The comparative study of Mesopotamian and Hellenic sacrifice confronts us finally with another problem belonging to the history of religions, and one of the greatest, the dogma of the death of the divinity and the origin and significance of that belief. For where the mystic type of communion-sacrifice is found, where the animal that is slain for the sacrament is regarded as the incarnation of the deity, the divinity may be supposed himself to die temporarily, doubtless to rise again to life, either immediately or at some subsequent festival. This momentous conclusion need not always have been drawn, for religious logic is not often persistently thorough, nor does the evolution of the idea belong necessarily to the sphere of totemism, as Robertson Smith supposed, and M. Reinach is still inclined to maintain. It is not my concern here to discuss the totemistic hypothesis, but I may point out that in the rare examples where the totem-animal is slain, it is not clear that it is slain as a divinity. Again, the belief in the periodic death of a deity might arise independently of the sacrifice, namely, from the essential idea of the godhead itself, when the divine life is identified with the annual life of vegetation: the phenomena of nature in autumn and spring may suggest to the worshipper the annual death and resurrection of the god or goddess. It is important to note that in this, as in the other source of the belief, the conclusion need not always have been drawn, for the vegetation-deity {250} might be supposed not to die in autumn or winter, but merely to disappear, and the story of his or her disappearance need not carry the same religious consequences as the story of the divine death. The immolation of the divine victim in a communion-service, wherein the worshippers partake symbolically or realistically of the divine flesh and blood, though suggested by a thought that we must call savage, may be pregnant with consequences momentous for higher religion, as the history of Christian dogma attests. And even the annual death of the nature-god may be raised to a higher significance than its mere nature-meaning, and may be linked with the promise of human immortality. We may note, finally, that a religion which expresses in its ritual the idea of the deity’s death and resurrection is likely to be charged with a stronger emotional force than one that lacks it; for the two events will excite an ecstasy of sorrow and of joy in the believer. As the phenomenon, then, is of such importance, it is necessary to be critical and unbiased in the collection of statistics. Our present field of inquiry is the Eastern and Western Mediterranean area; and here our conspicuous example is the ancient Sumerian-Babylonian ritual of Tammuz,[250.1] a folk-service of lamentation and rapture, psychologically akin in many respects to Christianity, and of most powerful appeal.