The psychologist studies the mind, and his ultimate appeal must be to introspection, to a direct observation of mental phenomena, and of their relations to external things.  Now, if the observation of mental phenomena were a simple and an easy thing; if the mere fact that we are conscious of sensations and ideas implied that we are clearly conscious of them and are in a position to describe them with accuracy, psychology would be a much more satisfactory science than it is.

But we are not thus conscious of our mental life.  We can and do use our mental states without being able to describe them accurately.  In a sense, we are conscious of what is there, but our consciousness is rather dim and vague, and in our attempts to give an account of it we are in no little danger of giving a false account.

Thus, the psychologist assumes that we perceive both physical phenomena and mental—­the external world and the mind.  He takes it for granted that we perceive mental phenomena to be related to physical.  He is hardly in a position to make this assumption, and then to set it aside as a thing he need not further consider.  Does he not tell us, as a result of his investigations, that we can know the external world only as it is reflected in our sensations, and thus seem to shut the mind up within the circle of mental phenomena merely, cutting off absolutely a direct knowledge of what is extra-mental?  If we can know only mental phenomena, the representatives of things, at first hand, how can we tell that they are representatives? and what becomes of the assumption that we perceive that mind is related to an external world?

It may be said, this problem the psychologist may leave to the metaphysician.  Certainly, it is one of those problems that the metaphysician discusses; it has been treated in Chapter IV.  But my contention is, that he who has given no thought to the matter may easily fall into error as to the very nature of mental phenomena.

For example, when we approach or recede from a physical object we have a series of experiences which are recognized as sensational.  When we imagine a tree or a house we are also experiencing a mental phenomenon.  All these experiences seem plainly to have extension in some sense of the word.  We appear to perceive plainly part out of part.  In so far, these mental things seem to resemble the physical things which we contrast with what is mental.  Shall we say that, because these things are mental and not physical, their apparent extension is a delusion?  Shall we say that they really have no parts?  Such considerations have impelled psychologists of eminence to maintain, in flat contradiction to what seems to be the unequivocal testimony of direct introspection, that the total content of consciousness at any moment must be looked upon as an indivisible, part-less unit.

We cannot, then, depend merely on direct introspection.  It is too uncertain in its deliverances.  If we would make clear to ourselves what mental phenomena really are, and how they | differ from physical phenomena, we must fall back upon the reflective analysis of our experience which occupies the metaphysician (section 34).  Until we have done this, we are in great danger of error.  We are actually uncertain of our materials.