It is unusually warm compared with the planet's surface (for reasons which are not understood), and this warmth has distended it so many hundreds of kilometers above Pluto's surface that it seems to be undergoing "hydrodynamic escape".

That is, it's being slowly swept away from the planet by the solar wind, a process which exists nowhere else in the Solar System today but was a major influence on the atmospheres of the inner planets during the Solar System's earliest days. Indeed, there is evidence that a little of it may be periodically tidally transferred to Charon and then back.

Then there's the question of how Pluto's atmosphere may relate to its surface -- whose features (as I said) seem to be seriously unlike those of Charon, and which in fact seems to be covered with different substances (frozen nitrogen, methane, carbon monoxide, and perhaps organic compounds) missing on Charon's surface (which is apparently mostly just water ice).

Why the difference?

To what extent may the periodic freezing and rethawing of the atmosphere have modified Pluto's surface?

And to what extent can a chemical analysis of Pluto's atmosphere further enlighten us as to its solid composition (which may include a thick layer of organic compounds)?

That surface composition, in turn, can tell us more about the extent to which the original object from which Pluto and Charon were formed was modified when (as is believed probable) it was split in two by a giant impact similar to the one, which is thought to have split our own Moon away from Earth.

Finally, Stern mentioned the fact that the very process of atmospheric freeze-out currently going on is meteorologically interesting itself -- no one is certain how fast it will occur.

It is being accelerated by the fact that not only is Pluto currently moving farther away from the Sun, but -- because Pluto's spin axis is titled on its side, and at perihelion it happens (by chance) to be almost side-on to the Sun -- a larger and larger area around one of its poles is now being cast into permanent nighttime shade, providing a cold patch where the atmosphere can freeze out more easily.

This is also steadily reducing the area of Pluto's surface, which a spacecraft can photograph -- another argument for an early Pluto probe launch which the SSES considers important.

And as the atmosphere starts to freeze into a layer of light-colored frost on Pluto's surface, more and more sunlight is reflected back into space without being absorbed to warm that surface -- producing a possible self-amplifying "runaway freeze out" process similar to the ones which may have had major effects on the climates of both Earth and Mars during their early days.

Because of this, Stern raised the serious possibility that at some unpredictable point in the next few years, Pluto's atmosphere -- instead of disappearing gradually -- could suddenly collapse without warning, increasing the risk that any delay in a Pluto mission would deprive it of the chance to observe the atmosphere.

However, a minority of scientists thinks other physical processes may be producing a "negative feedback" which will slow Pluto's atmospheric freeze out. We just don't know yet.

So, while a delayed Pluto mission would still have considerable scientific value -- and the SSES stated that scientific observations of smaller Kuiper Belt objects, which such a mission could also carry out, are more important than had been thought even a few years ago -- clearly an Outer Planets program in which a 2004 Pluto mission was launched before a Europa Orbiter will give NASA more scientific bang for its buck.

Such a program would, indeed, be a good deal cheaper in absolute terms, since a Pluto mission requiring a SEP or solar sail system will naturally be a lot more expensive than one that just uses a Jupiter gravity-assist. (And since Pluto is moving farther from the Sun, the longer a Pluto mission is delayed the longer the flight time to the planet becomes, further raising both its cost and the risk of failure.)

But how can the cost of both missions be minimized? Several proposals were made. One technique which is clearly viable -- and is now likely to end up being used -- is to catapult the Europa Orbiter out to Jupiter by having it make one or more gravity-assist flybys of Venus, Earth, and/or Mars, rather than launching it directly to Jupiter.

This would prolong the flight by a couple of years -- but it would also allow use of a much smaller launch vehicle, providing a big net cut in the mission's total cost. (Earth flybys may be unpopular, since the craft will be carrying plutonium -- but there are lots of possible inner-planet flyby plans.)

At the meeting, Planetary Society director Louis Friedman urged another possible solution -- launching the Pluto probe with a Russian Proton booster, which would actually have more capability than a Delta 4 while costing only half as much. But there is uncertainty as to whether current federal law would flatly forbid launch of this payload on a foreign booster -- especially since - in this case - nuclear regulations would unquestionably force the Russians themselves to provide the entire RTG for the spacecraft.

Another idea is to cancel the expensive development of the advanced, highly efficient Sterling nuclear generator currently planned for the Orbiter. This would not increase the craft's total weight, but it would require it to carry about three times more plutonium -- which produces problems of its own, since the stuff is so expensive.

In fact, it was revealed at the meeting that the Department of Energy currently does not nearly have enough plutonium-238 stockpiled to fuel such a Europa Orbiter. A decision will be reached soon as to whether NASA should ask the DOE to manufacture more, ask the Defense Department to provide part of its plutonium store instead, or even buy cut-rate plutonium from Russia.

One plan, which was firmly rejected, though, was to have the spacecraft simply orbit Jupiter itself and make a series of close flybys of Europa. While this would greatly reduce its needed supply of braking propellant, a JPL study revealed that such a craft would still be too heavy to be launched with anything short of a Delta 4 or Atlas 5 booster -- and so the cost reduction would really be quite small compared to the serious loss of science data about Europa.