It might almost seem as if the task the jury trial sets the judge has not been clearly perceived. A judge who thinks he has performed it when he has cast before the jury the largest possible mass of testimony, more or less reviewed, and who sees how people, who perhaps for the first time in their lives, are involved in a court of law, who perhaps see a criminal for the first time, and are under these circumstances the arbiters of a man’s fate,--a judge who sees all this and is satisfied, is not effective in his work. Nowadays more than ever, it is for the judge to test all evidence psychologically, to review what is only apparently clear, to fill out lacunae, and to surmount difficulties, before he permits the material brought together in a very few hours to pass into the jury’s hands. According to Hillebrand, much that seems “self-evident” shows itself dependent on definite experience attained in the process of hundreds of repetitions in the daily life; the very impression of self-evidence is frequently produced by a mere chance instinct about what should be held for true. Hume has already shown how the most complex and abstract concepts are derived from sensation. Their relation must be studied, and only when we can account for every psychic process with which we have to concern ourselves, is our duty properly fulfilled. Section 23. (2) Proof. Mittermaier[127] holds that “as a means of testimony in the legal sense of that term every possible source must be examined which may suffice the judge according to law. And from such examination only may the requisite certainties be attained from which the judge is to assume as determined, facts relevant to his judgment.” Only the phrase “according to law” needs explanation, inasmuch as the “source” of reasons and certainties must satisfy the legal demands not only formally but must sustain materially every possible test, whether circumstantial or logico-psychologic. If, for example, the fundamental sources should be a combination of (1) a judicial examination of premises (lokalaugenschein), (2) testimony of witnesses, and (3) a partial confession, the requirements of the law would be satisfied if the protocol, (1), were written or made according to prescribed forms, if a sufficient number of properly summoned witnesses unanimously confirmed the point in question, and if finally the confession were made and protocoled according to law. Yet, though the law be satisfied, not only may the conclusion be wholly false but every particular part of the evidence may be perfectly useless, without the presence anywhere of intentional untruth. The personal examination may have been made by a judge who half the time, for some sufficiently cogent reason, had a different conception of the case than the one which later appeared to be true. It need not have been necessary that there should be mixed therewith false information of witnesses, incorrect observation, or such other mistakes.