It i likley that a magician is cheating rather than disproving, say, the second law of thermodynamics but the reason for saying so is not rigorously the implied defience of the Second Law; it is that the cheating can only be truly identified and exposed as such by empirical observation of what the magician actually does.

Our confidence in the second law, confidence based upon a vast array of data of many kinds, merely serves to suggest that a bet on it would be worth placing. It is simply non-scientific to assert that what the magician did is absolutely known to be cheating because of the sanctity of the second law. People who think like that are not thinking as scientists..

To make this distinction is not to open the floodgates to a torrent of rubbish; it is to say that a hypothetical possibility cannot be dismissed solely on the ground that it is at variance with what we think we know. We may think wrong..

I imagine that in the 1480's there was no shortage of people rubbishing Columbus; "India is to our east so clearly it is rubbish to suppose that you can get there by sailing west - it will serve you right if you fall off the edge". And it is not only among pre-scientists that such attitudes can be observed. In 1808 John Dalton, the father of modern atomic theory, asserted to the effect that "you might as well try to put a new planet into the solar system as to split an atom." .

Quite so! .

By 1958, and granted a certain flexibility in the definition of 'planet', we had done both of these pre-rubbished things..

Scepticism, like charity, should begin at home but not, of course, end there; we should not be at our least sceptical when contemplating things that we are especially sure of - our current 'certainties'. .