Notes: Summary We report on experiments which were undertaken in an attempt to clarify mechanisms underlying the contrast effects of chromatic surround illumination on spectral responsiveness of cells in the parvocellular layers of the LGN (P-LGN-cells), that had been demonstrated under standard conditions in the preceding companion paper. The experiments were done in anesthetized macaques (Macaca fascicularis). In some neurons, S-potentials were recorded together with the post-synaptic action potentials, and all effects seen in P-LGN-cells were present already in their retinal afferents indicating their retinal origine. The responsiveness of the cells for center stimuli of different wavelengths and during illumination of the receptive field center or the outer surround was determined. Continuous outer surround illumination alered maintained dicharge rate (MDR), sensitivity and gain of P-LGN and retinal ganglion cells in the same way and empirically not distinguishable from direct illumination of the receptive field. Responses to surround flashes showed the same dependence on spectral composition as those to center flashes. Adaptation and excitation caused by outer surround illumination (inner diameter 5°, outer diameter 20°) were, in the average, ten times weaker than those exerted by light of the same spectral composition shone directly into the receptive field. Surround effects decreased proportional to r-2. Excitation by outer surround flashes was reduced by adaptation of the receptive field center in the same manner as responses to center flashes. The findings indicate that outer surround light has a direct excitatory and adaptive effect on the excitatory or inhibitory cones feeding into the receptive field. This indicates that straylight from the surround into the center could be responsible for the adaptive and excitatory effects of surround illumination. The straylight fraction from the remote surround into the receptive field must be higher, however, than that estimated from the psychophysically determined point spread function. It comes closer to earlier direct straylight measurements in excised eyes, but may be enhanced by chromatic aberration. If a surround of excitatory colour is flashed simultaneously with an excitatory center stimulus, additivity of center and surround excitation is observed only at low center intensities, while at higher center intensities the gain for center excitation is reduced similar to adaptive gain control. This could be explained by lateral interaction through horizontal connections in the retina, which decays within seconds, while adaptation of the cones feeding into the receptive field center is fully effective only after about 3 s. Our findings therefore suggest a two stage model for surround effects, a fast one mediated through horizontal connections controlling the gain of receptorbipolar transmission and a slow one through adaptation by straylight and controlling receptor gain. The fast process is receptor unspecific, i.e. pooling activity from all receptor types, while the second one is receptor specific. During real seeing both processes are simultaneous and complement each other because of continuous eye movements. Perceptual darkness and colour induction by remote surrounds are consistent with this model, which can also be applied to colour constancy. WM-cells (Yellow-minus-blue) show peculiar properties during surround or center illumination with blue light, suggestng an opponent mechanism different from that of other P-LGN-cells.