But there is now every reason to think that it will still be working by then -- and, sure enough, NASA has been putting great thought over the past year into still another mission extension.

The details of that planning have been mostly secret, although some tentative information has leaked out. This week, the beans were finally spilled when the Committee on Planetary and Lunar Exploration ("COMPLEX") of the National Academy of Sciences -- which provides NASA with a great deal of its mission planning information -- released two documents describing its initial recommendations for that third extended mission.

There are, however, still several alternative mission scenarios being considered for this period.

NASA is faced with a problem. Clearly, at some point in the reasonably near future, Galileo really will break down, either from radiation or simple old age. And after it loses its ability to maneuver, many things could ultimately happen to it,d depending on how its orbit is affected by the complex gravitational tuggings of Jupiter's four big moons.

It could, decades from now, make an accidental close flyby of one of them that would finally catapult it completely away from Jupiter and back into solar orbit (its least likely fate, according to preliminary Jet Propulsion Laboratory calculations), or it could make such a flyby that would slow it down enough to actually enter Jupiter's atmosphere and burn up. But its most likely fate would be to ultimately crash into one of the four Galilean moons.

And NASA very much wants to minimize any chance that it might hit Europa -- for that is one of the very few other places in the Solar System with a serious chance of possessing either past or present-day microbial life, and Galileo was not sterilized before launch.

Bacteria, as we now know, can be astonishingly tough. While the odds are small, the possibility cannot be ruled out that a small number of bacteria from Earth have survived years of Jupiter's radiation inside the craft's metal body, could survive even a crash at several thousand km per hour that would vaporize much of the craft -- and would then be embedded in Europa's surface ice deeply enough to avoid a fatal dose of radiation afterward, and be frozen into a state of suspended animation in that cryogenically cold ice for thousands or even millions of years (as some bacteria have been in Earth's permafrost).

Europa's ice is not completely static; its surface features are known to change over periods of a few million years, and perhaps a lot faster in some areas. Even its solid ice may be pliant enough to churn in very slow convection currents over periods of tens of thousands of years; Jupiter's tidal forces may cause blocks of it to slide past each other along fault lines; and if there is a subsurface liquid ocean, water from it may flood cracks -- or even larger "melt-through" events -- all the way up to the surface.

Thus there is a chance that any living germs from Earth in frozen suspended animation could eventually end up finding their way to Europa's ocean and thawing back to life. And (as the COMPLEX committee pointed out in a report on Europa exploration last year), if they ever do so, they would spread all over Europa's worldwide ocean quickly -- much more quickly than any Earth germs that found their way into the underground liquid water table kilometers below Mars' surface would spread across that planet.

Thus it's very important to avoid even small risks of biologically contaminating Europa -- in a long report on that subject that has just been released, COMPLEX recommends that NASA should reduce the chance that any one spacecraft will inject even one germ into Europa's ocean to less than one in 10,000.

And the only way to minimize any chance that Galileo might do so is to deliberately arrange for it, at the end of its mission, to either escape from Jupiter completely (which is impractical for other reasons) or crash into Jupiter itself or one of its other moons.

The new report recommends that crashing it on Ganymede or Callisto should also be avoided -- there is some evidence that they may also have much more deeply buried subsurface liquid water layers, and while present or past life on them is very unlikely, it can't quite be ruled out.

Impact on Io would be safe, though -- that rocky moon has been completely baked clean of liquid water by its continual violent volcanic activity and is even more hostile toward life than our own Moon.

And plunging the craft into Jupiter itself would also be safe -- even if some small pieces of it survived the fiery high-speed entry, it's now certain that Jupiter (despite the earlier hopeful speculations of Carl Sagan) has always been an extremely unlikely place for life to develop.

The entry probe that Galileo parachuted into the atmosphere in 1995 found only tiny traces of any organic compounds -- and any germs that did float in its atmosphere would find themselves being yanked down into Jupiter's roasting interior every few days by its powerful vertical convection winds. (The same is true of the other three giant planets -- nor has anyone come up with any convincing scheme as to how native life could have evolved there in the first place, since the giant planets' only liquid water is in the form of isolated cloud droplets. For these reasons, the Galileo Jupiter atmospheric probe was not sterilized.)

So NASA wants to end Galileo's mission by crashing it into either Jupiter or Io -- and since Jupiter is a much bigger target, it's easier to hit. But before doing so, NASA also wants to carry out still more scientific studies with Galileo, at two possible targets.

One is Jupiter's little inner moon Amalthea -- a potato-shaped lump of rock 250 km long that has only been photographed at a distance. While it orbits in a radiation environment even more intense than that around Io, a single flyby within a few hundred kilometers could provide not only close up photos but (through tracking of the spacecraft) a good measurement of its mass and thus its density, allowing an estimate of the overall rock type making it up.

The other is Io itself -- it may be some time before another spacecraft gets a chance for a close up look at it. While the problem with Galileo's near-IR spectrometer prevents it from making any close up maps of Io's minerals, more close up photos and detailed temperature maps of its spectacular volcanoes and other geological phenomena would be very valuable -- and scientists particularly want Galileo to make a low-altitude flyby of one of Io's poles, in order to settle the question of whether its molten interior allows it (like Ganymede) to generate its own magnetic field.

Galileo's flyby last November took it over Io's south pole at only 300 km, but unfortunately a safing event (due to a combination of a single radiation-damaged memory cell and a software bug that has since been corrected) kept it from making any measurements during its closest approach. In its new report, COMPLEX calls another polar flyby of Io the single most important thing that Galileo could do with its remaining life.