This article attempts to give a description of knowing as it actually occurs,--not how it originated nor how it is antecedently possible. The thesis is that an idea knows an external reality when it points to it, resembles it, and is able to affect it. The plan of exposition is to start with the simplest imaginable material and then gradually introduce additional matter as it is needed until we have cognition as it actually occurs. James postulates a single, momentarily-existing, floating feeling as the entire content, at the instant, of the universe. What, then, can this momentary feeling know? Calling it a 'feeling of _q_', it can be made any particular feeling (fragrance, pain, hardness) that the reader likes. We see, first, that the feeling cannot properly be said to know itself. There is no inner duality of the knower on the one hand and content or known on the other. "If the content of the feeling occurs nowhere else in the universe outside of the feeling itself, and perish with the feeling, common usage refuses to call it a reality, and brands it as a subjective feature of the feeling's constitution, or at most as the feeling's dream. For the feeling to be cognitive in the specific sense, then, it must be self-transcendent". And we must therefore "create a reality outside of it to correspond to the intrinsic quality _q_". This can stand as the first complication of that universe. Agreeing that the feeling cannot be said to know itself, under what conditions does it know the external reality? James replies, "If the newly-created reality _resemble_ the feeling's quality _q_, I say that the feeling may be held by us to be _cognizant of that reality_". It may be objected that a momentary feeling cannot properly know a thing because it has no _time_ to become aware of any of the _relations_ of the thing. But this rules out only one of the kinds of knowledge, namely "knowledge about" the thing; knowledge as direct acquaintance remains. We may then assert that "if there be in the universe a _q_ other than the _q_ in the feeling the latter may have acquaintance with an entity ejective to itself; an acquaintance moreover, which, as mere acquaintance it would be hard to imagine susceptible either of improvement or increase, being in its way complete; and which would oblige us (so long as we refuse not to call acquaintance knowledge) to say not only that the feeling is cognitive, but that all qualities _of feeling, so long as there is anything outside of them which they resemble_, are feelings of qualities of existence, and perceptions of outward fact". But this would be true, as unexceptional rule, only in our artificially simplified universe. If there were a number of different _q's_ for the feeling to resemble, while it meant only one of them, there would obviously be something more than resemblance in the case of the one which it did know. This fact, that resemblance is not enough in itself to constitute knowledge, can be seen also from