suggestibility, or perception, or memory, and so on, it would be rather easy to foresee the behavior of the individual in the act of decision, as every one of those other simple mental functions could be tested by routine methods of the psychological laboratory.  This consideration led me to propose ramified investigations concerning the psychology of decision in its relation to the elementary mental processes.  These studies by students of the laboratory are not yet completed.  But I soon saw that they would be unfit for the solution of my practical problem, as we recognized that these relations between the complex act of decision and the elementary functions of the individual seem to have different form with different types of men.[12] If I was to approach the solution of the practical problem, accordingly, I had to reproduce in an experimental form the act of decision under complex conditions.

It seemed necessary to create a situation in which a number of quantitatively measurable factors were combined without any one of them forcing itself to consciousness as the most important.  The subject to be experimented on then has to decide as quickly as possible which of the factors is the relatively strongest one.  As usual, here, too, I began with rather complicated material and only slowly did I simplify the apparatus until it finally took an entirely inconspicuous form.  But this is surely the most desirable outcome for testing methods which are to be applied to large numbers of persons.  Complicated instruments, for the handling of which special training is needed, are never so useful for practical purposes as the simple schemes which can be easily applied.  The form of which I finally made use is the following.  I work with 24 cards of the size of playing-cards.  On the upper half of every one of these cards we have 4 rows of 12 capital letters, namely, A, E, O, and U in irregular repetition.  On 4 cards, one of these vowels appears 21 times and each of the three others 9 times; on 8 cards, one appears 18 times and every one of the three others 10 times; on 8 cards, one appears 15 times and each of the others 11 times; and finally, on 4 cards one vowel appears 16 times, each of the three others 8 times, and besides them 8 different consonants are mixed in.  The person to be tested has to distribute these 24 cards as quickly as possible in 4 piles, in such a way that in the first pile are placed all cards in which the letter A is most frequent, in the second those in which the letter E predominates, and so on.  As a matter of course the result must never be secured by counting the letters.  Any attempt to act against this prescription and secretly to begin counting would moreover delay the decision so long that the final result would be an unsatisfactory achievement anyhow.  It would accordingly bring no advantage to the candidate.