This article features work on the "Bullet" Cluster, two colliding clusters of galaxies about 3.4 billion light years away. By combining information on where galaxies are, where hot X-ray emitting gas is, and how the total mass of the cluster is bending light from even more distant galaxies, the original team claimed that the only explanation that could simultaneously explain all the observations was that dark matter actually exists.
Dark matter is quite mysterious. When we look at other galaxies and clusters of galaxies, the stars and gas in those systems doesn't move like we think it should according to Newton's Law of Gravity. In most cases, the star's movements can be explained if there is matter we cannot see that obeys the laws of gravity, but otherwise doesn't interact with normal matter. Standard physics doesn't predict such matter, but some new physics hypothesizes that particles with these characteristics could exist. This is what we call dark matter.
Despite not knowing what dark matter actually is, including these basic properties into calculations of the formation and evolution of the Universe results in predictions that are pretty close to reality, although not exactly right. Dark matter has not yet been detected or made in physics laboratories, however. Until it is, I won't feel comfortable saying that dark matter has been "proven." I am pretty sure dark matter does exist, but I want proof.
In the 1980s, another hypothesis was put forward to explain the odd motions of stars and gas in other galaxies. This hypothesis is called "MOND," or MOdified Newtonian Dynamics. MOND says, in short, that gravity works just like Isaac Newton claims it does, getting weaker as you get further away from matter, at least until a certain point. After that point, gravity falls off more slowly than Newton would predict.
It is true that MOND can explain some parts of the observations behind dark matter as well as dark matter can. And about a decade ago, it was noticed that the Pioneer spacecraft leaving our solar system are slowing down at a different rate than Newton's Law would predict (although this signal is so small that there are lots of different tiny effects, like acceleration of the spacecraft due to the sun's light warming some parts more than others, and the extra "push" the craft gets when it sends radio waves back to Earth -- extraordinarily tiny effects that are hard to calculate exactly).
The problem most astronomers have with MOND is not (as some MOND supporters claim) that it would get rid of dark matter. The problem is that the MOND is continually being tweaked to explain new observations. And, as I've said before, no hypothesis can become a mature theory until it makes a prediction that can be tested. Explaining things in hindsight is not acceptable in that regard.
For MOND to be seriously considered by a large number of astronomers, it needs to make a prediction about something we can see that cannot possibly be due to dark matter or anything other than MOND. And then we have to go look and test that prediction. If the prediction holds true, MOND will get a big boost. But, if it is found not to hold true, then MOND supporters need to re-think their theory.
Vocal MOND supporters have pointed out how, in science history, there were times when widely accepted theories were held on to long after the evidence supported throwing them out. An example is the "ether," a mysterious substance that was hypothesized as necessary for light waves to travel. But tests showed that the ether did not exist, and new theories were developed to explain how light can travel in a vacuum.
However, historical precedence does not and should not mean that MOND is more likely than dark matter. As I said, dark matter still needs to be proven in a lab, and it may be that we find that dark matter, like the ether, doesn't exist. But, unlike the ether, dark matter has made testable predictions that are, for the most part, correct. And other "phantom" particles, like neutrinos, were hypothesized long before they were conclusively detected in the lab.
So, dark matter is not dead, nor even slightly ill. And if MOND wants to challenge the reigning theory, it needs to bulk up and fight according to the scientific rules of engagement: (1)develop a hypothesis, (2) make a testable prediction that would not be predicted by the prevailing theory, and (3) test the prediction.