confused. Thus some picture of mountain and lake in a country
which we have never visited, often recalls a vague sense of identity
with much we have seen elsewhere. Our recollections cannot be
disentangled, though general resemblances are recognised. It is also
a fact that the memories of persons who have great powers of
visualising, that is, of seeing well-defined images in the mind's eye,
are no less capable of being blended together. Artists are, as a
class, possessed of the visualising power in a high degree, and they
are at the same time pre-eminently distinguished by their gifts of
generalisation. They are of all men the most capable of producing
forms that are not copies of any individual, but represent the
characteristic features of classes.

There is then, no doubt, from whatever side the subject of memory is
approached, whether from the material or from the mental, and, in
the latter case, whether we examine the experiences of those in whom
the visualising faculty is faint or in whom it is strong, that the
brain has the capacity of blending memories together. Neither can
there be any doubt that general impressions are faint and perhaps
faulty editions of blended memories. They are subject to errors of
their own, and they inherit all those to which the memories are
themselves liable.

Specimens of blended portraits will now be exhibited; these might,
with more propriety, be named, according to the happy phrase of
Professor Huxley, "generic" portraits. The word generic presupposes
a genus, that is to say, a collection of individuals who have much
in common, and among whom medium characteristics are very much more
frequent than extreme ones. The same idea is sometimes expressed by
the word "typical," which was much used by Quetelet, who was the
first to give it a rigorous interpretation, and whose idea of a type
lies at the basis of his statistical views. No statistician dreams
of combining objects into the same generic group that do not cluster
towards a common centre; no more should we attempt to compose
generic portraits out of heterogeneous elements, for if we do so the
result is monstrous and meaningless.

It might be expected that when many different portraits are fused
into a single one, the result would be a mere smudge. Such, however,
is by no means the case, under the conditions just laid down, of a
great prevalence of the mediocre characteristics over the extreme
ones. There are then so many traits