made his Irish translation about 1071, which consequently represents the most ancient form of the ‘Historia’ now extant. The manuscripts fall into three principal groups: 1. The Cambridge, of which the chief, though not the most authentic, is Univ. Lib. Camb. Ff. i. 27; the manuscripts of this group, eight in number, represent the North-Welsh version, but have all been influenced by South-Welsh copies. 2. The Harleian group, comprising seventeen manuscripts, and representing the South-Welsh version; the chief manuscript is Harleian 3859, which dates from the tenth or early eleventh century, and is perhaps the oldest extant complete copy of the ‘Historia.’ 3. The Vatican group, comprising five manuscripts and representing the English version of 946; the chief manuscript being Vatican 1964. A manuscript at Chartres (No. 98), which may date from the ninth or tenth century, contains §§ 4–37, and represents the South-Welsh version. (For an account of the manuscripts reference may be made to Hardy, Descript. Cat. Brit. Hist. i. 318–36; De la Borderie, pp. 112–21; Stevenson, pp. xxi–xxix; cf. also Zimmer, pp. 36–42, 201, 277–82).
As an original authority the ‘Historia Britonum’ has little or no direct value. Skene, however, speaks of it as ‘a valuable summary of early tradition, together with fragments of real history which are not to be found elsewhere’ (Four Ancient Books of Wales, i. 40). The true interest of the ‘Historia’ is to be sought in its value for Kymric and Irish literary history from the sixth to the ninth centuries, for Kymric philology, British mythology, and the history of the Arthurian legend. The ‘Genealogiæ,’ however, possess a distinct historical value of their own, and are an important contribution to our knowledge of early British and English history.
The authenticity and value of the ‘Historia Britonum’ have been a fertile subject for criticism in the present century. Gunn, in his edition of 1819, first suggested the claims of Mark to the authorship, but himself regarded the true author as unknown (Preface, p. xv). Stevenson in 1838 regarded the ‘Historia’ as the work of an unknown writer, holding that the ascription to Nennius dated from the twelfth century, and that ‘the successive recensions which have manifestly been made rendered it impossible to satisfactorily ascertain its original form or extent’ (Preface, p. xv). Thomas Wright, in 1842, under the belief that there was no allusion to the ‘Historia Britonum’ older than the twelfth century, and that it claimed to be a work of the seventh century, says that ‘it contains dates and allusions which belong to a much later period, and carries with it many marks of having been an intentional forgery’ (Biog. Britt. Litt. p. 138). The publication of Todd's Irish version of the ‘Historia’ in April 1848 marks an epoch. Herbert, in his preface to this work, while recognising the genuine character of the ascription to Nennius, had no means to test the significance of such data as the genealogy of Fernmail, and concludes that ‘Marcus compiled this credulous book of British traditions for the edification of the Irish circa A.D. 822, and one Nennius, a Briton of the Latin communion, republished it with additions and changes circa A.D. 858’ (Preface, pp. 15, 18). Sir T. Hardy, writing later in 1848, regards the work as anonymous, and Nennius as the possible name of a scribe who in 858 interpolated and glossed the original work for his friend Samuel. He accepts the supposed evidence of the Vatican MS. in favour of a version which was at least as old as 674, and considers that there were later editions dating from 823, 858, 907, and 977 (Monumenta Historica Britannica, pp. 62–4, 107–14; cf. Descrip. Cat. of Brit. Hist. i. 318). Schoell in 1850 regards the authorship as quite unknown, and rejects all but §§ 7–49 and 56, and is doubtful as to the latter; he dates the various editions of the work in 831, 858, 907, 946, and possibly two others in 976 and 994. Skene in ‘The Four Ancient Books of Wales’ (1868) thinks the ‘Historia’ was written in Welsh in the seventh or early eighth century, and that it was afterwards translated into Latin. He observes the predominance of northern influence in parts of the work, ascribes an edition to Mark in 823, when the legends of SS. German and Patrick were added, and another to Nennius in 858, when they were finally incorporated. De la Borderie in 1883 for the most part follows Schoell, holding that the ascription to Nennius was a fiction, but that the original work dates from 822, and that there were six later versions in 831, 832, 857 or 859, 912, 946, and 1024 (L’Historia Britonum, pp. 19–24). Heeger in 1886 puts the date of composition in the early half of the eleventh century. The general attitude of scepticism was broken in 1893 by the ‘Nennius Vindicatus’ of Zimmer, whose arguments appear conclusive and have been adopted in this article.
The ‘Historia Britonum’ was first printed by Gale in 1691 in his ‘Scriptores Quindecim,’ iii. 93–139; the basis of this edition is the Camb. Univ. Lib. MS. Ff. 1, 27. It was included by Charles Bertram [q. v.] in his ‘Britannicarum Gentium Historiæ Antiquæ Scriptores,’ Copenhagen, 1757, which repro-
