the object and the other object or class of objects with which
we identify it. In the case of individual identification such illusions
are, of course, comparatively rare, since here there are involved so
many characteristic differences. On the other hand, in the case of
specific recognition there is ample room for error, especially in those
kinds of more subtle recognition to which I have already referred. To
"recognize" a person as a Frenchman or a military man, for example, is
often an erroneous process. Logicians have included this kind of error
under what they call "fallacies of observation."

Errors of recognition, both specific and individual, are, of course,
more easy in the case of distant objects or objects otherwise
indistinctly seen. It is noticeable in these cases that, even when
perfectly cool and free from emotional excitement, we tend to interpret
such indistinct impressions according to certain favourite types of
experience, as the human face and figure. Our interpretative imagination
easily sees traces of the human form in cloud, rock, or tree-stump.

Again, even when there is no error of recognition, in the sense of
confusing one object with other objects, there may be partial illusion.
I have remarked that the process of recognizing an object commonly
involves an overlooking of points of diversity in the object, or aspect
of the object, now present. And sometimes this inattention to what is
actually present includes an error as to the actual visual sensation of
the moment. Thus, for example, when I look at a sheet of white paper in
a feebly lit room, I seem to see its whiteness. If, however, I bring it
near the window, and let the sun fall on a part of it, I at once
recognize that what I have been seeing is not white, but a decided grey.
Similarly, when I look at a brick viaduct a mile or two off, I appear to
myself to recognize its redness. In fact, however, the impression of
colour which I receive from the object is not that of brick-red at all,
but a much less decided tint; which I may easily prove by bending my
head downwards and letting the scene image itself on the retina in an
unusual way, in which case the recognition of the object as a viaduct
being less distinct, I am better able to attend to the exact shade of
the colour.

Nowhere is this inattention to the sensation of the moment exhibited in
so striking a manner as in pictorial art. A picture of Meissonier may
give the eye a representa