I was sitting in a restaurant and the cars parked outside were parallel to the restaurant window. One car’s rear view mirror pointed directly at me about 30 degrees off perpendicular. As cars passed by some of them would reflect in the mirror. But every so often, every 12 or 15 cars, one would generate a spectacular burst of full color spectrum.
I’m trying to see what had to line up perfectly to do this. The sunlight bounces off a passing windshield then passes through the car’s far side window where it gets refracted. Then it reflects off the mirror (which probably has two panes of glass for dimming, so maybe refracted again) then refracts as it passes through the near side window and voila!
This is conjecture, but here goes: I think if the light is always passing through glass having parallel faces, and there’s nothing acting like a grating (such as a CD disk sitting on the dashboard), there’s no mechanism to create a spectrum. So something else must be involved. Perhaps the car rear view mirror is a glass wedge, using the aluminized back as a daytime reflector and the weaker reflection off the first glass surface as a nighttime reflector (that’s how the little lever on the bottom of the mirror does it). I’m not sure how to get the light to go through the wedge an odd number of times, though, and an even number of times should cancel this effect out. If there’s glass or plastic someplace with a bevel, that could do it.
1.) The light can pass through something that acts as a prism, in which the light refracts at least one time. Because of the variation of refractive index with wavelength, you can get a spectrum this way, because the different colors end up going in slightly different directions. This is how your classic prism works.
2.) You can have light travelling through two almost identical paths, but with a slight path length difference between them, so that you get constructive and destructive interference. Because the coherence length of sunlight is EXTREMELY short – only a few wavelengths of light – the path length difference cannot be any greater than this. This is how you get interference colors from oil films floating on water, or from a soap bubble, or from scratches on a windshield, or from Young’s Double Slit experiment, or from diffraction gratings. Because the path length difference has to be so very small, you can rule out any of those suggestions about other cars and interfering.
Sometimes the two cases aren’t so very different. Rainbows are often explained as being due to a prismatic effect – light refracts going into the raindrop and again on coming out. But a full explanation of the rainbow shows that it’s really an interference effect – light waves can take one of two different paths to come out at the same angle, and the paths are of unequal lengths. (although it’s actually even more complex than that, really)
So what effect were you seeing? My guess would be that something was acting like a prism. Retroreflectors are actually made up of tesselated prismatic structures.-- you can see a “rainbow” if you look at sunligh refl;ecting off one at an angle. Red refrlectors in a car won’t gibe you all colors, but a white reflector might. Or light might be reflecting from some other prism in the car.