LUCAN’S LOST GAULS: THE INTERPOLATION AT DE BELLO

This article discusses five spurious lines at Lucan 1.436-40. Reviewing the early printed tradition of De bello civili as well as examining the medieval manuscripts in which the lines are found, the study explores the extant evidence for the lines. In its search for the origin of the lines, the investigation comprises a discussion of the palaeography of the manuscripts, the poetic and contextual interpretation of the lines, and will venture a suggestion as to their date and presumptive author and the location where they were likely composed.

are conventionally printed: instead of Pictones immunes subigunt Accursius has Pictonis immunis subigit; he inverses lines 437 and 438; he prints Tricoros instead of Turonos, Medualle in place of Meduana, Adus instead of Andus, and Menabos for Genabos. Although he was aware of the grammatical and hermeneutical inconsistencies of the lines as he read them in his manuscript, Accursius left it to his readers to perform the necessary emendations. 8 It appears that it was Turnebus (1512Turnebus ( -1565 who first rearranged the lines in the way we saw them printed above (although he replaced 438 nebulis with ripis). According to a note in his Aduersaria, the French philologist saw them "in Belgica editione," but since he was only able to find the lines in a single manuscript (and one which had just the first four lines), he drew the conclusion that they were spurious. 9 The Belgian edition is mentioned in the context of another occurrence of the lines, hitherto unnoticed. At the foot of fol. 4v of the manuscript Berne, Burgerbibliothek, 45, a ninth-century copy of Lucan from Fleury, the five lines have been added in an early-modern hand, perhaps that of Pierre Daniel (1530Daniel ( -1603 or Jacques Bongars (1554-1612), the earlier owners of the manuscript, and connected with a line to their position after 435. The annotator claims to have found the lines "in Belgica editione" but regards them to be "valde suspecti"; however, the lines do not occur here as they are printed in the Belgian edition, as we shall see below, but according to the emended version as proposed by Turnebus, to whose Aduersaria the annotator refers with book and chapter.
is to be found in the fact that Accursius dates his apology for the work, entitled "Testudo," where he defends himself against charges of plagiarism, printed at the end of the volume, to November 1521, whereas the date of printing, found only a few lines below the date of the preface, is April 1524 (n.p.). 8 Accursius 1524: In Ausonii Mos. 468 (n.p.): "Hos autem, non aliter omnino referentes quam in codice ipso haberentur, et quod hic corrigendi locus non erat, legentium coniecturis discutiendos linquimus." 9 Turnebus 1599: col. 729: "Libr. i. Lucani quinque versus additi sunt in Belgica editione: quorum quatuor in vno duntaxat exemplari reperi, vt eo nomine mihi pene pro spuriis suspecti sint. confido autem me eos & melius ordinaturum, & fidelius exhibiturum …" The Belgian edition to which both Turnebus and the anonymous annotator in Berne, Burgerbibliothek, 45 refer ought to be Theodor Poelmann's 1576 Antwerp printing. 10 This contains the lines, numbered 436-40, printed almost exactly as Accursius did. 11 And indeed, a note to line 436 reveals that Poelmann read the lines in Accursius's Diatribae. In the same note, Poelmann also cites the lines from Willem Canter's Nouae lectiones, book III, printed at Antwerp in 1571. In this work, 12 the first four lines (436-39) occur in a discussion of other spurious or added lines in Lucan. In the note Canter quotes the lines with 436 Pictonus etc. in the singular, 437 and 438 in the order proposed by Turnebus, as well as displaying the forms Turonos, Meditana, and Andus. Canter does not mention a fifth line; his source must therefore have been different from that of Accursius. Gregor Bersmann, furthermore, in his Leipzig edition of 1589, prints the lines in the text (unnumbered) exactly as Poelmann; in a marginal comment, he also notes the readings of Canter. 13 The lines are also mentioned by Étienne Clavier in his 1602 edition of Claudian's opera. 14 Commenting on 32 Pictorum in the latter's panegyric of Emperor Honorius' fourth consulate, Clavier claims for reasons unknown that the first four verses (he was apparently unaware of 440) were inserted into Lucan's text by Jacques Cujas (1522-90). Since Cujas did not edit Lucan or comment on him directly, this assertion is difficult to substantiate, short of combing through his massive oeuvre of legal commentary. Hugo Grotius, however, in  Oudendorp seems to be the first editor to set off the lines typographically: in his 1728 edition he both brackets and italicizes them, and adds the comment that he found them neither Latin nor poetic ("neque latini, nec poëtici sunt"). He also expressed the sentiment, before launching into his own rather detailed commentary, that they were worthy neither of being emended nor explained ("addo, nec dignos, ut emendentur, aut explicentur"). Furthermore, the manuscript in which Oudendorp saw them, which had belonged to the mayor Hulst de la Haye, displayed the verses after 443 and, apparently, copied as part of the continuous text. 17 Oudendorp set the tone for future editors: henceforth, the lines are excluded from editions of Lucan's epic.
So far the printed tradition. Accursius claims to have seen the lines in a codex peruetustus, and Grotius, Oudendorp, and other critics mention that they have seen the lines in manuscripts. What, then, is the extant manuscript evidence for the spurious lines? II Whereas 440 is nowhere to be found outside the printed tradition, lines 436-39 are preserved by three manuscripts of Lucan's De bello civili: 18 The first is M (Montpellier, Bibliothèque Universitaire Historique de Médecine, H113), a manuscript from the second quarter of the ninth century, whose script "erinnert an Orléans-typ"; this belonged at one time to the abbey of Saint-Martin, Autun. 19 Used by all modern editors of Lucan, M is Hosius' "codex optimus," but Housman's "king of shreds and patches […], the manuscript which we could best dispense with." 20 Our spurious lines were added by a French seemingly twelfth-century hand mid-page in the left margin of fol. 7v, 21 where cropping of the page has resulted in the loss of the first five or six letters of each line. Their position after 435 Cebennas (spelled gebennas in the manuscript) and before 441 Tu quoque is indicated by an oblique line: 22 The next is R (Paris, Bibliothèque nationale de France, lat. 8040), an eleventh-century manuscript (either early or late, the dating varies), 23 which belonged to the abbey of St Benedict at Fleury-sur-Loire, 24 a gift of a certain Girard, according to an 'ex-libris', written in a hand contemporary with the main text, running across the opening of fol. 7v-8r and down part of the right-hand column of fol. 8r: "Hunc librum contulit Girardus monachus Patri Benedicto. Quem si quis tulerit anathema sit." 19 The placing and dating are Bischoff's in Bischoff & Ebersperger 1998-2014: vol. 2, 200, no. 2828. See also Munk Olsen 1982andGotoff 1971: 14. 20 See Hosius 1913: xlvii;Housman 1927: xiii. 21 Unless otherwise stated, the dating of scribal hands in this article is my own. 22 The tie-mark visible in the image over meduana is not picked up anywhere on the page. 23 Châtelain 1894Châtelain -1900  The same inscription is repeated in the same way at fol. 113v-14r. Our lines were added by a French late eleventh-or early twelfth-century hand at fol. 8r, immediately below the text, with a tie-mark indicating their position after 435 Cebennas (spelled gebennas in the manuscript).
Châtelain seems to suggest that if Girardus, the donor of the manuscript, is identified with a certain Giraldus, 25 "auteur de divers poèmes," he may be the author of our spurious lines; outside of the Histoire littéraire, however, this Giraldus is unknown. 26 R is normally not used by editors of Lucan.
The third and last manuscript is T (Paris, Bibliothèque nationale de France, lat. 8039) written, according to Bischoff, either late in the ninth or early in the tenth century, perhaps at Fleury. 27 The hands that copied the manuscript, which also includes Boethius' Consolatio, 28 are indeed French, but resist further and stricter classification. 29 Fabio Troncarelli suggests to me that the manuscript may have been housed at Saint-Evroult, 30 an abbey situated on the border of Normandy which had another copy of Boethius at the end of the ninth century, whose initials are similar to the ones in T. 31 What appears to be an ownership mark in the margin of fol. 61v, now almost entirely covered in ink, presumably by the manuscript's new owners, may provide further confirmation of this provenance. The text is written in two columns and surrounded and interweaved, for the first four and a half folios, by a plethora of glosses, which then begin to dwindle; they return at fol. 41r and continue to the end of the work at 49v. The last gathering is damaged (by fire?). Our additional lines were added in the upper margin of fol. 3v in a large, clear French tenth-or early-eleventh-century hand, not much different in age from that of the text hand, but in a higher grade script than the text itself. The cropping of the page has resulted in the loss of line 436. No visible indication in T signals the placing of the lines relative to the text (thus, unless it was explicitly stated in connection with the cropped-off line 436, the position of the lines could not have been gleaned by anyone who copied them from T). In fact, the large, round, elegant tenth-or early-eleventh-century hand of French pedigree that wrote the added lines on fol. 3v, with tapering descenders and feet devoid of serifs, is similar but not identical to a hand that relieves the main hand and writes part of the text on fol. 22r.

III
Based on the evidence of M and R, therefore, the position of the lines after 435 seems clear. However, aside from the obvious omission of 440, which is found in no medieval manuscript, the text as transmitted by these three manuscripts display significant differences from how they are conventionally printed.
First, the first three words of 436 are not in fact written Pictones immunes subigunt, but are found instead in the singular: pictonus immunis subigit (M and R; the line, as we saw, is missing in T). 32 Oudendorp gives as a reason for the change: "Numquam enim haec gens Pictoni, sed Pictones vocantur." 33 To be sure, Caesar uses the form Pictones (Gall. 7.4.6), as does Pliny the Elder (nat. 4.108 and 17.47) to describe the men of Poitou, otherwise referred to as Pictaui. Although it is used by Caesar and Pliny, however, the 'emended' reading Pictones is never used, to my knowledge, in Latin epic poetry, probably for the reason that the o is short (while the ending -es is long, thus yielding a cretic, impossible in hexameter). Pictonus, by contrast, may be scanned as a dactyl. Given that the words qualifying Pictonus (immunis and subigit) are both in the singular, the emendation to the plural is, to my mind, smacking of prurigo coniciendi: is it not more likely that Pictonus is a hapax than that all three words are wrong? I think the poet must be allowed some licence and wordplay in line with the other singular subject in these lines (439 Andus).
Usually, editors print the form Turones at 437; this is not, however, the reading of the three manuscripts, which all have Turonos. Grotius, Oudendorp (and others) were aware of the manuscript reading but emended it to Turonas, claiming to follow Caesar, whom they mistakenly thought used Turones ("sed Turones semper Caesar"). 34 Caesar in fact writes Turonos in the accusative plural (Gall. 7.4.6). Turoni is used by Tacitus in the nominative plural (ann. 3.41); 35 but Pliny the Elder writes Turones (nat. 32 As we saw above, Accursius, and therefore probably also the manuscript he was using, presented a variant of the singular reading (oddly substituting Pictonis for Pictonus Thus, the common reading of the manuscripts ought to be correct. Line 438 Meduana has been a source of some contention among critics. Thus, some editors print Meditana (Canter for example), others Medualle (Poelmann and Bersmann, presumably following Accursius); Meduana, however, is the Latin name for the river Mayenne, a branch of the Loire, as is attested by among others Gregory of Tours, 36 Theodulf of Orléans, 37 and other authors. It is thus from an early date a well-attested name for the Mayenne.
Anjou is the place (city or region) meant by Andus in 439. Since Caesar uses the form Andes to refer to the people of Angers or Anjou ('les Andécaves'), Grotius wanted instead to read Andis, but Andus is, again, a well-attested form, used by twelfth-century poets and prose-writers, ad propria / confusi cum mesticia, / Turoni cum leticia / sua revisunt menia. In occursum pontificis / vaditur a Turonicis / tam clericis quam laicis / cum vocibus hymnidicis." This 'Vita' is from Saint Martin of Pontoise (France, Val-d'Oise). The name of the author used to be in the last quatrain, but it was erased. According to Huygens, it may be supposed that his name was Guido. matum de Mosaicae historiae gestis 5.244); and it is a favourite word of Prudentius who, inter alia, uses it in the exact same position as our poet as a bridge between the fourth and fifth foot: Sic Lacedaemonias oleo maduisse palaestras (mam. 365); Sanguine iustorum innocuo maduisse recordans (c.Symm. 1.516). Because of these precedents, I wonder if maduisse should not be preferred to marcere, both because it is the lectio (paene) difficilior, and that it has epic precedent, but, most importantly, because it changes the rather awkward sense of marcere perosus, "loathing to be weak, languid or lazy," into "loathing to be wet" or even "loathing to be intoxicated." This seems to be a more fitting description of the effect of living near a river and its foggy banks. Although the evidence displayed by these sources is at times conflicting, the consensus opinion seems to be that the Mayenne carries with it negative connotations, whereas those of the Loire are positive. 43 Thus, on the whole, they chime well with the sentiment expressed in our lines. A critically updated version of the lines (not counting 440), taking full account of the manuscript evidence presented above, will read: Laon?) or Arnulf of Orléans who lived during the twelfth century." 46 Indeed, Arnulf, otherwise known for his exegesis of Ovid, wrote a comprehensive commentary on Lucan, the Glosulae super Lucanum, 47 and although no such commentary has been preserved in Anselm's name, the notes on Lucan (alongside those on Vergil and Statius) in the manuscript Berlin, Staatsbibliothek Preußischer Kulturbesitz, lat. fol. 34, have been associated with his school. 48 Neither of these scholars, however, seem to have been acquainted with the additional lines: Arnulf does not comment on them in his Glosulae, 49 and the commentary associated with Anselm also lacks mention of the lines. 50 Furthermore, these proposed attributions do not take into account the tenth-or early eleventh century date of the added lines in T, which provides a terminus ante quem for their composition. This terminus would also rule out Marbod's authorship of the verses; he was indeed a Loire poet, but his active years fell in the late eleventh-and early twelfth-century (he died in 1123). The tradition of attributing the verses to Marbod must have arisen from convenience. There is nothing that connects him to these lines other than his use of the words Liger and Meduana, in common with dozens of other poets. Guyet, or whoever it was who first made the claim, most likely suggested Marbod as the author of the lines without having seen T or without being able correctly to date the hand.
Châtelain's suggestion that it was Giraldus, the otherwise unknown donor of R to the library of Fleury, is impossible to substantiate: there are no surviving works by this man with which to compare the lines. It should also be pointed out that the ex-libris inscription bearing Giraldus's name is written in an entirely different hand than the one that added the spurious lines in R. In fact, the hand of 'Giraldus monachus' looks rather more like the scribe who wrote the main text in a French 46 Getty 1940: 135. 47 Marti 1958: 58. 48 See de Angelis 1997: 75-136. 49 Marti 1958 Staatsbibliothek zu Berlin, Preußischer Kulturbesitz, lat. fol. 34, fol. 3v, where the commentary jumps from commenting on 435 Gebennas (written Gehennae in the manuscript) to 441 tu … treuir.
eleventh-century hand. 51 In contrast, the addition of lines 436-39 is executed in a hand of the early to mid-twelfth century, spiky and with a crammed ductus. If Giraldus were the author of the lines, would it not be reasonable to assume that he wrote them in the same hand as he added the ex-libris? The least one would expect is that the two scripts dated to the same century. Consequently, it would be prudent to look beyond the suggestions of earlier scholarship to find the author of our added lines. Beginning by reviewing the evidence for where the lines could have originated, let us try to establish the ubi of our investigation.
First, collation of Lucan's text surrounding the additional lines in our three manuscripts (1.392-465) indicates that they are textually closely associated. It should be said at this juncture that Lucan presents one of the most complex textual traditions of the Latin classics, and no critic has been able to successfully bring order into the chaos offered by the manuscripts from the earliest stage in the tradition. 52 Whereas M is a staple ingredient in any critical edition of Lucan, the last editor to call on the authority of R and T was Lejay, 53 who assigns them to a different branch of the tradition than M. 54 A recollation of the larger passage in the midst of which the verses are found confirms the strong association between R and T, 55 but also puts Lejay's strict division into some doubt; indeed, it seems that M, at least for this passage, shares some readings with both R and T, 56 and was, at an early stage, corrected against a manuscript with readings similar to those furnished by R and T. 57 The three manuscripts are thus textually closely related. Second, the three manuscripts belong to the same geographic region. As was seen above, M was written in a script typical of ninth-century Orléans. Even if it is not known that R was written at Fleury, from an early date it belonged to that abbey. T is also French and may actually have been written at Fleury. If this is correct, all three manuscripts converge on a very specific area of north-western France, the region around Orléans on the Loire, Fleury being less than 35km upstream from Orléans. The additional lines are all about tribes, rivers, and towns in that region, the Loire valley. The source of the additional lines is likely to be found in the same area.
As to the date at which the lines were composed, the quando, it was seen above that T offers a terminus ante quem: the date of the lines cannot be more recent than the date of the hand that wrote them in T, that is the tenth-or possibly the early eleventh century. They could, of course, have been composed earlier -as early as Lucan's own time -and not entered into an extant manuscript until the tenth century. But this we can never know. Let us therefore continue to examine the quid -the contents of the lines and their possible source.
Lucan's main source for his catalogue of Gallic tribes was Caesar's De bello Gallico: seventeen of the twenty tribes mentioned by name are found in this work. 58 Although names of Gallic tribes are scattered among all seven of Caesar's books of De bello Gallico, at one point in book VII Caesar furnishes a list of eight tribes that he joins to himself in the fight against Vercingetorix (Gall. 7.4.6): "celeriter sibi Senones, Parisios, Pictones, Cadurcos, Turonos, Aulercos, Lemovices, Andes reliquosque omnes, qui Oceanum attingunt, adiungit." 59 Interestingly, the three tribes specifically mentioned in our four spurious lines -the Pictones, the Turoni, and the Andes -occur in the exact same order in Caesar's narrative (with other tribes interspersed, of course). Since, furthermore, as was pointed out above, Turonos is the form preferred by Caesar (and not Turones), and since Turonos is the form found in the manuscripts, it is likely that the composer of the verses had access to this form through Caesar. These two features could hardly be coincidental: our mysterious author was surely acquainted with Caesar's text. Is there a poet with a connection to the abbey of Fleury in the tenth century, who knew Caesar? V There was no shortage, as a matter of fact, of Latin poets at Fleury: Abbo (d. 1004) and Haimo (d. 1118) are the two most famous. Whereas Haimo's dates are too late for our poet, Abbo's would fit; but, judging by his writings, Abbo does not seem to have had a particular predilection for Caesar. 60 A much likelier candidate is Aimoin of Fleury, whom we have already encountered. Aimoin, who died in around 1010, not only wrote a history of the Frankish people, 61 but included in it a chapter entitled "De Gallia secundum Caesarem," in which the form Andus, central to the spurious lines, is used. 62 Dedicating his work to his abbot Abbo, he prefaces it with the following words: Admonitionis itaque tuae non immemor, qua saepissime hortatus es ut situm Germaniae vel Galliae, in quibus haec quae referentur acta sunt, non praetermitterem, ea quae in auctoribus Julio, Plinio ac Orosio invenire potui colligens, huic opusculo inserendo voluntati sublimitatis tuae satisfacere commodum duxi. His igitur adjunxi quae Julius de Germanorum Gallorumque moribus ac institutis in libro suae interserit Historiae (PL 139: 627B).
France, possessed a copy of De bello Gallico which, as a matter of fact, contains ample notes in Aimoin's own hand. 63 Unfortunately, comparing Aimoin's notes with the hand in T (or M and R for that matter) does not even yield a potential match: the hands are utterly different.
Nevertheless, in annotating De bello Gallico in Paris, BnF, lat. 5763 Aimoin takes particular interest in the names of the Gallic tribes mentioned by Caesar, especially on fol. 71r and 71v, where the Turoni and the Andus, amongst others, are mentioned, and he even repeats some of them in the margins. Although this is far from certain proof of the authorship of our spurious lines, it nevertheless provides testimony to Aimoin's interest in Gallic tribes, an interest that perhaps could have led to the composition of four occasional lines added to a copy of Lucan. Clearly, the fact that the lines are not written in Aimoin's hand in any of the surviving manuscripts does not exclude their composition by him. A more substantial objection would be that Aimoin has no reputation as a poet. Nonetheless, he did write poetry, although most of his works are written 63 Paris, BnF, lat. 5763, s. IX, from Fleury. It is BF1062 in Mostert 1989: 208; the volume also carries notes in the hand of Heiric of Auxerre; see Werner 1960: 83;Manitius 1911-31: vol. 2, 240. See also Brown 1979 Aimoin's hand in Paris, Bibliothèque nationale de France, lat. 5763, fol. 71v (detail).
in prose. Aimoin concluded his history of the Franks with a poem, translatio patris Benedicti, which, written in dactylic hexameters, confirms his skills in prosody. 64 Thus, while, technically speaking, any monk attached to the abbey school could have composed the lines, as an exercise in poetic composition if nothing else, the connection to Caesar tips the scale: Aimoin knew Caesar like no one else at Fleury in the tenth century. This leaves us with, if not a definite answer to the question quis, at least a plausible author of the four first spurious lines, who fits all the necessary criteria. It now remains to return to the question of where Accursius found the lines, and where 440 came from.

VI
Given the existence of 440 and the other differences between the text as transmitted by the manuscripts and as it was printed by Accursius, his codex peruetustus is not identifiable with any of the three surviving witnesses to 436-39. Indeed, aside from M, R, and T, none of the seventysomething manuscripts collated for this study transmit the additional lines. They do appear, however, in a copy of the editio princeps of Lucan kept at the Vatican Library, 65 not as part of the printed text, but added by an annotator. Until the end of Book 5 (fol. 54v), the volume is heavily glossed in a variety of cursive humanist hands, using both black and red ink. On fol. 9v, there is an insertion mark after 1.435 Gebennas, which is picked up by a note written in the lower margin in what appears to be a fifteenth-century humanist cursive hand (a different hand from the one that wrote most of the other glosses): Transcribing the note yields a familiar text: In antiquo codice erant haec carmina alia Pictonus immunis subigit sua rura: nec ultra In nebulis Medualle tuis marcere perosus Instabiles Tricoros circumsita castra cohercent Addus iam placida ligeris recreatur ab unda Inclyta Caesareis Genabus dissoluitur alis.
Indeed, the note proffers the lines in almost exactly the same way as Accursius prints them. All five lines are included, arranged in the same order as Accursius published them (with the inversion of 437 and 438): it has Medualle for Meduana, Tricoros for Turones, Addus for Andus. The only difference is that the note has Genabus for Accursius' Menabus. This latter form, however, appears as an interlinear gloss written in red ink just above Genabus; this could have been Accursius' source, and he could have picked the reading Menabus from the gloss.
The hand adding the lines has been identified as Pomponian, that is belonging either to Pomponius Laetus, a student of Lorenzo Valla, who founded an antiquarian and philological academy, the Accademia Romana, at his house on the Quirinal, or to one of his many emulators, the so-called 'Pomponiani'. As far as we know, however, Accursius was never a member of Pomponius Laetus' first Accademia Romana, 66 but he may very well have been a second-generation follower of Pomponius or otherwise inspired by their movement.
At this juncture, a few options seem possible. Either Accursius wrote the note himself, 67 copying the lines as he had discovered them in a codex peruetustus or antiquus, the whereabouts of which are still not traceable. Or he could have first found the lines already written in the incunable, and merely repeated the opinion of the scribe that they were found in codice antiquo, not having seen them himself but trusting the veracity of the reporter. A third option is that Accursius had nothing at all to do with this gloss, which could have been written by someone who had access to the same or a similar codex in which Accursius saw the lines, and wrote them down in the margin of the incunable.
Whatever the precise origins of this reference, the text in the three manuscripts that preserve the additional lines indicate that the manuscript, if there ever was one, in which Accursius saw the lines, contained a version that had already expanded on the tradition. The absence of 440 from the manuscript tradition, as well as the inversion of lines 437 and 438 make this clear. But without the later emergence of other evidence, this is where the trail runs cold. Whether or not Accursius penned the notes in the Vatican incunable, there is still no trace of his codex peruetustus, and the origin of line 440 will remain a mystery. The only thing that can be said for certain is that 440 was never part of the medieval tradition of the spurious lines as they are found transmitted by the Loire manuscripts. Anyone who saw the first four lines in one of the extant French manuscripts (or in a manuscript that has subsequently been lost) could have made up the fifth line and added it in his own copy. He or someone else could have taken this copy with him to Italy, where it was later discovered by Accursius and/or the unidentified scribe in the Vatican incunable. 68 Whether he saw them in the Vatican incunable or elsewhere, Accursius printed the five lines in his notes on Ausonius, whence Poelmann picked them up and printed them at the place where they, eventually, acquired their conventional numeration. With the exception of the Hulst manuscript (whatever this was) and the codex peruetustus of Accursius (whatever this was), it seems clear that the four or five spurious lines were never regarded as authentic in the medieval period; in the manuscripts in which they are extant they are always copied outside of the main text, never as part of it. It was early modern scholars who were responsible for adding them to Lucan's text.
Summing up the evidence brought forth in the preceding pages, let me conclude: lines 1.436-39 are first recorded, and were probably crafted, sometime in the tenth or eleventh century, in the area around Orléans, probably at Fleury, by someone who found Lucan's list of Gallic tribes lacking. Perhaps he was a local patriot who wanted to add the tribes in his own immediate vicinity -the men of Poitou, Touraine, and Anjouto Lucan's original Bituriges, Nervii, Arverni, Sequani, Suessones and others. Alternatively, and more probably, he wanted to complete Lucan's catalogue -which relied on Caesar's account -by supplying the 'missing' peoples from Caesar, with whose text he was intimately familiar. Perhaps it was a combination of both. And perhaps the poet was Aimoin. To the four 'original' lines were added, at some other unknown point in time, line 440, perhaps by a native of the Orléanais -Cenabum is the Latin name for Orléans; 69 this 'extended' version, however, was never part of the French tradition, but was either brought to or composed in Italy, where the lines were discovered some three or four hundred years later and became part of the tradition of Lucan's De bello civili. 68 The only thing we know about the incunable's earlier provenance is that at one point it was owned by Fulvio Orsini (1529-1600), but none of the hands writing the notes surrounding the text appears to be his. 69 Caesar, Gall. 7.11.5.