Without a glossary explanatory of Kant's terminology, this passage would be hardly intelligible in a translation; but it may be paraphrased thus: All knowledge is founded upon experiences of sensation, but it is not all derived from those experiences; inasmuch as the impressions of relation ("reine Anschauungen"; "reine Verstandesbegriffe") have a potential or _a priori_ existence in us, and by their addition to sense-experiences, constitute knowledge. CHAPTER IV. THE CLASSIFICATION AND THE NOMENCLATURE OF MENTAL OPERATIONS. If, as has been set forth in the preceding chapter, all mental states are effects of physical causes, it follows that what are called mental faculties and operations are, properly speaking, cerebral functions, allotted to definite, though not yet precisely assignable, parts of the brain. These functions appear to be reducible to three groups, namely: Sensation, Correlation, and Ideation. The organs of the functions of sensation and correlation are those portions of the cerebral substance, the molecular changes of which give rise to impressions of sensation and impressions of relation. The changes in the nervous matter which bring about the effects which we call its functions, follow upon some kind of stimulus, and rapidly reaching their maximum, as rapidly die away. The effect of the irritation of a nerve-fibre on the cerebral substance with which it is connected may be compared to the pulling of a long bell-wire. The impulse takes a little time to reach the bell; the bell rings and then becomes quiescent, until another pull is given. So, in the brain, every sensation is the ring of a cerebral particle, the effect of a momentary impulse sent along a nerve-fibre. If there were a complete likeness between the two terms of this very rough and ready comparison, it is obvious that there could be no such thing as memory. A bell records no audible sign of having been rung five minutes ago, and the activity of a sensigenous cerebral particle might similarly leave no trace. Under these circumstances, again, it would seem that the only impressions of relation which could arise would be those of co-existence and of similarity. For succession implies memory of an antecedent state.[24] But the special peculiarity of the cerebral apparatus is, that any given function which has once been performed is very easily set a-going again, by causes more or less different from those to which it owed its origin. Of the mechanism of this generation of images of impressions or ideas (in Hume's sense), which may be termed _Ideation_, we know nothing at present, though the fact and its results are familiar enough.