The dual nature of taboo as the avoidance of both the sacred and the unclean is noted by authorities on the subject who differ in other respects as to the definition of taboo, such as in the relation of taboo to the magical ceremonies by which man undertook to mould his environment to his wishes. Whether the tabooed object be regarded in one light or the other, the breaking of taboo is associated with dread of the unknown--besides the fear of infection with the qualities of the tabooed object according to the laws of sympathetic magic. There is also the fear of the mysterious and supernatural, whether conceived as the mana force or as a principle of "bad magic." Dr. J.G. Frazer has collected into the many volumes of "The Golden Bough" a mass of evidence concerning the taboos of primitive society. On the basis of his definition of magic as "a misapplication of the ideas of association by similarity and contiguity," Dr. Frazer divided magic into "positive magic," or charms, and "negative magic," or taboo. "Positive magic says, 'Do this in order that so and so may happen.' Negative magic or taboo says 'Do not do this lest so and so should happen.'"[4, p.111, v.I.] But Dr. Frazer's conclusion, which he himself considered only tentative, was not long left unassailed. Prof. R.R. Marett in his essay "Is Taboo a Negative Magic?"[5] called attention to the very evident fact that Dr. Frazer's definition would not cover the characteristics of some of the best known taboos, the food taboos of Prof. Tylor to which we have previously alluded in this study, as a consequence of which "the flesh of timid animals is avoided by warriors, but they love the meat of tigers, boars, and stags, for their courage and speed."[3, p.131.] Are not these food taboos rather, Dr. Marett asks, a "misapplication of the ideas of association by similarity and contiguity" amounting to the sympathetic taboos so carefully described by such writers on Magic as MM. Hubert and Mauss of L'Année Sociologique? Still another kind of taboos mentioned by Dr. Frazer but amplified by Mr. Crawley in "The Mystic Rose," the taboos on knots at childbirth, marriage, and death, are much better described by the term "sympathetic taboo." Moreover, if taboo were a form of magic as defined by Dr. Frazer, it would be a somewhat definite and measurable quantity; whereas the distinguishing characteristic of taboo everywhere is the "infinite plus of awfulness" always accompanying its violation. As Dr. Marett observes, there may be certain definite results, such as prescribed punishment for violations against which a legal code is in process of growth. There may be also social "growlings," showing the opposition of public opinion to which the savage is at least as keenly sensitive as the modern. But it is the "infinite plus" always attached to the violation of taboo that puts it into the realm of the mystical, the magical. It would seem that Dr. Frazer's definition does not include enough.