences in response characteristics as described by Granit are 

 certainly present, but no such clear dichotomy as the terms "E-retina" 

 and "I-retina" imply has been demonstrated. Ultimately the bases for 

 such distinctions may well be found in the functional or anatomical pre- 

 dominance of rod or cone receptors and in the distribution of retinal ele- 

 ments responding in various ways to the onset and cessation of light. 



An interesting hypothesis supported by Granit (1938, 1947) is to the 

 effect that rods and cones are rivals competing for possession of the "final 

 common pathways" in the form of optic nerve fibers. Therman (1939) 

 has appealed to this hypothesis to account for his observation that a 

 given red light may actually arouse a larger action potential in a fight- 

 adapted frog than in a dark-adapted one. Red light is relatively (though 

 not absolutely) more effective for cones than for rods in the human eye 

 (see Wald, 1945). If this is also the case in the frog, Therman's finding 

 may lend some support to the rivalry hypothesis. 



Of interest here are observations by Granit et al. (1939) on the cat. 

 They found that the 6-wave was dependent upon the presence of a rela- 

 tively high concentration of visual purple. Even a moderate reduction 

 in this concentration (as by relatively short exposures to light) results in 

 a marked reduction in the 6-wave. In line with these observations, Riggs 

 and Johnson (1949) have shown that in the human eye very moderate 

 levels (less than 1 ft-lambert) of fight adaptation are sufficient to abolish 

 the principal component of the action potential except for stimufi of 

 extremely high intensity. A tenfold increase in level of adapting light 

 is sufficient to cause an increase of nearly a hundredfold in the intensity 

 of a test flash necessary to arouse a given magnitude of action potential. 

 The reduction in rod-initiated responses by light adaptation is well 

 estabfished. There is no direct evidence for Granit's further hypothesis 

 that cones assert themselves at high intensity levels by taking over the 

 pathways used by rods at low intensities. 



It is not possible in this short chapter to present a detailed account of 

 the many experiments on the action potential in various animals. Since 

 this is the only type of electrical recording which has been applied to the 

 human eye, we shall devote particular attention to some of the facts with 

 which it has provided us. Hartline (1925) obtained conclusive evidence 

 that the vertebrate electroretinogram is of the same form whether it is 

 recorded from the excised eyeball or from corneal and neutral leads to 

 the intact animal. He also obtained records of the human electroretino- 

 gram and showed that it was grossly similar to that of other mammals. 



