That Magic should depend upon the assumption that “things connected in thought are connected in fact” seems unintelligible, until we consider that the thought in which they are connected is a judgment concerning the facts. But how is it possible to judge so foolishly? Absurd association is intelligible; it cannot be helped; but judgment is not a passive process. Well: a logician might interpret the case as a fallacy in applying the inductive canons: the Esquimo had two instances: (_a_) no ham-bone, no seal; (_b_) ham-bone, seal: or, again, into the circumstances of his expedition he introduced a ham-bone, and the seal followed. Therefore, the carrying of a ham-bone was the cause of getting a seal. He had not read Mill, and did not know what precautions should be taken before adopting such conclusions. The mind must have begun in this way; and after many thousands of years, the opportunities of error in empirical judgment began to be appreciated, and the canons were formulated. The experience of every man at every stage of life, personal or social (probably evoking an inherited disposition), continually impresses him with the belief in some connexion of antecedent and consequent, that each event arises out of others. But what it is in the antecedent that determines the event can never, in practical affairs, be exactly known. In definite cases, where method is applicable, we may analyse the consequent into the tendencies of forces in amount and direction; or we may sometimes reproduce it by an experiment exactly controlled. But where these resources fail us, as they often do in practical affairs, we are little confident of grasping all, or even the chief conditions of any event that interests us. The savage knows nothing of method, and, therefore, feels less distrust. He learns something of the conditions of success in hunting; this is biologically necessary to him as it is to a tiger; only those survive that learn them. He knows that he must have weapons as efficient as possible, must go to the habitat of his prey, must not be heard, or seen, or smelt. But with all precautions he sometimes fails; therefore there must be some other condition which he does not understand. Hence anything observed, or any word or action, that happens to be followed by success, may have been a means to that success.