Even in the case of Mill, it sometimes appears as if an inkling of the immense difference had begun to dawn upon him, as when, in a note to his _Analysis of the Phenomena of the Human Mind_ (vol. i., chap. xi. p. 407), in criticizing his father’s theory, he says: “If belief is only an inseparable association, belief is a matter of _habit_ and accident and not of reason. Assuredly an association, however close, between two ideas is not a sufficient _ground_ (the italics are his own) of belief; it is not evidence that the corresponding facts are united in external nature. The theory seems to annihilate all distinction between the belief of the wise, _which is regulated by evidence_ and conforms to the real successions and co-existences of the facts of the universe, and the belief of fools _which is mechanically produced_ by any accidental association that suggests the idea of a succession or co-existence to the mind; a belief aptly characterized by the popular expression, believing a thing because they have taken it into their heads.” This is all excellent. But it is robbed of its most essential worth, when, in a later note (vol. i. p. 438. note 110) we hear J. S. Mill say: “It must be conceded to him (the author of the _Analysis_) that an association sufficiently strong to exclude all ideas that would exclude itself, _produces a kind of mechanical belief_, and that the processes by which the belief is corrected, or reduced to rational bounds, _all consist in the growth of a counter-association_ tending to raise the idea of a disappointment of the first expectation, and as the one or the other prevails in the particular case, the belief or expectation exists or does not exist exactly as if the belief were the same thing with the association,” and so on.