http://marsrovers.jpl.nasa.gov/gallery/press/opportunity/20040202a.html

Finally some new images. A high-res color pan from Opportunity, and even more interestingly, some spectral data from Pancam. Right now only spectra from Opportunity, though they showed some from Gusev, too. There is also a convenient reference chart showing spectra of some relevant minerals (in the 400-1000nm range).

One thing -- those "light toned" outcrops are not *that* light. We're talking a total albdeo of maybe 0.25, and as Nick showed, clearly reddish.

The soil, by contrast, is between 0.1 and .15, and the dark "blue-grey" cobbles 0.05-0.1

A few thoughts:

That puts the "light toned bedrock" into the range of mafic minerals like olivine and pyroxene -- the spectra isn't right but for those in particular but my point is that they are fairly dark, which I think makes them a little less exotic than they seemed at first (to me at least).

The rocks are still too red to just be basalt or something similar. They are more red than the soil, even (especially if you include near IR, as Nick's images showed us), or perhaps more salmon-orange would be a better way of putting it, since their spectral peak starts in the yellow-red. That early peak and the falloff after 800nm looks not quite right for hematite. Maghemite is better but still doesn't quite seem to do it.

The JPL people included a few other minerals that look like they could be components of that . Interesting implications for each:

Goethite: hydrated hematite, basically. Yet another of these iron oxides, but this one with hydrogen included.

Jarosite: a hydrated iron sulfate, potassium bearing too.

Palogonite: "basalitic obsidian". Basically, glassy, sometimes tuffaceous igneous rock, but with a basaltic composition rather than the silica-rich composition of most glassy rocks. On Earth, it seems to form most often when basalt interacts with water (eg, the glassy rinds of pillow basalt underwater).

Of course, I suspect the sample minerals they included are related to the theories they have about what's at the site -- there are probably others that would be just as relevant spectrally but that aren't part of the paradigm.