Removing the overburden of a rock will result in cracks. In mining, this is a rock burst, a serious problem to the industry.

A meteor impact creates fractures.

Repeated cycles of thermal expansion amd contraction, over time, will cause fractures, even without water.

For the most part, water infiltration and subsequent freezing exploits pre-exisiting fratures. The water has to get into the rock in the first place.

The shape of many of the cobble and smaller size rocks on the surface of mars are called ventifacts (http://en.wikipedia.org/wiki/Ventifact), semi-triangular shaped rocks that have been polished by the saltation of sand grains driven by the wind. Such shapes are common here on terra, and we saw lots via the Spirit rover.

Given the lack (?) of lower temperature felsic rocks that host mica, this mineral is unlikely to be our suspect. Mafic rocks, such as basalt and gabbro, do contain plagioclase feldspar, which has two cleavage planes at nearly right angles to each other. When looking for feldspar in a hand sample, it helps to turn the rock in the sun and look for little sparkles coming from the cleavage planes.

So that’s my first guess: a phenocryst (a larger crystal) of sodium rich plagioclase feldspar, specifically labradorite, which also displays an irradescence due to light bouncing around internal crystal twinning planes. (That’s a longer discussion.)

OR, it might be a piece of volcanic glass, like obsidian, which would also not be unexpected in this setting. However, glass breaks along a scallop shape (concoidal fracture) so the reflection tends to sweep across the surfae as opposed to being a single plane.