LIBRARY OF THE UNIVERSITY OF CALIFORNIA. Class ^^^1 CHAPTERS ^ ON ALLITERATIVE VERSE i;y JOHN LAWRENCE, D.Lix., M.A. (Lond.) LKKTOK OF ENGLISH IN THE UNIVERSITY OF I'KAGrK A DISSERTATION JN CANDIDATURE FOR THE DEGREE OF D.LIT. {LOND.) ACCEPTED BY THE EXAMINERS DEC. 1892 LONDON: HENRY FROWDE OXFORD UNIVERSITY PRESS WAREHOUSE AMEN CORNER, E.C. 1898 Price Three Shillings and Sixpence net CHAPTERS ON ALLITERATIVE VERSE BY JOHN LAWRENCE, D.Lit., M.A. (Lond.) LEKTOR OF ENGLISH IN THE UNIVERSITY OF PRAGUE A DISSERTATION IN CANDIDATURE FOR THE DEGREE OF D.LIT. (LOND.) ACCEPTED BY THE EXAMINERS DEC. 1892 t'NJVepr^/jy J 1 LONDON: HENRY FROWDE OXFORD UNIVERSITY PRESS WAREHOUSE AMEN CORNER, E.C. 1893 HORACE HART, PRINTER TO THE UNIVERSITY PREFACE The dissertation herewith submitted for approval varies very much from my original design, which was to have written a critical review of recent German treatises on the subject of alliterative verse. In January last, whilst engaged upon this task, my attention was drawn to certain unusual metrical points which I had noted in Cod. Junius XI in the summer of 1889, thinking them at the time simple errors on the part of the scribe, but which I now saw to be an invaluable clue to the construction of the half- verse. Following this out, by the aid of Prof. MoUer's work on alliterative poetry (Zur ahd. Alliterationspoesie), I was soon led to the con- clusion, stated in Chap. I. p. 30, that the half-verse is constructed on the same plan as the whole verse, that as the latter is divided into two sections by a caesura, each of which may begin either with an accented, or an unaccented syllable, so is the former. Unfortunately, my notes were confined to the Exodus and Daniel poems, and it was impossible for me, as I was then stationed at Prague, to consult the MS. further. This circumstance, and the know- ledge that during the present year we might expect another article on Metrik from Prof. Sievers (to be published in Paul's Grundriss) led me to abandon my plan of dealing with the whole subject of alliterative verse-structure, and eventually I decided, until I should return to England, ' "778 iv Preface. to confine my attention to the laws of alliteration simply, using as my chief guide the admirable treatise by Eieger : Die Alt- und Angelsdchsische Verskunst. Here again, however, I found it impossible to pursue a straight course. The question of crossed alliteration first encountered me. Like other people I had accepted Frucht's calculation as probably decisive. It turned out to be based on false premises. I had therefore to under- take an investigation of my own, and this cost me more weeks of labour than I quite care to confess. When it was at length concluded a fresh subject for research presented itself in the shape of vowel alliteration. I began this with no thought of examining more than Piers Plowrtian, with a view to finding in what respect, if any, its vowel alliteration differed from that of Beowulf. I was led on to examine a number of other alliterative poems of the fourteenth century. For myself the work has been very instructive. I can only hope that it will turn out of some use to others also. The tables which I have drawn up will, at any rate, provide a key to the construction of the verses in these poems where it is most difficult to de- termine. This digression employed me, with two short intervals, until the end of August. Meanwhile, in July, I had returned to England and made the collation which is given in Chap. I. I found the Genesis even richer in exceptional pointing than I had expected, and am sorry that the discussion of vowel alliteration has left me so little time for working up my materials. The instances in which ordinary points have been omitted in the MS. have had to be left unconsidered. It is possible that many of these are not due simply to neglect and would, if examined, yield fresh light upon the scribe's theory of O.E. verse. The results given in Chap. I are however sufficient to show what this was in the main. Among the conflicting theories which now hold the field it Preface, v agrees most with Prof. Holler's, the simplest and most easily applied of all. The true rhythm of the Old English verse is not a matter of mere antiquarian interest. Until it is understood the development of English prose-rhythm cannot be pro- perly explained. Probably we shall have to wait some time yet before anything like unity of opinion can exist upon the subject. We have yet to hear what answer Prof. Sievers can make to his various critics, and to receive a more extended treatment from his pen than that contained in P.B.B. X, xii, and xiii. My own conviction, however, is that truth lies on the side of Prof. MoUer, to whom our Junius scribe now comes as an ally. I would fain hope that my own attempt ' to copy fair what time hath blurr'd ' may give some fresh stimulus to the already awakening interest in O.E. verse-lore, and in particular may call increased attention to Prof. Moller's treatise, which even in Germany has been too much neglected. JOHN LAWRENCE. 6, RiLLBANK Terrace, Edinburgh, September 21, 1892. POSTSCRIPT. Since the above was written several months have elapsed, during which I have kept the type of these ' chapters ' standing in the hope of giving them a thorough revision before publication. Circumstances have hitherto conspired to hinder this being done, and I am unwilling longer to delay sending them forth. The chapter on Crossed Alliteration has received some changes, and a few foot-notes (indicated by an asterisk) have been added here and there. vi Preface. In the interval the 'more extended treatment' by Prof. Sievers has appeared in the shape of his Altgerma- nische Metrik (Halle, November 1892) but without, as far as I can see, much affecting the position of the rival theories. My own discussion of Prof. Sievers's system begins, rather abruptly, on p. 27 below. Tabn House, Ilkley, April 8, 1893. CONTENTS. CHAP. I. — Metrical Pointing in Codex Junius XI : its relation TO theories of O. E, Verse-Structure : collation WITH Thorpe CHAP. II. — Crossed Alliteration CHAP. III. — Vowel Alliteration in the Fourteenth Century COMPARED WITH THAT OF BEOWULF . Vowel Alliteration in Beowulf . Morte Arthure Destruction of Troy William of Palerne Alexander Sir Gawayne and the Green Knight . Cleanness Patience . Piers Plowman Concluding Remarks .... 38 56 60 65 77 82 89 91 93 94 lOI III CHAPTER I. THE METRICAL POINTING IN CODEX JUNIUS XI. * In der in metrischer Beziehung besonders schlechten Hs. Junius XI' (P. B. B. x. 458). In this way does Prof. Sievers characterize the manuscript containing the poems once attributed to Csedmon. The words sound like a re- flection upon the scribe, but can scarcely have been so meant, for never was a slur less merited. Junius XI alone of 0. E. MSS., as far as I am aware, gives us any material help in determining how to read alliterative verse. In the Beowulf^ as Guest observes (p. 312), and as anyone may see from the E. E. T. S. photographic copy, or from Holder's Abdruch^ ' the point was used merely to close a period, and the versification had nothing but the rhythm to indicate it.' This appears to be the case with the Exeter Book also, if I may judge from a short examination (which is all I have been able to make) of the transcript of that MS. in the British Museum (Addi- tional MSS. 9067, 154 f ). How it stands with the Vercelli MS. I am unable to say. The scribe of Junius XI has, however, been at pains to insert the metrical points in most cases. Where mid-points are omitted it is generally after a short first half- verse with single alliteration followed by a second half- verse without Auftakf^; e.g. pd Boedtvode Boyppend ure {Gen. 206). End-points have seldom been forgotten ; less than forty times in the more than 4000 verses of the Genesis, Exodus, and Daniel poems. The points are, with very few excep- ' I use this term throughout in preference to the ponderous term anacrusis and to Mr. Skeat's catch (Essay on All. Poeiry), B 2 Dissertation on Alliterative Verse. tions, correctly placed, much more so than in Thorpe's edition. Guest (p. 313) tells us that, having marked twenty or thirty cases of ' doubtful prosody in the latter, he after- wards found on examination that in every instance but two the text had been altered.' Our own examination will, I think, confirm the ' doubtful prosody ' of Thorpe's verse divisions in most cases where he has deviated from the MS. The metrical imperfections of Junius XI must therefore lie at the door of the authors themselves who wrote the poems it contains, or possibly of some intermediate scribe, and not of our scribe, who has indeed struggled manfully with the difficulties of the verse, which here and there are considerable. In my short article on the text, in Anglia xi. I myself made, it must be confessed, some slight imputation upon his carefulness. This called forth from Prof. Stoddard, of New York University (whose article on the MS. in Anglia x. will be jemembered), some expressions which he will, I trust, forgive me for inserting here. 'That scribe,' he wrote, ' established with me a very good repu- tation. I am rather jealous of his honour.' Let me hope Prof. Stoddard", will feel that in the present chapter I have made the amende honorable for anything said earlier to the disadvantage of one whom we have so much cause to thank as the writer of the Csedmon MS. In the following table, I give, on the left the instances in which Thorpe has placed a point after words which have none in the MS., and on the right, those in which points have been omitted. The two columns together exhaust the disagreements between Thorpe's edition and Junius XI, as regards the metrical pointing. Verbally, Thorpe's edition is extremely correct, though not absolutely so. The num- bers of the verses concerned are given from Grein. The index a stands for first half- verse ; h for second half-verse, as usual. The Metrical Pointing in Codex Jimitts XL 3 Grein. Point added by Thorpe. Grein. Point omitted by Thorpe. Genesis Page Line Genesis Page Line I* 1, I micel 4a 1, 7 ealra 9a 1, 17 so^Sfaest. 30» 3, 3 onjan 30a 3, 3 unrsed. 36^ -, 16 werlo^an 47b 4, 2 mode 48a -, 3 woldan 50a -, 7 cininj 53* -,13 msBra 53^ 4, 13 mod. 71b 5, 15 sy^^an. 72» 5, 15 seomodon 72^ -, 16 si^e 72^ -16 swearte. 73* -17 |)orfton. 79a -, 29 J>eawas 82b 6, 2 buan 133* 8, 32 seresta. 133^ 8, 32 jeseah 135* 9, I jewat 141* -,13 jesundrode 145* -21 -flode 155^ 10, 12 jyt 156a 10, 13 lond. 165b - 32 cyninj 170a 11, 4 lenj 171* - 6 won^es 197a 13, 3 -jrene 206a -, 21 sceawode 227a 15, 2 selestan 262a 17, 19 onjan 262a 17, 19 enjyl. 284b 19, I stride. 285b -, 3 me. 290a 19, 12 oleecan 291b - 15 lens. 297b -, 27 jehwilc 298a -, 28 waldend 317* 20, 30 jeswinc 319a 21, 4 si=Se 345^ 22, 24 sweartan 345^ 22, 25helle. 356* 23, 10 styde. 395* 25, 17 jemearcod. 398* -, 23 adame. 401a 26, 3nu\ » This word is omit B ted in Gre 2 in V. 401. Dissertation on Alliterative Verse, Grein. Point added by Grein. Point omitted by Thorpe. Thorpe. Geneais Page Line Genesis Page Line 403a 26, 7 Sodes. 403^ -, 8 nu. 405a -, II onwendon. 408a -,17 beam. 425^ 27, 30 mode. 474^ 30, 29 habban 474b 30, 29 worulde. 475* -, 30 witode 475* -, 30 waeron. 475** - 31 SeJ>m5l)0. 486^ 31, 16 he 487^ - 19 fyre 487^ 31, 19 sweartost. 507a 32, 22 drihten. 507^ -, 23 bine. 511^ 32, 31 wite^ 548* 35, I eor^rice 555^ - 15 he 555" 35, 15 swa. 588b -, 22 inc. 560a -,25 willende 561a -, 27 rume 567* 36, 5 |)ineB. 567^ 36, 5 habban 568a -, 7 adame 573» -,17 andwyrde 583* 37, I jeare 593^ -, 22 beames 606^ 38, 14 scea^Sa 606b 38, 15 jeorne. 6i8a 39, I craefta 6i8«* 39, I cime. 6i8b -, 2 jiet. 640* 40, 16 morS 643^ -, 23 waestm 643^ 40, 24 an. 649" -, 35 onjan. 6si- 41, 3 jode 651a 41, 3 bysene. 659^ -, 20 unc 659^ -, 21 betere. 683* 42, 35 je^nunja 694^ 43, 22 scea^a 702b 44. 3 jodes. 706b -, 9 jehate. 711* 44, 18 bodan 712* -, 20 hyldo 713b -, 23 were 713b -, 24 swelce. 730b 45, 22 cwyde. 748* 46, 22 his 748* 46. 23 eft. 755^ 47. 4 mor^res. 759* -, 11 heortan. The Metrical Pointing in Codex Junius XL 5 Grein. Point added by Grein. Point omitted by Thorpe. Thorpe. Genesis Page Line Genesis Page Line 759b 47, 12 uncre. 760b -, 14 eft. 763^ -, 20 lijas. 764* 47, 20 secan 787^ 49, 5 jodes 8loa 50, 17 heofnum 824b 51, 10 adam 827» -, 15 adam 828* - 17 waldendes 836^ -, 34 J)eodne8 836^ 52, I hafa. 839*^ -, 7 innan. 851a 52, 29 leohte 852» -,31 feran 855* 53, 2 neosian 863* -, 18 ahsian 867» -, 26 her 872a 54, 4 sod , 882a - 24 adam 882» 54, 25 eft. 893^ 55, 13 sealdest. 906^ 56, 3 |)inum. 907» 66, 3 breostum 913b - 17 fiersna 913b -, 18 ssetan. 914a -, 18 tohtan 914^ -, 20 jemsene. 9i8» -, 26 jod 927» 57, 12 o^erne 955* 58, 32 frofre. 955*^ 58, 32 let 965* 59, 17 I'a 968* -, 23 twa 969* 59, 25 cain. 972* -,31 eor)y 1895^ -,31 secan 1901a 114, 8 jebyrdum 1912^ -, 31 wanjas 1956*' 117, 20 a 1956b 117, 20 aefter. 1960* 118, 4 jefrajjn 1964^ - 13 t'a 1964b 118, 13 feower. 1967a -, 18 herjum 1972b 118, 29 bennum. 1973^ 119, 2 5u=6|)r8ece 1974* -, 3 foran 1974* 119, 3 fife. * Grein i 768, Assxjri 1. Dissertation oti Alliterative Verse. Grein. Point added by Grein. Point omitted by Thorpe. Thorpe. Genesis Page Lino Genesis Page Line 2016a 121, 27 loth. 2027a 122, 15 aner. 2027*' - 16 escol. 2037^ 122, 36 him 2037b -, 36 torn. 2047b 123, 19 hum. 2091b 126, 7 lenior. 2092* 126, 7 mseje^ 2096a -, 16 J)anon 2IOI» -, 26 solomia 2135* 128, 31 sceoldon 2142b 129, 12 feoh. 2146* 129, 19 wurde 2149a - 24 heonon. 2172* 131, 6 |)a 2173a 131, 8 daedrof. 2182b - 27 mynte^. 2183a - 28 sie. 2207* 133, 7 ejypta 2219a 134, 3 forwyrade 2220a -, 5 msejburje 2223a -, II e^ylstaef 2228* -, 21 recene 2229» -, 23 afanda 2229^ -, 24 wille 2229b 134, 24 frea 2241a 135, II sarran 2247* -, 23 «e 2247a 135, 23 ajar. 2249* - 27 dojora. 2250* - 29 unarlice. 2250>> - 29 ))8Bt 2253a 136, 4 drihten 2253b 136, 5 dema. 2257a -, 12 lease 2269a 137, 5 )>u 2302a 138, 35 ¥eoden 2327b 140, 14 J)ae8. 2335* 140, 29 cyninjas 2336a -, 31 }>a 2345* 141, 15. metode. 2351* 141, 27 feejere 2352^ -, 29 drihten 2356* 142, 3 "ismael 2368» -, 27 fremede The Metrical Pointing in Codex Junius XL 9 Grein. Genesis 2397a 2463a 2490b 2512a 2559a 2565a 2573a 2623b 2649a 2690a 2693* 2759b 2767'' 2770a 2783b 2784* 2789b 2826a 283i» 2845* 2847* 2889a Point added by Thorpe. Page Line 144, 29 aedre 148, 27 arones 150, 12 8ona 151, 21 idesum 154, 21 jcomorra - 33 anlicnesse 155, 15 jeworhte 158, 26 his 160, 12 unfricjendum 162, 32 ))a 163, 4 baldor 167, 1 feran -, 17 on -, 23 cynne 168, 16 8i¥'6an -,17 ellor -, 28 lice 171, 10 alwalda -, 20 abraham 172, 16 rinces -, 20 sB^elinJes 175, 2 sweord 2929a I 177, 13 jenam Exodus Grein. Genesis 2463b 2490b 2491* 2544b 2577^ 2584^ 2628a 2628« 2662a 2693* 2693b 2769* 2771a 2784* 2893^ 291 1* 2912a 2926b 2933* Exodus 9b 178, 10 sylfes 25» 179, 7 worhte 34» -, 24 jedrenced Point omitted by Thorpe. Page Line 148, 28 snytra. 150, 12 here. -, 13 burhwarena. 163, 26 lean. 155, 25 up. 156, 5 hatne. 159, 2 brinjan. -, 2 him. 161, 8 J)eawf8eat. 163, 4 jumena. - 5 Su=S. 167, 21 wuldortorht. -25 weox. 168, 17 ajar. 175, II swa. 176, 13 ofstum. -, 15 wuldorjast. 177, 8 rom. -, 21 lac. lO Dissertation on Alliterative Verse. Grein, Point added by Grein. Point omitted by Thorpe. Thorpe. Exodus Page Line Exodus Page Line 56* 181, 4 folce 56a 181, 4 j'y- 61* —> 14 mearchofu. 64«* -, 20 feondum 70a 182, 3 forbserned. 86* 183, 4 wuldre. io8b 184, 17 wundor. lo9» 184, 17 syllic 142b 186, 22 injefolca. 158^ 187, 33 tredan l6ia 188, I hwsel. 197'^ 190, 10 jemynted. 197b 190, 10 ))am 209a 191, 3 healfa. 288*^ 196, 7 ece 288b 196, 7 y^e. 34i» 199, 19 simeones 364* 200, 29 deopestan 376* 201, 22 heofonum 381^ 202, I feor 381^ 202, I neah. 391a -, 26 jetimbrede 393^ -, 24 on 412^ 204, 2 reodan. 413* 204, 2 majan 413a ~i 3 mece. 447^ 206, 4 dea¥e. 509a 210, 2 heoro 5 x3a 210, 10 -bodan. 570* 214, 16 feonda. 572a "1 20 brimu. Daniel Daniel 35^ 218, 8 8er))on. 38^ 218, 12 ))aBre 43* -,22 israela 6ia 219, 27 jestreona 61a 219, 27 jestmdan. 62» -, 29 eorlas 66» 220, 4 freos 69^ -, II wejas 79a -, 30 secan Io6a 222, 17 ejesful. 134* 224, 10 unbli^e 154^ 225, 15 jife. 163a 225, 32 daniel 163'' -,33 micelne The Metrical Pointing in Codex Junius XL 1 1 Geein. Point added by Thorpe. Grein. Point omitted by Thorpe. Daniel rage Line Daniel 164a 182a 193a 203^ 203^ 206* 208b 214'* Page Line 225, 34 bleed. 227, 5 wurSedon. -, 27 wseron (?). 228, 17 ])ider. -,17 hweorfan. -, 22 hearan. -, 25 ))e ¥;u |)e. 229, 8 fyres. 214^ 229, 9 wolde 219^ -, 19 gelaeste 224* "' 26 yrre 227b 230, 3 jrimmost. 233^ 230, 15 nydde 233^^ 234a -, 15 jrome. -, 16 fsB«m. 252^ 231, 25 waeron 254* 232, 2 utan 254b - 3 jehwearf 254^ 232, 3 alet. 26l» 16 |)e 261a 266* 274b -, 16 fas^me. -, 26 fyrscyde. 233, II him. 275a 233, 12 inne 276* -, 14 |)onne 280* -, 23 azarias 292a 234, 14 nu 2928- 234, 14 jeorne. 293a -, 16 help 299b 235, I sittendum 303* 235, 8 lif. 303^ -, 8 jeond 315* 236, 2 jacobe 320a "' 12 maenijeo 334* 237, 7 halja. 345» 237, 29 feondas 366b 239, 7 jesceaft 396b 240, 34 sellende. 397* 240, 34 eallum 398b 412b 413^ 241, 3 adzarias. 242, I fela. - 3 t'ly. 430* 243, 2 la^e 430^ 431* 436* 243, 3 lens- -, 4 cyninj. -, 14 leoda. 444a 244, 7 jewur'^od 444* 244, 6 wundre. 12 Dissertation on Alliterative Verse. Grein. Point added by Thorpe. Daniel 449* 475* 48 1» 495* 505^ 528^ 57o» 589^ 61 2«^ 674* 676* 676* - 683» 708^ 7i6» Satan 5^ 22* 55^ 65* 79b 95b II2» 112^ Page Line 244, 16 brsesna 246, 5 he -18 jastum 247, 10 -hyd -,31 scylde ^ 249, II ealle 251, 27 onhweorfe^ 252, 25 rice 253, 5 |)onne 254, 15 e^el 258, II eorla -,15 a woe -, 17 baldazar -, 29 babilone 260, 12 judea -27 mihtijran 265, 10 uton -, 24 rim 266, 14 mode 268, 14 ahtest -, 36 -^ae^n 269, 27 dream 270, 24 heofon 271, 28 flyje -, 29 ^rajum Grein. Daniel 444b 450* 450b 452b 456* 467* 474'' 475* 489^ Point omitted by Thorpe. 509'' 512^ 523' 641* 650^ 668b 676* Satan 5' 55' Page Line 244, 7 |)e. -, 18 sinum. - 19 aldre. -, 23 morSre. 245, I duju^Se. -, 22 ejesan. 246, 5 jesawon. -, 6 cwealme. -, 35 sel. 248, 7 ufan. - 13 wej. - 33 "f^ite. 256, 15 eft. -, 34 wilddeorum. 257, 36 cwelm. 258, 15 ^eode. 112^ 265, II S3B. 268, 14 anum. 271, 29 flylite. The Metrical Pointing in Codex Junius XL 1 3 G^REIN. Point added by Grein. Point omitted by Thorpe. Thorpe. Satan Pago Line Satan Page Line 113* 272, I neosan 113* 272, I earda. 113^ -, 2 ma. II4» - 2 \>e 125^ -, 26 jelomp 213^' 277, 32 seo. 236^ 279, II -maelum 280* 282, I j^nornedon 280* to Points from p. 282, 1. I 286b to 1. 14 inclusive, all 287b to omitted in Jun. xi ; 295* also from 1. 16 to I. 296a to 31 ; and from 1. 33 to 297a 1-35- 299* 283, 3 eor^an 301* -, 7 jastlice 301^ -, 8 cuma^ 302^ -, 10 moton 303*^ The points of lines 11, 304*; 12, 13, and of 22, 23, 308b . . 24, 25, 27, and 29 not 3io» in Jun. xi. "~"- 311b; 312b 371b 287, 22 swearte 371b 287, 23 je|)ohte. 372» -, 23 wolde 378b 288, 10 andwlitan 442* 292, 17 ordfruma 489* 295, 20 jereaw. 490* -, 21 carcernea. 497* 296, 4 fela 497b 296, 5 teonan. 53o» 298, 8 stod 530* 298, 8 upp. 554^ 299, 24 ajan 559^ 300, 4 folja^. 594* 302, 6 hycjan. 61 5» 303, 18 jesena^ 648* 305, 17 torht 648* 305, 17 swesle. 664* 306, 14 aldor 666» -, 18 ?e|)rowode 679' 307, 12 liht^ 679' 307, 12 lifijendum. 704* 309, 3 sid 731^ 310, 23 sastas 731^ 310, 23 werijan. » Gr€ )in. leoht. 14 Dissertation on Alliterative Verse. Character The metrical points in Junius XI, pp. i to %\% (the part Points. of the MS. containing the Genesis, Exodus, and Daniel poems), are lozenge-shaped and of a good size. They were evidently put in by the same bold hand which wrote the MS. Those in the second part of the MS. containing the Satan, which is written in a smaller and inferior hand, are mere dots. In many cases the dots are rendered more noticeable by the mark / above them (compare those in Sir F. Madden's Layamon). Occasionally this mark ap- pears without the dot, when two words belonging to different half- verses have been written close together. The lozenge-shape of the points in the first portion of the MS. is often imperfect, but for the most part comes out distinctly enough when examined under a microscope. Sometimes the points are small, especially when they occur at the end of the finishing stroke of an a, &c. In these cases it is difficult to decide whether the scribe meant anything by them or not. I have, however, thought it best to include all instances in which the point is clearly an ink mark, and not a mere stain or speck in the parch- ment. Small points are those at mod, (Genesis 53^) ; engyl, (Genesis 362*) ; snytra, (Genesis 2463^) ; y3^e, (Exodus 288^) ; halga, (Daniel '^^t^^) ; sel, (Daniel 489**) ; Jyndde, (Daniel 6^6^) ; folgad, (Satan 559^). These have all been omitted by Thorpe. Wherever in the MS. there is any fair semblance of a point to agree with Thoi-pe's pointing I have allowed it to pass, although in some cases, e.g. rice, (Daniel 589) the point is of the smallest. Here and there a point has apparently been erased, e. g. after mode, (Genesis, 47^) ; lease, Genesis, 2257* ; tredan, (Exodus, i^S);gelceste, (Daniel 219) ; also perhaps at gyt, (Genesis 155^), where after the t there are the partly erased letters of a word written in error. At si^e, (Genesis 319*) and his, (Genesis 2623^), instead of the usual points, we have commas to indicate that the succeeding words fylde and sweostor respectively, which follow too closely, are separate words. The point ^ifiry, (Daniel 413), is perhaps not metr The Metrical Pointing in Codex Junius XI, 1 5 rical. It was customary to mark off numerals with dots. na Thus, in the Beoiuulf MS. we have -xv- (ver. 207), -xii* (ver. i868),-xxx- (ver. 379). So in Junius XI we have -firy-^ (Daniel 272) and in the Exeter book -iii*, {Az. 171), -iiii-, {Az. 174), 'J)ry; (Az. 155). (Cp. also the numbers of the Cantos in the Beowulf US,) I have not included among the metrical points omitted by Thorpe a curious mark after J)urh, (Daniel 595), which looks like the commencement of some letter, perhaps of the down stroke of a t. On the other hand I have included the case of wceron, (Daniel 193), but with a query. The mark here, when looked at through' a microscope, is seen to be pear-shaped, altogether different in appearance from the usual lozenge-shaped point. The point at fosidm^ (Daniel 234) is much the same shape. § I. In discussing the contents of the foregoing table I Discussion propose to leave the Satan poem out of the account. ences be- Taking first the left-hand columns, it will be seen that we Thoroe have 1 01 references to first half- verses (marked a) and 77 and Ju- nius XI. to second half- verses (marked 6). The former number includes all the cases in which a mid- point omitted in the MS. has been supplied by Thorpe. To ascertain the exact number of MS. omissions the following deductions must, however, be made: — 45 for verses in which the MS. has also a mid-point though not in the same position as Thorpe ; 7 for cases in which Thorpe has erroneously placed the end-point of a verse after the commencement of the one next following, according to Grein's division, viz. Genesis 764% 907", 1857*, 2092*, Exodus 109% 413*, Daniel 475* ; and i for Daniel 276* where a quite super- fluous point has been inserted by Thorpe in the Auftakt of a verse, so creating an entirely disconnected half- verse : in all 53. We have remaining 138 instances in which the scribe of Junius XI neglected to put in a mid-point, not a great number out of the 4289 verses in the three poems, being less than 3*25 per cent. In the case of the end-points i6 Dissertation on Alliterative Verse, the total is much smaller still, viz. 37 (or less than iTo), since from the number 77 above given we must subtract 40 for cases in which the MS. gives an end-point but not after the same word as Thorpe. § 2. Where the MS. and Thorpe are at variance as to the right position of an end-point, Grein's division generally agrees with the former. With regard to the csesurse, Grein, as a rule, gives no indication, but there can be no doubt that here again Thorpe has, for the most part, changed the manuscript pointing for the worse. In some instances, how- ever, it is apparently at fault, coming too early in Genesis 618,651, 1255, 1401, 1974, ^247» 2693, 2784, Daniel 61, 6y6y and too late in Genesis ^^^ 1521, 1697, 2463, Daniel 430, 512, in all of which cases the scribe has overlooked the main caesura and given a point at a sub-caesura. Upon this sub-caesura, or division of the half- verse, something will be said later on, when we shall have occasion to notice the verses in which both it and the main caesura have been marked in the MS. What seems a curious mistake in pointing has been made in Exodus 56 where the MS. reads : Of erf or he mid Jty- tolce fcestenCb worn* I can only suppose that the scribe was in a careless mood for the moment, as indeed the character of the accents alsd seems to show. Possibly some reminiscence of such a verse as Genesis 47 1* : swd him defter fy- may have been floating in his mind as he wrote. It is noteworthy that in verse 21* Ofercom he mid Jty campe- there is also a point after the article, though not of the normal shape and size ; moreover, in this case, the point at the main caesura has not been forgotten. In the other cases of difference between Thorpe and Jun. XI. in regard to the main caesura, the MS. division, if not correct, is at least defensible. They are as foUows : — i. Genesis 30, 262, 955. In these three cases the syntax is clearly in favour of the MS. division, to which, moreover, no objection can be made on the ground of metre, since first half- verses of precisely the same character as it gives The Metrical Pointing in Codex Junius XI. 1 7 are to be found also in Beowulf (cf. P.B.B. x. 284 ff.) ; on the other hand, against Thorpe's pointing of Genesis 30 com- pare P.B.B. X. 242, § 16, where objection is made to a word like unrced in a B. verse. ii. Genesis 72. Here Thorpe's division (with which, strange to say, both Grain and Bouterwek agree) gives ua two alliterating words in the second half-verse to one in the first, a thing which Eieger (p. 9) finds to occur only once in the Genesis, viz. ver. 2321. The MS. division at sweoHe and porfton is the correct one (cp. P.B.B. x. 512). Against ihQTQ^vliuig enjanihement (of 72^ 73*) nothing can be said. For a sentence to terminate, as at si^e, in the middle of a half-verse is certainly unusual, but occurs again Genesis 2567^ 2568*, /cer hie Strang hegeat- wite, Jms heo wordum, iii. Genesis 133. Thorpe divides as Rieger (p. 39), so obtaining an impossible half- verse dceg cbresta geseah : — — — X X— (For the secondary accent in ceresta cp. P.B.B. x. 228, 229. For the length of dceg in this position see Paul in P.B.B. viii. 1 84, note.) iv. Genesis 475. Thorpe and Grein divide at ^uitode, the latter supplying tires (!) in the second half- verse (in the Glossar, p. 472, wuldres is suggested), so avoiding the confusion into which Thorpe is led by his false pointing. The MS. division is unobjectionable, the half- verse him to wceron- being quite a possible A^ verse (cp. P.B.B. x. 284) even without habban (see § 4, i). V. Genesis ^^5. Thorpe divides at he, Grein and Bouterwek, apparently, at cerende, where they place a comma. Thorpe is clearly wrong. The scribe points at stva, taking the word closely with hwilc in the sense of whatso, whatsoever (cp. swa hwylce daga : Ps. 1374=^1 quacunqne die : Grein iv. 499, § 11), which is the best way. vi. Genesis 1601. This is really a long verse of three half- verses : — xxx-x I x-x-| X x-x-|. Thorpe divides at the end of the first, Grein and the MS. at the end of the second. (Op. P.B.B. xii. 476,) 1 8 Dissertation on Alliterative Verse, vii. Daniel 392. Here there is a conflict between syntax and prosody as in Genesis 72. Thorpe and Grein point according to syntax, obtaining a first half- verse of the unusual form -^ x x |-x-||. The scribe separates the adverb from its clause by the metrical point, which is certainly awkward. viii. Daniel 261. Thorpe and Grein (Glossar, p. 574, § 2), take ^ces . . J^e SiS the genitive of the relative se Jbe, but I query if the enclitic ^e can so be separated, and in any case if it is strong enough to bear the second chief accent of the half- verse, which it must do according to their arrangement. The MS. points at/oe^me, taking ^e no doubt as a conjunc- tion and J^ces as demonstrative : The young men pt^aised God that under his shelter ivas driven away, &c. ix. Genesis ^^6, 398 ; Daniel 206. Thorpe neglects the esesural point of the MS. in these three verses, taking them as merely half- verses ; Grein, on the other hand, as long, agreeing with the MS. Exodus 413. Point again neglected by Thorpe: Grein divides as MS. Genesis 2628* : regarded as merely a half- verse by Thorpe and Grein ; by MS. as a full verse. According to Sievers, however, heht hringan is too short for a half-verse. More- over the alliteration of the governing verb instead of the infinitive is irregular. X. The following instances of deviation on the part of Thorpe need no discussion, the MS. pointing being obviously correct in every case ; viz. Genesis 567, 748, 882, 1023, 2250 ; Exodus 86, 197, 513 ; Daniel 266, 303. xi. Genesis 2253 ^^^ Exodus 288 are corrupt passages. § 3. The instances in which Grein has restored the MS. division where Thorpe had altered the position of an end- point are as follows — Genesis 71, 345, 475? 606, 643, 659, 702, 713, 76^, 836, 893, 906, 913, 914, 1023, 1856, 2047, 2091, 2577 ; Exodus 108, 142, 412 ; Daniel ^^y 396, 398, 474. These we need not go into. The Metrical Pointing iji Codex Ju7iius XL 1 9 The MS. divides too early in Genesis 487, 1832, 1964, 2037, 2229, 2490 ; Daniel 233, 254 (in each case pointing at the middle pause of the half- verse) ; and too late in Genesis 131 6, 1759. ^s regards Genesis 1964, it is worth remarking how often the scribe omits a metrical point after the word /a ; cp. Genesis 965, 1697, 2172, 2336, 2690. The pointing of Genesis 131 6, 1759 is indefensible if it was intended to indicate where verses 1315, 1758, respectively, close. Prob- ably, however, we have here the extra point before allitera- tion of which examples will be mentioned below. In a few other cases Grein (as well as Thorpe) differs from the MS., viz. in the end-division of Genesis 72, 474, 1446, 2182. As to the first of these cp. § 2, ii ; the others will need a few words. i. Genesis 474. Thorpe and Grein divide at habban, the MS. at worulde. The former division makes 474^ a Schwell- vers - X I w X X | — x , though Sievers does not notice it as such (P.B.B. xii. 454 ff.) ; the latter leads to an awkward break in the syntax of 475* (cp. however § 2 ii. and vii.). ii. Genesis 1446. This case is much the same. Dividing at eft^ as Thorpe does, we get an irregular half- verse, viz. B with minor accent in middle thesis : x — w x — ; (cp. P.B.B. X. 242 § 16). Can-ying eft into the next half- verse as the MS. does produces a strong syntactic pause in the Auftakt of 1446^. iii, (Je-nesis 2182, 2183. Thorpe and Grein arrange these as two verses, each with a Schivellvers in the second half ; the MS. as three normal verses thus : — fcegen freobearmim. fceste mynte^ . ingepancum .p me cefter sie • eafo ran sine - yrfeweardas.— {Junius xi, p. loo, lines i, 2, 3.) The scribe's division is of course the right one. Prof. Sievers (in P.B.B. x. 476) says that in Germ. x. 417 Grein has corrected his former arrangement. I have no means of consulting this, but no doubt it will give the verses as above *. * It does do so. C 2 20 Dissertation on Alliterative Verse, Our examination so far has, I think, made it clear that the scribe of Junius XI possessed a more accurate know- ledge of the ars "metrica of our fathers than any of his editors has had. In the next section I hope to show that had his pointing been in all cases strictly attended to a vast amount of mis- directed ingenuity which has been spent over metrical theories might have been spared. § 4. We have now to consider the verses in which the MS. has extra pointing beyond what it usually gives at the end of each half- verse. It will be seen that in many cases these verses are of an abnormal length belonging to the so-called ISchwellverse. These I must refrain, for the present, from discussing fully. As, however, Prof. Sievers's system of scansion is the one generally accepted to-day, I shall classify them, as far as possible, according to it. The ordinary verses with extra pointing will be treated at more length. I give first the instances in which the extra pointing appears in a first half- verse : — X. 1. Genesis 9*: soa/cest-and awi^fet'om^- 2. f, 156* : vrid lond . ne vre%as nytte. 3. „ ^g^^: hehcef^nu "^em-earcod-amie JXl^dda7^'^ea}'d^ ^. ,, 401* : ne '^elyfe ic me nu -pees \eohtes furdor. 5. „ 403* : pcBt we m.ihti'^es y)des . mod onwcecen. 6. „ ^o^^'.pwt hie pcBt onvfendon pcet he mid his worde behead. 7. „ 4oS^ : fira beam .on pissum fcBstiini clo7nme. 8. „ 507* : hcefstpe wi^ dtihten . dijrne '^eworhtne. 9. ,. y^g^: hyj^e ymbheortan.^erume' 10. „ 969*: cain . and abel. 11. „ 1077*: Bida- and sella. 12. ,, 1400* : pam at melistan . wcbs. 13. „ 1547*: olla.ollira. 14. ,, 1617^ ; ohus ' and dham. ' I regret that I have not in all cases noted the exact page and line in the MS. where these examples occur. They will be easily found, how- ever, by aid of Thorpe's edition, which gives the MS. pagination in bracketed numbers. Grein gives that of Thorpe similarly ; so that with the two editions and the MS. combined, no trouble will be experienced in verifying the subsidiary pointing to which attention is here called. The Metrical Pointing in Codex Jtmius XL 2 1 fjddelhiyi . beam- para pe \ceddon . loth. a.ner . and manre. ac pu most heonon . hui^e Icedan- deed rof . drihtne sinum. drehta dojora . "^ehani. he pa unetode . oncwcei^. ■Wi/lmhatne . li'^. peatvfcest . and "^epyldi"^- 'wtildor torht . ymh wucan- cniht tveox - and pa%' him pa of stum . to- vvuldor%ast . yydes. onhleotpcet \ac . ';^ode- TDLearchofu . vnorheald. forbcerned . hurhhleoou' on hwcel . htvreopdn. peahpe him on "heal/a . "^ehwam. feorh of teonda . dom£. ealle him hrimu . hlodi^e puhton. &^esful . ^Ida heamum- hlced . in Bahilonia. wur^edon . vrihpjld. frecne . fyres • tvylm. in fce^m . fyres li-^e- het pa se oym7i'^ . to him- la^searo leoda • cynin^es- hcefde onpam wundre . "^ewuroodpe . Btvi^mod sinum . leodum. dom ivear^ mfter duj^u^e • "^ecyoed' wid pees e'^esati . pyre- pcet he wi^ civealme . ^ebearh. vrite ' wealde^. pa wees eft . "^eseted- The following have extra pointing in the second half- verse : 15. Genesis 1654* 16. >» 20168' 17. ?> 2027a 18. ,, 2149a 19. ,, 2173a 20. „ 2249a 21. >» 2345* 22. ij 2584a 23. ii 2662a 24. „ 2769a 25- >» 2771* 26. yi 2911a 27. >» 2912a 28. »» 2933a 29. Exodus 6ia 30- ^j 70a 31. >» l6ia 32. »» 209a 33- j» 570a 34. »> 572a 35. Daniel I06a 36. >> 164a 37. »» 182a 38. >» 214a 39. »> 234a 40. >? 431* 41. >> 436a 42. „ 444* 43. ?> 450a 44. >» 456a 45- >? 467a 46. „ 475a 47. ,, 523* 48. ,, 641* 1. Genesis 284^ : pa ne tvilla^ me cat p am stride ■ j^eswican. 2. „ 425^ : pcet me is on mode . minum stca sar. (Thorpe, Grein, Bouterwek, all read : on minum mode.) 3. „ 706^ : pcet hep am ^ehate - "^etruwode- 4. „ 730'' : nu hie wordcwyde . his. 5. „ 755^ : Siva hivcet stca wit her inor^res -poliao. 6. „ 839^' : uton %an on pysne weald . imtan. 7. „ 1476^ : pa ^yt se eade^a . uer. 22 Dissertation on Alliterative Verse. (Grein : siva us, &c.) •ilk. Bern, and cham. onianpa his selfes . beam, ac hie earm-Uce' pees pe us aec^ea^. bee- drohta^ . secan- wlatice . moni'^e- hennum . seoce. escol .priddan. (cp. 2027*) nis woruldfeoh .pe ic me aian him pees lean.fot-^eald. he ^eseah from foldan . up. to him . selfum- ■^ud^-bordes swen"^. and him peer vom . %eseah- eac pan neah . andfeor. I^eofon deaoe . htveop- pape 071 fruman • cer^on. %if^ ' of heofmim- swa he grimmost . mihte* peet eower tela . 'geseah- pcet se weere his aldre . scyldi%. (cp. 450*) se hie ofpam jnordre . alysde- no py Bel.dijde. Vifan . of roderum. pe he mid vrilddeorum . ateah- od poet him Gwelm . ^esceod. In Genesis 2893 : ^cet he ^edcecle' swa- hine drihten het- it is difficult to see to which half- verse the scribe in- tended Siva to belong ; but I imagine it was to the first. Indeed I am not sure that there is a point at ^edcede, the end of which, as well as the s of swa, is partly obscured by a stain. In Daniel 413^ ^oit we hry* sendon- the extra point is probably numerical. So in Daniel 272 we have -Sry- with the ordinary two points: (cp. above, page 15, lines i -5). The additional point occasionally occurs before the chief letter in the second half-verse, but only when this is of an extended character. In first half- verses, with the exception of Genesis 1316% 1759* (already mentioned), I have not noticed anything similar except in : 1. Genesis 1712^ : /a wear^- tArone. 2. Daniel \<)-^^: pa wceron.ed^elum' -. 8. Genesis 1551^ 9. >» 1593b 10. „ 1692^ II. ,, 1723b 12. >> i8i8b 13- 5? 1825b 14. >» 1972b 15. 5> 2027b 16. >> 2142b 17. It 2544^ 18. 5» 2577b 19- ,, 2628b 20. ?> 2693b 21. ,, 2926b 22. Exodus 381'' 23- ,, 447^ 24. Daniel 35^ 25. 11 154b 26. „ 227b 27. >i 412b 28. 5» 450b 29. ,, 452b 30. „ 489^ 31. ,, 509b 32. n 650b 33- ,, 668b vi^i rcKOf / r The Metrical Pointiti^^'if^^Cdde^f Junius XL 23 both of which cases are doubtful. The point at weard is very small, though in the usual ink, and not merely a speck in the parchment. That at wceron as we have said (p. 15), is not of the usual lozenge-shape, but is shaped like a pear with a stem. There is a mark somewhat like it, but apple-shaped, under the a of forgeaf^ Exodus ii^ The examples I have found in second half- verses are : Z. I. Genesis 28 5 'M hie hdbba^ me ■ to "hearran ^ecorene- „ 291*' : we toille ic lerv^-his geon%ra wurdan. „ 403^ : Uton o^wendan hit nu . inonna hearnum- „ 507 b : ic ^ehyrde hine .pine deed and word. »> 558^ : /y ic wat pcet he inc . aholj^en wyr^. „ 649^ : pcet heo on%an . his wordum truwian. „ 759^ : ealle sijnt uncre . hearmas %etvrecene. „ 760^ : nu mile ic eft .pam li'^e near. „ 1 107^ : and his ■ yldrumpah. „ 1546^: nemde wceron . jfcoha. „ 1726^': no hwcei^re . gifeoe wear^ „ 2327^ : ne pearf pe pees • eaforan sceomi^an. Daniel 203^ : pcet hie pider . htveorfan wolden. „ 208^ : pe ^u pe . to wundrum teodest. „ 274^ : him .peer owiht ne derede. Besides these we have Genesis i^^^^^ pe him cefter-mund- byrde a, where mundbyrde is Grein's emendation, the passage being plainly corrupt. Mentioned as Schwellverse by Sievers in his article (RB.B. xii. 454 . . 482) are :— X. 2. Genesis 156*: Type C-^||...x-^|v^x. with zweisilbige innere Senkung, Nebentcn (lond) and Auflosmig (wegas = v^ X ). Cp. P.B.B. xii. 469. 3- » 395* '- Type B -^ || . . . x -i | . x -^, or better ^ x . . . || -^x . 1 — . (p. 471), with viersilhiger Auftakt (p. 480). 6. „ 405* : B verse {typus : 4 Auftakt + s + 2 !) — (p. 480). 7. „ 408* : Type A — x..||-^x. |-^x. with Nebenton (beam) in erster Senkung which is viersilbig (P- 479)- 18. „ 2149*. A corrupt passage (p. 476). 19. „ 2i73^=A a I with Nebenton (rof) in erster Senkung (459) 33. Exodus 570*= A a i (p. 459). 34. „ 572* : Type A, .the words ealle him being suppressed (476). 24 Dissertation on Alliterative Verse, 39. Daniel 234* : Either A, with extra foot (prefixed) x — or (?) lengthened C x-i|-^x . ||-^x . (468). 41. „ 436* : A a 2 a with Nebenton (searo) in erster Senkung (461). 42. „ 444*: Taken as a second half-verse by Sievers, and incorrectly given under sub-type A a 2 b, in- stead of under A a 4b (462). 43. „ 450* : A a I with Nebenton (mod) ; (459). 44. „ 456*: Abga: with ^m/osmw^ (duju^e) ; (464). Y. I. Genesis 284^: Ab? (p. 479). 16. „ 2142^ : Can be reckoned as A, but is perhaps corrupt (476). 26. Daniel 227^ : A a i (p. 459). 28. „ 450^ : A a 4 b (p. 462), 29. „ 452^: Ab9b (p. 464). Z. 3. Genesis 403^ : A a with dreisilhiger Auftakt (p. 480). 1 2. „ 2327^ : A a 2 b with Auftakt and Aufiosung, auf. 2. and 3. Hehung (461). 13. Daniel 203^: A a 3b (p. 462). 14. „ 208^: A a 3b (p. 462). These are all of the verses with extra pointing that I can find mentioned in Prof. Sievers' s article on the Schiuell- vers. Some of the others had certainly an equal title to appear, whilst many of those which are not actually >S^ctoe^^- verse belong to what Prof. Sievers calls gesteigerte types. There remain, however, many which are simply ordinary verses. In what follows I shall attempt to give such a description of each verse as Prof. Sievers himself would approve, referring, where suitable, to his articles in vols. X. and xii. t)f Paul unci Braunes Beitrdge. X. I. Genesis g^=SL gesteigerter E verse : E 8 a (P.B.B. x. 310). 4. „ 401a = Y I, Gen. 284 (P.B,B. xii. 479). 5. „ 403'* = a complete long verse, of two regular half- verses B + A:xx— I xxwx II -^x |-^x. 8. „ 507* = the same : composed of A^ + A: — x x | — x 1| — X X I — X . 9. „ 7 59* = a SchiveUvers : Ab 7a with Aujlosung: {P.B.B. xii. 463). 10. ,, 969*=an ordinary A verse — x I -^ x . 11. „ 1077*= „ „ J-x X \-^y. 12 ,, 1400^ = 3 Schwellvers (?) B with single alliteration cp. B 2 a (xii. 472). The Metrical Pointing in Codex Junius XI. 25 13. Genesis 1547* 14. »» 1617*= 15. 1654*= 16. 2016*^ 17. 2027*= 20. 2249a = 21. 2345*= 22. 2584*= 23- 2662a = 24. 2769*= ■■2b gesteigerter D : — x | -^-^ x . (x. 302). a doubtful A verse — x | — x {cha-am ; cp. x. 480 e.) an ordinary E verse : v^ x — x | — . a Schwellvers (?) B i, 2b (xii. 471). an ordinary A v. — x x | — x . a Schwellvers : B i e (xii. 471 and cp. x. 294 § 6). a Schwellvers (?) cp. B 2 a (cf. xii. 472 : Seef. 32). an ordinary E v.—— x | — . a doubtful gesteigerter A verse with dreisilhige Senkiing (No example in Beowulf, but cp. P.B.B. x. 310). an E verse with * Erweiterung des ersten Fusses.* (x. 309, 310). an ordinary E verse, a Schwellvers (?) cp. B 2a (xii. 472). an E verse of form — x -^ | w x . (x. 266). :a Schwellvers : C i f with Auftakt (xii. 469). gesteigerter A ; A 19 a (x. 280). „ with Auftakt : A 20 (x. 280). an ordinaiy C verse, but the passage is corrupt, a Schivellvers : B 2 with Senhung 3 + 2 (xii. 471). „ „ Aa I with Aufl(5sung (xii. 460). a Schwellvers'. C i a „ „ (xii. 469). N. B. Babilonia to be read — lonja. ■gesteigerter A : A. 17 a (x. 279). D : D. 12 (x. 305). a Schivellvers B with single alliteration ? :an ordinary B verse. -2i, Schwellvers B with single alliteration, and S'en- kung 2 + 2 (cp. x. 40). = an ordinary A verse — x | — x . „ B„ XX— IxeSx. =a full verse of four accents: A'^ plus -Lx x- (cp. P.B.B. X. 312, § 2). a Schwellvers: Ab 9b (xii. 464),— taking -truwode as -irmvde (Sievers, Grammatik, § 412, 2), or as -treoivde. B verse with second factor of compound in Sen- kung (x. 242). a Schwellvers : Aa 3b with Auftakt and Aujldsung (xii. 462). a Schwellvers: C2c with Auftakt (of two syl- lables)— (xii. 469). a Schivellvers: B 2 e (xii. 471). Cp. X. 14. 25. „ 2771*: 26. „ 2911*= 27. „ 29 1 2^= 28. „ 2933*= 29. Exodus 61 *= 30. „ 70*= 31. „ i6ia = 32. „ 209a = 35. Daniel 106*= 36. „ 164* = 37- j5 l82» 38. >» 2I4»^ 40. >5 431*= 45. „ 467* 46. M 475a 47. »? 523a 48. M 641a ". 2. Genesis 425 ■■ 706^= 4. 730^ = 5. 755^ = 6. . 839'' = 7. 8. n I476»' » 1551" 26 Dissertation on Alliterative Verse. Y. 9. Genesis 1S93^ = B. Schwellvers: B 2b (xii. 471). 10. 1 692^ = an ordinary C verse. II. 1 723^ = an ordinary B verse. 12. 1818^= „ „ A „ 13. 1825^= „ „ „ „ 14. 1972^= „ „ » » 15. 2027b = „ „ „ „ 17. 2544b = „ „ B „ 18. 2sy7^ = si Schwellvers: B2 a with Auftakt (of two syl- lables): (xii. 471). 19. 2628b = an ordinary C verse. 20. 2693b = an ordinary E verse. The point at %u^ is re- 21. »» 2926b 22. Exodus 38ib = 23- ,, 447b 24. Daniel 35^ = 25. >> 154b 27. »> 412b 30. >» 489b 31- j> 509b 32. >> 650b 33- ^ >) 668b Z. I. Genesis 285b = 2. 29lb = 4. )) 507b 5- » 558b 6. »» 648b 7^ 8. 9- 759b 760b 1107b markable. The word occurs at the end of a line in the MS. (p. 128, 1. 3), which may have something to do with the misplacement of the point. Or did the scribe wish by his pointing to bring out the subordinate crossed allitera- tion in the verse : ac ic me gumena • haldor gu^ . hordes stven"^- ? =an ordinary B verse. A » »» B »» >? A. an ordinary C verse, plus x — . Schwellvers (or ? ordinary) B (cp. xii. 471 and X. 292). a Schwellvers: A b 9 b withAuftakt and Auflosung. (xii. 464). a Schwellvers made up of an ordinary B verse (but without alliteration) + an A verse with Auftakt ; nearest example in Sievers, Exodus^ 572 (xii. 476). a Schwellvers, B 2 with Auftakt of two syllables, and Senkung 5 (? 3) + i (xii. 472). a B verse without alliteration + a B verse with alliteration. Cp. Z. 2. a Schwellvers: gesteigerter D with Auftakt of three syllables. No example in Sievers. a Schwellvers : A b 10 b with Auflosung (xii 464). Cp. Z 2 and Z 5. an ordinary B verse ; but the scribe apparently intended the words and his. to belong to the The Metrical Pointing in Codex Junius XL 2 7 first half verse, which however they cannot do without violence to Rieger's rule {Vers- hunst, p. 43). Z. 10. „ i$46^ = '? erweiterterJ). cjp. X. 2$$: Beowulf, 172^. II. „ 1 726^ = a Schwellvers: B 2b with Auft'osung (xii. 472). 15, Daniel 274^=a Schwellvers : Ab 7 a with Nebenton (or ? oht for dmht; cp. P.B.B. x. 485) and Aufldsung{xii.463). Altogether, including the examples given by Prof. Sievers, it appears that there are among the verses with extra pointing some 50 Schwellverse, 9 verses of a gestei- gerter type, and about 31 ordinary verses. It is not always easy to determine in the case of B verses the limit between a Sievers ordinary, and a Sievers Schwell- vers. Why, for example, should Beoivulf 971^ hwce^ere he his folme forlet be reckoned (P.B.B. x. 241) as an ordinary B verse, and Andreas 51^ : hwce^re he in hreostum Jja gyt as a Schwellvers (xii. 472) ? This discrepancy in the treatment of B verses has been pointed out by Cremer (Metrische Untersitchung, Bonn, 1888, p. 25) and emphasized by Hirt (Verskunst, pp. 114, 115). There is some disagreement between Prof. Sievers and his follower Dr. Luick as to the nature of the Schwell- versy the former considering that extension takes place through the prefixing of a foot to the normal verse, the latter through the addition of a foot after the csesura (cp. xii. 458). Nevertheless on p. 468 of his article Prof. Sievers hints at the possibility of a verldngertes C X- |-x-||-x- A clear example of this seems to be Y. 32, Genesis 650^, where the additional foot is of the form x - instead of - x , as usual. This verse might perhaps be referred to the B type with second factor of compound in the Senkung (x. 242), but not without a certain degree of violence. It is impossible, however, not to feel that it is only by violence that certain of these verses can be brought under Prof. Sievers's five-type classification. The first of those which he mentions, viz. X. 2, Genesis 156% wid lond'ne 28 Dissertation on Alliterative Verse, vregasnytte- is an example in point. This is put down by Prof. Sievers as a C verse with Nebenton in the Senkung (xii. 469) i.e. as -||^XwX | -x ^, in spite of his own principle, enunciated on a following page (471), that ' Nebentone schliessen sick einer vorausgehenden Hebung leichter an als eiTter folgenden,^ It is clear that the MS. division of the verse is the natural one. The same remark will apply to the division of the other half-verses, the pointing of the MS. coinciding in almost every instance with the natural division. In no case is an enclitic separated from the word preceding it, and in only one instance — Gen. lop^"]^ — is a proclitic separated from the word which follows it. Cp. X. i, 10, 48, &c., Y. 25, 509, 668, &c. In this respect it must be admitted that the scribe's sub-division of the half-verse as a rule answers better than that of Prof. Sievers, to what Dr. Heusler (Ljdpalidtir^ p. 9) says is always: Die letzte Frage der Metrik, viz. 'How have I to read?' (Wie hah' ich zu lesen?) For example Prof. Sievers would scan Dan. 154^, gife- of heofnuni' as an A verse, vS x x | - x , and again, Gen. 14CO* : /jam at niehstan- ^vces- as a B verse, x x — | x — ; in the former ease placing a proclitic syllable in the same foot with the word preceding it, in the latter separating a final syllable from its own word to connect it with the word following. This might be excusable if the alliterative verse caesura resembled that of the Greek and Latin hexameter, which divides the verse in a place where the end of a word coincides with the middle of a foot ; but as the O. E. long verse is never divided in this way, we have no reason to suppose that the half-verse is either. Prof. Sievers, it is true, never, as far as I am aware, makes actual mention of the sub-csesura, though by the way he represents the various types of half- verse (-x |-x,-|-xx &c.) he seems to imply its theoretical existence. Practically, both csesurse — principal (including middle and final) and se- * Corrected in Alfgermamsche Metrik § 95, 10 to Typus AC - x ...-- x . The Metrical Pointing in Codex Junius XI. 29 condary — would be, I suppose, ignored in the delivery if, as Prof. Sievers thinks, this resembled our modem method of recitation, and was not performed in measured time. For my own part, when listening to Racine's Androniaque at the Theatre rran9ais, I lost all sense that it was verse which was being recited ; metre and rhyme both disappeared in the flow of declamation^. Such must have been a for- tiori the fate of the alliteration, also, if the Old English verse was recited in the manner that Prof Sievers supposes, not to speak of unnatural accents, such as that on woes in Beowulf 82^, ^cet his aldres wees ende ^egdn^en. But even in France this mode of recitation is a thins: of modern introduction. M. Larroumet (Andromaque, Notice Historique, p. 37) tells us that before the time of Moliere, on chantait les vers tragiques plus qu'on ne les disait, and he quotes a remark from Talma to the effect that the actor Lekain (admitted to the Comedie Fran^aise in 1 750) follow- ing out the reform in part introduced by Baron, combated in his turn ' cette declamation redondante et fastueuse, cette diction chantante et niartelee, oil le profond respect pour la cesure faisait tomber regulierenient les vers en cadence.^ If this was the style of recitation in France as late as 1 750, it seems most unlikely that in England, in the centuries before the Conquest, the modern oratorical method of de- livery was already practised, especially when we remember that our old poems were recited by scops in some cases if not invariably themselves poets {leoHvyrhtan). Such men would have a tender regard for the form of the verses they sang — to give the correct word, and the one which we have the authority of the Beowulf itself for using (cp. vv. 90, 496) — and would certainly not deliver them so as to make what had cost so much care to construct indistinguishable from the prose of ordinary life. It is proverbial that a poet, * Since printing the above I find in the Contemporary for September, p. 388, a description of Sarah Bemhardt's mode of reciting, entirely- bearing out what I say. The writer thinks it < reprehensible (at any rate from the author's standpoint).* 30 Dissertation on Alliterative Verse. especially when giving forth his own effusions, is particular to bring out the metre. The poet Coleridge was an example of this. Mr. Collier (Lectures on Shakspere, p. 62) tells us Coleridge insisted that poetry should be read with intona- tion, and that his own reading of Spenser (for example) ^almost amounted to a song.' And the reason given by Coleridge for this was that ' a poet %vHtes in measure, and measure is best made apparent by reading with a tone ' (ib. p. 6^). Prof. Sievers's hypothesis is therefore antecedently so improbable, that until much stronger reasons have been brought forward in support of it, than we have as yet heard, we are justified in setting it aside. Prof. MoUer of Copenhagen has argued against it with a force and fulness which leave nothing to be desired (Zur ahd. Al- literationspoede, pp. 146, 149. See also Dr. Heusler's Ljdpahdttr, pp. 14, 17). Into his argument I cannot enter now, but to what has already been said I may add that the parallelism so common in O. E. alliterative verse must have compelled a strict attention to the csesurse on the part of the scop. How otherwise could such passages as Exodus 1-7, or Elene 99-104, have been made intelligible to an audience ? The occurrence now and then of enjambeme nts cannot, I submit, weigh against such a consideration as this. It seems to me, therefore, safe to assume that the ordinary points in MS. Junius XI represent actual metrical pauses which were made in chanting the verses, and further, that the extra-ordinary points which we have been considering represent the same thing. In this case it is not difficult to see on what principle, in the scribe's view at least, the half- verse was constructed. His pointing is a complete endorse- ment of the view that the half-verse is zweigliedrig (Heusler : Altd. Vershunst). The extra point marks the innere Cdsur of which Prof. MoUer speaks {Zur ahd. Alliterationspoesie, p. 120), by means of which der ger- Tnanische Halbvers zerfiel in zwei Filsse (on p. 171 note, by preference, Olieder) wie der Vers in zwei Halbverse. The Metrical Pointing in Codex Junius XI. 3 1 In other words, the half- verse is symmetrical with the whole verse. As the latter is divided by a caesura into two parts, each of which may, at will, commence with an accented syllable, or with one or more unaccented syllables forming an Auftakt, so is the former, though in this case the unaccented syllables (der innere Auftakt) will of course be light. This being so, we may arrange the ordinary half- verses (in X and Y) as follows : Inner Auftakt. I St Arsis. Auftakt. and Arsis. X. 10. Ge7i. 969a - Cain . and Abel . 12. „ 1400a })am at niehstan . - wses . 21. „ 2345a he ])a metode • on- cwse^ . 22. „ 2584*^ - wylmhatne • - lis- 23. „ 2662* - J^eawfaest • and je- J?yldij. 26. „ 2912a him |)a ofstum . - to . 48. Dan. 641* \>a. waes eft . Se- seted . Y. 10. Gen. 1692^ ac hie earm . - lice . M. » 1723^ })8es \>e us secjea^ . - bee . 12. „ 1818^ - drohta^ • - secan . 25. Dan. 154^ - Sife . of heofnum . &c., &c. Similarly can be arranged some of the verses mentioned above (§§ 2 and 3), in which the pointing comes earlier or later than we should have expected it, viz. — Inner Auftakt. ist Arsis. Auftakt. and Arsis. Gen. 651a })8et he })a bysene . from Sode „ 1401a nym^e hea waes . a- hafen „ 2247a si^^an Ajar . - \>e „ 2693a ac ic me jumena . - baldor » 53^ - mod . Se- twaefde „ 1521^ - serest . be- grinded » 1832'' J)8et J)u sie sweostor . - min &c., &c. The most striking evidence of the pause in the middle of the half- verse is that given in Gen. i6g2\ ac hie earm-lice- I may say that both points are perfectly distinct and normal in the MS. The example is unique (except for the some- 32 Dissertatioft on Alliterative Verse. what confusing point in Gen. 3693^, zud-hordes siven^-), but we may compare with it Destruction of Troy v. 3097^, halfe so luffahle, where Dr. Luick considers the doubling of the / a sign that the consonant was to be pronounced without voice, and therefore that the syllable luff- was to be followed by a pause. He says {Anglia, xi. p. 408) ' Man musste offenhar nach der ersten silbe etwas absetzen, so dass das wort fast in zwei getrennt wurde; die spirans kavi in den auslaut : daher die seltsame schreibung ^.' I quote this passage as being interesting rather than con- clusive. A stronger confirmation of the Ziveigliedrigkeit for which we are arguing is the existence of such a half- verse as the first of Exodus 118: hdr h(B^ . holme'^um wederum • Prof. Sievers, indeed, P.B.B. x. 513 says that Ex. 1 18* is kein moglickes Hemistich, which is certainly true if the five types (A, B, &c.) exhaust the forms of verse that were at the disposal of our old poets. It must be admitted that so curt a half- verse is rare in O. E. poetry. In Old Norse it was more frequent, e.g. H^v. 77, Deyr fe, H^v. 52, niihet eitt (cp. Heusler, Ljdpahdttr pp. 81, 18). But I cannot perceive that its introduction is at all hurtful to the rhythm of the passage in Exodus where it occurs ; on the contrary, it gives an impressive effect which is quite lost by substituting the genitives hares, hMes, as Prof. Sievers proposes. The open- ing verse of the well-known Scotch song 'All the Blue Bonnets are over the Border' has very much the same rhythm, viz. March, march, \\ Ettrick and Teviotdale which is sung, and can just as easily be recited, in the following time : — 1 t \ > ^ > > > ^ \ * A friend has expressed to me doubts whether Dr. Luick's inference is here sound ; I may mention, however, that Mr. Sweet (Hist. Eng. Sounds, p. i66j remarks. 'In efffei- the doubling of the/ may mean voicelessness.' The Metrical Pomting in Codex Junius XL 33 Here the monosyllable Diarch is sung in the same time as the three syllables Et-trick and, and there is no difficulty in supposing that in the verse from Eaxxius the mono- syllables hdr, hce^, were intended to occupy the same time as the trisyllables holmegUTYi, wederum. We may remark too that holnie^urn is plainly equivalent to wederum (though the first syllable of the former is long), in accordance with the principle laid down by MoUer (Althd. Alliterationspoesie, p. 11^), th&ijede Folge von zwei Silben (w plus w, — plus w) wenn nur die Nehenhebung nicht aufdie zweite, soiidern erst auf die folgende Silhe fdllt, gleichwertig mit — (ist). The remarks just made trench upon the difficult question of Taktgleichheit, which is too large for the compass of this chapter. In the one following, I shall endeavour to show that the construction of those verses in Beowulf, which have ' crossed alliteration,' lends some strength to the view so ably maintained by Prof. Moller (and accepted by Dr. Heusler) that alliterative verse is metrical in the strictest sense of the word. At least one example of a half- verse as condensed as hdr hw^ occurs in Beowulf, but it has been obscured by an emendation which has brought ruin into the most pathetic outburst in the whole poem, viz. the wail of Hro^gar over the death of iEschere (vv. 1323-1330). Here the original is incomparably fine, not a word is lost from the startling, Ne frin piX cefter scelum ! at the outset (in response to the cheery morning greeting of Beowulf), to the closing eulogy over the dead warrior : Simple scolde eorl wesan ebr-idd, sioylc ^sc-here woes. The editors have introduced ce^eliT}^ (!) from vv. 130* 2343* {ce^eling dkr^dd), in which the word is quite in place, whereas in the passage above it would be the merest padding"^. * Another instance in which the strength of the original appears to me to have been grievously watered down is B. 2489*, hreas hide [Editors ; 34 Dissertation on Alliterative Verse. Another question which is affected by the existence or not of the inner csesura, but which I must only glance at here, is that of the number of accents in a half- verse. This was the point at issue in the famous controversy between the Vierhehungstheorie and the Zweihehungstheorie, which originated in 1848, with Wackernagel {Litteraturgeschidde, § 25), who first controverted the view of Lachmann that the alliterative half-verse contains four accents. Lach- mann, it is true, excepted O. E. verse from the rule, but Schubert {De Anglosaxonwni arte 7)ietrica, 1870) endea- voured to bring this also under it. His argument was supposed to have been triumphantly set aside, and the Zweikehungstheorie finally established by Vetter in 1872 {Zum Muspilli, pp. 3-25); but the question has again raised its head and opinion is still unsettled with regard to it, as may be very clearly seen from a single example. The half- verse, Beoiuulf, ic^, h^ran scolde, which is the first given by Prof Sievers under the A- or Grundtypus, possesses according to Moller (p. 128, note), and Fuhr (Metrik des westgerTn. Allit.verses, p. ^y) four accents {—k I - ^ ) ; according to Hirt ( West germ. Verskunst, p. 8 1 ) three accents, — x | — | — ; according to Sievers (P. B. B. X. 222) two accents, —x | - x , &c. ; according to Hinze {Zum Andreas, p. 10) only one accent, - x x x ! Dr. Hirt (cf p. 49) sees a proof that there is an accent on the second syllable of scolde, in the existence of A verses with ' zweisilhige Senkung im zweiten Fuss,' e. g. o^res do- gores (cp. P. B. B. X. 233), inasmuch as we must neces- sarily read dS-gores, and do-gres and do-gores are metrically equivalent. I should see it rather in the pause at the end of the half- verse. Final syllables which have naturally no hreas hewoAAkfi]. Retaining the MS. reading, and breaking the half- verse with a pause of 2 Morm (see p. 36) at hreas, the hide is brought out with a force that reminds one of \irg\V& procumhit humi bos (Aen. v. 481) ; or we may compare Tennyson's The Prince wWiout a word from his Juyrsefell (Geraint and Enid, ii. 519^. Cp, also Beoimdf 653% 3087*. The latter verse, if I interpret it rightly, is as dramatic a touch as anything in the poem, and at the same time a fine example of reticence. The Metrical Pointing in Codex Junius XI. 35 accent easily acquire one before a pause in measured recitation, as we may notice whenever we hear the English liturgy read in our churches. So again in a Roman Catholic church, during the reading of a homily of St. Augustine's, I have heard the second syllable of hofiio pronounced in all three possible ways : with a principal accent (hdmd) before a full stop, with a secondary (hdmb) before a comma, and with no accent at all {hdmo) when not followed by a pause. And this is, I believe, the immemorial practice. Prof. MoUer gives practically the same explana- tion as my own, saying: * Meine Accentsetzung bezieht sich uherall auf den gesungenen oder im Takte recitierten Vers' (p. 128, note). If, then, the half- verse h'pran scolde was recited with two pauses — a middle and a final — as we believe, it must have been recited also with four accents, two primary and two secondary, thus : h{/rdn scdlde. I may conclude this chapter by giving an extract from Beowulf with full points and accents inserted : — — Beowidf ig4-2i6. pset fram ham . gefrgegn . Higelaces . \>egn . 195 ^6d . mid Geatum . Grendles . dseda . se W8es mon . cynnes . msegenes . strengest . on p^m . dsej^e . J?;fsses • Iffes . ge'Sele • ond eacen . Het him y^ . lidan . jo-dne . J^e^yrwan . cwse^ he Ju^ • c;^iiin^ . 200 ofer swan . rade . s^cean . wolde . mserne • J)eoden . J; a him waes manna • |)earf . ))one sfS-fset . him . snotere . ceorlas . 1^-hwbn . logon . Jjeah he him leof . wsere . hwetton . higerbfne • hs^l . sc^awedon . 205 Hsefde • se goda . Geata . leoda . cempan . jecdrone . ))ara \>e he cen-oste .^ findan . mihte . fiftena • siim . siind-wudu • sohte . secg . wisade . lagu-creeftig . mon . land . ^em^rcu . 210 Fyrst . forS jewat . flota webs . on y^um . bat • under beorge . Beornas • gearwe . on stefn . stigon . streamas . wiindon . Siind . wi^ sande . secgas . bseron . D 2 ^6 Dissertation on Alliterative Verse. o on bearm . nacan . beorhte • frjfetwe . 215 ju^-searo • jeatolic . jiiman . ut sciifon . weras . on wil-si^ . wiidu . biindenne . It will naturally be asked : Are the pauses here represented by dots to be all of equal duration % This again raises the question of Taktgleichheit, which cannot in this place be fully dealt with. Some answer must, however, be given. According to Prof. MoUer (p. ito) the half- verse consists theoretically of two Takte, each of four Moren: )< x x x (a Moi^e, X , being the time required for one short syllable) ; and therefore the whole verse of four Takte, thus : )^xxx I )kxkx\\:^xkx \ :< x k x \\ Where^ in a verse, the Moren are not filled by actual syl- lables, their time must be occupied by rests (represented by r) in reciting, by holding on the note in singing. A long syllable, — , is equivalent to two Moren. Thus, verse 208 above would be symbolically represented as follows : — kx |-^xr|| — rr|— Jcx. According to this system the pause at sec^ will be twice as long as that at sohte, whilst at wudu there will be no real pause, and the point will merely indicate the end of the Takt. In like manner the pause at hcefde 205^ is merely nominal, the time being filled up by the word in Auftakt {se) following it. Another point of great difficulty, which I must postpone for the present, is the proper treatment of the Auftakt '^ when this extends beyond a single syllable, as in 194*, I98^ 20 ^^ Hirt and Fuhr make systematic feet out of the alternation of accented and unaccented syllables which necessarily exists in such cases. But this cannot be done without disturbing the alliteration hitherto regarded as the hinge upon which the whole verse turns. Hirt how- ever insinuates that its role has been over-estimated. ' Meaning by this term all the syllables preceding the first alliterating letter. Prof. Sievers distinguishes between Auftakt for A and D verses, and Eingangssenkung for B and C verses (cp. P.B.B. x. 215). The Metrical Pointing in Codex Junius XL 3 7 Speaking of the various ways in which verses like 205* (Sievers A^) have been regarded, he says (p. 41): — ^ Dieae Gegenilberstellungen zeigen aber, dass die Alliteration bei der BestimTnung dieser Verse nicht allein massgebend ist, ja sie wagen es, utis zuzuraunen, dass sie wohl ilberhaupt nicht eine so bestirriTnende Bedeutung habCf als man bisher anzunehmen geneigt war,^ A dangerous suggestion this, to cut the thread upon which the verse is strung, and we shall do better, I think, to con- sider that the accents in this introductory portion of the verse were simply ignored, and the words themselves hurried over as quickly as might be. Poets, after all, must work with the materials they have, and were we never to allow a natural accent to be neglected, what would become of our English anapsestic verse, e. g. of such a line as Browning's, And grow one | in the sense I of this world's | life.— And then | the last song I CHAPTER II. Crossed Alliteration in Beowulf ^ The normal O. E. verse contains three alliterating syllables, two in the first half and one in the second, as in Beowulf 4 : Oft Qeyld Soe/in'^ aoea^ena preatum. Often the first half'-verse has only one alliterating syllable, which may be in the first arsis as in Beowulf 2 : \e6d-cynin'yi prym %efrunon or in the second, e. g. Beowulf 22 : poet hine on -ylde eft ^ewuni'^en. These types of alliteration may be S3rmbolically represented thus : — (i) a I a II a J - (ii) a | - I| a | - (iii) - | a 1| a | -. The only syllable which invariably alliterates is the one in the first arsis of the and half-verse, and its initial has accordingly been called the chief-letter, as governing the alliteration of the verse. The name is taken from the Icelandic hgfu^stafr, which is the term used by the old gi-ammarian Snorri Sturluson. The fourth arsis of the verse is excluded from the regular alliteration. Occasional examples, however, occur of an alliterative scheme a | b || a | b, as in Beowulf i : Hwcet I we Gar-Dena in gear-da'^um and also of a scheme b | a || a | b as in Beowulf ^^^ : VTit pwt "j^eQWcedon cniht-wesende * The edition used = the Heyne-Socin, 1888. Crossed Alliteration. 39 (cp. Rieger, Verskunst, p. 5). Rieger found some 60 (einige sechzig) of the former type and 1 8 of the latter in Beowulf. My own counting gives the numbers as 82 and 24 respect- ively. It has been supposed by most writers since Rieger that these types with crossed alliteration exist only in appearance. Thus Horn (P.B.B. v. 1 66) contended that they are entirely contrary to the nature of alliterative verse, and must therefore be quite unintentional, and Frucht (Zv, Cyneiuulfs Elene, &c., p. 75) endeavoured to show that the occurrence of such coincidences between second and fourth arses as -d.ena, -da^uni, is less frequent in Cynewulf, at all events, than the doctrine of Probabilities would lead us to expect. Frucht's calculation has been, so lately as this year, accepted as probably correct by Fuhr {Metrik des westg. Alliterationsverse8,-p. 25), but it involves the extraordinary fallacy of treating all possible initials (Anlaute) as equally likely to occur in the final arsis of a verse. Of such An- laute there are 1 9, viz. h, c, d, /, ^, h, I, m, n, p, r, s, sc, sp, at, t, /, w, and any one of the vowels. It therefore follows, says Frucht, that the chances of such a coincidence as that in Beowulf I are one in nineteen^. Now there are 1406 verses in Cynewulf in which crossed alliteration is possible (that is, in which the first half- verse has only single allitera- tion); therefore we should expect to find 14064-19 or 74 instances of it among them. There are however only 64. Thus there are fewer examples of coincidence than we should naturally expect, apart from any design in the writer, and we must therefore infer that such as do occur are accidental. This is Frucht's argument. To my mind it would be quite inconclusive, even if it were mathe- matically sound. Who will suppose that the crossed allitera- tion is undesigned in the powerful line of In Memoriam ^Vl. 4 j : j£-g lieavy-Bhotted hammoch-Bhroud ? ' Clearly one in eighteen, as one of the nineteen Anlmde is wanted for the chief letter. 40 Dissertation on Alliterative Verse. Yet, strongly alliterative as the poem is, such an example does not, to my knowledge, occur again in it. (Though compare Canto xxi. 4, 2.) The calculation, however, as we have said, rests on a false assumption. A very short examination is enough to make it clear that some initials (e.g. h and w) occur much oftener in the final arsis of a verse than others (e. g. than r and \\ and that, consequently, to place all the letters on the same footing, as Frucht has done, must be erroneous. If we are to apply the doctrine of Probabilities with any safety we must take the letters one by one. Even were we to encounter this task, however, it is probable that Mathe- matics would refuse to give a verdict upon data so scanty and so little rigorous as we possess. It will be more to the purpose to collect the verses in which crossed alliteration occurs and see what impression they produce when read together, and if they have any peculiar characteristics marking them off as a class by themselves. Taking first the type a | b || a | b, we have in Beo%vul,f the following examples : — i. a = Vowel, b= consonant : viz. V : b ^<5 %c%-hanan kn'^an brewer, Beow. 1 263 ; also V : c. v. 70a; V : h 1 183; V : 1 566, 1489; V : m 908 ; V : r 1202; V: s 2516; and V : w^ 131 5 ; in all 9 examples with vowel-alliteration in ist. and 3rd arses, and consonant-alliteration in 2nd and 4th. ii. (a) b + consonant : viz. b : c sed pe hdn-cofan heor^an Gu^e 1446, also b : h 3148 ; b : n 343 ; b : st 1920 ; b : w 2163, 2908, 3067 ; in all 7 examples. (b) b + vowel: v. 1969 i ex. iii. Similarly c : h 98 ; c : w 282, 2067 . . . . 3 ex. iv. d : h 88, 2727 ; d : w 2187 3 ex. v. f : b 2466 ; f: h 1850; f: m 3163 . . . . 3 ex. vi. g : d I, 1825 ; g : f 592 ; g : h 3075 ; g : 1 830 ; g : m 2268 ; g : n 804 7 ex. vii. h : b 654, 1444 ; h : g 64, 374 ; h : 1 1 144 ; h : st 32, 1 133; h :w39, 1938,2999 10 ex. viii. 1 : h 2031 ; 1 : m 209 ; 1 : w 972 ; 1 : }? 34 . . . 4 ex. ix. m ; c 418, 2182 ; m : g 1302 ; m : h 731 ; m : s 2092 ; m : }? 201 ; m ; w 2480 7 ex. Crossed Alliteration, 41 X. n : h 2398 ; n : s 1204 2 ex. xi. s : f 1476 ; s : g 1343 ; s : h 920, 1017 ; s ; 1 3090 ; s : m 1612, 2955 ; s : w 2187, 2875 9 ex. xii. sc + vowel 19 ; sc : m 1940 2 ex. xiii. t : m 1141 I ex. xiv. ]? : c 2971 ; f> : h 2236 ; |:> : 1 2638 ; 1? : m 365, 1342, 2722 ; fJ : n 2225 7 ex. XV. w : b 2568 ; w : d 590 ; w : f 2262 ; w : g 1996 ; w : 1 2746 ; w : s 1404 ; w : J? 525 7 ex. In all 82 examples. Whether the poet sought for it or no, the crossed allitera- tion in these verses has a sensible aesthetic value, as it also has in Tennyson's line quoted above. The * singularly beautiful effect' of the 'double' alliteration occasionally found in Milton is pointed out in English Lessons for English People by Messrs. Seeley and Abbot (p. 182). Schipper quotes Hamlet, I. 2, 65, A little more than kw, and less than laind, as an example from Shakespeare, but erroneously, as the emphatic accent is on Tnore which is in antithesis to less^ whereas little being an adverb of degree is proclitic (Rieger, p. 39). The example which Seeley and Abbot give from Shakespeare is equally defective : viz. Our dreadful marches to delightful measures. Richard III, Act I. sc. i. v. 8. as delightful can properly alliterate only with the 1. Zeu- ner in his treatise on alliteration in modern English poems gives better examples of the scheme a | b || a | b, which he considers entschieden heabdchtigt. If, therefore, we recognise design in the scattered in- stances where this form of alliteration appears in modern poetry it is not reasonable to think that the author (or authors) of Beowulf, a poem which, as Rieger well says (p. 13), appears under every aspect with the stamp of a classic, fell merely by accident into crossed alliteration in the cases cited above, or that so artistic a composer as 42 Dissertation on Alliterative Verse, Cynewulf was not fully aware of its effect in the instances mentioned by Frucht. A circumstance which strengthens this assertion is the tendency of an initial to repeat itself at the close even of successive long verses. The commonest initial in the fourth arsis is w, its appearances in this position being distributed very evenly throughout the poem. Evenly, that is to say, for large sections of 500 verses, but within these sections the occurrences are very irregular, often coming in patches of three or four, and then ceasing for intervals of 20, 30, or 40 verses. Thus the 35 verses from 247 to 281 do not once contain this ending, whilst the next eleven verses, 282 to 292, have it five times. It is entirely missing again between 1 1 40 and 1 1 70 and between 1600 and 1645. Often after an interval it comes in couplets, and sometimes in triplets as in vv. 1795, 1796, 1797; 1864, 1865, 1866; 2161, 2162, 2163 ; 2185, 2186, 2187 ; 2780, 2781, 2782. We see the same inclination of the poet to harp, from time to time, on one string in the case of other letters. There is a quartette of c's in 2045, 2046, 2047, 2048 ; a triplet of I's (2935, 2936, 2937); of Ks (1722, 1723, 1724) and of st's (1373, 1374, 1375), the last a by no means com- mon initial. It must be borne in mind that such alliterative end- rhymes are not a mere ocular effect, as we might be tempted to imagine when looking at the verses in a modern edition of Beowulf. The ' poet's eye in a fine frenzy roUing ' was probably not disturbed by such mechanical details as pens and paper. However, even if they were written down at the time of composing, the verses would not be arranged in lines, if we may judge from our MSS. in which they run on just like prose. But if this rhyming tendency could act after the in- terval of a whole verse, it could certainly do so when merely a half- verse had intervened. We have now to answer the a 'priori objection of Horn, Crossed Alliteration, 43 that crossed alliteration is entirely contrary to the nature of alliterative verse. So indeed it seems, at first sight, to be. The final arsis is ordinarily excluded from the alliteration because the expiratory force, which naturally declines towards the end of the verse, is not strong enough, without a fresh exertion, to give effect to it ^. We may however argue that the extra alliteration, though real, is of course less emphatic than the principal, and in the second place we may enquire whether there cannot be something in the construction of the verses which have it to justify their partial deviation from the ordinary rule. The following table supplies an answer to this question. The verses are divided into sections according to the arses. The first column of each half-verse contains the words (if any) which precede the first alliterating letter : — a. b. AuflaU. \st Arsis. iznd Arsis. Auftakt. 3rd Arsis. 4th Arsis. (a)B. I Hwaet! we Gar- Dena in gear- dasum V 972 t6 lif- WraSe . last Weardian » 1263 t6 ecs- banan . . ansan breSer „ 1343 se J)e sefter Sinc- gyfan on Sefan gredteO „ 1404 aefter Wald- SwaSum Wide gesyne „ 1446 8e6 Jje ban- cofan beorsan case „ 1996 J)8et J)(i J)one wael- gsest wihte ne grette » 2031 sefter le<5d- hryre lytle hwlle „ 2262 sefter wls- fruman wide feran „ 2466 on l)am feorh- bonan fsfehSe Sebetan „ 2480 J)aet mses- wine mine Sewrsfecan » 2516 of eortJ- sele lit SeSeceS „ 2638 5*1 f him pys lieu pearf Selumpe n 2971 sySSan pe<5d- Cynins pyder oncirde » 3075 nses hS gold- hwaet gearvvor hffifde » 3090 swses- lice siS ^yfed » 3148 o»>8etheJ)a ban- hfts se brocen hffifde M 374 y^va. td ham forgeaf HreSel Ge^ta „ 908 swylce Oft I owe this ( bemearn jxplanation to Prof. M5ll< &nan 3r. maeluni 44 Dissertation on Alliterative Verse. a. b. AujiaU. 1st Arsis. 2nd Arsis. Auftakt 2rd Arsis. ^th Arsis. (a) B.I 141 Sifhe torn- Semot J)urh tedn mihte (b)„ 64 J)& waes Hr6S- gare . . . here-sped gyfen „ 282 ond J)a Cear- wylmas . colran wurSaS „ 566 be n- lafe uppe l^Son „ "44 ))onne him Htin Lafins hilde ledman 5, 1315 hwaeSrebim al- ■walda ^fre wille ,, 1444 scolde here- byrne . hondum Sebroden „ 2067 aefter Cear- Wselmum coh-an weorSaS „ 2182 ach§ man- Cynnes mseste crsefte J, 2225 ac for ]>re^- nedlan l>e(6w) nat-hwylces „ 2727 )5sethe cUes- hwila Se drosen haefde „ 2908 ofer Bi<5 wulfe byre Wihstanes „ 2955 J>8ethd S£b- mannum on sacan mihte » 3067 swa waes Bid wulfe J)ahe biorses weard „ 343 . bedd geneatas . Beowulf is min naiiia » 1938 • • • haud gewriSene • • • hraSe se- oSSan waes ,» 2163 . . . bre<58t- Sewafedu. • • • Brtic ealles well (c),, 920 * * * switJ- hicsende td Sele pkm hean (cp.1017) „ 1204 nefa Swertinses nyhstan SiSe „ 1969 bonan On5enJ)e6es bursuin in innan „ 2236 J)anc- hycsende . p^r sehydde T> 2999 • • • ham- weorSunse hyldo to wedde (d) „ 32 ])sfer set hySe St6d . . ■ hrinsed- Stefna » 34 a- ledon pa leofne pedden n 365 J)set hie pedden min wis pe mdton „ 418 forjjan hie maesenes craeft minne ctfSon » 700 J)urh £ines craeft . ealle oferCdmon » 1302 aefter maSSum- gife m^runi Ge^te n 1342 ),*8>e ]>incean maes . . pesne monesum „ 1489 ondI)k his m6d ahl65 „ 804 . gflS-billa nan • . gretan nolde Crossed Alliteration. ■ 45 a. b. AufUiM. \8t Arsis. 2nd Arsis. Auftakt. Srd Arsis. 4th Arsis. (e) B. 830 Qe^t-mecsa Ie6d . gilp Sel^sted ,, 3164 . . . fore-snotre men . . . fin dan mihton (f)„ 590 ^ werhSo dredsan pedh >in wit duse M 654 . HrdSsar Bedwulf ond him hsel abesid ,1 1 183 . . a,rum healdan gjfn &r Jjonne he „ 1612 ssbla ond m^la J>8etis 868 metod » 1825 gumena dryhten ])onne ic sft dyde ,, 1850 folces hyrde ond J)(i J)in feorh hafast „ 2568 ■winia bealdor J)ase wyrm sebe^h „ 2746 . . . Wlslaf ledfa nt se wyrra liseS (g)., 39 hilde- Waepnum ond heaSo- W^dum V 98 eynna Sehwylcum J)ara J>e cwice hwyrfaW „ 1911 bunden- stefna ofer brim- stre^mas ,,2187 drihten wereda se ddn wolde „ 2188 Sw^Se (■w6n)don ])8ethe sleac W^re '!i)„ 20 r mserne pedden ]ja him waes manna pearf „ 1202 Eonnen rices Sece^s 6cne r^d n 1476 Snottra fensel n(!l ic eom BifSes ffts „ 2092 manisra Sumne hyt ne mihte Swa ,. 2876 sisora Waldend J)8et h$ hyne sylfne SeWraec „ 19 Scyldes eaferan . . . Scede- land-um in 0) V "32 hrinsed- Stefnan . , holm storme wedl „ 2722 pedden m^nie . . . pesn unsemete till We will begin the consideration of the foregoing examples with some remarks which Frucht makes upon double and single alliteration, i. e. upon the regular types (i), (ii), and (iii), viz. a | a || a | — &c., mentioned at the outset. He says (p. 76) ' Was den Gehranch der doppelten resp. ein- fachen Alliteration im ersten Halhverse hetrifft, so ist derselbe nicht willkurlich ; . . . in Idngeren oder durch Nebenicten gesteigerten Verstypen erscheint die einfache Alliteration bei weitem nicht so hdufig wie in den kiirzeren oder leichteren Formen; sie wurde hier eben ungenilgend befunden, Wollte man nun die gekreuzte Alliteration a Is 46 ■ Disseidation on Alliterative Verse. eine besorulere Kunstform von der einfachen scheiden, so niilsste Tnan naturlich verrtiuten, dass audi die gekreuzte Alliteration hauptsdchlich in gewichtigeren Versen ihre Stelle kdtte ; aber gerade das Unigehehrte ist der Fall; wo die doppelte Alliteration sich Tnehrt, schwindet ivi allgemeinen die gekreuzte^ sie steigt und fdllt mit der einfachen Alliteration, sie ist eben ein Teil dieser letzterenj The facts here stated are correct, but the commentary upon them is misleading^ and seems due to some confusion of thought. Where double alliteration is present in the first half-verse, crossed alliteration is obviously by that very circumstance excluded, and the verses with crossed alliteration must therefore be a part of those which have single alliteration in the first half-verse. The question is, do they form a separate class amongst these 1 Verses with double alliteration are as a rule heavier than those with single, as Frucht rightly observes. But in saying that we should therefore expect the verses with crossed alliteration, if such a thing exists, to be likewise heavier than those with single alliteration, he makes a questionable analogy. Crossed alliteration involves alliteration in the final arsis where, as we have said, the force of expiration is naturally weakest. If therefore a verse has alliteration in this place, we ought to expect that for some reason there has been less than the usual expenditure of breath in the earlier portion of it, so leaving a reserve of force of which the final arsis has taken advantage, in other words, that the verse has been lighter than the average. In the table, the examples are arranged according to Professor Holler's takte (see p. 36), which brings the alliterating words out into strong relief. It will be seen that for the half-verses a a, bf, both takte are of the mini- mum weight, and that for the other sections ab, ac, &c., only one takt is so in each half-verse. Nowhere is this minimum takt absent throughout the whole verse ; in one instance (B i) it occurs all four times. So light a verse is Crossed Alliteration. 47 very rare ; we have one such, however, in Cynewulf (Juliana 384), also with crossed alliteration, viz. : void tldn prcece, nele feor ponan. In the a half- verses of sections a, b, c, d, e, there is, as far as I can find, no peculiarity, as regards lightness or anything else, to distinguish them from ordinary first half- verses with single alliteration. The case is different, how- ever, as regards the b half-verses. The most striking thing about these is the preponderance of verses of the Sievers A type and the fewness of the relatively heavier verses of the D and E types. For whereas, taking the whole poem, the D and E types together are more than half as numerous in the second half- verse as the favourite A type (cp. P.B.B.x.T^ip. 262, 268, 290), here they are together little more than an eleventh part as numerous as the A verses. And not only are the heavier D and E types avoided, but the A verses themselves in these sections are either of the simplest possible form, — x | — x , or of the form — X X I - X with a light proclitic, most often simply ^e-, as the second syllable of the senkung after the first arsis. In only one instance (B 700) have we the form -XXX I -X, and in this elision is possible. For the whole poem the relative frequency of the type — X | — x to -XX ] -X is 720 : 297, in the sections (a) . . , (e) it is 34 : 9. Similarly the a half- verses of sections (/), (g), (h), and (i) are almost without exception of the simplest A type. These considerations may serve to show that the crossed alliteration of the type a | b || a | b in Beowulf is not a merely haphazard thing, but that its occurrence is subject to certain rules of rhythm which the poet, consciously or not, has obeyed. Whether the scheme b | a || a | b, also mentioned by Rieger ( Verskunst, p. 5), has any real existence is a more difficult and at the same time a less important question. 48 Dissertation on Alliterative Vei'se. Less important, because in any case the effect of such a secondary alliteration can only be slight, inasmuch as it takes place, if at all, between the weakest and most widely separated arses in the verse ; and more difficult, for its occurrences are much less frequent and more open to question. In some cases, nevertheless, the rime seems fairly strong, e. g. in — B. 1223 Qolne wtde-ferh^ weras ehti%a^ „ 1934 ncenii, poet dorste deor jene^aw, and especially in — B. 2616 hrun-fd'^ne helm h.rin'^de hyrnan. The last verse should, however, probably be read hrHn-fdyie "helm hyrnan h.rin'^de (Rieger, p. 21.) [since briln-fdgne cannot stand where it does without alliterating with the chief letter — Rieger, p. 19], and this gives us a verse of the a | b || a | b type. Similarly B 780^ should be ^nig manna, which would give us another example of the same type. The following are all the apparent examples of the second form of crossed alliteration which I have noticed in Beowulf: — b. Type of each Half- verse a. iSievers). . 355 pemese goda dgifan pence^ (a) A3 (b)C 535 '^it pc^i gecwcedon cniht wesende „A3 » D j^o{^)pcBt hit a, mid gemete jsianna mnig „B „A 1 185 wene ic, poet he mid gode gyldan 'wille „A3 „A 1223 ealne wMe-ferh^ weras ehtigacf „ D gest' •„D 1483 Bwylce pu pa madmas pepu me aealdest „A' „c 1574 Yiwear-f pa he vrealle w&pen hafenade „A» „ D gest'. 1722 \>cet he pees gewinnes weorc prowade „A» „D 1729 hwtlum he on \ufan \cete^ hworfan „A« „A 1733 gede^ him swd gewealdene worolde dMas „A3 »A 1 827 gif icpcet gefricge ofer fldda hegang ,,A3 „B 1934 ncenigpcet dorste deor gene^an „A3 „ A 2021 htmlum for dugu^e dohtor iLro^gdres „A3 „Dgestr. 2I59(?) Qwced pcet hyt hcBfde 'S.iorogdr cyning „A3(?) „A » 2338 heht him pa gewyrcean -wtgendra hied „ A3 „E >} 2378 hww^re he hiiie on folce freond-ldrumheold,, A^ „E >» 2386 hep^rfor feorme feorh-utcnde hledt „A3 V E Crossed Alliteration. 49 V. 2407 s« wees on J^dm predte /)reotted^a Becg a. A' b. E ,, 26i6{?)'bf'un-fdgne helm hringde hyrnan , E mA „ 2708 ond hi hynepd hegen dhroten hcpfdon , A» „c „ 2974 ac M him on \iedfde helm ser gescer , A^ mB „ 3082 \>cet he ne grette gold-weard pone , , A' „A ,,3166 eall sivylce hyrsta swylc on horde ser . , A-' mB „ 3182 cwMonJ^cet he w^re woruld-cyninga , A' „D Twenty-four in all. I have omitted one of those cited by Rieger, viz, Beowulf 2983, Jfd wceron mowije /e his m«j wri^on, as I prefer, with Sievers (P. B. B. x. 285), to accent ])ct wceron m.6ni^ey &c. On similar grounds some of the examples in the above list would, no doubt, be omitted by others. Modern English verse furnishes some examples of the b I a II a I b scheme, which Schipper (Englische Metrik, II. I, 74), following Zeuner, calls ' Umschliessende Alliteration,'' a term for which it is difficult to find a good English equi- valent ; Seeley and Abbott (p. 182) call it a kind of verbal proportion. Instances quoted by the latter are : The hallow'd taper trembling in thy hand. (Pope.) One laced the helm, another held the lance. (Dryden.) But such alliteration is rare and, as in the cases just given, of small value. Zeuner, who considers the Reinistellung abab in modem English poetry certainly designed (ent- schieden beabsichtigt), is doubtful as to the arrangement baab which he says is kauni beabsichtigt (pp. 34, ^^). Horn (P.B.B. v. 167) remarks that the objections to the scheme a | b || a j b hold a fortiori with regard to b | a II a I b ; nor is it possible to reconcile the latter with the nature of alliterative verse by the argument used in the case of the former, as the first half- verses of the verses with the arrangement b | a || a | b, instead of being lighter, are considerably heavier than the average. Cp. B- 1733 S«d«<^ him swd ^ewealdene vrorolde dcelas „ 1 185 wene ic, pcei hS mid gode gyldan Vfille E 50 Dissertation on Alliterative Verse. whilst no less than eleven of the second half- verses belong to the D, E types. Before passing away from this subject I will mention that in some verses there is apparent alliteration between a syllable with secondary accent in the first half-verse, and the final arsis of the second, e.g. Beowulf (^'^^: (a) \e6da \and-ieweorc Id^um beweredon or between ai;sis in the first half- verse, and secondary accent in the second, e.g. Beowulf 1495: (b) hidan •wolde 'brim--wylm (ynfSn%, or between secondary accents in both half- verses, Beowulf 251 1 : (c) "Beowulf ma^elode hedt-wordum sprcec. and once between secondary accent in first half-verse and chief-letter, viz. Beowulf 445 : (d) m-cegen Hre^manna nd pu minne pearft. Like a are verses 33, 4^, 6^, 208, 288, 305, 350, 490, 505, 756, 1162, ii64,.i2oi, 1244, 131^. i3^o» 1401, 1421, 1455, 1461, 1595, 1651, 1703, 1706, 1800, 1853, 1938, 2109, 2121, 2148, 2162, 2171, 2232, 2286, 2301, 2536, 2589, 2591, 2732, 2820, 2988, 3085, 3106, 3150, 3167; like 6, 1067, 1484, 1495, 2101, 2171, 2283, 2483, 2591, 2959; like c, 131, 236, 988, 1066, 2594, 2966. Whether such rimes were designed or not by the poet, and emphasized or not in the delivery, it is impossible that they should have been without some effect. They must have been heard by ears far quicker and more trained than ours to catch alliteration, as a kind of under-current in the melody of the verse. Close examination would, I am convinced, reveal a far greater wealth of alliteration and assonance in the verse of Beowulf than we notice in casual reading. It is upon these two elements, as Mr. R. L. Stevenson {Contemporary Review, April 1885) so eloquently teaches us, that the Crossed Alliteration. 51 beauty of the contents of a phrase, or of a sentence, im- plicitly depends. ' The vowel demands to be repeated ; the consonant demands to be repeated ; and both cry aloud to be perpetually varied. You may follow the adventures of a letter through any passage that has particularly pleased you ; find it perhaps denied awhile, to tantalize the ear ; find it fired against you in a whole broadside (cp. what was said about w in final arsis) or find it pass into congenerous sounds^ one liquid or labial melting away into another.'' Mr. Stevenson gives an exceptionally fine example of this from our greatest poet, ' exceptional in Shakespeare — ex- ceptional indeed in literature : ' The harge she sat in like a hnmished throne Burn^ on the water: the poop was "beaten, gold T?nrple the sails and so pvafumid that The tvinds were lovesick with them. After these splendid chords an3rthing from our old epic must sound a little flat, but I will venture to place along- side of this passage Beowulf 320-324. In order to give greater prominence to the subordinate alliteration, the principal alliteration is here not marked. Street wees stkn-jWi sti'^ uAsode juwum cBtyBdere. Gti^-ljrne sctn heavd hond-loceny hrin^'iren sdr son^ in seavtvum, pd hie td sele fuvoum in hyra "^ryre-^eati^im j.a,ngan cwomon. In these verses the chiming of liquids and nasals is even more distinct than in the passage from Shakespeare, as are also the vowel assonances, and though the whole effect is no doubt less harmonious, yet for the purpose in hand, the representation of warriors marching in their armour, nothing could be better calculated. The balanced swing of the metre also aids, so that we are reminded, in Mr. Sk cat's words, of the ' regular tramp of an ai*my on the march ' (Essay on All. Poetry, p. 35). Nevertheless, it was apropos of this very passage that Mr. E 2 52 Dissertation on Alliterative Verse, Wadham remarked, certainly with more wit than discretion, ' Surely the Anglo-Saxon lyre must have been a gridiron or some instrument not more tunable.' {English Versification, p. II. Longmans, 1869.) Finally, there is one case in which we may without doubt regard the repetition of an initial as purely casual, and needing to be as much as possible slurred over in delivery : viz. where it occurs before the chief letter, as in Beowulf '^^^. Jesaja him eac vrordum pcet hie sint wil-cuman Here if any accent were given to sint we should produce a kind of crossed alliteration of a type b | a || b | a, with the minor alliteration in the first place, which would pro- duce an intolerable confusion in the minds of an audience. This seems to be an insurmountable difficulty in the way of those metricians who introduce accented syllables before the ones which alliterate. Fuhr (p. 138) accents Beowulf 1497 thus: ^r hS pbne "^rAndwdn^ dn'^ytan mehte. With accents so laid listeners would catch a quasi-allitera- tion between cer and dn before the chief alliteration was heard, which would make a serious discord in the verse. In like manner when Hirt (p. 1 1 2), accents Beowulf 29 : BW(kse ^est^as Bwa hi silfa heed what is to prevent a listener from supposing that in swdb he has already heard the chief letter? To say that the accent upon the word is only a minor one is no answer, for it would be crediting an audience with hearing of preter- natural acuteness to suppose that they could in a moment detect the fact when the alliteration was present to mislead them. So again in Beowulf 19 : l&n%e hwtle him JfCBS l^-/red, the sub-alliteration coming where it does must inevit- ably disturb the chief. Crossed Alliteration, 53 As regards the Auftakt, therefore, the investigation in this chapter tends to confirm the ordinary view that it is in theory without accent. It must be admitted, however, that such a half- verse as Beowulf 744* : seems to show that the apparent participation of a syllable on which a merely secondary accent falls in the alliteration may be ignored. CHAPTER III. Vowel Alliteration in the Fourteenth Century COMPARED with THAT OF BeOWULF. The following investigation was suggested by Vietor's remark {Elemente der PhoTietik^ § 30, Anm. 1) that, whereas in modern English the spiritus lenis is not heard before an initial vowel, we may suppose, from the evidence of vowel alliteration, that Anglo-Saxon possessed it. / It struck me, in this case, as not unlikely that the sound was lost in the confusion of language which followed the Norman Conquest, inasmuch as being one of comparative difficulty, and unprotected by any orthographic sign, it would not easily hold its ground during a lengthened period in which there was no standai'd of speech and no contemporary literature. If my conjecture were correct, it seemed probable that the vowel alliteration of the fourteenth century, when the genius of the old poetry woke again to life with astonishing freshness after his long sleep, w^ould show some trace of the change which had come over the initial vowel. On my looking into Piers Plowman this appeared to be the case. I found it hard to discover any vowel alliteration at all ; and, after looking through the whole Vision of 2400 verses (Clar. Press Ed.), had collected only about thirty ex- amples, and in these there seemed to me a tendency to alliterate with the same vowel all through a verse (cp. Pien^s Plowman, vii. 132 : 'Abstinence ^e Abbesse,' quod jpieres, ' myne A b c me taw^te')^ a thing diametrically opposed to the ancient rule of Snorri Sturluson (cp. Rieger, p. 14), observed also by the Old English poets. Vowel Alliteration in the \^th Century. 55 But I was disconcerted to find that Mr. Skeat had no word in support of my idea, either in the Introduction to Tiers Plowman (Clar. Press, p. xxxvii), or in his Essay on Alliterative Poetry (see Bishop Percy s Folio MS. vol. iii. Introduction), or in his chapter on Metre, prefixed to the E. E. T. S. edition of the Morte Arthur e. In all three places we read that if the chief letter be a vowel, ' it is sufficient that the sub-letters be also vowels ; they need not be the same, and in practice are generally different.' So far as respected the quality of vowel allitera- tion in the fourteenth century : as to its quantity nothing was said. The joint-editor of the Destruction of Troy, however, Mr. Donaldson (vol. ii. p. 508, note), incidentally notices the rarity of vowel alliteration in Piers Plotuman and William of Palerne as contrasted with its comparative frequency in the Destruction of Troy and in the Morte Ar~ thure. Moreover, Rosenthal in his article ' Die allit. Lang- zeileim XlV.JhdtJ (Anglia, i. 439) seems to imply that the existence of vowel alliteration in the fourteenth century requires proof, as he brings forward instances to establish it, though by whom doubt has been thrown upon it I am not aware. As regards the frequency of vowel alliteration, therefore, there can be no doubt that the poetry of the fourteenth century differs toto ccelo from the Beowulf for whilst in the latter it occurs in 15 per cent, of the verses, and meets the reader at every moment, in Piers Plowman, one must have a vigilant eye to find it at all, putting aside the cases in which h enters (which it does very often twice in a verse), that is those in which there is, according to the usual description, alliteration between the spiritv^s lenis and spiritus asper. Even the Morte Arthure, by far the richest of the fourteenth century poems in vowel alliteration, contains it less often by half than Beotvulf. For some reason the later poets seem to have been re- luctant to alliterate with the vowel, perhaps considering, 56 Dissertation on Alliterative Verse. like the writer on Alliteration in the Encyclopcedia Britan- nica (Mr. E. S. Dallas), that 'alliteration is never effective unless it runs upon consonants.' K so, the effect of a verse with vowel alliteration in the later period must have differed radically from that pro- duced by the same thing earlier on. Certainly a metrical resource which was a favourite with the Beowulf poet, must have been good enough for so careless a versifier as Langland, for, as Mr. Skeat says, * William was not very particular about his metre.' (Cp. too Luick, ' Die englische Stabreimzeile/ Anglia, xi. p. 430.) The subjoined examples will, I hope, throw some light upon the nature of the difference. They are taken from the following poems : — 1. The Beowulf. 2. The Morte Arthure, held to have been the work of Huchowne, a Scotch poet of the fourteenth century, by Sir F. Madden (cp. his edition of Syr Gawayne), Panton and Donaldson (Introduction to Destruction of Troy^ vol. ii. pp. xxxiii and Ixiii), and Trautmann (Anglia, i. p. 139); on the other hand, considered by Mr. Morris (Preface to Alliterative PoeTns, E. E. T. S. 1864, p. vi) to have been 'composed in one of the Northumbrian dialects spoken south of the Tweed.' The MS. of the poem dates from the middle of the fifteenth century, and was written by Robert Thornton of Oswaldkirk, Yorkshire. Mr. Morris says it is ' not the original copy, nor even a literal transcript of it. It exhibits certain orthographical and grammatical peculiarities introduced by a Midland transcriber.' Panton (Preface to Destruction of Troy, E. E. T. S. 1874, p. xvii) contends that mere spelling is a very unsafe criterion ; read aloud more Scottico it becomes plain that the language of the poem is truly Scottish or Northern. Trautmann {Anglia, i. 139 ff. 1878) seeks to prove from metrical peculiarities that the Morte Arthure is Scotch : Luick (* Die englische SPxhreimzeile im XIV. <&:c. J hdt.' Anglia, xi. p. 586) Vowel Alliteration in the \/^th Century. 57 considers Trautmann's arguments not fully conclusive {^nickt vollig henveisend '). Morris and Skeat {Specimens of Early English, Pt. II. p. xxxix) give 1360 as the date of composi- tion of the poem: Trautmann (Anglia, i. p. 148) assents; Hugh of Eglintoun (Huchowne) having died about 1380. 3. The Destruction of Troy (E. E. T. S. 39, 56, published 1869, 1874), considered by the editors Panton and Donaldson to be a work of Huchowne's earlier than the Morte Arthure (Preface to Vol. ii. pp. xxxiii and Ixiii) : as also by Brandes (Englische Studien, viii. p. 410): allowed to be of Scotch origin, but not Huchowne's by Bock (Zur Destruction of Troy, Diss. Halle, p. 58) ; denied to be Huchowne's by Trautmann (Anglia, i. 126), who gives no opinion as to whether or not the dialect is Scotch. Donaldson (vol. ii. p. Iv) says that the MS. is of ' about the middle of the fifteenth century,' and that the scribe, 'evidently a West Midland man, adapted the work to his own dialect.' 4. William of Palerne (E. E. T. S. Extra Series, No. 1), translated from the French by a certain ' William ' about 1350. Dialect, Midland, possibly Shropshire (Morris and Skeat : Spec. ii. 138, Clar. Press). 5. The Gestes of Alisaunder, printed along with 4, supposed to be of the same authorship, by Sir F. Madden and Mr. Skeat; of different authorship, by Rosenthal (Anglia, i. 439). 6. Sir Gaiuayne and the Green Knight (E. E. T. S. No. 4), by Huchowne, according to Sir F. Madden ; not hy Hu- chowne, according to Tr&utmEimi— (Anglia, i. p. 123). Composed about 1360 (Morris and Skeat, Spec. Early Eng. ii. 39); between 1370 and 1380 (Trautmann, Ueher Verfasser . . . einiger allit. Gedichte, Hab.-schrift, Leipzig, 1876, p. ^^). Written in the 'West Midland speech' (Morris E. E. T. S. No. i. Pref. ix). 7. Cleanness and Patience (E. E. T. S. No. i) unanimously assigned to same author as 6, 58 Dissertation on Alliterative Verse. 8. The yim)n of Piers the Plowman (E. E. T. S. Nos. 28, 38, 54), by William Langland : arranged in three texts, A., B. and C, by Mr. Skeat, assigned respectively to the years 1362, 1377, and 1399. Dialect of A. Southern, with Mid- land peculiarities (Clar. Press, Specimens, 11. p. 175), of B. and C. 'mixed, but on the whole that of Langland himself (Skeat, E. E. T. S. 54, Pref. Ixxi, Ixxiv). A very convenient parallel edition of all three texts (not at my disposal till my lists had been made out) has been given by Mr. Skeat (Clar. Press, 1886). I. Vowel Alliteration in Beowulf. The examples have been taken from the Heyne-Socin edition (Paderborn and MUnster, 1888). They are arranged in alphabetical order, thus :a.a:a;a.a:8e; &c. Short vowels precede long, and simple vowels diphthongs. Ex- amples with alliteration in all the first three arses of the verse are given first, then follow examples of verses with only two alliterative lettei*s. Single alliteration in the first arsis is expressed thus : a . - : a, in the second, thus : - . a : a. ii. a & :& 733. a :y 1566. ( 510. 719, a & : 165. a : f 1003. a . -:6. 1315, 1503, a £b : e 1333. a 6 lib 1372. ( 1588. a & ea 593, 817. a ti : y 1662. a . - : e 823, 1525. a e :& 785. a y : ed 1767. a . - : ea 681, 1797. a e :e6 597. a f :ea 849. a. - red 1936. a ea : e 56. a f : eo 2835. a . — : eo 303, 669. a ed : 89 392. a - : a 929. a . - : u 1309. a eo :£b 1677. a - :& 1273. a . - itL 346. a eo : e 2696. a . - : » 354. - . a : » 960. a i : 6 2671. Vowel Alliteration in Beowulf, 59 d. d. . a : a 956. ft. i ea 1849. ( 283, 375, eaJ 700, 1459, (2542. ft . ft : ea 2462. ft. e 2784. ft. - : ft.aa :i 1108. ft. u ea 2740. ft . e : e 2877. ft. y i 1737. ft. - eo 2535. ft.e :i 1 169. ft. - a 1884. ft. - eo 317, 635. ft . e : ea 1764. ft.e : fib 3076. ft. ft. : ft 425, . SB 1270. JOII. ft. - ( 256, 556, "" I2840. ft . ea : ea 1548. ft . ea : u 2269. ft. - I2906. 1183, ft. ft. ~ d 219. tl 382, 458. ft . eo : i 2065. 1 135, 296, ft. - y 740. ft . eo : u 3139. ft . - : e I 894, 1763, - .ft §b 2607. ft . i : 89 2716. - I2521, 2593. .ft eo 431. 8B. SB . ft : ea 2668. ae . ^ : e 2343. flB . 88 . u : e 263. : e 2189. 8B . - 1 159,2166, • (2372. 8B . A : u 130. m . u : i 330. 89 - :ea 3172. ae . e : e 2507. 8B . u : ea 2444. 89 . - :eo 92, 983. SB. eft: a 1226. ffi . - : ft 889. 89 - : i 2227. sB.ea:u 1501. 8B . - :A 2375. 89 . - : 1950. «B.eft:^ 198. sa . eo : u 2008. 8B . - -1.80^: 1597, 89 89 . - :y 2598. 'f 1245- sb. & . a : e6 251. ^ u :aB 3136. £b - : ^ 656, 1357. dfe . 8B : ti 3107. & u : ea 1866,2625. ^ - : eo 1051,3016. A.e :i 1618. &.eft:§ 617. & & y : 1 2082. : ea ^^, & - (3158,2845, • 13036. A . eo : i 803. & ^ : u 46. & - :o 2501,3126. £B . eo: 6 1229. & - :a 15. ^ - : u 760. A . : a 2499. & - : ft 647, 2238. A - :y 9- A.o :y 1887. & - :fiB 1330,2732. - & :eo 628. A.d :u 1 166. - e. e.a :ft 2315. e . 8B : e 1788. e . eo : eo 1968. e . ft : e 1530. e . 8b : i 643. e . eo : i 1282, 1650. e . ft : ea 1378. e . ea :^ 693, 2655. e . eo : 2918. e.ft :u 449,1411. e . eo :A 6, 1892. e . eo : ft 2677- e.& :6 1680. e . eo : ea 2826. e . e6 : ea 1739- 6o Dissertation on Alliterative Verse, e .1 : & 1460. e .1 : ea 2779. e . io : 11 2201. e u : a, 1098. e u 1 988,2565, •"^1(3170). e u :i 2215. e u : 444, 2272. ^. e e . ed S . eo : i eo 1513. y 2379. e 2862. e a 1494. ea 1 195. i 2720. 6 / 340, 662, ) . - : a ^ 806, 1288, e . - (1754,2112 • I2897. (1542. { 84, 108, e . - : eal «^' ^^4 12737. ) . - : ft < 901, 1263, e . - : ed 829. (2400,2644. e . - : 2388. (2123. e . e . — : 6 1350, 1472 : u 877. ». - :^l 241,2591, •^\2788. . e : ea 261 1. e. . y : ea 2150. d. - :o 1761,2797. . - : a 1780. e. - :6 3102. . - : e 1693. §. - : u 410. . - : ea 577, 914. - .§ :e 1775. . - : eo 2894. - . d :eo 1731. . - : eo 1961. ea. ea a : k looi. ea . eo : & 2958. ea .- : ffi 12, 1658. ea ft. : e 2369. ea .eo:e(^559,298o, (3175. ( 758, 1383, ea 8d : a 907. ea . - : &< 1467, 2304, ea 8B : ea 2043. ea eo:ea 1728. (3004,3027. ea fib : tl 2075, ea . eo : 2617. ea . - :e 373, 808. ea e:ft, 1711,2775. ea . i : fib 2450. ea.- :§ 946,2331. ea e : & 2350. ea . i : eo 1 1 56. ea •- : ea 870, 1718. ea e : e 2791. ea i : i 999. ea . - :eft, 1782,2471. ea ea e : eo 903. e : 2930. ea ea . i : 2479. i : 6 1875. ea .- 1 884,2339, • ^''12823. ea e : 6 2452. ea 6 : ea 2764. ea .- : i 1777. ea e : u 603. ea u:eo 357. ea . - : i 1609. ea ea e : ft 2814. d:ft 2211. ea ea u: 2761. u:ft 961. ea.- .^ (1327,2868, (2939,2973. ea e : eo 1703. ea . ft:^ 2298. I 481, 1030, ea 6 : eo 2886. ea . t : ^ 534. ea .- : 6 1 1339, 1689, ea. § : d 2199. ( 472, 1715, (1946,2062. ea ea:A 973. ea . - : ft, ^2600, 2658, ea .— : ? 2230. ea ea : ea 836. (2921. :? 2362. ea ea : 1664. Vowel Alliteration in Beowulf. 6i ed. e& . ft : 6 433. ed . ea : e 1622. ed . - : ft. 100, 882. ed . e : u 276, 2141. ea . ti : eo 2335. ed . - : e 138. eo. eo . a : y 690. eo.i :a 1448. eo.- :6 1202. eo . SB : a 369. eo . i : 1045. eo . — I ec f 796, 1036, (1443,2728. eo.& :^ 2249. eo.i :8B 1113. eo.^ :e 1758. eo .0 :u 2952. eo . - : ec ► 1329,3168. eo.e :» 2817. eo.6 :u 2909. eo.- :i |i 142, 1867, I2191. eo.e :8b 2623. eo . tl : eo 762. eo e -e ( ^^8, 2536, eo.y :o 112. eo . - : 1958. ^°-^ -^13064. eo.e :ea| 358, I2134. 2022, eo.-:.{4---^. eo.- :d eo.- :u 860. 1073. eo.- :8B 2235. eo.- :tl 664. eo.e :o 3008, eo.- :ife| 792,1236, (2233,2713. eo . - \f 1209. eo.e :u 224. - .eo:e 1 146. eo.ea:i 3124. eo . ea : y 770. eo.- :e| 753,2143, I2892, 3047. - .eo:d - . eo : ou sees pat pe "Ewperour es angerde a lyttille. 72 Dissertation on Alliterative Verse. In the last two verses es stands before the mid caesura and receives an accent from the pause. The caesura in these cases is, it must be confessed, a purely artificial one. Luick (p. 596) observes that the boundary between the two half- verses is occasionally obliterated in the M.A., as in other XIV century poems. But such an artificial caesura is not un- known even in Beowulf. Cf. v. JZ.Jjwm eafera wees, tefter ceyined. In the second of the above examj^les es is clearly in the first arsis, and therefore may fairly be held to alliterate. Similar is the case of are in V, 1327* ' je are ivith my fie enemy, ^ and again in V. 2289* J)er are of the Orient. The possessive pronoun could alliterate by itself before a following noun in Beovjulf, at least in the second half verse. Cf. vv. 345, 2532 &c., and see Rieger, p. 31. In the M. A. such a thing is rare, but occurs in v. 2803* And Arthur e one ou.re syde, and so probably v. 2368^ none of oure elders, and probably too in v. 1016^ or any of hjs elders (see Remark iii). viii. The same word does not always alliterate on the same syllable. Thus we have: alias 3966, alias 1153; aYaiunnvarde 3169, 3765 and usually, but avawniwarde 2830: Gaiuayne 1^79 &c., but Gawayne 3920; undone 3967, but unddne 3753; eriYerounes 4125, but envei^de 1694; enamelde (probably) in 765, 2027 &c., but ennelled 1294; withowttyne 2^oo,hui withowttyne ^^go, withoiuttyne 114. Verbal compounds with en- usually alliterate with the second syllable, e.g. encroche 3213, endordid 199. An exception seems, at first sight, to occur in v. 307 To entycc the Eniperour to take overe the niounttes. Judging from the first half- verse we must pronounce entyce ; the second half- verse, however, leans rather towards entyce ; since, in accordance with what has been said in Remark vi, it is Vowel Alliteration in Morte Arthure. "j}^ more natural for the verb take to alliterate, than for the following adverb to do so. The alliteration therefore seems to waver in this verse, a strange phenomenon peculiar, I think, to the Morte Arthure. It occurs again : V. 910 Tlie vesare pe arventaile enarmede so /aire V. 1096 'Ruyd B.rmes as an Sbke ivith rusclede sydes V. 4137 / see syr Tlwaytie overaette ivith Sara'^enes kene V. 3717 Sir Arthure &^ Gawayne asryede them bothene. Compare also vv. 2868, 3590, c^^y. In this connection may be mentioned one or two ex- amples of crossed alliteration : (i) of ab : ab in v. 2628 That he usede in werre alle this SLiighte wyntfer (where vowel and half-vowel alternate) ; also in vv. 1 1 y^, 3444 ; and (ii) of ba : ab in vv. 1259, ^^33- ix. The adverb anon usually alliterates with the n in the Destruction of Troy, with the a in William of Paler ne. In the M. A. it seems to do so with the 0, cp. vv. 2717, 2841, 4264; and notice too ondne, vv. 571, 2006, 2067. X. The peculiar /i-alliteration of the Destruction of Troy, cp. D. of T. loi, ^n yle enablt nobli and wele, noticed by Trautmann {Anglia, i. 126), and Schipper (i. 207), seems to peep out in V. 1905 Be an aumtere of armes Joneke has nommene, in which case this verse is another instance of the wavering alliteration mentioned above. A similar end-alliteration seems to occur with w in 2987 Thorow otvte be rerevrarde he holdes wayes. The alliteration of the t in the phrase ^e tother also be- longs here : compare 234 Sir Ow^htreth on pe tother syde of Turry tvas lorde ; but 2530 The toper irously answers hyme sone. Cp. Dest of Troy, 6198* . . pat other ; 7905^ and pat tothir. We have toper again in 3410 The toper Ector of Troye, the chevalrous gume, 74 Dissertation on Alliterative Verse. a. verse in which the chief letter seems at first sight missing, but must be sought for in ch{ = tsh) ; a singular alliteration found again in ^ir Gawayne and the Green KnigJd, 151:2 : J}at chaunce so hy-tydez hor cheuysaunce to change. That the ch- in chevalrous was pronounced tsh, and not, as often now, sh, can be seen from such alliterations as chefe chevalrye v. 18 (cp. chosyne, cheftayne, cheefe, 530; and see also vv. 1362, 1404, 1540, 1619, 1832, &c.). The case in question is interesting as an evidence that the poet was guided in his alliterations by the ear, and not by the eye. In these days of reading it is the alliteration to the eye which catches our attention first, so that even Messrs. Seeley and Abbott can quote Shakespeare's — Our dreadful marches to delightful measures as an instance of crossed alliteration ab : ab (See Engl. Lessons for Engl. People)^ as if there were alliteration between dreadful and delightful ! (Cp. ch. II. p. 4.) xi. The chief letter is unquestionably missing in v. 4080 : Ayers with Arthure, with sevene score of knyghtes, where it seems obviously right to read with aughte score, cp. 278, 3189. Less easily emended is 4344. That was of Ectores hlude the kynge sone of Troye. We seem compelled to alliterate Ectores : Troye, giving, it must be confessed, a very unnatural accent to the former word (though compare Owghtreth, v. 234). In the Destr. of Troy, where the name is very frequent, it is always the E which alliterates. For the chief letter to be in the final arsis, as in Troye, is a common occurrence in the M. A., more so, according to Trautmann (Anglia, i. 123), than in any other O. E. poem. We have instances with a vowel in vv. 1 1 22, 2543, 2848, 2869, 3246, and perhaps in 2848. Vowel A Hit. in the Destruction of Troy. 75 Destruction of Troy. a. I / 181, 273, 762, a . aw : a 276. S55, 1090, 1 164, a . aw : ai 5707. 1588, 1744, 2023, a . haw : a 12937. 5568, 5674, 5727, a . e : a 161 2, 1765. a a -\ (5843), 6072, 6108, a . e : e 9107. 6289, 7206, 7105, a . ew : 1479. 7241, 7994, 9104, a . : au 2856. 10809, 10145, 10201, a . : 6356, 9288, 9797. 10302, 10723, 14024, C 537, 2040, 7949, 14026. au . a :a<94i5, 9419, 9527. I 67, 1 83 1, 1899, ( 9801, 10602, 10612. 5770, 5791, 6060, au . a : e 6640. a a : au s i 1 6114, 6183,10635, au. e : a 6216, 6399. { 10644- au . e ^^1 ^'t < 7306. 7085, 7245. a a : aw 1862. a a : ay 1577. au . e : 7254- a a : e 1087,(6249), 7329. au . : a 6319. a a : 5692, 8903. au . :au 5. a ai : e no. au . :o 9375- a ai ( 5784. 1940, 6159, 6625, au . u ay . a :a 9745- : a 9088. a au \ : a < ( 6861, 9515, 9541, 10063, 10538, 10608, ' 10692, 10804. ay . e ay . ay . u : a 175, : e 173- :au 531. 7314^ ( € 46, 1441, 6874, ), e . au e 6725, 6810. Q a 1 7229, 7384, 8895, e . au *o 1121, 7391. j 1 8919, 10668, 10760, e . e : a 5602, 6424. 1 10865. e . e au 6490. ; 314, 517, 6617, / 126, 143, 264, e a : au 7006, 10667, 1082 1, ^ 14035- 294, 5920, 1002, 6413, 1390, 6456, e . a : e 121, 1491, 7056, 6599, 6712, 6803, 7322, 7344, 7947. e . e : e ^ 6835, 6843, 6855, e . a :i 7119. 6934, 6993, 7128, e . a : 852, 9185, 10590. 7300, 7404, 7424, e . a : u 1053- 7919, 9234, 10320, e . a : y 6222. ' 10493. e . au : a 6392. e . e ea 7988. e . au : au 6796, 6S89. e . e i 749- 76 Dissertation on Alliterative Verse. e . e : y 1633. e . ho : a 7187. e . ea : e 9067, 9661. e . ho : aw^ 9164. e . hei : e 394. e . ho -1 5950, 6065, 7233, e . i : e 1936. 1 1969. e . iu :i 923. e . ho :he 5950, 6065, 7233. e . o : a 819, 8901. 9343- e . o : au 61 17, 6359, 7250. e .ho : 6157,14031- e . o : e 2, 410, 961, e . ow : a 6088. 6198, 7219, 7907. e . ow : au 6166. G . o :hei 1476. e . ow : e 6792. e . o : o , 6194, 6708, \ 6859, 9203. 6745, e . ow : 6021, 6237. e • y : e 1330- e . o :ho 9537- hei . e : e 115. e . o : ou 6404. 1. i . a : o 4313 ; i . e : a 6013 ; i . o : a 6998 ; f . i : f . e 7345. . a :a i 617, 7215, 7367, : u 877- \ 9381,14012. ho : 1919. . a : au 742, 6055,10615. oy :ho 6169. . a : aw i960. ho a : a 11831. . a : e 1 47 1, 1600. ho a : u 11661. . a : 281. ho e : a 6303,10806,11840. . a : ow 5565. ho e : au { 6930, 9362, 9725, ( 14006. . au : a 6376. . e : e 6048, 6704, 7136, / 288, 3007, 6677, 9239- ho e : e ] 7379, 91 10, 9258, S 6179, 6189, 7388, ' 10482. . e .0 j 8955. ( 6305, 7074, 9093, . e :ho 9219. ho e : j 9121, 9768,10331, .he : u 328. ' 10469,10839. . : e 7358. ho : a 9743- . : ey 123. ho : 12944. ( 48, 151, 165, ho u : u II753- . -1 553, 7205, 10312. 9096, oy a : 6172. Vowel A Hit. m the Destruction of Troy. 77 U. u .a : a 6894. u y :ho 3447. u . au : a 1679, 6911. u y : u 7981. u . u : u 1545. 5801, 5808, 5819. - y. y .a : a 6402. y e : 4426. y .a : 2858. y : a 9298. y • a :y 6531- y • : 6202. y . ai : a 897. REMi y ^RKS ho : 1646. i. The metrical regularity of the Destruction of Troy can escape no one. It was pointed out by Trautmann {Anglia, i. 125), and again by Schipper (i. 211). Luick takes this poem as the basis of his metrical treatise, ' Die englische Stabi^emizeile irti XIV. <&c. Jhrhd.' (Anglia, xi. 392.) The alliterative scheme a . a : a :- is carried out with even monotonous exactness. Only twice in the first 2000 verses does the variety -. a : a- occur, viz. in verses 453 and 806 (Luick, p. 394). The last verse of Book xvi. describing the third battle, which is another exception, gives us by way of compensation a particularly good example of crossed alliteration a b : a b, viz. 7345 : ffore to pere Innes us the tight endit. Thus the task of determining the alliterating words where there is vowel alliteration is extremely easy for this poem. ii. Trautmann (ib. 1 26) remarks that the alliteration in this poem is often upon the same vowel in all three arses. This peculiarity again, though directly contrary to ancient usage, is unmistakeable, as a glance at the above lists will show. Of the 167 examples given, all three vowels are alike in 35 cases, two out of the three in 91, and in jS Dissertation on Alliterative Verse, only 41 instances are all tke vowels different. Where the vowels are not all alike, however, the difference is often made merely by a pure diphthong, the first element of which is the same vowel as in the other arses e.g. a . a : au, verses 6']^ 1831, &c., and as the stress of pronun- ciation was upon the first element of a diphthong the variation from identity of vowel is in such cases only apparent. Complete dissimilarity as in the combination e . a : o, verses 9185, 10590, occurs in only 23 of the 167 examples given above. iii. There has been some question as to whether in the Destruction of Troy the aspirated h ever alliterates with a vowel. Trautmann {ih. 126) says not ; Bock (p. 16) affirms that it does very frequently in the case of romance words. Luick (p. 428) lightly says that in these the h is dis- regarded. He explains away the apparent instance in ver. 9343 : Whethur "Elan he so honerable, or of so hegh prise, by saying that here no doubt the poet intended the original aspirated form of the proper name. In another case, cited by Bock, hawne, ver. 12937*: Agamynons "hawne sone we have doubtless a mere scribal peculiarity. iv. Luick (pp. 397, 425) falls into error with regard to the word other, saying that it never alliterates, and appears to have been always unaccented. Instances to the contrary are: 48a Ouyd and othir (cp. 819*, 961*, 7388*) ; 7205* Without ordinaunce of other (cp. 9203*) ; 6198* And "Esdras pat other ; 7219* Tlctor on the other side. Luick also says, no doubt correctly, that all usually does not alliterate in the Destruction of Troy. It occurs how- ever in the chief arsis in ver. 8895 : I haiie not errif in anything of all my tyme yet, Vowel A Hit, in the Destruction of Troy. 79 and in verses 10590, 10760, it takes the alliteration from a following substantive, in each case, however, a weak one : Er?es and all men oght of astate And "Elan, of sill tvemen, a,ngardly fast. In ver. 328 : With all odotir of herhis pat on vrtJie springes, the metre is in favour of accenting odour in preference to all. Sometimes, as in the Morte Arthure, we have to decide between the claims of all and other to alliterate, e.g. V. 852 Entris unth armur and all his other gere. V. 9185 Noght Ector of all other, oddist of knightes. In the former verse rhythm and sense are both in favour of other ; in the latter rhythm at least favours other (cp. Morte Arthure, 530^). V. Redundant alliteration is less frequent than in Morte Arthure, but occasionally occurs, e.g. ver. 1330 : "Ercules yreful euer vponone (cp. 1390), where every word begins with a vowel, and again ver. 48 : Ouyd and othir J?at onest were ay. There is too in the Destruction of Troy a very peculiar tendency for the last word of a verse with vowel alliteration to begin with a vowel even when the vowel is not accented : e. g. 897 Bothe of ymur &^ sdre, after I-wise 1 121 Then "Ercules the Avntrus onswarid Anon — cp. verses 961, 3447, 8955, 8989, 9096, 9743, 9831, 10.590. vi. As in the Morte Arthure, the preposition can alliterate to the disadvantage of a following noun or even adjective, e. g. V. 1053 Entrid into Aries vnder a signe V. 8901 Ne entnd not the ost till after tiro yeres. So Dissertation on Allite7'ative Vei'se. vii. In one instance the indefinite article a takes the alliteration ! viz. ver. 1441 : • Hit ivas euyn hot a ve7iture of Angur to come, where it seems to be joined on to the following word in the same way as the n of the other form of this article in the already mentioned 7i-alliteration of this poem (cp. Morte Arthure^ Remark x). We have possibly a similar case in Morte Arthure, ver. 903 : Ahoven one pat a jeryne of Acres owte over, but as the occurrence of two alliterating syllables in the first half- verse is by no means so regular in this poem as in the Destruction of Troy^ it is difficult to feel sure in the matter. (For earlier examples of alliteration on the indef. art. cp. B. I886^ Gen. 88o^ 2226^) viii. The word mn alliterates, in 7949 Jf ^'^^ntur he, sir AchtUes, I aw the to sle. William of Palerne. f 59, 235, 629, a . a :ha 4921. 645, 649» 813, a . a :hi 864. 834, 861, 874, a . a : 136, 1306, 3381. 913, 941, 958, a . a : - 3769- 967, 976, 1015, a . e : a 3950- 1028, 1041, 1048, a .he : e 1469. 1064, 1414, 1517, a . : a 568. 1528, 1532, 1667, r 586, 1499, 4248, 1701, 1734, 1760, a : a < 4816, 4863, 5222. 1993, 2cx)4, 2062, L 5229, 5277, 5446 3220, 3278, 3332, a . - : au 921. 3423, 3476, 3670, a . - :hu 363. 3784, 3885, 3963, 4358, 4452, 4508, 4537, 4565, 4595, 4672, 4708, 4737, 4885, 4958, 4991, 5040, 5155, 5308, U409. William of Palerne, 8i e . a e . a e . e e . e a 273, he 1445. i 882, a 1881, r 168, \ \ 1093. e * ] 1632, ^ 5338, he 276. i 3755- hi 1261. o 5400. hu 2515. 56 1429. 404. 2479. 1906. 205, 272, 1 1 30, 1417, 5260, 5269, 1 5412. e , ei : e 709, e ■ey : e 4641. e . :ho 1 106. e . ou :hi 3229. he .ha : a 441. he . hei : a 1875. he ,hi : 389. he . : a 3032. he . :ha lOIO. he .hu :i 688. ei . e : e 1612. e . - : a 166. e : e 2519. e . - : e 232, 1474. 3105- . a : a 3220. . a :he 5403. . he :ha 3633- . : a 1069. . :i 3791- • oy : a 139- ho . a : he 1 103. ho . e .he 1768. ho . :he 295> 3797 . - : a 3814. . - :i 4952. - . : 1402. ou . e : e 3646. ou . : a 1080, 3745 i. In regard to vowel alliteration, William of Palerne presents a more striking contrast to the Beowulf than any poem we have yet examined. I have found onl y 131 c ertain instances of it in a total of 5540 verses, or abouO •35 per cent., whereas in Beowulf we have about 490 instances m 3184 verses, or 15-4 per cent. Again, in William of Pa- lerne, the alliteration turns almost entirely on the three vowels a, e, o, and a strong preference is shown for keeping the vowel the same in all three arses of the verse, by far the most favourite combination being a . a : a, which occurs 54 times. The alliteration e . e : e occurs eleven times ; so that together these two combinations are found in 6^ out of the whole number of 131 examples of vowel alliteration in the poem. Scarcely ever do we find a different vowel in G 82 Dissertation on Alliterative Verse. all three arses. Excluding such cases as ho . a : he, ver. 1 103, where the rough breathing enters, actually or in appearance, I have observed only two instances of complete variation, viz., e . a : i, vv. 882, 2479. ^^ ^^ ^^'^t of these, ver. 882 : and eft red as rose . in a litel while, the alliteration is of the weakest, and seems in the first half- verse to waver between the vowels and the consonants ; in the second example, ver. 2479 • but eche man a,l ni^t . inned him where he mip it is fairly strong. This tendency to preserve the alliterating vowel un- changed was noticed in the Destruction of Troy. It is seen at its height in the Gestes of Alisaunder^ in which there are but four instances of alliteration between different vowels. This was pointed out by Kosenthal {Anglia, i. 439), who finds in the fact an argument against Mr. Skeat's opinion that William of Palerne and the Alexander are by the same author. The Alexander is, however, only one-fourth as long as William of Palerne, and, if we take into account the difficulty which restricting the alliteration to one vowel would impose on the poet, the difference between the two poems in regard to vowel alliteration seems insufficient to justify any argument as to difference of authorship. ii. There is a great sameness among the instances with the combination a . a : a, most of which in the earlier part of the poem contain the name alisaundrine, whilst later on that of alphouns is of frequent occurrence. The formula onnon d^nswered and saide is as common as the roV8' a-naixu- fi6fx€i'os of Homer: cp. vv. 645, 958, 967, 1048, 1667, &c.; so, too, we find again and again anon atteled ; cp. vv. 813, 861, 941, &c. ; &7ion after ^at ilk, 629, 1041, &c., and other such stereotyped expressions. iii. The alliteration falls occasionally on words which have naturally very slight or no sentence stress. Thus in William of Palerne, 8 o ver. 882, quoted above, the preposition in has the chief letter even before a noun, and so again in 3791, 4952. The same thing occurs with on^ 3381, 4565 ; and at, 166, 1414, 5040; also with to, 59^; fromfi, 3614; wip, 3000, 4498, &c. The word and seems to alliterate in 5277^ and again in 3950'', unless in the latter case it is the indefinite article a which does so, as in the Destruction of Troy, 1441. The conjunction as alliterates in 273^ 976^, 1881'', &c., and the verb am in 1532^ Speaking of such cases, Mr. Skeat says {Essay on All. Poetry, p. 7), ' By a very bold licence, the cliief letter even occurs in the initial catch of the second line.' Rosenthal (Anglia, i.) questions this, preferring to ignore the allitera- tion in such half-verses, and Luick also {Anglia, xi.) overlooks it, scanning, for example, William of Palerne, 4^ J)at was a couherde, as a C verse (Sievers) ; cp. p. ^68, also 581. That the alliteration in thorn was intention^al is clear from their frequency ; to what extent it was real is not an easy question. Some stress the words with the quasi chief-letter must have had, otherwise the alliteration would have been without effect. It may help towards a solution, if we compare some half-verses in which a naturally proclitic word (for example, with) has the alliteration, with others similarly constructed, in which it has not. Thus we have : 604^ -wip pi mede tvordes 1646^ ivip pi taire wordes 1060^ "wip riche dr° wip pore 5442'' wip \asse 6r^ wip more 2606^' 'wip ful ^ode houndes 1060^ wip ful taire 'yftes 409^' wip him he ladde 3667^ tvip hem pan sche tvent Cp., too, 2109 and 1261 ; 3311 and 1337 ; 1083 and 3232. In the case of 409^ no doubt the word wij? really does alliterate, since even in Old English verse prepositions could do so before a pronoun (cp. Rieger, p. 32). The other half- verses in the left-hand column, one is disposed at first sight to scan similarly to the corresponding verses on the right, disregarding the alliteration. Two reasons, however, G 2 84 Dissertation on Alliterative Verse. may be urged against this course, viz. (i) that in the case of individual words varying alliteration, and therefore, presumably, varying accentuation, is not uncommon ; for example : aJisaundrine 629, 645, &c., &c., but alisaundrine 1979, 2028, 2065. atepmoderes 130, but stepmoder 2640 (compare appeltre 753). p&mperour "zyi^ 272, &c., but \emperour 429, 1 1 27; and even Jjemperour 1180, 1390. Q^kes 235, 2062, &c., but askes 69. a>nsivered 645, 958, &c., but ansivered 70. ahoute 89, 220, &c., but a.houte 4358. Sitir 4537, but atir 3884. and (2) that words naturally of the strongest accent, sub- stantives and adjectives, are occasionally found not only in the thesis succeeding the first alliterating word of a half- verse, but also in that preceding it, as in the following examples : 137^ J?at euel chance hire tide. 381^ gret help per-of after, 368^ pemperour nam hede. 315^ hut god do Ms wille. 59^ pat childern ofter tvilnen. 420^ pe testes aeppen ware. 463* min ei^en Borly aren aogettes. 1800b fQ meliors so hende. 344b pe worse hestow neicere. 190^ a-day wip his howe. 307b as lord "wip his owne. In every one of these instances the strict rules of Old English alliteration are violated (cp. Rieger, pp. 19 and 25, § iv, I and 4), most flagrantly so in the last two, but also in the others, in an ascending scale of enormity. It is manifest that, as regards alliteration, they stand or fall together, for it is impossible to draw any line between them. Moreover, except in the Auftakt, these verses are perfectly regular, presenting the types, A(x)-x x-x, C X x-^x,BC X x-^x-x, which Luick (Anglia, xi. 568) finds to occur in William of Palerne in the ordinary man- ner ; whereas if we ignore the alliterating word and place William of Palerne, 85 the preceding noun in the first arsis, we get in some in- stances, particularly in 463*, sadly misshapen verses. I conclude, therefore, that the poet intended wip to be accented in vv. 190^. 307^^, just as he meant sorly to have the accent in 463*, and that the words lord and ei^en should be somewhat slurred in the reading. In modern verse, in anapsestic verse especially, we are familiar enough with having to hurry over words of similar weight in the thesis, as in the second of the following well- known lines of Byron : And the sheen | of his spears | was Hke stars | on the sea Where the blue | toave rolls night \ ly on deep | Galilee. We even admit in this position words made emphatic by antithesis, as in Mr. Gilbert's amusing — You like my features I suppose /'m disappointed with my nose — in the * Mountebanks J In cases like these there is no doubt a conflict between the ' ideal rhythm and the real ' (cp. Scherer, Zur Gesch, d, deutschen Spr. ^ 628) greater than would have been tolerated by the ancient ear, more intent to catch the music, less absorbed in the meaning, of a verse than the modern. But the author of William of Palerne is one of the least skilful, even in the Silver Age of allitera- tive poetry, and it is not surprising if the licence we have been discussing is more common with him than with his contemporaries, excepting Langland ^. The same arguments will hold a fortiori for the preposi- tions wipoute and wipinne, which alliterate not unfrequently with the ^(;; cp. vv. 259, 316, ^^^, 542, 1639, 1900, 2041, 2350, 2464, 2685, 3660, 4608, 4731, 4813, 4807, 5457, and 685, 2956, 3652, and will also cover the case of other pre- positions and particles. * The same licence of alliterating the preposition before a substantive is found even in the Old English version of the Psalms, and occasionally in other poems of the decadence. Cf. Rieger, p. 33. 86 Dissertation on Alliterative Verse. The postposition ^viih alliterates in 47 9 7^^ ferwith wold he slaked (cp. 138*) quite according to rule ; just so in Beowulf we have fandjjd J)(et inne ( 1 1 8^),/^ he him of dyde (67 (j^) ; and in Genesis 1053* eastlanclum on. Probably the irregular alliterations we have been discussing may be traced in part to these regular uses. iv. Alliteration of the rough with the smooth breathing does occur in William of Palerne, but not very frequently. Apart from merely apparent cases in which the initial h is merely orthographical, as in heritage, 1445^, written also eritagey 4641^, we have hors and armes, 1103*; to bi-hold the estres, 1768* ; /)er is holli al here ost, 3797% &c. V. The word an^n alliterates usually with the a, but also with the n ; cf. 353 7^ It appears to do so in ver. 136* : a noynement anon sche made • of so grete strengthe, but a noynement is of course to be read an oynement, the word of in the second half- verse clearly showing that in the first half- verse we have vowel, and not n-, alliteration. I have noticed indeed only two instances of a word alli- terating by help of the end-consonant of the word preceding it, viz. 1127 : hotv pemperour wip ost . pider urns come, and 3435 : pat pe same seg hade slawe . hia em per-to-fore, but there are probably several others. vi. The word alone alliterates with the a in vv. 15:^8, 1532 {allone) (cp. olive, 4816 ; algate, 1064, 141 4 ; alas, 388 ; but alway, 345). The indefinite al alliterates before a sub- stantive, 151 7, and similarly a^Ze, 59, 1064,3381,3728, 3423, &c. ; so also does the indefinite ani, vv. 568, 404, 1080. vii. Crossed alliteration^ abab, occurs in vv. 108, 278, ^5^5i 2628, 3398 ; and baab in 888, 1908, 3259. Of the latter examples, ver. 888, at any rate, seems due to art and not accident : Bot olisaimdrine wiste ivel . what pat him mjUd. The Alexander, 87 viii. The pointing is at fault in 5446, and to &lisaundri7ie anon . r^t pei sayde. The csesural point should be placed after alisaundrine, and not after anon as in the text. Cp. ver. 4816 : and 'yf &lisaundrine . were panne eUiue. ix. The expression ei/jer . . qper presents some difficulty in ver. loio : pan eiper hent oper . hastely in armes. From ver. 1908, however: pan eiper lapped oper . ful loueli in armes, it is plain that the verb hent above occupies one arsis, whilst from 3391^ epir m.ette oper we see that the eiper can take a subordinate position to the oper since here eifer is in the Auftakty oper in the second arsis of the half-verse. We may therefore conclude to alliterate ver. ioio«': pan eiper hent oper. and so 3032 : and hendli eiper oper • pan colled in armes. In 3854* Grimli eiper oper gret both words are in thesis (cp. too 3814). If this view of the relation of the two words in this expression be correct it must be confessed they offer an exception to the rule for Old English verse, that of two pronouns in the same half- verse the one which comes first takes the alliteration ; a rule however which was occasionally broken even by the older poets (cf. Rieger, P- 31)- Alexander. a . ai ( 22, 27, 268, a . e : e 240. a 461, 500, 1094, a . : au 290. ' 1 148. au . a : a 498 aui ^^' < 547, 230, 41 5> au . e : e 592. 902, 1017. a . - : a 491. au 1027. 88 Dissertation on Alliterative Verse, e. e . SB : e 454- e . e yi 280. . 408, 424, 464, ei . e e 338. e . e : e 486, 552, 595, e . - y 1086. ( 908, 1091, eu . — u 34. 1. i .i :i 616. . o . o : o 695. : a 177. : o 664. 738. — : a - : o 135. 576. Remarks. i. Vowel alliteration in the Alexander fragments is almost as scanty as in William of Palerne. Rosenthal {Anglia, i. 414), found only 33 examples. He however omitted ver. 1094, a . a : a ; ver. 908, e . e : e ; verses 109, 902, a . a : au ; and ver. 34, eu .- : u. These bring the total up to 38, which gives a percentage of 3 out of the whole number of 1249 verses in the poem. ii. As Rosenthal points out, the alliteration in a verse usually runs upon the same vowel throughout, or if there is any variation it is due to a diphthong the fundamental vowel of which is the one alliterating e. g. a . a : au, 230 ; ei. ere, 338. Rosenthal finds only four instances of allitera- tion between wholly different vowels, viz., a . o : au, 290 ; . a : o, 695 ; e . e : y, 280 ; and e .- : i, 1086. To these 1 should add 177 : Olympias pe onorahle . over 2M hue hy^ht, taken by Rosenthal as an example of o . o : o. (In Sir Gawayne and the Green Knight these two words appear as a compound— owr-a^, which is no doubt to be accented Sir Gawayne and the Green Knight. 89 on the first syllable. Cp. Bir G. and the G. K,, ver. and 6uer-al 6nker y^ine, an iambic line with three accents.) Moreover, I prefer to take ver. 1086 : Syn pou aH Qrthlij thyself^ . in an yll tyme, as indicated, and not, as Rosenthal, in an yll tyme. Again ver. ^'jG: Too looke on Olympias . pe onorahle Queene, is better so taken than as Too looke on, &c., and therefore as an example of o .- : o, not of o . o : o, as Rosenthal gives it. iii. Rosenthal is correct in saying that there is no instance of alliteration between the rough and smooth breathings in the Alexander, Sir Gawayne and the Green Knight. a. a . a : a 95, 171, a . :ho 208. 241. a .ho : au 948. a . a : ha 330- a . hou :ha 309. a . a :he 2335. a . hou :he 2102. a . a : no, 1463. a . u :ha 185. a .ha : a 536, 2277. ha . a '. a 136, 491 ( 681, 1388 a . ha :hu 1543. a . aw :ha 836. ha . a :ha 590. a • hay :he 972. ha . a :he 2S1. a . e : a 85. ha .ay :he 438. a . e : 356, 2416. ha .he : a 323, 2331 a .he :he 26, 467. ha . : e 657. a .he :ho 289. ha . :he 2288. a .he :hy 349, 2271. ha .ny : a 2297. a . i : 300. au . e : a 27. a . hi : a 250. ay . e : 551. a .hi :ha 2287. ay . ei : e 629. a . : a 2466. hay . :he 223. 90 Dissertation on Alliterative Verse. - . a : a 1654. a : u 535, 1262. a . - : au 2522. ~ • ha : 2123. - . a : o 275. ha . - :y 649. a ■" : u 93. a (-?wo):i 42. e. e . a : a 567. e .hu :ha 788. e . a :(a)i 660. he . a : aw 2301. e . a :hi 5. he .ha : e 427. e . a : au 2467. he . a :i 477- e .ha : e 1641. he . e :ha 210. e .ha : how 379. he . e : e 1808. e . au : a 2464. he .he : ou 1 1 84. e . e :he 723. he . : a 253- e . e :ho 140, 881. he . hou : 2481. e .he :ha 496. he .ha : u 2079. e . o :he 2423. e : a 1474. e . u : e 184. i . i .a : ha 69? ;. C ). o .ha :ha 1914. ho . e :ay 2180. o .ha :ho 2016. ho . i :he 1930. o . au :he lOOI. ho .i : 2182. o .he : e 630. ho .ho : a 1257. o . o :ho 2412. ou . a : a 29. o .ho :he 593. hou .ho : ow 408. o . u : e 2190. : a 841. ho .ha : e 2520. : e 150. ho . aw :he 1038. : 385. u. 608. y hy u 113. 2232. hy . ha 1547. Cleanness and Patience, 91 Cleanness. a. ( 411, 7i3> a . : 237, 698. a . a : a < 1035, 1301, a .ho : a 643- ( 1470. a . ho :ha 626. a . a :ha 42, 688. a .ho :ho 1602. a . a : i 1314. a . ou : a 19. a . a : 761. a • y : au 1411. a . au : a 207. ha . au :he 1707. a . au :ha 1443- ha .i :ha 321. a . ay : a 650. av . a : a 1477. a .ay : u lOIO. au . a : ew 1457. a . e :ha 573. au .ha : ou 941. a .he :he 678. au .i ; 108 1. a . e : 601. au . : a 937, 1276 a . e :ho 485. au . ou : au 795- a . e : u ICX)I. ay .ho : e 1684. a .he :hu I2II. a . - : a 924. a .i : a 572. a . - :i 1 173. a . :ha 141. au . - : a 258. a . :hi 745- ^- • € >. e . a : a 329. e .ho : 705. e . a :he 317. e . hou : 602. e . a : V 303. e . - : a 647, 1698 e . au :ha 895. e :ha 562. e . e : a 1323. he . a : 1749. e . e : e i 241, \ 1006. 277, he .ha : u 524. he . au :ha 506. e . e : u 747. he . e :ha 1336. e . e :y 1240. he .he : u 710. e . e :hi 946. he .i :he II79- e .he :ha 349- he .1 :hy 1294. e .hi : e 1332. he .ho :au 240. e . : e 402. hy .hy : a 380, 1098 i .ha :he 577. i I. hi . e : a 479. i . e : au 782. hi .hi :y 430. 9^ Dissertation on Alliterative Verse, 0. o a :he 1086. ho . a : ha 335. o e : ha , 448. ho .ha : e 148V o he : a 1704. ho . au : a 1451. o .hi : au 1600. ho . au :he 1222. o o : a 1044. ho . au : u II. o o : ou 923- ho .ho : a 1544- o hou :hy 623. ho . u : i 14. o u : o 1092. ho • y : V 1695. o - : o 299. ou . a : a 357- o - :ho 1799. ou : a 345- o : e 1 1 44. hou . ou :hy 1290. I I. V he :ha 1520. hu . a : 426. u o : a 30. hu . e :ha 874. u hou : a 1284. hu .i : au 1659. u u : 6 892. hu . : e 1204. u 11 : u 267. V . - :ho 927. J u hy . a : a 656. hy .he : au 1664. hy . e : u 45I; Path hy ENCE. I. . : a 604. a he :he 411. ha . a :hy 29. a hy : V II. ha . ay : hou 450. ha a :ha 17, 25. ha . e :hy 434- ha a : hau 15- ha . : a 34. ha . a :he 21, 23, 27. ha . ha : 217. hay . e : e 438. ha . a :hu 19. e. e . he : i 205 ; he . a : hi 380 ; he . e : e 463 ; he . he : a 373 eu . a : e 133. Remarks ohs^j^d^^m'fne , &c. 93 0. . hay : ho 394- . V : V 213. .he : ho 335- •hy :he 408. .0 : hy 512. ho .he : e 28, 123. u. 40. Remarks on the G.C.P. Group. i. The most characteristic peculiarity of the vowel alliteration in these three poems is the frequency with which words with initial h take part in it. This is the case in Sir Gawayne and the Green Knight 69 times out of a total of io:i verses with vowel alliteration, in Cleanness 50 times out of loi verses, and in Patience no less than 24 times out of 27 verses. Moreover, in all but a few in- stances the words in question are of English origin. The only romance words I have noted in this connection are honour (Sir G. 1038), honoured {Sir G. 593, 2412, 2520), harnays {Sir G. 590, 2016), hoge {Sir G. 208), honest {C. 14), and haunte (P. 15). It is clear, also, from the form of the indefinite article, where it is prefixed, that the h was really sounded. In every instance we find the form a; thus: a hundreth {Sir G. 1543), a hole (Sir G, 2180), a hepe {C, 1211, P. 380), a horse {C. 1684), a Ml {G. 946), even a hoge — query, by analogy — {Sir G. 208). On the other hand we have an ox {Sir G. 208), an errand {Sir G. 1808), an usage {G. 710), &c. The same difference is shown where the negative adjective is prefixed, the most interesting example being Patience, ver. 394 : Ne non oxe to no hay ne no horse to water. ii. In Sir Gawayne and the Green Knight, it is not un- 94 Dissertatio7i on Alliterative Verse, common for the alliterating letter to occur three times in a first half- verse, e.g. ver. 698 : Alle pe iles of Anglesay on lyft half he haldez, and ver. 2423 : Exellently of alle pyse oper, under heiien-ryche. Luick (§ 84) remarks that in such cases, although the secondary accent bearing the alliterative letter is sometimes very strong, we must not suppose that three arses, that is to say three alliterations proper, are present. This is no doubt true, and with regard to the first of the above instances, similarly constructed verses with consonant alliteration, e. g. : 1778 Alle pe apeches of a-pecialte pat sprang of her mouthe. 1786 For alle pe lufes upon \yue, layne not pe sope, make it clear that the share of the word alle in the alliteration of vv. 698, 2423, is only apparent ; and so again for 1038* : Al pe honour is your siwen, and88i»: Alle of ermyn in erde. The case of the second of the two verses quoted above, viz. 2423, is less easy, and there is a similar difficulty in deciding as to which words alliterate in ver. 24 : pen in any oper pat I wot, syn pat ilk tyme. When two pronouns occur together the usage varies even in the classic period of alliterative verse. The most regular proceeding is for the first to alliterate (cf. Rieger, p. 31), e.g. Christ 324% fed ncenig S^er, or for both to do so, Christ 291% 2dnig o^er, but the second sometimes al- literates alone: Menologiuvi, 197^, swd nan o^er nd de^. In Beowulf 534^ we have Jjonne cenig oder man, which Sievers {P.B.B. x. 239) takes as a B verse, that is to say, as ponne denig 6der man, the only arrangement possible under his five^ype system. But denig can alliterate even before Remarks on Sir Gawayne^ &c. 95 a noun, cp. B. 510, ne inc ebni^ mon, B. 792^ sbni^e J)in^ai B. 3128^, sbni^ne ddl, therefore I should prefer, in the above half-verse, B. 534'', to take wni^ as alliterating, and similarly any in Sio' Gaiuayne and the Green KnigJd, ver. 24. The alliteration in the Destruction of Troy, ver. 1532 : Mijnors of vnarhull ston ^^ Tdony oper thinges appears to justify this, with which verse we may contrast Sir Gatuayne and the Green Knight 551 : Aywan Ssr^ Tirrik &^ oper f id monxj. and with this again ver. 1036 : Boiifre sour &^ Baundyuer, S^ oper auch mony. Also, in Sir Gawayne a7vd the Green Knight 2423, 1 should assign the alliteration to alle^ inasmuch as to give it to oper would involve a double variation from Rieger's rule, and would make the middle thesis abnormally long and heavy. We have probably an exception to the rule in Cleanness 705 : 'Elle'} pay mo'^t Yionestly ayper oper welde. Here the inversion in the second half-verse {oper coming before its governing verb) speaks for giving the alliteration to the second pronoun. Cp. too what was said above on this phrase — (William of Palerne, Remark viii) and its use in an iambic three accent line in Sir Gawayne and the Green Knight, ver. 841 : Ayper oper in 6,rmes con felde. iii. The same difficulty from redundance of the alli- terating letter occurs also in Cleanness, e. g. : C. 42 Oper ani on of alle pyse he schulde he halden utter. „ 1749 He^est of alle oper, saf onelych tiveyne. which, for reasons above given, I have counted as exam- ples of a . a : ha and he . a : o respectively. On the other hand, I have taken the alliteration in ver. 601 : Olde Ahraham in erde one'^ he sytte^, 96 Dissertation on Alliterative Verse. as a . e : o since the adjective olde is here merely an orna- mental epithet, with scarcely more force than such titles as s^yr, liing^ kc, which never (I think) alliterate. Similarly we have good as an epic epithet (gode Gawayn, cp. €a6\ol eratpoi, II. xvi. 327) with g'uas^-alliteration in vv. 109, 2491, but with real alliteration in 2214. Such an epithet, excluded from alliterating, we probably have in Sir Gawayne and the Green Knight 1388 : He has2?pe5 his fayre haZs Ms sarnies ivyth-itme. This seems to justify our disregarding the initial of apel in ver. 258 : J?e Opel aunceteres sune^ pat Adam wat^ called, where to take it as alliterating would give the half- verse a final thesis of inordinate length and weight. Whether in 207*: Of allejjyse ajjel aungelejy we lay the stress on alle or on of el is a matter of no consequence as far as the form of the half-verse is concerned, but I should prefer to let the former have it. The case is different with regard to vv. 1222 : and Yiolkked out his Bxnen y^en heterly hope, and 1707 hagerly in his aune hive his lnQued was couered, where I have treated y^en and hwe as being in the thesis, disregarding their initial letters (cp. v. 1 1 : pay hondel per his BMne body &' Ysen hit hope). In these instances the adjectives are emphatic and must therefore, from their position, alliterate in preference to the nouns following, whilst to consider both adjective and noun as alliterating would unduly extend the Auftakt, and is moreover discountenanced by the case of ver. 1 1 . Similarly in ver. 380 : and lieterhj to pi hyje hylles pay oJed on faste, the alliteration is no doubt as indicated, the context Re7narks on Sir Gawayne, &c, 97 showing that the adjective hy^e is not simply otiose; so too of course ver. 927 : here vtter on a rdunde hil hit howe^ hit one, and V. 1 799 Jbe omementes of gddde^ hous pat lioly were makedj despite the apparent alliteration in hil and hous. And so again (with consonant alliteration), ver. 493 : "Myryhj on a fayr morn monyth pe fxjrst, where the crossed alliteration is evidence for laying the stress on fayr. A verse which seems at first sight to be upon the same footing is 573 : And al wat^ for pis ilk eiiel, pat vnhappen glette. Above, however, we read, ver. 571: FmI felly for pat ilk faute for-ferde a kxjth ryche. The adjective ilea could dispense with alliterating before a noun even in the classical period, e. g. Genesis 1083 : ou pa ilcan tid Tubal Cain (cp. Rieger, p. 30), though this was not common; the alliteration in ver. 573, is therefore at our discretion, since the form of the verse is indecisive ; and similarly in ver. 782*: in pat ilk euentyde. In ver. 1 600 on the other hand : To open vch a hide pyng of Sbunteres uncowpe, the metre speaks clearly for not alliterating with the distributive vcha (=ilka). Cp. too Cleanness, vv. 333,334, 335> 33^^ ^^33 ; Patience 361, 414. Analogous to the cases above is that of vv. 874*: pat a'^ly hurled in his eres, and 1204* and harde hurles pur^ pe oste, but the decision is here easier, since even in the oldest English verse the adverb could precede the verb without alliterating [Rieger, pp. 27, 28], and therefore we may 98 Dissertation on Alliterative Verse, without hesitation ignore the initials of a^ly and harde. Still more certainly may we ignore that of al in C. 643 : Abraham, al hodles imth arme'^ vp f olden, remembering the naturally proclitic character of adverbs of degree (Rieger, p. 39). In the same way we very often meet with the adverb fvl not alliteratiQg (cp. Sir Gawayne and the Green Knight 788). In C. 345 on the other hand, ^Now Noe' quoth o\xre lorde, ^ art J?ou al redy,* (a quaint line with unusually weak alliteration), al has the chief letter. It must however be acknowledged that the most we can attain to with regard to cases like the foregoing is pro- bability, more or less high, since the author of these poems, though a maker of some skill does not always conform to the strict rules of rhythm, obeyed by the poets of a better day. For instance in Cleanness, ver. 40* : With rent cokre'} at the kne, and Patience 531* J>at pacience is a ndbel poynt, there is no justification for the adjectives' not alliterating. iv. As interesting cases of alliteration may be mentioned : (i) One instance of the so-called Ti-alliteration, viz. Cleanness, 1779 Withinne an cure of the ny^t an entr4 pay hade. (2) The use of asked in Patience 195 penne ascryed pay hym skete, dr' aaked (!) ful loude. cp. William of Paleme 69 sche "kolled it ful "kindly 6-» askes (!) is name. (3) The contrast between the alliteration in Cleanness, ver. 14 : As he "honest sri-wyth, &" in-with all^ fy^P^'i^ and that in 20 Bope "with-inne dr^ "with-outen, in "wede^ ful bry^f. (4) The rare instances of un- alliterating in Cleanness. viz. unhap, ver. 892 (unhap, vv. 143, 150) ; ungoderly, Piers Plowman. 99 ver. 1092 {u7igoderly, ver. 145) ; nnj/ryftly, ver. 267, and possibly also Xinclannesse, ver. 30. (5) Lastly, the alliteration of vpso-doun in Patience, ver. 362. Piers Plowman. A-text. Prologue to Passus viii. B-text. „ to „ vii. C-text. Passus i. to ,, . x. A. iii. 247 = B. iii. 260 =C. iv. 418; A. viii. 119 — B. vii. 132 ; C. vii. 79. C. iv. 462 (cp. B. iii. 304 a . ha ; e) ; C. vi. 197. B. vi. 224 = C. ix. 231. A. iii. 204 = B. iii. 210 =C. iv. 268. j B. iii. 284 = C. iv. 442 (cp. A. iii. 266 a . he : ha) ; I C. iv. 237. B. iii. 304 (cp. C. iv. 462 a . a : e). A. iii. 64 (cp. B. iii. 72 a . hi : he).-^_ B. iii. 346 = 0. iv. 499, A. V. 220= B. v. 390=0. vii. 440. B.v. 366 = 0. vii. 417 (cp.A. V. 210 a . - : a) ; B. V. 612 = 0. viii. 250 (cp. A.vi. 93 a . e : e) ; 0. ii. 289 (cp. B. i. 189 a . hi : a) ; 0. vi. 177 ; 0. ix. 283. A. i. 63 = B. i. 65 = 0. ii. 61. B. V. 543 = 0. viii. 181. 0. i. 3 (cp. A. Pr. 3 = B. Pr. 3 ha . he : ho) ; 0. i. 30 (cp. A. Pr. 28 =B. Pr. 28 a . he : ho). A. iii. 266 (cp. B. iii. 284=0. iv. 442 a . ha : ha). 0. ii. 46 (cp. A. i. 47 = B. i. 49 a . he : who). B. vi. 147=0. ix. 146 (cp. A. vii. 134 a . he : ho). 0. ii. 108. 28 =B. Pr. 28 (cp. 0. i. 30 a . e : ho) ; A. vii. (cp. B. vi. 147 = 0. ix. 146 a . he : e). A. i. 47 = B. i. 49 (cp. 0. ii. 46). 0. ii. 64 (cp. A. i. 66 e . - : ho ; B. i. 68 - . e : ho). B. Pr. 140. B. i. 189 (cp. 0. ii. 289 a . e : a). B. iii. 72 (cp. A, iii. 64 a . ha : hu). wha i ^' ^^^' *^'* ^^P* ^' "^* ^^^ w. - : w : 0. iv. 422 ( ha . o : a). Jl 2 a . a : e a . a : a . a :ho a .ha :ha a .ha : e a .ha :hu a . au : a a . au :ha a . e : e a . e : i a . e :ho a .he :ha a .he : wha a .he : e a .he : a .he :ho a .he : who a .he :liy a . hei : a a .hi : a a .hi :he { A. Pr. ^ 134 ioo Dissertation on Alliterative Verse, ha . ha ha . ha ha . au ha . e : ay ha . e : ye ha . e : ey ha ha ha ha ha ha . o : a a .ho : ha B. v. 603 (cp. C. viii. 242 ho . a : ho). a . hu : he A. viii. 6 (cp. B. vii. 6 ha . e : ss ; C. x. 6 hu . e : ss). a . huy : e C. iv. 303. ha . a : ha A. i. 171 = 6. i. 195 - C. ii. 193. ha . a : whi A. vii. 135. ha . ha : a A. v. 221 ; B. v. 391 = C. vii. 441. B. Pr. 179 (cp. C. i. 194 ho . ha : e) ; B. iii. 180= C. iv. 227 (cp. A. iii. 174 ho . - : e) ; C. iv. 300. ha . ha : u A. iv. 106. ha . au : ha C. viii. 94. C. iv. 435 (cp. A. iii. 263 ha . e : hei ; B. iii. 277 ha . e : ey). ha . e : ye B. i. 152 (cp. C. ii. 151 ha . e : 50). B. iii. 277 (cp. A. iii. 263 ha . e : hei ; C. iv. 435 ha . e : ay). , . J A. iii. 263 (cp. above) ; B. vi. 4 (cp. C. ix. 2 ( ha . e : hy). ha . e : 50 C. ii. 151 (cp. B. i. 151 ha . e : e). ha . e : hy C. ix. 2 (cp. B. vi. 4 ha . e : hei). B. vii. 6 (cp. C. X. 6 hu . e : s . s and A. viii. 6 a . hu : he). ha . he : ey C. x. 102. C. iv. 422 (cp. A. iii. 252 w . - : w ; B. iii. 264 a . o : wha). ha . ho : a C. iv. 302. / C. viii. 76; 267 = (cp. B. v. 624 = A. vi. 104 ha. ha . ho : e | hu : e) ; C. x. 5 (cp. B. vii. 5 ho . ho , e ; A. viii. ' 4 ho . ho : he). ha . hu : e A. vi. 104 (cp. C. viii. 76 ha . ho : e). ai . he : a C. vi. 59. ay . ho : ha C. vii. 255 (cp. B. v. 265 y . a : ha). J A. V. 56 (cp. B. V. 73 m . m : ?) ; A. v. 210 (cp. B. ( V. 366=0. vii. 417 a . e : a). .a : au B. v. 640 (cp. viii. 284 - . a : ou). a . - : e C. iv. 345, 397. a . - : hei B. i. 156 (cp. C. ii. 155 a . - : hy). a . - : hey B. v. 300 (cp. A. vii. 347 a . - : hy). .a : e B. vi. 266 (cp. A. vii. 251 - . a : ^e). .a : 5e A. vii. 251 (cp. above). a . - : o B. iv. 38. a . - : ho A. V. 234 = 6. v. 461 =C. viii. 68 ; C. iv. 411. a . - : V C. iv. 338. r C. ii. 155 (cp. B. i. 156 a . - : hei) ; C. vii. 347 (cp- a . - : hy -^ r^ r. \ I B. V. 300 a . - : hey). lia . - : a C. vii. 264. au . - : a B. v. I53=C. vii. 127, Piers Plowman. lOI . au : whe au au au ei ey who C. vii. 224 (cp. A. V. 132 au . - : who ; B. v. 218 ow . au : who). A. V. 90 (cp. B. V. 109 au . - : ey). B. V. 109 (cp. above). A. V. 132 (cp. B. V. 218 ow . au : who ; C. vii. 224 - . au : whe). ay : ha C. iv. 324. e. e . a : a 6 . a : e .ha : a e .ha :he e .ha : hey e .ha :ho e .ha :how 6 .ha : hoy e .ai :he e .ai e . e : a e . e e e . e •he e . e i e . e e . e y e . e hy e .he : a e .he : e e .he : ho e . eo : a e . i : who e .hi .he e . • a e . he e . e .ho :he e .ho : hu e . u : ha e . V V e .hu : a C. viii. 173 (cp. A. vi. 19 y . a : o ; B. v. 535 e . a : o). B. v. 535 (cp. above). B. vi. 5 = C. ix. 3. B. v. 466=0. vii. 313 (cp. A. V. 239 V . ha : he). C. X. 188. B. vi. io8 = C. ix. 113 (cp. A. vii. 99 he . ha : he). B. vi. 118 (cp. C. ix. 123 e . ha : hoy ; A. vii. 109 ho . he : hey). C. ix. 123 (cp. above). C. ii. 127 (cp. A. i. 114 ei . eo : he ; B. i. 123 ey . e : he). C. vii. 188. B. ii. 83 (cp. A. ii. 63 e . e : e ; C. iii. e . e : y). B. iii. 2I2 = C. iv. 270 (cp. A. iii. 206 e . eo : a). A. ii. 63 (see above). A. vii. 126. A. i. 120 = B. i. 129. C. X. 214. C. iii. 88 (cp. A. ii. 63 e . e : e ; B. ii. 83 e . e : a). C. X. 203. A. V. 59 = B. V. 76 = 0. vii. 63. C. ii. 133. A. Pr. 50= C. i. 51 (cp. B. Pr. 53 he . he : ho) ; A. v. 235 = B. V. 462 (cp. C. viii. 69 hu . e : ho). A. iii. 206 (cp. B. iii. 212, C. iv. 270 e . e : a). B. V. 40. A. vi. 99 (cp. B. V. 619- C. viii. 262 e . hy : he). B. ii. 173. C. i. 123. B. iii. 237. C. iv. 243. A. vii. 248 = B. vi. 263 - C. ix. 273. C. vii. 307. C. vii. 87 (cp. B. V. 121 e . y : y). C. i. 1 4 (A. Pr. 1 3 heo . e : hei ; B. Pr. 1 3 he . e : hie). e .hu :he e • y : y e .iiy :he he . a : a he . a : ay he .ha : e he . e : a he . e :ey . 6 :hie . e :ho 1 02 Dissertation on Alliterative Ve^-se. C. ix. 263. B. V. 121 (cp. C. vii. 87 e . V : v). B. V. 619 = C. viii. 262 (cp. A. vi. 99 e . hi : he). A. vii. 198 (cp. B. vi. 212 he . a : ay=C. ix. 219). B. vi. 2i2 = C. ix. 219. B. ii. 228 = 0. iii. 238 (cp. A. ii. 204 heo . ha : e). C. iv. 235. C. iv. 94 ; C. ix. 66 (cp. A. vii. 60 he . he : e ; B. vi. 67 he . e : he). B. vi. 67 (cp. above) ; C. ii. 17 (cp. A. i. 17 hi . eo : v ; he . e : he ^ B. i. 17 hy . e : he) ; C. ix. 183 (cp. A. vii. 177 he . he : he = B. vi. 190). C. vi. 153. B. Pr. 13 (cp. A. Pr. i3heo . e : hei ; C. i. 14 e . hu : a). A. v. 249. ( B. iii. 129^0. iv. 166 (cp. A. iii. 125 he . he : eo) ; he .he : e < A. vii. 60 (cp. B. vi. 67 he . e : he ; C. ix. 66 V he . e : e). A. iii. 125 (cp. above). B. iii. 241. C. iv. 252. A. i. 114 (B. i. 123 ey . e : he ; C. ii. 127 e . ai : he). A. vii. 300 (cp. B. vi. 314 hei . huy : e ; C. ix. 336 hey . hy : e). A. ii. 204 (cp. B. ii. 228 = C. iii. 238 he . ha : e). A. Pr. 13 (cp. B. Pr. 13 he . e : hie ; C. i. 14 e . hu : a). B.i.123 (cp.A.i. ii4ei. eo : he; C.ii.i27e.ai: he). B. Pr. 128. C. ix. 336 (cp. A. vii. 300 hei . hu : e B. vi. 314 hei . huy : e). B. vii. i6o=C. X. 309 (cp. A. viii. 146, 147). A.i. 66 (cp. B. i. 68 - . e : ho ; C. ii. 64 a . he : hy). B. i. 68 (cp. above). C. X. 315. C. i. 55 (cp. A. Pr. 54= B. Pr. 57 - . he : e). A. Pr. 54= B. Pr. 57 (cp. above). A. V. 91 (cp. B. V. no ho . e : ha). he .he : eo he .1 :ha he .ho : e ei . eo :he hei .hu : e heo .ha : e heo . e :hei ey . e :he ey .hie : a hey .hy : e e . - :hai e . - :ho - . e :ho e . - :hy - . e : ei - .he : e hey . - : a 1. i .a : a A. iii. 249. i .a : au B. v. 640 (cp. C. viii. 284 i . a : ou). i .a : ou C. viii. 284 (cp. above). hi . hei : e A. viii. 4 (cp. B. vii. 4 hy . hei : e ; C. x. 4 hy . hey : e). hi . eo :v A. i. 17 (cp.B. i. 17 hy . e : he ; C. ii. 17 he . e : he) Piers Plowman. 105 hi . o : he I ^' ^' ^^^ ^^P" ^- ^- ^^^ hi . ho : ha ; C. vii. 389 ( hi . ha : ha). !A. V. 172 (cp. B. V. 329 = C. vii. 378 hi . ha : ho) ; A. V. 185 (B. V. 341 hi . ho : ho ; C. vii. 391 ha . ho : ho), hi . o : V B. V. 164 (cp. C. vii. 149 - . ho : of). i . - : i C. iv. 373. O. o .a : hu A. vi. 109 = B. v. 629 = 0. viii. 272. o . e : ha B. v. 409=0. viii. 18. o . e : wha A. ii. 96. o . e : e B. ii. 14. o . e : he B. Pr. 164. o .he : he B. Pr. 106 ; B. v. 233 = 0. vii. 235 ; 0. iv. 340. o .0 : e 0. X. 211. o . ho : ha A. vii. 76 = B. vi. 85 = 0. ix. 92 ; 0. vii. 193. o . ho : he A. viii. 83 = B. vii. 99=0. x. 175 ; 0. vii. 19. , . , I 0. iii. 228 (cp. A. ii. 194 o . hu : ho ; B. ii. 218 I o . how : ho). o . how : ho B. ii. 218 (cp. above). o . hu : ho A. ii. 194 (cp. above). ..__ o . hy : hi B. vi. 133 (A. vii. 124 ho . hy : ou). ho . a : ho 0. viii. 242 (cp. B. v. 603 a . ho : ha). ho . ha : a A. vii. 283 = B. vi. 298 = 0. ix. 320. ho . ha : e 0. i. 194 (cp. B. Pr. 179 ha . ha : e). , , \ A. vii. 175 (cp. B. vi. 188 hu . ha : he ; 0. ix. 181 ho .ha :v ^ -Z \ \ ho . ha : he). ho . e : ha B. v. no (cp.A.v. 91 hey. - : a); O.vii. 306; x. 190. ho . e : hy 0. x. 201. ho . he : e A. v. i68=B. v. 325 = 0. vii. 374. ho . ho : a 0. x. 219 ; B. v. 227 (cp. below). . j A. V. 141 = 0. vii.233(cp. B.v. 227ho . ho : a; B. I vii. 5 (cp. 0.x. 5 ha. ho :e; A, viii. 5 ho. ho: he). , ^ A. vii. 190 (cp. B. vi. 203 ho . ow : ho ; 0. ix. 207 ho . ou : ho < , f \ ( ho . ow : ha). ho . ow : ha 0. ix. 207 (cp. above). ho . ow : ho B. vi. 203 (cp. above). ho . hu : e A. vii. 2i8=B. vi. 233=0. ix 238. ho . hy : ou A. vii. 124 (cp. B. vi. 133 o . hy : hi). , j B. V. 218 (cp. A. V. 132 au . - : who ; 0. vii. 224 I - . au : whe). a 0. X. 235. ow . au o . - : hy 0. viii. 249 (cp. A. vi. 92 = 6. v. 61 1 w . w w I04 Dissertation on Alliterative Verse. ho . - : G A. iii. 174 (cp. B. iii. i8o=C. iv. 227 ha . ha : e). - . ho : o C. vii. 149 (cp. B. v. 164 hi . o : v). ho . - : o C. ii. 72 (cp. A. i. 73 = B. i. 75 ho . — ou). ho . - : ou A. i. 73 = B. i. 75 (cp. above). y u. u .a : a B. ii. 86 (cp. A. ii. 66 y . v : a ; C. iii. 91 v . a : o). u . a : o A. viii. 40; C. iii. 91 (cp. above). V . ha : he A. v. 239 (cp. B. v. 466 = C. vii. 313 e . ha : he). u . u : a B. V. 240=0. vii. 239. V .V : ho A. viii. 98 = 0. x. 290 (cp. B. vii. 114 y . y : ho), hu . a : ha A. vi. ii=B. v. 527 (cp. C. viii. 165 ho . ha : ha), hu . ha : a B. vi. 298 = 0. ix. 320 (cp. A. vii. 283 ho . ha : a). hu . e : ho C. viii. 69 (cp. A. v. 235 e . he : ho = B. v. 462). hu .he : e C. iv. 310. y- y .a : ha B. v. 265 (cp. C. vii. 255 ay . ho : ha). A. vi. 19 (cp. B. V. 536 e . a:o; C. viii. i73e.a:a) ; B. V. 410= C. viii. 19. y .8 : o C. X. 152. y .V : a A. ii. 66 (cp. B. ii. 86 u . a : a ; C. iii. 91 v . a : o). y . y : ho B. vii. 1 14 (cp. A. viii. 98 = 0. x. 290 v . v : ho), hy . e : he B. i. 17 (A. i. 17 hi . eo : v ; C. ii. 17 he . e : he). hy . hei :e B. vii.4 (cp. A.viii. 4hi.hei:e; C.x.4hy. hey : e). hy . hey : e C. x. 4 (cp. above). y . - : e C. vii. 20. y . - : i A. i. 48 = B. i. 50. Remarks. i. In the foregoing table the sign = indicates that the alliteration is the same for the references so connected. Variations are enclosed in brackets. Where a reference stands by itself, the verse referred to has no counterpart in the other texts. ii. The alliteration is identical for all three texts in four- teen instances. The A. text contains eight verses with Piers Plowman, 105 vowel alliteration not found in the other two texts, the B. text ten, and the C. text forty -four. Vowel alliteration occurs in the A. text in about ^i'^ per cent, of the whole number of verses in the Vision, in the B. text in about 3-8 per cent., and in the C. text in 3-9 per cent. Thus Lang- land became somewhat more free in his use of this kind of alliteration as he grew older. Rosenthal calls attention to the care which he bestowed in his later versions to remedy verses defective in his earlier ones. We may see this from the following examples : A. V. 210 And after al this sutfet . a7t accesse he hedde B. V. 366 Atid after al this excesse • he had an accidie (cp. C. vii. 417). Here clearly, excesse in B. v. ^66 replaces surfet in A. v. 210 in order to produce an alliteration with al. For accenting the word on the first syllable we have the evidence of Skccesse in the verse from A., which word also shows that we must give the alliteration in ^66^ to accidie. Compare again : A. vi. 93 Tho that Adam &= Ewe . eeten heore bone. B. V. 61 3 Tho Adam 6r» "Eue . eten apples unrosted. C. viii. 250 That Adam &' 'Eue . a^ens ous alle shutte, in which an amusing conflict goes on between sound and sense, which are at length satisfactorily harmonized in the third example (cp. below, A. iii. 174, &c.). Further : B. V. 543 Axen after hym er . til note in this place ; ) C. viii. 181 Asken after hym er noio . in thys like place, ) also B. ii. 8^ And the erldome of etiuye . and wratthe togideres. ) C. iii. 88 The erldom of emiye . and yre he hym graunteth. ) and A. i. 17 And for he hihte the eorthe . to semen ow •srchone. ) B. i. 17 And therfore he hy^te the eHhe • to help ^otv vchone. ) and A. iii. 174 TTiow hast Yionged on my nekke . enleue tymes. \ B. iii. 180 Thow hast hanged on myne half • elleuene tymes ' C. iv. 227 Thotv hast hanged on myn hols - elleuen tymes ' and A. V. 91 IIou Heyne hath a newe cote . and his wyf another. ) B. V. 1 10 And biholde haiv "Eleyne . hath a newe cote. ) io6 Dissertation on Alliterative Verse. all cases in which deficient or defective alliteration in the earlier version has been remedied in the later. Cp. also : A. vii. 60, B. vi. 67, C. ix. 66 ; B. i. 89 and C. ii. 189 ; B, i. 6^ and C. ii. 64 ; B. v. 164 and C. vii. 149 ; B. iii. 264 and C. iv. 422. A needless repetition has been avoided and at the same time the form of the verse gi^eatly im- proved in C. i. 193*: 'Ne haue it Taongid ahoute hus ha,ls, for B. Pr. 179 Ne hangen it ahoute the cattes hafe, where cattes is quite out of place in the thesis. However it is doubtful if this consideration weighed with William for a little earlier (B. Pr. 170; C. i. 185) we have this word in the same anomalous position in both texts. Some of the verses in A. and B. afterwards di-opped altogether were probably left out as past mending, par- ticularly B. iv. 38 (a hopeless verse), perhaps also A. v. ^6 ; vii. 12,6 ; iii. 249. On the other hand A. viii. 40 which has been omitted from B. and C. is metrically faultless and is an unusually good specimen of vowel alliteration, viz. u . a:o: "Vsure and auarice • and othes I defende. In some cases the alliteration has certainly been changed for the worse, e. g. in C. viii. 173 : In "Ermanie. in AUsaundre . and in 'Damascle. Cp. B. V. 535 /w 'Ermonye, in AUsaundre • in many other places, although the latter verse is perhaps not faultless, seeing that the natural stress in reading would fall on Ttiany as in the already quoted Destruction of Troy 1532 : ULynors of marbull ston &' mony oper thinges (cf. Luick, p. 397). So again A. vii. 175^ -vpfor to loke has been, with doubtful taste changed in B. vi. 188^ into on hem /or to loke. iii. Langland is the least scrupulous of the fourteenth century poets in admitting words of slight sentence stress to alliterative rank. An example of this is his frequent Piers Plowman, 107 use of liave (and its forms) in alliteration, and that not only as a principal but also as an auxiliary verb. As a princi- pal verb have occasionally alliterates in the classical period, thus, h.abban twice in Beoivulf 446*^, 46 2^ and once in the Elene 621^, and Jiafa (imperative) Beowulf 659^. It is however usually unaccented (cf. Hinze, Zum Andreas, p. 5), and as an auxiliary never alliterates in the best period, as far as I know. In the Vision the word alliterates as an infinitive in A. vii. i'^^'^, C. vii. 255^, B. v. 466*, &c., as an imperative in C. vii. 307^, and as an auxiliary in A. v. 220^, B. V. 391* and B. i. 152*. Occasionally it receives ad- ditional weight by an inversion. Thus in C. x. 187 : And alle holy hermites . h&ue schal the same, and B. v. 391 Afid ^it hSiue I hated hir . al my lyf-tyme. The latter verse has been changed from And ^it ich&ue i-hated hire, &c. (A. text v. 221*) Such a change is I think contrary to Langland's usual practice, which was rather to substitute direct for earlier inverted constructions. This he does in C. viii. 242*: The hokes aren Almys-dedes, for Of Almes-dedes ar the hokes, B. v. 603 ; also in C. ix. 320* Hunger eet al in haste, for B. vi. 298^ Al Hunger eet in hast ; and in C. viii. 69* Til ich haue hurd euesong, for B. V. 462 Til ich haue euensonge herde. Another word of weak force which sometimes alliterates, but more often not, is the indefinite all. In A. iii. 247^ and al his peple aftur (cf. B. iii. 260) we might doubt whether al or aftur has the chief letter, were it not for the variation in C. iv. 418^ and aZ hus lyge puple, which verse is again a guide for the alliteration in B. v. 227 a,l hire lyftyme. Where the expression alle that occurs, the word alle for the most part does not alliterate, cp. B. i. 11 8, B. vii. 6, C. ix. 66, &c. We have however C. vi. 59^ a,lle that ben croun^de and A. viii. 6* and oil that euere hulpen him, the word euere in the second example being of secondary io8 Dissertation on Alliterative Verse, force to aZ, the sense of which it generalizes, as of wlto in our modern whoever. With regard to the alliteration of such slight words as in (A. i. 128^), and (A. iii. 249^) I have nothing to add to what has been already said (cp. William of Palerne^ Remark iii). iv. The expression either other comes before us again in C. iv. 340, C. vii. 149, B. V. 164. The last two cases are similar to William of Palerne, ver. loio (cp. Remarks ix). C. iv. 340 is doubtless to be scanned thus : And ayther ys operes heZ^ . of hem cometh retribution^ for (i) it is clear that help is in one of the arses of the first half-verse and therefore (2) that ojjeres must be in the other, inasmuch as preceding help, which governs it in the geni- tive, it has naturally a still stronger stress. Thus in no case of this combination that we have examined does the word either alliterate. V. I have noticed but one case of 7i-alliteration common to all three texts, viz. that in A. iii. 265 In auenture hit nuy^ed me . an ende ivol I make. B. iii. 279 An auenture it noyed men . none ende wil I make. and C. iv. 437 An aunter hit nuyede me - non ende tvol ich make. The A. text contains another example, viz. vi. 119 : * No,' quath an apeward . '■for nout that I knowe ! ' which has been altered in B. v. 640 to ' Ne I,' quod an apewarde . * bi a,u^te that I knotve ! ' and similarly in C. viii. 284. Beyond these instances I have observed no other in any of the texts. Ordinary alliteration with n is also rare in the Vision, occurring in the A. text about fifteen times in 1833 "^^-5 ^i^* ^^- 97' ^^5> 144; iii. 182; V. 181; vi. 61, 64; vii. 14, 136, 138, 228 231, 293; viii. 139 ; ix. 106. (For the alliteration of the pronoun 'I' in B. v. 640, C. viii. 284 see Morte Arthurs, Remark vii.) vi. Vowel alliteration is mixed with /^-alliteration to a still greater degree in the Vidon than in the G. C. P. Concluding Remarks. 109 group (=Sir Gawayne and the Green Knight, &c.). In the latter poems we find this combination in about 62' 3 per cent, of the verses alliterating with the vowel, in the Vision we have it in as many as 74-3 per cent. Unlike the G. C. P, group, however, the Vision makes no distinction between words commencing with a vowel, and words with initial h (see G. C. P. Remark i). In the G. C. P. group the indefinite article a is prefixed even to a romance word beginning with h {a hoge G. 208) ; in the Vision we invariably find the form an even before English words with initial h. Thus, just as, before initial vowels, we have : an archangel, an auenture, an aunte, an accesse, non ale, non ende (in A. and B., but in C. — viii. 68, ii. 129—710 ale, no ende). So we have before h : — an hermite A. Pr. 3 (cp. B. Pr. 3) ; an hep A. Pr. 50 (cp. B. Pr. 53 C. i. 51) ; and again A. v. 168 (cp. B. v. 325, C. vii. 374), also A. v. 178 (cp. C. vii. 385) ; an hundred A. Pr. 84, A. vi. 11, A. vi. 104 (cp. B. Pr. 210, V. 527, 624 ; C. i. 159, viii. 165, 267) ; an heui hridel A. iv. 20; an hote heniest C. vi. 7 ; an home C. iv. 16 ; an heire C. vii. 6 (cp. A. v. 48. B. V. 66) ; an half aJcer A. vii. 4 (cp. B. vi. 4, C. ix. 2), and similarly, none hors A. ii. 144, non halfpeny ale A. vii. 293 (cp. B. vi. 307, C. ix. 329). Concluding Remarks. Vowel alliteration in the fourteenth century was evidently in a moribund condition : for how long past it has been practically dead would require a further investigation. In one poem only — the Morte Arthure — does it show any- thing like the vigour and variety with which it appears in Beowulf. In Piei^s Plowman, except when associated with h, its presence is almost imperceptible ; in the Destruction of Troy, William of Palerne, and especially in the Alexander, it has become for the most part a monotonous repetition of the same vowel. In Beoumlf, on the other no Dissertation on Alliterative Verse. hand, there is only one example, viz., ver. 836 (ea . ea : ea) of identically the same vowel in all three arses of the verse. Such a difference points to a difference of feeling with regard to vowel alliteration. It is clear that, whilst the Beowulf poet sought variety, the poets of the fourteenth century sought identity of vowel. The explanation seems to be this. That which alliterated for the Beoivulf poem was, as Rieger says (p. 16), the spiritus lenis, or, better, the glottal catch. In the fourteenth century this was, in the South of England at least, no longer or but faintly heard, and therefore the poets, feeling that alliteration is entweder genau or gar nichts (Rieger, 1 6), strove to alliterate as far as possible with the same vowel. To do so altogether would, of course, have been intolerably irksome. I hazard the conjecture that the reason why vowel alliteration still has much of its old character in the Morte Arthure is, that the glottal catch was retained longer in Scotland, where this poem is generally allowed to have been written, than in England. Whether any trace of it is still to be heard in the language of the people I cannot say, but it is significant that its companion sound — the h or spiritus asper — is much better preserved north of the Tweed than south. This does not seem to be the effect merely of the better education which has prevailed among the lower classes in Scotland. Indeed in one case an h has been retained which was given up by orthography centuries ago, viz., in the expression ' ThaVs it ' still, I am told, pronounced ' Thafs hit * by many people. The retention of the guttural in loch, and other words is a similar phenomenon. In the Morte Arthure the so-called alliteration between spiritus lenis and spiritus asper is very rare though not — as Trautmann asserted — wholly unknown. In the G.C.P. group it is remarkably common, and that the h was not mute in such cases, the form of the indefinite article used before it shows. In Piers Plowmaii, where this alliterative Concluding Remarks, 1 1 1 combination is still more common, but where the indefinite article has the same form before }i as before a vowel, I strongly suspect the h was pretty much a dead letter. In one respect vowel alliteration may be said to begin with the fourteenth century, for the glottal catch is, like the h, a throat consonant, differing from it in possessing no symbol. It is not then the vowels themselves which al- literate in Beoiuulf, but this unwritten consonant which precedes them. On the other hand, in the a . a : a combination so common in the fourteenth century, it is really the vowels which alliterate, according to the view expressed above. We have already mentioned that where vowels alliterate (or appear to do so) in Beowulf, it is usually different ones which are used. Similar variety in the vowels of alliterating syllables is seen when these commence with a consonant. I noted above (Chap. II. p. 56) a passage in Beowulf (320-324) where vowel assonances were strongly marked. The vowels following the alliterating letters are, however, ^varied, e. g. : str&t—st§i.n—stig; gum. — gad. — gtii^; hesird — hond — hring ; song. — sear.— se^e ; gryre—geat. — ga.ngan. We have only to glance down a page of Beowulf to see that assonance between alliterating syllables is not sought for. To show this clearly the vowels in the alliterating syllables of the first fifty verses of Beowulf with consonant alliteration, are here added : — I. & . - : e& V. 17. u ea V.30. . eo : i 2. eo - : y . » 18. eo § : ab M 31- eo . a : a 3- y g : ea „ 19- y - : e „ 32. f • - : i 5- ss : eo „ 21. eo : 8B „ 34- § - : eo 7. ed u : d „ 23. i . - : i M 35- ed y : ea 8. eo : eo „ 24. eo . £8 : » 36. cb ad : a 10. - : f „ 25. 89 . 8b : a M 37. eo . - : SB II. • y : 6 „ 26. y . - : SB „ 38. y - : eo 13- eo . ea : „ 27. e e : ed ,. 39- i - : ea 14. . 6 : y „ 28. - . fib : i ,. 40. i y : ea 16. a . - : i >, 29. & . i : e „ 41- & . ae : i 112 Dissertation on Alliterative Verse, V.42. 6 M 43- - » 44. eo » 45- u „ 47. - „ 48. ea ea : eo V.49. : ^ 5, 50. : ^ » 51- : »» 52. : e ,, 53- : M 54- a : ed V.55. 8b : 8B 6 : e ,, 57- ea ea : eo 6 : e „ 58. a a : 8B eo : 8B M 59- eo - : u : eo „ 60. . 6 : eo eo : The variety here is not far short of what has been seen in the case of syllables with initial vowels. I conclude, therefore, (i) that Vietor is right in taking vowel alliteration (so-called) as an evidence that in the oldest English the glottal catch was heard before initial vowels, and (2) that the altered character of the vowel alliteration in the Early English poems which we have examined is a reason for supposing that this sound was already fast disappearing in the fourteenth century, and in the south of England, at all events, was abeady extinct. The very small place which vowel alliteration holds in Lay anion suggests, in the same way, that in Gloucester- shire the glottal catch had been discontinued early in the thirteenth century. In the first 4000 lines of the poem (which are all that I have examined), such scanty instances of vowel alliteration as occur are mostly confined to the letter a. K. Kegel's article ' Die Alliteration in Lajanion ' {Germanistische Studien, i. 171), amidst all the wealth of alliterative phraseology which it exhibits, as inherited by Layanion from the older poets, gives only the' following examples of alliteration upon vowels : V. 5202 (Calig. MS.) „ 15770 „ „ 16234 5 „ 19104 20045 Parallelisms. mid orde and mid ^gge. ord fram pan ende. Old Formulae. J?er wes pe B^ele eorl. Aldolf ihaten. and pe eorl Aldolf scat beon ure aXdre. to gaines Ardure B^elest kingen. (cp. 22983.) (cp. 16559.) Concluding Remarks. V. 29476 (Calig. MS.) pa seiden men Anglisce SL^ele ibome. „ 29481 ,, „ Iwis ge heod Anglisce "Englen iUcchest. The Ormulum is not an alliterative poem, but, never- theless, contains a good deal of alliteration, and (in keeping with its more northern origin) has more variety of vowel alliteration than Lay anion. The best examples among the first 10,000 verses are the following : 6775 WippMtenn ord and ende. Cp. 9676. 3577 and d beoJ> hutenn ende. (Cp. Satan 315 ; also Orm. 409, 4208, 4827, 8879.) 7570 Att nre lifess ende. Cp. 10256; also 2512. 4803 purrh an full sdell a,dle. 5055 And Sin full apell kinde. 5146 Forr lufe off &nig operr mann. Cp. Beowulf $0^, 534. 8174 patt anig mann nagg e^ghenn. 4187 Off ehhte daghess elde. Cp. 4229, 8396. These appear to be undoubted specimens: others which I have noted are more doubtful. Indeed, it is only by comparison of several verses that we can in some cases be sure that alliteration was intended. Thus, in Ormulurri 3385 purrh Godess enngless a.vrtvnedd there is apparent alliteration between the last two words, and comparison with 3395, 3429, 3507, 4004, 9608, confirms it. Again the alliteration in 3699 Wipp enngless eche fode ^ is attested by vv. 3867, 4369, and so on. a RETURN TO the circulation desk of any University of California Library or to the NORTHERN REGIONAL LIBRARY FACILITY BIdg. 400, Richnnond Field Station University of California Richmond, CA 94804-4698 ALL BOOKS MAY BE RECALLED AFTER 7 DAYS 2-month loans may be renewed by calling (510)642-6753 1-year loans may be recharged by bringing books to NRLF Renewals and recharges may be made 4 days prior to due date DUE AS STAMPED BELOW SEPl 1993 YC ' ^'^Co / r}?t^^^:s^i