How and when the witness is to be told things there is no rule for. The wise adjustment between saying enough to awaken interest and not too much to cause danger is a very important question of tact. Only one certain device may be recommended--it is better to be careful with a witness during his preliminary examination and to keep back what is known or suspected; thus the attention and interest of the witness may perhaps be stimulated. If, however, it is believed that fuller information may increase and intensify the important factors under examination, the witness is to be recalled later, when it is safe, and his testimony is, under the new conditions of interest, to be corrected and rendered more useful. In this case, too, the key to success lies in increase of effort--but that is true in all departments of law, and the interest of a witness is so important that it is worth the effort. Topic III. PHENOMENOLOGY: STUDY OF THE OUTWARD EXPRESSION OF MENTAL STATES. Section 10. Phenomenology is in general the science of appearances. In our usage it is the systematic co-ordination of those outer symptoms occasioned by inner processes, and conversely, the inference from the symptoms to them. Broadly construed, this may be taken as the study of the habits and whole bearing of any individual. But essentially only those external manifestations can be considered that refer back to definite psychical conditions, so that our phenomenology may be defined as the semiotic of normal psychology. This science is legally of immense importance, but has not yet assumed the task of showing how unquestionable inferences may be drawn from an uncounted collection of outward appearances to inner processes. In addition, observations are not numerous enough, far from accurate enough, and psychological research not advanced enough. What dangerous mistakes premature use of such things may lead to is evident in the teaching of the Italian positivistic school, which defines itself also as psychopathic semiotic. But if our phenomenology can only attempt to approximate the establishment of a science of symptoms, it may at least study critically the customary popular inferences from such symptoms and reduce exaggerated theories concerning the value of individual symptoms to a point of explanation and proof. It might seem that our present task is destructive, but it will be an achievement if we can show the way to later development of this science, and to have examined and set aside the useless material already to hand. Section 11. (a) General External Conditions.