Why is there "ringing" at the violet end of a rainbow but not the red end? I've recently noticed that if you look closely at the violet end of a rainbow, you can see a sort of "ringing" effect where there are alternating bands of color and lack of color. You can apparently photograph this:
 
Zoom in closely to the violet end of the rainbow to see it.
What causes this? And why doesn't the same effect produce "ringing" on the red side of the rainbow?
 A: On rainbow formation: https://en.wikipedia.org/wiki/Rainbow#Supernumerary_rainbow . See a bit more about supernumeraries here http://www.atoptics.co.uk/rainbows/supers.htm, and http://www.atoptics.co.uk/rainbows/supers.htm . 
Classical rainbow is explained by geometric optics, which applies on big enough droplets. When these approach the scale of light wavelength, geometric optic get less and less valid, replaced by wave optics. Smallest droplets lies at the limit, giving the quite complicated Mie scattering profile https://en.wikipedia.org/wiki/Mie_scattering 
Look as this nasty profile - and it's a log plot ! ( The 2 lobes at 180 +- 40° are the classical geometry rainbow. )

As a wave diffraction effects it oscillates, through interference bands that are very wavelength-dependant and thus appear colored. On general clouds are made of drops having different sizes, which tend to blur out these oscillations at averaging. But when droplets do have quite the same size, then the oscillation stands and show as the supernumerary.
