Humphrey Wanley, who transcribed a few lines of the Beowulf manuscript at the end of the seventeenth century and published them in 1705, is the only source for a few lost letters. At the end of the eighteenth century, Grimur Jonsson Thorkelin and his hired scribe, probably James Matthews, a British Museum staff member, together saved about 2000 letters that were subsequently lost by fire damage as a consequence of the Cottonian Library fire in 1731. John J. Conybeare and Frederic Madden, in the early nineteenth century, produced two collations, which together help test the accuracy of the Thorkelin transcripts and of their own collations.
For an extensive introduction (<%= link_to '"Thorkelin's Discovery of Beowulf"', 'http://www.uky.edu/~kiernan/Thorkelin/Th_1/', :target => "_blank", :class => 'links' %>), a comprehensive collation, and a critical analysis (<%= link_to '"The Reliability of the Transcripts"', 'http://www.uky.edu/~kiernan/Thorkelin/Th_3/', :target => "_blank", :class => 'links' %> and <%= link_to '"Conclusion"', 'http://www.uky.edu/~kiernan/Thorkelin/Th_3/#conclusion', :target => "_blank", :class => 'links' %>) of these indispensable copies of Beowulf before fire-damage severely harmed the manuscript, see Kiernan, The Thorkelin Transcripts of Beowulf (Copenhagen 1986). The restored Thorkelin readings are all incorporated in the text and presented in the texutal notes; users may examine details by double-clicking any reading in round or square brackets when 'Options' are set for 'Critical Edition'.
<%= image_tag "tha-thb1.jpg" %>
To set viewing options for both transcripts, go to Views in the top menu, select 'Change View,' and in the window that opens choose a 'Layout' (here one transcript above the other) with 'Browse Mode' set to 'Free.' This setup lets the user compare the two Thorkelin transcripts section by section.
<%= image_tag "tha-thb2.jpg" %>
Thorkelin A used large paper and copied continuously, rather than copying the Beowulf MS page by page and line by line, as Thorkelin did. A 'Top/Bottom' layout permits wide views of Thorkelin A on smaller screens.
<%= image_tag "tha-3.jpg" %>
Thorkelin A's procedure of copying continuously, instead of following the exact format of the manuscript, makes it difficult to collate with either the manuscript or with Thorkelin B. Both Conybeare and Madden use the MS foliation, too, as do most modern editors.
<%= image_tag "tha-2.jpg" %>
In 1983, at the request of the Assistant Keeper of Manuscripts at the Royal Library, I supervised the penciled entry of Beowulf manuscript foliation numbers in Thorkelin A to facilitate collations by scholars. The keeper also penciled in diagonal lines where the new manuscript leaf starts.
After he had learned to distinguish thorns and eths from p's, Thorkelin A began proofreading his work and made some interlinear corrections with a smaller, more practiced imitation of insular script. He did not get very far in his proofreading, however. James Matthews died in 1787.
<%= image_tag "tha-1a.jpg" %>
Thorkelin A's skill in copying insular script improved radically in the course of his work, and it is usually easy to detect later additions and corrections by him. If they involve the restored text, these later changes are listed in the textual notes. (Note that 'Layout' is set to 'Single' view here.)
<%= image_tag "tha-2a.jpg" %>
An 'Edition Search' under 'Apparatus' > 'Early Restorations' > 'Transcripts' includes in its results complete lists of later restorations.
The reliability of Thorkelin's own transcript presents many difficulties for an editor, in part because Thorkelin had Thorkelin A to consult as he made the second transcript, in part because he was devising a first edition and Latin translation.
<%= image_tag "thb-1.jpg" %>
Thorkelin began but later erased an interlinear Latin translation in A (pp. 2-6). His occasional dependence on A in later proofreading, rather than on the manuscript, is here evident in his transcript of fol. 130r17, where he begins to omit the same line as A.
<%= image_tag "thb-3.jpg" %>
Thorkelin B is at its best when A does not provide readings, either on purpose or by accident. Although he oddly abandons following the manuscript line by line on the last page, B alone attests to the ð in bið 198v17 (original d crossed later, however) and to the rare macron over g in freoge 198v18 (erroneously expanded to freogen).
For an introduction to these collations, see Kiernan, <%= link_to '"The Conybeare-Madden Collation of Thorkelin's Beowulf."', 'http://www.uky.edu/~kiernan/Cony_Mad/', :target => "_blank", :class => 'links' %>.
Madden first copied Conybeare's collation into his copy of Thorkelin's Beowulf. He later added his own, more accurate, collation.
<%= image_tag "cony-mad.jpg" %>
Both Conybeare and Madden interleaved their copies of Thorkelin. In the course of collating the edition with the manuscript, they frequently noted important information on the interleaves.
When an interleaf contains a note, the O-button on the top menu will be green <%= image_tag "Green-O.jpg" %>, and the drop-down menu is live. Click the arrows to open a drop-down menu and gain access to the interleaf or interleaves.
To view an interleaf correctly facing the page it annotates, first go to Views > Change Views, and then choose Left/Right layout and Free browse mode to display the same page in both frames. Here Conybeare is chosen for both frames.
<%= image_tag "freecoll.jpg" %>
Then click the arrow on the appropriate frame, here the left one, to open the interleaf:
<%= image_tag "collpick.jpg" %>
The collation and its interleaf will now be facing pages, as in the copy of Thorkelin's edition:
<%= image_tag "freecol2.jpg" %>
Conybeare discovered the misplacement of text at fol. 131 and noted on the interleaf facing p. 9 that fol. 132 should follow fol. 130.
On page 58 Conybeare correctly relocated the misplaced fol. "131" between fols. 146 and 147; to keep track of its former incorrect and now correction location, it is numbered fol. 147A(131) in the MS foliation.
Madden duly copied Conybeare's note identifying the misplaced leaf, 131, now 147A(131), and added to Conybeare's note on p. 58 "see p. 9 addition."
On his interleaf to p. 9, Madden provided a transcription in Old English script, showing how many letters were lost along the right edge when he collated.
Madden made a skillful tracing, now the frontispiece of his collation (Harvard University, Houghton 28286.24.3*), of the first nine lines of the 'Beowulf' manuscript.
Madden also made many careful facsimiles of damaged sections of the manuscript, revealing what appeared to him the exact state of the manuscript in 1824, long before the leaves were inlaid in their protective frames.