For some time I have been preparing a general work on primitive superstition and religion. Among the problems which had attracted my attention was the hitherto unex plained rule of the Arician priesthood; and last spring it happened that in the course of my reading I came across some facts which, combined with others I had noted before, suggested an explanation of the rule in question. As the explanation, if correct, promised to throw light on some obscure features of primitive religion. I resolved to develop it fully, and, detaching it from my general work, to issue it as a separate study. This book is the result. Now that the theory, which necessarily presented itself to me at first in outline, has been worked out in detail, I cannot but feel that in some places I may have pushed it too far. If this should prove to have been the case, I will readily acknowledge and retract my error as soon as it is brought home to me. Meantime my essay may serve its purpose as a first attempt to solve a difficult problem, and to bring a variety of scattered facts into some sort of order and system.