estion therefore became superfluous. On the other
hand the signs for head-movements on the part of the horse, were given
while the question was being put. I ask, which way is "upward", and at
the same time I look upward. In this case therefore the question itself
is not entirely insignificant.--I experienced greater difficulty in
getting Hans to respond with the head-movement to the left. After much
practice I was able to evoke this movement by means of giving the
command aloud, but never by means of the "silent" command. Accidentally
I hit upon a device by means of which I attained this end also. I asked
the horse aloud "Which direction is left?",--whereupon he reacted
properly; then I immediately repeated the question silently, and was
successful every time. My mental attitude here was still the same as
when I put the question aloud. What sort of an attitude this was, I
could not, of course, have stated explicitly at the time. I could not,
therefore, awaken it at will,--and if I allowed but a minute to elapse
between the spoken and the silent question, the vivid after-effect (the
so-called "primary memory image") soon disappeared and the test was
wholly unsuccessful. Practice, however, soon helped me to overcome this
last difficulty also. I believe that my inability to evoke this specific
reaction on the part of the horse, lay in the unfavorable position which
I assumed, for it did not allow the horse to perceive my movements
easily. For the same reason, Hans would at first indicate "no" and
"zero" by turning to the right, seldom to the left.

As in the case of counting, a high degree of concentration was also
necessary here, but with this difference, that here attention was
directed to ideas present to the mind, ("yes", "no", etc.), whereas in
the counting process attention was directed toward expected sensory
impressions (i. e., the taps of the horse).

All that has been said thus far is readily understood psychologically.
The following curious fact, however, is noteworthy. Hans used the
head-movement to indicate two such different concepts as "zero" and
"no"; it appeared therefore that in both cases he was receiving the same
kind of directive. Observation proved that such was the case and the
directive in question was none other than an imitation in miniature, or
rather a movement anticipatory of the expected head-movement of the
horse. Now, whereas the signs for "up", "down", "right", and "left" were
natural expr