Briefly, I would also mention a few of the more interesting introspective observations which were made in situations in which the horse responded with movements of the head for answers such as "yes" and "no", "up" and "down", etc. From the very beginning I put questions to Hans which would have to be answered by a shake of the head. It often happened that instead of indicating "0", Hans would begin tapping some number. But the wonder of it was that, in many cases, he responded properly. I knew only that I inwardly pronounced the word "null" (zero), and that I looked expectantly at the horse's head. In the case of questions to which I expected the answer "yes" or "no", I imagined myself enunciating the answer, i. e., I used motor imagery. The tests failed, the moment I employed only visual or auditory imagery, whereas, motor imagery was always effective in calling forth correct reactions.[M] When the proper response was "up" and "down" I would think of those directions in space, and likewise with "left" and "right" in which case also I would put myself in the horse's place. [Footnote M: Thus it is possible to think of the word "no" in three different ways. I may get a visual image of the written or printed word, or the auditory image of the word as spoken by another person, or finally I might think of it in terms of images of the sensations of movement which would arise if I myself were to enunciate or write the word. And so, in like manner, I could think of any other word in terms of either visual or auditory or motor imagery. In all probability the auditory and motor always occur together,[6] but still it is possible to make the one or the other predominate. It appears that the imagery of most persons is a mixture of auditory-motor and visual elements, with a predominance of one or the other kind. Individuals who utilize almost exclusively the visual (as does the author, as a rule), are rare. But rarer still is the pronounced motor type.]