of a compound text, which properly represents none of the versions.
Kristensen's work, therefore, has the great merit of bringing us as close to the actual narrators as can be done by the medium of a printed page. It is not an attempt to present a digested statement or critical account of their tales and beliefs, nor to discover what lies behind them all; but for directness of method and honest adherence to the facts it leaves nothing to be desired, and, rightly understood, is of the highest value.
I have already mentioned that Kristensen's serious collecting began with ballads [Folkeviser). Of these he has now published four collections, forming volumes i., ii., x., and xi., of his Popular Traditions of Jutland (Jyske Folkeminder), and containing in all 475 versions of popular ballads, many of them previously unprinted in any form. For the value of his work in this department one could not wish a higher authority than that of Grundtvig, from whose appendix to the first collection a few particulars may be profitably extracted. In 1844 Grundtvig made an appeal to those interested in the matter, to note down and send to him any versions of ballads which they could discover still living in the mouths of the people. Perhaps he expected that little of the kind was to be found at that time, for even in 1814 an editor of Danish ballads had solemnly assured his readers that they were no longer a living thing. The result of Grundtvig's appeal, however, was that during the next 27 years he received oral versions of 130 old ballads. It took 170 different persons to collect these. Within three years (1868-70), and almost entirely within the limits of a single parish, Kristensen had taken down versions of 150 different ballads, 75 of which had not been found in living tradition elsewhere, and 14 of which were previously unknown in Denmark in any form. The significance of these finds is strongly emphasised by Grundtvig; "it is," he says, "a real old Danish Hercula-
