Streak tests are easy tests helpful in mineral identification. The streak is simply the color of the powdered mineral. It doesn't matter how the mineral is powdered - you can scrape off some with a nail or pound the mineral to bits with a hammer. More commonly mineralogists use a streak plate, a piece of unglazed porcelain usually cut in a square or hexagon a few inches across. Streak plates have a hardness of about 6.5, so if you want to test the streak of anything harder, get out the hammer! They can be bought from most mineral supply houses. For example, the latest Ward's Natural Science Establishment catalog lists them at 10 for $2.90. When they get dirty they can be cleaned by scrubbing them off with an old toothbrush. I often use some sand with the water to scour off resistant streaks. If they get too dirty - heck, toss them out - they cost less than 30 cents each. When I was a kid, I used the back of old bathroom tiles to make an even cheaper streak plate.
Why do a streak test instead of just looking at the color of the bulk mineral? The color of a larger chunks of mineral can really vary, depending on what trace element impurities may be present. Calcite, for example, can be any color of the rainbow ( and a few that aren't on any rainbow). But calcite always has a white streak. So why don't the impurities color the streak? They do, but only to a slight extent. This is because light going through a small grain of a mineral has less chance to interact with the impurities than light going through a big chunk of the material. Powdering the material thus minimizes the effect of the impurities.
Streaks are most useful in the oxides and sulfides. Silicates and carbonates generally have white or light colored streaks. The oxides are fun to streak. Hematite's red streak is distinct from goethite's yellow-brown streak and pyrolusite's coal black streak. Sphalerite is another mineral that can be lots of colors, but gives a yellow streak.
The streak of rocks is generally not distinctive. They usually give a light streak that reflects their dominant silicate or carbonate composition. If they give a red or brown streak, it suggests the presence of iron oxides. Of course, if the rock is coarse grained, you can try the streak test on the individual mineral grains.
Mineral databases and texts sometimes list the streak colors and some times don't. It depends on the tastes of the author and the data available. All minerals have streaks (you can powder anything if you put your mind to it) but they may not be too distinctive (hundreds of minerals have white streaks). I think that when a new mineral is described, the streak should always be included. After all, the material had to be powdered in order to do its microprobe or x-ray analysis, so all some one needs to do is remember to record the color. That would be a real help to those of us who don't have well-equipped analytical labs in our basements.Mineralogical Meanderings