This is an electronic version of Kiernan, “Part One: Thorkelin's Discovery of Beowulf,” The Thorkelin Transcripts of ‘Beowulf’, Anglistica XXV (Copenhagen: Rosenkilde and Bagger, 1986), pp. 1-41.

Part One: Thorkelin's Discovery of Beowulf

We know that Grímur Jónsson Thorkelin (1752-1829) published the editio princeps of Beowulf (1815) and that his transcripts of the poem preserve for us hundreds of readings now lost to the manuscript because of fire-damage. But despite his primary and still crucial role in Beowulf studies, we know very little about him and his epoch-making trip to the British Museum. It has long been assumed that he went to England in 1786 to study the Beowulf manuscript and that for an unexplained reason he then hired someone else to copy the poem before he himself made a copy. Neglected documents in the Rigsarkivet in Copenhagen now reveal that he did not know of the existence of Beowulf until after he began doing research at the British Museum.¹

The poem first came to his attention in England through Humfrey Wanley's misleading catalogue description, which presumed that Beowulf the Dane of the opening lines was the hero of the poem.² It was a happy fault in this case, for Thorkelin's special interest in Beowulf, reflected in the title of his edition, De Danorum Rebus Gestis ...Poëma Danicum Dialecto Anglosaxonica,³ was that it represented for him a new source of Danish antiquities, the real objective of his research trip. Once he found out about the poem, Thorkelin presumably hired a copyist because he himself was occupied with his broader research interests. Documents at the Rigsarkivet and the Kongelige Bibliotek in Copenhagen and at the British Museum and British Library in London, moreover, give us reason to believe that Thorkelin's copy of the manuscript was made not in 1787, as he says, but sometime between 1789 and 1791, perhaps after April 1790, when he may have been a member of the staff at the British Museum. We may have two copies of the Beowulf manuscript because an edition of an Old English poem about Danish Scyldings was an ideal project for an antiquary who had not yet decided whether to work in England or in Denmark.

At the time of his trip, Thorkelin was Regius Professor of Antiquity in the University of Copenhagen and Keeper of His Majesty King Christian VII's Privy Archives.⁴ In 1785, by virtue of his position and his special qualifications, he was granted a handsome yearly stipend by Christian VII (1766-1808) “to travel through Great Britain, Ireland, and the Isles, for two years in order to collect and record all the extant Danish and Norwegian Monuments, Deeds, and Documents ... on his promise to deliver on his homecoming to Our National Archive and the great Library all the Collections he in such manner may procure.”⁵ Thorkelin's career as a professional archivist began in 1777, when he became Secretary of the Arnamagnean Commission, Árni Magnússon's legacy to the University of Copenhagen of “a collection of 1761 manuscripts, and several thousand original Charters relating to the history of Scandinavia.”⁶ In 1780 Thorkelin became assistant in the Gehejmearkivet, now the Rigsarkivet, and in 1783 he was appointed “Professor extraordinarius” at the University, next in line for a chair on the Philosophical Faculty.⁷

A similar but more exhalted honor came his way in the following year, when Christian VII promised him the post of National Archivist as soon as it became available. A copy of the King's letter is among Thorkelin's personal papers in the Rigsarkivet:

Since our Professor Thorkelin has served in Our National Archive, and in addition has worked on editions of Snorre Sturleson and other Works, and has both the Age and the Knowledge, and the right Credentials, We with justice can entrust him, when the National Archivist's Position falls vacant, with this Post: So give we to the foresaid, Our Professor Thorkelin, by this Our Written Pledge in the Office of the National Archivist, and with it the established Stipend, as soon as the same becomes vacant, at the death of the current National Archivist.⁸

The presiding archivist in 1784 was Christian Eberhard Voss, who at the age of seventy-five was appointed in the same year, 1780, that Thorkelin joined the service.⁹ Perhaps Thorkelin's research trip to Great Britain and Ireland, planned and funded in 1785, was intended as much for Voss's peace of mind as for Thorkelin's contribution to the Royal archives. At any rate, Thorkelin did not return from abroad until Voss's death in 1791.

There was respected precedent for Thorkelin's trip. Danish antiquarianism in the eighteenth century was aimed at retrieving a knowledge of the Danish presence in medieval times. Thorkelin saw his own research project as an essential supplement to the work of a former Danish National Archivist, Jakob Langebek (d. 1777), whose Scriptores Rerum Danicarum Medii Ævi included the results of extensive research trips to Sweden, Finland, Russia, and the other Baltic lands the Danes had influenced. Langebek had visited Great Britain, too, but his results there were comparatively sketchy, as Thorkelin naturally took pains to suggest in his own research proposal. In fact, Langebek twice cites Wanley's catalogue description of Beowulf in Volume I of Scriptores Rerum Danicarum, the second time marvelling that no Englishman had yet produced an edition:

In Bibliotheca Cottoniana asservatur Codex membran. qvo continetur Poema Anglosaxonicum vestustum & egregium, in qva descripta videntur bella, qvæ Beovulfus qvidam Danus, ex Regia Scyldingorum stirpe ortus, gessit contra Sveciæ Regulos, testante Hicksio T.II.p.219. Miror, neminem eruditorum Angliæ curiosum fuisse de tante antiqvitatis editione. qva & suis propter Poesin & nostris Historiæ causa infinite gratificarentur.¹⁰

Despite Langebek's evident enthusiasm, there is no indication here that he had seen the poem or even knew that the Cotton Library had become part of the holdings of the newly founded British Museum. And while Langebek's comments, which stress the potential interest of the poem to both Englishmen and Danes, may have ultimately led to Thorkelin's decision to edit Beowulf, there is decisive evidence that Thorkelin first learned about the poem the same way Langebek did, by reading Wanley's catalogue.

The drafts of Thorkelin's research proposal are among his personal papers in the Rigsarkivet.¹¹ In his proposal Thorkelin stresses his expertise in diplomatics,¹² the work he had already accomplished at the Gehejmearkivet in locating and arranging royal documents, and the large number of additional documents that could be recovered in Great Britain and Ireland, following the example of Langebek, who “at royal expense had travelled for some years” in the lands surrounding the Baltic. Thorkelin specifically mentions the need to search the libraries in Oxford, Cambridge, Edinburgh. and Dublin, as well as Cathedral churches, the Tower of London, and the British Museum. He alludes to deficiencies and omissions, so far as Danish and Norwegian interests are concerned, in the existing English publications of historical documents, and he ends by asking “that it might be most graciously allowed and commanded me to make a trip for two years through England, Scotland, Ireland, and the neighboring Isles, in order to record and collect everything that is found relating to the Acts of our Forefathers, be it in Charters, Antiquities, and Annals. Whatever Collection at the end of the trip shall be put in Your Majesty's National Archives or the Great Library.” ¹³ It was an extremely ambitious proposal.

The official duration of the trip sanctioned by Christian VII was from 1 July 1786 to 30 June 1788, but the initial grant was renewed three times, allowing Thorkelin to remain abroad until 1791. Partial records of the renewals are printed in Fonden ad Usus Publicos under the dates 1 December 1790 and 16 January I793.¹⁴ The first of these renewals allowed him to stay in Great Britain an additional year, until the end of June 1789, because “his plan of study was considerably disturbed by one of those two members of Parliament, [William Wyndham] Grenville or [Humberston] Mackenzie,” who had recommended that Thorkelin accompany them to Iceland — “whereby he was held in uncertainty for three months, caused considerable expense, and finally in the month of June 1788 was informed that he would not be taking the trip.”¹⁵ Other reasons cited for the requested extension were that the libraries of Dublin were closed during the summer and would not reopen until the convening of Parliament in January(!), and that “notwithstanding all his diligence” the wealth of relevant material at the British Museum and the Tower of London still required considerable time for him to transcribe.¹⁶ The second extension does not mention the reason for Thorkelin's extended stay for two additional years, but merely records the King's command to reimburse Thorkelin for funds granted on 25 January 1791. However, an official letter from Copenhagen to the Danish ambassador in London, dated 11 May 1790, and copied by Thorkelin on 22 May 1790, reveals that a one-year extension was granted at this time so that Thorkelin could procure a British process for making potash from seaweed.¹⁷ Thorkelin's research in Great Britain was indeed far-ranging.

As this letter illustrates, it is possible to reconstruct to some extent the nature of Thorkelin's research projects in Great Britain and Ireland from his personal papers in the Rigsarkivet. The papers tell us, for example, that in addition to manuscripts, he was interested in trees and trade, farming and manufacturing, contemporary literature and politics, as well as folklore, runology, and archeology. Though London was his home base,¹⁸ he travelled constantly, as his grant stipulated, to Scotland, Ireland, and the Isle of Man; and if he had had his way he would have gladly gone to his native Iceland, too. One folder in his archive at the Rigsarkivet remarkably documents some of his travels with “Skotske Brittiske Borgerbreve” (with seals still attached) awarded between 1787 and 1790 by Aberdeen, St. Andrews, Dornoch, Dundee, and Wick.¹⁹

In a notebook he took to Ireland in 1789 he describes St. Stephen's Green, and makes sketches of runic crosses with “intertwisted serpents,” and relates an anecdote that may have reminded him of the ending of Beowulf:

The Countess of Moira told me that the common people of Ireland still continued to follow their deceased friends to the grave with mournful tales - they even hired some poor people, who recited over the coffin the virtues of the dead, & put questions to him why he had left his relations & dear acquaintances. I was first surprised at this, but recollected soon that but short time had elapsed since mourning or funeral sermons had been in use.... The funeral Psalms sung in Danmark & Iceland during the deposition of the remains into the grave are perhaps nothing more than a supply of those songs, our heathen ancestors composed to the memory of their heroes after their death.²⁰

We know from the BL Reading Room Register that Thorkelin used the Beowulf manuscript just before he left for Ireland.²¹ In any event it is more than likely that he had been studying the poem in Thorkelin A since early 1788.

Thorkelin's personal papers, Rigsarkivet 6431.5 and 6, together with the Reading Room Register for the years of Thorkelin's trip, allow us to reconstruct to some extent his modus operandi, the way he prepared for his work as well as the way he went about some of his most important research, including his discovery of Beowulf. Some revealing notebooks survive among his personal papers. The box marked 6431.6 E.3 contains two folders, one titled “Thorkelins Reise til Stor Britannica og Irland, 1785-1791,” and the other, in Thorkelin's own handwriting, “Memorandum til min Engelske Scotscke og Irland Reise. 1785- 1791."²² The latter contains a small packet identified as “Thorkelinske Noticer fra hans Englandsrejse,” with three notebooks inside. One, headed Biblioth. Cotton., provides a contemporary record of Thorkelin's eventual discovery of the Beowulf manuscript in the course of his research in England,²³ while the Reading Room Register gives us the actual date of the discovery.

The Cotton Notebook shows that Thorkelin prepared, either before he went to England or in the course of his research, a list of Cotton manuscripts that appeared to him to have some possible bearing on Danish antiquities. To prepare his list, he used Thomas Smith's Catalogus Librorum Manuscriptorum Bibliothecæ Cottonianæ, a carelessly prepared work that omitted all mention of Beowulf in the otherwise flawed account of BL MS Cotton Vitellius A. xv.²⁴ The Nowell Codex, the part of Cotton Vitellius A. xv containing Beowulf, actually consists of five items: a St. Christopher fragment, The Wonders of the East, Alexander's Letter to Aristotle, Beowulf, and the Judith fragment.²⁵ Yet Smith's catalogue records only two items: Translatio epistolarum Alexandri ad Aristotelem, cum picturis de montrosis animalibus Indiæ and Fragmentum de Juditha & Holopherne. It is not surprising then that Thorkelin failed to include Cotton Vitellius A. xv when he first prepared the list of manuscripts he wanted to study at the British Museum. Fortunately, he later supplemented his list by reference to more reliable sources than Smith.

At the British Museum Thorkelin's general method was to study all the manuscripts he had listed in his Cotton Notebook to see if they did indeed contain material relevant to Danish history. This stage of his work is partly recorded in the Reading Room Register and is further supplemented in the many notations he added in his personal notebooks as the case warranted. In his Cotton Notebook, if a manuscript had something worth transcribing he would write “NB” in the margin, as he did for Cotton Julius A. xi; if a manuscript proved to be of no value to him, he would write “nothing” in the margin, as he did for Julius E. vi, also on the first page of his notebook. After reading Wulfstan's Sermo Lupi ad Anglos contra Danos in Nero A. i, he wrote “Nihil” in the margin (fol. 3v), betraying some forgiveable bias in his selection of relevant material. Other typical notations illustrating Thorkelin's work-in-progress in London are, “Charta Canuti” (often), “nothing/ on paper.” “illegible from burning,” “incip. post conquestum,” and “Not yet perused.” To aid his search for unpublished chronicles, he noted those manuscripts that had not been used by Edmund Gibson in the Chronicon Saxonicum.²⁶ For example, he first listed “Chronicon Saxonic. ab Christo nato ad 977” from Tiberius A. vi; he later added the note, “not used by Gibson”; later still he crossed out the entire entry, writing “used” in the left margin and “deliv[ero]” in the right. His repeated use of some manuscripts is reflected, too, in the Reading Room Register: it shows, for instance, that Thorkelin first studied Tiberius A. vi on 7 January 1788, and later borrowed it from 10 February to 20 March 1788. Once in London he supplemented Smith's catalogue with Wanley's Catalogue and David Casley's A Catalogue of Manuscripts of the King's Library: An Appendix to the Catalogue of the Cottonian Library.²⁷ There are many entries in the notebook like the one added to Julius A. ii, “Vid. Casley p. 318.”

But for us the most interesting addition to Thorkelin's original list of Cotton manuscripts occurs on the fifth leaf of the notebook. Between Vitellius A. xiii and Vitellius A. xvii, Thorkelin added the entry Vitellius A. xv, squeezing in the note, “See Wanlej in Hickes.” He had found out about a Dane named Beowulf who fought wars against the Swedish princes.

[Fig. 1: Thorkelin's Discovery of Beowulf]

The Reading Room Register shows that he made this discovery on 3 October 1786, about two months after he began working at the Museum. Thorkelin took another look at Beowulf during the day of 16 October 1786. But according to the Register he did not use the Beowulf manuscript at all in 1787, when both transcripts were supposedly made, and if the Register is reliable, he did not even consult the manuscript again until 1789. It should be kept in mind that the notebook entry merely adds Beowulf to a long list of manuscripts Thorkelin intended to study later. Although he did go out of his way to locate the poem, abandoning the order of the entries in his Cotton Notebook to look for it, and probably commissioning Thorkelin A soon after finding it, the date of his discovery should not be confused with the date of Thorkelin B, which all of the available evidence suggests was made considerably later than 1787.

A transcription of the Vitellius entries on this page of his Cotton Notebook, with Thorkelin's later notes bracketed, can help us keep his discovery of Beowulf in his rather than our perspective:

Vitellius A x Annales p. 82
   Alphabetum Norwegicum
[nothing] p. 83 xiii Effigies Regum Angliæ
[xv See Wanlej in Hickes]
[Vide] xvii Annales Wintonenses
[p. 84] xx Historica [nothing]
p. 86 C viii Quædam Chronica
[nihil]
p. 87 [NB] ix Chartæ Regum [NB]
p. 91 D vii Collectanea Noelli
[nothing] inter quæ Chartæ [23. Ch. Canuti)
xviii Annales et Regista
   Stæ Mar. de Cumba

Thorkelin's later notes can be approximately dated by reference to the Reading Room Register. The addition of the “Wanlej” entry must have been made in October 1786, when Thorkelin learned about Beowulf. The other added notations were all apparently made on 28 May 1789, when Thorkelin in the course of his research came to the Vitellius entries in his notebook. On that day, Thorkelin first borrowed Vitellius A. xiii, xv, xvii, and xx, and it must have been then that he wrote “nothing” next to A. xiii and A. xx, and “Vide” next to A. xvii. Later in the day he returned all four manuscripts, including the Beowulf codex, and borrowed Vitellius C. viii and ix, D. vii, and E. v. After studying them, he wrote “nihil” after C. viii, “NB” two times beside C. ix, and “nothing” and “23. Ch. Canuti” beside D. vii. He returned these four manuscripts on the same day, too.

The Reading Room Register, then, forces us to question the 1787 date Thorkelin gives for his transcript of Beowulf. Though it tells us that he discovered the poem late in 1786, and that he had enough interest in it to borrow it twice, the Register also tells us that he was more interested in Icelandic manuscripts at the time, and that he did not acquire his transcript in 1786, for the manuscript was in his possession for only two and a half days at the most. It tells us that he did not borrow the manuscript at all in 1787, or even in 1788, and that in 1789 when he did borrow it again, he concentrated instead on other Vitellius manuscripts. At this stage in his research, Vitellius C. ix was manifestly more important than Beowulf to Thorkelin. In fact, he pinned to the page in his Cotton Notebook a note that covers up the reference to Vitellius A. xv. The note reads:

      Bibl. Cottoniana
Otho. A xii
   de Brithnoto Northan-
          humbrorum Duci
              Thorkelin
          Burnt together with
              9.10.11. & 13 —
Vitell. C ix Vita Griffini quædam
            & Danis & Norwegis

He had learned about the Battle of Maldon in the course of his research, but then found out that the manuscript had been destroyed in the fire of 1731. He also had learned that Vitellius C. ix contained, in addition to royal charters, a Life of Griffin, which he duly transcribed along with the charters, the annals, and of course Beowulf. The Register helps us date these later entries, too, since it shows that Thorkelin first studied Vitellius C. ix on 28 May 1789. He later borrowed this manuscript from 1 to 21 September 1789 to make his transcripts of the Vita Griffini and the Ystoria Gruffudd ap Kynan Brenin Gwynedd in it.

The Beowulf transcripts, then, represent a relatively small part of the “Collectio Thorkelini” gathered in Great Britain and Ireland between 1786 and 1791. Thorkelin's benefactors expected him to return to Denmark with a wealth of Danish and Norwegian antiquities, not with only two copies of the same poem, and he did not disappoint them. He found and transcribed many pertinent texts. Unfortunately, his “collectio” no longer exists as a discrete collection. It is now difficult to know for certain how much material Thorkelin copied and had copied for him on his long research trip. Some of his manuscripts were recatalogued under new bindings, some were dismembered and rebound with new shelf marks without records of the changes, and some were recopied by him in Copenhagen for at least one special collection of charters. The four-volume Catalogus Manuscriptorum in Collectione Nova Regia at the Kongelige Bibliotek inaccurately lists fifteen different Thorkelin manuscripts under “Historia Magnæ Britanniæ & Belgii foederati.” They are Ny kgl. Saml. 261c, 261d, 266c Fol., 511, 511b, 512 and 513 (the two Beowulf transcripts), 513b, 513d (including the Life of Griffin transcripts), 513e, 513f, 514b, 516b, 518d, and 518e.²⁸

In fact, Ny kgl. Saml. 511 and 511b no longer exist as such, much of their contents having been rebound in the reign of Frederik VII (1848-63) as the three manuscripts now numbered 261c, 261d, and 266c Fol. and the six numbered 513d to 518d. Notes in pencil on the first pages of these manuscripts identify the item numbers of the original manuscript, so we still have some notion of how many of Thorkelin's transcripts from Ny kgl. Saml. 511 have yet to be found as well as how many are extant. Though Thorkelin does not date any of these composite collections of his transcripts, nor for that matter any of the individual transcripts in them, it is possible to determine termini a quo from the Reading Room Register.

Thus, the earliest date we can assign for No. 31 in Ny kgl. Saml. 261d is between 10 February and 20 March 1788, when Thorkelin first consulted Domitian A. viii. Other possible dates are between 24 February and 9 April 1789, and between 8 October and 24 December 1789, when he was also using this manuscript. We can be fairly sure, in fact, that he copied No. 31 sometime between 8 October and 24 December, since for this item he also copied parts of Claudius D. ii, which he used only between 2-12 December 1789. Dating No. 32 is problematic, for Thorkelin says he used Harleian 746 (“Incipit lex que Anglice Danelage est vocata”) for this transcript, yet there is no record in the Reading Room Register that he ever used this manuscript. The lack of such a record may mean that the transcript was made after April 1790, when it appears that Thorkelin was no longer required to use the Register. Another possibility is that 746 is an error for 785, another copy of the same text, which he signed out from 24 November to 24 December 1789. His transcript contains collations from Domitian A. viii, which as we have seen he was using at the same time as Harleian 785. Ny kgl. Saml. 513d was probably made in 1789, for Thorkelin first consulted Vitellius C. ix on 28 May 1789, and then borrowed the manuscript from 1-21 September. There are no further entries of this manuscript in the Register.

Dating Ny kgl. Saml. 513e is considerably more complicated: the earliest transcript, No. 13, probably dates from 27-29 September 1786, when Thorkelin borrowed Julius A. i; the other transcripts are much later. Thorkelin first used Domitian A. xiii for No. 14 between 8 October and 24 December 1789; for No. 16 he used Vespasian A. xxii from 17 June to 3 August 1789 and again from 22-24 September 1789; he first used Faustina A. viii on 4 April 1787, but he probably made his transcripts for No. 17 sometime in the months after 16 February 1790, when he signed the manuscript out for at least two and a half months; for No. 18 he used Tiberius E. iv from 19-20 August 1788; for No. 19 he borrowed Titus A. xiii between 24 September and 22 October 1789; he studied Harleian 3763 on 22 July 1788 and borrowed Nero A. iv from 22-28 May 1789 for No. 20; for No. 21 he used Domitian A. i on 8 October 1789; there are no Register entries for Claudius E. viii (for no. 24), which indicates that he must have used this manuscript after April 1790, when he was no longer using the Register.

For the transcripts in Ny kgl. Saml. 513f he first used Cleopatra D. ii on 15 February 1790, Claudius C. ix from 21-26 May 1789, Harleian 3776 on 22 July, and then again from 23 July to 13 August 1788; Harleian 2188 on 21 July 1788, and then again from 24 November to 24 December 1789. For those in Ny kgl. Saml. 514b he used Tiberius A. ix on 3 March 1790, Cleopatra D. iii on 15 February 1790, Vespasian A. xviii from 17 June to 3 August and from 22 September to 24 December 1789, and Faustina A. viii from 16 February beyond the end of April 1790, as well as on 4 April 1787. The Register enables us to date his use of only two of the four manuscripts copied for Ny kgl. Saml. 516b (Vespasian A. vi, from 22 September to 24 December 1789, and Domitian A. vii, from 24 November to 24 December 1789), so the other two must have been used after April 1790.

Unfortunately, Thorkelin did not identify the manuscripts he used for Ny kgl. Saml. 518d. O'Flannagan dates his transcript, Ny kgl. Saml. 518e, in 1789, while the second copy in Ny kgl. Saml. 266c Fol. (also numbered 47) is precisely dated 30 July 1789. The material from the Isle of Man in Ny kgl. Saml. 261c is undated, but there is a second transcript from Man, Ny kg. Saml. 268 Fol. (without an original number from Ny kgl. Saml. 511), “A true Copy of the Spiritual Laws and Customs of the Isle of Mann & Compared with the original by J A O'Cubbon, Archdeacons Register” that is dated “Sept. 21st 1789.” The evidence shows, then, that nearly all of these transcripts were made between 1788 and 1790. As we shall see, the same is certainly true of one other transcript Thorkelin dates in 1787, and probably true of his Beowulf transcript, Thorkelin B, as well.

Not all of the item numbers from Ny kgl. Saml. 511 and 511b are accounted for, of course, in the rebound manuscripts. There are extant numbers from 3 to 52 in them, but even in the unlikely event that No. 52 in Ny kgl. Saml. 513f was Thorkelin's final transcript, reflecting the total number of transcripts he made, twenty-one are still missing (Nos. 1, 2, 4, 12, 15, 25, 26, 30, 36-42, 44, 46, and 48-51). We can account for two of the missing transcripts by assuming that the unnumbered Ny kgl. Saml. 268 Fol. either preceded or followed 261c (No. 43), since both transcripts come from the Isle of Man; and that one of the copies numbered 47, both from Ireland, should be numbered 46 or 48. That would leave nineteen transcripts still missing. It is more likely, however, that one 511 collection was restricted to folio-size transcripts, like the two transcripts in question - 266c Fol. (number 47) and 268 Fol. (unnumbered). That would mean that at least twenty-one transcripts were missing from one 511 collection, while at least forty-five were missing from the other. In any event. Thorkelin probably recopied many of these missing items for Ny kgl. Saml. 495b fol., Fortegnelse paa diplomer,²⁹ now housed in the Rigsarkivet as Register 122.

This register includes many of the same charters from the reign of Cnut that Thorkelin had so carefully noted in his Cotton Notebook. It is reasonable to suppose that he originally kept these transcripts with the other ones in Ny kgl. Saml. 511. Moreover, this assumption is supported by the fact that two charters from Harleian MSS 258 and 3763 remain, apparently by oversight, in Ny kgl. Saml. 513e. In any event, the entries in Register 122 record his use of many manuscripts.³⁰ Some of the transcripts from this register can be dated relatively early, since Thorkelin first used Cotton Augustus ii on 3-4 October 1786, Vespasian C. xiv from 8 November to 19 December 1786, Harleian 7007 from 10 November to 21 December 1786, Cotton Nero B. iii from 28 February to 4 April 1787, and Rymer's Foedera all of the time that he was working at the British Museum in 1787.

Thorkelin surely made many other transcripts, which will perhaps be identified in their turn.³¹ We know from the Reading Room Register, as well as from the first renewal of his grant and from extant letters to his patron Johan Bülow, that in 1787 he had found and intended to transcribe material relating to the Palatinate and to Iceland. A small part of his work on the former is preserved in Ny kgl. Saml. 1019, a 900-page transcription from the “Letter-Books of Sir Thomas Roe.”³² Most of the material he must have copied relating to Iceland has yet to be found. They may eventually turn up among the transcripts at the Arnamagnæan Institute in Copenhagen or Reykjavik,³³ or among the Icelandic manuscripts Thorkelin sold in 1812 to the Advocates' Library, now in the National Library of Scotland.³⁴ There are preserved, as well, a few uncatalogued transcripts in his working notebooks.³⁵ Nonetheless, it appears that the bulk of Thorkelin's research on the Danish and Norwegian presence in great Britain and Ireland during the Middle Ages was brought back to Denmark in the now dismembered and dispersed 511 and 511b. which Thorkelin describes in the Catalogus Manuscriptorum in the following way:

511: Duo Volumina tractatuum variorum ad historiam, imprimis antiquam, Magnæ Britanniæ facientium ex Codicibus manuscriptis in bibliothecis britannicis asservatis, exscriptorum, opera & manu Gr: Joh: Thorkelini.    

Primum continet Annales & excerpta ex Annalibus Chronicisque inprimis de Danorum gestis & imperio in Anglia; Secundum continet Annales, &c: historiam Manniæ & Irlandiæ illustrantia. - Præmittitur cuivis Volumini elenchus contenti.

511b: Varia ad historiam Anglosaxonum: a, Excerpta ex brevi chronico de britannicis insulis; b, Prafecturæ comitatus & insula quæ pertinent corniæ regni Britanniæ; c, Excerpta ex Britonium Historia (Anglosax.) d, Chronica Anglosax. e diversis codicibus (Anglosax.) e, Compendium historia de regibus Anglosaxonicis (lat.) f, Duo hymni de victoriis Regis Athelstani contra Scotos & Hibernos (Anglosax.); g, Comites Northumbria.

Without the lost elenchus contenti we cannot be sure whether Ny kgl. Saml. 511b is merely volume two of 511, or a third volume of Thorkelin transcripts. All of the manuscripts now numbered 513d through 518d, as well as 261c, 261d, and 266c Fol., derived their contents, according to the notes, “Af No 511;” none of the notes mentions “511b.” Moreover, at least one item from 511b, Duo hymni de victoriis Regis Athelstani contra Scotos & Hibernos (Anglosax.) - two versions of the Battle of Brunanburh - has not yet turned up among Thorkelin's extant transcripts. It is possible that Thorkelin ultimately suppressed this transcript because it recorded a decisive Scandinavian defeat, but it is also possible that all of the transcripts in 511b survive somewhere.³⁶

At any rate, it seems safe to say that Ny Kgl. Saml. 511 and 511b had been cheaply bound in England to prevent their many transcripts from being lost or misplaced. The books were then dismembered in Copenhagen to form more coherent collections. We may infer that Thorkelin was responsible for the dismemberment, since it was he who recopied so many items for Register 122 now in the Rigsarkivet. Presumably he himself arranged the remaining items into the smaller units in which they now exist, but when he died in 1829 they still had not been rebound. In addition to the manuscripts from 511 and 511b, Thorkelin brought back four bound transcripts that are also listed under “Historia Magnæ Britanniæ” in the Catalogus Manuscriptorum: i.e., 512 (Thorkelin B), 513 (Thorkelin A), 513b (Lydgate), and 518e (Annals of Innisfallen). None of the bindings are Danish; all are almost certainly English.³⁷ They vary greatly in quality, perhaps suggesting that Thorkelin had different intentions for their use.

Ny kgl. Saml. 513 and 513b were sumptuously bound, with the Kongelige Bibliotek evidently intended as their destination. The former, Thorkelin A, or the transcript of Beowulf made by Thorkelin's copyist, has a fine, marbled, brown-calf binding, with gilded borders and spine, titled on the spine, Hist. Dan. Anglosax. The latter, Lydgate's poem on the Life of St. Edmund, was transcribed by Thorkelin and beautifully illuminated by a “Mr Oneal of Clarkenwell,” copying the manuscript illuminations.³⁸ It is an imposing book with an expensive, straight-grained, red morocco binding, displaying gilded neoclassical designs on borders and spine. At the top of the spine, one of the vase-like ornaments has been effaced and replaced by the royal cipher of Frederik VI (1808-39). The paste-downs and flyleaves are marbled in the English fashion.

Ny kgl. Saml. 512 and 518e, on the other hand, were inexpensively bound, perhaps because Thorkelin intended them for his private use, with editions ultimately in mind. Ny kgl. Saml. 512, Thorkelin's own transcript of Beowulf, has an ordinary half-binding in calf, with marbled paper boards, and no elaborate gilding except for the title on the spine, Poema Anglos de rebus Daniciis. Ny kgl. Saml. 518e, the transcript of the Annals of Innisfallen made for Thorkelin by Theophilus O'Flannagan, has marbled paper boards and only a cloth spine.

If Thorkelin originally intended these last two transcripts for the Kongelige Bibliotek, it seems odd that he had them bound so plainly. Some evidence supports the view that they were in fact for his own use. The O'Flannagan transcript was evidently deposited in the library after Thorkelin's death in 1829: it was certainly numbered 518e long after his death, in the reign of Frederik VII (1848-63), when Ny kgl. Saml. 511 and 511b were rebound as manuscripts 513d-518d. Thorkelin's transcript of Beowulf was undoubtedly placed in the Kongelige Bibliotek during Thorkelin's lifetime, but presumably not until he had finished his edition. This conclusion is based on the title-page of Thorkelin B, written in Thorkelin's late, shaky handwriting; by contrast, the title- page for Thorkelin A was written while Thorkelin's hand was still steady.³⁹ The current shelf marks for Thorkelin B and Thorkelin A, Ny kgl. Saml. 512 and 513 respectively, would appear to refute the evidence of the title-pages, but in fact the paste-downs on the front covers of the two transcripts reveal that A was formerly numbered 512, and B was 513, reflecting the true order in which they were copied. It would seem to follow, then, that the transcript of Lydgate's poem, now Ny kgl. Saml. 513b, was formerly numbered 513. If so, we can make some intelligible order of the number sequence of the main transcripts Thorkelin originally deposited in the Kongelige Bibliotek when he returned from his research trip in 1791: 511 and 511b (the miscellaneous transcripts), 512 (Thorkelin A), and 513 (Lydgate's poem). When Thorkelin added his own transcript of Beowulf to the royal collections, he naturally gave it some pride of place, numbering it 513, and altering the original 513 to 513b. Ultimately not satisfied with second place for his transcript, though, he then reversed the call numbers of Thorkelin A and B.

Apparently it was Thorkelin himself who, at this time and in this connection, mutilated the title-page of the A transcript. The current state of the page, at any rate, seems to fit the proposed circumstances. The bottom third of the page has been cut off with scissors, eliminating a substantial part of the original text. Part of what remains of the title has been rather sloppily emended by Thorkelin in another ink. At least four lines of text were excised at the bottom of the page: bits of the first line can be seen in the facsimile along the cut edge, while in the manuscript the beginnings of three additional lines can be seen along a narrow strip (not reproduced in the facsimile) that was left to preserve the stitching in the lower half of the page. We cannot now know what these lines read, but it is likely they were somehow related to Thorkelin's emendation of the remaining title. The page now reads,

Poëma Anglosaxonicum de Rebus gestis Danorum ex Membrana Bibliothecæ Cottonianæ Vitellius A. in Musæo Britannico. fecit exscribi Londini A.D. MDCCLXXXVII. Grimus Johannis Thorkelin. LL.D.

Thorkelin's late emendation, fecit exscribi for original exscripsit, removed the false claim that Thorkelin himself had made the first transcript. The cutting of the page after binding and the late emendation both make sense if, as the evidence implies, Thorkelin B was a relatively late addition to the Kongelige Bibliotek. Thorkelin naturally wanted credit for making a transcript of Beowulf on his research trip, and presumbably he took credit for it on the title-page of Thorkelin A until he added his own transcript to the collections.⁴⁰

If so, Thorkelin's wish to give priority to his own transcript of Beowulf, to list it first in the library catalogue and to suggest by its shelf marks that it was copied first, may have induced him to falsify the real date of his transcript. There is no need to ascribe any motive for an inaccurate date here, however, for Thorkelin is repeatedly wrong or uncertain about dates in his extant notes. In his personal papers in Rigsarkivet 6431.5.D, for example, he states that his Sketch of the Character of His Royal Highness the Prince of Denmark was published in 1790, rather than 1791. As noted above, the Reading Room Register has no record of his use of Cotton Vitellius A. xv in the year 1787. Perhaps Thorkelin merely decided to date both transcripts in 1787, when Thorkelin A was finished, even though B was not finished until much later. In any case, there is no reason to trust the testimony of the title-pages. Thorkelin styles himself an LL.D. on both, yet he did not receive his honorary degree from the University of St. Andrews until 26 October 1787.⁴¹ The Reading Room Register reveals that he did not return to the Museum from 25 May 1787 until 7 January 1788, so neither title-page is a contemporary record.

Though the Reading Room Register does not have an entry at all for the A transcript, the procedure for obtaining hired transcripts at the British Museum in the 18th century helps explain the silence of the Register. Thorkelin regularly had transcripts made for him on his trips to Ireland and the Isle of Man, but the only commissioned transcript from the British Museum that has so far come to light is Thorkelin A, the first Beowulf transcript. The reason is that professional transcripts from the Museum were expensive, as Thorkelin explained to his patron, Johan Bülow, in a letter dated 3 April 1787:

It is infinitely difficult to gain access to the archives here, and copies are not permitted except by the custodians, who therefore demand unheard of prices. A single diploma can cost a whole guinea. Despite all this I still have the best hope, notwithstanding the great cost.⁴²

Thorkelin's remarks are clarified and confirmed by two rules in the earliest Reading Room regulations:

1. That no one be admitted to make use of the Museum for Study, but by leave of the Trustees, in a General Meeting, or the Standing Committee: and that the said leave be not granted for a longer term than half a year without a fresh application.

4. That such persons be allowed to take one or more Extracts from any printed Book, or Manuscript: and that either of the Officers of the department to which such printed Book or Manuscript belongs, be at liberty to do it for them, upon such Terms, as shall be agreed on between them.⁴³

We can be reasonably sure, then, that the anonymous A transcript was made by a member of the Museum's staff, able to make a copy of Beowulf without having to sign out the manuscript in the Reading Room Register. Thorkelin probably ordered the A transcript shortly after he discovered the poem, and his letter to Bülow is a veiled reference to the only hired transcript for him known to exist from the British Museum.

The absence of records for Thorkelin A and B notwithstanding, the documentation of Thorkelin's research at the British Museum is quite revealing. The first reference to him appears in BL Add. MS 45870, a record of “Persons admitted to Reading Room / March 2d 1781 to June 9th 1795.” On fol. 29, “Mr Grimer Johnson Thorkelin from Iceland had Leave till the next Meeting of the Committee, when he desires to have it extended to six months. August 8th 1786.”⁴⁴ On the next folio a note signed by Joseph Planta states, “At a Committee Oct 27. 1786 — All the above leaves applied for were granted.” Thorkelin's first days in the Museum were doubtless marked by some linguistic frustrations, for he made his debut in the Reading Room Register as “Mr. Geo. Thorkelin,” borrowing Royal MS 14.C.1 (containing Geoffrey of Monmouth's De gestis Britonum) from 10 to 31 August 1786.⁴⁵ The next time his name appears in the Register it shows that he is beginning to study Old English manuscripts in the Cotton collection, starting with Ælfric's Grammar and Glossary in Julius A. ii, which he used with Julius A. i, and two other Julius manuscripts,⁴⁶ from 27 to 29 September. On 2 October he made two of his earliest transcriptions of Old English material by copying at the end of his Cotton Notebook the Life of Furseus in Julius A. x and the continuous glosses to Latin hymns and canticles from Julius A. vi; at the same time he copied a Latin Vita Oswini also from the martyrology in Julius A. x. The appearance of these transcripts in the Cotton Notebook itself seems to suggest that he had not yet embarked on a systematic program of making transcriptions. His interest in the grammar and the glosses early in his work implies that he was, however, preparing himself for reading Old Engish texts.

At the same time he was obviously studying catalogues of manuscripts, hunting for Danish antiquities, for on the next day the Register records his discovery of Beowulf in Vitellius A. xv. He must have been hunting in Wanley's Catalogue. In the first part of the day, though, he read Norman history in Julius E. vi, Ælfric's Homilies in Julius E. vii, sixteenth- and seventeenth-century letters in Julius F. vi, and a Latin chronicle in F. vii; he also began searching for useful charters in Augustus ii.⁴⁷ Later in the day he requested Vitellius A. xv. He returned it and the manuscript of charters the next day, 4 October. Since the Reading Room was only open from 9:00 a.m. until 3:00 p.m. he had at the most a day and a half in which to study Beowulf, time he was still sharing with the charters in Augustus ii.⁴⁸

How well Thorkelin came to know Beowulf on this first occasion cannot be known, though it is likely that his knowledge of Old English was still too rudimentary for him to have read the poem in a sitting. Certainly he did not have time to make a transcript. Since he also had the manuscript of charters to work with, it is possible that he could do no more than browse through the Beowulf manuscript at this time. At any rate when he wrote to Johan Bülow a month later, he seemed uncertain about his discovery. He said only that “among the manuscripts are not a few pertaining to Denmark; I hope to find something worthwhile, including an unknown Anglo-Saxon [work], containing a collection of songs about the exploits of the Danish kings in the 3rd century, and by which the good Saxo is increased not a little.”⁴⁹

This vague description of Beowulf, probably gleaned from the two sections of the poem transcribed by Wanley, suggests that Thorkelin had not yet read the poem. He knew he had found an important text, but he was unprepared to give it his full attention. On 16 October he again consulted Vitellius A. xv, but with four Icelandic manuscripts from the Banks collection.⁵⁰ He returned all of these manuscripts the same day, but the Icelandic material seems to have won his greater interest, for when he returned to the Museum on the 19th, it was to sign out two Icelandic-Latin dictionaries.⁵¹

His interest in Old English manuscripts is not again reflected in the Register until 1788, and his interest in Beowulf not until much later. From the end of October until 21 November 1786, he studied instead the Icelandic sagas in Add. MS 4861; from 8 November to 19 December he had out the sixteenth- and seventeenth-century papers of Cotton Vespasian C. xiv (containing a Danish ambassador's complaints of English fishing in Icelandic waters); from 10 November to 21 December he borrowed a half-dozen seventeenth-century manuscripts from the Harleian collection, none of which relates to Anglo-Saxon;⁵² and from 13 to 21 December he studied the Irish law codes in Harleian 432 and the saints' lives in Cotton Vespasian A. xiv.

In 1787, the year in which both transcripts of Beowulf are supposed to have been made, Thorkelin did not call for a single Old English manuscript. The Reading Room Register shows that Thorkelin's investigations took quite different turns in this year and that he borrowed books for only the first five months. On 3 January he used the Annals of Ulster in Add. MS 4795 from the Milles collection. By the end of the month, and for the remaining four months, his interest switched for the most part to English-Scandinavian affairs in early modern times. He pursued this fascination by studying manuscripts from the Rymer collection on 29 January, 27 February, 28 February (Add. MSS 4628, 4629, 4630), and 20 March (Add. MS 4574), and by borrowing Cotton Nero B. iv and v from 28 February to 4 April. On 28 March he used five more Nero manuscripts,⁵³ presumably for the information they contained relating to the Hanseatic League.

On 3 April he again studied Ayscough's catalogue of manuscripts, and then signed out Add. MSS 4168 and 4172 from the Birch collection, the “Letter-Books of Sir Thomas Roe.”⁵⁴ Thorkelin wrote to Bülow the same day to say that he had begun to transcribe the five volumes of Roe's journal, which he thought would come to about “900 Ark,” or 1800 pages. Only half of this work survives in the mammoth Ny kgl. Saml. 1019, from Add. MSS 4168 and 4172; the rest, presumably from Add. MSS 4169-71, has yet to be found. On the next day, Thorkelin consulted the De regibus Angliæ in Cotton Faustina A. viii, but continued reading the Roe Letter-Books. Then, from 16 April to 25 May he borrowed the entire Roe collection (Add. MSS 4168-4172) along with Add. MS 4861, containing some of the Icelandic material he published the following year in Fragments of English and Irish History. The last entry of his name for 1787 in the Reading Room Register shows that from 9 to 25 May he again borrowed Add. MS 4630 from the Rymer collection, a manuscript he had first studied on 28 February. Thorkelin spent the summer and fall of 1787 in Scotland, where he received his honorary degree from the University of St. Andrew's on 26 October. The next time his name appears in the Register is on 7 January 1788.

The evidence of the Register is that Thorkelin did not transcribe Beowulf in 1787. The possibility that the Register merely fails to record his use of the manuscript in that year is quite remote: he was by his own testimony busy with other manuscripts in the first part of the year and busy in Scotland in the second part. The Register suggests also that he had studied almost no Old English by then, and none at all at the British Museum.

Things were different after he returned from Scotland in 1788. On 7 January he took out Cotton Tiberius A. iii with its many Anglo-Saxon texts, including continuous glosses, prayers, and homilies; and the Anglo-Saxon Chronicle in Tiberius A. vi.⁵⁵ He used, too, Tiberius A. ix, presumably for its Annales de gestis Britonum, Saxonum, Danorum. Two days later he took out Tiberius A. ii, A. xiii, and a second version of the Anglo-Saxon Chronicle in Tiberius B. iv. On 9 January he also studied Tiberius B. v and C. xiii (Gesta regum Saxonum). From 10 February to 20 March he signed out the Anglo-Saxon Chronicle in Tiberius A. vi (the manuscript with the poems he had first studied on 7 January) and in Domitian A. viii (lacking the Chronicle poems). From 1 to 28 April he again borrowed Tiberius A. xiii, probably still for its charters, rather than for the Old English homily.

Toward the middle of the year Thorkelin was drawn from Cotton to Harleian manuscripts, and was led away from what seems to have been rather determined interest in Old English.⁵⁶ He borrowed Harleian 2278, Lydgate's Life of St. Edmund, for seven months, from 27 May to Christmas Eve 1788. His use of this manuscript provides a useful analogue to his use of the Beowulf manuscript. As in the case of the Beowulf transcripts, Thorkelin says on the title page of Ny kgl. Saml. 513b that he copied The Life of St. Edmund in 1787. In this case, however, the Reading Room Register shows that he did not borrow the manuscript until May 1788. Moreover, since he wanted to copy the entire manuscript and to hire an outside artist to copy the illuminations, Thorkelin was obliged to get permission from the Trustees. One of the rules of the Reading Room was “That no whole Manuscript, nor the greater part of any, be transcribed, without leave from the Trustees, in a General Meeting, or the Standing Committee.”⁵⁷ Thorkelin complied with this regulation in a letter dated 14 March 1788 to Joseph Planta, who was then Keeper of Manuscripts:

Among a great number of Manuscripts in the British Museum relative to the Danish History, I find the Life of Saint Edmund written by John Lydgate. Before I attempt to copy it, it will be necessary for me to apply for Leave, invested in the Trustees.
   The Danish Litterature will be for ever much obliged to you, if you on my account would please to intercede for the purpose of getting permission of copying the said manuscript & of admitting Mr Oneal of Clarkenwell, who has undertaken the drawings.⁵⁸

The Trustees' Minutes confirm that Thorkelin was granted permission to copy the manuscript on the same day.⁵⁹ Here, then, is a clear instance of Thorkelin's having antedated a transcript.

Thorkelin required seven months to complete his transcript of the Lydgate manuscript, and we might expect to find in the Register that he borrowed Beowulf for a considerable period of time, as well. The professional artist Thorkelin hired to copy the Lydgate illuminations was perhaps at work for some of the seven months, but Thorkelin again borrowed Harleian 2278 from 5 January to 9 April 1789, and the illuminations could have been copied during this period. In any case the Lydgate transcript provides at least a rough equivalent for the two transcripts of Beowulf, insofar as the Register shows when Thorkelin could have acquired both his own transcript and his commissioned illuminations for Harleian 2278.

There is no comparable documentation for the Beowulf transcript. This lack of documentation, however, can be reasonably explained. Neither transcript would have required specific permission from the Trustees, for neither was a copy of a “whole Manuscript, nor the greater part” of Cotton Vitellius A. xv. Indeed, the regulations specifically permitted Thorkelin, as an admitted reader, “to take one or more Extracts from any ... Manuscript.” Though the Reading Room Register does not indicate when Thorkelin made his transcript, we must attribute the absence of a record either to the clerk who kept the Register or to the unusual circumstance that Thorkelin was no longer required to use the Register after April 1790.

Despite the silence of the Register, there is reason to think that Thorkelin's scribe made his copy, as Thorkelin says on the title page, in 1787. The same regulation that permitted readers to take extracts from any manuscript states “that either of the Officers of the department to which such ... Manuscript belongs, be at liberty to do it for them, upon such Terms, as shall be agreed on between them.” We know from his letter to Bülow, cited above, that on 3 April 1787 Thorkelin was in the process of engaging a member of the staff to make transcripts for him. We also know that he left for Scotland early in June and did not return to the Museum until January 1788. He was apparently arranging for a transcript of Beowulf to be made while he was “out of the house.” It seems no coincidence, then, that at a General Meeting of the Trustees held on 24 February 1787 it was ordered “That Mr Ayscough [of the Department of Manuscripts] be at liberty to make extracts from printed books or Manuscripts for persons out of the house who may apply for them on being permitted so to do by the Officers of the Departments to which the said books or MSS belong.”⁶⁰

While these minutes might seem to identify Samuel Ayscough as Thorkelin's scribe, Ayscough's inartistic hand and exorbitant charges (“unheard of prices,” to judge by Thorkelin's letter) effectively eliminate him. The scribe who so skillfully copied the capital letters from the Beowulf manuscript and who learned to copy insular script so impressively cannot have been Ayscough, who has left behind many examples of his distinct lack of artistry. The draft of his catalogue of Additional Manuscripts provides on its title page numerous capital letters in a roughly contemporary example.⁶¹ What remains interesting about this catalogue, however, is that it was prepared on exactly the same kind of paper as that used for Thorkelin A. The watermark shows a crowned fleur-de-lis and the manufacturer, J. Whatman. This paper must have been Museum issue, confirming that Thorkelin's scribe, though not Ayscough, was a member of the staff, who presumably was allowed to make a copy of Beowulf without using the Reading Room Register. On the basis of handwriting, we can eliminate most of the known officers of the Department of Manuscripts, including Ayscough, who like most of his colleagues wrote in a conventional eighteenth-century style, similar to the handwriting Thorkelin uses in B. The A transcript is written in a small, neat, remarkably vertical script, and it is reasonable to suppose that the scribe's ordinary handwriting was also small, neat, and vertical, an unconventional style for the period.⁶²

Throughout the two large volumes of the Museum's “Original Letters and Papers” for the years 1743-1809, only three letters (nos. 583, 585, and 588, all dating from 1780-81) exhibit a handwriting that might be said to correspond to the “small, neat, vertical” script of Thorkelin A. All three letters are signed by a Mr. Matthews, who persistently, and ultimately successfully, petitioned the Trustees for a raise in salary.

[Fig. 2: Matthews's Letter No. 583, 1781]

His third letter, dated January 1781, shows a watermark of a crowned fleur-de-lis, the same paper used for Thorkelin A. The Trustees' Minutes identify him as “Mr James Matthews of Wardour Street” when he was appointed Accomptant to the Trust on 20 December 1776.⁶³

[Fig. 3: Matthews's Letter No. 585, 1780]

Though nothing short of a signed transcript of another Old English manuscript could prove that James Matthews was Thorkelin A, there is reason to suspect that indeed he was. Matthews was apparently both an artist and a skilled copyist. The Trustees' Minutes reveal that he was first employed by the Museum in 1771 because he had distinguished himself as a scribe: “Dr Watson and Mr Harper having mentioned Mr Matthews, a Person who has been employed under Mr Dalton in making a Catalogue of the Library at the Queen's Palace; as one who was well qualified, and willing to assist in making the Catalogue of Printed Books. Ordered. That Mr Harper be empowered to employ him...”⁶⁴ Matthews worked on this Catalogue until 1779, when Harper reported at a meeting of the Standing Committee that his services were no longer needed. Matthews had once again made a good impression, however, for at the same meeting Joseph Planta asked for Matthews' help for a proposed Catalogue of Medals.⁶⁵

Planta's request suggests that Matthews was an artist, needed for illustrations of the medals. Though unfortunately he declined Planta's request, the inference that Matthews was an artist is reinforced by a gift he made to the Museum on 9 June 1780. The minutes of the Standing Committee report that “Mr Matthews presented an Etching and explication of a bronze Statue of Hercules, belonging to him, for which thanks were ordered.”⁶⁶ Since part of the business of this meeting was to consider Matthews' petition for a raise in salary, it is highly unlikely that Matthews hired an artist to make the etching. Rather, since he himself owned the statue, we may assume that Matthews himself made the etching, probably as an inducement for the Trustees to increase his salary. The issue of his salary was postponed, for the second time, to a later meeting.

It seems plausible that Planta, having specifically asked for his services for a Catalogue of Medals, would think of Matthews again when Thorkelin needed a talented scribe to copy a manuscript for him while he was in Scotland. The letter to Bülow implies that Thorkelin had already approached Ayscough about the project, that Ayscough wanted too much money for his work, and that Thorkelin was in the process of making other, less expensive arrangements. The circumstances may even explain why there is no record of Thorkelin A in the Reading Room Register. In order to keep peace in his department, Planta may have allowed Matthews, who had not been specifically authorized by the Trustees to make transcripts, to copy Beowulf in a private room, where he would not be seen by Ayscough.

If Matthews was indeed Thorkelin's copyist, it helps explain some odd features of Thorkelin A. An officer in one of the departments of the British Museum might be expected to know about insular script, or to ask questions of his colleagues if he did not. This consideration alone tends to eliminate Ayscough as the scribe. But the Accomptant to the Trust, while he may well have been more artistic than officers of the departments, would probably not be familiar with insular script, especially if much of his previous work had been with printed books, not manuscripts. More important, if Matthews was the A scribe it explains why he stopped proofreading his copy against the manuscript after a few hundred lines. The Trustees' Minutes reveal that Matthews died sometime before the meeting of the Standing Committee on 1 December 1787, when the executrix of “the late Mr Matthews” is mentioned in connection with “Arrears due to him as Accomptant to the Trust.”⁶⁷

If, as the evidence suggests, Thorkelin had a full transcript of Beowulf ready for him when he returned to the British Museum in 1788, there would have been little point in his sitting down to make another copy. It is more likely that he continued his investigations in the British Museum while perhaps studying Thorkelin A at home in his spare time. The Reading Room Register shows clearly that he was interested in numerous other manuscripts in 1788. Working with the beautiful Harleian MS 2278 doubtless whetted Thorkelin's interest in the Harleian collection. On 25 June 1788, he borrowed a third version of the Anglo-Saxon Chronicle in Cotton Tiberius B. i (a late addition to his Cotton Notebook, with all 6 chronicle poems). But for the next month and a half, from 26 June to 8 August, he immersed himself in the Harleian collection, which only contained a few Anglo-Saxon pieces for him to study.⁶⁸ Thorkelin moved at such a pace through the Harleian collection that his name is the only one to appear for long stretches in the Register.

By 19 August he was ready to begin a similarly relentless investigation of the Cotton collection, this time following the order of entries in his Cotton Notebook, but also adding relevant items as he worked. A systematic study of Old English texts, then, continued to be deferred, though every now and then some Old English texts came to hand. Thorkelin began his Cotton searches by borrowing some manuscripts he had studied before: Augustus ii (charters). Tiberius B. iv (the Anglo-Saxon Chronicle) and C. xiii (Gesta regum Saxonum), as well as two Harleian manuscripts, 283 and 353. He also looked at Tiberius E. i, v, and vii, all “burnt to a crust” according to the Cotton catalogue, and made a transcript from Tiberius E. iv. The Cotton catalogue says “This once valuable and elegant MS. on vellum, consisting of historical tracts, the chief of which is ‘Beda de temporibus,’ is so much damaged as to be useless.”⁶⁹

On the same day, he signed out chronicles from Tiberius E. viii and from ten Caligula manuscripts,⁷⁰ including Caligula A. vii (containing, among other things, the Old Saxon Heliand), x, and xiv, and returned them on 2 September. He studied Caligula B. i and ii on 28 August and again on 1 September, along with Caligula B. iii-x. The next day he consulted Caligula C. i-viii, as well as Harleian MSS 4376 and 4430. From 3 to 8 September he signed out eight Caligula manuscripts and Julius D. x. From 10 September to 15 October he used eight more Caligula manuscripts and Otho C. xiii.⁷¹ The final entry of Thorkelin's name in Add. MS 46513 of the Reading Room Register (recording from 16 September 1784 to 24 October 1788) shows that on 14-15 October he used seven Harleian manuscripts, none of which he had consulted before.⁷²

The first entry of his name in Add. MS 46514 of the Register (recording from 27 October 1788 to 8 August 1792) shows that on Christmas Eve 1788 he used Harleian 7004, a manuscript he had first used on 8 August. He began the New Year by covering old ground, signing out Caligula A. vii (the Heliand manuscript he had used in August), Caligula A. ix, Harleian MSS 2278 and 7004, the Roe Letter-Books in Add. MSS 4168-72, and Add. MSS 4795-97 from the Milles collection, all from 5 January until 9 April. From 24 February to 9 April he examined again the three versions of the Anglo-Saxon Chronicle in Tiberius B. i, Tiberius B. iv, and Domitian A. viii. On 12 March he used a half-dozen Harleian manuscripts, and on 20 April he reviewed, among other items, some of the Roe Letter-Books⁷³ and either the Anglo-Saxon Chronicle or Cnut's Laws from Domitian A. viii.

Thorkelin then resumed his systematic use of the Cotton Notebook, picking up where he had left off, at the end of the Caligula manuscripts and the beginning of Claudius. He was especially industrious in one week at the end of May: from the 21st to the 26th he signed out eight Claudius manuscripts, including A. ii, B. vi, and C. ix.⁷⁴ From the 22nd to the 28th he used six Nero manuscripts, including Nero A. i (containing Wulfstan's Sermo lupi ad Anglos), ii, and iv; from the 25th to the 28th he continued his investigation of Nero manuscripts by borrowing eight more; and on the 26th he studied five Galba manuscripts.⁷⁵ He doubtless learned at this point that the Otho manuscripts he was interested in had been destroyed in the Cottonian Library fire and that Vitellius C. ix contained the two versions of the Life of Griffin. The note he attached to his Cotton Notebook thus dovetails with the undoubted course of his research at this time. To be sure, on 28 May he reached the Vitellius entries in his notebook. On that day he signed out four Vitellius manuscripts, A. xiii, xv (Beowulf), xvii, and xx, returned them all, and then signed out four more, Vitellius C. viii. ix, D. vii, and E.v.

The entries in the Register show, then, that when Thorkelin finally looked at the Beowulf manuscript again, after his discovery of it in 1786, it was not to make a transcript, but only to review its contents along with those of other Vitellius manuscripts. In fact, another of his transcripts, the Life of Griffin from Vitellius C. ix, was evidently the first tangible product of his work on 28 May. It was probably then that he attached to the Vitellius entries the note identifying C. ix as “Vita Griffini quædam & Danis & Norwegis,” which the notebook entry identified only as “Chartæ Regum.” In any event the Register records that Thorkelin borrowed this manuscript later, from 1 to 21 September 1789, when he doubtless made his transcript. Once again, it is noteworthy that we find no similar evidence in the Register indicating when he might have acquired his Beowulf transcript.

Thorkelin was in Ireland in the summer of 1789, and the only entries of his name in the Register are on 15 June, when he borrowed Vespasian A. iv and viii, and 17 June, when he borrowed Vespasian A. xv, xviii, and xx, all of which were returned on 3 August, before he returned from his trip. Thorkelin forgot to return these manuscripts when he left for Ireland, and they remained at his assigned seat in the Reading Room until a member of the staff returned them at the beginning of August, when the entire Museum customarily went through its annual cleaning. After he left Ireland in the middle of August, Thorkelin went on to the Isle of Man until the 20th.⁷⁶

As we have seen, he probably made his copies of the Vita Griffini in the first three weeks of September. From the 16th to the 18th he studied four manuscripts relating to early Irish history in the Milles collection.⁷⁷ In the last week of September, he concentrated on Vitellius and Vespasian manuscripts again. On the 21st and 22nd he borrowed five Vitellius manuscripts, and when he returned these books on the 22nd, he took out five Vespasian manuscripts, including A. i (the Vespasian Psalter), vi, xv, and xviii, all of which he kept for three months, returning them on Christmas Eve 1789.⁷⁸ Later on 22 September, he took out seven more Vespasian manuscripts, which he returned on the 24th; later still in the day he took out five more Vespasian manuscripts, returning them the next day.⁷⁹

For the rest of 1789, Thorkelin was mainly studying Cotton manuscripts. He finished reading the Vespasian manuscripts between 23-25 September.⁸⁰ In the process he began checking out Titus manuscripts in three installments between 24 September and 8 October.⁸¹ On the 8th he studied Domitian A. i, ii, iii, and iv, returned them, and yet again studied either the Anglo-Saxon Chronicle or the Leges Canuti in Domitian A. viii. as well as five other Domitian manuscripts,⁸² not returning any of them until Christmas Eve. Before then his name appears in the Reading Room Register only two more times in 1789, when he borrowed Domitian A. vii and Harleian MSS 307 (Leges Canuti), 785 (Danelage), and 2188 on 24 November. and Claudius D. ii on 2 December, all of which he also returned on Christmas Eve.

His name does not again appear in the Register in 1790 until the middle of February and his name disappears from the Register after April. During the short period in 1790 that his name still appears in the Register, Thorkelin was not studying Old English manuscripts.⁸³ The last entry of his name is on 6 April 1790, when he used several manuscripts from the Harleian and Birch collections, some of which he kept out, returning them at the end of April. We know, however, that he worked in the Museum after April and that he did not return to Denmark until April or May 1791.⁸⁴

The reason the Register falls silent at this point is best explained by assuming that, for some reason, Thorkelin was no longer required to use the Register after April 1790.⁸⁵ If so, he may well have acquired his copy of Beowulf after April 1790, in the year remaining before his return to Copenhagen.

To be sure, a 1787 date for Thorkelin B should be doubted on general grounds, now that we have the negative evidence from the Reading Room Register and know something about Thorkelin's wide-ranging research projects. It is highly doubtful that Thorkelin would have felt free to make a superfluous transcript of 70 manuscript folios, 140 pages, as early as 1787, particularly since he was in Scotland for the last half of the year. After all, he was obliged by his grant to collect widely dispersed historical documents and, as far as he knew before leaving for Scotland, he had only until June 1788 to carry out his ambitious project. Indeed, toward the end of his original grant he was forced to request an extension, because he had so much material left to copy in the Dublin libraries, in the Tower of London, and in the British Museum. Before he left for Scotland, he could not be certain that his grant would be automatically renewed, and he would not have wanted to return to Denmark in June 1788 with little more to show for his efforts than two copies of the same unknown Anglo-Saxon poem. We may well he glad to have two transcripts of the poem, but eighteenth-century Danes would have been perfectly satisfied with one.

We also know now, from his working notebook for the Cotton collection, that Thorkelin did not plan to copy Beowulf at all when he left Denmark, and that even when he learned about the poem and so added Cotton Vitellius A. xv to the notebook, his knowledge of the poem was still restricted to Wanley's catalogue description. In 1787, not yet fully knowing its contents, but obviously recognizing its potential value, he presumably ordered a transcript by a professional copyist, so that he could continue his search for charters, annals and other historical documents, of unquestioned value to his benefactors, in various collections scattered throughout Great Britain, Ireland, and the neighboring isles. The letter he wrote to his patron about his hopes of acquiring transcripts despite their high cost suggests that he commissioned Thorkelin A sometime after 3 April 1787, the date of the letter. After he returned from Scotland at the end of the year he was then able to study the poem at his leisure in the copy he obtained. We may be sure that Thorkelin did not personally supervise the first transcript, for his copyist had to learn by himself how to distinguish such insular letterforms as thorn, wynn, ash, and eth, which Thorkelin, as an Icelander and a student of Old English, already knew well.

The paper he used for his transcript of Beowulf strongly supports a post-1787 date for Thorkelin B. He copied the poem on paper typically gathered in bifolia, measuring approximately 18 x 23 centimeters, and bearing the watermark of a crowned posthorn. Like the make with the crowned fleur-de-lis used for Thorkelin A, the one with a crowned posthorn is ubiquitous in Museum documents. In relevant documents it was used, for example, in the Trustees' Index to Minutes, Vol. 1, 1754 to 1817 (where the crowned fleur-de-lis also appears on the flyleaf), in General Meeting, Vol. IV (1776-1806), and in Committee, Vol. VII (1779-1789). In R. 6431.6.E.3, some of the Thorkelin B paper, in bifolia, was used by Thorkelin for notes about his travels. One of these bifolia is dated by him “August 8th 1789?”. This shaky clue, a questionable date challenging another questionable date, does not, of course, prove that Thorkelin copied Beowulf in 1789, but it does provide a far more acceptable date than 1787. Beginning in 1788, the same paper was regularly used by Thorkelin in his letters to Bülow.

The exigencies of time alone strongly suggest that Thorkelin B was made in the later years of the research trip. We now know that Thorkelin did not learn about the poem until the end of 1786; that he was not studying Old English manuscripts in 1787; that he stopped working at the British Museum less than halfway through 1787; and that he did not again use any manuscripts in the British Museum until 7 January 1788. The earliest we can date the first transcript, on the basis of all available evidence, is at the end of 1786, but most likely after Thorkelin left for Scotland in June 1787; the earliest we can reasonably date Thorkelin's own copy is 1789.

Whenever Thorkelin A was made, it took considerable time to transcribe, with the hired scribe painstakingly mimicking the unfamiliar insular script, and at least beginning to proofread his work in collation with the manuscript. Appreciably more time must have elapsed before Thorkelin, having studied the first transcript, decided that he wanted to make another full transcript himself. His preliminary work with A is reflected in both transcripts. In A, Thorkelin wrote an interlinear translation, which he later erased, for most of the first two fitts,⁸⁶ made frequent conjectural emendations, and regularly underlined proper names. His familiarity with the proper names in Beowulf when he made his own copy can be seen in the fact that he regularly capitalized them as he copied. There is no question that Thorkelin B is not a mechanical copy of Beowulf. Thorkelin thought he understood what he was copying, for he had already studied the poem in Thorkelin A.

There is no evidence whatever that Thorkelin had so thoroughly immersed himself in the study of Beowulf as early as 1787. The Trustees' Minutes, which are demonstrably incomplete for the years 1786-1791,⁸⁷ unfortunately do not include a record of the Beowulf transcripts. But this is not surprising because the regulations allowed Thorkelin to make a transcript without seeking special permission, and Thorkelin's scribe, as a member of the staff, would likewise be allowed to make a transcript at the discretion of the Keeper of Manuscripts. Nonetheless, the Minutes do support the negative evidence from the Register that 1787 is too early for Thorkelin B, by showing that Thorkelin was a year off in his dating of another one of his transcripts, Lydgate's Life of St. Edmund. The Register, on the other hand, shows that Thorkelin began to study Cotton manuscripts in the last three months of 1786, but not in the order of his notebook, and that he was at once sidetracked to Icelandic manuscripts. He did not study Old English at all in 1787, and the only Cotton manuscripts he used were Nero manuscripts relating to late Scandinavian affairs. All of the records, including the Register, show that his work at the British Museum began in earnest in 1788, when he started studying Old English rather systematically and also began using his Cotton Notebook in an orderly way.

In addition to the Register, there are numerous references to him, from 1788 on, from other sources, which may have some bearing on why Thorkelin's name no longer appears in the Register after April 1790. The first reference occurs in the British Museum's “Original Letters and Papers, Vol. II (1785-1809).” In January 1788 Planta applied to the Trustees for a leave of absence, and was asked to report to a Standing Committee of the Trustees how his duties might be provided for should his request be granted. In a letter dated 18 January 1788, Planta recommended, among other things, “that a Deputy be appointed [in the Department of Manuscripts] on the footing of an Assistant, and to act as such under Mr. Southgate.” Planta goes on to say,

The choice of such a Deputy I would gladly leave to the recommendation of the Committee or to the absolute nomination of the electing Trustees; but as it may be required of me to propose a proper person, I think myself bound to suggest that I know of no one better qualified than Mr. Thorkelin a Gentleman well known to many very distinguished Characters in this Country, and who in the annexed letter declares his readiness to accept of the offer.

Planta even goes on to offer his own apartment in the Museum to accommodate his replacement. To merit such a warm and confident recommendation, Thorkelin must have been well known to Planta. Harvey Wood has shown that Thorkelin had powerful friends associated with the Museum, not the least of whom was Sir Joseph Banks, one of the Trustees. When John Pinkerton heard rumors of Planta's possible resignation and applied to Banks for his support, Banks replied in a letter dated 11 February 1788 that he was “positively engaged to Mr. Thorkelin, should Mr. Planta resign.”⁸⁸

Thorkelin's letter to Planta also survives in the “Original Letters and Papers” for 18 January 1788. His response is even more surprising than Planta's recommendation:

In answer to the Proposal you have been pleased to make me concerning the deputation in the British Museum in case you should obtain the leave of Absence, you have applied for; I herewith acquaint you, that I am very willing to accept of it on the terms, you mentioned viz: Being on the footing of a Assistant: that I am perfectly at Liberty to do so for the Space of two years to come; my leave of Absence from the Court of Copenhagen extending to the end of the Year 1789: & that should I be approved of, I shall at the expiration of that term, apply for a farther Leave of Absence, or if I have any prospect of a permanency in this country, resign my employment in Denmark, my inclination, as well as the object of my favourite studies rendering an establishment in this Kingdom highly desirable.”⁸⁹

It comes as a shock that Thorkelin would he willing to give up his future as the Danish National Archivist for a minor position in the British Museum. Perhaps Thorkelin's enthusiastic response should not be taken at face value, though, but rather seen as an effort to establish the most advantageous way to carry on his research. A post in the British Museum would give him direct access to the best resources in Great Britain, easy access to the Tower of London, and an open door to cathedral and university archives throughout Great Britain and Ireland. As we have seen, his candidacy for a post in the British Museum was not taken lightly. How seriously Thorkelin himself regarded “a permanency” in England is a matter of speculation, though it is worth noting that Bülow worried about the possibility as early as 8 March 1787.⁹⁰ It seems safe to conclude that Thorkelin was quite anxious to secure a position in the British Museum, because he wanted to make many transcripts there, and that he was willing to extend his stay abroad indefinitely. But it also seems likely that he always intended to return to Denmark if the post of National Archivist became vacant.

In any case, there are indications that Thorkelin was employed by the British Museum in some capacity between the years 1788 and 1791. His employment would explain why the Reading Room Register fails to record his use of the Beowulf manuscript while he was making his transcript. If Thorkelin became a member of the Museum staff during these years, it would help explain why the Register no longer records Thorkelin's use of any manuscripts whatever after April 1790. The staff lists of the British Museum do not go back beyond the mid-nineteenth century, but three entries from the “Ledger Book of Accompts of Major Edwards's Fund” have some possible bearing. Major Edwards's Fund paid for purchasing additions to the Cotton collection, and the ledger book reveals (fols, 6, 7, and 8) that Thorkelin was paid from the fund on three occasions:

19 Mar. 1788. “paid Thorkelyn for Sundry” £9.10s.
28 Mar. 1788. “paid Thorkelin for Manuscripts” £16.10s.
28 Mar. 1789. “paid Thorkelyn for Icelandic Manuscripts” £16.10s.

Ms K. Janet Wallace, Archivist in the British Museum, infers that because “he is referred to informally in the account book as ‘Thorkelyn’ instead of ‘Mr Thorkelyn’... he may have been in the Museum's employ at that time."⁹¹ The entries from the ledger book further suggest that he may have been hired to purchase manuscripts for the Museum on his wide-ranging trips through Great Britanin and Ireland in search of Danish antiquities.⁹² In this way Thorkelin may have reconciled service to both Danish and British archives.

But a more prestigious post at the British Museum was offered to him in May 1790, which most likely explains why his name no longer appears in the Reading Room Register after April. On 9 May Dr Charles Woide, assistant librarian in the Department of Printed Books, died unexpectedly, leaving a position that needed to be filled. The day after Woide's death, Planta wrote to Thorkelin, who was not in London at the time, informing him of the opening and urging him to apply for it at once. Thorkelin not only applied for the position but was promptly appointed. As he says in his autobiographical notes for the year 1790 (R. 6431. 5a), “The Archbishop of Canterbury [who officially chaired the Trustees' General Meetings] offered me a librarianship in the British Museum, which combined with another post will give me a yearly income of 300 pounds, free rooms, lighting, and a fire.”⁹³

The speed with which this appointment was made appears from the fact that Thorkelin wrote to his patron Bülow on 18 May, about a week after Woide's death, to say that if he took the job he would only hold it until his position at the Gehejmearkivet opened. However, only three days later he wrote again to Bülow to say that he had decided after all not to take the job, and that he was returning to Copenhagen at once.⁹⁴ Yet on the next day, 22 May, he copied the letter from Copenhagen to the Danish ambassador in London, giving him an additional year's leave of absence, and shortly thereafter wrote to Bülow to thank him for interceding on his behalf.⁹⁵ At any rate, by May 1790 Thorkelin almost certainly enjoyed privileged access to the collections in the British Museum. Under these special conditions, it is less cause for wonder that records of his use of manuscripts in the Reading Room Register cease after April 1790.

All combined, the evidence on Thorkelin's trip to Great Britain and Ireland between 1786 and 1791 strongly suggests that Thorkelin copied the Beowulf manuscript sometime between 1789 and 1791, perhaps in the summer or fall of 1789, or perhaps after April 1790, in view of the negative evidence of the Reading Room Register. Although he certainly did basic research at the British Museum before 1788, the extant records, notably the Register, suggest that he had not yet made a transcript of Beowulf. Since we have ample evidence that Thorkelin was often unreliable about dates, there is no need to seek out a special reason for his misdating his Beowulf transcript. However, if he had not made in 1786 or 1787 any substantial transcripts relating to the Danish presence in Great Britain during the Middle Ages, he may have been tempted to antedate his choicest transcripts in the period of his original two-year grant. This harmless motive would at least explain why he later added the date 1787 to his title-pages of Lydgate's poem and of the Beowulf transcripts, even though he never dated any of the other extant transcripts.

In his letter to Planta in January 1788, Thorkelin had claimed that he would resign his employment in Denmark if he could obtain a permanent position at the British Museum. We now know that he declined a permanent post there when faced with the actual offer. Yet perhaps his decision to edit Beowulf and thus to make a superfluous copy of the manuscript reflects his late dilemma over whether to stay in England or return to Denmark. The problem was seemingly resolved for him in 1791, when the old Danish National Archivist died. In any event, Thorkelin then returned to Denmark with two copies of Beowulf that replace the manuscript itself for well over 1900 letters now lost by fire-damage.

Notes to Part One

1. Rigsarkivet, “Privatarkiv Nr. 6431, G.J. Thorkelin, 1752-1829.” This archive includes four volumes of letters to him (6431. 1-4), and until 1980 a huge bundle of “Personlige Papirer” (6431.5). See my article, “Thorkelin's Trip to Great Britain and Ireland, 1786-1791,” The Library 5 (March 1983), 1-21. In 1980 the personal papers were newly placed in folders and boxed, so that what used to be only 6431.5 is now 6431.5 and 6431.6. Most of the material relating to Thorkelin's trip to England is now in 6431.6.E.3, a box labeled “Vedr. G.J. Thorkelins rejse til Storbritannien og Irland 1785-91 / Heri ansøgninger, rejsepas, optegnelser m.m.”

2. Wanley says, “In hoc libro, qui Poeseos Anglo-Saxonicæ egregium est exemplum, descripta videntur bella quæ Beowulfus, quidam Danus, ex Regio Scyldingorum stirpe Ortus, gessit contra Sueciæ Regulos.” Antiquæ Literaturæ Septemtrionalis liber alter ...Catalogus historico-criticus (1705), reproduced in facsimile as no. 248 in the series English Linguistics: 1500-1800 (Menston, 1970), p. 219.

3. Ex Bibliotheca Cottoniana Musæi Britannici edidit versione lat. et indicibus auxit Grim. Johnson Thorkelin. Dr. J.V. (Copenhagen, 1815).

4. There is a list of his positions and attainments on the title-page of his Fragments of English and Irish History in the Ninth and Tenth Century. In Two Parts. Translated from the Original Icelandic, and Illustrated with some Notes by Grimr Johnson Thorkelin, LL.D. (London, 1788).

5. “for at reisse igiennem Storbritanien, Irland og Øerne i 2de Aar for at samle og optegne alle der forefindende Danske og Norske Monumenter, Handlinger og Documenter...imod at han ved sin Hiemkomst afleverer til Vores Geheime Archiv og det store Bibliotheque alle de Samlinger, han i saa Maade kan tilveyebringe.” See “7. Decbr. 1785. Rejseunderstøttelse til Prof. G.J. Thorkelin,” Fonden ad Usus Publicos: Aktmæssige Bidrag til Belysning af dens Virksomhed, udgivne af Rigsarkivet (1765-1800), 3 Vols. (Copenhagen, 1897-1947), I, p. 83.

6. Fragments, p. x, note. The direct descendants of this Commission are the Arnamagnæan Institute at the University of Copenhagen and the Stofnun Árna Magnússonar in Reykjavík.

7. Thorkelin defines his title in this way in his anonymous Sketch of the Character of His Royal Highness the Prince of Denmark, to which is added, A Short Review of the Present State of Literature and the Polite Arts in that Country (London, 1791), pp. 57-61.

8. “Da vores Professor Thorkelin saa længe har tient i Vores Geheime Archive, og tillige arbeidet ved Udgaven af Snorre Sturleson og andre Værker, og har baade den Alder, og den Kundskab, og det gode Vidnesbyrd, at Vi med föie Kand betroe ham, naar Geheime Archivarii Tienesten bliver ledig, dense Post: Saa give Vi til bemelte Vores Professor Thorkelin herved Vores skriftlige Löfte paa Geheime Archivarii Embedet, og den dertil lagte Gage, saa snart samme ved nu værende geheime Archivarii Död vorder ledig.” Rigsarkivet 6431.6.E., number 42 (formerly 23).

9. Dansk Biografisk Leksikon, pp, 313-14.

10. “Langfeðgatal fra Noe til varra Konunga,” Scriptores Rerum Danicarum Medii Ævi, partim hactenus inediti, partim emendatius editi, qvos Collegit, Adornavit, et Publici Juris fecit Jacobus Langebek, 9 Vols. (Copenhagen, 1772-1878), I, 9 note R; see p. 44 note E, and p. 2 Table I.

11. Rigsarkivet 6431.6.E., now numbered 49 and 50, after several changes in the numeration. The numeration was not changed in 1980, however, and it is still not always possible to cite numbers for particular items.

12. See Diplomatarium Arna-Magnæanum (exhibens Monumenta Diplomatica ex Bibliotheca Legati Arna-Magnæani) 2 Vols. (Copenhagen, 1786), which he published shortly before leaving for England.

13. “Paa Kongelig Bekostning reisde...i nogle Aar”; “At mig maatte allernaadigst vorde tilladt og befalet at giøre en Reise i tvende Aar giennem England, Scotland, Irland og derhenhørende Øer for at optegne og samle alt, hvad som findes herhørende til Forfædrenes og Handlinger, det være i Diplomatiqven, Oldsager og Historie. Hvilken Samling ved Reisens Slutning skulle nedlegges i Deres Majestæts Geheime Archive og det store Bibtioteqve.” The full proposal is printed in Fonden ad Usos Publicos, I, pp. 83-85.

14. See pp. 124-26 and 145-46.

15. “hans Studerings Plan meget var bleven forrykket ved en de tvende Parliaments Raader Grenville og Mackenzie...hvorved han med Uvished blev opholdt i 3 Maaneder, foraarsaget adskillige betydelige Udgifter og endelig i Junii Maaned 1788 blev underrettet om, at Reisen ikke gik for sig.” Fonden ad Usus Publicos, I, p. 125. His visa for this aborted trip survives as Rigsarkivet 6431.6.E. number 57: “Reise-Pas for Professor Grim Jonson Torkelin [sic] fra Engeland til Island.” It carries the Kings's seal and is dated 25 April 1788. Thorkelin knew that his grant had been extended (and his stipend increased) by 13 January 1788, when Johan Bülow, his patron, wrote to him saying “Det veed De vel allerede, at De har Tilladelse at blive Et Aar længere borte end efter Deres første Tilladelse, Men at Kongen har forøget deres Indtægter med 200r af Fondet ad pios usos(!), til Hielp til det Aars Udgifter, det tror jeg, at ingen før har sagd Dem.” See Alfred Glahn, “Mæcen og Klient: af en Brevveksling mellem to Bogvenner, 1785-1790,” Aarbog for Bogvenner (Copenhagen, 1925), pp. 66-67, hereafter cited as Glahn. This edition of the correspondence between patron and client is based in part on the letters from Bülow to Thorkelin in Rigsarkivet 6431.1-4, in part on the letters from Thorkelin to Bülow in the “Sorø Akademis håndskriftsamling, Johan Bülow (d. 1828), 86.A.1-120,” herafter cited as Sorø.

16. “uagtet al hans anvendte Fliid,” Fonden ad Usus Publicos, I, p. 125.

17. Rigsarkivet 6431.6.E. number 60 (formerly 43).

18. Letters to him from this period that survive in the four volumes of letters in the Rigsarkivet (6431.1-4) and the three volumes in the library of Edinburgh University (La. III. 379) show that his permanent lodgings were in London, at the home of a Mrs. Woods, on “Brownlow Street No 5, near Drury Lane.” There is an excellent study of the letters at Edinburgh in E.H. Harvey Wood, “Letters to an Antiquary: The Literary Correspondence of G.J. Thorkelin, 1752-1829” (unpublished Edinburgh University Doctoral Thesis, 1972), hereafter cited as Harvey Wood. So far, Glahn is the only published work on the letters from the Rigsarkivet and Sorø,

19. Rigsarkivet 6431.5.C.2, “Diplomer fra Universiteter, lærde skoler m.m. 1780-1825.” One of these is his LL.D., the honorary degree of Doctor juris utriusque from the University of St. Andrews.

20. On the eleventh leaf of the notebook titled “Julius 1789?” in Rigsarkivet 6431.6.E.3.

21. BL. Add. MSS 46513 (“Reading Room Register of MSS., Sep 1784 to Oct. 24. 1788”) and 46514 (“Oct. 27. 1788 to Aug. 8. 1792”), hereafter cited as Reading Room Register.

22. Though Thorkelin did not leave Denmark until the middle of July 1786, he always dates the trip from 1785, when the research grant was made. A brief diary of his voyage in 6431.6.E. shows that he set sail from Elsinore on 20 July, and after a rough voyage arrived at London on 3 August.

23. Hereafter cited as Cotton Notebook. The other two notebooks were used in Ireland, though one was prepared for use in England, as the headings “Ex Rogeri Dodesworthi,” “Ashmole's Collection,” “ In Biblioth. Joh. Mori Episcopi Norwic,” and “Lambeth Library” (fols. 2-4) all show. The second notebook is the one titled “Julius 1789?” (see note 20).

24. (Oxford, 1696), p. 83. Except where Thorkelin specifically says otherwise, the page references in the Cotton Notebook refer to this catalogue. The copy now in the Kongelige Bibliotek was evidently used by Thorkelin to prepare his list, for many of the items in the Notebook have been marked “NB” by him in this copy of Smith's catalogue.

25. For a study of the history and construction of BL MS Cotton Vitellius A. xv, see Kiernan, Beowulf and the Beowulf Manuscript (New Brunswick, 1981, rpt. 1984). [Now the Revised Edition (British Library Publications and University of Michigan Press, 1997).]

26. seu Annales Rerum in Anglia Præcipue Gestarum, A Christo nato ad Annum usque MCLIV. deducti... (Oxford, 1692).

27. Together with an Account of BOOKS burnt or damaged by a late fire (London, 1734).

28. The lists are in Vol. I, Manuscripta in Folio, D, 79, and Vol. III, Manuscripta in Quarto, I, 111-112. In the card catalogue the same items, except for Ny kgl. Saml. 511 and 511b, are listed under Normannerne. Ny kgl. Saml. 511 is listed in the card catalogue as Tractatus ad Hist. Autog. M. Brit. 4^o under “Thorkelin, Gr J.”.

29. henhørende til den Danske Historie Efter Allernaadigst Kongelig Befalin[g] udskrevne af Originalerne i d[e] Engelske Bibliothequer og indleverede til det Geheime Archive ved G.J. Thorkelin.

30. Including Cotton MSS Augustus ii, Claudius B. vi and C. ix, Cleopatra F. iv, Galba E. ii, Nero B. iii (very often), Tiberius A. vi and B. iv, Titus C. vii and viii, and Vespasian B. xxiv, C. xiv, and F. iii and iv, as well as Harleian MSS 57, 58, 61, 66, 283, 353, 358, 596, 3763, 6995, 7007, 7008, and 7012. There are also frequent references to “Rymeri Anecdota,” Thomas Rymer's Foedera, Conventiones, Literæ, et Cujuscunque Generis Acta Publica, Inter Reges Angliæ, et Alios quosvis Imperatores, Reges, Pontifices, Principes... ab Anno 1101, ad nostra usque Tempora... (London, 1704), the autographs of which were still kept in the Tower of London.

31. He is thought to have brought back original manuscripts, too, like Ny kgl. Saml. 167b, the Waldere fragment, and Ny kgl. Saml. 268b Fol., “Poemata heroica lingva hibernica - Codex membranaceo chartaceus,” ed. David Greene, Duanaire Mhéig Uidhir: The poembook of Cú Chonnacht Mág Uidhir, Lord of Fermanagh 1566-1589 (Dublin, 1972). His name appears on the inside front flyleaf of this manuscript.

32. See Glahn, pp. 62-64, and Sorø A.86.8. While complaining about how little time he had left to make transcripts, he told Bülow on 3 April 1787 that he had found and had begun to transcribe the five folio volumes of this Journal.

33. Thorkelin drew salaries from both the Geheimearkivet and the “Arna Magnæiske Stiftelse” while he was in Great Britain.

34. Advocates' MSS 21.2.1-13., 21.3.1-18., 21.4.1-18., 21.5.1-8., 21.6.1-9., 21.7.1-19., 21.8.1-16.

35. Including some excerpts from the Old English gloss to Latin hymns and canticles in Cotton Julius A. vi, and from the Latin life of St. Oswin and the Old English life of St. Fursey in Cotton Julius A. x. See items 160 and 161 in Neil Ker, Catalogue of Manuscripts Containing Anglo-Saxon (Oxford, 1957), pp. 202-206.

36. Three of the four extant versions of the Battle of Brunanburh are preserved in Cotton MSS Tiberius A. vi, B. i, and B. iv. The Reading Room Register shows that Thorkelin used these manuscripts repeatedly in 1788 and 1789, yet the Cotton Notebook shows that he later deleted the entries, ostensibly because Gibson had already used them in his Chronicon Saxonicum (fol. 2).

37. For my identification and description of these bindings, I am indebted to Raemund Joergensen, the binding expert at the Kongelige Bibliotek.

38. The artist is identified in Thorkelin's letter requesting permission to copy the manuscript (British Museum “Original Letters and Papers, Vol. II, 1785-1809,” number 656. The central archives of the British Museum are in the process of being listed and given new reference numbers; however, it will remain possible to identify BM items cited here by their original volume titles).

39. See The Thorkelin Transcripts of Beowulf in Facsimile, pp. C and B respectively.

40. Harvey Wood argues (pp. 81-82) that the title page for A was originally the one for B, but A's page is too large (c. 19cm wide x 25cm long, counting the strip along the binding) to have been in B (c. 18cm x 23cm). The pages are not produced exactly to scale in the facsimile, but their comparative sizes can be seen on facing pages (A, p. 90, B, p. C).

41. The Dansk Biografisk Leksikon (pp. 610-611) mistakenly claims that Thorkelin received the degree in 1788. The document itself in Rigsarkivet 6431.5.C.2 gives the date “vicesimo sexto die mensis Octobris Anno Domino millesimo septingentesimo octogesimo septimo,” which archives at the University of St. Andrews confirm.

42. “Til Archiverne her er det uendelig vanskeligt at erlange adgang, og Udskrifter icke tilladte uden ved Betienterne, som derfor kræve uhørte Priser. et kort diplom koster kun en heel Gvinea. Uagtet alt dette har ieg dog det beste Haab, uagtet med icke smaa bekostning” (Glahn, p. 62).

43. “The First Reading Room Regulations” in Arundell Esdaile, The British Museum: A Short History and Survey (London, 1946), p. 329.

44. An entry for 30 August on the same page (fol. 29) states that “Mr Thorkelin desires to present to the Museum the following Books, of which he is the Editor, ...Diplomatarium Arna-Magnæanum & - 2 vols Qu[art]o - Dania et Norvegia in Sigillis. Sec. 13. He has also sent a List of several printed Books, relating to the Antiquities of Denmark, Norway & Iceland, which he intends to present to the Museum, if they are not there already.”

45. In a letter to Bülow on 26 August 1786, Thorkelin states that he arrived in London on 7 August. He implies that he was detained by customs officials until the 25th, though surely the detention refers only to his baggage (see Glahn, pp. 53-55, and Sorø A.86.4). Perhaps the letter reflects Thorkelin's embarrassment for not writing to his patron sooner.

46. Julius A. xi and C. vi. To show the range and schedule of Thorkelin's research, the following discussion incorporates in the text and notes a full record of his use of manuscripts based on the Reading Room Register.

47. Some of the charters from Augustus ii show up in the former Ny kgl. Saml. 495b Fol., now Register 122 in the Rigsarkivet.

48. And Thorkelin assured Bülow that “ieg lever daglig [at the Museum] fra 9 til 3 Eftermiddagen” in a letter of 3 November 1786 (Glahn, p. 57, Sorø A.86.5).

49. “Iblandt Manuscripterne ere icke faae henhørende til Dannemark; ieg haaber at finde noget værdigt, hvor i blandt et ubekiænt Anglosaxisk, indeholdende Samling af Sange om de Danske Kongers Bedrifter i det 3ie Aarhundrede, og hvor ved den gode Saxo forøges icke lidet" (my italics; see Glahn, p. 57, and Sorø 86.A.5).

50. BL Add. MSS 4877 (the Elder Edda), 4880 (an Icelandic-Latin lexicon), 4881 (Icelandic law codes), and 4895 (Icelandic prayers). Thorkelin found these manuscripts by studying Samuel Ayscough's catalogue of manuscripts acquired by the British Museum between 1756 and 1782. In the Register these “Additional MSS” appear with the rubric “Don[ation],” but their numbers still identify them.

51. BL Add. MSS 4871 and 4880.

52. Harleian MSS 6986, 6988, 1760, 4439, 7004, 7007.

53. Nero B. ii, vi, viii, ix, and x. The next day he used Sloane MSS 2876, 2877, and 4046.

54. “containing copies of his correspondence and papers during his negotiations as Ambassador-Extraordinary to Hamburg, and, afterwards, to Regensburg and Vienna, for the settlement of a general peace, with the plenipotentiaries of the Emperor, France, Sweden and Denmark; 1638-1642.” The British Library Catalogue of Additions to the Manuscripts, 1756-1782, Additional manuscripts 4101-5017 (London, 1977), p. 43

55. With the Chronicle poems, Battle of Brunanburh, Capture of the Five Boroughs, Coronation of Edgar, and Death of Edgar. Some short transcripts from this manuscript survive in Register 122. If Thorkelin transcribed the poems or other parts of the chronicle, his copies have not yet been found.

56. The Reading Room Register is a bit confusing at this point because of an error in the sequence of pages. A page evidently came loose and was reinserted in the wrong place, so that folios 176 and 177 in the Register are reversed

57. See Esdaile (p. 330).

58. BM “Original Letters and Papers, Vol. II (1785-1809),” number 656.

59. “Trustees' Minutes: Standing Committee 14 March 1788,” p. C 1982.

60. “Trustees' Minutes: General Meeting 24 February 1787,” p. GM 878.

61. BL MS 111f. (currently uncatalogued): “A CATALOGUE of all the Manuscripts Bequeathed by the Rev THOMAS BIRCH D.D. to be preserved in the BRITISH MUSEUM to which is added a Catalogue of all the other Manuscripts Bequeathed Presented or Purchased to the Present Time by the Rev. Samuel Ayscough 1781.”

62. Malone makes the comment that, “Like any beginner [A] made his characters large at first (that is, large for him), but as he became more proficient they grew smaller. His insular hand in its maturity is markedly small, agreeing in this respect, one may conjecture, with his ordinary hand” (p. 4).

63. “Trustees' Minutes: Standing Committee” 1539 (20 December 1776). “Trustees' Minutes: General Meeting” 761 (15 February 1777).

64. “Trustees' Minutes: Standing Committee” 1314 (13 December 1771).

65. “Trustees' Minutes: Standing Committee” 1681-82 (3 December 1779).

66. “Trustees' Minutes: Standing Committee” 1713 (9 June 1780).

67. “Trustees' Minutes: Standing Committee” 1971, (1 December 1787).

68. During this period he called for Harleian MSS 27, 40, 54, 55, 56, 61, 66, 84, 86; 105, 108, 112, 114; 200, 201, 202, 208, 247, 257, 258, 261, 270, 279, 283, 293, 294, 298; 310, 312, 315, 322, 353, 358, 362, 377, 382, 383, 392; 296, 391, 432, 438, 447, 464, 532, 588, 596; 624, 638, 651, 655, 661, 662, 667, 694, 695, 643, 646, 776, 787; 841, 980, 1178, 1217, 1244, 1275, 1300, 1320, 1355, 1389; 980, 1178, 1217, 1244, 1275, (cancelled and replaced by 1423, 1509, 1514-1519, 1526); 1760, 1761, 1782, 1802, 1808, 1813, 1867, 1901, 1924, 1965, 1967 (cancelled and replaced by 1582-1584, 1616, 1623, 1706, 1707, 1725, 1728, 1729, 1755); 2110, 2188, 2264, 2277, 2334, 2371, 2383, 2386, 2788, 2895, 3640, 3721; 3641, 3643, 3648, 3650, 3666, 3669, 3671, 3673, 3688, 3697, 3723, 3725, 3759, 3763, 3764, 3776, 4363, 4432-4434, 4569-4580, 4583, 4586, 4787-4790; 3846, 3891, 3895, 3908, 3911, 3960, 3970, 4080, 4163, 4242, 4344, 4439, 4456, 4461; 4471, 4488, 4515, 4535, 4536, 4542-4545, 4637, 4660, 4714, 4799, 5176, 5192; 4009, 4843, 4875, 4876, 4976, 4994, 5018, 5133; 6069, 6157, 6455, 6602, 6670, 6760, 6826; 6988, 6994, 6995, 7004, 7007, 7010-7012, 7015. 7016.

69. The comment that the manuscript “is so much damaged as to be useless” is not correct, and someone has crossed out the phrase in the copy in the BL Student's Room.

70. Caligula A, iii, and vii-xiv.

71. Caligula D. i, ii, iv-ix; and E. vi-xiii.

72. Harleian MSS 1691, 1968, 2144, 2876, 2877, 4277, 4300.

73. Harleian MSS 612, 1760, 6986-6988, 7502; and Add. MSS 4170-4172, 4188 (from the Birch Collection); Add. MS 4796 (Milles Collection); Cotton Caligula A. ix.

74. Claudius A. iii, vi, xiii, and xiv. B. iii and vi, and C. ix and x.

75. Nero A. i, ii, iv, vi, viii, and xiii; C. i, v, vii, ix, and x, and D. i, iii, and viii; Galba A. xviii, D. iii, vii, and x, and E. ii.

76. See Harvey Wood (p. 63).

77. Don[ation] 4783, 4784, 4787, 4789.

78. Vitellius E. xiv and xvii and F. vi, x, and xvi; Vespasian A. i, iv, vi, xv, and xviii. For the possibility that Thorkelin was copying Beowulf at this time, see my article “Madden, Thorkelin, and MS Vitellius/Vespasian A. xv,” forthcoming in The Library.

79. Vespasian A. xix, xxii, B. xi and xxiv; C. xiv, xv, and xvi; D. iv, vi, x, xii and xiii.

80. Vespasian D. xviii and xxi; E. ii-v and ix-xxvi; and F. i, iii, xiii, and xv.

81. Titus A. x, xiii, xiv, xix, and xxiv, from 24 September to 22 October; Titus B. i, ii, and vi, and C. vii and x from 24 September to 8 October; and Titus C. xviii and xix, and D. xxvi from October 6-8.

82. Domitian A. ix, x, xii, xiii and xviii.

83. On 15 February he used eight Cleopatra MSS, C. i-iii, vi, and x; D. ii, iii and vi; and Harleian 359 (Giraldus Cambrensis). On the following day he used Faustina A. vi and viii. From 23 February to 19 March he borrowed Cleopoatra F. iv, while from 25 February to 31 March he borrowed Royal MSS 13.B.i and ii. On the 3rd of March he consulted Tiberius A. ix, Sloane MSS 1898, 3103, and 3371, and Add. MS 4789 from the Milles collection, returned them, and then looked at Sloane MSS 260 and 2442, and Add. MSS 4193 and 44 from the Birch collection. On the 19th, the day he returned Cleopatra F. iv, he studied two more manuscripts from the Birch collection and two from the Harleian (Add. MSS 4115 and 4176; Harleian MSS 539 and 862). The remainder of the entries show him using Harleian, Sloane, and Additional manuscripts (Add. MS 4158 and Sloane MSS 2144, 2876, 3322 on March 22; Sloane MSS 2046, 2877, 4047, 4048 on March 23; Sloane MSS 4044, 4046, 4054, 4058 from 23-26 March; Add. MS 4159 from 24-26 March; Add. MSS 4559, 4564 from 26 March lo 31 April; Add. MSS 4300, 4310, 4318 on 31 March, and again on 6 April, when Harleian 7003 was entered as if it was a Birch manuscript, too; and Add. MSS 4277, 4293, and Harleian MSS 7011, 7012, 7016, from 6-31 [sic] April).

84. Thorkelin wrote to George Dempster from London on 14 April 1791 (Harvey Wood, p. 324), and to Bülow from Copenhagen on 23 May 1791 (Glahn, p. 87), so he returned home sometime in between.

85. This deduction is to some extent confirmed by the Register. Allhough he borrowed Cotton Faustina A. viii on 16 February, there is no record that Thorkelin ever returned the manuscript. The manuscript was returned of course, and Thorkelin may have returned it when records were no longer kept for him. It is possible, too, that a clerk simply failed to record the return of the manuscript.

86. It is plausible that Thorkelin based this first translation on the excerpts from Wanley's Catalogue, since it begins with fitt I, omitting (like Wanley) the part of the text that follows the genealogy.

87. Members of the staff took turns as secretary. It was Planta's turn when Thorkelin arrived in 1786, and he entered the general note in Add. MS 45,870 stating that “At a Committee Oct 27. 1786 - All the above leaves applied for were granted. J Planta Sec[retar]y” (fol. 30). Although readers were required to reapply for leave every six months, none of Thorkelin's renewals is specifically recorded. Nor is there any mention among the minutes of the “List of several printed Books” Thorkelin sent on 30 August 1786 “which he intends to present to the Museum, if they are not there already” (Add. MS 45,870 fol. 29). There is, however, a form letter dated 1 September 1788 from “[Paul] H. Maty, then secretary, thanking Thorkelin for gifts of books and adding the note, “Mr Maty has also the honour to inform Mr Thorkelin that the Trustees will with great pleasure receive the other books he proposes to present them with as it does not appear that any of them are in the house”(Rigsarkivet 6431.6.E, unnumbered).

88. See Pinkerton's Literary Correspondence, ed. Dawson Turner (London, 1830), Vol. I, p. 180; cited from Harvey Wood (pp. 67-68).

89. The letter illustrates Thorkelin's tendency to mistake or misrepresent dates. According to the Fonden ad Usus Publicos his grant did not extend to the end of 1789, but only to the end of June 1788; and even the first extension he was granted allowed him to stay only until the end of June 1789 (pp. 125-126). But Thorkelin apparently knew in late 1787 that his grant had been extended, as the letter from Bülow dated 13 January 1788 shows (see note 15 above and Glahn, pp. 66-67).

90. “Et Rygte har sagd, at De ville blive i Engelland” (Glahn. p. 64).

91. In a letter to me dated 28 August 1980. It is perhaps worth mentioning that as early as 5 January 1788 Danish visitors to the Museum were allowed to use the Reading Room on the strength of Thorkelin's recommendation. BL Add. MS 45,870, the document recording “Persons admitted to [the] Reading Room,” reveals that this function was usually restricted to members of the staff. Planta's letter recommending Thorkelin as his replacement (“on the footing of an assistant”) was written less than two weeks later.

92. Harvey Wood discovered that Thorkelin was in all probability commissioned by King George III to collect Scandinavian books for the King's Library (pp. 40-41).

93. “Tilböd Ærkebiskopen af Canterbury mig Bibliothecat i det Britiske Museum, som forennet med et andet Embede skulle give mig aarlig Inkamst af 300 pund, frie værelser lys og Ildebrand.” The Museum staff lived on the premises in the eighteenth century. The other post Thorkelin alludes to here could be a commission for book-collecting for the Museum.

94. See Glahn (pp. 76-77) and Harvey Wood (pp. 68-71).

95. “Da jeg havde giørt mig færdig at gaae om bord til fæderne Landet, i følge Deres allerhøystærede Skrivelse, og den besluttning Jeg havde taget i overeensstemmelse med mit Brev til Dem af 21 May næstafvigte, erholder Jeg den angenemme Efterrettning fra Hr Græve Wedel Jarlsberg, at Deres Majestæt Kongen, og D.K. Høyhed Chron Printzen, havde allernaadigst behaged at tillade mig, at blive endnu et Aar Længere i England; og følgeligen, at De paa den allerkraftigste Maade tage Deel i alle Ting som angaae mig” (Glahn, p. 79, undated, but “21 May næstafvigte” means that it was written in June).