When this occurs, many different things can happen. If the un-evolved star is not very massive, and the separation between the two stars is large, maybe nothing will happen; the two stars may remain detached. However, it is possible that part of the surface of the giant star may find itself closer (gravitationally speaking) to the companion (secondary) star. If this occurs (i.e., if the red giant overflows its Roche Lobe) mass will be lost from the red giant and accreted onto the secondary. (This is sometimes called a semi-detached system.) Alternatively, the red giant may become so large that it engulfs its companion and forms a contact binary system. In a contact binary, the secondary star orbits within the atmosphere of the primary!

In contact binaries, things get crazy, and the results can be astounding. For instance, the friction caused by one star orbiting inside the atmosphere of the other can splash the atmosphere right off the red giant, leaving a naked core. (This is akin to stirring up the water in a bath-tub so much that it splashes out of the tub.) At the same time that this is happening, the secondary star will lose some of its orbital energy, due to friction, and spiral in towards the core of the primary. The result may be either a merged core, or two stars that are very close together. Note that, if the outside of the red giant is stripped away, what's left is nothing more than a white dwarf star.

Now things get interesting. Suppose the result of the contact binary system is a white dwarf and a companion star that are very close together. It is possible that the companion star will overflow its Roche Lobe and accrete matter onto the white dwarf. To some extent, this accretion is similar to the accretion of a semi-detached binary. However, there is a fundamental difference between the two events. To understand this, you have to consider what accretion really means.

In an accreting binary system, the secondary star is not standing still in space -- it is orbiting the white dwarf. As a result, when matter comes off the star, it, too, will orbit the white dwarf. So, if the matter is orbiting, how does it wind up accreting onto the other star? The answer is that is doesn't, unless something can take away its orbital motion. Is there something that can do that? Yes, there is --- its the accreted matter itself. If alot of gas from the primary star is orbiting about the secondary, there will be friction, as some of the particles rub up against each other. This friction will heat up the gas, and also slow the gas down, so that it gradually spirals in towards the white dwarf. The gas will form an accretion disk.

Now note: the amount of friction occuring in the accretion disk depends on how fast the gas is orbiting around the star. But the orbital speed, of course, depends on the gravity (Kepler's and Newton's laws): the more massive the body, the faster the orbital speed; the closer to the body, the faster the orbital speed. If a secondary star is accreting onto a compact object, such as a white dwarf, the orbital speed and friction can be extremely high. The friction will heat the gas to extremely high temperatures -- in fact, from the blackbody law, the accretion disk will glow in the x-ray portion of the spectrum. When a ``star'' is observed to emit x-rays, it's very likely that it's not the star that's emitting the x-rays, but an accretion disk.

Finally, let's consider some of the strange objects that are formed by accreting systems. First, let's accrete onto a normal white dwarf. White dwarfs are white dwarfs because they have no hydrogen to burn. However, through accretion, the star can find itself with a substantial layer of hydrogen on its surface. After a while, as more matter is piled on it, bang -- the accreted matter (explosively) fuses. We call this a nova.

Now, let's fast forward a few years. White dwarfs are white dwarfs because electron degeneracy holds up the star against gravity. But electron degeneracy only works if the star is less than 1.4 solar masses. Normally, this is not a problem, but, over time, an accreting white dwarf may get enough material to put it over the Chandrasekhar limit. If this happens, crush, ka-boom, the white dwarf collapses to a neutron star in a supernova type event. (Sometimes this is called a Type Ia supernova.)

Even a neutron star can be affected by accretion. Just like a person who strokes a basketball and makes it spin faster and faster, the accretion of matter onto a neutron star can make the star spin faster and faster. In the process, the neturon star/pulsar can be spun-up to rotate at thousands of times per second, Moreover, all this time, the accretion disk is radiating hard x-rays. This high energy light will blast the secondary star, and the added heat may literally evaporate it! The result would be a bare, millisecond pulsar.

And, of course, the binary systems with the most energetic x-rays involve accretion onto a black hole. Near the even horizon, the orbital speed of the gas is enormous, so extremely high temperatures are achieved. By looking for the brightest and most energetic x-ray sources in the sky, astronomers find black hole candidates.