In the 1970s astronomer Vera Rubin was observing the Andromeda galaxy and discovered something very curious. Andromeda is a spiral galaxy, like our own, and spins like a pinwheel. The “spinning” is comprised of all the individual stars (and gas and dust, but the stars are what we can see) revolving around all the mass within their orbit. If you run the numbers, as stars get further away from the galactic center they should revolve more slowly. The relationship between distance from the galactic center and each star’s velocity is called a galactic rotation curve. Rubin and others predicted the curve should be decreasing in a linear fashion (after an initial rise because of the increase in mass at the galactic center).

What Rubin found, however, was that the rotation curve of Andromeda increased at first as expected but then did not decrease with distance but remained largely flat. This difference between prediction and observation was a genuine anomaly and required an explanation. The results were verified with other large spiral galaxies, and yes, they all have flat rotation curves. The stars on the outskirts of these galaxies, according to Newton, should be flying away. They are moving to fast to be held by the gravity of the galaxy they are orbiting.

Unless…

Perhaps there is more mass in the galaxy than we can observe. There is matter that is not giving off light like stars or even reflecting or glowing from the light of stars like gas clouds. There must be matter we cannot see, some dark matter. How much dark matter would it take to explain the observed rotation curves? Quite a lot – about six times the mass of the stuff we can see. If true, then some 84% of the matter in the universe is dark matter. And we do not know what dark matter is – we don’t know what most of the universe is made of.

This is a great scientific story, earning Vera Rubin a place among the scientific greats (but apparently not a Nobel Prize). But how confident are we that Dark Matter is the answer to the anomaly of the galactic rotation curves? Another approach to solving this anomaly is called Modified Newtonian Dynamics, or MOND. The overall idea here is that the equations used by Rubin and others are not the real gravitational equations (and here we are not referring to general relativity), but are approximations. They work precisely on the scale of a solar system, but on the scale of a galaxy tiny factors in the equations become significant and can explain the galactic rotation curves.

So either we invent a new form of matter or we fudge the equations to make them work. Both approaches are hypotheses, they are both proposed as a way to resolve something we don’t know, and in part are arguments from ignorance. What we would need to confirm either hypothesis is positive evidence that it is actually true. If, for example, we discovered the dark matter particle, with all the right properties, that would clinch the dark matter hypothesis. For MOND it would be difficult because by definition it operates only at galactic scales, but perhaps some precise experiments might validate any particular MOND theory.

But also, for either approach, we can make further observations about how the large scale universe moves and see which hypothesis it fits better. This has already happened, and the evidence is piling up on the dark matter side of the scale. The most impressive evidence, perhaps, is the bullet cluster. These are two galaxy clusters in the process of colliding. Galaxies are mostly empty space, so the stars mostly just pass by each other but do interact gravitationally. Gas clouds, however, have pressure. When the gas of the two clusters collided they significantly slowed down. The gas in these clusters actually represents most of the visible mass. So we have a situation where two galaxy clusters collide and most of the visible mass dramatically slowly down at the collision point. If dark matter is real, and it behaves the way we think it does, then it should have passed through with the stars.

Even though we cannot directly see dark matter we can see its gravity, such as its effect on the rotation curves of galaxies. We can also look at the gravity of very massive objects through gravitational lensing, the bending of light by gravity. Astronomers used gravitational lensing to map the gravity and therefore the mass of the bullet cluster – and most of the mass just kept on going, leaving the gas behind. That is a pretty solid confirmation of dark matter, and leaves MOND proponents with a serious problem.

While it is true that dark matter is inferred to exist and we have not confirmed it directly, that inference is very strong and based on multiple observations. It is also currently the best explanation for all these observations. But yes, there is a question mark next to dark matter until we actually discover the dark matter particle. That is why it is an area of active research.

Still, new MOND theories keep cropping up. Recently an astronomer proposed that relativistic frame dragging might explain the galactic rotation curves. Until this hypothesis can explain the Bullet Cluster, however, it is not likely to go anywhere.

As a side note, I was able to (mostly) resolve a question I had about the galactic rotation curves. The story I told above explains why the rotation curves are different than predicted without dark matter, and why stars are moving faster than they should. But I was never clear on why the curves were mostly flat. Is that a coincidence? I don’t like such coincidences in science, they give me the uneasy feeling we are (or at least I am) missing something. I was recently offered an explanation by Ken Walczak who works at the Adler Planetarium. It has to do with the distribution of Dark Matter throughout galaxies and the mathematical effect that has on the rotation curves. If you map the rotation curves of the halo of dark matter they will be flat over a very large scale of mass, and since dark matter is most of the matter in a galaxy, the rotation curve of the galaxy is flat. Another way to look at this is that the distribution of dark matter means that the mass of a galaxy increases linearly with radius. The bottom line is – this is not a cosmic coincidence, it’s math.

Chances are, given the current evidence, dark matter is real. Unless a new better hypothesis comes along (and it’s not MOND), or someone makes an observation incompatible with dark matter, it will remain the leading contender. But the controversy is not likely to end until we actually discover what dark matter is.