When Vera Rubin measured the spin of galaxies, she found their outer stars moving so fast that visible matter alone could not hold them in place — one of the clearest early signs that most of every galaxy is made of something nobody has ever seen.

Beginning with the Andromeda galaxy in the late 1960s, the astronomer Vera Rubin and her colleague Kent Ford measured how fast stars and gas clouds orbit at different distances from a galaxy’s centre. They expected the outer material to move slowly. It did not. In Andromeda, and then in galaxy after galaxy, the orbital speed stayed high all the way to the edge of what they could measure. The visible stars, gas and dust could not supply enough gravity to hold matter moving that fast in place.

Rubin and Ford published their Andromeda result in 1970, in a paper in the Astrophysical Journal. Over the following decade they extended the work, and by 1980 had measured the same pattern across twenty-one spiral galaxies. The consistency was the point. One odd galaxy could be explained away. Twenty-one could not.

What the rotation curves showed

The expectation came from the solar system. Mercury orbits the Sun far faster than Neptune, because almost all the mass sits in the middle and the pull weakens with distance. A galaxy, where most of the visible light is concentrated toward the centre, was expected to behave broadly the same way: stars near the edge should lag well behind stars near the core.

A plot of orbital speed against distance is called a rotation curve. For a system following that expectation, the curve rises and then falls away. What Rubin and Ford found instead was a curve that rose and then stayed roughly level, far out into each galaxy’s faint outskirts.

A flat curve carries a specific implication. If outer stars orbit as fast as inner ones, there must be far more mass at large radii than the starlight accounts for, pulling on them. The discrepancy was not small. The galaxies behaved as though they held something like five to ten times more mass than everything visible in them combined.

Rubin and Ford were not the first to notice missing mass

Rubin’s rotation curves are often described as the first evidence of dark matter. That is a simplification worth correcting.

In 1933, the astronomer Fritz Zwicky studied galaxies moving within the Coma cluster and found them travelling too fast for the cluster’s visible mass to hold together. He inferred a large amount of unseen material and used the term dunkle Materie, dark matter. In 1939, Horace Babcock measured the rotation of Andromeda and found its outer regions implied an unexpectedly high ratio of mass to light. From the 1950s, radio astronomers tracing neutral hydrogen extended rotation curves beyond the visible disc and saw them stay high.

What Rubin and Ford added was not the first hint. It was something harder to set aside: clean optical rotation curves, measured the same way across many ordinary spiral galaxies, all showing the same thing. Their work is better understood as the evidence that moved the missing-mass problem from a recurring curiosity to a problem mainstream astronomy had to take seriously.

What the curves do not settle

It is worth being precise about what a rotation curve does and does not establish.

The observation is robust. Outer regions of spiral galaxies orbit faster than the visible matter can account for. That has been confirmed many times over and is not seriously disputed.

The interpretation is a separate question. The leading explanation is dark matter: a form of matter that does not emit or absorb light and makes itself known through gravity. On this account, every spiral sits in an extended halo of it. The case has grown well beyond rotation curves. Gravitational lensing, the pattern of the cosmic microwave background, the way galaxy clusters such as the colliding pair known as the Bullet Cluster behave, and the large-scale arrangement of galaxies all fit a universe containing several times more dark matter than ordinary matter.

Two caveats still belong in any honest account. Dark matter has not been detected directly. Decades of increasingly sensitive experiments built to catch a dark matter particle have so far returned nothing. And the rotation curves on their own admit an alternative. In 1983 the physicist Mordehai Milgrom proposed that gravity itself behaves differently at the very low accelerations found in galactic outskirts, an idea known as modified Newtonian dynamics, which reproduces flat rotation curves without invoking unseen matter. Most astronomers favour dark matter, largely because modified-gravity models struggle with the cluster and cosmological evidence that dark matter handles well. But the rotation curves alone do not decide between the two.

Where it stands

Rubin died in 2016. In 2019 the large survey telescope then under construction in Chile was renamed in her honour. The Vera C. Rubin Observatory released its first images in June 2025 and has since begun its principal task, a ten-year survey of the southern sky. Among its goals is mapping the distribution of dark matter across the sky through weak gravitational lensing, the slight distortion of background galaxies by intervening mass.

The situation Rubin’s work helped create is, in one sense, stable. The discrepancy between how galaxies move and how much matter can be seen in them is established beyond reasonable doubt. What that discrepancy is remains the open question. A telescope now carries her name while the thing her measurements pointed to has still not been identified.