stant of heroic tradition. As Brunhild she plays a great part in the most famous hero-tale of the Germans. But certain characteristics clearly differentiate the Irish from the German representatives of the part. There is an unhuman independence of, or indifference to the mortal wooer, a divine abandon when she decides to yield, a callousness to the fate of the particular mortal on whom she bestows her favours, that stamp her of the kin of the immortals, that place her on a different level from such beings, transcendently endowed with valour and high-heartedness, yet women all the same, as Sigrun or Brunhild. These characteristics are clearly marked in Saxo's heroine, whose conduct after Amleth's death moves the worthy chronicler to one of his familiar outbursts of rhetorical commonplace about the fickleness of woman. Note, too, that Amleth's first wife is as ready to subordinate herself to her rival as are Guildeluec or the second wife in Gold-tree.
It may be urged that the name Hermutrude is non-Celtic, but I do not think this point is of the slightest importance. Saxo would almost certainly give his personages a recognisable name, even if, as is not likely, his Danish informants had retained and correctly rendered an alien Celtic one.
So far, then, the consideration of allied stories has strengthened my general proposition by showing, both: that another possible offshoot from the original of Eliduc exists, and that the derivatives of Eliduc show no tendency to revert to the folk-tale type. A close examination of the lai and the recently collected folk-tales may further support the contention that the Scotch-Gaelic tale probably represents the original of Marie's poem, and almost certainly is not derived from the continental versions.
With regard to the date of the lai, a terminus ad quem is furnished by that of Ille et Galeron, finished, as Prof Förster shows, in 1167. By this time, then, Marie's poem, or one closely resembling it, must have enjoyed wide favour. But
