the Carolingian. The old kings and knights gave way to new kings and knights, but only the names were changed, the rest remained almost unaltered. The same process of transfer from older and more or less forgotten heroes takes place continually; new names are substituted for the old, and local considerations play a decisive rôle in the transfer. The trouvère who sings the exploits of the ancestor of this baron will use the same language and ascribe the same exploits to the ancestor of another baron when he sings in his hall and at his banqueting feast. The same tendency prevails everywhere and at every time. We meet with it at almost every turn in the epical poetry of the East, and of the West. It is one of the constant factors in the development and evolution of the ballads. Sufficient attention has not been paid to this point. Here and there this transfer and change has been admitted, but not recognised as an universal law, only as an exceptional incident. I on the contrary find in this practice of constant substitution, the very key to the problem of the sources from which the ancient writers have drawn their inspiration. Their skill consisted in giving a thorough local character to a tale borrowed from elsewhere and in so changing the colouring as to impress their contemporaries and to win their applause.
This, then, is my starting-point in the investigation of the sources of the Merlin legend. Our earliest authority for it is the Chronicle of Geoffrey of Monmouth.[1] I therefore ask myself:—Given a monk at a local shrine, endowed with some poetical imagination, to what kind of literature could he have had access in England at the end of the eleventh or twelfth century? What mass of tradition could be floating about him, to be caught up and fixed in his writings? and for what kind of audience did he
- ↑ See my article on Jewish Sources and Parallels to Early English Metrical Romances, 1807.
