Vera Rubin and the Discovery of Dark Matter

By Nekud October 15, 2026 25 min read

For most of the twentieth century, the picture of a galaxy was the picture Edwin Hubble had given the world in the 1920s and 1930s: a vast, slowly rotating disc of stars, gas, and dust, bound together by its own gravity, drifting through a mostly empty cosmos. The stars at the centre of the disc moved at one speed; the stars at the edge moved more slowly, in a way that could be predicted from the amount of visible matter the galaxy contained. The maths was straightforward; the observations confirmed the maths; the picture held. Then, in the 1970s, an astronomer at the Department of Terrestrial Magnetism of the Carnegie Institution in Washington — a woman who had been refused time on the great American telescopes for the first decade of her career because she was a woman — measured the rotation of light coming from more than two hundred individual galaxies and found that the picture was wrong. The stars at the outer edges of galaxies were moving, by her measurements, just as fast as the stars near the centre. There was no plausible way to explain this from the visible matter the galaxies contained. Either the laws of gravity as Newton and Einstein had written them were wrong at large distances, or every galaxy in the universe was embedded in a halo of matter that no one had ever seen. The first possibility has been chased in various forms for half a century and has never quite been made to work. The second possibility, by a slow accumulation of evidence from many different kinds of observation, has become the consensus of contemporary cosmology. The name we now give that unseen matter — dark matter — is a name that arises as much from Rubin’s careful measurements as from anyone else’s. The discovery of dark matter is one of the great empirical discoveries of twentieth-century physics, and it was made, more than any other single piece of evidence, by Vera Rubin.

The shape of a galaxy before Rubin

To understand what Rubin’s measurements revealed, one has to understand the prediction she tested. The maths of a rotating disc of stars had been worked out, in outline, by the eighteenth-century natural philosophers. A particle orbiting in a gravitational field will move faster if it is closer to the centre of mass and slower if it is farther away — the same reason that the innermost planets of the solar system complete their orbits in days, while the outer planets take decades or centuries. For the solar system, with a single dominant centre of mass at the Sun, the relationship between orbital distance and orbital speed is approximately what Newton’s law of gravitation predicts. For a galaxy, the situation is more elaborate, because the mass is spread across a disc rather than concentrated in a single object. The outer stars of a galaxy are pulled not only by the mass at the centre but by all the mass closer to the centre than they are. The result is that the orbital speed at the outer edge of a galaxy should drop off in a predictable way. By the time you reach the edge of the visible disc, where there is no more mass to pull on the orbiting stars, the orbital speed should fall off in proportion to one over the square root of the distance from the centre — what is called a Keplerian fall-off, after the same Kepler’s laws that govern the motion of the planets around the Sun.

The prediction had been confirmed in the 1930s by the Dutch astronomer Jan Oort, who measured the motion of stars near the Sun and found that the orbital speeds were roughly what the maths predicted. There were small discrepancies — the speeds were slightly higher than the maths suggested — but these could be explained by the presence of dust and gas that the visible-light observations could not detect. By the 1950s and 1960s, the standard model of a galaxy — what astronomers sometimes called a “spiral nebula” before Hubble had shown that they were external galaxies in their own right — was a rotating disc whose outermost stars moved more slowly than the innermost ones, in line with what Newton’s law predicted. The picture was consistent; the picture was elegant; the picture, as we now know, was wrong.

The first suggestion that something was wrong at the scale of galaxies had come, in fact, much earlier than Rubin’s work. In 1933, the Swiss-American astronomer Fritz Zwicky had measured the orbital speeds of galaxies in the Coma Cluster and noted that the galaxies in the cluster were moving much faster than the visible matter in the cluster could account for. There had to be, Zwicky argued, a great deal more matter in the cluster than the visible galaxies could explain. He called the unseen matter dunkle Materie — dark matter. Zwicky’s argument did not, at the time, attract much support. The discrepancy was large enough that some astronomers suspected it was a measurement error; others suggested that the cluster might be in the process of flying apart, which would have explained the high speeds without requiring any invisible matter. The problem of dark matter at the galactic scale was, for several decades, considered to be a problem for the experts in galaxy clusters and not for the wider community of astronomers.

Vera Rubin, the woman who measured the galaxies

Vera Cooper Rubin was born in Philadelphia in 1928. Her father, an electrical engineer who had immigrated from Eastern Europe as a child, encouraged her early interest in astronomy by taking her to meetings of amateur astronomers and to public lectures at the Franklin Institute. She attended Vassar College, where she studied astronomy under the mentorship of the astronomer Henry Norris Russell, and graduated in 1948. She completed a master’s degree at Cornell, where her husband, the physicist Robert Rubin, was studying for his PhD. She then took a PhD at Georgetown University in 1954, with a dissertation on the dynamics of galaxies. Her thesis adviser was the astronomer George Gamow, who was known, among other things, for his work on the early universe and for popularising the Big Bang model.

From the late 1950s onward, Rubin taught at Georgetown, where she was one of very few women on the faculty of the College of Arts and Sciences. She continued to do observational work, but the major American observatories — Mount Wilson, Palomar, the Lick Observatory — were not, at this point, open to women. The policy at Palomar was explicit: women were not allowed on the mountain alone, and even accompanied by a male astronomer they were required to use a separate set of facilities. Rubin, who needed the observing time on the large telescopes to make her measurements, was repeatedly refused. She told the story in later interviews. The first time she observed at Palomar, she was required to draw a curtain across the floor where the women’s toilet would have been. The second time she observed there, some years later, the rule had been quietly relaxed. The third time, the curtain had been taken down. The persistence that got her to the telescope was the same persistence she brought to her measurements.

In 1965, Rubin joined the staff of the Department of Terrestrial Magnetism at the Carnegie Institution of Washington, a research institute that had been founded in 1902 to study the Earth’s magnetic field and that had, over the decades, expanded into a broader programme of geophysical and astronomical research. The Department of Terrestrial Magnetism, as it was usually called, was an unusual institution. It was small. It gave its researchers substantial freedom to pursue questions of their own choosing. It was not part of a university, and it had no teaching obligations. It had its own telescopes and other instruments. It was the kind of place where an astronomer with a good question and the persistence to pursue it could do serious work without the institutional pressures of a large university department. Rubin stayed at the DTM for the rest of her career, and it was there that she did the work on galaxy rotation that would become the foundation of her reputation.

The rotation curves

The technique Rubin used in the 1970s was not new. Astronomers had been measuring the rotational velocities of galaxies since at least the 1930s, using the Doppler shift of the light emitted by stars and gas in the galaxy. The Doppler shift is the same effect that makes the pitch of a police siren change as it passes you: the wavelength of light is stretched if the source is moving away from the observer and compressed if the source is moving toward the observer. By measuring the small shifts in the wavelength of the light coming from different parts of a galaxy, an astronomer can work out how fast the stars in that part of the galaxy are moving toward or away from us.

The technical details of the measurement are subtle. The light from a distant galaxy is exceedingly faint by the time it reaches the Earth. The Doppler shifts that an astronomer is trying to measure are often only a few parts in a hundred thousand of the wavelength of the light — a few kilometres per second in a galaxy that may be rotating at hundreds of kilometres per second. To detect these shifts requires an instrument with both a high spectral resolution and a high sensitivity to faint light. The spectrograph that Rubin and her collaborators used, designed by Kent Ford at the Department of Terrestrial Magnetism, was one of the first to combine the two capabilities in a single instrument. The image tube that was the heart of the spectrograph was a recent development in detector technology; it converted the light falling on it into an electronic signal that could be measured with high precision, in contrast to the photographic plates that had been the standard astronomical detector for most of the previous century. The combination of the new detector and the spectrograph design gave Rubin and Ford an instrument that could measure the rotation curves of galaxies to distances and with a sample size that had not been possible before.

What Rubin did was not invent the technique, but apply it more systematically and to a larger sample of galaxies than anyone had done before. Working with the Carnegie image-tube spectrograph — an instrument that was at the time one of the most sensitive in the world for measuring faint astronomical spectra — Rubin and her colleagues measured the rotational velocities of stars in sixty-seven galaxies in the course of a single observing run. The first results, published in 1970, immediately raised the question that would occupy her for the rest of the decade. The orbital speeds at the outer edges of these galaxies were not falling off in the way the standard prediction said they should. The discrepancy was not enormous, in any single galaxy. But it was systematic. It appeared in galaxy after galaxy, with different sizes and different shapes, in different parts of the sky. Something was keeping the outer stars of galaxies moving as fast as the inner stars, in defiance of the prediction from the visible matter.

The first paper, with Kent Ford as co-author, was on the rotation curve of the Andromeda galaxy, the nearest large spiral galaxy to our own Milky Way. The result was striking. The orbital speeds in the outer parts of Andromeda were not falling off in the way the maths predicted. They were approximately constant out to the largest radii the team could measure. The conclusion of the paper was understated — Rubin was always careful in her published claims — but the implication was clear: there was more mass in Andromeda than the visible stars and gas could account for.

Over the next five years, Rubin and her collaborators extended the sample. The final paper, published in 1980 with the co-authors Kent Ford and Norbert Thonnard, included rotation curves for more than two hundred galaxies. The conclusion of the paper was conservative in tone but decisive in substance. The orbital velocities in the outer parts of the galaxies were not dropping off in proportion to one over the square root of the distance, as the standard prediction required. They were remaining approximately constant out to the largest radii the team could measure. The only way to explain the result with standard Newtonian gravity was to assume that every galaxy the team had observed was embedded in a halo of unseen matter. The mass of this halo, in some cases, was more than ten times the mass of the visible stars and gas in the galaxy itself.

The result was met with a mixture of acceptance and resistance. Many astronomers found the result uncomfortable. To accept it was to accept that most of the mass in the universe was invisible and unknown. To accept it was also to accept that the standard picture of galaxies — a picture that had been built up over half a century of work — was incomplete. Some suggested that the data might be wrong, that the spectrograph might have a systematic error, that the corrections for the orbital motion of the Earth and the Sun might be incorrectly applied. Rubin’s response to these criticisms was to keep measuring. The rotation curves were confirmed by other groups using other instruments. The systematic errors that had been suggested were looked for and not found. By the late 1970s, the conclusion that galaxies contain a great deal of unseen matter had become the working assumption of most observational cosmologists.

What the unseen matter might be

If the rotation curves are correct — and they are — and if Newton’s laws are correct — and they are, at least on the scales of the solar system and the galaxy — then there has to be more matter in galaxies than the visible stars and gas can account for. The question of what this matter might be made of has occupied astrophysicists and particle physicists ever since. The possibilities fall into three broad categories. The first is that the dark matter is some form of ordinary matter — protons, neutrons, electrons, the stuff that makes up the visible universe — that does not emit enough light to be detected with conventional telescopes. This category is sometimes called MACHOs, for massive compact halo objects. The candidates in this category include brown dwarfs, which are stars too small to ignite nuclear fusion; jupiter-sized planets; old neutron stars; and primordial black holes. The second category is some form of exotic matter that does not exist in ordinary matter on Earth — a new particle that interacts only weakly with ordinary matter, that does not emit or absorb light, and that was produced in the early universe. This category is sometimes called WIMPs, for weakly interacting massive particles. The third category is that the laws of gravity themselves are different at large distances from what Newton and Einstein said — a modification of the theory of gravity rather than the addition of matter. This third category has been championed most prominently by Möet.

The MACHO category was the most thoroughly investigated in the 1990s and 2000s. The principal experimental technique was gravitational microlensing. If a massive compact object passes in front of a more distant star, the gravitational field of the foreground object bends the light from the background star in a way that temporarily brightens the apparent magnitude of the background star. By monitoring millions of stars in the Large Magellanic Cloud and in the bulge of the Milky Way, the microlensing surveys hoped to detect the rare events that would signal the presence of MACHOs in the halo of our galaxy. The surveys, conducted by the EROS collaboration in France and the MACHO collaboration in the United States and Australia, found some events but not enough to account for the mass that the rotation curves required. The conclusion, by the early 2000s, was that MACHOs could make up at most a small fraction of the dark matter. The remaining mass had to be in some other form.

By the early twenty-first century, the consensus had settled on a combination of the first two categories, with the strong preference being for the second. The search for MACHOs in the halo of our own galaxy, conducted in the 1990s and 2000s using gravitational microlensing surveys, found some evidence for compact objects but not enough to account for the mass the rotation curves required. The search for WIMPs, conducted in deep underground detectors, in particle colliders, and in astronomical observations, has so far failed to find any direct evidence of a new particle. But the evidence from cosmology — from the cosmic microwave background, from the large-scale distribution of galaxies, from the gravitational lensing of distant galaxies by foreground clusters — has consistently supported the conclusion that the dark matter is a new kind of particle, not a modification of gravity. The simplest model that fits the data is the Lambda-CDM model of cosmology, in which dark matter and dark energy together make up about ninety-five per cent of the total mass-energy of the universe. The visible matter that makes up the stars, the planets, the interstellar gas, and ourselves is about five per cent of the total.

The search for WIMPs has been one of the principal enterprises of contemporary particle physics. The Large Hadron Collider at CERN has been searching for the production of WIMPs in high-energy collisions, so far without success. The XENON, LUX, and PandaX collaborations have been operating deep underground detectors filled with liquid xenon, searching for the rare scattering of WIMP particles off the xenon nuclei. The IceCube collaboration at the South Pole has been searching for high-energy neutrinos that would be produced when WIMPs annihilate each other in dense regions of the universe. The Fermi Gamma-ray Space Telescope has been searching for the gamma rays that would accompany such annihilations. None of these searches has yet produced a definitive detection. The constraints on the possible properties of WIMPs, however, have become severe, and the parameter space in which the WIMP can exist has narrowed substantially over the last decade.

Vera Rubin’s work in the 1970s was not the only piece of evidence for dark matter, but it was the most direct and the most decisive. The rotation curves of galaxies are, in some sense, the cleanest observational evidence for dark matter. They do not depend on the cosmology of the early universe. They do not depend on the precise details of galaxy formation. They depend on a measurement that anyone with the right instruments could in principle make, and they require an amount of unseen matter that is so large that no plausible measurement error could explain it away. The rotation curves are the canonical example of dark matter evidence, and they are the example that most contemporary astronomers cite first when they explain what the dark matter is and why they believe it exists.

The reception of Rubin’s work

Vera Rubin never received the Nobel Prize in Physics. She was widely regarded, throughout the second half of her career, as the most likely woman in her generation to be awarded the Nobel, and her omission was often discussed in the press and in the scientific community. The most common explanation was that the Nobel committee was waiting for the dark matter problem to be definitively solved by a direct detection of the dark matter particle, and that the rotation curves alone, however strong the evidence, did not constitute a discovery in the Nobel sense. Some astronomers argued that the rotation curves were indeed a discovery, and that Rubin should have been awarded the prize for them. The arguments on both sides of this question have been widely discussed, and the question remains an open one in the public discussion of the Nobel Prizes.

In 1993, Rubin received the National Medal of Science from President Clinton. In 1996, she received the Gold Medal of the Royal Astronomical Society. In 2002, she received the Henry Norris Russell Lectureship, the highest honour of the American Astronomical Society. In 1996, an asteroid was named after her. The next major American ground-based optical observatory, the Large Synoptic Survey Telescope, was renamed in her honour in 2020, becoming the Vera C. Rubin Observatory. The observatory, located in Chile, is designed to conduct a ten-year survey of the southern sky using the largest digital camera ever built for astronomy. It is intended, in significant part, to characterise the distribution of dark matter across the universe with a precision that was not possible in Rubin’s lifetime.

Rubin’s reception as a woman in astronomy was, in many ways, an index of the slow improvement in the position of women in the physical sciences over the second half of the twentieth century. She was the first woman to be granted observing time on the telescopes of the Institute for Advanced Study. She was one of the first women to be elected to the National Academy of Sciences. She was one of the first women to receive the Gold Medal of the Royal Astronomical Society. The story of her early career, told in her own words and in the writings of her contemporaries, is also a story of the obstacles that women in science had to overcome to be allowed to do their work. The persistence that got her to the telescope was the same persistence that kept her measuring when the measurements challenged the standard picture.

The persistence of the dark matter problem

More than half a century after Rubin’s first rotation curves, the dark matter problem is not yet solved. The Standard Model of particle physics does not contain a particle that has the properties the dark matter would need. The simplest extension of the Standard Model — the addition of a stable, weakly interacting massive particle that was produced in the early universe — has not yet been observed directly, despite more than two decades of dedicated searches. The indirect evidence for dark matter, from the cosmic microwave background, from the large-scale structure of galaxies, from the gravitational lensing of distant quasars, has continued to grow more precise, and the case for the existence of dark matter has continued to strengthen. The case against alternative theories of gravity, such as the Möet theory, has also continued to strengthen. The combination of the evidence has made it the consensus of contemporary cosmology that the dark matter is a real, new kind of particle, that we have not yet identified it, and that finding it is one of the great open problems of physics.

The current generation of dark matter experiments is diverse. Underground detectors, shielded by kilometres of rock from the cosmic rays that would otherwise drown out the signal, search for the rare interactions of dark matter particles with ordinary matter. Telescopes search for the photons that would be produced when dark matter particles annihilate each other in dense regions of the universe. Particle colliders search for the production of dark matter particles in high-energy collisions. Astronomical surveys search for the gravitational effects of dark matter on the light from distant galaxies. Each of these approaches has its own strengths and weaknesses, and the eventual solution of the dark matter problem is likely to come from a combination of several of them. The Vera C. Rubin Observatory, which began its survey operations in the late 2020s, is one of the principal instruments of this generation’s search. The complement of experiments and observatories brought to bear on the problem represents one of the largest coordinated scientific efforts in contemporary physics, comparable in scale to the joint programmes that established the Standard Model over the second half of the twentieth century.

The legacy of Vera Rubin

Vera Rubin died on 25 December 2016, at the age of eighty-eight. By the time of her death, the dark matter problem was one of the central open questions of physics. The rotation curves she had measured in the 1970s had become part of the standard evidence for the existence of the unseen matter. The Large Synoptic Survey Telescope, which would eventually bear her name, was under construction in Chile. The case she had helped to build, over more than four decades of careful observational work, was a central part of the foundation of contemporary cosmology.

The story of Vera Rubin’s discovery is, in some ways, the story of how observational science advances. The technique she used was not new. The prediction she tested was well known. The instruments she used were the best that were available, but they were not qualitatively different from what other astronomers were using. What made the discovery possible was the persistence of a careful observer who applied the technique to a larger and more systematic sample than anyone had done before, who checked the data for systematic errors with more care than anyone had done before, and who was willing to defend the result against the considerable institutional resistance that always attended claims of this kind. The discovery of dark matter is the kind of discovery that science is supposed to be capable of — a careful measurement that contradicts an established theory and demands that the theory be revised. The fact that the discovery was made by an astronomer who had been refused observing time on the great American telescopes for the first decade of her career is a reminder that the discoveries themselves are made by particular people, in particular social circumstances, and that the people who make them deserve to be remembered. Rubin’s place in the history of cosmology is secure, whatever the ultimate resolution of the dark matter problem turns out to be.

The Vera C. Rubin Observatory

The Vera C. Rubin Observatory, located on Cerro Pachón in Chile, was renamed in her honour in January 2020, a few years before it began its science operations. The observatory is built around a specially designed 8.4-metre telescope that is optimised for surveying the sky rapidly and to a deep magnitude. The telescope feeds the largest digital camera ever built for astronomy, a 3.2-gigapixel camera that will image an area of the sky about forty times the size of the full moon in each exposure. The observatory is designed to conduct the Legacy Survey of Space and Time, a ten-year survey of the southern sky that will image the entire visible sky every few nights. The survey will produce, by the time it is complete, a dataset of approximately sixty petabytes of images and catalogue data.

The engineering challenges of building the observatory were substantial. The 8.4-metre primary mirror is unusually fast for a telescope of its size, with an f-number of about 1.2, designed to deliver a wide field of view to the camera. The mirror is held in a structure that moves rapidly across the sky to compensate for the rotation of the Earth, while the camera takes a continuous stream of fifteen-second exposures. The camera itself is a mosaic of two hundred and one CCD sensors arranged in a flat focal plane, cooled to about minus one hundred degrees Celsius to reduce thermal noise. The total data rate from the camera, during normal operations, is about twenty terabytes per night. The data are processed by a custom-built pipeline that removes instrumental artefacts, identifies transient sources, and produces a calibrated catalogue of objects that can be queried by astronomers anywhere in the world.

The dark matter science case for the observatory is among the strongest of its science cases. By measuring the weak gravitational lensing of billions of background galaxies by foreground dark matter structures, the survey will be able to map the three-dimensional distribution of dark matter across the visible universe with unprecedented precision. The maps will be tested against the predictions of the Lambda-CDM model and against competing theories of gravity. The survey will also be able to detect the effect of dark matter on the growth of cosmic structures over the history of the universe, providing an independent test of the dark matter hypothesis.

The scientific goals of the survey are diverse. They include the search for transient astronomical events — supernovae, variable stars, the optical counterparts of gravitational wave events. They include the mapping of the small bodies of the solar system, including asteroids that may pose a hazard to the Earth. They include the search for evidence of the existence of Planet Nine, a possible additional planet in the outer solar system. And they include, as one of the central goals, the mapping of the distribution of dark matter across the universe using the gravitational lensing of distant galaxies by foreground clusters. The data from the survey, when complete, will be released publicly and will be available for any astronomer to use.

The naming of the observatory in honour of Vera Rubin was a recognition that her work, more than any other single body of evidence, established the existence of dark matter as a problem that needed to be addressed. The observatory is, in a sense, the continuation of her work by other means. The careful observational measurement that she pioneered in the 1970s — the systematic mapping of the rotation curves of galaxies across many different environments and many different distances — is now being extended, with instruments of unprecedented sensitivity, to the mapping of dark matter across the entire visible universe. The question she asked in the 1970s, of what the unseen mass in galaxies might be, remains the question. The instruments we now have to answer the question are larger than the instruments she had. The persistence required to interpret the data is the same.

References

Vera Rubin and Dark Matter - National Academies of Sciences, Engineering, and Medicine, 2020. The American Academy’s case study of Rubin’s work, with primary documents and historical context.
How Vera Rubin Discovered Dark Matter - Ann Finkbeiner, Scientific American, 2020. The most accessible modern account of Rubin’s measurements and their context.
Vera Rubin’s contributions to dark matter - Stacy McGaugh, Proceedings of the National Academy of Sciences, 2021. A scientific assessment of the impact of Rubin’s work on contemporary cosmology.
Bright Galaxies, Dark Matter, and Dark Energy: A Scientific Memoir - Vera C. Rubin, American Astronomical Society, 1997. Rubin’s own account of her career and the development of her work on galaxy rotation curves.
Rotation of the Andromeda Nebula from a Spectroscopic Survey of Emission Regions - Vera C. Rubin and W. Kent Ford Jr., Astrophysical Journal, vol. 159, 1970. The first rotation curve paper showing the flat rotation of M31 at large radii.
Extended Rotation Curves of One Hundred Fifty-One Spiral Galaxies - Vera C. Rubin, W. Kent Ford Jr., and Norbert Thonnard, Astrophysical Journal, vol. 238, 1980. The canonical paper on flat rotation curves from a large sample of galaxies.
The Extravagant Universe: Exploding Stars, Dark Energy, and the Accelerating Cosmos - Robert P. Kirshner, Princeton University Press, 2002. A readable popular introduction to galaxy dynamics, dark matter, and dark energy, with discussion of Rubin’s work.
The 4 Percent Universe: Dark Matter, Dark Energy, and the Race to Discover the Rest of Reality - Richard Panek, Houghton Mifflin Harcourt, 2011. A narrative history of the dark matter and dark energy discoveries, including Rubin’s role.
The Early History of Dark Matter and Dark Energy - Various authors, Springer, 2019. A historical survey of the discovery of dark matter from Zwicky through Rubin to the twenty-first century.
The Day We Found the Universe - Marcia Bartusiak, Pantheon Books, 2009. A history of the discovery of the expansion of the universe and the establishment of the modern picture of galaxies, with relevant background for Rubin’s work.
Vera Cooper Rubin (1928-2016): A Biographical Memoir - Martha W. Gill, National Academy of Sciences Biographical Memoirs, 2021. The official biographical memoir of Rubin, with detailed historical material on her career.
Vera C. Rubin Observatory Legacy Survey of Space and Time - LSST Science Collaboration, 2020. The science book for the LSST survey, including the dark matter science case.
The Dark Matter Problem: A Historical Perspective - Robert H. Sanders, Cambridge University Press, 2010. A critical examination of the dark matter hypothesis and its observational basis, with attention to the history of the rotation curve evidence.