428 Comparative Animal Physiology 



death of the animal, the response changed in character regularly twice daily, 

 even though the animal was maintained in total darkness. It is possible to 

 maintain Chlaenius in complete darkness, except for test flashes, for a week 

 and to record a diurnal variation in the electrical responses. 



During the night phase the eye is more sensitive, in that the light in- 

 tensity necessary to produce a minimal detectable electrical response is about 

 1/1000 of that necessary during the day phase. This diurnal variation in sen- 

 sitivity is not correlated with iris pigment migration, for in Dytiscus such mi- 

 gration occurs only as a response to changes in the state of light and dark 

 adaptation. ^"^ 



The electrical activity of the neurones of the optic ganglion of insects 

 may also be detected by measurements made with electrodes on the cornea. 

 The activity of the ganglionic neurones is often rhythmic and appears as 

 ripples superimposed on the electroretinogram. If the optic ganglion is 

 excised the oscillations can no longer be recorded from the eye. Excision of 

 the brain, however, has no obvious eff^ect on these oscillations,'^'*' ^'*^ The 

 frequency of this oscillatorv activitv is usually 8 to 40 or more cycles per 

 second and apparently results from the synchronized activity of the ganglionic 

 neurones. Oscillatory activity occurs when the light is on, or for a short pe- 

 riod of time just after illumination has ceased."* The frequency and magni- 

 tude depend on the temperature, on the state of adaptation, on the intensity 

 and exposure period of the light, and on the species. In many respects the 

 oscillatory activity of the optic ganglion is very similar to that of the verte- 

 brate cereberal cortex which is recorded in the electroencephalogram. 



SUMMARY. The measurement of visual functions by means of the retinal 

 action potential agrees with measurements of the same visual functions by 

 other techniques. Moreover, the correlations lead inevitably to the conclusion 

 that the retinal action potential is controlled by a photochemical reaction. 

 The previous discussion does not reveal any glimpse of the transitional mech- 

 anism between the photochemical event and the electrical event in the retina; 

 this remains to be demonstrated in the future. Nor does the foregoing dis- 

 cussion prove irrevocably that the retinal action potential is a critical event 

 in the peripheral visual mechanism; it merely indicates the possibility that it 

 is. Consideration of the relation between the retinal action potential and the 

 discharge of nerve impulses in the fibers of the optic nerve is of interest in 

 this connection. 



Electrical Activity of the Optic Nerve: Nature of Optic Nerve Activity. 

 One of the most useful techniques in the study of vision is that of recording 

 the impulses from individual fibers of the optic nerve, either from the nerve 

 itself^^ in invertebrates, or from the fibers on the inner surface of the retina 

 before they form the nerve^^ in vertebrates. This method of investigation 

 has yielded much of our knowledge of the action of the sensory cells, both as 

 individual units for the detection of light and as units for the discrim- 

 ination of wave length (color vision). 



Most aff^erent nerves, when stimulated via their end-organs, respond with 

 a train of action potentials rather than with a single action potential. The 

 fibers of the optic nerve are no exception. An example of the response of a 

 single nerve fiber of Limidus to prolonged illumination of the sensory ending 

 is shown in Figure 133. In Limidus each fiber is connected to a single sensory 



