In all theories, the real problem is, how did the early groups get their totem names? The names, once accepted and stereotyped, implied a connection between each kindred, and the animal, plant, or other thing in nature whose name the kindred bore. Round the mystery of this connection the savage mind would play freely, and would invent the explanatory myths of descent from, and kinship with, or other friendly relations with, the name-giving objects. A measure of respect for the objects would be established: they might not be killed or eaten, except under necessity: magic might be worked by human Emus, Kangaroos, Plum-trees, and Grubs for their propagation, as among the Arunta and other tribes; or against them, to bar their ravaging of the crops, as among the Sioux. As a man should not spear a real Emu, if the Emu was his totem, so he does not, for reasons to be adduced, marry or have an amour with a woman who is also of the Emu blood. That is part of the tabu, resulting from the circumstances presently to be explained. All these things, given the savage stage of thought, would inevitably follow from the recognised but mysterious _connection_ between men and the plants and animals from which they were named. All such connections, to the savage, are blood-relationships, and such relationship involves the duties which are recognised and performed. _But how did the early groups come to be named after the plants and animals_; the name suggesting the idea of connection, and the idea of connection involving the duties of the totemist to his totem, and of the totem to the totemist? NO ‘DISEASE OF LANGUAGE’ The names, I repeat, requiring and receiving mythical explanations, and the explanations necessarily suggesting conduct to match, are the causes of Totemism. This theory is not a form of the philological doctrine, _nomina numina_. This is no case of disease of language, in Mr. Max Müller’s sense of the words. A man is called a Cat, all of his kin are Cats. The language is not diseased, but the man has to invent some reason for the name common to his kin. It is not even a case of Folk Etymology, as when a myth is invented to explain the _meaning_ of the name of a place, or person, or thing. Thus the Loch of Duddingston, near Edinburgh, is explained by the myth that Queen Mary, as a child, used to play at ‘dudding’ (or skipping) stones across the water: ‘making ducks and drakes.’ Or again, marmalade is derived from _Marie malade_. Queen Mary, as a child, was seasick in crossing to France, and asked for _confiture_ of oranges; hence _Marie malade_--‘marmalade.’ In both cases, the name to be explained is perverted. There is no real ‘stone’ in Duddingston--‘Duddings’ town,’ the _ton_ or _tun_ of the Duddings; while ‘marmalade’ is a late form of ‘marmalet,’ a word older than Queen Mary’s day.