incursions of the irresponsible amateur. I would add that what I have to say is not intended in any way to abrogate Dr. Frazer's contrast between magic and religion. On the contrary, I consider it to embody a working distinction of first-rate importance. I merely wish to mitigate this contrast by proposing what, in effect, amounts to a separation in lieu of a divorce. A working principle, if it is to work, must not be pushed too hard.
The question, then, that I propose to discuss is the following: Does the spell help to generate the prayer, and, if so, how? Now the spell belongs to magic, and the prayer to religion. Hence we are attacking, in specific shape, no less a problem than this: Does magic help to generate religion?
Perhaps it will make for clearness of exposition if I outline the reply I would offer in what follows to this latter question. First, I suppose certain beliefs, of a kind natural to the infancy of thought, to be accepted at face value in a spirit of naive faith, whilst being in fact illusory. The practice corresponding to such naive belief I call "rudimentary magic." Afterwards I conceive a certain sense of their prima facie illusiveness to come to attach to these beliefs, without, however, managing to invalidate them. This I call the stage of "developed magic." Such magic, as embodying a reality that to some extent transcends appearance, becomes to a corresponding extent a mystery. As such, on my view, it tends to fall within the sphere of religion. For I define the object of religion to be whatever is perceived as a mystery and treated accordingly. (Dr. Frazer, however, defines religion differently, and this must be borne in mind in estimating the pertinence of such criticisms as I may pass on his interpretations of the facts.)
Let us now turn to the Golden Bough to see what light it throws on this same problem, viz. whether magic is a factor in the genesis of religion. If I understand Dr. Frazer aright—and of this I am by no means sure—his position
