With the following comment the chapter will be concluded. Introspections are necessarily subjective in character. If they are to possess general validity, they must be borne out by evidence furnished by others--and this to a greater extent than is necessary for other forms of observation. It was hardly possible to get corroboration from the other persons who had worked with Hans, for, although some of them were excellent observers of external natural phenomena, few of them had had the necessary amount of practice in introspection. The necessary confirmation, however, was had in laboratory tests, which we shall presently describe. CHAPTER IV LABORATORY TESTS The tests which are to be briefly reported here, were begun in November, 1904, and were carried out at the Psychological Institute of the University of Berlin. The purpose was twofold: first, to discover whether the expressive movements noted in Mr. von Osten, Mr. Schillings, and others, were to be regarded as typical and to be found in the majority of individuals,--and secondly, to ascertain in how far the psychical processes which I had noted in my own case and which I believed to lie at bottom of these movements, were paralleled in, and confirmed by, the introspections of others. The effort was made to make the experimental conditions as nearly as possible like those under which the horse had worked. The affective atmosphere which colored the situations in which the horse took part, could not, of course, be transferred, but this was in some respects an advantage. One person undertook the role of questioner, another--myself--that of the horse. The experiments fall into three groups, corresponding to the types of the horse's reactions: 1, tests in counting and computation; 2, tests in space reactions; 3, tests in fetching or designating objects.