There is still another great danger which one may beware of, optima fide,--the danger of knowing something untrue. This danger also is greatest for the greatest talent and the greatest courage among us, because they are the readiest hands at synthesis, inference, and definition of possibilities, and see as indubitable and shut to contradiction things that at best are mere possibilities. It is indifferent to the outcome whether a lie has been told purposely or whether it has been the mere honest explosion of an over-sanguine temperament. It is therefore unnecessary to point out the occasion for caution. One need only suggest that something may be learned from people who talk too much. The over-communicativeness of a neighbor is quickly noticeable, and if the why and how much of it are carefully studied out, it is not difficult to draw a significant analogy for one’s own case. In the matter of secrets of other people, obviously the thing to be established first is what is actually a secret; what is to be suppressed, if one is to avoid damage to self or another. When an actual secret is recognized it is necessary to consider whether the damage is greater through keeping or through revealing the secret. If it is still possible, it is well to let the secret be--there is always damage, and generally, not insignificant damage, when it is tortured out of a witness. If, however, one is honestly convinced that the secret must be revealed--as when a guiltless person is endangered--every effort and all skill is to be applied in the revelation. Inasmuch as the least echo of bad faith is here impossible, the job is never easy. The chief rule is not to be overeager in getting at the desired secret. The more important it is, the less ought to be made of it. It is best not directly to lead for it. It will appear of itself, especially if it is important. Many a fact which the possessor had set no great store by, has been turned into a carefully guarded secret by means of the eagerness with which it was sought. In cases of need, when every other means has failed, it may not be too much to tell the witness, cautiously of course, rather more of the crime than might otherwise have seemed good. Then those episodes must be carefully hit on, which cluster about the desired secret and from which its importance arises. If the witness understands that he presents something really important by giving up his secret, surprising consequences ensue.