I conjecture that this feeling first showed itself within the family, and led the families of a band to exchange their daughters; and that later it extended to all members of the same band, and that wives were then sought from other bands. Probably wives were obtained sometimes by capture or enticement, sometimes by exchange so far as amicable relations with neighbours prevailed: for the evidence is far from showing that “marriage by capture” was ever a general custom; whilst our knowledge of the Australians and Boschmans shows that neighbouring bands of savages are not always hostile. Such a state of things may have existed for ages before the rise of Totemism, and amongst races that never adopted Totemism. When names of animals were first given to, or adopted by, various groups, Totemism somewhere (perhaps in several places) resulted, as a belief in the magical or spiritual connexion between men and animals of the same name. But the bands that adopted Totemism were already exogamous, having an aversion to marriage with others of the same band; and the practice of Exogamy was thus brought under the mysterious sanction of totemic ideas. It would then be further extended to bar marriage with those of the same Totem in other groups. Totemic sanctions may have been useful in confirming the Exogamy of some groups, but all superstitious aids to right conduct are liable to perverse issues, and at best they are second best. Totemism prevents consanguineous marriages, but also prevents marriage with thousands of people where no blood-relationship is traceable. If any races were able to perceive that the interest of Exogamy was the prevention of marriage between near kin, and then to keep account of kinship and govern themselves accordingly, they chose the better part.