Among the quaint conceits with which primitive cultures abound is that of attaching to particular numbers a peculiar character of sanctity. “Everything in the universe,” a Crow Indian once told me, “goes by fours.” As a matter of fact, most things in Crow religious life are adjusted to this conception. An important ceremonial act is thrice feigned so as to be actually performed at the fourth attempt; religious processions halt four times; songs are sung in sets of four; in mythic tales it is the fourth trial that carries an heroic feat to a successful issue. Now this cultural fact very largely eludes psychological interpretation. The first thing that strikes us is that this feature is no peculiarity of the Crow, but is rather widely distributed among their immediate neighbors and even remote Indian tribes, though jointly occupying a continuous area. Since outside of this region other numbers figure as mystic, we cannot regard the view of the sacredness of Four as a general trait of human psychology but must assume that the concept was borrowed by most of the tribes now holding it. A wider survey teaches us that corresponding, though not identical, conceptions are very common. _Seven_ figures in parts of Asia, _Three_ in European folklore, _Five_ in Oregon and northern Nevada, _Six_ among the Ainu of Yezo, _Nine_ among the Yakut, _Ten_ among the Pythagorean philosophers of ancient Greece, very much as _Four_ does among the Crow. Now the fact that a particular Crow Indian regards Four as a sacred number does not mean that this is an individual peculiarity of his any more than the Christian’s reaction to a cross is a proof of some psychological idiosyncrasy. Individually the Crow Indian may be quite indifferent to the number and yet he would view it as sacred because he has been taught so to regard it. This is, of course, the vital difference between ethnology and psychology which has already been emphasized. Nevertheless, the association must at one time have been formed in an individual mind, whether among the Crow or elsewhere, and the question arises as to what such an association means. Francis Galton showed some time ago that such associations of definite personal characteristics with numbers occur by no means infrequently among Europeans. The phenomenon we are dealing with is thus linked with a group of related phenomena and in so far forth is explained.