This is rather a good example of it.
Darwin looking into whether blondes were more or less likely to get married and have children than darker haired women. Thus, was the population going to become darker haired over time?
Now, from memory, this is before we (that is, the world in general) found out about Mendel\’s work on peas and thus knew about the inheritance characteristics of dominant and recessive genes (looking it up I see I\’m right, and that Darwin\’s alternative theory was wrong).
So, he tries to collect the data by teaming up with a doctor and studying the women who attended a hospital. Then:
"Eventually Darwin came to the conclusion that the experimental basis was not good enough. Both Beddoe and Darwin came to the conclusion that the original results were misleading and didn\’t make sufficient allowance for the darkening of hair with age."
That\’s good science that is….theory, collect evidence, realise that the evidence cannot be used to prove the theory one way or the other….then drop theory.
Sure, it would be wonderful to be able to prove such things one way or another, but the science consists of the methods by which you provide such proofs. And if you can\’t provide the proof in a scientific manner than you\’re not able to provide the proofs.