Notes on Ballad Origins. 157
lady (who has usually helped him out of her father's prison, or enabled him to escape from the wrath of her father, a giant, ogre, elephant, French king, Turk, Moor, or w'hat not) in a far country. He reaches his home, and (by dint of magic or other constraint) forgets his love. She comes to him, revives his memory, and weds him, though he has just taken a new bride, whose wedding-feast she generally interrupts. Every folklorist knows many versions of this far-travelled tale, found in Samoa as in the Jason cycle, where the conclusion is tragical. Often the forgetfulness of the hero is absent.
Mr. Henderson replies (i) : " For anything that Mr. Lang actually shows, there is nothing remarkable in dif- ferent countries having this story about the return of an old true love, since the return of old true loves is quite common."
Agreed ! But that their ladies had released the men from their father's prison ; that the men took other brides under magical constraint, and that the old true loves returned during the bridal, and wedded their first flames, cannot be a very usual sequence of events. Even if it were common, what then ? The ballads and stories would all the more need no source in literary romance, as on Mr. Henderson's theory ; if I understand it.
(2) " If the similarities in the different versions are of such a character as cannot be accounted for by a theory of mere chance, the ballad could not have been evolved by the ' folk fancy,' whatever that may mean, for it is quite incapable of constructing a minutely detailed story."
By the folk fancy, I mean here the fancy of peoples who have no professional literary class, say the Eskimo, Zulus, Algonkin, or Samoyeds. I would add European peasantry, but it is not inconceivable that literary romances might reach them, and be degraded by them, though, in fact, Perrault's printed coxites, for instance, are know^n to be derived from French peasant tales, which still exist, un-
