Thanks for the references. One thing had always puzzled me, and the
Wikipedia answered it. In New Mexico we saw not just one, but 3 or possibly
4 rainbows. The primary and secondary ones were very bright, but inside the
primary one there was another fainter one or possibly 2 fainter one.
A colleague insisted this was impossible, and cited the classical analysis
of rainbows. After looking in some books, I never saw any references to
extra bows on the same side as the main bow. Well, the colleague was right
that classical geometric optics does not predict it, but perhaps this
insistence is why he was never a Nobel candidate. The extra bows are called
supernumerary and are interference effects. It is strange that none of the
texts mention it at least in passing. After seeing this and knowing that
classical analysis did not predict the extra bows I wonder "what would
Feynman do?"