An open notebook. The evidence here is rock-solid; the explanation is not. I'm writing down where the weight of it currently pushes me, and what would push me off.

The problem, stated plainly

Galaxies spin. If the only mass in a spiral galaxy were the stuff we can see — stars, gas, dust, all crowded toward the bright center — then the outer stars should orbit slowly, the way Neptune dawdles compared to Mercury. Gravity ought to weaken far from the visible mass.

It doesn't. Vera Rubin and others clocked the orbital speeds of stars far out in galaxies — read off the Doppler shift of their light — and found the curve goes flat: the edges move just as fast as regions much closer in. Something massive and unseen reaches well beyond the glowing disk — a vast halo whose gravity keeps those fast edges from flinging off.

Two more clues:

• Gravitational lensing. Mass bends light. Galaxy clusters bend the light of background galaxies far more strongly than their visible matter could — there must be more mass than meets the eye.
• The cosmic ledger. The afterglow of the early universe and the pattern in which galaxies clustered both demand roughly five times more matter than the ordinary kind can supply.

So something is there. The honest question is what.

The suspects

MACHOs — Massive Compact Halo Objects. Ordinary matter that simply doesn't shine: faint dwarf stars, stray planets, old black holes. Tidy, and it needs no new physics. But surveys watching for the brief brightening as such an object drifts in front of a background star haven't turned up nearly enough. Mostly ruled out as the bulk.

WIMPs — Weakly Interacting Massive Particles, and their cousin the axion. A new kind of particle, left over from the early universe, that feels gravity but barely touches anything else. This fits the broad data beautifully. The catch: a generation of exquisitely sensitive detectors, buried deep underground, has been waiting for one to nudge a nucleus — and so far, silence.

Modified gravity — maybe nothing is missing, and our law of gravity is just wrong at galactic scales. MOND, the best-known version, nails those flat rotation curves with a single tweak. But it strains on clusters, and stumbles on the Bullet Cluster, where two clusters collided and their lensing mass sailed right through, cleanly separated from the visible gas. That looks like matter that barely interacts — exactly what a particle predicts and a gravity-tweak struggles to explain.

Where I currently lean

Toward a new particle — some flavor of cold dark matter. It is the only candidate that fits galaxies, clusters, lensing, and the early-universe data all at once. And to be clear, dark matter is a separate puzzle from the force accelerating the expansion; they share the word "dark" and nothing else.

What nags me: we have never caught a single particle directly, and decades of looking is a long silence.

What would change my mind: a confirmed direct detection would settle it overnight. So would a modified-gravity theory that explains the Bullet Cluster and the early-universe pattern without hand-waving. Either way, whatever this stuff is, its gravity is part of what shapes the expanding cosmos we live in.