¹ Bliss, A. J., The Metre of Beowulf, (1958; rev. ed., Oxford, 1967), §106.Google Scholar

² Pope, J. C., The Rhythm of Beowulf (1942; rev. ed., New Haven, Connecticut, 1966), p. 99.Google Scholar

³ ‘Verbal Stress in Old English Verse’, Anglia 93 (1975), 307–34, at 328.Google Scholar

⁴ The Later Genesis, ed. B. J. Timmer (Oxford, 1948), pp. 51–4.Google Scholar Timmer's metrical analysis is too vague to be of much benefit. The suggestion that the translator was continental Saxon rather than English was first made by Ten, Brink, Geschichte der englischen Literatur (Strassburg, 1877), p. 106, n. 1.Google Scholar

⁵ Capek, M. J., ‘The Nationality of a Translator: some Notes on the Syntax of Genesis B’, Neophilologus 55 (1971), 89–96.CrossRefGoogle Scholar

⁶ Bethel, A. C. P., ‘A Reconsideration of Various Aspects of Genesis B: Editions, Historical Background, Theology, Vocabulary, Metre and Syntax’ (unpubl. PhD dissertation, London, Univ., 1977)Google Scholar. For my comments on aspects of her work, see below, nn. 17, 110 and 113. For an early study, see Friedrich, Graz, Die Metrik der sog. Caedmonschen Dicbtungen, Studien zum germanischen Alliterationsvers 3 (Weimar, 1894)Google Scholar. His work is, however, little more than a listing of verse types following Kaluza's development of the Vierbebungstheorie (ibid. pp. 97–108) and of no use for my purposes.

⁷ Kuhn, H., ‘Zur Wortstellung und -betonung im Altgermanischen’, BGDSL 57 (1933), 1–109.Google Scholar

⁸ E.g., Creed, R. P., ‘A New Approach to the Rhythm of Beowulf’, PMLA 81 (1966), 23–33CrossRefGoogle ScholarStevick, R. D., ‘The Meter of The Dream of the Rood’, NM 68 (1967), 149–68Google Scholar and Hieatt, C. B., ‘A New Theory of Triple Rhythm in the Hypermetric Lines of Old English Verse’, MP 67 (1969), 1–8.Google Scholar

⁹ See Bliss, Metre, §§1–8, 41–5 and 88–98.

¹⁰ A debt of gratitude is due to the late Professor Alan Bliss and Dr Peter Lucas of University College, Dublin, who read this paper in typescript and saved me from many errors. Dr Lucas, in addition, very kindly allowed me to read his forthcoming article on enclisis (see n. 90) in typescript, and I benefited greatly from his discussion. Many thanks also to Mary Murphy, my wife, who helped me in the thankless task of checking the verse citations and the statistical tables.

¹¹ Unless otherwise indicated all quotations are from the text of Gen B in The Junius Manuscript, ed. G. P. Krapp, Anglo-Saxon Poetic Records 1 (New York, 1931).Google Scholar All other Old English citations are from Krapp, and Dobbie, E. V. K., ASPR (New York, 1931–1953).Google Scholar Short titles of texts follow those in ASE 4 (1975), 213–15.Google Scholar Quotations from the Old Saxon Genesis (indicated by OS) are from E. Schröder's edition in his Anhang to Heliand, ed. E. Sievers, Germanische Handbibliothek 4 (Halle, 1935). The text of Heliand (Hel) is that of the Cotton manuscript in Sievers's edition, except in those instances where the text is corrupt, in which instances that of the Munich manuscript (Hel M) is quoted.

¹² Similarly 805b: OS 14b; and 808b: OS ₁₇b.

¹³ Genesis, ed. Timmer, pp. 49–50.Google Scholar Capek (‘Nationality’, p. 94, n. 8) is probably right in suggesting that gisuuerek (OS 16b) is the object of dribit, which has uuind (OS 15a) as subject, but the syntax is ambiguous and may have appeared so to the translator, who treated the verb as intransitive in his translation.

¹⁴ Similarly 802b: OS 12; and probably a number of instances where the Old Saxon text is corrupt (8_13a: OS _22a; and 8₁₄b: OS ₂₃b).

¹⁵ With eight syllables in the first thesis: 507b; with seven syllables: 368b, 499b, 558b, 619b and 679b. On similar verses in Old Saxon, see Kauffmann, F., ‘Die Rhythmik des Heliand’, BGDSL 12 (1887), 283–359, at 316–18 and 328–9.Google Scholar

¹⁶ Timmer, (Genesis, pp. 49–50)Google Scholar was surely wrong in claiming that sc(i)ene is used only in connection with persons. Such is the case in most instances, but in two it refers to bodily form (265b and 503a) and, more significantly, to the Tree of Knowledge in 467b, where the meaning is ‘beautiful’ rather than ‘bright’. For similar use of sc(i)ene with inanimates, see Christ B 695, Christ C 1147, Pa 26, And 766 and Jud 316.

¹⁷ Substitution of the uninflected acc. in belle, as in 331a (probably due to Old Saxon influence), gives a normal verse. However, the translator avoids uninflected forms in all other instances, and the verse is best left as it stands. Bethel's scansion (‘Aspects of Gen B’, p. 176) of 792b, stressing both Gesyhst and sweartan, is to be rejected. Her view (ibid. p. 199) that double alliteration is found in several b verses is unconvincing, as all instances can be shown to be accidental alliteration; see below, pp. 84 and 87.

¹⁸ Timmer, (Genesis, p. 49)Google Scholar was wrong in supposing that there is no Old English reflex of OS tuom. See Christ C 1211.

¹⁹ ibid. p. 50.

²⁰ E.g. giongorscipe (249a), betesprace (263b), hearmscearu (432a) and wradmod (547a and 815a). For a discussion of bapax legomena in Gen B, see Timmer, ibid. pp. 27–33, but also Capek's discussion (‘Nationality’, p. 90) and his revised list on p. 95, nn. 19 and 20. A number of compounds with bealo as first element are attested in Old English; in one instance (bitre bealodæde, Res 20a) the sense is identical to OS balouuerek. See also Res 35 and Christ C 1301.

²¹ Bliss (Metre, §§84–5) uses the term ‘remainders’ for those verses in Beo which cannot be accommodated within the five main types. For remainders in Gen B, see below, pp. 100–1.

²² Genesis. p. 54.

²³ Kuhn (‘Wortstellung’, pp. 4–5) calls the three classes respectively ‘Satzteile’, ‘Satzpartikeln’ and ‘Satzteilpartikeln’. See also Slay, D., ‘Some Aspects of the Technique of Composition of Old English verse’, TPS (1952), 1–14.Google Scholar In this section (pp. 72–88) and in the following section on Kuhn's Law (pp. 88–95) the alliterating letter(s) appears in bold in verses with single alliteration in order to avoid ambiguity. In verses with double alliteration, however, there is no necessity to indicate this.

²⁴ These positions, thus, correspond to both of Kuhn's normal positions, i.e. before and after the first stressed element respectively.

²⁵ Metre, §10.

²⁶ ibid. §§19–23. On verbal alliteration in Gen B, see below, pp. 80–8.

²⁷ This reading for 491b was proposed by Graz, (Metrik der Caedmonschen Dichtungen, p. 99)Google Scholar without discussion. All editors, following the manuscript pointing, print: and wand him pa ymbutan pone deaoes beam (491b–2a). Pointing in the Junius manuscript, however, is not infallible (see also below, n. 125), and I am grateful to Dr Peter Lucas for pointing out to me that ymbutan is a preposition governing pone…beam and clearly belongs to 492a.

²⁸ See Willard, R. and Clemons, E. D., ‘Bliss's Light Verses in the Beowulf’, JEGP 66 (1967), 230–44.Google Scholar I cannot, however, agree with them in assigning stress to all finite verbs in Pos. 1, regardless of whether alliteration is present or not. Such treatment leads to some very odd results in the scansion, such as b verses of Type 1D* and Type 2A verses with anacrusis, although the evidence against such verses is overwhelming. Old English verse illustrates a situation whereby a verb may or may not be stressed, and this probably represents a mixture of archaic and new practices, as suggested by Lehmann, W. P. and Tabusa, T., The Alliterations of the Beowulf (Austin, 1958), pp. 5–8.Google Scholar

²⁹ With two exceptions (321b and 748b), alliteration occurs on the first stressed element in Gen B. On lack of alliteration in 238, see below, p. 100 and n. 129. In 249b, MS fyligan (so also Krapp and Timmer in their editions) should be emended to ful(l)gan ‘perform, accomplish’, as proposed by earlier editors; for a full discussion, see The Junius Manuscript, ed. Krapp, p. 164, note on 249.

³⁰ Other instances: (ænig) 409a and 503b;(eall) 314a,488b, 604a, 670b and 759b; and (sum)432a and 636 (2x).

³¹ Other instances: (eall) 583a and (feala) 322b, 457b and 788b.

³² other instances: (eall) 337b, 351b, 550a, 756b and 804b and (monig) 634a, 728a and 738b.

³³ other instances: (eall) 306b, 432b, 565b and 674b and (fela) 445b, 479a, 579b and 708b.

³⁴ Perhaps surprisingly, there are only three certain instances of double hypermetric verse; for details, see below, App., Table 12.

³⁵ See above, n. 29. For a full discussion of indefinite adjectives of quantity, see Fakundiny, L., ‘The Art of Old English Verse Composition’, RES ns 21 (1970), 129–42 and 257–66.Google Scholar

³⁶ Other instances of ful are 634b, 705a and 728a.

³⁷ In Beo swa and ful are generally restricted to Pos. 1, but there are a few instances of swa as proclitic in Pos. 2, e.g. pæt næfre Grendel swa fela (Beo 591a).

³⁸ In 748b eft stands in Pos. 2 and must be stressed, and postponed alliteration in the verse must be accepted; see below, p. 101.

³⁹ The auxiliary verb late has non-functional alliteration and is unstressed; see below, pp. 84‣6.

⁴⁰ Likewise beonan (666b), innan (723b), inne (436b), nider (343a), ufan (375a and 513a) and up (415a and 497a).

⁴¹ 374b, 436b, 525b, 735b, 751b, 755b, 806a, 811b and 817b.

⁴² See ‘Wortstellung’, p. 9.

⁴³ For further examples, see Bliss, Metre, §19.

⁴⁴ In GenB double alliteration in hypermetric verses is found on the first two stressed elements, never on the first and third.

⁴⁵ Both verses are, nevertheless, remainders; see below, p. 108.

⁴⁶ ‘Wortstellung’, pp. 12–13.

⁴⁷ Confirmation that sniomor is stressed at the expense of no is afforded by the alliteration.

⁴⁸ Both verses are Type 2A2 Compare similar adverb combinations in Beo 853a, 1762b and 3044b.

⁴⁹ That nu receives primary stress is confirmed by early Middle English nouthe etc. See Middle English Dictionary, ed. H. Kurath and S. M. Kuhn (Ann Arbor, 1952–), s.v.Google Scholar

⁵⁰ Metre, §13.

⁵¹ This group would be very much larger if such verses as ellen fremedon (Beo 3b) were included. Bliss (ibid. §14) includes only those verses standing at the beginning of the verse which are preceded by a stressed element in the verse clause, and I have followed his practice. It is noteworthy that this is the only group in which the verb stands in Pos. 2; not surprisingly, it is the only group, besides group 9, in which the verb alliterates in all instances.

⁵² 459b, 504a, 551a, 606a, 607a, 619a, 713a, 749a and 777a. In 504a and 606a the negative particle ne precedes the verb; nevertheless, these verses may be included in this group.

⁵³ 276a, 445a, 614a, 724a, 771a and 846a.

⁵⁴ 241b, 312a, 372a, 432b, 447b, 448a, 564a, 617a, 765b, 809a, 811a, 841a and 842a.

⁵⁵ 235b, 242b, 263b, 272b, 320b, 322a, 332b, 443b, 449a, 686b, 725b, 808a, 842b and 848b. I agree with Stanley (‘Verbal Stress’, p. 330, n. 54) in taking bidan (842b) as preterite pl., like the preceding verbs, rather than as infinitive. However it is regarded, bidan is unstressed.

⁵⁶ 248b, 386a, 430a, 453a, 479a, 540a, 554b, 699a, 849a and 850a.

⁵⁷ 255b, 260b, 280b, 294b, 297b, 315a, 338a, 388b, 401b, 452b, 489b, 494a, 519b, 520b, 564b, 603a, 626a, 740b, 816b and 821a.

⁵⁸ 236a, 236b, 237a, 263a, 268b, 284a, 306b, 353b, 382b, 406a, 446b, 447a, 517b, 547a, 579a, 581a, 684a, 684b, 687a, 688a, 762b, 830a and 840b.

⁵⁹ 235a, 300b, 313a, 342b, 343b, 368b, 371a, 445b, 493a, 493b, 507b, 518b, 543b, 614b, 676b, 704a and 802b.

⁶⁰ In Beo there are never more than two particles between the finite verb and the first stressed element, whereas in GenB there are four verses with three particles in this position. However, I have included all instances under the same heading for the convenience of abiding by Bliss's classification.

⁶¹ 240a, 259a, 389a, 405b, 491a, 491b, 575b, 588a, 621a, 647a and 792b.

⁶² 304a, 619b, 669a and 708a.

⁶³ 265a, 274b, 276b, 344a, 500a, 500b, 503b, 549b and 581b.

⁶⁴ 405a, 433a, 435a, 483a, 563b, 635a and 797a.

⁶⁵ 432b, 519a, 662a, 724b and 817b.

⁶⁶ The figures for Beo are those of Bliss, Metre, §§14–25.

⁶⁷ Metre, §26. For a more radical view of non-functional alliteration in finite verbs, see Kendall, C. B., ‘The Metrical Grammar of Beowulf: Displacement’, Speculum 58 (1983), 1–30.CrossRefGoogle Scholar

⁶⁸ Metre, §§20 and 22.

⁶⁹ See above, pp. 73–4.

⁷⁰ (Group 1) 459b and 606a; (group 3) 241b, 432b, 617a and 765b; (group 4) 248b, 386a, 453a, 540a, 554b, 699a, 849a and 850a; (group 5) 268b, 306b, 353b, 382b, 446b, 517b, 684a, 684b, 762b and 840b; (group 6) 491b and 575b; (group 8) 433a, 435a, 483a, 563b, 635a and 797a; and (group 9) 662a, 724b and 817b. There is a very useful discussion of the question of verbal stress in Andreas by Stevens, K. (‘Some Aspects of the Metre of the Old English Andreas’, Proc. of the R. Irish Acad. 81C (1981), 1–27, at 2–15)Google Scholar. His conclusions concur largely with my own.

⁷¹ So also onwendan (400a), which is probably pres. subj. pl., with confusion of endings similar to bidan in 842b (see above, n. 55). Compare the two non-alliterating verbs in the same group (gelyfe 679b and wende 712a), both of which introduce a subordinate clause.

⁷² See Bliss, Metre, §§12 and 27–9. Habban, onginnan and witan also occur as lexical verbs. As a lexical verb babban is never stressed, and witan is stressed only once (386a) in Pos. 1.

⁷³ ‘Verbal Stress’, pp. 321–34.

⁷⁴ ibid. pp. 322–3. It is not clear to me why Stanley regards prefixes and copulatives as Satzteilpartikeln.

⁷⁵ ibid. pp. 330–1.

⁷⁶ ibid. pp. 324–5.

⁷⁷ 235a, 322a, 335a, 368a and 500a.

⁷⁸ ibid. pp. 330–1.

⁷⁹ 567b, 625b, 837b and 843b. In the last two instances the verb must be stressed or the verse lacks alliteration.

⁸⁰ Examples of unstressed infinitives in Pos. 2: Gen A 2820, Wife 42b and Jud 90a; and OS 44a and 143a. Unstressed infinitives seem to have been quite common in Old Saxon (see Stanley, ‘Verbal Stress’, pp. 331–4), and this probably accounts for the rather high proportion in Gen B.

⁸¹ ‘Wortstellung’, pp. 8–9. For convenience, breaches of his ‘Satzspitzengesetz’ (ibid. pp. 43–5) are included in my list. See also Slay, ‘Technique’, pp. 12–13.

⁸² ibid. pp. 9–10.

⁸³ Metre, §20, n. 1.

⁸⁴ Verses such as Gewat pa ofer wagholm (Beo 217a) are not included here, as pa in Pos. 2 is stressed; see above, p. 78. For a contrary view, see Kendall, ‘Metrical Grammar’, pp. 1–9.

⁸⁵ Verbal prefixes other than those found in GenB are also possible, e.g. on- (Beo 2640a), purb- (Dream 46a) and unider- (Hel 1452a).

⁸⁶ There are apparently no examples of this type in Old English verse, and the examples in GenB probably reflect Old Saxon usage.

⁸⁷ ‘Verbal Stress’, p. 327.

⁸⁸ The date of the translation is difficult to determine with any degree of certainty. The language of the text is late West Saxon, and that of the exemplar is irrecoverable. Israel, Gollancz (The Cædmon Manuscript of Anglo-Saxon Biblical Poetry (Oxford, 1927), p. liv)Google Scholar dates the poem between the middle of the ninth century and the last quarter of the tenth. Timmer, (Genesis, pp. 43–5)Google Scholar shows fairly convincingly that it can hardly be so early, although his dating of 900 is not altogether convincingly either. It is fair to assume that the Old Saxon original arrived in England during the second half of the ninth century and that the translation was made at any time after that, but perhaps even as late as the first half of the tenth century.

⁸⁹ Metre, §21, n. 3.

⁹⁰ In the nom. 389a, 579a and 830a; in the acc. and dat. 263a, 284a, 379a, 394a, 400a, 406a, 491b, 647a, 662a and 684a. In Dream a number of instances in which a pronoun is followed by another particle should be treated in similar fashion (bim pær 31a, bie dær 63a and 64a, bie dæron 67a, and us man 75a). There are some instances with two pronouns in Pos. 2, e.g. he him (Beo 1711a) and hi me (Dream 46a). The case for enclisis in Old English is put forward by Peter Lucas in his forthcoming article, ‘Some Aspects of the Interaction between Verse Grammar and Metre in Old English Poetry’, in which such instances briefly referred to here are treated fully.

⁹¹ There are two examples of stressed pronouns in Pos. 1: forpon wit him noldon (741a) and hine bædon (Beo 780b).

⁹² 389a is a remainder (see below, p. 108). On491b, which is Type ₃B₁, see above, n. 27.6_47a is Type 3A*3 with monosyllabic anacrusis (hereafter designated ‘+’).

⁹³ Hypermetric Type ₃B₁(₃B₁).

⁹⁴ Cf., e.g., Fela sceal gebidan (Beo 1060b) Weorod wæs on wynne (Beo 2014a).

⁹⁵ Hypermetric Type I AI b(2AI). The possessive is postposited and must be stressed; see Fakundiny, ‘Verse Composition’, pp. 136–8.

⁹⁶ 25 5a is hypermetric Type + 1A*1b(1A1), 318b is Type + + 1A1band65 5b is Type + 1A1b.

⁹⁷ Both are Type +1A*1b.

⁹⁸ See also, e.g., 270b, 275b, 317b and 344b. For a full discussion of the syntax of auxiliary verbs in Beo, see Bliss, A., ‘Auxiliary and Verbal in Beowulf’, ASE 9 (1981), 157–82.Google Scholar

⁹⁹ See also, e.g., 282b, 291b, 328b and 391b. The auxiliary must be stressed in only one instance; see above, p. 85.

¹⁰⁰ This ‘extended’ type is apparently avoided in Beo, but appears in other poems, e.g. Christ A 271 (sculon), Dan 36 (wæron) and Sat 429 (wolde).

¹⁰¹ It may be possible to extend this licence to other verbs in Pos. 2, such as the celebrated ageafondsware formula (Slay, ‘Technique’, pp. 12–14). Other examples of lexical verbs in this position are GenA 2878a (geseah), Phoen 524b (seceô) and And 904a (spræc).

¹⁰² Out of a total of 130 verses of Type D with secondary stress in Beo, only four verses, two of which are suspect, appear in the b verse; see Bliss, Metre, §61.

¹⁰³ Type a comprises 4.3% of the total verses(Beo 5.6%); Type d comprises 8.0% (Beo 10–9%).

¹⁰⁴ Bliss, Metre, §§67 and 73.

¹⁰⁵ ibid. §§76–83.

¹⁰⁶ Type ₂A₃b is attested as a normal verse in 241a (stiðferhð cyning) and 702b (handweorc godes). Heavy verses of this type are very common; see, e.g. Beo 519b, 782b and 1619b and Exo 248b and 462b. In one example, however, up takes precedence over the verb and has primary stress: Yð up færeð (Exo 282a).

¹⁰⁷ Type 1D*6, for which there is just one example as a normal verse: oncyð eorla gehwæm (Beo 1420a).

¹⁰⁸ Type 1D6, for which there are two examples in normal verses: seon sibbegedriht (Beo 387a) swefon sibbegedriht (Beo 729a). The type is also attested as a heavy verse (Beo 848a, 2527a and 2774a).

¹⁰⁹ All Type 3A* verses (see below, App., Tables 8 and 9), except 236a and 240a, have monosyllabic anacrusis. In 621a the final vowel in wite should be underdotted and disregarded in scansion. Bliss's claim (Metre,§108) that Type 3A is ‘scrupulously avoided’ in Beo depends on his view of accidental alliteration in finite verbs of groups 5 and 6; see above, pp. 81–4. If, on the other hand, all alliterating verbs in these groups are stressed, then the following verses are Type 3A/A*: (Type 3A) Beo 2247a and (Type 3A*) 590a, 1711a and 2623a.

¹¹⁰ Bethel (‘Aspects of GenB’, pp. 175–6) scans 379a as Type + 1A* 1b and 240a as Type 1 Aic, but her placement of the caesura immediately after the verb in these and other verses of this kind is difficult to understand. I fail to appreciate how pronouns can be proclitic upon a following stressed element which is not a verb.

¹¹¹ Type 2Ara(ii); see Bliss, Metre, §55.

¹¹² For resolution of the sequence (img #)× in Beo, see ibid. §38.

¹¹³ 459 and 765b are Type 2A1a(iii). 848a is Type 1A*1a(ii). 263 is Type 1A*1b; Bethel's scansion (‘Aspects of GenB’, p. 175) as Type 1A1d not only misplaces the caesura (see above, n. 110), but also fails to recognize resolution of hine.

¹¹⁴ ‘Zur Rhythmik des germanischen Alliterationsverses’, BGDSL 10 (1885), 209–314 and 451–545, at 255–6 and 312. 744bGoogle Scholar is Type 2A1a(iii); 360a and 800a are Type 1A*1b(ii).

¹¹⁵ There is no ambiguity in inflected forms. So, e.g., in heardes irenes (383a), the vowel of the second syllable of irenes must be underdotted and ignored in scansion. See Sievers, ibid. pp. 480–2.

¹¹⁶ Permissible in 262a and 293a and precluded in 246b, 349b, 582b and 657a.

¹¹⁷ Respectively, the sequences and , with resolution of godes).

¹¹⁸ Obligatory in 507a; impossible in 240b, 247b, 251b, 295b, 386b, 490b, 515a, 637a and 642a; and optional in 267b and 662b.

¹¹⁹ Obligatory in 497b and 665a and optional in 557a and 658a. Underdotting would normally be precluded in arendel-ian, since the medial vowel represents an original full vowel. However, it seems plausible that underdotting could have been extended by analogy to such forms, as evidenced by early West Saxon forms such as arndian; see Campbell, A., Old English Grammer (Oxford, 1959), §393.Google Scholar

¹²⁰ Obligatory in 249a and 283a, impossible in 267b and 662b and optional in 743a.

¹²¹ E.g.,aldor (639a), ellor (773a), morðor (342b) and tacen (540b, 653a and 774a; optional in 714a).

¹²² If wæstm is monosyllabic, 236a is Type e, but this type is very poorly represented in the poem; see above, p. 95.

¹²³ 255a, 466a, 520a and 594a.

¹²⁴ Type 3B2. See Bliss, Metre, §59.

¹²⁵ So Sievers, (Der Heliand und die angelsächsische Genesis (Halle, 1875))Google Scholar and Klaeber, F. (The Later Genesis and Other Old English and Old Saxon Texts Relating to the Fall of Man (Heidelberg, 1913)).Google ScholarKrapp, (Junius, p. 168)Google Scholar follows the manuscript pointing and places swa in the a verse, claiming that it is ‘logically…closely connected with hwilc’, whereas Timmer (Genesis, glossary, s.v.) regards it as co-relative. Ericson, E. E. (The Use of Swa in Old English (Göttingen, 1932), p. 10)Google Scholar lists it among conjunctive adverbs in initial position, and this seems most likely. Thus this is one of the few instances in which the manuscript pointing is inaccurate (see above, n. 27, and Junius, ed. Krapp, pp. xxii-iii).

¹²⁶ See Bliss, Metre, §32 (7).

¹²⁷ Heliand und Genesis 730–1. In this instance the manuscript pointing, with a dot after wordcwyde and bis, is no guide to the verse division. The adopted reading in 730b gives Type d3.

¹²⁸ 239b and 735b are Type 2C1; 806a is Type 2C2 (cymð= (img #) ×). 574b is the sole example of Type a in the b verse. Other instances of the scansion requiring uncontracted forms are hean (300b, 358a and 545b) and near (760b), and are so marked by Klaeber in his edition. Klaeber, however, wrongly marks onfon (697b) as uncontracted; in fact the verb is contracted and the verse is Type 2E1. See also Pope, J. C., ‘On the Date of Composition of Beowulf’, The Dating of Beowulf, ed. Chase, C. (Toronto, 1981), pp. 187–95, at 193Google Scholar and n. 27.

¹²⁹ For a discussion of proposed solutions and references, see Junius, ed. Krapp, p. 164, n. on 238. None of the suggested emendations is convincing; the proposal that ea alliterates with palatal g is supported by doubtful instances (Exo 33, 190 and 339).

¹³⁰ A number of hypermetric verses have a similar initial sequence; see below, p. 108.

¹³¹ Figures for Types A, D and E include the corresponding light verses associated with these types (a, A;d, D;e, E). Figures for Beo are those of Bliss, , Metre, pp. 122–3Google Scholar (App. C, table 1).

¹³² See below, pp. 105–6.

¹³³ For full details, see below, App., table 10. The figures for Beo represent a simplification of Bliss, , Metre, p. 128Google Scholar (App. C, table IV).

¹³⁴ ibid. §46.

¹³⁵ ‘Rhythmik’, p. 234. The number of verses has been variously calculated; for a summary and references, see Cable, T., The Meter and Melody of Beowulf (Chicago, 1974), p. 33.Google Scholar

¹³⁶ Metre, §§46–50.

¹³⁷ ‘A Metrical Examination of the Poems Guthlac A and Guthlac B’, Proc. of the R. Irish Acad. 71C (1971), 91–137, at 94.Google Scholar

¹³⁸ Meter and Melody, pp. 33–5.

¹³⁹ 323a, 385a, 411a, 428a, 462a, 481a, 673b, 676a, 795a and 807b.

¹⁴⁰ (a-) 294a, 379a and 406a; (be-) 528b and 529a; (for-) 452a, 579a, 598a, 630a, 647a and 692a; and (to-) 841a.

¹⁴¹ (se) 272a, 479a and 496a; (pæt) 770a; (pis) 655b; (min) 542a; (pin) 521a; and (bis) 659b, 731a and 770a. On the reading adopted in 731a, see above, p. 100.

¹⁴² 379a and 647a are Type 3A* and 41 1a is Type 2A1. Cf., e.g., OS 16b, 90a, 121b and 157a, all of which are Type 2A with monosyllabic anacrusis.

¹⁴³ It can now be seen that this confirms Bliss's rejection (Metre, §50) of Type E with anacrusis. None of these verses contains those prefixes or proclitics characteristic of anacrusis.

¹⁴⁴ For details, see App., Table 9. In hypermetric verses anacrusis is characterized by elements identical to those found in normal verses; see App., Table 13.

¹⁴⁵ OS 5a, 21a, 23a, 37a, 69a, 72a, 90a, 148a, 157a, 193a, 198a, 222a, 264a, 272a, 279a, 289a and 321a; and 16b, 97b, 121b and 247b.

¹⁴⁶ See Sievers, , Altger manische Metrik (Halle, 1893), §89.Google ScholarBliss, (Metre, pp. 162 and 167)Google Scholar lists only three single hypermetric verses in Beo (1167b, 2173a and 2297a) and in GenA (156a, 1552b and 2857a), a fairly typical average for most Old English poems.

¹⁴⁷ Metre, §110. Max I is remarkable, for instance, in that, of thirteen double hypermetric verses in Old English verses (not including GenB), nine occur in the poem, as do five of six remainders listed by Bliss.

¹⁴⁸ Metrik, §120.

¹⁴⁹ OS 58b, 66b, 78a, 81b, 98b, 177b, 191a, 229a, 231b, 233b, 244b and 295a.

¹⁵⁰ See above, pp. 69–70.

¹⁵¹ ‘Der angelsächsische Schwellvers’, BGDSL 12 (1887), 454–82, at 455.Google Scholar

¹⁵² See Bliss, Metre, §107.

¹⁵³ Other varieties not found in Old English are Types 3A*(1A), 3A*(1A*) and 3B(3B). For details, see App., Tables 12 and 13.

¹⁵⁴ Bliss, Metre, §§74–5 and 105–6.

¹⁵⁵ See above, pp. 97–8.

¹⁵⁶ See Sievers, ‘Schwellvers’, pp. 479–80.

¹⁵⁷ Two verses in Max I (35a, 111a) also have the configuration (3B) at the end of the verse (Bliss, Metre, §110), but it is difficult to see how any of these should be classified.

¹⁵⁸ See above, pp. 78–9.

¹⁵⁹ 588a and 679a, with the sequence see above, p.101.

¹⁶⁰ 389a, 392a, 395a, 400a and 401a.

¹⁶¹ ‘An Argument for an Interpolation in the Old English Later Genesis’, NM 72 (1971), 224–45.Google Scholar

¹⁶² 397a, 403a and 408a.

¹⁶³ For references, see Junius, ed. Krapp, p. 166, n. on. 370.

¹⁶⁴ For a recent proposal, see Bliss, , ‘The Origin and Structure of the Old English Hypermetric Line’, N(img #)Q 217 (1972), 242–8.Google Scholar

¹⁶⁵ Genesis, p. 39.

¹⁶⁶ Above, pp. 68–72.

¹⁶⁷ ‘Nationality’, pp. 91–3.

¹⁶⁸ ibid. pp. 89–90.

¹⁶⁹ This impression is strengthened by personal contact with native English speakers who have lived in Germany for several years; a greater or lesser degree of interference is evident in their English, despite the fact that the two languages have diverged to a much greater degree than was evident in the ninth century.

¹⁷⁰ Heliand und Genesis, p. 16.