and swim away. That is to say, it has several ways of
reacting to the stimulus and seeking equilibrium, and if
one fails it tries another. But now, when it anchors itself
again, it seems to have learnt something, for if again
touched it does not go through the stages of contracting
and of bending aside. It keeps to its more radical remedy,
and moves off again. The effect is very transient, but it
is there, and is amply verified for Protozoa and for several
of the lower Metazoa. Observe that precisely what is
learnt is not to avoid some object previously sought, or
to seek some object previously indifferent. What is tem-
porarily learnt is to prefer one type-reaction to another --
the stronger, the more effective, but also that which the
animal is normally least ready to put into action. The
relation of data -- which, if not apprehended by the organ-
ism, has somehow affected it -- is the failure of response
A to remove an annoyance B, or conversely the success of
response C in this respect. It need not learn anything
about the qualities or relations of B. What has affected it
is the success of one of its type-reactions in relation to B.
The stimulus B becomes the basis of a different type-
reaction from that which it at first called forth. Correlation
is effected between a stimulus and a certain type-reaction.
Such correlation is observable in human action (though
often in combination with higher modes of response) in the
acquisition of skill, which is essentially a modification of
the means by which a result is attained. We saw that in
skill of any kind, though consciousness plays a part, the
process of acquisition is in detail unconscious. We can
now see that the general method is that of selective correla-
tion. The too violent response lands us in a fall or
punishes us with a cut. The insufficient movement ends
in a bruise on the other side. If there is no bump or cut,
there is at least the disappointment of failure, while the
chance response that hits the mark is encouraged by the
sudden joy of success. The result is that the successful
movements are selected and the failures eliminated. It is
reached by a process which we seek in vain to bring within
our consciousness. Yet conscious experience has some-
how operated to make this accuracy of execution possible.

-63-