For the same reason that everything interests us more according to the degree it involves us personally, we do not examine facts and completely overlook them though they are later shown to be unshakable, without our being able to explain their causal nexus. If, however, we know causes and relationships, these facts become portions of our habitual mental equipment. Any practitioner knows how true this is, and how especially visible during the examination of witnesses, who ignore facts which to us seem, in the nature of the case, important and definitive. In such cases we must first of all not assume that these facts have not occurred because the witness has not explained them or has overlooked them; we must proceed as suggested in order to validate the relevant circumstances by means of the witness--i.e., we must teach him the conditions and relationships until they become portions of his habitual mental machinery. I do not assert that this is easy--on the contrary, I say that whoever is able to do this is the most effective of examiners, and shows again that the witness is no more than an instrument which is valueless in the hands of the bounder, but which can accomplish all sorts of things in the hands of the master. One must beware, however, of too free use of the most comfortable means,--that of examples. When Newton said, “In addiscendis scientiis exempla plus prosunt, quam praecepta,” he was not addressing criminalists, but he might have been. As might, also, Kant, when he proved that thinking in examples is dangerous because it allows the use of real thinking, for which it is not a substitute, to lapse. That this fact is one reason for the danger of examples is certain, but the chief reason, at least for the lawyer, is the fact that an example requires not equality, but mere similarity. The degree of similarity is not expressed and the auditor has no standard for the degree of similarity in the mind of the speaker. “Omnis analogia claudicat” is correct, and it may happen that the example might be falsely conceived, that similarity may be mistaken for equality, or at least, that there should be ignorance of the inequality. Examples, therefore, are to be used only in the most extreme cases, and only in such wise, that the nature of the example is made very clearly obvious and its incorrectness warned against.