The Arunta, on the other hand, suppose that the souls of animals which evolved into human beings, are reincarnated in each child born to the tribe. ‘Two clans of Western Australia, who are named after a small opossum and a little fish, think that they are so called because they used to live chiefly on these creatures.’[18] This myth has some support in modern opinion: the kins, it is argued, received their totem names from the animals and plants which mainly formed their food supply; though now their totems are seldom eaten by them. These legends, and others, are clearly ætiological myths, like the Samoan hypothesis that gods are incarnate in the totems. The myths merely try to explain the original connection between men and totems, and are constructed on the lines of savage ideas about the relations of all things in the universe, all alike being personal, and rational, and capable of interbreeding, and of shape-shifting. Certain Kalamantans of Sarawak will not eat a species of deer, because ‘an ancestor became a deer of this kind.’[19] All such fables, of course, are valueless as history; and, in the savage state of the intellect, such myths were inevitable. MODERN THEORIES Mr. McLennan himself at first had a theory, which, as far as I heard him speak of it, was more or less akin to my own. But he abandoned it, says his brother, Mr. Daniel McLennan, for reasons that to him appeared conclusive. I ought to mention that Mr. A. H. Keane informed me, several years ago, that he had independently evolved a theory akin to mine, of which, as it then stood, I had published some hint. (For a statement of Mr. Keane’s theory see our Preface.) In 1884[20] I wrote, ‘People united by contiguity, and by the blind sentiment of kinship not yet brought into explicit consciousness, might mark themselves by a badge, and might thence derive a name, and, later, might invent a myth of their descent from the object which the badge represented.’ But why should such people mark themselves by a badge, and why, if they did, should the mark be, not a decorative or symbolic pattern, but the representation of a plant or animal? These questions I cannot answer, and my present guess is not identical with that of 1884.