Gary,

Great questions! Maybe someday we'll have great answers... :)

Seriously, it is kind of wierd, but let me toss out a few possibilities.

First, I think it is generally agreed now that the floor of Gusev has mostly been covered with flows of basalt that are more recent than any of the lake sediments. Many thought that before MER landed, too. That was why Nick wasn't that happy about it as a landing site. Those mesas off toward the mouth of Ma'adim are presmably related to the water flow, but unreachable. In the landing elipse, you'd have to either find a crater that drilled down below the basalt (which is what they hoped for at Bonneville), or, you'd have to find bits of old sedimentary terrain that for some reason were high-standing and were not buried by the basalt. We got lucky and got the latter: the Columbia Hills (move the landing point a few miles further downrange, and this would have been a very boring mission)

As for where the basalt came from, they haven't found that yet, but with fluid "flood basalts" like this it can be hard to tell; they often leave much in the way of edifices around their eruption point, maybe a subtle low ridge where a fissure was. Then add a bunch of cratering and wind erosion.

I think everyone agrees the hills are older than the basalt: they look more cratered, and it looks like they are embayed by the basalt.

So why are they sticking up there? Some ideas:

1. The original fill in Gusev was much thicker, and at least a few hundred meters of it eroded away before the lava flows. For some reason, this set of hills was resistant. Maybe better cemented than most. Maybe there was a crater above them (since eroded away) that shock-bonded the rocks. Maybe there was some early vulcanism here and there was chunk of lava or a cinder cone on top of them when that old surface started to erode.

2. The original fill level in Gusev was below the hills, and then they were raised up somehow, before the lava flows happened. To me, seems the most likely is that the hills represent a crater rim segment, or more likely, a spot where several large crater rims came together, and thus may have formed an especially high/solid point. Then the basalt came in and filled in all those old craters, except this one high point.

3. A combinationof 1 and 2 (which seems most likely of all to me, though I'm definitely not a geologist!).

In any of these cases, the lake fill is *old* -- the basalt is old enough, and the deposition must have happened long enough ago to fill floor and then erode back down, and probably far enough back that the cratering rate was still quite high.

4. I suppose the hills could also be volcanic in origin (but predating the basalt). An intrusive mass, a chain of cones, whatever.

The 40% salt/sulfate levels do sound hard to acheive other than evaporites. The Clovis numbers aren't that high, I think, since it has a lot less sulfur than Meridiani (but just as many salts, it seems). I'm not sure if that requires deposition in water.

Given that a lot of other rocks up there do look like regular basalt, I'm not sure that would make much sense for a lake bed -- why jumbled up evaporites/lava? Maybe that's an argument for the old crater rim theory, in which case either the impact mixed up old laval flows and salty sediments, or, it hit pure basalt and there was aqueous alteration of the fractured rock *after* the cratering event (but long before the top level flood basalts on the plain). Of course, that requires getting water up to the top of a hill, which could be problematic.

Or, picking up on 4, maybe they are volcanic and there was a hydrothermal system.

Bottom line: it's fun theorizing but I have a paper to present at a confernce in two weeks -- and it ain't about geology so I'd better get back to work!