> I don't remember the citation on this--it was a few years back--
> but Eva-Lise is _not_ making this up out of whole cloth (and,
> if you'd read her post, is not attempting to discredit Mendel
> or discard his life's work). Someone relatively recently went
> back and looked at Mendel's data, and they are suspiciously
> good. What Mendelian genetics actually tells us is that, *on
> average*, if you have a parent with two genes AA and one with
> two genes aa, and they have four children, you'll get one AA,
> two Aa, and one aa. That doesn't mean you'll get that exact
> assortment with every set of four children, just as it's entirely
> possible for a fair coin to come up heads five times in a row
> (and any statistician will be suspicious if you claim to have
> tossed a coin 100 times, and had it come up heads, tails,
> heads, tails, and so on 50 times).
> As an earlier poster said, I don't think Mendel was consciously
> cheating, and he was right about the broad point: but he seems
> to have fudged the data.
Mendal did fudge his data a *little*. That was because Mendal
didn't know anything about recessive and dominant genes. His
still valid point was that reproduction was not random and gene
research is still confirming that it is not random.