In drawing up this rough sketch of the illusions of introspection, I have had no practical object in view. I have tried to look at the facts as they are apart from any conclusions to be drawn from them. The question how far the liability to error in any region of inquiry vitiates the whole process is a difficult one; and the question whether the illusions to which we are subject in introspection materially affect the value of self-knowledge as a whole and consequently of the introspective method in psychology, as many affirm, is too subtle a one to be fully treated now. All that I shall attempt here is to show that it does not do this any more than the risk of sense-illusion can be said materially to affect the value of external observation. It is to be noted first of all that the errors of introspection are much more limited than those of sense-perception. They broadly answer to the slight errors connected with the discrimination and recognition of the sense-impression. There is nothing answering to a complete hallucination in the sphere of the inner mental life. It follows, too, from what has been said above, that the amount of active error in introspection is insignificant, since the representation of a feeling or belief is so very similar to the actual experience of it. In brief, the errors of introspection, though numerous, are all too slight to render the process of introspection as a whole unsound and untrustworthy. Though, as we have seen, it involves, strictly speaking, an ingredient of representation, this fact does not do away with the broad distinction between presentative and representative cognition. Introspection is presentative in the sense that the reality constituting the object of cognition, the mind's present feeling, is as directly present to the knowing mind as anything can be conceived to be. It may be added that the power of introspection is a comparatively new acquisition of the human race, and that, as it improves, the amount of error connected with its operation may reasonably be expected to become infinitesimal. It is often supposed by those who undervalue the introspective method in psychology that there is a special difficulty in the detection of error in introspection, owing to the fact that the object of inspection is something individual and private, and not open to common scrutiny as the object of external perception. Yet, while allowing a certain force to this objection I would point out, first of all, that even in sense-perception, what the individual mind is immediately certain of is its own sensations. The relatively perfect certainty which finally attaches to the presentative side of sense-perception is precisely that which finally attaches to the results of introspection.