¹ ASNSL 94 (1895), 149–60Google Scholar. (Henceforth cited as Schepss.) For a brief account of the most important work on Alfred and the commentaries, see Whitelock, Dorothy, ‘The Prose of Alfred's Reign’, Continuations and Beginnings, ed. Stanley, Eric G. (London, 1966), pp. 82–3Google Scholar. The best study is that by Otten, Kurt, König Alfreds Boethius (Tübingen, 1964), esp. pp. 4–19 and 119–57CrossRefGoogle Scholar. (Henceforth cited as Otten.) See also Schmidt, Karl Heinz, König Alfreds Boethius-Bearbeitung (Göttingen, 1934)Google Scholar. (Henceforth cited as Schmidt.) Brian Donaghey, S., ‘The Sources of King Alfred's Translation of Boethius' De Consolatione Philosophiae’, Anglia 82 (1964), 23–57Google Scholar, is in many ways a very judicious essay; but it suffers from having appeared in the same year as Otten's book and from the fact that its author was unable to see Schmidt's dissertation or to examine any of the commentaries at first hand.

² Whitelock, Dorothy, ‘William of Malmesbury on the Works of King Alfred’, Medieval Literature and Civilisation. Studies in Memory of C. N. Carmonsway, ed. Pearsall, D. A. and Waldron, R. A. (London, 1969), p. 89Google Scholar. Whitelock's language is deliberately cautious (‘related to’, ‘normally ascribed to’) for reasons which will be considered shortly. As she has written in another essay, ‘There is a need for a study of the relationship between Alfred's version and the Latin manuscripts of the De Consolatione Philosophiae, especially the glossed manuscripts’ (‘The Prose of Alfred's Reign’, p. 83).

³ I am indebted to Professor Dorothy Whitelock and Dr Michael Lapidge for having read this paper and for having made several suggestions for its improvement.

⁴ Mythological lore is something for which Alfred has been thought particularly indebted to the glosses: see Schepss, p. 149. For previous discussions of the Orpheus metre in Alfred and the commentaries, see Schepss (pp. 152–3), Schmidt (pp. 47–8) and Often (pp. 132–3).

⁵ These have yet to be edited in full. For a list of manuscripts and of the sigla here employed, see the Appendix. My colleague Professor Petrus W. Tax is preparing an edition of the anonymous of St Gallen, and I am indebted to him for help with this commentary. Diane K. Bolton and I are working toward an edition of Remigius. Partial editions of the commentaries are referred to in subsequent notes: those by Naumann (n. 6), Stewart (n. 7), Silk (n. 20) and Bolton (n. 12). Schepss, Schmidt and Otten printed extracts they thought relevant to Alfred from a few manuscripts.

⁶ Schepss compared Alfred's translation to glosses in MSS M1 and Ma. In Johannes Scottus (Munich, 1906) Rand called attention to MS T (see pp. 82–106)Google Scholar.

⁷ Notkers Boethius: Untersuchungen über Quellen und Stil (Strassburg, 1913)Google Scholar. Manuscripts are listed pp. ix–x, extracts printed and discussed pp. 1–23 and 34–59.

⁸ ‘A Commentary by Remigius Autissiodorensis on the De Consolatione Philosophiae of Boethius’, JTS 17 (1916), 22–42.Google Scholar

⁹ ‘Étude critique sur les commentaires de la “Consolation” de Boèce (IX^e-XV^e siecles)’, Archives d'histoire doctrinale et littéraire du Moyen Age 14 (1939), 5–140.Google Scholar

¹⁰ See especially his discussion on pp. 119–21.

¹¹ La Consolation de Philosophie dans la tradition littéraire (Paris, 1967), pp. 241–97.Google Scholar

¹² I find Courcelle's arguments for authorship (esp. pp. 242–4) and date (pp. 254–9) informative but not compelling. The former comprise a number of resemblances – some of them rather general-between the CPh glosses and those in other commentaries thought to be by Remi. However close these resemblances may seem to be, and however acceptable the attribution of the other glosses in question, it is at least a possibility that a different commentator, perhaps even working at a school where Remi taught, used the latter's work as a basis for his own. As to date, this depends chiefly upon the dating of the commentary on Martianus Capella to 901–2 (when, on fairly good evidence, Remi gave a course of lectures on the De Nuptiis), and upon the CPh commentary's being based upon, and thus subsequent to, that on Martianus. As Courcelle notes (pp. 254–5) such arguments are to be made with caution, there being no certain evidence about even the relative chronology of Remi's work. The dating, he concludes, is ‘au moins probable’.

¹³ Courcelle, , La Consolation, pp. 261–3Google Scholar. Since Courcelle's book appeared, Fabio Troncarelli has argued that many ‘Remigian’ glosses can be found in manuscripts which antedate Remi's supposed work on CPh: ‘Per una ricerca sui commend altomedievali al De Consolatione di Boezio’, Miscellanea in Memoria di Giorgio Cencetti (Turin, 1973), pp. 363–80Google Scholar. And Diane K. Bolton, in an important article on the Anglo-Saxon manuscripts of the Remigian commentary, has pointed out that there are more ‘versions’ of the commentary than Courcelle seems to have noticed: ‘The Study of the Consolation of Philosophy in Anglo-Saxon England’, Archives d'histoire doctrinale et littiraire du Moyen Âge 44 (1977), 33–78Google Scholar. On Rémi's sources, see Courcelle, Pierre, ‘La culture antique de Remi d'Auxerre’, Latomus 7 (1948), 247–54Google Scholar, and Bolton, Diane K., ‘Remigian Commentaries on the “Consolation of Philosophy” and their Sources’, Traditio 33 (1977), 381–94.CrossRefGoogle Scholar

¹⁴ Dr Michael Lapidge reminds me that both Remi and Grimbald of St Bertin's were protégés of Archbishop Fulco and may in fact have been at Rheims at the same time. Philip Grierson has argued that Grimbald came to Alfred, most probably from Rheims, between 885–6 and 893–4 (probably c. 887); see ‘Grimbald of St Benin's’, EHR 55 (1940), 546–51Google Scholar. Remi was there perhaps as early as 883; see Lutz, Cora, Remigii Autissiodorensis Commentum in Martianum Capellam, 2 vols. (Leiden, 1962) t, 5Google Scholar. One might conjecture, therefore, that Grimbald brought with him to England a glossed CPh which contained the accumulated tradition of such a school. But, since the following pages argue the lack of close correspondence between Latin commentaries (Remigian or otherwise) and Alfred's translation, it seems more plausible to suggest that such learning as Grimbald and Remi might have shared at Rheims simply helps to account for the rather general similarities which exist between B and Latin glosses.

¹⁵ One such manuscript is L4; and cf. Courcelle, , La Consolation, pp. 259–67.Google Scholar

¹⁶ Alfred ‘habebat ex Sancto Dewi Asserionem quendam, scientia non ignobili instructum, quern Scireburniae fecit episcopum. Hic sensum librorum Boetii De Consolatione planioribus verbis enodavit, quos rex ipse in Anglicam linguam vertit’ (Willelmi Malmesbiriensis Monachi de Gestis Regum Anglorum Libri Quinque, ed. Stubbs, William, Rolls Ser. 90, 2 vols. (London, 1887–1879)Google Scholar, bk 11, ch. 122 (1, 131, my italics)). Practically the same words occur in the Gesta Pontificum: “Asserus, ex Sancto Dewi evocatus, non usquequaque contempnendae scientiae fuit, qui librum Boetii de Consolatione Philosophiae planioribus verbis elucidavit.’ William adds immediately, ‘labore illis diebus necessario, nostris ridiculo’! (Willelmi Malmesbiriensis Monachi de Gestis Pontificum Anglorum Libri Quinque, ed. Hamilton, N. E. S. A., RS 52 (London, 1870), bk n, ch. 80 (p. 177)).Google Scholar

¹⁷ La Consolation, pp. 268–9.

¹⁸ ‘Cum igitur ad eum advenissem in villa regia, quae dicitur Leonaford, honorabiliter ab eo susceptus sum, et cum eo ilia vice octo mensibus curto mansi, in quibus recitavi illi libros quoscunque ille vellet, et quos ad manum haberemus. Nam haec est propria et usitatissima illius consuetudo die noctuque, inter omnia alia mentis et corporis impedimenta, aut per se ipsum libros recitare, aut aliis recitantibus audire’ (Asser's Life of King Alfred, ed. Stevenson, William Henry, repr. with contr. by Whitelock, D. (Oxford, 1959), ch. 81 (p. 67, my italics).Google Scholar

¹⁹ Whitelock, ‘William of Malmesbury on the Works of King Alfred’, pp. 89–90.

²⁰ It is essential to note that all attempts to link Asser with a Latin commentary on CPh begin from the supposition that a commentary was indeed used by Alfred. It was in this context that Courcelle asked, rhetorically, if Asser might have been the author of the glosses preserved in Vatican City, Bibliotheca Apostolica Vaticana, lat. 3565, whose text hand is s. ix¹ (on the glosses see below). Though he judged them too illegible to allow a clear verdict, what he could read of the commentary on III.ix seemed too blemished by astrological speculation to permit attribution to Asser (La Consolation, pp. 269–70). Subsequently Troncarelli pointed out that these glosses are much more legible than Courcelle had thought and edited some of them (‘Per una ricerca’, pp. 371–8); and Bolton reported Troncarelli's suggestion that Asser or one of his associates was their author (‘The Study of the Consolation’, p. 36). After my article was written, Troncarelli's own arguments for Asser's authorship appeared in Tradizioni perdute: la “Consolatio Philosophiae” nell’ alto medioevo (Padua, 1981), pp. 137–51Google Scholar. Troncarelli's description of the manuscript and its glossing hands (pp. 139–41) is extremely valuable. That the manuscript was in England at least by c. 950 seems beyond doubt. Troncarelli goes on to claim that the earliest of the three insular glossing hands is, in fact, s. ix^cx and writes Asser's commentary or a recension thereof. One may wonder, first, about some of Troncarelli's palaeographical conclusions. On the basis of a letter from Bernhard Bischoffin 1970 he says the text of CPh was written at Fleury; he is sure that the first insular glossing hand is ‘Welsh’ and wrote, he suspects, at Glastonbury. But Bischoff's published judgement about the provenance of 3363's text is less dogmatic. It and the text of Vatican Ottobon. lat. 35 (Sedulius, Carmen paschale) come, Bischoff says, from a scriptorium in the Loire region, and both contain ‘Glossen schmaler keltischer Hände, über die ich noch kein definitives Urteil wagen möchte’ (by schmale Bischoff presumably means to exclude the two larger and heavier insular hands of 3363): see ‘Irische Schreiber im Karolingerreich’, Jean Scot Érigine et l'histoire de la philosophic (Paris, 1977), p. 48, n. 3Google Scholar. Troncarelli's dating of the first insular glossing hand, moreover, is heavily influenced by the prevenient conviction that its glosses were available to Alfred. But quite apart from any palaeographical questions, that glosses in a Welsh hand of s. ix^cx must have Asser as their author can be argued only along one of two lines: from the premise that the only author of such glosses must be Asser; or from evidence showing that the glosses are clearly reflected in Alfred's translation. The premise is patently absurd (and Troncarelli does not articulate it); the evidence for influence on Alfred consists of two details in the Orpheus metre (pp. 147–8) neither of which withstands scrutiny. (One concerns Charon – see below, p. 170 and n. 37; the other the Furies/Fates-see below, p. 172 and n. 41.) I have, in any event, edited the glosses of 3363 here (see App., siglum V1- and now cf. Troncarelli, pp. 186–7). Judging by their content, I see no reason for connecting them with Alfred's circle. They show even fewer points of similarity with B than do the glosses of the Remigian commentaries, so much so that Troncarelli is moved to suggest (p. 147) that 3363 must represent an abbreviated version of Asser.

²¹ As already mentioned, Schepss compared Alfred's translation with two manuscripts, M1 and Ma. In an earlier study he had offered a few very brief extracts from what is now known as the St Gallen commentary in G1, M2 and Ma (this last the text of CPh with interlinear and marginal glosses on 4r–57V, written by Froumund and now lost); see ‘Handschriftliche Studien zu Boethius de Consolatione Philosophiae’, Programm der Königlichen Studien-Anstalt Würburg (Würburg, 1881), pp. 35–9Google Scholar. Naumann knew of twelve manuscripts (Notker's Boethius, pp. ix–x), but his extracts came from just two-thirds of these: for R (Remi) from M1, Ma and T; for ‘X’ (St Gallen) from E2, G, G1, M2 and Ma (4r–57v). Stewart's selections of Remi were printed from C2, Ge, Ma and T. Schmidt used, in addition to the extracts published by Naumann and Stewart, three manuscripts: M1, Ma and T. Another partial edition of Remi appeared as an Appendix (pp. 305–43) to Silk, E. T., Saeculi Noni Auctoris in Boetii Consolationem Philosophiae Commentarius, Papers and Monographs of the American Acad. in Rome (Rome, 1935)Google Scholar; Silk's selections are taken from four manuscripts: C2, Ma, T and V5. (Courcelle argues that the ‘ninth-century’ commentary which makes up the body of Silk's edition is not earlier than the twelfth century; see La Consolation, pp. 250–3, 304 and 411–12. According to Courcelle, this commentary is a compilation drawn from Remi and Adalbold of Utrecht and perhaps from William of Conches as well.) Otten used Silk's Appendix, Schmidt's study, and consulted directly E1 (for St Gallen) and T (for Remi). Donaghey used the material published by Schepss (1895), Stewart and Silk. Thus just three manuscripts of the Remigian commentary have been examined by those working on Alfred, just six counting those available in the partial editions. Before Otten only Schepss (1881) and Naumann had printed even extracts of St Gallen; and, so far as I can see, of those studying Alfred's version only Otten has examined even one manuscript of the St Gallen commentary. It is worth mentioning that naturally enough the printed extracts have omitted glosses which do not represent the ‘commentary’ being edited and so have much oversimplified the diversity of the glossing.

²² All quotations of B are taken from King Alfred's Old English Version of Boethius, ed. Sedgefield, Walter J. (Oxford, 1899Google Scholar; repr. Darmstadt, 1968). The Orpheus metre is on pp. 101– and is cited by page and line. I consider here only the prose version. It has long been established that the verse was based on the prose; moreover the Cotton manuscript has a verse rendering of just the opening four lines of the Latin metre (corresponding to 101, 19–22). For the sake of simplicity I have put all quotations of the Old English in romans: both 101, 19–20 (where the prose is found only in the Bodleian manuscript and so is italicized by Sedgefield), and the rest of the section (thereby eliminating Sedgefield's indications of damaged or variant readings). Where a variant from B(odley) might have some impact on the argument, it is given in parentheses. All quotations of Boethius's Latin are from the edition by Bieler, LudwigCorpus Christianorum, Series Latina 94 (Turnhout, 1957)Google Scholar; the Orpheus metre, on pp. 62–4 of this edition, is cited by line, while other passages of CPh are cited by book, section and line. I have used small roman numerals for the metres, arabic numerals for the proses.

²³ So Schepss and Otten. The latter thought that lexical glosses especially helped the translator (see p. 121), but he hastened to add that similar solutions to difficult words could have come independently from similar sources.

²⁴ Alfred's rendering of boni, which comes at the end of the clause in B, will be taken up below, for he expands on it. The only really comparable instance of an extended word-for-word translation in the metre is at lines 42–3, B 103, 4–6.

²⁵ For other instances of paraphrase which simplify the Latin diction and syntax, compare: 40–1 with 103, 3–4; 44–6 with 103, 6–9; and 47–8 with 103, 9–10.

²⁶ I find no such gloss on querens in the commentaries, which offer such synonyms as causans, murmurans (see App.). Scribes of the period generally would have spelled both verbs with an e, though often (but not always) ae was written ξ.

²⁷ Alfred, unlike the commentators, declined to render the religious sense uates (6), and his omission of Orpheus's divine origin is consistent with this (see App.). Lines 22–3 are the most difficult syntactic crux in the metre; the commentaries offer help with them, suggesting that they are an instance of epexegesis.

²⁸ One manuscript does have a gloss which amounts to a similar generalization: at 11 An has ‘cerva species pro genere’. But such suggestions are frequent in older commentaries, such as those by Servius.

²⁹ Thus, as mentioned above, Alfred may very well have read quaerens for querens, an interpretation without parallel in the glosses. At 45 (B 103, 7) Alfred translates dum as [si]ððan (fiforðam); the glosses have donec, ‘until’. Mention has also been made (above, n. 27) of the explanation of the syntax of 22–3 of the Latin, which might have helped Alfred and his scholars if the lines resisted comprehension. Other instances where Alfred's translation runs contrary to the glosses are his treatment of Tityos (39) and arbiter (40), both of whom Alfred, but no commentary, renders as cyning (103, 1 and 4).

³⁰ See the glossary, p. 247. Moreover, in the Latin, God has been equated with Good (III. 12.31–2); there B is slightly more equivocal: ‘God wære) þurh hine selfne good’ (97, 7).

³¹ ‘summum huic bono’ (III. 12.52–3), ‘swa heaum Gode’ (98, 18, Sedgefield's capital); 'summum … bonum’ (54), ‘þæy hehste god’ (98, 21); ‘summum bonum’ (74), ‘þæt hehste god’(100, 13); ‘in summodeo’ (75), ‘on ðæm hehste goode’(100, 14);‘summum … bonum’ (76), ‘þæt hehste good’ (100, 14).

³² Boethius, and his readers, were by no means the first to speak of the fabula of Orpheus. Servius, for instance, at the beginning of his commentary on the fourth Georgic, related how Vergil took out a panegyric to Gallus and substituted there the fabula Orphei: Servii Grammatici in Vergilii Carmina Commentarii, ed. Georg Thilo and Hermann Hagen; vols. 1 and 11, Aeneid, rev. G. Thilo (Leipzig, 1923), vol. III, Eclogues and Georgics (Leipzig, 1887)Google Scholar. The classical authors regularly treated the Orpheus story as a fabula: thus Vergil says as he tells it in Georgics iv, perhibent, that is, ‘people say’ (507). In retelling it, later authors use such verbs as ferebatur, narrat, aestimatur and fingitur (see the passages quoted about Orpheus, below). Such diction and the attitude underlying it account for Alfred's additions of phrases such as ‘ongon mon secgan’ (101, 26) and ‘sædon hi’ (102, 1) and helps to explain his use of sculan as ‘to be supposed to’.

³³ In the following discussion two sorts of sources are considered: those to which Alfred's circle plausibly had direct access (Vergil, Servius, Isidore and perhaps the Metamorphoses) and those which represent more general medieval traditions about Orpheus and the underworld. Knowledge of some of the latter in ninth-century England is problematic. But, since Alfred's circle included Asser, with his Welsh background, and Grimbald, with his continental training, the king's scholars had contact with an intellectual tradition not represented in extant English manuscripts (any more than CPh or its commentaries) until the tenth-century revival. Thus some familiarity with Martianus Capella among ninth-century Welsh scholars is argued by a manuscript like Cambridge, Corpus Christi College 153; the circulation of Macrobius's In Somnium Scipionis in ninth-century France is well attested; and the Culex was part of the Carolingian collection of the pseudo-Vergilian ‘luvenalis Ludi Libellus’. Indirect transmission by such means as oral communication may arguably have played some rôle in preserving and disseminating information, as it seems to have done in the case of the late-ninth-century West-Saxon Orosius; see The Old English Orosius, ed. Bately, Janet, EETS s.s. 6 (London, 1980), Ixi–lxiiGoogle Scholar. Very much more doubtful is knowledge of Seneca's tragedies; reference to their two accounts of Orpheus and the underworld has been included for what they might have contributed to a composite tradition. No claim is made here for direct borrowing by Alfred's translation from such problematic or doubtful sources.

³⁴ A few manuscripts have a gloss on captus (30) which might be thought to approach the sense of onjagnian; see App.

³⁵ Cerberus, as the three-headed dog who guards hell, is a regular feature in the Orpheus story: e.g. Met. x.21–2 and 64–6, Culex 270 and Here. Fur. 783–5.

³⁶ Etymologiarum … Libri XX, ed. Lindsay, W. M. (Oxford, 1911), xi.iii.33.Google Scholar

³⁷ Troncarelli, (Tradizioni perdute, p. 148Google Scholar, n. 46) sees in V1 the traces of a misunderstanding which led to the introduction of Charon. He argues that V1 has two glosses on ‘tergeminus … ianitor’ which referred, or could have been misunderstood as referring, to two separate figures. But the gloss seems clearly to be one, not two: ‘Tergeminus ipse est cerberus et canis est et est tria capita habens ianitor inferni. ipse autem mirabatur de luctu orphei.’ Troncarelli thinks a second gloss begins with ianitor (see his edition of it, p. 187). But there is no indication of this in the manuscript: the word begins in midline with the hand's characteristic i. It does not, moreover, account for the name Charon. Most importantly, it would leave one to wonder why Asser, whose assistance with the translation led Troncarelli to connect the manuscript with Alfred's circle in the first instance, would have permitted such a banal misreading of what we are asked to believe was his own, simple Latin.

³⁸ I have not found such a variant, though the manuscripts sometimes preserve alternative readings and the commentaries frequently discuss them. Scribes were even willing at times to record additions to the text itself, and so it is not impossible that Alfred's manuscript had one for Charon. For example, MS P1 has, beside the lines describing Ixion, ‘Nee prae pectore nititur / saxum volvere sisyphus’. (Seneca's catalogue of famous sufferers includes Sisyphus in Hercules Oetaeus, where Orpheus's descent is also recounted (1081–2).) Were the reading portitor or the misinformation that Charon was a three-headed door-keeper in the underworld widely current, one might very well have found a gloss such as ‘vel aliter. tergeminus portitor, id est Charon qui etiam tria capita fertur habere'; but 1 have seen no such gloss.

³⁹ Charon is alluded to in both Vergil's and Ovid's accounts of the Orpheus story. After Orpheus has looked back and lost Eurydice a second time, the portitor will not permit him to pass into hell again (Geo. iv.502–3 and Met. x.72–3). Describing Orpheus's descent, Vergil told how Cocytus and Styx restrain the shades (Ceo. iv.478–82), and in Culex we are told that Orpheus feared neither Cerberus nor the burning waters of Phlegethon (270–2); thus the necessity for the boatman is clearly implied. Alfred's calling him a geatweard (102, 16) may reflect an inappropriate etymology of Charon as portitor (in contrast, commenting on Aen. vi. 298, Servius explained: ‘PORTITOR… qui portat’). But in Georgia iv and Metamorphoses x there is no explicit reference to Orpheus's going by boat; the portitor simply prohibits the visitor from returning (transire), and so the translators in the Loeb volumes, for instance, render the epithet ‘warden’ and ‘keeper’, respectively. It is just possible that Alfred similarly adapted the function of Charon to a context where the rivers of the pagan underworld were neither mentioned nor appropriate. A hint that he was aware of Charon's real function appears at the end of the description: ‘7 hine gesundne eft ðonan brohte’ (102, 19–20).

⁴⁰ In Hercules Furens Hercules encounters Charon at the Styx, a senex of repellent appearance (764–72). In Aeneid vi Aeneas and the Sibyl meet the wretched old man just before they come upon Cerberus: ‘Portitor has horrendus aquas et flumina servat / terribili squalore Charon, cui plurima mento canities inculta iacet …’ (298–300, my italics).

⁴¹ Troncarelli's chief piece of evidence that V1 represents Alfred's source and so is Asser's commentary is the claim that the confusion of Fates and Furies derives from the gloss in V1, which strings together the names of Fates, and Furies, (Tradizioni perdute, pp. 147–8)Google Scholar. But Troncarelli does not mention Isidore's conflation of the two groups and underestimates the currency of the confusion in the Latin glosses: it occurs in MSS G, N, L1, P7, P10 and V5 as well as in V1 (see App.). Of these he mentions only G (St Gallen 844) and P7 (BN 15090). I do not find it, as Troncarelli purports to, in C1 (Cambridge, Gonville and Caius College 309) or in O1 (Oxford, Corpus Christi College 59), and so the notion of ‘un errore ben comprensibile in ambiente inglese’ (Tradizioni perdute, p. 148, n. 45) seems to be mistaken.

⁴² Geo. iv.469: Orpheus, ‘ingressus, manisque adiit regemque tremendum’; cf. Aen. vi.106, 252, 396 etc. Ovid has Orpheus approach Pluto, umbrarum dominum (Met. x.16), ‘qui regit ima’ (47). At 28 of the Latin metre, umbrarum dominos was probably meant to refer to Pluto and Proserpina, who are regularly present in the classical accounts of the Orpheus descent.

⁴³ Servius goes on to quote Horace, Satires 1. i. 69–70. Horace has been discussing people deceived by cupido falsa (61) and names Tantalus as an example, concerning whose plight he adds the two lines: ‘quid rides? mutato nomine de te / fabula narratur.’

⁴⁴ Though Vergil is generally held to have been much more widely known than Ovid in the early Middle Ages, Servius might have drawn attention to Ovid's account by his reference to it when commenting on Geo. iv. 522. The accounts in Culex (268–95) and in Seneca (Herc. Fur. 569–91 and Herc. Oet. 1031–99) have been referred to in the preceding notes. As Bieler observed, Herc. Fur. 582 is quoted by Boethius at 40–1 of the metre; although some of the Remigian commentaries identify 8–9 as a quotation from Vergil and quote Geo. iv.467 on Taenara (26), no commentary identifies the Seneca passage.

⁴⁵ ‘si potuit manes nititur exemplis quae inferiora sunt per comparationem, ut [Aeneas] videatur iustius velle descendere: nam Orpheus revocare est conatus uxorem … Orpheus autem voluit quibusdam carminibus reducere animam coniugis: quod quia implere non potuit, a poetis fingitur receptam iam coniugem perdidisse …’

⁴⁶ De Nuptiis Philologiae et Mercurii, ed. Dick, Adolf, with addenda by J. Préaux (Stuttgart, 1970), pp. 480Google Scholar, line 14–482, line 5. There are numerous other references to Orpheus in the De Nuptiis. On the manuscripts and medieval circulation of Martianus Capella, see Leonardi, C., ‘I codici di Marziano Capella’, Aevum 33 (1959), 443–89Google Scholar, and 34 (1960), 1–99 and 411–524.

⁴⁷ Commentarii in Statii Thebaida, ed. Jahnke, Richard (Leipzig, 1898), p. 382Google Scholar. For a preliminary assessment of the surviving manuscripts of this commentary, see Clogan, P. M., ‘The Manuscripts of Lactantius Placidus’ Commentary on the Thebaid‘, Scriptorium 22 (1968), 87–91.Google Scholar

⁴⁸ Chronicorum Libri, ed. Migne, Patrologia Latina 27, col. 310.Google Scholar

⁴⁹ De Civitate Dei xviii.14; ed. Bernard Dombart and Alphonse Kalb, CCSL 48 (Turnhout, 1955), 605.

⁵⁰ Chronica, ed. Mommsen, Theodor, Monumenta Germaniae Historica, Auct. Antiq. 11 (Berlin, 1894), 121.Google Scholar

⁵¹ Ch. LXVI, ‘De sex huius saeculi aetatibus’, ed. Theodor Mommsen, MGH, Auct. Antiq. 13 (Berlin, 1898), 259. Cf. Tiro, Prosper, Epitoma Cbronicon, ed. Theodor, Mommsen, MGH, Auct. Antiq. 9 (Berlin, 1892), 389Google Scholar, and lunior, Isidorus, Chronica Maiora, MGH, Auct. Antiq. 11, 437.Google Scholar

⁵² Such brief references can be found in the works of Ausonius (fourth-century), Sidonius Apollinaris (fifth-), Ennodius (fifth-sixth-), Cassiodorus and Venantius Fortunatus (sixth), Alcuin and Theodulf of Orleans (eighth-ninth-), Sedulius Scottus (ninth-) and Eugenius Vulgaris (tenth-).

⁵³ Epistula 117, PL 22, col. 957.

⁵⁴ Commentarii in Somnium Scipionis, ed. Willis, James (Leipzig, 1970), p. 105, lines 13–16.Google Scholar

⁵⁵ Etym. III.xxii.9. Similar references to the power of Orpheus's song can be found in: Dracontius, , Romulea, ‘Praefatio’ 1–11 (ed. Vollmer, Friedrich, MGH, Auct. Antiq. 14 (Berlin, 1905), 1933)Google Scholar; Venantius Fortunatus, ‘Ad Gorgonem’ 1–10 (ed. Friedrich Leo, MGH, Auct. Antiq. 4.1 (Berlin, 1881), 153); Engelmodus of Soissons (d. 865 – ed. Ludwig Traube, MGH, Poet. Lat. Aevi Car. 3 (Berlin, 1896), 37, lines 87–8); the anonymous of Salzburg (c. 855–9), Carmina VII. 10–11 (ed. Dümmler, Ernst, MGH, Poet. Lat. Aevi Car. 2 (Berlin, 1884), 642)Google Scholar; Scottus, Sedulius, Carmina 11.vii. 11–12 (MGH, Poet. Lat Aevi Car. 3 (Berlin, 1896), 173)Google Scholar; and Abbo of St Germain (d. 896), ‘Praefatio’ to Bella Parisiacae Urbis (ed. von Winterfeld, Paul, MGH, Poet. Lat. Aevi Car. 4.1 (Berlin, 1899), 77, lines 20–3)Google Scholar. There are also accounts of the Orpheus and Eurydice story in the early mythographers, but these show more relationship to the versions of the Remigian commentary than to Alfred's treatment: Scriptores Rerum Mythicarum Latini Tres, ed. Bode, Georg (Celle, 1834Google Scholar; repr. Darmstadt, 1968), ‘First Vatican Mythographer’, ch. 76, pp. 26–7 (the entry is reproduced by the gloss in MSS A and C4 – see App.) and ‘Second Vatican Mythographer’, ch. 44, p. 90 (for Courcelle's argument that this mythographer was Remigius see La Consolation, pp. 244–8); Fulgentius, , Mitologiarum Libri Tres, ed. Helm, Richard (Leipzig, 1898), III.x, pp. 77–9Google Scholar (in part reproduced by the gloss of MS C4 – see App.); Hyginus Astronomus, ed. Serra, Fabricio (Pisa, 1976), pp. 39–41Google Scholar; and Hyginus, , Fabulae, ed. Rose, H. J. (Leiden, 1953)Google Scholar-Rose prints an extract from the edition of Micyllus (Basel, 1535) which seems to have been taken over from Fulgentius (p. 115).

⁵⁶ A further point must be noted briefly. Though Orpheus sings laments after his first loss of Eurydice in both Georgics iv (464–6) and Metamorphoses x (11–12), these two accounts are concerned to stress the harper's songs of grief after the fateful backward look (Geo. iv. 507–20 and Met. x. 72–90 and xi. 1–47). Boethius and his English translator write with a different purpose, emphasizing Orpheus's powerful grief and song before his descent. Thus details found after the descent in the classical versions, or in detached descriptions of his powers, appear prior to the descent in Boethius and Alfred.

⁵⁷ See the passages from Jerome and Isidore quoted in the text, above. In Georgics iv Orpheus charms the trees after losing Eurydice the second time (510), and the same is true in Ovid's account (Met. x. 86 ff. and xi. I ); cf. Here. Fur. 572 and Here. Oet. 1043–4; in Culex the silvae are moved before the descent (280–2). Claudian, De Raptu Proserpinae 11, ‘Praefatio’, describes Orpheus's power over the trees (21–4), and Quintilian speaks of it in the Institutio Oratorio (i.x.9). It is mentioned in the account by Martianus Capella (De Nuptiis, ed. Dick, , p. 480Google Scholar, lines 22–3) and in the allusions by Venantius Fortunatus, Engelmodus of Soissons, the Salzburg anonymous and Abbo of St Germain referred to above, n. 55.

⁵⁸ See, once again, the passages from Jerome and Isidore. It is not mentioned by Vergil, but it appears in Ovid, Met. xi. 2, ‘et saxa sequentia ducit’. Cf. Here. Fur. 572, Institutio Oratoria i.x. 9 and De Raptu Proserpinae 11, ‘Praefatio’, 19.

⁵⁹ Orpheus's power over animals is referred to by Jerome, Macrobius and Isidore. It appears in Geo. iv. 510 and several times in the Metamorphoses, where the singer is once said to be surrounded by a gathering of beasts and birds: ‘Tale nemus vates attraxerat inque ferarum/concilio medius turba volucrumque sedebat’ (x. 143–4). The same detail is in Culex: ‘et turba ferarum/blanda voce sequax regionem insederat Orphei’ (278–9). Cf. Hercules Furens (‘ad cuius sonitum constiterant ferae’, 574), Here. Oet. 1055–60, De Raptu Proserpinae ii, ‘Praefatio’, 25–9, Institutio Oratorio i. X. 9, Capella, Martianus, De Nuptiis, ed. Dick, , p. 480Google Scholar, line 21, and the allusions in Dracontius, Venantius Fortunatus and Abbo of St Germain referred to above, n. 55.

⁶⁰ Above, n. 56.

⁶¹ There is another passage in Vergil which, along with Servius's gloss, could help to place Orpheus's song in the Thracian mountains. Ec. vi. 30 reads, ‘nec tantum Rhodope miratur et Ismarus Orphea’, and Servius comments, ‘RHODOPE ET ISMARUS montes Thraciae, in quibus Orpheus consueverat canere’.

⁶² These sixteen items are as follows (listed in order by the line number of the Latin metre, with the lemmata of CPh, the key word of the gloss and the corresponding words of B): i boni summi dei/boni, hehstan godes. 4 Soluere abicere, aweorpan. 6 uates citharista, hearpere. 6 THREICIUS a thracia, Dracia. 11 CERUA species pro genere, ne nan neat. 16 CUNCTA … saxa, stanas. 28 DOMINOS deos, godu. 29 TERGEMINUS tria capita, prio heafdu. 30 CAPTUS delectatus, onfægnian. 32 DEAE parce, Parcas. 34 IXIONIUM rex lapitharum, Levita cyning. 34 (IXION) religatus ad rotam, wheol Se Ixion wæδ to gebunden. 34–5 NON PRAECIPITAT stetit, osstod. 37 TANTALUS rex, se cyning. 37 TANTALUS avaritia, gifernesse. 52 ff. (closing moral; see App.).

⁶³ P7 lacks, e.g., abicere/aweorpan, citharista/hearpere, that Ixion was king of the Lapiths and that Tantalus was a king. C₄ lacks most notably that Orpheus charmed stones and that the deae are the Parcae.

⁶⁴ Vi: gives Cerberus three heads, confuses Furies and Fates, has Ixion ‘confixus in rota’ and says Tantalus was a king. P11: Threicius a thracia, Cerberus with three heads and Tantalus nimium avarus. On: Orpheus a harper, Threicius de thracia.

⁶⁵ See above, n. 29.

⁶⁶ See esp. Otten, pp. 125–8 and 155–7.

⁶⁷ The following discussion is based chiefly on Otten, pp. 119–57, but I have also considered the passages quoted by Donaghey, who took up Schepss's parallels and added published glosses to those Schepss had used. Schmidt, as Otten notes (p. 125), often adduced instances which are much less convincing.

⁶⁸ The gloss Otten quotes (p. 151) is as follows (T, 122v): ‘Homines non tenentur ea lege qua conditi sunt, quia, cum omnia teneant propriam legem, homines sua adinventione semper faciunt.’ For a similar instance, see Otten, p. 123, where again T is said to explain B. Of the four Old English phrases italicized in Otten's quotation, three have their direct equivalent in CPh; the fourth is an expansion by Alfred, perfectly explicable from context, with a rather inexact equivalent in the gloss: ‘ærδæm pe hit gefremed weorse’ (128, 12), qualiter perficiatur (T, 154 r ).

⁶⁹ ‘Ita et homo, qui ad veram tendit felicitatem et ad summum bonum, quod est deus, debet prius eradicare truncos vitiorum et liberare animum suum a superfluis curis’ (Otten, p. 122; T, 133v).

⁷⁰ Otten, p. 152.

⁷¹ Ibid.

⁷² Ibid. p. 140.

⁷³ Ibid. p. 121.

⁷⁴ Ibid. pp. 134–41.

⁷⁵ Ibid. p. 126.

⁷⁶ The anonymous West Saxon translator says that Nero ‘ongon wyrcan scopleos be δæm bryne’ (The Old English Orosius, ed. Bately, , p. 137, lines 24–5).Google Scholar The Latin original is even closer to Alfred, supplying the explicit connection with the Troy story: Nero ‘tragico habitu Iliadem decantabat’ (quoted Ibid. p. 324).

⁷⁷ Otten, p. 127.

⁷⁸ Commentarii in Somnium Scipionis, i.x. 16 (ed. Willisxy, , pp. 44–5).Google Scholar Macrobius states, as T does not, that the sword is nacod (‘gladium vagina raptum’) and that it is suspended by a thin thread (‘de filo tenui’), and he has the tyrant address his familiar directly: ‘talis est … vita quam beatam putebas: sic semper mortem nobis imminentem videmus. aestima quando esse felix poterit qui timere non desinit.’ Cf. Alfred: ‘Eala, hwæt, δæt bis geδælig mon pe him ealne weg ne hangas nacod sweord ofer δæm heafde be smale præde, swa swa me git symle dyde’ (65, 28–66, 1).

⁷⁹ Otten, p. 122.

⁸⁰ E1 reads, ‘An tu i. h. u. s. in mundum. Scenam. umbram. calamitatem. Nunc primum. Adverbium. Subitus i. subitaneus. ut novum aliquid subito tangat quod antea aliis non contigerit. Hospesque v. quia fortunam ut alii solus fore despicis s. quod novum tibi quasi hospiti fortuna impendit. quod non aliis impendent ante’ (p. 117). T has, ‘vitae scenam fortune foeditatem. subitus subitaneus. Quia fortuna ut alii ferre despicis.’ Then T gives a long marginal gloss on scenam: ‘grece umbra dicitur. Unde scena dicitur umbraculum sub quo fortuna agitur. Hie autem scenam vocat hunc mundum in quo fortuna ludos suos agit. quosdam humilians quosdam exaltans.’ The gloss continues with a discussion of the stage, ‘versilis et ductilis quae ideo vertebatur. ut quemadmodum diverse tragedie in ea recitabantur …’ (127r). Mi also has this gloss on the reversible platform, at the end of which come the lines Otten quotes.

⁸¹ In this connection it is worth noting what Janet Bately has observed with reference to the only known extant commentary on the Latin Orosius: the fragment preserved (on 1.i.13 – 11.xii. 5) ‘has no more than a handful of details in common with’ the Old English version (The Old English Orosius, p. lxii, n. 4).

⁸² Bately's researches into the additions to the Old English Orosius have suggested that the translator of that text had at least second-hand access to a fairly broad range of classical and patristic sources; see esp. ‘The Classical Additions in the Old English Orosius’, England before the Conquest. Studies in Primary Sources presented to Dorothy Whitelock, ed. Clemoes, Peter and Hughes, Kathleen (Cambridge, 1971), pp. 237–51Google Scholar, and The Old English Orosius, pp. lx–xiii.

⁸³ Anglo-Saxon England, 3rd ed. (Oxford, 1971), p. 270.

⁸⁴ Asser's Life of King Alfred, ch. 77 (ed. Stevenson, , p. 62).Google Scholar

^{page 187 note 1} For this study I have examined all the Remigian manuscripts known to me. The thirty-two listed here have at least some glosses for 111.xii and I have collated them all. Seven other manuscripts contain some Remigian glosses but have no comments on this metre. (One of them, L, has some non-Remigian comments which are reported – see below, under ‘Other early manuscripts with glosses’; another, Cambridge, Corpus Christi College 214, has interlinear Old English glosses on 111.xii, but these shed no light on B.) Six manuscripts record only the commentary on 111.ix. (One, P4, has a few glosses on 111.xii which are reported; see below, under ‘Other early manuscripts with glosses’.) Courcelle's list of Remigian manuscripts (La Consolation, pp. 405–6) has been of enormous help; but, as is not surprising in a work of such scope, it is neither complete nor fully accurate.

^{page 187 note 2} This list includes eleven manuscripts of known English provenance (A, C₂, C₃, C₄, Es, Ge, O₁, O₂, P, P6 and P9); all but O₁ are discussed by Bolton, ‘The Study of the Consolation’, pp. 51–8. Four others she lists are among the fragmentary copies of the commentary and have no glosses for 111.xii.

^{page 187 note 3} This manuscript was lost during World War II. It contained a text CPh accompanied by the St Gallen commentary, copied by Froumund (4r–5 7v), and a continuous copy of Remi written by Froumund's students. I am indebted to Professor E. T. Silk for allowing me to use his photographs of the Remigian portion.

^{page 188 note 4} Courcelle (La Consolation, pp. 403–4) lists four ‘complete’ manuscripts of this commentary and ten which contain ‘fragments’ of it. I have reported three of the complete copies here (E1, G1 and N) but have not included Paris, BN lat. 13953 (which is, as Courcelle says, an abbreviation of St Gallen). L₄ also contains a fairly full version of the commentary, as does W1. Of the fragmentary copies Courcelle lists, I have collated E, G, M₂ and P₇. (The last has some St Gallen glosses mixed with what is chiefly a version of the Remigian commentary.) Einsiedeln 302 lacks the folios containing 111.xii. Four other manuscripts were lost in World War II: Bonn 175, Chartres 59, Maihingen and Metz 377. (Since these seem to have been compilations of St Gallen and Remi their glosses are probably well represented by the extant manuscripts.) M₄, which also contains some St Gallen glosses mixed with Remigian comments, has been collated. There remains one manuscript of St Gallen which I have not seen, Einsiedeln 322 (according to Courcelle, ‘quelques glosses seulement’); I have assumed that it is adequately represented by the manuscripts which I have seen.

^{page 188 note 5} The glosses collated for 111.xii come from the text with commentary; the manuscript also contains a continuous copy of Remi, but this ends with the glosses on 111.ix.

^{page 189 note 6} This manuscript contains a continuous copy of Remi on 111.ix inserted into a glossed text of CPh; the glosses reported here come from the latter.

^{page 189 note 7} Reporting variants would render this Appendix intolerably complex. But it must be pointed out that many of the glosses, especially those consisting of short phrases, show considerable variation. One example will suffice. For the gloss on 1–2, BONI FONTEM, a full record would read:

BONI FONTEM: bonitatis, vel sapientiam summi dei (P₉ A P V₃ P6 Ge C₄)

— vel: om Ge C₄sapientiam: sapientiae P₉ A dei: boni P₉ P V₃

I have judged that for the present purposes my simplification is acceptable.