In our solar system, almost all the mass is in the Sun, so the farther a planet is from it, the more slowly it orbits. Mercury races along at about 47 km/s; Neptune crawls at about 5 km/s. Astronomers expected galaxies to work the same way. Most of a spiral galaxy's visible light comes from its bright center and inner disk, so stars far out past the bright disk should orbit more slowly.
A graph of orbital speed against distance from the center is called a rotation curve. In the 1970s, Vera Rubin and Kent Ford measured rotation curves for Andromeda and many other spirals, using redshifts and blueshifts of the gas on either side. The curves didn't drop. They stayed flat — stars far out at the edges orbit about as fast as stars close in, and gas beyond the visible edge does too.
The dashed line is what the galaxy's visible stars and gas predict. The dots are what's actually measured. Add an invisible halo of mass until the prediction matches the data.
Here's a way to measure the missing mass without any big equations. Pick a distance r₀ at the edge of the bright disk, where nearly all the visible matter is already inside the orbit. If there were nothing else out there, then farther out, orbital speed would drop like this:
And since the mass needed to hold stars in orbit grows with speed squared, you can compare what's actually there to what the visible matter provides:
| Step | Worked example |
|---|---|
| Data | At the edge of the bright disk (r₀ = 15,000 ly), stars orbit at 200 km/s. At 60,000 ly — four times farther out — they still orbit at 200 km/s. |
| Expected | v = 200 × √(15,000 ÷ 60,000) = 200 × √0.25 = 200 × 0.5 = 100 km/s |
| Ratio | (200 ÷ 100)² = 2² = 4 — there's 4 times as much mass inside 60,000 ly as the visible matter provides |
| Dark share | If the total is 4 parts and visible matter is 1 part, then 3 of 4 parts — 75% — is dark matter |
Real galaxies give the same kind of answer. The Milky Way's rotation curve stays flat at roughly 230 km/s far beyond the Sun, and its total mass, most of it in a dark halosurrounding the disk, is several times the mass of all its stars and gas. (That's the halo from 5.1.1 — it's much bigger and heavier than its old stars suggest.)
One strange measurement might be a mistake. What makes dark matter convincing is that completely different methods, on completely different scales, all point to the same thing.
Four independent lines of evidence — different methods, different scales, same answer. Click each one.
A small number of scientists explore a different idea: that gravity itself works differently over huge distances. That's how science should work — but so far, evidence like the Bullet Cluster, where the mass and the normal matter ended up in different places, is very hard to explain without dark matter.
By the 1990s, astronomers expected gravity to be gradually slowing down the expansion of the universe. Two teams set out to measure how much, using Type Ia supernovae — white dwarfs that blow up when they pass the 1.4-Sun limit you met in Activity 4.2.2. Because they all explode at about the same peak brightness, how bright one looks tells you how far away it is, even billions of light-years out.
In 1998, both teams got the same surprise: the most distant supernovae were fainter than expected — farther away than a slowing expansion allows. The expansion isn't slowing down. For the last several billion years, it has been speeding up. Whatever is pushing it is called dark energy. The discovery earned Saul Perlmutter, Brian Schmidt, and Adam Riess the 2011 Nobel Prize in Physics.
Put all the evidence together — rotation curves, clusters, lensing, supernovae, and the afterglow of the Big Bang itself — and you get today's recipe for the universe:
What the universe is made of today. Click a slice.