of the tribal divisions; indeed, the old totem names have lost their former cult significance and have become mere clan names. Thus, the conception here advanced differs from the usual one in that it gives a different significance to the totem names on the two levels of development. In the case of the Australians, we regard them as the names of _cult groups_; in America, where the totem cult proper has receded or has disappeared, we regard them as _mere clan names_. But the extension of totem names to the entire clan organization in the latter case is not, as it were, indicative of a more developed totemism, but rather of _a totemism in the state of decline_. The totem animal, though here also at one time an object of cult, is such no longer, but has become a mere coat of arms. In support of this view of American totem names, we might doubtless also refer to the so-called totem poles. Such a pole consists of a number of human heads representing the ancestors of the clan, and is crowned by the head of the totem animal. This is obviously symbolic of the idea that this succession of generations has as its symbol the totem animal that surmounts it--that is, the totem pole is an enlarged coat of arms. Because of the great regularity of its occurrence, the dual form of tribal division must be regarded as everywhere due to the same cause. Concerning its origin there can scarcely be any doubt. Obviously it has no real connection with totemism itself. This explains why the tribal divisions originally derived their names, not from the totem, but from localities or from other external sources, as the conditions among the Australians would seem to indicate. A phenomenon which recurs in widely distant regions with such regularity as does dual division, is scarcely intelligible except by reference to the general conditions attendant upon the spread of peoples. A tribe leading the unsettled life of gatherers and hunters must of inner necessity separate as its numbers increase or as the food-supply begins to fail. It is but natural that the tribe should first separate into two divisions on the basis of the hunting-grounds which the members occupy; the same process may then repeat itself in the case of each division. The fact that when deviations from the principle of dual division are found, they are most likely to occur in the subordinate groups, is also in harmony with the view that the divisions are due to the natural conditions of dispersion. For, in the case of the subordinate groups, one of the smaller units might, of course, easily disintegrate or wander to a distance and lose its connection with the tribe.