A rule you already trust
In the last two guides we mapped the anatomy of the Milky Way — its disk, bulge and halo — and read the galaxy's history out of its oldest stars. Now we put the whole thing in motion. Everything in the disk orbits the centre: the Sun included, dragging Earth and us along at roughly 220 kilometres every second, completing one lap in about 230 million years. Ask a simple question about that motion and the answer cracks open one of the biggest puzzles in all of science.
The question is just this: how fast should a star move on its orbit, given where it sits? You already know the answer from the gravity rung. Kepler's third law, in the form Newton gave it, says the orbital speed is set entirely by how much mass lies inside the orbit. In the Solar System almost all the mass — the Sun — sits in the middle, so the planets thin out their speeds with distance: Mercury races, Neptune crawls. Speed falls off as one over the square root of radius. That graceful decline is the signature of a central, dominant mass, and we trust it because it predicts the planets to the digit.
Reading speed from starlight
Before we can be shocked, we have to actually measure how fast things orbit at each radius — to draw what astronomers call the rotation curve: a graph of orbital speed against distance from the galactic centre. We cannot stopwatch a star creeping across the sky; over a human lifetime it barely budges. Instead we use the trick you met in the light rung: the Doppler shift of spectral lines. Light from gas moving toward us is squeezed bluer, light from gas moving away is stretched redder, and the size of the shift tells us the speed along our line of sight — to a kilometre per second.
Stars are useful close in, but the disk's outer reaches are thin on bright stars and thick with cold hydrogen gas. The key tool there is the 21-centimetre radio line of neutral hydrogen — a faint hum that hydrogen atoms emit, which sails straight through dust that would block visible light. Radio astronomers from the 1950s onward used it to clock gas all the way to the disk's ragged edge and far beyond, in clouds well outside the glowing starlight. That reach is exactly what the puzzle needs: it lets us measure speed where there is barely any visible matter left to do the pulling.
There is real subtlety in turning these line-of-sight speeds into a clean curve. We sit inside the disk, sharing its rotation, so every measurement is a velocity relative to our own moving frame; untangling it requires knowing our distance to the galactic centre and our own orbital speed. For other galaxies, seen more or less edge-on from outside, the job is cleaner — one side of the disk sweeps toward us, the other away, and the rotation curve almost draws itself. The Milky Way and its neighbours tell the same story, which is why we trust it.
The curve that refuses to fall
Here is the shock. As you move out from the centre, the speed climbs at first — no surprise, because you keep enclosing more disk and bulge. But then, where the visible galaxy thins toward nothing and Kepler's gentle decline should set in, the curve simply does not fall. It flattens out and stays there, kilometre for kilometre, far beyond the last bright stars. The gas in the dim outskirts is orbiting just as fast as gas deep in the lit disk. By the Kepler rule we trust everywhere else, that is impossible unless there is far more mass out there, pulling, than anything our telescopes can see.
What Kepler predicts vs. what we see, far from the centre:
orbital speed v(r) = sqrt( G * M(<r) / r )
M(<r) = mass enclosed inside radius r
EXPECTED (visible matter only):
past the bright disk, M(<r) stops growing
-> v ~ 1 / sqrt(r) ... speed should DROP with distance
OBSERVED (the rotation curve):
v(r) ~ flat, ~200-250 km/s, far beyond the stars
-> for v to stay flat, M(<r) must keep rising as ~ r
Conclusion: unseen mass keeps piling up where light does not.How big is the gap? Add up everything that shines or that we can otherwise account for — stars, gas, dust, the lot — and it falls short of the mass the rotation demands by a factor of several. Roughly speaking, the luminous Milky Way appears to be embedded in something whose total mass is a handful of times larger. The visible galaxy, the part that took our whole subject to understand, turns out to be the minority partner. This was not a faint statistical hint; once the 21-centimetre data were in, the flat curves of galaxy after galaxy made it undeniable through the 1970s, above all in the careful work of Vera Rubin and Kent Ford.
An invisible halo
What kind of unseen mass would flatten the curve? Work backwards from the graph and it tells you its own shape. To keep the orbital speed constant as you go out, the enclosed mass must keep rising in step with radius — which means the extra stuff cannot be a disk like the stars. It must be a roughly spherical, puffed-up cloud that extends far beyond the bright disk, growing thinner outward but never quite ending. Astronomers call it the dark-matter halo: an enormous, near-invisible sphere of mass that the luminous Milky Way sits inside like a coin dropped into a beach ball.
It is worth being honest about a leap we just made. The rotation curve proves that something with mass is out there pulling — gravity is the only thing we are really measuring. Naming that something a 'halo of dark matter' is one hypothesis for what it is. It earns confidence not from this one curve alone but from a stack of independent clues all pointing the same way: the speeds of whole galaxies swarming in clusters, the bending of background light by foreground mass, and the pattern of ripples in the early universe's relic glow. We will meet those in the cosmology rungs. The flat rotation curve is simply the first, closest, most homely member of that family of evidence.
What it is — and what it is not
If the halo is made of matter, why have we never seen it? Because, the leading idea goes, it neither emits nor absorbs light at all — it is dark in the strict sense, touching ordinary stuff only through gravity. That alone rules out the easy answers. It is not just dim stars, cold gas, or rogue planets: those are ordinary matter, and we can total them up, and there is nowhere near enough. It is not a great cloud of black holes of the everyday kind, which careful searches have largely excluded. The favoured candidate is some new, non-baryonic particle — outside the familiar roster of protons and neutrons — that simply slips through everything, including detectors built to catch it. Decades of those experiments have so far come up empty, which is an honest and important part of the story.
There is also a respectable rival worth naming. Maybe nothing is missing at all, and instead our law of gravity is slightly wrong at the feeble accelerations found in a galaxy's outskirts. That is the idea behind modified Newtonian dynamics, and it fits the flat rotation curves of individual galaxies strikingly well — better, in places, than dark matter does without fine-tuning. But it struggles to explain the cluster-scale and early-universe evidence that dark matter handles naturally. Most astrophysicists therefore favour dark matter, while keeping the modified-gravity option honestly on the table. This is a live debate, not a settled verdict.
Whichever way it resolves, the rotation curve has already done something profound. We set out in this rung to learn the Milky Way as our home, and we found that the home is mostly made of an unseen something we measure but cannot name. The next guide turns inward instead of outward, to the very centre of all this rotation — the dense, dark heart where the stars whip around fastest of all, circling a supermassive black hole.