The EMC Effect — the inside changes with the neighbourhood

In 1983 the European Muon Collaboration measured the deep-inelastic structure function of iron and divided it by that of deuterium, expecting a ratio flat at unity. They found instead a depletion of 10–15% near : the quarks in a bound nucleon do not carry momentum as they do in a free one. Forty-three years later no single mechanism is agreed. This paper tests whether a packing-geometry variable derived from a nuclear-structure model — a cavity-filling measure (in standard notation a rescaled binding-energy-per-nucleon proxy, , equal to 1 for a free proton and falling as binding tightens), termed snugness in the model and discussed in §2 — tracks the measured EMC slope across seven nuclei, and reports the outcome without softening it. Snugness correlates with the EMC slope at on nuclei (a description at this ). It does not survive its own control: snugness is a smooth, monotone function of mass number A, and three of five arbitrary saturating functions of A match or beat it. The correlation carries no information beyond A itself, and is recorded here as a null. The informative outcome is the target the run exposes. The variable that genuinely tracks the EMC effect is the short-range-correlation (SRC) scaling factor , and is not monotone in A: it rises steeply to about and then saturates, with gold sitting below iron despite being 3.5 times heavier. No monotone function of A can reproduce that shape. We therefore state an exact constraint that any future candidate must meet — non-monotone in A, saturating above , and computed from geometry rather than fitted — and hand that shape forward. This record holds one paper on its own and is refreshed every couple of months. The complete deposit, 10.5281/zenodo.21788037, holds every paper with its code and data and is updated first; for the newest version of this paper, go there.