1. Intracellular recordings have been made from the blowfly (Calliphora erythrocephala) retinula cell; apart from the transducer mechanism, these cells also feature a pupil mechanism. 2. At several mean intensity levels, within the apparently linear range of response, frequency characteristics of amplitude and phase and responses to 'delta'-flashes and 'delta'-flash pairs have been obtained. 3. Fourier methods have shown these responses to be mutually compatible, confirming linearity in these circumstances. 4. Non-linear behaviour can be made to appear at the lower frequencies when the modulation depth is increased. 5. Non-linearities can also appear through application of the superposition test: a low frequency sine wave, modulated so as to elicit an apparently linear response, and a high frequence sine wave which does not give rise to non-linearity even at the highest modulation depths can, when superimposed, yield a greater response to the latter when situated at the minima of the former than at its maxima. 6. At frequencies above approximately 1 Hz these superposition non-linearities are attributed to the transducer mechanism gain control. Below this frequency the pupil mechanism takes part considerably in the retinula cell's total observed gain control: its characteristics remain yet to be cleared up. 7. The transducer's linear and non-linear properties fit in closely with those of the Fuortes-Hodgkin model which couples increases in gain and time constants. 8. The Fuortes-Hodgkin model will probably require some quantitative modifications in the originally treated case of Limulus, on account of its pupil. 9. Finally, the merits of Veringa's diffusion model, and the possibility of eventually joining this model with the Fuortes-Hodgkin one are pointed out briefly.
展开▼
