This is an electronic version of an essay in Beatus vir: Studies in Anglo-Saxon and Old Norse Manuscripts in Memory of Phillip Pulsiano. eds. Kirsten Wolf and A.N. Doane. Tempe, AZ: Medieval & Renaissance Texts & Studies (MRTS), 2005, pp. 85-106.

Odd Couples in Ælfric’s Julian and Basilissa
in British Library Cotton MS Otho B. x*

Kevin Kiernan

The unusual marriage of Julian and Basilissa features two saints-to-be who vow on their wedding night in their wedding bed not to consummate their marriage.1 Their connubial plans extol chastity, Old English clænnysse, ensuring immediate female sanctity for Basilissa2 and a more perfect union for the couple in heaven by the end of the story. Julian superfluously earns his sainthood through more active, manly virtues, including a violent martyrdom by beheading. Notwithstanding its chaste setting, the fragmentary, appropriately headless, version of this story semi-saved in British Library Cotton MS Otho B. x has led to a productive and even reproductive marriage of text and technology. I use the term “technology” in a broad sense, alluding to Anglo-Saxon writing or copying, modern printing, medieval and modern textual encoding, early-modern and modern bookbinding and rebinding of burnt manuscript fragments in individual paper frames, as well as the more obvious modern technology of computers and digital cameras with all their coupled hardware and software. My goal, to illustrate how medieval and modern technology can unite to help editors deliver newly born Old English texts, may seem remote from the lives of saints. Yet in the end these burnt fragments of manuscripts can rejoice with the martyrs and say, We ferdon þurh fyr 7 wæter, 7 þu us læddest on celincge.3 

Cotton Otho B. x was grievously injured in the disastrous fire in 1731 that destroyed or seriously damaged nearly one quarter of the magnificent collection of over a thousand manuscripts amassed by Sir Robert Cotton following the dissolution of monasteries in the sixteenth century.4 Very little of the Passion of St. Julian and His Wife Basilissa survived. Walter Skeat, its modern editor, wrongly thought that only one folio made it through the fire. His edition is based on the complete manuscript of Ælfric’s Lives of Saints preserved in British Library Cotton MS Julius E. vii, but he collated this fragment with his edition (lines 27-91) and recorded nine variant readings from “Otho B. x. fol. 7 (as now numbered)” (his italics, p. 92). To help scholars find their way among the extremely disordered fragments of Cotton Otho B. x,5 I have devised a “virtual” foliation, which keeps track of the correct order of the surviving bits of leaves, but also gives the official British Library foliation number in parentheses. In this case, fol. 10(7) means that the leaf is numbered 7 in the British Library foliation of the disordered manuscript, but that it should come tenth in the correct order of the surviving leaves. The line references provide both folio lines and Skeat’s verse lines. Thus fol. 10(7)r3:29 identifies a reading on line 3 of the recto, which corresponds to Skeat’s verse line 29.6 As we shall see, without a special foliation and dual lineation, it would be extremely difficult for anyone to keep track of the other surviving portions of this text. 

[Fig. 1: fol. 10(7), recto and verso]

Figure 1 provides a greatly reduced overview of fol. 10(7), recto and verso, as it is now framed in its nineteenth-century paper binding. Far superior to the limitations of print facsimiles, an electronic facsimile allows one to study a dynamic image of the folio in detail at high resolution, in full color, enhanced by ultraviolet, and much enlarged.7 At its normal size one is able to see where Sir Frederic Madden identified the text in pencil as “SS Julianus and Basilissa” in the upper-left frame, recto. As the shape of the burnt leaf indicates, the greatest losses of text are concentrated in the beginning of the first nine lines on the recto, and the ends of the first nine lines on the verso. 

A substantial part of the folio survives and most of what survives is relatively easy to read, except for sections obscured by gauze and tape, which were somewhat recklessly used, after Skeat’s time, to forestall further deterioration. These obscure sections are made more legible with ultraviolet fluorescence.

[Fig. 2: fol. 10(7) recto, enhanced by ultraviolet]

In the following transcription the surviving text is in bold, with Skeat’s restorations bracketed in italics. The folio boundaries for lines 1-9 are conjectural.

          7 him þa begeaten.] Þa wurd[on gegearcode þa
          gyftu æfter gewunan] 7 hi butu beco[man on anum
          bedde tosomne.] Hwæt þa iulianus [hine georne
          gebæd to ðam hælende] criste þæt he [hi]ne gehælde [wið
05      ealla onten]dnyssa 7 yfele costn[unga. Ða wearð
          þæt brydbed] mid bræþe afylled swy[lce þær lægon
          lilie 7 rose. Ð]a cwæð basilissa to [þa]m c[lænan
          brydguman, Hit is winte]rtid nu 7 ic wundrie
          þearle hwanon] þes [wyrt]bræð þus wynsumlic[e
10      steme 7 me nu ne lyst nanes sinscipes. ac þ[æs
          hælendes geþeodnyss[e] mid geheald[enre clen-
          nysse. Iul[ianus and]wyr[de] þam æðelan [mædene 
          Þes wynsuma bræð þ[e ðu w]undrast þ[earle næfð
          nan anginn . ne eac n[ænne] ende. Þes br[æð is of
15      criste seþe is clænnysse lufigend gif wy[t þurh-
          wuniað on anwealgum mægðhade. 7 hine [clænlice
          lufiað þonne cume we to his rice. 7 wyt ne beo[ð to-
          twæmede ac a to worulde blyssiað. Basilissa [cwæð
          þæt heo on clænum mægðhade þurhwunian wolde
20      for [ða]m wynsuman behate 7 habban þæt éce lif
          7 þone hælend to brydguman. Þa cleopo[de
          Iulianus on cneowgebedum þus. Confirma hoc d[eus
          quo]d operatus est in nobis et reliqua. Gefæstna þis
          hæl]end. þæt þæt ðu on ús gewyrcst. 7 basilissa sona
25      swa gelice dyde. Þa astyrede þæt brydbedd 7 beor[ht
          leoht þær scéan. 7 Crist sylf wearð þær gesew[en
          mid scinendum werode. 7 his modor m[aria mid
          hyre mædenlicum heape. Crist clyp[ode þa to ðam
          clænan cnihte 7 cwæð þæt he hæfde ofers[wiðod 

For the recto Skeat in his collation records only five variants, none of which are scribal blunders, and two are arguably superior to the readings in Cotton Julius E. vii. The five variant readings Skeat notes are anwealgum for ansundum r16:43; we for wit r17:44; wyt for wit r17:45; the omission of á r17:45; and the expansion of the phrase, crist wearð, to crist sylf wearð þær r26:54.

Although ansund, “sound, entire, unhurt,” appears often in his writings, it is plausible that Ælfric, in keeping with the alliterative style of this text, here used anwealg (i.e., onwealg), “sound, uninjured, uncorrupted”: gif wyt þurhwuniað on anwealgum mægðhade.8 The use of we, “we,” for wit, “we two” is unremarkable (Ælfric often uses both), as is the spelling variation, characteristic of both manuscripts, of y for i in wit. The omitted a, “ever,” is almost certainly a mistake, because Ælfric frequently uses a to worulde, but a is actually redundant in the phrase to worulde, “for ever.” The most significant variant reading is crist sylf wearð þær [gesewen], which explicitly states that “Christ himself was [seen] there,” rather than the more ambiguous crist wearð [gesewen], which Skeat translates, “Christ was [visible].” Skeat fails to note 15 additional variant readings on 10(7) recto: beco[mon? (beco-, not co-, follows butu) for coman r2:28; ontendnyssa (partitive genitive plural) for ontendnysse r5:31; bræþe for bræðe r6:32; sinscipes for synscipes r10:37; clen]nysse for clennisse r11-12:38; wynsuma (weak masculine nominative singular) for wynsuman r13:40; anginn for angin r14:41; ende for ænde r14:41; seþe for seðe r15:42; wyt for wit (again) r15:43; þone for ðone r21:48; cleopode for clypode r21:49; þus for ðus r22:49; brydbedd for brydbed r25:53; and modor for modur r27:55. 

As is usually true of the fire-damaged Cotton manuscripts that Henry Gough inlaid, the paper frames cover bits of the text along the damaged edges of the fol. 10(7) verso, and the use of fiber-optic backlighting will reveal additional covered readings.

[Fig. 3: fol. 10(7) verso, enhanced by ultraviolet]

In the following transcription of fol. 10(7) verso the surviving text is in bold, with Skeat’s restorations bracketed in italics. The folio boundaries for lines 1-9 are still of course conjectural.

          woruldlice gælsan 7 þone g]ram[lican feond. Of Marian
          werode wæ]s þus gec[lypod,  Eadig eart þu Basilissa
          forþan] þe ðu gebigdest  [þin mod to halwendum
          myne]gungu[m] 7 middan[eardlice swæsnysse
05      mid eall]e forsihst 7 þe sylf[e gearcost to wuldre.  Þa
          com to ða]m bedde boc fram [þam Hælende. 7 twegen
          his hal]gan mid twam kyn[ehelmum arærdan hi þa
          upp 7 he]ton hi rædan.  Þa r[ædde Iulianus þas
          word on þæ]re bec.  Se þe for mi[nre lufe middan]ear[d
10      fors]yhð  he bið soðlice [ge]teald to þam unbesmiten[um
          halgum, þe] næran on h[eo]ra life besmitene mid
          wifum. Ba]silissa bið [get]eald to þær[a] mædena getæ[le 
          þe Marian] folgiað þ[æs hæl]endes meder.  Æfter
          þissere] rædinge 7 oðrum tihtingum.  gewendon
15      þa hal]gan to þam hælende upp.  Hwæt þa iulianus
          unge]wæmmede heold his bryde.  7 hi wæron geðeodd[e
          mid s]oþre clænnysse gástlice þeonde on godes
          gewyt]nysse .  Heora fæderas wǽron gefyrn cristene .
          hit] gelamp ða hraþe þæt hi of life gewitan 7 læfdon
20      h]eora æhta þam æþelum mannum .  Iulia[nu]s þa
          dælde be his drihtnes willan heora l[and a]re .  þe heo[m
          gelæfed wæs .  7 arærde him mynster . 7 his mædene
          o]þer .  He wearð þa fæder ofer fela muneca .  7 [Basi]lissa
          m]odor ofer manega mynecena .  7 hi þa gas[tlican
25      we]rod under gode well gewissodan .  on dæ[g]hwamlicre
          lar]e to heora drihtnes willan .  Oð þæt seo reþe ehtnys[se
          on Egypta] lande becom .  fram þam wælhreowan
          casere þe w]æs geciged dioclytianus .  Þa gebædon
          þa hal]gan hi to þam hælende .  7 he him asende

For the text on the verso Skeat notes only four variants, hraþe for raðe v19:79, willan for wyssunga v21:81, fela for fæla v23:84, and well v25:86, which the Julius scribe omits. Skeat overlooks 19 additional variant readings on 10(7) verso: ðu for þu v3:60; gebigdest for gebygdest v3:60; Þa for Ða v8:67; fors]yhð for forsihð v10:68; s]oþre for soðre v17:76; ða for þa v19:79; gewitan for gewytan v19:79; æþelum for æðelum v20:80; heo[m for him v21:82; gelæfed for læfed v22:82; oþer for oðer v23:83; gewissodan for gewyssodon v25:86; drihtnes for dryhtnes v26:87; willan for wyllan v26:87; for v26:88; seo for se (ehtnys is feminine nominative singular, not masculine) v26:88; reþe for reðe v26:88; ehtnys[se for æhtnysse v26:88; wælhreowan for welhreowan v27:89. In short, Skeat lists only nine of the 43 variant readings on folio 10(7).

To make the results of any and all collations accessible for later investigation and analysis, an editor of an electronic edition must encode or tag everything (variant readings, pointing or punctuation, abbreviations, erasures, corrections, ultraviolet enhancements, descriptions of damage and misplacements, and any other physical features of the manuscript record). Comprehensive tagging or encoding in conformance with what is called a Document Type Declaration (or DTD)9 paves the way for later interactive searching and analysis by scholars who may hold diametrically opposite views of the evidence. Basically, in order to help, a computer must have some encoded way to locate the things we want it to find. A traditional scholarly edition collects textual data and normally displays it in textual notes, as Skeat does here. But Skeat’s notes do not make it possible to discover, for example, how often the Otho scribe used y where the Julius scribe used i in any or all words, or where the scribal pointing in both manuscripts coincides. It is customary in Old English editions to pass over many minor spelling variants, such as y for i or þ for ð, from collations with other manuscripts, as Skeat may have done deliberately here, and to ignore such manuscript features as scribal pointing in another manuscript, as Skeat does, and even scribal abbreviations, which Skeat however records in his text by means of italics. 

Exhaustive textual details, however, are prohibitively expensive and drastically clutter traditional print editions. In an electronic edition, on the other hand, it is advantageous to tag or encode exhaustively, because textual details cost nothing but the editor’s time and trouble, and there is no need to display the results unless someone is actually looking for them. For example, it is a simple matter to encode scribal pointing as metrical or syntactic or both, which would allow scholars to weigh the manuscript evidence for each kind of pointing. Then, with the aid of electronic search facilities, users of the edition will be able to fashion their own individual investigations, looking in a moment, for example, for differences in scribal practice in the use of thorn and eth or i and y, for the use or non-use of abbreviations, for different spellings of the same words in the same or different manuscripts, for Old English synonyms like ansund and anwealg meaning “sound” and “uncorrupted,” or for adjectives modifying mægðhad, “virginity, maidenhood.” 

An edition of the fragmentary text of Julian and Basilissa, consisting of only parts of three folios of an original seven,10 is part of a larger project to provide an electronic edition of what survives from Otho B. x as a collection. The use of fragments, usually reduced by collation to a few footnotes as in Skeat’s edition, as a primary text has the effect of reversing the goals of a traditional standard edition like Skeat’s. The fragment becomes the main text, using Skeat’s text to restore lost sections, but relegating his text to variant readings wherever text survives in the fragment. In order to understand the fragments in their original context, it is necessary to create a full glossary of the new composite of Skeat’s text with the fragments embedded to replace it wherever readings survive. To create a readable and ultimately searchable electronic edition of sometimes almost completely illegible manuscript fragments, I first collated the readable bits of text with Skeat’s online edition available through the Complete Corpus of Old English in Electronic Form.11 Using specialized mark-up or tags conforming to a “DTD” designed for damaged manuscripts from the Cotton Library, I next make transcripts of the fragments as the primary texts, and encode Skeat’s text as variant readings. The first stage of tagging, then, separates the text into folios and folio lines, while tagging as well Skeat’s verse lines. Using style sheets designed for the purpose, I can then transform the encoded files into Hypertext Markup Language (HTML), and use a web browser to display both transcripts and modern editions from the same tagged document. Graphical interfaces we are developing will allow users to integrate images, transcripts, modern editions, comprehensive glossaries, editorial annotations, ancillary materials, and search facilities to provide unprecedented access to these previously unedited texts.

After the first stage of tagging has identified the folios and folio-lines, tagged the verse lines of Skeat’s edition, integrated any readings that differ from Skeat’s text, tagged the latter as variants, and added the Cotton Library DTD, I use this “lightly tagged” document to generate an exhaustive word list in preparation for making a glossary. We have developed a very useful Electronic Editing tool for the purpose.12 The tool processes the encoded file and presents a working environment, with wordlist and corresponding text on one side and templates for all parts of speech on the other, as shown in Figure 4: 

[Fig. 4: Glossary Tool]

When the editor chooses a word from the text to gloss (in this case anwealgum), the word is simultaneously highlighted in the text in the lower left to provide its context, in the wordlist in the upper left, and as default entries in the “glossed word” and “location” fields on the right side of the Glossary Tool. Since anwealg is an adjective, the editor then selects “adjective” from the templates, and fills in the required information to define the word and describe its grammatical properties. All of these fields are automatically tagged in XML for use in the database, making it possible to search and otherwise access all features and any combination of them throughout the glossary.

Standard editorial work proceeds on the text, as well, and everything is likewise encoded for use in the database (these collations serve to encode variant readings, emendations, conjectural restorations, ultraviolet readings vs. daylight readings, etc.). The glossary production tool ensures that any editorial changes to the text must be updated in the glossary. A browser interface with programmed applications displays image and text together, allows for precise linking between them, and accesses the database for apparatus, glossary, translation, and user-defined searches. The browser interface used while editing the text will ultimately be included in the user’s version of the edition to allow users to emend or develop the glossaries. By compiling a single comprehensively encoded file with multiple style sheets, users may display text as a diplomatic transcript, or as prose or verse with or without editorial brackets and italics, and without losing access to textual information of any kind. Using a collection of customized style sheets and a transforming processor, users can also develop their own transformations of the master document and display the results in a browser.

Surely one of the strangest looking folios in all Old English manuscripts is folio 67, the last “leaf” of Cotton Otho B. x, as its charred remnants were mistakenly ordered in a mid-nineteenth-century restoration. It looks like a surreal jig-saw puzzle of a domestic disturbance involving a raised frying pan (or perhaps a pie).

[Fig. 5: BL fol. 67, recto and verso]

It is so difficult to read that Skeat understandably failed to realize that “folio 67” belonged to Julian and Basilissa when he collated readings from Otho B. x with Cotton Julius E. vii. N.R. Ker in his Catalogue of Manuscripts containing Anglo-Saxon was apparently the first scholar after Frederic Madden to recognize not only that “folio 67” comes from Julian and Basilissa, but that it is actually the remnants of two different folios.13 In his Catalogue entry, Ker tersely observes:

5. [f. ‘32v’ Idus Ianuarii. Passio Sancti Iuliani Et Sponse Eius Basilissa. ‘Iulianus wæs gehaten...’.] As Skeat, i. 90. ‘þa ... asende’ and ‘þystru[m] ... ðam’ (Sk., ll. 29-91, 219-333) remains in part (ff. 7: 67 fragm. 2, 67 fragm. 1).

He notes elsewhere without further comment, “67 fragm. 1 ... reversed” (p. 225). Given Ker’s extremely taciturn style, it will be helpful here to explain and illustrate what he means. His catalogue entry records that the Passio Sancti Iuliani Et Sponse Eius Basilissa, celebrated on the Ides of January (13 January14), was the fifth item in Otho B. x, and that it began on the verso of folio 35 in the foliation recorded by Humfrey Wanley. The reference to Skeat is to volume 1 of his edition, beginning on page 90. The remaining fragments correspond to lines 29-91 for fol. 10(7) and to lines 219-333 for the two folios combined on “BL fol. 67” of Skeat’s edition. It will be easier to visualize the correct order of the fragments by graying out sections that are out of sequence.

[fol. 11(67R)r]            [fol. 11(67L)v]            [fol. 12(67Rv)r]            [fol. 12(67Lr)v]

[Fig. 6: fols. 11(67R) recto and 12(67Rv) recto in order]

Figure 6 illustrates that the part of so-called “folio 67” that should be next after fol. 10(7) verso is the right (R) side, instead of the left, as Gough mistakenly thought when he framed the two fragments. Its virtual foliation number is therefore 11(67R) recto. But when the page is turned, the verso of this fragmentary leaf is now on the left (L) side, hence properly called fol. 11(67L) verso. As Ker indicates, the next fragmentary leaf was “reversed,” that is, pasted in back to front. Thus the real recto is on the right side of the verso (Rv). I therefore call it fol. 12(67Rv) recto. The last page is the one erroneously framed first in the sequence, here called fol. 12(67Lr) verso. 

As Figure 7 illustrates, a diplomatic transcription alone is relatively useless:

[Fig. 7: transcript of fol. 12(67Lr) verso, enhanced by ultraviolet]

When filled in by collation with Skeat, lines 304-333, the transcript of the folio assumes what must have been close to its form before the fire:

          binnan þam cwearterne 7] geba[d mid þam Cristenum,
          oð þæt Martianus hi to þam] mart[yrdome gefette.
          Hi wurdon þa gebroh]te gebunde[ne on racen-
          teagum  ealle to þam tun]num 7 to [þære ontend-
05      nysse.  Þa cw]æð [martia]nus m[id mycelre
          angsumnysse  to] þam halgum weru[m 7 to his
          agenum suna,  eala] hwilc an[wilnys 7 geortruwad
          wylla,  þurh ða þeo]s fægere g[eogað nu for-
          wurðan sceall.  Eala] þu iulian[e, þe awendest
10      minne sunu  swa þæt] he min [ne ræcð ne eac
          þære mede]r .  Þa com [seo modor mid
          mycelre sarnyss]e .  7 ealle heo[re hyredmen to
          þære heofun]ge  7 manega [oðre menn] to [þære mycclan
          wæfers]yne .  Þa cwæ[ð þæs] deman su[nu to his
15      dreorigan fæder,  Ne þur]fe ge [us bemænan
          ne urne si]þ .  bew[epan be]wepað eo[w sylfe, we siðiaþ to
          heofonum] .  we [farað] þurh ðæt fir [unforhte
          þurh God ] 7 we ansunde becumað ef[t] to eo[wrum
          gesihþum.  Þon]ne þu me eft gesihst ge[sundne
20      of þam fyre,  ge]ðafa þæt myn modor m[e ge-
          spræca]n  7 sume þreo n[iht on] min[um ræde
          beo]n .  ic w[ene] þæt ðu ne f[orleosa naðor ne
          hi ne] me .  [Þa we]arð [seo] modo[r on mode
          geblissod,] .  7 se []der [cwæð þæt h]e þæs cnapan
25      willan  wold]e gefremman gif he of  ðam [fyre come.
          Þa het se] dema his g[i]ngran þis [don,  7 eode mid
          his wife] aweg to his huse .  fo[r]ðan ð[e he] ne [mihte
          gese]on hu his sunu forburne .  Þa het s[e] unde[r-
          gerefa hi ealle gebri]ngan  into ðam tunn[um 7

[Fig. 8: fol. 12(67Lr) verso, ultraviolet, detail]

A thorough encoding of the damage to and restoration of this folio would include tagging for cause of damage (fire, hole, tear, water, other), its description, extent, and source of transcription (high-intensity light, microfilm, or ultraviolet image), in addition to sources of restoration (here Skeat’s edition along with ultraviolet fluorescence and image processing).

The restoration of fol. 11(67R), corresponding to lines 219-245 in Skeat, presents many interesting textual and technical problems. Ker indicates that the first surviving recognizable word on the fragment is þystru[m] r4, but there are other bits of text preceding it that image processing discloses. At the tip of the fragment an e and most of m from [c]em[pan] survive; below that in line 2 is -um þær- from [weardmann]um þær[a], and in line 3 -ft gecirr- from [æ]ft gecirr[on] is still visible. And the final -um from [blind]um remains before þystru[m] in line 4:

[Fig. 9: 11(67R)r1-4, ultraviolet processed]

A painstaking collation with Skeat’s text continues to recover, line by line, expected bits of the succeeding text until line 8, when for five consecutive lines on a postage stamp-sized bit of vellum, completely unexpected and uncollateable letters appear: min 11(67R)r8:223, iste 7 ða r9:224, eofodo r10:225, or he r11:226, and ser r12:227.

[Fig. 10: fol. 11(67R)r8-11]

In view of the obvious mismatch of the text on this fragment, it seemed most probable that Henry Gough had pasted the fragment on the wrong folio. A futile search for combinations of these letters in close proximity anywhere in Julian and Basilissa seemed to prove, moreover, that they came from another text entirely. The mystery was temporarily compounded when the expected and collateable letters, hte 11(67L)v8:223, urdon v9:224, oðra v10:225, and æs v11:226, were found on the other side of this fragment (less text is visible on the verso, because of the retaining edges of the paper frames).

[Fig. 11: fols. 11(67R)r8-11 and 11(67L)v8-11 are reversed]

          nysse his dri]hte[n þæt he foresceawode hu
          hi gefullode w]urdon [Þær wæron binnan þære
          byrig seofan gebr]oðra [Cristena, þæs caseres cynnes
          7 heora fæder w]æs [Cristen,  þam alyfde se casere

[fol. 11(67R)r8-11 restored with misplaced 11(67L)v8-11]

With the assurance that the fragment was in its correct place on the recto, it became apparent that the unrecognized text from the recto that could not be collated with anything must contain spelling variants. Moreover, the fragment reading iste 7 ða was now easy to find in line 9 of the verso:

me to his Cr]iste 7 ða  [seofan cnihtas þe be

With this line as a guide, min became easy to place with minne in line 8, and ser likewise fell in with casere in line 12. Knowing from the collation of folio 10(7) with Skeat’s edition that the Otho scribe used the spelling variant cleopode for clypode and heom for him, I then realized that the Otho scribe had spelled lyfedan as leofodon at line 10, and hyra as heora at line 11:

          ealle tobrytte, 7] min[ne sunu gebygde fram
          me to his Cr]iste 7 ða [seofan cnihtas þe be
          þinre leafa l]eofodo[n buton ehtnysse on
          þyssere byrig f]or heo[ra mycclum gebyrde.
          Þa asende se ca]ser[e þisne cwide ongean, gif

[fol. 11(67L)v8-12 restored with misplaced 11(67R)r8-12]

There was yet another unsolved (or rather, unnoticed) mystery on this folio, attesting to its practically unreadable condition without special lighting, image processing, and magnification. At lines 18 and 19 the bit of vellum containing what appears to be a couple of words in the same ink but in some unknown language with a peculiar script.

[Fig. 12: fol. 11(67R)r18-19]

What happened is that Gough pasted the fragment in the right place but upside down:

[Fig. 13: fol. 11(67R)r18-19 rightside up]

Once set aright, the text fits in where it belongs:

          ge[fullode] . Hi eodon ða [on niht 7 Godes ængel
          hi [lædde] 7 ð[æt] cwearter[n geopenade mid his handa

The same problem occurs, of course, on the verso, where the words þa godes and hi are upside down at lines 18 and 19:

[Fig. 14: fol. 11(67L)v18-19 upside down and right side up]

When set right side up, the appropriate text emerges, although the top of insular s is gone in godes, and only the tops of ah in aht survive:

          Martianus] þa gode[s me]n gef[eccan 7 axode
          Iulianum hweðer] hi [aht sme]adon [ymbe hyre.

At the bottom of the same folio a collation of fol. 11(67R)r26-29 with Skeat’s edition (lines 241-245) reveals that the Otho scribe has omitted exactly two lines of verse. To my knowledge, no one has noticed this omission. The missing text comes from the last two lines, a section of the damaged leaf that is tolerably well preserved and legible, although the left side of the fragment is confusingly misaligned with the right side:

[Fig. 15: fol. 11(67R)r27-29, enhanced by UV]

Skeat’s text reads,

                              woldon for cristes naman on þam cwearterne þrowian .
          Þa het se cwellere hi . of þam cwearterne gelædan .
          and axode hwi hi woldan butan ehtnysse þrowian .
          Þa cwæð se yldesta broðor to ðam arleasan deman15

The version in Otho B. x (here digitally realigned) appears to be a deliberate revision, removing the potentially merciful action of the judge: “They were willing for Christ’s name to suffer in the prison. Then said the eldest brother to the impious judge....”:

[Fig. 16: fol. 11(67R)r27-29, with left side realigned]

          ....woldon for cr[istes
          naman on ðam cwearterne ðrowian. Þ[a cwæð se
          yldesta broðor to ðam arleas[an de]man....


Restored by collation with Skeat and with the misplaced bits in their proper places, the remaining text on the fragment of fol. 11(67R) recto is:

          Þa cwædon þa c]em[pan þe se cwellere gesette þam halgum
          to weardmann]um þær[a wæron twentig,Unrihtlic
          us bið, þæt we æ]ft gecirro[n fram þysum
          beorhtan leohte to blind]um þystru[m fram
05      life to deaðe, fram soðe to l]easu[n]ge . Hi f[eollon
          þa ealle to Iulianes fotum] mid gelea[fan herigende
          þæs Hælendes nam]an . Iulianu[s þa bæd mid onbryrd-
          nysse his dri]hte[n þæt he foresceawode hu
          hi gefullode w]urdon [Þær wæron binnan þære
10      byrig seofan gebr]oðra [Cristena, þæs caseres cynnes
          7 heora fæder w]æs [Cristen,  þam alyfde se casere
          heora cristendom to healdenne butan ælcere
          ehtnysse for ðam arwurðan cynne. Hi hæfdon
          ænne mæssepreost swiðe mæres lifes, Antonius
15      gehaten, þ]e him mæ[ssan gesang. Þas geneosode
          se Hæ]lend 7 het hi [gan to þam cwearterne mid
          heora] mæssepreoste [þæt þa men wurdon
          ge[fullode] . Hi eodon ða [on niht 7 Godes ængel
          hi [lædde] 7 ð[æt] cwearter[n geopenade mid his handa
20      hrepunge. Þ]a sædo[n þa] geb[roðra þæt se
          ]lend hi asænde mid [heora mæssepreostæ
          þa men to fu]lligenne. Iulianus þa son[a þæs
          ða[ncode God]e 7 wearð þa gefu[llod se fore-
          sæda cna[pa 7 his f]æder cempan on [Cristes
25      naman en[demes.] Þis wearð þa ge[cyd þam
          cwellere m[arti]ane þæt ða seofon ge[b]r[oðra
          bu]ton ælcere [eh]tnesse woldon for cr[istes
          naman on ðam cwearterne ðrowian . Þ[a cwæð se
          yldesta broðor to ðam arleas[an de]man

The restored text on fol. 11(67L) verso is:

          Andsæte bið þæt treow þe æfre grewð on leafum 7 næfre
          nænne wæstm his scyppende ne bringð swa synd we
          Cristene gi]f ure c[ristendom ne bið acunnod.
          Þa het se c]wellere [hi to þam cwearterne ge-
05      lædan, 7 sen]de his gewr[it to þam wælhreowan
          casere, Ge]help uru[m godum 7 hat to þe gefeccan
          þisne dry] Iulianum [þe ure goda anlicnysse mid
          ealle tobrytte, 7] min[ne sunu gebygde fram
          me to his Cr]iste 7 ða [seofan cnihtas þe be
10      þinre leafa l]eofodo[n] [buton ehtnysse on
          þyssere byrig f]or heo[ra mycclum gebyrde.
          Þa asende se ca]ser[e þisne cwide ongean, gif
          Iulianus þurhwunað mid his geferum on þysum
          nim fela tunnan 7 do hi þær on innan, onæl hi
15      siððan e]alle oðr[um mannum to bysne. 7 gif he
          þurh his] dr[y]cræft [þæt fyr adwescan mæg
          gewitna h]i ealle loc[a hu þu wylle. Þa het
          Martianus] þa godes [me]n gef[eccan 7 axode
          Iulianum hweðer] hi aht [sme]adon [ymbe hyre
20      agene þearfe o]n [þæ]re h[wile oðþæt. Iulianus
          sæde, ure] geþanc is s[wa s]wa h[it wæs
          gif þu æn]ig wite beþohtest we [synd gearwe to þam.
          Þa mi]d þissere spræ[ce bæron menn] on
          ðære stræt] anes hæð[enes mannes] lic 7 se
25      heardhe]orta dema [het beran þo]ne deadan
          to his dom]setle . cwæð þa [to Iuliane] eower
          Crist] arærde þa deadan [to] life læt nu [ge
          swu]telian gif he soð g[od] sy 7 ge þisne aræron . 
          Þa andwyrde iulianus þam] arleasan deman 

The strange case of British Library “folio 67” and its companion in Julian and Basilissa teaches a valuable lesson about odd couples and the difficult nature of collaboration. We know that Sir Frederic Madden had correctly identified the bits and pieces of two distinct folios from Julian and Basilissa. It is unlikely in the extreme that he instructed Henry Gough to frame them together as if they were one folio, much less to put 12(67Lr)v back to front and frame it before 11(67R)r, or to place one small piece of 11(67R) back to front and another small piece upside down. While Gough achieved a restoration that has kept all of these pieces intact for a century and a half, “folio 67” also represents the breakdown of an otherwise fruitful collaboration of an odd couple, Madden and Gough, keeper and conservator, scholar and technical expert. Just as Madden did not have the technical expertise to rebind these fragments, so Gough did not have the expertise to read Old English. They stopped working together too early on this puzzle. A patient and sustained collaboration is likewise essential between another truly odd couple, humanities and computer science, for digital libraries to achieve their full potential for restoring, searching, and editing currently inaccessible humanities collections.16

Saint Simons Island, June 2002


*My continuing work on this badly damaged codex is part of an interdisciplinary project, “The Digital Atheneum: new techniques for restoring, searching, and editing humanities collections,” supported by the National Science Foundation, IBM’s Shared University Research grant, the British Library, and the University of Kentucky’s Center for Computational Sciences. The principal investigators are Kevin Kiernan for the humanities (with the invaluable assistance of Linda Cantara) and Brent Seales and James Griffioen in Computer Science.

1The “Passion of St. Julian and His Wife Basilissa” is edited by Walter W. Skeat, Ælfric’s Lives of Saints ... edited from British Museum Cott. MS. Julius E. vii with variants from other manuscripts, Vol. 1, EETS OS 76 & 82, pp. 90-115.

2As Clare Lees observes, “the transformation of sexuality into the gift of chastity is the prime component of the female saint’s life.” “Chastity and Charity: Ælfric, Women, and the Female Saints.” Tradition and Belief: Religious Writing in Late Anglo-Saxon England. Medieval Cultures 19. Minneapolis: U of Minnesota P, 1999, p. 147.

3“We have passed through fire and water, and thou hast led us into a cool place,” line 340.

4Andrew Prescott, “‘Their Present Miserable State of Cremation’: the Restoration of the Cotton Library.” Sir Robert Cotton as Collector: Essays on an Early Stuart Courtier and His Legacy. Editor C.J. Wright. London: British Library Publications, 1997. 391-454.

5To read the surviving leaves in the correct order using the official British Library foliation, one must begin with fol. 60v, then 36v, then 49, 1, 3, 5, 4, 50, and 6, before getting to the misplaced folios of Julian and Basilissa. There is another misplaced leaf following BL fol. 4 at the Bodleian Library, Oxford.

6For a discussion of this virtual foliation and the disorder of the manuscript that makes it essential, see Kevin Kiernan, W. Brent Seales, and James Griffioen, “The Reappearances of St. Basil the Great in British Library MS Cotton Otho B. x,” Computers and the Humanities 36:1 (February 2002), 7-26.

7All images from Otho B. x were digitized by David French and Kevin Kiernan for the Digital Atheneum project and are used with permission of The British Library Board.

8For anwealg in reference to chastity, see Bosworth-Toller under onwealh ii. The Toronto Dictionary of Old English cites the Julius E. vii reading from Julian & Basilissa for ansund, onsund, “whole, sound, having integrity,” in its figurative sense, b.ii, “of virginity, chastity: uncorrupted, unblemished, immaculate.”

9I will refrain in this essay from discussing in detail or resorting to the terminology of the XML technologies we are using. As Ælfric says at one point in Julian and Basilissa, Þeos race is swiðe langsum fullice to gereccenne, / ac we hit sæcgað eow on þa scortostan wisan, “This story is very tedious, to tell it all, but we tell it you in the briefest way” (lines 139-140). 

10Approximately a page (26 lines of verse) is missing before fol. 10(7). As Wanley indicated, Julian and Basilissa began with its rubric on a verso, following the Passio Sancti Mauri Abbatis, a text totally destroyed in the fire of 1731. Two folios are missing between 10(7) and 11(67R), and one and a half following 12(67Lv)r. With 101 verse lines missing at the end, Julian and Basilissa probably ended a few lines into the verso, followed by the Passio Sancti Sebastiani Martyris, which as Wanley says began on 39 verso of the old foliation before the fire. For Wanley’s foliation before the fire, see Ker (p. 225). 

11ed. Antonette di Paolo Healey, with Richard Venezky and Peter Mielke (Dictionary of Old English Project, Centre for Medieval Studies, University of Toronto, January 2000).

12The tool was programmed by my research assistant, Ionut Emil Iacob, and extensively used by my students, Kenneth Hawley, Demorah Hayes, and Jocelyn Taylor, as well as by students in my Spring 2002 seminar on Electronic Editing.

13Ker §177-178. Madden must have known that the fragments were two folios, not one, in order to reconstruct the text even to the point that he did, but it remains a mystery why he allowed his talented restorer, Henry Gough, to frame them as if they were parts of the same folio.

14As Skeat notes (p. 91 note 1), the actual feast day is 9 January.

15“They were willing for Christ’s name to suffer in the prison. / Then the tormentor bade them be led from the prison / and asked why they wished, without persecution, to suffer. / Then said the eldest brother to the impious judge...” (lines 242-245). The Otho version omits verses 243-244.

16I presented an early version of this paper as “Electronic Editions of Damaged Manuscripts” for a special session on “Whose Standards? MLA Committee on Scholarly Editions” for the Modern Language Association in December 2000. My student, Jocelyn Taylor, made a related presentation on “Creating a Glossary for a Damaged Old English Manuscript” for the National Conference for Undergraduate Research in Lexington, Kentucky, in March 2001.