The Earth's moon has an unusual surface. The near side is dominated by large 'seas' (lunar maria) of basalt whereas the far side is covered by lighter uplands of lighter coloured silicon-rich rocks. It used to be thought that the Earth's gravity had somehow caused the eruption of flood basalts on the near side, but it was eventually realized that centrifugal forces caused by the moon's obit about the Earth would cause an equal pull on the far side, so this could not been the case; since this revelation it has generally been assumed that the lunar maria are flood basalts that coincidently erupted on the Earth-side of the moon.
The Near (left) and Far (right) sides of the moon.
The origin of the moon has long been a subject of scientific curiosity. In 1975 two papers (one by William Hartmann and Donald Davis of the Planetary Sciences Institute in Tuscon, Arizona and the other by Alfred Cameron and William Ward of the Harvard-Smithsonian Centre for Astrophysics) separately proposed that the moon was formed by the collision of a large body with the early earth, throwing a large amount of ejecta into space, which eventually consolidated to form the moon. Neither of these papers received much attention at the time, but after the Conference on the Origin of the Moon in Kailua-Kona, Hawai'i, in 1984, the idea became popular, and over the next two decades a series of increasingly sophisticated computer models seemed to confirm this theory.
Computer animation of the formation of the moon.
In the 4 August 2011 edition of the journal Nature Martin Jutzi and Erik Asphaug of the Earth and Planetary Science Department at the University of California Santa Cruz propose a new theory on lunar formation. They noted that the computer models used to calculate that the moon could be formed by the impact of a large body into the early Earth were as likely to result in the formation of two moons as one.
They then decided to try to model the development of a two moon system to see what would happen. This had surprising results; their models suggested that two moons in similar orbits would eventually collide, but not in a large spectacular explosion. Instead they would collide more slowly, at only 2-3 kph. If one body was considerably smaller than the other (as was likely) this smaller body would be unlikely to form a solid core, but instead be a loose collection of rubble. Such a small rubble moon colliding with the larger moon at a low speed would not impart enough energy to smash the moon apart, but rather would accrete onto one side of the moon - forming the lunar highlands.
As predicted by Galileo, mass does not effect acceleration due to gravity.
Under this theory the Lunar Maria become the older, original surface of the moon, with the highlands of the far side overlying them, a more recent occurrence.