Thus far I have been concerned to show how, starting from a principle such as that like produces like, about which there is nothing magical in the eyes either of those who believe in magic or of those who have left the belief behind, man might evolve the conception of magic as being the lore or the personal power which enables a man to do what ordinary people cannot do. A few words are necessary as to the decline of the belief. The first is that the belief is rotten before it is ripe. Those applications of the principle that like produces like which are magical are generally precisely those which are false. The fact that they are false has not prevented them from surviving in countless numbers to the present day. But some suspicion of their falsity in some cases does arise; and the person who has the most frequent opportunities of discovering their falsity, the person on whose notice the discovery of their falsity is thrust most pointedly, is the person who deals habitually and professionally in magic. Hence, though it is his profession to work wonders, he takes care as far as may be not to attempt impossibilities. Thus Dr. Haddon (_l.c._, p. 62) found that the men of Murray Island, Torres Straits, who made a "big wind" by magic, only made it in the {101} season of the southeast trade wind. "On my asking," he says, "whether the ceremony was done in the north monsoon, my informant said emphatically, 'Can't do it in northwest.' That is, the charm is performed only at that season of the year when the required result is possible--indeed when it is of normal occurrence. In this, as in other cases, I found that the impossible was never attempted. A rain charm would not be made when there was no expectation of rain coming, or a southeast wind be raised during the wrong season." The instance thus given to us by Dr. Haddon shows how the belief in magic begins to give way before the scientific observation of fact. The collapse of magic becomes complete when every one sees that the southeast trade wind blows at its appointed time, whether the magic rites are performed or not. In fine, what kills magic regarded as a means for producing effects is the discovery that it is superfluous, when for instance the desired wind or rain is coming, and futile when it is not. And whereas morality and religion only condemn the end aimed at by magic, and only condemn it when it is anti-social, science slowly shows that magic as a means to any end is superfluous and silly. {102}