It is true that, on ordinary questions, the truth is always uppermost in the subjective mind. A hypnotic subject will often say, during the hypnotic sleep, that which he would not say in his waking moments. Nevertheless, he never betrays a vital secret. The reason is obvious to those who have followed the line of argument in the preceding pages of this chapter. The instinct of self-preservation, always alert to avert any danger which threatens the individual, steps in to his defence. Instinctive auto-suggestion here plays its subtle _rôle_, and no suggestion from another can prevail against it. If the defence involves falsehood, a falsehood will be told, without the slightest hesitation; and it will be told with preternatural acumen, and with such plausible circumstantiality of detail as to deceive the very elect. Neither will there be any variance or shadow of turning after repeated experiments, for the memory of the subjective mind is perfect. This rule holds good, not only with regard to secrets which involve the personal safety of the individual, but in all matters pertaining to his material interests, his reputation, or the interests of his friends, whose secrets are confided to his care. That this is true is presumptively proved by the fact that in all the years during which the science of hypnotism has been practised, no one has ever been known to betray the secrets of any society or order. The attempt has often been made, but it has never succeeded. The truth of this assertion can be demonstrated at any time by experiment. Such an experiment has a greater evidential value in establishing the rule than almost any other laboratory experiment. A subject might plunge a paper dagger into an imaginary man, or he might draw a check, sign a note, a contract, or a deed, in obedience to experimental suggestions, when he would not commit a real crime, or sign away his birthright, in obedience to criminal suggestion. But when a subject is asked to betray the secrets of a society to which he belongs, it is quite a different matter. In the one case a compliance with the suggestion proves nothing, simply because it is a laboratory experiment. In the other case his refusal to comply with the suggestion proves everything, because his betrayal of such a secret in the laboratory is just as vital as to betray it elsewhere.