With 3 figures
The revival within the past decade of a Fourier treatment of flicker (DE LANGE, 1954 and others) has owed much of its impetus to the possibility that the small size of the flicker sensation at flicker threshold allows the non-linearities of the eye to be disregarded. The present papier summarizes how electrical flicker responses from a simple eye do show that the approximation to linearity near flicker fusion may be very good indeed. The simple eye chosen belongs to the wolf spider Lycosa baltimoriana (KEY-SERLING); it appears to lack intra-ocular synapses (WIDMANN, 1908) and so may be suitable for recording electrical potentials from primary visual receptor cells alone. To bring the eye to a state like that at flicker fusion, experiments reported here all took place with eyes of intact animals light adapted by some amount of constant background illumination.
It is most convenient to begin with the usual sorts of rectangular shaped periodic stimuli. Tracings of averaged retinal action potentials elicited by 8 % incremental flashes are shown in Fig. 1 as solid lines; the simplicity of the responses is evidenced by the absence of on-or off-effects. This simplicity was a reason for choosing the spider eye for this work.
Next, consider that all these periodic stimuli be described as trains of flashes with the appropriate separations; then these flicker responses, if linear, ought to be simple superpositions or sums of the responses to all the component flashes of the trains. Such a predicted relation between flash and flicker responses is typified by the filled circles in Fig. 1, which are the superpositions obtained after repeatedly drawing the response to the single flash shifted each time along the time axis by a distance equal to the period of the stimulus. Clearly, the courses of the flicker responses are well predicted, even for such small flash separations that the response amplitudes approach 'fusion'.