We are then inclined, for these and other reasons, to explain the ceremonial eating of the totem plant or animal in Australia by the analogy of the ceremonial eating of first-fruits elsewhere, and to regard the ceremony as being in all cases an act of worship, { 188} in which at harvest time the worshippers of a god seek communion with him by means of sacrifice and prayers of thanksgiving. But if we take this view of the sacrifice and prayers offered at harvest time, we shall be inclined to regard the rites which are performed at seed time, or the period analogous to it, as being also possibly, in part, of a religious character. In the case of agricultural peoples it is beyond doubt that some of the ceremonies are religious in character: where the food plant is itself regarded as a deity or the mode in which a deity is manifested, not only may there be at harvest time a sacramental meal in which, as amongst the Aztecs, the deity is formally "communicated" to his worshippers, but at seed time sacrifice and prayer may be made to the deity. Such a religious ceremony, whatever be the degree of civilisation or semicivilisation which has been reached by those who observe the ceremony, does not of course take the place of the agricultural operations which are necessary if the fruits are to be produced in due season. And the combination of the religious rites and the agricultural operations does not convert the agricultural operations into magical operations, or prove that the religious rites are merely pieces of magic { 189} intended to constrain the superior power of the deity concerned. Indeed, if among the operations performed at seed time we find some that from the point of view of modern science are perfectly ineffectual, as vain as eating tiger to make you bold, we shall be justified in regarding them as pieces of primitive science, eventually discarded indeed in the progress of advancing knowledge, but originally practised (on the principle that like produces like) as the natural means of producing the effect desired. If we so regard them, we shall escape the error of considering them to be magical; and we shall have no difficulty in distinguishing them from the religious rites which may be combined with them. Further, where harvest time is marked by the offering of sacrifice and prayers of thanksgiving, we may not unreasonably take it that the religious rites observed at seed time or the period analogous to it are in the nature of sacrifice and prayers addressed to the appropriate deity to beseech him to favour the growth of the plant or animal in question. In a word, the practice of giving thanks to a god at harvest time for the harvest creates a reasonable presumption that prayer is offered to him at seed time; and if thanks are given at a period analogous to { 190} harvest time by a people like the Australian black fellows, who have no domesticated plants or animals, prayers of the nature of petitions may be offered by them at the period analogous to seed time.