Beowulf fills ff. 129 (132)a to 198 (201)b of the British Museum MS Cotton Vitellius A. XV.
Beowulf is written in two hands, the first of which goes to l. 1939. This hand was identified by Prof. Sedgefield (Beowulf, Introduction, p. xiv, footnote) with that of the piece immediately preceding Beowulf in the MS, and by Mr Kenneth Sisam, in 1916, with that of all three immediately preceding pieces: the Christopher fragment, the Wonders of the East, and the Letter of Alexander on the Wonders of India. The pieces preceding these, however (the Soliloquies of S. Augustine, the Gospel of Nicodemus, Salomon and Saturn), are certainly not in the same hand, and their connection with the Beowulf-MS is simply due to the bookbinder.
From l. 1939 to the end, Beowulf is written in a second hand, thicker and less elegant than the first. This second hand seems to be clearly identical with that in which the poem of Judith, immediately following Beowulf, is written. This was pointed out by Sievers in 1872 (Z.f.d.A. XV, 457), and has never, I think, been disputed (cf. Sisam, p. 337; Förster, p. 31). Nevertheless the two poems have probably not always formed one book. For the last page of Beowulf was apparently once the last page of the volume, to judge from its battered condition, whilst Judith is imperfect at the beginning. And there are trifling differences, e.g. in the frequency of the use of contractions, and the form of the capital H.
This identity of the scribe of the second portion of Beowulf and the Judith scribe, together with the identity (pointed out by Mr Sisam) of the scribe of the first portion of Beowulf and the scribe of the three preceding works, is important. A detailed comparison of these texts will throw light upon the characteristics of the scribes.
That the three preceding works are in the same hand as that of the first Beowulf scribe was again announced, independently of Mr Sisam, by Prof. Max Förster, in 1919. Sievers had already in 1871 arrived at the same result (see Förster, p. 35, note) but had not published it.
It seems to me in the highest degree improbable that the Beowulf-MS has lost its ending, as Prof. Förster thinks (pp. 82, 88). Surely nothing could be better than the conclusion of the poem as it stands in the MS: that the
casual loss of a number of leaves could have resulted in so satisfactory a conclusion is, I think, not conceivable. Moreover, the scribe has crammed as much material as possible into the last leaf of Beowulf, making his lines abnormally long, and using contractions in a way he does not use them elsewhere. The only reason for this must be to avoid running over into a new leaf or quire: there could be no motive for this crowded page if the poem had ever run on beyond it.
There is pretty general agreement that the date of the Beowulf-MS is about the year 1000, and that it is somewhat more likely to be before that date than after.
The Beowulf-MS was injured in the great Cottonian fire of 1731, and the edges of the parchment have since chipped away owing to the damage then sustained. Valuable assistance can therefore be derived from the two transcripts now preserved in the Royal Library of Copenhagen, made in 1787, when the MS was much less damaged.
A. Poema anglosaxonicum de rebus gestis Danorum ... fecit exscribi Londini A.D. MDCCLXXXVII Grimus Johannis Thorkelin.
B. Poema anglosaxonicum de Danorum rebus gestis ... exscripsit Grimus Johannis Thorkelin. Londini MDCCLXXXVII.