Begin with the sky. From every direction, always, space delivers a converging press — the influx from the boundary of the visible universe, a sphere that grows two light-years wider every year. Nothing needs to be added for this press to exist; it is what space is doing. Stand anywhere and the whole sky pushes inward on that point, evenly, from all sides at once.
Evenly — unless something blocks a part of it. Occlusion — one body blocking part of the pressure converging on another, so the unblocked side wins — matters because that imbalance is every force in this theory. Blocking is not a side-effect here. Blocking is the work.
What does the catching? A form. The proton is a loop of moving substance tied as a trefoil — a knot of three crossings, the simplest loop that cannot be undone without cutting; topology, the study of shape that survives bending, matters here because an untieable knot is a particle that cannot decay away. Follow the path and you cross six arcs: under, over, under, over, under, over.
And the rope itself is built, not asserted. Take the electron's standing loop in hydrogen, drawn as a tube of the electron's own width. Pack one thousand eight hundred and thirty-six of those volumes into the knot's length — the proton-to-electron mass ratio, written as geometry. The strands travel together in a slow lay, like paired wires, weaving as one group.
Bring two forms together and ask what binding is. Not glue. Gearing. A proton and a neutron — a proton carrying one electron inside — mesh at their rims, and the rims are narrow: the poloidal flow allows only a small window of contact, so angle is the premium, not area. For the windows to mesh at all, neighbours must turn in opposite senses — proton one way, neutron the other. Complementary, or grinding.
Spin the meshed pair fast and each knot blurs into the tight ring it truly occupies at speed. Meshed, they work parallel, in synchrony — but hear this exactly: the total movement of entwined parts is identical to their movement apart. Every part spends its whole budget at c, meshed or free; nothing is saved, nothing is lost, and no bookkeeping can find a missing tick. The only thing that can differ — the only thing there is to differ — is how the pair answers the arriving flood together: their cooperative resistance to drag. Two forms meshed answer the push as one built thing, and the resistance of the union is not the sum of the resistances of its parts. That difference has a value. Hold that thought.
Here is the question this page exists to answer: when two forms interleave without interfering, whose width is the shared region? It belongs to neither separately. Any measure that assumes each nucleon keeps a private, separable width must fail for bound nucleons — and the literature's "charge radius" assumed exactly that, on top of assuming a charge-substance the theory does not contain. It doesn't compute. It never could.
The bookkeeping that survives interleaving is the union. Light the packed nucleus and read its one shadow: where forms overlap along the light, the shadow counts the shared region once. Sum the shadows of the isolated forms, subtract the shadow of the union, and the difference — call it ΔA — is precisely what the private-width picture overcounts. That overcount is the mass defect: the resistance the union no longer carries — the parts' movement untouched, their cooperative answer to drag changed. The shadow is not a picture of the defect. The shadow's arithmetic is the defect's arithmetic.
The alpha — two proton–neutron pairs — sits at the centre as four interleaved poles turning on one shared axis. Around it, satellites take the most-spread positions opposition allows: one lone valence riding above (lithium); a tense dyad at opposite poles, chasing each other around the core (beryllium); then trigonal, then tetrahedral — carbon's four deuterons standing where its chemistry says they should — then the five-fold bipyramid, the octahedron, and so on, until twelve close the icosahedral shell and the next shell opens. The 720° rule — a closed shell must carry exactly 720 degrees of angular defect — is why twelve sixty-degree caps close a sphere, why the wire model pulls itself into a bowl, and why nuclei want to be round.
CALIBRATED One scale converts shadow-area to movement: κ, fixed once by least squares across the table — 12.9 MeV per square femtometre on the opposition packing. One dial, declared.
MEASURED With that single dial, the shadow union tracks the measured binding of 217 isotopes at R² = 0.981, mean error under nine percent — and the law held when the packing geometry was completely rearranged, which is what you would expect if the defect belongs to the overlap and not to any one arrangement.
OPEN What is owed, plainly: the light alpha-cluster nuclei bind a little less than pure overlap suggests, and the iron peak's fine structure does not fall out of geometry alone — both point at the depth of the shared-electron well, the angular-window ledger still to be built. A gap named clearly is a kind of teaching too.