We are at a somewhat similar juncture in our understanding of Martian geochemistry now as we were with lunar geochemistry after the Surveyor program. There are parallels, I think.

Without samples in hand, and with 1960s technology, the best we could do for in-situ examination of lunar surface rocks by unmanned landers was the Alpha Particle Backscatter device, a direct ancestor of the APXS units flown on Pathfinder and the MERs. With just the element distributions readable by those devices, we were able to infer the basaltic nature of the maria, but the anorthositic nature of the highlands was less well defined. It was one possible solution to the aluminum-rich distributions, but not the only one. It took until the fourth manned lunar landing for an actual chunk of anorthosite to be found and brought back to Earth, confirming the anorthositic nature of the Moon's original crust. (Small flakes of anorthosite had been found in the soils returned by earlier missions, but there was so little of it, even in the mostly basaltic breccias returned from Fra Mauro, that alternate theories for the Surveyor VII data weren't completely ruled out until the J missions returned significant percentages of anorthositic samples.)

We are lucky to have additional sensors on Mars -- the Mossbauer and mini-TES results are invaluable in identifying actual mineral signatures. But there are still more questions than answers when it comes to things like soil chemistry. The Viking biology results are poorly constrained at best -- hopefully the soils at the Phoenix landing site will be similar enough to the Viking soils that its results can better constrain the possible causes for the Viking results. But we're still in the same situation in re Martian soils as we were in those delicious months between the final Surveyor results and the first returned lunar samples; the peroxide theories mostly account for the observed results, but they're certainly not the only possible causes.

I'm hoping we can get enough information from Phoenix to be able to better constrain our ideas of Martian soil chemistry, because it's going to be a long time -- a lot longer than the 18 months between Surveyor VII and Apollo 11 -- before we can examine samples... *sigh*...

-the other Doug

When discussion of Mars sample return comes up I find myself wondering ... what about the Martian meteorites? Certainly they're not pristine, but to some extent that can be compensated for. There's also the problem that (so far) they only sample a few sites, but this would be somewhat true of samples returned as well. Presumably we're missing soil samples, unless they're incorporated in the rock. Hunting meteorites in Antarctica is presumably fairly inexpensive compared to a sample return mission.

Just playing devil's advocate here, I don't particularly oppose a sample return mission, though I'd rather see swarms of MSL's instead .

I though about that too. But then I thought, (as you also mention) 'if the aim is specifically to study
martian soil and its intresting characteristics as revealed by Viking, the meteorites are of little help.'
Also, the significance of samples specifically picked for return based on a host of parameters determined
by a science rover is huge compared to the "beggers can't be choosers" samples we have now. As well,
there is the importance of knowing where a sample came from, it's original placement (where it was formed)
in the geological framework of Mars.