There is one manuscript in which Beowulf has survived to the present day. The earliest known owner of the manuscript is an early Anglo-Saxon scholar known as Laurence Nowell, Dean of Lichfield. Some time later, it entered into the manuscript collection of Sir Robert Bruce Cotton (1571-1631) and was shelved under the bust of Roman Emperor Aulus Vitellius shelf A, position 15; hence the name Cotton Vitellius A. xv. It is a composite manuscript consisting of two codices (the Southwick Codex and the Nowell Codex) and nine different works between them. Beowulf follows three prose works in the Nowell Codex and precedes the poem Judith.

Cotton Vitellius A. xv.
The Southwick Codex 1 - The Soliloquia of St. Augustine 2 - The Gospel of Nicodemus 3 - The Debate of Solomon and Saturn 4 - St. Quintin Homily (ending lost)
The Nowell Codex 1 - The Life of St. Christopher (beginning lost) 2 - The Wonders of the East 3 - Alexander's Letter to Aristotle 4 - Beowulf 5 - Judith

In 1700, Cotton's collection was donated to the British people. By 1722, Cotton's house had deteriorated and the collection was moved to Essex House in Strand. Seven years later, it was moved again to Ashburnham House in Westminster. In 1731, Ashburnham House caught fire. Cotton Vitellius A. xv. was badly burned around the edges when it was saved by being thrown from the window with many other manuscripts.

G.J. Thorkelin, an Anglo-Saxonist from Iceland, and a hired scribe made two transcripts of Beowulf in 1787. It was not until the next century that the British Museum went about systematically repairing the books damaged by the fire. By that time, much of the text of Beowulf had crumbled away from the edges of the pages. By 1845, Cotton Vitellius A. xv. was rebound mounted on paper frames that help slow the deterioration of the edges of the pages. In 1882, Julius Zupitza produced a black-and-white facsimile and transcription of Beowulf, followed by Kemp Malone's in 1969. In 1990, work on the Electronic Beowulf, a collection of high quality digitizations with fibre-optic and ultra-violet lighting headed by Kevin Kiernan, continues at the British Library and the University of Kentucky.

When the 1969 edition of Zupitza's facsimiles went to press, the manuscript measured 195 mm high by 115-130 mm wide with a written area 175 mm high by 105 mm wide.

On the whole, the manuscript remains fairly readable, but some folios in particular had seen much neglect. Folio 179/182 is argued to be palimpsest, that is, it had been erased in preparation for reuse or revision. Here is a fragment from a preliminary scan from the Electronic Beowulf project:


Fig: Beowulf Manuscript Fragment

This fragment actually corresponds to what appears to be the description of the dragon. It is possible that we may never know what an Anglo-Saxon dragon would have looked like.

Some disagreement arises over the foliation, or page numbering, of Cotton Vitellius A. xv. Evidence shows that this has been attempted at least six times. At least two attempts were made before the fire of 1731, which have since burned away. The subsequent four attempts can be found on the outside corners of the manuscript, and on the paper frames. Two numbering schemes persist as more canonical. The manuscript foliation, performed sometime between 1793 and 1801, was done with two quires of the Nowell Codex out of order, and two folios of Beowulf out of place. The official foliation done in 1884 corrects these errors, but becomes somewhat problematic when converting between the two systems. The following table illustrates them:

1793-1801 Old Foliation	1884 Official Foliation
129, 130	132, 122
132 to 146	134 to 148
131	149
147 to 188	150 to 191
197	192
189 to 196	193 to 200
198	201

Cotton Vitelius A. xv. was assembled in gatherings or quires of three to six vellum leaves tied together with thread. Following the fire damage and subsequent remounting of the manuscript, it is not longer possible to be certain which folios went together in quires. While it is speculated that Cotton Vitellius A. xv. is a collection of codices, with Beowulf being one of them, one might conclude that Beowulf might have been a complete work in itself at one point. However, without knowing the quires that make up the collection, Beowulf may only be an integral part.