tioner. If such is the case, they certainly cannot
be considered astonishing. Thus, when to the question, "How many of the
gentlemen present are wearing straw hats?" the horse answers correctly
in accordance with the wording of the question and omits the straw hat
of a lady, then Mr. von Osten is the guide. It is no wonder that Hans
never showed the slightest excitement when confronted with difficult
problems, nor that it apparently took no time whatever to solve them.

Hans, however, was also a faithful mirror of all the errors of the
questioner. Aside from mistakes due to occasional interruptions on the
part of visitors, these errors had two sources: faulty computation and
inadequate concentration--i. e., aside from arithmetical errors on the
part of the questioner, were his premature or belated movements. Since
both of these factors might be operative, the following three
possibilities arise.

(_a_) The questioner computes correctly but does not move at the proper
moment. Nearly all the errors which had been accredited to the horse,
were of this kind.

A part of these errors had the appearance of being significant, that
is, they might be interpreted as a misapprehension of the question. If,
for instance, instead of a sum only one of the quantities was given, or,
if instead of a product only one of the factors was given, it might be
interpreted that the horse simply wished to repeat the problem. Thus,
Mr. von Osten in response to the question: "How much is 3 times 5?",
twice in succession received the answer, "3", and upon my question, "How
much is 3 plus 4?" he answered, "3", and to "How much is 2 times 6?" he
tapped 6, and to "What is one-fourth of 36?" 4. In part (certainly in
the second and third example cited) an individual quantity or factor had
been emphasized in the consciousness of the questioner (cf. page 105)
and in part the reactions were due to chance. Thus, when Mr. Hahn asked
the question: "What is one-half of 10?", he received the following
responses: 2 and 10, and then 17 and 3. To this class belong also, the
tests made by the Commission of September and reported in Supplement
III. (See page 255).

Other errors, even though they may not have appeared to be significant,
might yet have been characterized as mistakes due to speed; as when, e.
g., Hans made an error of one unit--and sometimes, though less
frequently, of two units--too much or too little in his response. One
might be led to believe tha