This was first noticed in 1863, in the Angers Library, where it was found degraded into the binding of a number of devotional works and a treatise on metric, dated 1459, and once the property of a priest at Alencon. In 1877 M. Gaston Paris called the attention of the learned to it, and the result was that the Danish Government received it next year in exchange for a valuable French manuscript which was in the Royal Library at Copenhagen. This little national treasure, the only piece of contemporary writing of the History, has been carefully photographed and edited by that enthusiastic and urbane scholar, Christian Bruun. In the opinion both of Dr. Vigfusson and M. Paris, the writing dates from about 1200; and this date, though difficult to determine, owing to the paucity of Danish MSS. of the 12th and early lath centuries, is confirmed by the character of the contents. For there is little doubt that the Fragment shows us Saxo in the labour of composition. The MSS. looks as if expressly written for interlineation. Besides a marginal gloss by a later, fourteenth century hand, there are two distinct sets of variants, in different writings, interlined and running over into the margin. These variants are much more numerous in the prose than in the verse. The first set are in the same hand as the text, the second in another hand: but both of them have the character, not of variants from some other MSS., but of alternative expressions put down tentatively. If either hand is Saxo's it is probably the second. He may conceivably have dictated both at different times to different scribes. No other man would tinker the style in this fashion. A complete translation of all these changes has been deemed unnecessary in these volumes; there is a full collation in Holder's "Apparatus Criticus". The verdict of the Angers-Fragment, which, for the very reason mentioned, must not be taken as the final form of the text, nor therefore, despite its antiquity, as conclusive against the First Edition where the two differ, is to confirm, so far as it goes, the editing of Ascensius and Pederson. There are no vital differences, and the care of the first editors, as well as the authority of their source, is thus far amply vindicated.