for her death, he found her one night in a deep and solitary dale amid a number of women. With great joy he seized her, and, carrying her off, lived with her again for many years and had a numerous progeny. Not a few of her descendants were living when Map wrote, and were known as the children of the dead woman. This, of course, is not a Swan-maiden story at all. At the end of Chapter V. I have referred to some similar tales; and what we learned during our discussion of the subject of Changelings may lead us to suspect that we have here in an imperfect form a story of the exchange of an adult woman for a lifeless image, and her recovery from the hands of her ravishers. This is by no means the same plot as that of the stories recounted by Liebrecht in which the wife or the betrothed is rescued from the grave. Those stories, at least in warm climates where burials are hurried, and in rude ages when medical skill is comparatively undeveloped, are all within the bounds of possibility. There does not appear in them any trace of mythology,—hardly even of the supernatural; and he would be a bold man who would deny that a substratum of fact may not underlie some of them. To establish their relationship with the group we are now considering, links of a much more evident character are wanting. The fact that they are traditional is not of itself sufficient. The fairy of the Forest of Dean had not revived after death, or supposed death; nor had she been recovered from supernatural beings who had stolen her away. Map's account, to whatever his expression from the dead may point, is inconsistent with either the one or the other. Rather she was stolen from her own kindred, to become the wife of him who had won her by his own right arm.
But a single instance, and that instance either inconsistent with the analogous traditions, or unable to supply a cogent or consistent explanation of them, is not a a very safe basis for a theory. What is it worth when it
