There is, however, a further consideration which we should perhaps take into account. Man, when he has a desire which he wishes to realise,--and the whole of our life is spent in trying to realise what we wish,--takes all the steps which experience shows to be necessary or reason suggests; and, when he has done everything that he can do, he may still feel that nothing is certain in this life, and the thing may not come off. Under those circumstances he may, and often does, pray that success may attend his efforts. Now Dr. Frazer, in the second edition of his _Golden Bough_, wishing to show that the period of religion was preceded by a non-religious period in the history of mankind, suggests that at first man had no idea that his attempts to realise his desires could fail, and that it was his "tardy recognition" of the fact that led him to religion. This tardy recognition, he says, probably "proceeded very slowly, and required long ages for its more or less perfect accomplishment. For the recognition of man's powerlessness to influence the course of {159} nature on a grand scale must have been gradual" (I, 78). I would suggest, however, that it cannot have taken "long ages" for savage man to discover that his wishes and his plans did not always come off. It is, I think, going too far to imagine that for long ages man had no idea that his attempts to realise his desires could fail. If religion arises, as Dr. Frazer suggests, when man recognises his own weakness and his own powerlessness, often, to effect what he most desires, then man in his most primitive and most helpless condition must have been most ready to recognise that there were powers other than himself, and to desire, that is to pray for, their assistance. Doubtless it would be at the greater crises, times of pestilence, drought, famine and war, that his prayers would be most insistent; but it is in the period of savagery that famine is most frequent and drought most to be feared. Against them he takes all the measures known to him, all the practical steps which natural science, as understood by him, can suggest. Now his theory and practice include many things which, though they are in later days regarded as uncanny and magical, are to him the ordinary natural means of producing the effects which he desires. But when he has taken all {160} the steps which practical reason suggests, and experience of the past approves, savage man, harassed by the dread of approaching drought or famine, may still breathe out the Manganja prayer, "Hear thou, O God, and send rain." When, however, he does so, it is, I suggest, doubly erroneous to infer that this prayer takes the place of a spell or that apart from the prayer the acts performed are, and originally were, magical. These acts may be based on the principle that like produces like and may be performed as the ordinary, natural means for producing the effect, which have nothing magical about them.