m any one part of the cortex to many other parts.
These fibers are provided by native constitution, but probably
terminate rather loosely in the cortex until exercise has developed
them. They may be compared to telephone wires laid down in the cables
through the streets and extending into the houses, but still requiring
a little fine work to attach them properly to the telephone
instruments.

[Illustration: Fig. 64.--Diagram for the learning of the name of an
object, transformed into a neural diagram. The vocal movement of
saying the name is made in response to the auditory stimulus of
hearing the name, but when the neurone in the "speech center" is thus
made active, it takes up current also from the axon that reaches it
from the visual center, even though the synapse between this axon and
the speech neurone is far from close. This particular synapse between
the visual and the speech centers, being thus exercised, is left in an
improved condition. Each neurone in the diagram represents hundreds in
the brain, for brain activities are carried on by companies and
regiments of neurones. (Figure text: object seen, visual center name
heard, auditory center, speech center, name spoken)]

The diagrams illustrating different cases under the law of combination
can easily be perfected into neural diagrams, though, to be sure, any
diagram is ultra-simple as compared with the great number of neurones
that take part in even a simple reaction.

The reader will be curious to know now much of this neural
interpretation of our psychological laws is observed fact, and how
much speculation. Well, we cannot as yet {417} observe the brain
mechanism in actual operation--not in any detail. We have good
evidence, as already outlined, for growth of the neurones and their
branches through exercise.

[Illustration: Fig. 65.--Control, in multiplying. The visual stimulus
of two numbers in a little column, has preformed linkages both with
the adding response and with that of multiplying. But the mental set
for adding being inactive at the moment, and that for multiplying
active (because the subject means to multiply), the multiplying
response is facilitated.]

We have perfectly good evidence of the law of "unitary response to
multiple stimuli" from the physiological study of reflex action; and
we have perfectly good anatomical evidence of the convergence and
divergence of neural paths of connection, as required by the law of
combinatio