Hope I dont burn anyone with excessive posting and speculation... but here goes:

These new pictures from the Columbia Hills do show spheres. Not in the abundance as at Meridiana, and not as well formed, but with many simularities. The rocks containing them also have simularities to those of Meridiana, but also important differences. The most obvious difference is the elevated rims seen in many images. This imagehas caused me to conclude that a good possibility is that we are seeing sediments from much shallower water than the moderately deep water limestone-like sediments of Meridiana. (only Burns Cliff has shown cross bedding, so far). My conclusion is that these here at Columbia are from the reef top or shoreline facies itself. The rims are formed from algal strands, similiar to encrusting red algae or stromatolites, and formed independant coral like colonies, or formed rims around pre-existing rocks. Unlike most stromatolites that form domes, these formed rims in the hypersaline, precipitating environment. They may be very close to, or immediately on top of igneous shoreline rocks, which is why there is not much of this type of deposit at Gusev. This could be the shoreline deposits of the lake that filled Gusev itself! Only this lake was not as deep, or the thicker Sulfate deposits are in a deeper part of the basin. If these are older, basement deposits predating Gusev crater, then there should be more Sulfate containing deposits above, probably making a transition to deeper water environments.

The encrusting algae could have formed into small enclosing pools along the tidal zone. These were over run by waves, allowing coarser sediments to infill them. Thus we see the same spheres as at Meridiana, but more frequently eroded and detached from the associated algal strands. Many have been formed into lines by the current. Bedding is much coarser, or absent. There is a preponderance of channel like sets of dual rims, thickened at the entrance side of the channel only. (I need to collect the examples of this, sorry for not linking more)

Some spheres can be seen, apparently attached to these segmented strands, as is the case at Meridiana, these are particularly evident toward the center of this stromatolite pool- as if they grew upright there. Indeed, that is what I think. Some of these nodules imaged by the micro cam do not seem to be like the usual perfect berries. I admit that. These in this next image seem to be nodules made out of encrusting algae or coral. There is a nice complex segmented microfossil right at the tip of the main nodule at lower right center, I propose.

The eroded rock in the first linked image shows a number of spheres on narrow upright "stems". I would like to again point out the incongruity that there is not rain on Mars to cause vertical "hoodoos." Wind, and even more so, particle laden winds, come from the side, not from above. In the upper right quadrant, some long shadows are a give away for some long strands tipped by spheres. These strands have bends, dispelling any explanation as hoodoo type ventifacts. Thus these lengthy, often vertical, narrow sections are clearly an incongrous element of the rock matrix. They are harder, like the spheres, and in such a close association with them, that there is likely some relationship there - that might be described as a "stem". Some micro photos, please... please?