different nature from those whose tranquillity it menaces. To be sure, in the sentiment which the believer feels for the things he adores, there always enters in some element of reserve and fear; but this is a fear sui generis, derived from respect more than from fright, and where the dominating emotion is that which la majesté inspires in men. The idea of majesty is essentially religious. Then we have explained nothing of religion until we have found whence this idea comes, to what it corresponds and what can have aroused it in the mind. Simple souls of men cannot become invested with this character by the simple fact of being no longer incarnate.
This is clearly shown by an example from Melanesia. The Melanesians believe that men have souls which leave the body at death; it then changes its name and becomes what they call a tindalo, a natmat, etc. Also, they have a cult of the souls of the dead: they pray to them, invoke them and make offerings and sacrifices to them. But every tindalo is not the object of these ritual practices; only those have this honour which come from men to whom public opinion attributed, during life, the very special virtue which the Melanesians call the mana. Later on, we shall have occasion to fix precisely the meaning which this word expresses; for the time being, it will suffice to say that it is the distinctive character of every sacred being. As Codrington says, "it is what works to effect anything which is beyond the ordinary power of men, outside the common processes of nature."117 A priest, a sorcerer or a ritual formula have mana as well as a sacred stone or spirit. Thus the only tindalo to which religious services are rendered are those which were already sacred of themselves, when their proprietor was still alive. In regard to the other souls, which come from ordinary men, from the crowd of the profane, the same author says that they are "nobodies alike before and after death."118 By itself, death has no deifying virtue. Since it brings about in a more or less complete and final fashion the separation of the soul from profane things, it can well reinforce the sacred character of the soul, if this already exists, but it cannot create it.
Moreover, if, as the hypothesis of the animists supposes, the first sacred beings were really the souls of the dead and the first cult that of the ancestors, it should be found that the lower the societies examined are, the more the place given to this cult in the religious life. But it is rather the contrary which is true. The ancestral cult is not greatly developed, or even presented under a characteristic form, except in advanced societies like those of China, Egypt or the Greek and Latin cities; on the other hand, it is completely lacking in the Australian societies which, as we shall see, represent the lowest and simplest form of social organization which we know. It is true that funeral rites and rites of mourning are found there; but these practices do not constitute a cult, though this name has sometimes wrongfully been given them. In reality, a cult is not a simple group of ritual precautions which a man is held to take in certain circumstances; it is a system of diverse rites, festivals and ceremonies which all have this characteristic, that they reappear periodically. They fulfil the need which the believer feels of strengthening and reaffirming, at regular intervals of time, the bond which unites him to the sacred beings upon which he depends. That is why one speaks of marriage rites but not of a marriage cult, of rites of birth but not of a cult of the new-born child; it is because the events on the occasion of which these rites take place imply no periodicity. In the same way, there is no cult of the ancestors except when sacrifices are made on the tombs from time to time, when libations are poured there on certain more or less specific dates, or when festivals are regularly celebrated in honour of the dead. But the Australian has no relations of this sort with his dead. It is true that he must bury their remains according to a ritual, mourn for them during a prescribed length of time and in a prescribed manner, and revenge them if there is occasion to.119 But when he has once accomplished these pious tasks, when the bones are once dry and the period of mourning is once accomplished, then all is said and done, and the survivors have no more duties towards their relatives who exist no longer. It is true that there is a way in which the dead continue to hold a place in the lives of their kindred, even after the mourning is finished. It is sometimes the case that their hair or certain of their bones are kept, because of special virtues which are attached to them.120 But by that time they have ceased to exist as persons, and have fallen to the rank of anonymous and impersonal charms. In this condition they are the object of no cult; they serve only for magical purposes.
However, there are certain Australian tribes which periodically celebrate rites in honour of fabulous ancestors whom tradition places at the beginning of time. These ceremonies generally consist in a sort of dramatic representation in which are rehearsed the deeds which the myths ascribe to these legendary heroes.121 But the personages thus represented are not men who, after living the life of men, have been transformed into a sort of god by the fact of their death. They are considered to have exercised superhuman powers while alive. To them is attributed all that is grand in the history of the tribe, or even of the whole world. It is they who in a large measure made the earth such as it is, and men such as they are. The haloes with which they are still decorated do not come to them merely from the fact that they are ancestors, that is to say, in fine, that they are dead, but rather from the fact that a divine character is and always has been attributed to them; to use the Melanesian expression, it is because they are constitutionally endowed with mana. Consequently, there is nothing in these rites which shows that death has the slightest power of deification. It cannot even be correctly said of certain rites that they form an ancestor-cult, since they are not addressed to ancestors as such. In order to have a real cult of the dead, it is necessary that after death real ancestors, the relations whom men really lose every day, become the object of the cult; let us repeat it once more, there are no traces of any such cult in Australia.
Thus the cult which, according to this hypothesis, ought to be the predominating one in inferior societies, is really nonexistent there. In reality, the Australian is not concerned with his dead, except at the moment of their decease and during the time which immediately follows. Yet these same peoples, as we shall see, have a very complex cult for sacred beings of a wholly different nature, which is made up of numerous ceremonies and frequently occupying weeks or even entire months. It cannot be admitted that the few rites which the Australian performs when he happens to lose one of his relatives were the origin of these permanent cults which return regularly every year and which take up a considerable part of his existence. The contrast between the two is so great that we may even ask whether the first were not rather derived from the second, and if the souls of men, far from having been the model upon which the gods were originally imagined, have not rather been conceived from the very first as emanations from the divinity.
IV
From the moment that the cult of the dead is shown not to be primitive, animism lacks a basis. It would then seem useless to discuss the third thesis of the system, which concerns the transformation of the cult of the dead into the cult of nature. But since the postulate upon which it rests is also found in certain historians of religion who do not admit the animism properly so-called, such as Brinton,122 Lang,123 Réville,124 and even Robertson Smith himself,125 it is necessary to make an examination of it.
This extension of the cult of the dead to all nature is said to come from the fact that we instinctively tend to represent all things in our own image, that is to say, as living and thinking beings. We have seen that Spencer has already contested the reality of this so-called instinct. Since animals clearly distinguish living bodies from dead ones, it seemed to him impossible that man, the heir of the animals, should not have had this same faculty of discernment from the very first. But howsoever certain the facts cited by Spencer may be, they have not the demonstrative value which he attributes to them. His reasoning supposes
