Although it will no longer be returning samples to Earth, the drill might still be worthwhile, since the onboard instruments can look at and analyze the ends of extracted rock cores from below the layer of surface weathering that coats Mars rocks.

In addition to its panoramic science cameras, Athena would navigate using pairs of cameras on its front and rear, as well as a belly-mounted camera to provide closeup views of the sampled surface. It could, with luck, drive as much as 100 meters across the Martian surface each day.

On the previous night, its Earth controllers would study the previous day's photos, pick out the next day's desired destination (as well as one or more intermediate "waypoints"), and send the rover its marching orders.

On the next day, it would proceed automatically to each waypoint along a straight-line path, veering around detected obstacles automatically. Instead of Sojourner's laser obstacle detection system, it would actually analyse the stereo photos from its cameras directly to detect dangerously large rocks or holes.

This technique sounds doubtful, but it has been developed -- and tested -- for several years on test rovers here on earth, with a very high success rate. (Only last month, Athena's earthly prototype "FIDO" underwent a long and very successful test in the Nevada desert.)

What is uncertain is whether this mission can be prepared in time. The Pathfinder landing system has of course been successfully tested on Mars itself -- and since Athena was already scheduled for a 2003 launch, its own development doesn't need to be speeded up. But there are still possible problems.

NASA is currently very closemouthed about the details of its 2003 Mars plans, but it's possible that there might be problems in making the total spacecraft light enough to be launched on a Delta 2 booster -- especially since the rover, which originally was supposed to communicate with Earth only through orbiting Mars relay spacecraft, will now have an X-band radio system with a pointable 30-cm dish antenna added to it to allow a backup communications link directly to Earth. (This system by itself could add about 20 kg to the rover's weight.)

There may also be problems in designing Athena to survive Pathfinder's high-speed bouncing landing, since it was originally designed to make a gentle touchdown aboard a full-fledged soft lander.

But if it is possible, this could be a mission of major scientific importance -- especially since the Pathfinder system, since it can make a safe landing in almost any kind of rough terrain, could plant Athena in a far greater number of scientifically interesting landing sites than the original soft-landing carrier spacecraft could.