lore until we know something of its place in folk-lore analysis. Working this out in some examples chosen almost at random from my collections, let me first note some customs of considerable significance which allow us at once to penetrate beneath the stratum of Christianity into the paganism beneath.
Remember, I am trying to show the importance of analysing the component elements of folk-lore. Baptism, an essentially Christian ceremony, might off-hand be supposed to contain nothing but evidence for Christianity. It might at most be expected that the details of the ceremony would contain relics of adapted pagan rites, and this we know is the case. But my point is, that we can go beyond even this, and discover in the popular conception of the rite very clear indications of the early antagonism between Christianity and paganism—an antagonism which is certainly some eighteen hundred years old—in this country, and though so old is still contained in the evidence of folk-lore.
An analysis of baptismal folk-lore shows us that its most important section is contained under the group which deals with the effect of non-baptism.
In England we have it prevailing in the border counties, in Cornwall, Devonshire, Durham, Lancashire, Middlesex, Northumberland, and Yorkshire, and in north-east Scotland, that children joined the ranks of the fairies if they died unchristened, or that their souls wandered about in the air, restless and unhappy, until Judgment Day. Various penalties attended the condition of non-baptism, but perhaps the most significant is the Northumberland custom of burying an unbaptised babe at the feet of an adult Christian corpse—surely a relic of the old sacrifice at a burial which is indicated so frequently in the graves of prehistoric times, particularly of the long-barrow period. In Ireland we have the effect of non-baptism in a still more grim form. In the sixteenth century the rude Irish used to leave the right arms of their male children un-
