does not, therefore, establish a supernatural dynasty, as does the chief of divine descent.
And now, having considered the several factors which coöperate to establish political headship, let us consider the process of cooperation through its ascending stages. The truth to be noted is, that the successive phenomena which occur in the simplest groups habitually recur in the same order in compound groups, and again in doubly compound groups.
As, in the simple group, there is at first a state in which there is no headship, so, when simple groups which have political heads possessing slight authorities are associated, there is at first no headship of the cluster. The Chinooks furnish an example. Describing them, Lewis and Clarke say: "As these families gradually expand into bands, or tribes, or nations, the paternal authority is represented by the chief of each association. This chieftain, however, is not hereditary." And then comes the further fact, which here specially concerns us, that "the chiefs of the separate villages are independent of each other": there is no general chieftainship.
As headship in the simple group, at first temporary, ceases when the war which initiates it ends, so, in the cluster of groups which severally have recognized heads, a common headship at first results from a war, and lasts no longer than the war. Falkner says, "In a general war, when many nations enter into an alliance against a common enemy," the Patagonians "chose an apo, or commander-in-chief, from among the oldest or most celebrated of the caciques." The Indians of the upper Orinoco live "in hordes of forty or fifty under a family government, and they recognize a common chief only in times of war." So is it in Borneo. "During war the chiefs of the Sarebas Dyaks give an uncertain allegiance to a head chief, or commander-in chief." It has been the same in Europe. Seeley remarks that the Sabines "seem to have had a central government only in war-time." Again: "Germany had anciently as many republics as it had tribes. Except in time of war, there was no chief common to all, or even to any given confederation."
This recalls the fact indicated when treating of political integration, that the cohesion within compound groups is less than that within simple groups, and again that the cohesion within the doubly compound less than that within the compound. What was there said of cohesion may here be said of subordination; for we find that, when by continuous war a permanent headship of a compound group has been generated, it is less stable than the headships of the simple groups. Often it lasts only for the life of the man who achieves it; as among the Karens and the Maganga, and as among the Dyaks, of whom Boyle says: "It is an exceptional case if a Dyak chief is raised to an acknowledged supremacy over the other chiefs. If he is so raised he
