with the left foot, then as a matter
of fact, this form of reaction came about as was described on page 61,
with this difference that there we had to do with voluntary controlled
movements on the part of the questioner, whereas here, they are the
result of an unsuitable degree of tension which expressed itself in
frequent and disconcerting jerks. Besides the answer 3, this so-called
floundering was the only reaction the average person could obtain from
the horse in the absence of Mr. von Osten and Mr. Schillings. It would
however occur also in the case of these gentlemen and would be received
by them with resentment when in truth it was Hans's greatest feat, for
he showed his extremely keen reaction upon every movement of the
questioner. To this group belong also the errors in the case of higher
numbers, the sole cause of which lay in the difficulty with which
tension could be maintained and the body kept motionless for so long a
period. These errors occurred in accordance with a certain law. If, for
instance, a certain test repeatedly evoked incorrect responses, the
questioner would gradually increase the duration of tension and would
thus come a little nearer to the desired goal with every test. In this
way, Mr. von Osten desiring 30 as an answer obtained consecutively the
responses, 25, 28, 30; and I, myself, for the answer 20, received
consecutively the responses 10, 18, 20 (see also the laboratory tests,
page 105). Sometimes too, the questioner would flag in his efforts
before the goal was reached. Thus in one of my first tests, I received
for the answer 11 the following responses: 1, 4, 5, 7, 4. I was unable
to get beyond 7. In other instances, the horse responded first with too
few and then with too many taps. The correct response therefore could
only be obtained after an appreciable amount of gauging of tension, as
in target practice there must be a gauging of distance. (See page 92).
In this way Mr. von Osten obtained for 10 the responses 8, 8, 11, 10,
and Mr. Schillings for 17, received 9, 16, 19, 18, 18, 14, 9, 9, and
finally, after some efforts, 17 taps. Thus there was a rise from 9 to
19, then a fall back to 9 and after eight tests the correct response. As
long as we attempt to explain this fact as error on the part of the
horse, so long will it remain inexplicable, but the moment we regard it
from the point of view of the psychology of the tension of expectation,
it becomes perfectly plain.

The same holds