If mutilations originate as alleged, we may expect to find some connection between the extent to which they are carried and the social type as simple or compound, militant or industrial. On grouping the facts as presented by fifty-two peoples, the connection emerges with as much clearness as can be expected. In the first place, since the development of mutilation as a custom goes with conquest, and resulting aggregation, it is inferable that simple societies, however savage, will be less characterized by it than the larger savage societies compounded out of them, and less than even the semi-civilized societies. This proves to be true. Of peoples who form simple societies that practise mutilation either not at all or in slight forms, I find, among races wholly unallied, eleven—Fuegians, Veddahs, Andamanese, Dyaks, Todas, Gonds, Santals, Bodo and Dhimals, Mishmis, Kamtchadales, Snake Indians; and these are characterized throughout either by absence of chieftainship, or by chieftainship of an unsettled kind. Meanwhile, of peoples who mutilate little or not at all, I find but two in the class of compound societies; of which one, the Kirghiz, is characterized by a wandering life that makes subordination difficult, and the other, the Iroquois, had a republican form of government. Of societies practising mutilations that are moderate, the simple are relatively fewer, and the compound relatively more numerous: of the one class there are ten—Tasmanians, Tannese, New Guinea people, Karens, Nagas, Ostiaks, Esquimaux, Chinooks, Comanches, Chippewyans; while of the other class there are five—New-Zealanders, East Africans, Khonds, Kukis, Calmucks. And of those it is to be remarked that in the one class the simple headship, and in the other class the compound headship, is unstable. On coming to the societies distinguished by severer mutilations, we find these relations reversed. Among the simple I can name but three—the New Caledonians (among whom, however, the severer mutilation is not general), the Bushmen (who are believed to have lapsed from a higher social state), and the Australians (who have, I believe, similarly lapsed); while, among the compound, twenty-one may be named—Feejeeans, Sandwich-Islanders, Tahitians, Tongans, Samoans, Javans, Sumatrans, Malagasy, Hottentots, Damaras, Bechuanas, Caffres, Congo people, coast negroes, inland negroes, Dahomans, Ashantees, Fulahs, Abyssinians, Arabs, Dakotas. Social consolidation being habitually effected by conquest, and compound and doubly-compound societies being, therefore, during early states, militant in their activities and types of structure, it follows that the connection of the custom of mutilation with the size of the society is indirect, while that with its type is direct. And this the facts show us. If we put side by side those societies which are most unlike in respect of the practice of mutilation, we find them to be those which are most unlike as being wholly unmilitant in organization, and wholly militant in organization. At the one extreme we have the Veddas, Todas, Bodo and Dhimals; while, at
