The hypothesis of 'power' has been really the source of 'illusion.' The tendency to personify leads us to convert metaphor into fact, to invent a subject of this imaginary 'power,' and thus to create a mythology of beings to carry on the processes of nature. In other words, Brown here follows Hume or even anticipates Comte. As J. S. Mill remarks,[471] this erroneous identification of 'power' with 'will' gives the 'psychological rationale of Comte's great historical generalisation'; and, so far, Brown, as a follower of Hume, is clearly on the way to positivism. The world, then, is a vast aggregate of 'loose' phenomena. A contemplation of things reveals no reason for one order rather than another. You may look at your loadstone as long as you please, but you will find no reason for its attracting iron. You may indeed interpolate a number of minute intervening sequences, and the process often suggests a vague something more than sequence; but this is a mere illusion.[472] Could we, in fact, see all the minute changes in bodies we should actually perceive that cause means nothing but 'the immediate invariable antecedence of an event.'[473] Brown especially argues against the attempts of d'Alembert and Euler to deduce the first laws of motion from the principle of 'sufficient reason.'[474] That, as he argues in detail, is merely begging the question, by introducing the principle of causation under an alias. What, then, is the principle? We believe, he says,[475] that 'every event must have a cause,' and that circumstances exactly 'similar must have results exactly similar.' This belief, though applicable to all events, does not give us the 'slightest aid' to determining, independently of experience, any particular event. We observe that B follows A, but, for all we can say, it might as well follow any other letter of the alphabet. Yet we are entitled to say in general that it does uniformly follow some particular letter. The metaphor which describes cause and effect as a 'bond' tying A and B together is perfectly appropriate if taken to express the bare fact of sequence;[476] but we fall into error if we fancy there is really any bond whatever beside the events themselves.