> Genetic changes do not occur in response to selection.  They occur
>  randomly, as a result of internal errors and exposure to mutagens.

There's an interesting article in the December 1999 Discover magazine
concerning "transposons," which are bits of DNA that can relocate
themselves in the chain, changing the properties of the resulting DNA.
It thus appears that genetic variability is not an accident or error;
it's built into the normal function of DNA. The particular results, on
the other hand, are more or less random, and many genetic diseases are
thought to have arisen this way. Given the way the world works,
changes that give a selective advantage are no doubt exceedingly rare.

Bill Dooley