S. Reinach, who successfully traced the relics of the totemic system in the cult and customs of later periods, though attaching from the very beginning only slight value to the factor of descent from the totem animal, once made the casual remark that totemism seemed to him to be nothing but “_une hypertrophie de l’instinct social_.”[157] The same interpretation seems to permeate the new work of E. Durkheim, _Les Formes Élémentaires de la Vie Religieuse; Le Systême Totémique en Australie_, 1912. The totem is the visible representative of the social religion of these races. It embodies the community, which is the real object of veneration. Other authors have sought a more intimate reason for the share which social impulses have played in the formation of totemic institutions. Thus A. C. Haddon has assumed that every primitive tribe originally lived on a particular plant or animal species and perhaps also traded with this food and exchanged it with other tribes. It then was inevitable that a tribe should become known to other tribes by the name of the animal which played such weighty rôle with it. At the same time this tribe would develop a special familiarity with this animal, and a kind of interest for it which, however, was based upon the psychic motive of man’s most elementary and pressing need, namely, hunger[158]. The objections against this most rational of all the totem theories are that such a state of the food supply is never found among primitive men and probably never existed. Savages are the more omnivorous the lower they stand in the social scale. Besides, it is incomprehensible how such an exclusive diet could have developed an almost religious relation to the totem, culminating in an absolute abstention from the referred food. The first of the three theories about the origin of totemism which Frazer stated was a psychological one. We shall report it elsewhere. Frazer’s second theory, which we will discuss here, originated under the influence of an important publication by two investigators of the inhabitants of Central Australia[159]. Spencer and Gillen describe a series of peculiar institutions, customs, and opinions of a group of tribes, the so-called Arunta nation, and Frazer subscribes to their opinion that these peculiarities are to be looked upon as characteristics of a primary state and that they can explain the first and real meaning of totemism. In the Arunta tribe itself (a part of the Arunta nation) these peculiarities are as follows: 1. They have the division into totem clans but the totem is not hereditary but is individually determined (as will be shown later). 2. The totem clans are not exogamous, and the marriage restrictions are brought about by a highly developed division into marriage classes which have nothing to do with the totems.