n others
and that they would recognize a manifoldness of variations where the
others would see only uniformity. In that I silently presupposed that
the perception of the uniformity must be something disturbing and
disagreeable and the recognition of variations something which
stimulates the mind pleasantly. But when I came to examine the
question experimentally, I became convinced that such a hypothesis is
erroneous, and if I interpret the results correctly, I should say that
practically the opposite relation exists. Those who recognize the
uniformities readily are not the ones who are disturbed by them.

I proceeded in the following way. To make use of a large number of
subjects accustomed to intelligent self-observation, I made the first
series of experiments with the regular students in my psychology
lecture course in Harvard University. Last winter I had more than four
hundred men students in psychology who all took part in that
introductory series. The task which I put before them in a number of
variations was this: I used lists of words of which half, or one more
or less than half, belonged to one single conceptional group. There
were names of flowers, or cities, or poets, or parts of the body, or
wild animals, and so on. The remaining words of the list, on the other
hand, were without inner connection and without similarity. The
similar and the dissimilar words were mixed. The subjects listened to
such a list of words and then had to decide without counting from the
mere impression whether the similar words were more or equally or less
numerous than the dissimilar words. In other experiments the
arrangement was that two different lists were read and that in the two
lists a larger or smaller number of words were repeated from the first
list. Here, too, the subjects had to decide from the mere impression
whether the repeated words were in the majority or not. In every
experiment the judgment referred to those words which belonged to the
same group and which were in this sense uniform, or to the repeated
words, and it had to be stated with reference to them whether their
number was larger, equal to, or smaller than the different words. If
all replies had been correct, the judgment would have been 40 per cent
equal, 30 per cent smaller, and 30 per cent larger, as they were
arranged in perfect symmetry. As soon as I had the results from the
students, we figured out for every one what number he judged equal,
smaller,