The notion of analogue in science I think is absolutely fundamental. Two systems are analogous if, indeed, they do obey the same equations. Now, how did I know that the mechanical and the electrical system obeyed the same equations? Essentially because one knew enough about electricity to... to know how electric circuits behaved. I mean, let's face it, the... the behaviour of a tuning circuit in your radio or in the television, the mathematical equations describing it are identical to a weight bumping about on the end of a spring. And they are identical because we know the... the physics of electrical circuits, we know the physics of springs and weights, and when you write down the equations, they turn out to be the same. So you can... you can draw... and what mathematics does for you is to enable you to recognise the analogy between different systems. I think there's almost nothing a scientist can know which is more important than that; to know, you know, that it's really... once you know the mathematics, you get the behaviour, and you can use one system to tell you how another system will behave. I remember being so cross - this was years later after I was in London working in biology - I learnt about Cybernetics and what they'd done at MIT during the war, who had formalised all this, and that what cybernetics is, is a mathematical formulisation of the analogies between machines. And I said, 'Christ, I know that.' But I only knew it in the particular, you see. It's different because I knew there was an analogy between mechanical behaviour and electrical behaviour, but I'd not make the generalisation to the fact that you can draw an analogy to genetical behaviour, or you can have an analogue of psychological behaviour, I just knew it in one particular case. And what the people at MIT had done is to generalise and say: Analogues can be made between any two systems provided you know the equations are the same. But it's a very important idea, and I was so cross I hadn't thought of it.