MOND: An alternative to Dark Matter
MOND stands for MOdified Newtonian Dynamics. It is an alteration to Newton's law of gravity, made to fit the observations of galactic movement and to do away with the dark matter problem. It was first proposed by Mordehai Milgrom in 1983. The theory was looked down on by most physicists because it went against Newton's laws, which have been proven to work in nearly every case imaginable. If proven to be true, it would take all the fun out of solving the dark matter mystery.
The way that MOND works is that for very small accelerations (about 10^-10 meters per second squared), force is proportional to acceleration squared (f=ma^2 instead of f=ma). The theory accurately predicts the movements of galaxies, doing away with the need for a halo of dark matter. The theory also eliminates the need for extra mass in the universe to slow the universal expansion.
The mechanics of nearly every galactic system found in the universe agree with MOND's predictions, with the exception of galaxy cluster cores. In order for cluster cores to fall in line with MOND, there must be some dark matter present. This fact makes MOND insufficient to explain dark matter, but there are arguments for why it is still a viable theory. For all we know, cluster cores could be more susceptible to the formation of black holes and other collapsing stars.
MOND can not be tested experimentally because the accelerations involved in the theory cannot be simulated here on Earth. The theory can only be compared with observations, and in that matter it performs incredibly well. However, many argue that MOND is not a true scientific theory. It only explains what it was meant to explain, and even then there are exceptions. The fact that it agrees with observations doesn't necessarily mean it is grounded in reality; rather, that it predicts what it was designed to predict and nothing more..
For more information on MOND, read Milgrom's article in the August 2002 issue of Scientific American.