Movement-sensitive neurones in the toad's retina.

Abstract

1. Neuronal classes. Recordings from single optic nerve fibers of the common toad Bufo bufo (L.) have revealed three types of retinal ganglion cells which correspond to classes II, III and IV in the frog. 2. Receptive field organization. All neurones have a central excitatory field, ERF, surrounded by an inhibitory one, IRF (Fig. 10a, b). The ERF for each cell class was ERFII ≈ 4°, ERFIII ≈ 8°, ERFIV = 12–15°. 3. Illuminance change over the entire visual field had no effect on class II neurones, produced an “on — off” response in class III and an “off” response in class IV (Fig. 2b–c). 4. Angular size of stimuli moved through the central rezeptive field (constant angular velocity and stimulus background contrast). (i) Squares: As the edge length approached the ERF diameter, the discharge rate of all neurones increased, then decreased for larger squares which activated the inhibitory surround (Figs. 3, 4a). (ii) Extending a vertical stripe in the horizontal direction of movement had a similar effect (Fig. 4b). (iii) Elongating a horizontal stripe by more than 2° in the direction of movement produced no change in the discharge rate (Fig. 4c). (iv) Simultaneous movement of two stimuli a and b through the ERF caused a greater discharge than for either allone. Responses to a in the ERF were inhibited if b was in the IRF (Fig. 5). 5. Increased angular velocity (constant contrast and angular size) produced increased activation of all neurones (Fig. 6a–d). The degree of increase was different in each neuronal class (Fig. 7A). 6. Stimulus background contrast (constant angular size and velocity). The discharge rate generally increased for increasing contrast between stimulus and background (Fig. 8). A white stimulus against black background produced maximal activation of class II neurones; black on white was maximally effective for the other two classes (Fig. 9a–c). 7. Input-output functions. A power function best describes the relationships between impulse frequency and (i) stimulus angular velocity, and (ii) stimulus background contrast (Eqs. 9, 11). Impulse frequency is logarithmically dependent on stimulus area (Eqs. 3–6). 8. Retinal output and visual behaviour. The neurophysiological findings are compared with quantitative results previously obtained from corresponding behavioural experiments concerning visually induced prey-catching and avoidance reactions.