It might seem at a first glance to follow from this four-fold scheme of
immediate or quasi-immediate knowledge that there are four varieties of
illusion. And this is true in the sense that these four heads cover all
the main varieties of illusion. If there are only four varieties of
knowledge which can lay any claim to be considered immediate, it must be
that every illusion will simulate the form of one of these varieties,
and so be referable to the corresponding division.

But though there are conceivably these four species of illusion, it does
not follow that there are any actual instances of each class
forthcoming. This we cannot determine till we have investigated the
nature and origin of illusory error. For example, it might be found that
introspection, or the immediate inspection of our own feelings or mental
states, does not supply the conditions necessary to the production of
such error. And, indeed, it is probable that most persons, antecedently
to inquiry, would be disposed to say that to fall into error in the
observation of what is actually going on in our own minds is
impossible.

With the exception of this first division, however, this scheme may
easily be seen to answer to actual phenomena. That there are illusions
of perception is obvious, since it is to the errors of sense that the
term illusion has most frequently been confined. It is hardly less
evident that there are illusions of memory. The peculiar difficulty of
distinguishing between a past real event and a mere phantom of the
imagination, illustrated in the exclamation, "I either saw it or dreamt
it," sufficiently shows that memory is liable to be imposed on. Finally,
it is agreed on by all that the beliefs we are wont to regard as
self-evident are sometimes erroneous. When, for example, an imaginative