choices being characterized by deliberateness and definite search for
the right box. Uncertainty was plainly indicated, and in this the
behavior of the animal differed markedly from that in the concluding
series of the regular experiment.

It seems safe to conclude from the results of these control series that
Sobke has no free idea of the relation of secondness from the right and
is chiefly dependent upon memory of the particular settings for cues
which lead to correct choice.


_Problem 3. Alternately First at Left and First at Right_

For four successive days after the last control series in connection
with problem 2, Sobke was merely fed in the apparatus according to
previous description (p. 43). He exhibited a wonderfully keen appetite
and was well fed during this interval between problems.

The method of experimentation chosen for problem 3 in the light of
previous experience was that of confining the monkey for a short time,
ten to fifteen seconds, in the wrong box, in each of the first ten
mistakes for a given trial, and of then aiding him to find the right box
by the slight and momentary raising of the exit door. Aid proved
necessary in a few of the trials during the first four days. Then he
worked independently. As work progressed it was found possible and also
desirable to increase the period of confinement, and in the end, sixty
seconds proved satisfactory. It was also thought desirable to increase
the number of trials per day from a single series during the early days
to two or even three series from June 29 on. Often three series could be
given in succession without difficulty. During the early trials on this