NASA's Stardust mission has just reminded us — for the second time — that we don't know as much about comets as we thought we did.

Comets are thought to be nearly unaltered relics from the formation of the solar system, and Stardust's target, 81P/Wild 2 (pronounced vilt) was considered a perfect target. These days Wild 2 can venture fairly near the Sun (just beyond Mars's orbit). But prior to a close brush with Jupiter in 1974, its orbit was much more distant.

So cosmochemists would have bet money (and maybe did) that the dust and gas coming off Wild 2 when Stardust flew through it in 2004 would be primordial building blocks, released from their deep-freeze confinement after 4½ billion years, that would unlock the key to the solar system's formation.

The smart money, it turns out, was wrong. Not long after Stardust returned to Earth in January 2006 with its precious cargo, it became clear that some of those building blocks came from a very hot kiln. Researchers found an array of minerals that formed at temperatures of at least 2,000°F (1,100°C). Not exactly "deep-freeze" conditions!

Now Stardust has stumped the scientists again. In Science for January 25th, a team led by Hope Ishii (Lawrence Livermore National Laboratory) reports that the captured cometary grains are closer matches to "an inner solar-system asteroid than an outer-solar-system comet with primitive unaltered dust."

In other words, silicate minerals that formed near the Sun were somehow transported to the outer system, where they mingled with ices to create Comet Wild 2. There are no mineral clusters untouched since the Sun first blazed to life, no infusions of exotic isotopes from distant supernovas — if it weren't so dynamically implausible, you might think Comet Wild 2 was an escapee from the asteroid belt.

Did all comets form this way? Probably not. High-flying jets have captured a host of interplanetary dust particles drifting into our atmosphere from space. These tiny treasures have the kind of primitive compositions and unique isotopic mixes that scientists had expected to find in the Stardust samples.

So maybe we just need to go sample another comet (Rosetta, a European-built craft, will do so in 2014). Or maybe our theories of solar-system formation need retooling. Whatver the reason, Stardust has lots of cosmochemists scratching their collective heads right now.