One of the greatest mysteries of astronomy is the problem of the missing mass: All of the matter scientists can see in the universe accounts for only a small percent of the observed gravity.

Astronomers often invoke the concept of dark matter to explain this discrepancy, but some researchers say the problem is really our understanding of gravity. These scientists tout an idea called MOND - Modified Theory of Newtonian Dynamics - to explain why the universe seems to behave as if there's much more matter in it than we think.

Instead of assuming that this missing mass exists in the form of dark matter, which scientists have yet to detect directly, MOND advocates say we must alter Einstein's theory of General Relativity.

Under MOND, mass is much more effective at bending space-time than under General Relativity, so it takes less stuff in the universe to account for all the gravity we measure.

Fudge factor still needed

Though no one has yet proven or disproven either dark matter or MOND, supporters of the latter are in the minority. And MOND may be becoming even more of a long shot, according to cosmologist Pedro Ferreira of Oxford University in England. Ferreira wrote a review article in the Nov. 6 issue of the journal Science assessing the current state of MOND ideas.

"My personal view at the moment is that dark matter is a far simpler theory than any of the modified theories that I've seen," Ferreira said. Nonetheless, he said MOND shouldn't be discounted out of hand just because it's the less popular idea, nor because many physicists are loathe to tamper with Einstein's General Relativity.

"Very few people have worked on MOND; a very large number of people have worked on dark matter," said Jacob Bekenstein, a physicist at Hebrew University in Jerusalem who has researched MOND. "To compare them is kind of silly because we don?t really know enough of whether MOND is working well or not. Just too little effort has been going into MOND."

Bekenstein admitted that MOND was not yet a fully fleshed-out theory: It cannot make physical predictions on all scales of the universe.

When applied to just galaxies, MOND can predict very well the behavior that astronomers observe. But when MOND is applied to larger structures like clusters of galaxies, it fails. To make MOND work for clusters, it must include more complicated concepts, such as entities called dark fields, which are different from dark matter, but work in a similar way to alter the amount of gravity present.

"It seems like if you want to build a proper theory of MOND, you bring in something like dark mater through the back door," Ferreira said.

This fudge factor seems to defeat one of the primary purposes of MOND when it was first proposed, which was to avoid having to invent a mysterious unseen entity acting in the universe, such as dark matter.

Even Bekenstein admitted that involving dark fields in MOND is not ideal.

"If you work only on galaxies then MOND doesn?t need any help," he told SPACE.com. "But if you go up to clusters it needs some help. This is one of the things I hold against MOND."

However ...

Bekenstein pointed out that dark matter isn't perfect either. Thirty years after it was proposed, scientists have yet to find the stuff out there in the universe, and the idea isn't yet ideal at predicting all manner of situations, either.

"In the models of galaxies with dark matter, you have to carefully adjust the distribution of dark matter," he said. "Since you don?t see the dark matter you're kind of free to adjust what you want, but it's not very credible in my opinion. It's too free an idea."

Ferreira said some kind of answer may come soon with the advent of new satellites set to observe the distribution of mass in the universe more precisely.

"I think things are going to really heat up over the next 10 years," he said.