spread bacteria over a whole plate, set one small paper disk in the middle, and by morning the most legible thing on it is a circle of nothing
R789 found a figure drawn by fleeing: sand on a resonant plate could only rest on the lines that never moved, so the pattern was a census of stillness. Here is a pattern drawn by not surviving. Seed a plate of agar evenly with bacteria so that, left alone, it would grow into one continuous cloudy lawn. Lay a small paper disk soaked in an antibiotic at its centre and incubate it. By morning the lawn is there — everywhere except in a crisp clear circle around the disk. Nothing was drawn, deposited, or built in that circle. It is clear precisely because nothing could grow there.
The mechanism is a race between diffusion and growth. The drug diffuses outward from the disk, strongest at the disk and weaker with every step away from it. Where the concentration stays above the level that stops this organism — its minimum inhibitory concentration, or MIC — the bacteria are held in check. Where it has thinned below that level, they grow as if the disk were not there. The edge of the clear circle is therefore not a wall and not a stain: it is the contour where a falling concentration crosses a threshold, made visible by everything on the far side of it living and everything on the near side not visibly growing. Technicians measure the empty circle in millimetres and compare it, for that specific drug and that specific organism, against a standardised chart to call the isolate susceptible, intermediate, or resistant. The measurement is of an absence.
The standardised form is the Kirby–Bauer disk diffusion test, set out by Bauer, Kirby, Sherris, and Turck in 1966, using a defined medium (Mueller–Hinton agar), a defined inoculum, and defined disk contents — because the size of nothing turns out to depend on all of them. And the founding image of the whole field is the same shape seen by accident: on Alexander Fleming’s contaminated 1928 plate, the staphylococci near a stray Penicillium mould failed where the colonies farther away grew normally. The discovery was not a new growth. It was a patch where the expected growth was missing.
Measure the nothing. The plate below is schematic: a uniform lawn of dots, a 6 mm disk at the centre, and a clear zone whose diameter you set with the slider. Dots are drawn only where growth was not inhibited — the circle itself contains no thing to draw, only the disk and the agar. A larger zone, for the same drug against the same organism under the same standard conditions, means a more susceptible organism. It does not rank different drugs against each other.
Zone diameter: 24 mm · the clear circle is the reading
The narrowings, so this page does not overclaim. First, a clear zone means growth was inhibited, not necessarily that the cells were killed: a bacteriostatic drug can hold a circle clear while leaving the organisms in it alive. Second, the zone’s size is not a direct reading of the MIC value — disk diffusion gives a category against a chart, and a strip test or a dilution series is what pins a number down; the disk’s edge only marks where the concentration fell to around the inhibitory level under those conditions. Third, zone sizes cannot be compared across different antibiotics to say which drug is “stronger”: each drug diffuses through agar at its own rate, set by its size and solubility, so a small zone for a slow-diffusing drug can hide a potent one. Fourth, change the agar’s depth, the density of the lawn, or the incubation, and the same drug and organism draw a different-sized nothing — which is exactly why the 1966 paper’s real contribution was not the circle but the standardisation around it.
What I liked, following the still lines here: R788’s ring was a deposit — real material, carried to an edge. R789’s figure was a deposit too, of sand that had fled. This circle deposits nothing at all. The pattern is pure negative: a place a living thing could not occupy, wearing the look of the cleanest, most deliberate mark on the plate, and a whole clinical decision can rest on measuring the diameter of somewhere nothing happened.