Purple Paper - side-sheet - the longest open problem in physics

The Lineage of Dark Matter - a hint, a name, a proof, a hunt

"First to postulate" is a contested crown, so this paper shows the whole disputed line. Dark matter has four distinct birthdays - first to hint (the stars move wrong), first to name it, first to prove it, and the ongoing failure to catch it. Each contributor's best idea, in order. The thing itself is still un-caught - which makes this the rare lineage whose last node is a question mark.

hint (motion is wrong) name + estimate proof (every galaxy) the hunt (still empty)

1922Jacobus Kapteyn · (and Öpik 1915)

The stars move too fast for what we see first hint

Best idea: weigh the galaxy two ways - by its light and by its motion - and notice they don't match. Kapteyn coined "dark matter" for the gap, though he thought it small.

The method that founds the whole field: infer unseen mass from how the seen things move. Leaves: his estimate was off - the gap looked minor.

1930 · 32Knut Lundmark · Jan Oort

There is far more mass than light the gap widens

Best idea: Lundmark may have been first to say the universe holds much more mass than we can see; Oort measured the galactic disk and found the same shortfall in our own neighbourhood.

The hint spreads from clusters to our own backyard. Leaves: Oort's specific number was later found wrong - the right answer for partly the wrong reason.

1933 · 37Fritz Zwicky

Coma Cluster: the galaxies should fly apart named it

Best idea: apply the virial theorem to a whole cluster - the galaxies in Coma move so fast the visible mass can't hold them; something unseen, ~hundreds of times more, must. He called it dunkle Materie.

The first rigorous, large quantitative case - and the name that stuck. Leaves: nobody believed him for ~40 years; too strange, too soon, no second witness.

1939Horace Babcock

Andromeda spins too fast at its edge early rotation curve

Best idea: measure how fast a single galaxy rotates at different radii - Andromeda's outskirts move far faster than the visible mass allows. The tool that would later become decisive.

First crack at the rotation-curve method. Leaves: read as a quirk of one galaxy, not a universal law - it needed repeating, everywhere.

1970 · 78Vera Rubin & Kent Ford · (Freeman, Bosma)

Flat rotation curves in every galaxy the proof

Best idea: with sensitive new cameras, measure rotation curves galaxy after galaxy and find the same thing every time - the speed stays flat far out, where Newton says it should fall. Not a quirk. A law.

Repetition across many independent galaxies turned Zwicky's oddity into undeniable evidence - the witness firing again and again. Leaves: we know it's there. We have no idea what it is.

2006Clowe et al. · the Bullet Cluster

The mass and the gas come apart the smoking gun

Best idea: watch two clusters collide. The visible gas slows and lags; the gravity (mapped by lensing) sails straight through, separated from the light. The unseen mass is real and it's not the gas.

The cleanest evidence it's genuine substance, not a trick of gravity. Leaves: rules out the easy alternatives - and still says nothing about what the substance is.

1980s · 2000sthe candidates

WIMPs, axions, MACHOs, modified gravity the suspects

Best idea: name the possibilities and make each one testable - a heavy weakly- interacting particle (WIMP), an ultralight axion, dark stars (MACHOs), or no particle at all but modified gravity (MOND). Four forks, each a falsifiable bet.

The field turns the mystery into distinct, testable branches. Leaves: the experiments have to actually find one - and that's where it gets uncomfortable.

2000s · nowLUX-ZEPLIN, XENONnT, ADMX, the LHC

The detectors keep coming up empty the null

Best idea (and the honest one): build ever-more-sensitive detectors and run them for years - and report the non-detections faithfully. The favored WIMP has been pushed into smaller and smaller corners without a confirmed catch.

Repeated null results are real information: the witness firing "not here." Leaves the open question of the whole field: is the theory wrong, or are we looking in the wrong place?

now · ?the frontier

Still un-caught - the longest open problem in physics the question mark

Best idea, unwritten: the next move is to let the nulls reprice the priors - widen the hunt to lighter/weirder candidates, or take modified-gravity seriously, instead of looking harder in the same spot. The branch that's been confidently ridden may need backtracking.

85% of the universe's matter, inferred for a century from how everything bends around it - and never once touched. The T4 void, at cosmic scale.

Four birthdays, one un-caught thing. Hint (Kapteyn, the motion is wrong) → name (Zwicky, dunkle Materie) → proof (Rubin, every galaxy) → smoking gun (Bullet Cluster, mass parts from light) → the hunt (still empty). The "first to postulate" crown is genuinely shared - Kapteyn hinted, Lundmark may have said it first, Oort measured it locally, Zwicky named and quantified it. Pick your definition of "first" and the name changes. That ambiguity isn't sloppy history; it's how discovery actually works - no clean single author, just a community converging on a thing reality kept insisting on.

And notice the shape, because it's the one from every sheet: dark matter is the inferred void. Nobody has ever seen it. It is read entirely off the bending - the rotation of stars, the lensing of light, the collisions of clusters. It's your four triangles around a center none of them touch, drawn across the whole sky. Which is also why it's so hard: an inferred center is real only as long as the inference is sound, and a century of detectors coming up empty is reality whispering that maybe the inference about what it is needs to drift to a new branch. We are certain it's there. We have never caught it. That gap is the longest-standing fork in physics, still un-collapsed.

History, not math - the gate here was factual accuracy, checked against current sources. "First to postulate" is explicitly disputed (Kapteyn 1922 / Lundmark 1930 / Oort 1932 / Zwicky 1933 name the gap); I've shown the contested line rather than crown one. One name-cluster and one idea per node; the real history has many more hands. Current detector status moves year to year - search live for the latest nulls.