(ii) The sub-type B2, or B with disyllabic second thesis, is rarely found when the first thesis has only one syllable, þe drýhtnes bibṓd Cri. 1159, þū wá̄st gif hit is Beow. 272, þām wī́fe þā wórd Beow. 640; with resolution of the first arsis, þurh dároða gedrép An. 1446, and of the second, þurh níhta genípu Gū. 321; it is commoner with a disyllabic first thesis, þā of wéalle geséah Beow. 229, hē þæs frṓfre gebā́d Beow. 76; with resolution of the first arsis, mid his hǽleða gedríht Beow. 663, ofer wároða gewéorp An. 306; with trisyllabic first thesis, þonne hē ǣ́r oððe sī́ð El. 74, wes þū ūs lā́rena gṓd Beow. 269; with resolution of the first arsis, þēah hē þǣr mónige geséah Beow. 1614, and of the second arsis, þæt nǣfre Gréndel swā féla Beow. 592; with first thesis of four and five syllables, hwæðre hē in brḗostum þā gít An. 51, þæs be hire sē wílla gelámp Beow. 627.

Verses with trisyllabic second thesis are extremely rare and doubtful.[76] It should be noticed that, in this second type too, the thesis seldom consists of a second part of a compound, as hine fýrwìt brǽc Beow. 232, the exceptions are proper names, as nū ic Bḗowùlf þéc Beow. 947, ne wearð Héremṑd swā́ Beow. 1710.

Type B, according to Sievers, occurs 1014 times in Beowulf, of which 293 are in the first hemistich and 721 in the second.

§ 27. The Type C has three sub-types: (i) C1, the normal type ×–́|–́×, without resolution, as oft Scýld Scḗfing Beow. 4, gebū̀n hǽfdon 117. Here too the first thesis may consist of two, three, four, or five syllables, þæt hīe ǣ́ghwýlcne An. 26, þone gód sénde Beow. 13, ofer hrónrā́de Beow. 10, ǣr hē onwég hwúrfe Beow. 264, mid þǣre wǽlfýlle Beow. 125, þe ic him tṓ sḗce El. 319 þāra þe mid Bḗowúlfe Beow. 1052, oð þæt hine sémnínga An. 821, þāra þe hē him míd hǽfde Beow. 1625, swylce hīe ofer sǣ́e cṓmon, An. 247. (ii) C 2 is the normal type C with resolution of the first arsis, and is of such frequent occurrence that it may be looked on as a special type, on hérefélda An. 10, forscrífen hǽfde Beow. 106, in wórold wṓcun Beow. 60; a less common form is that with resolution of the first and second arsis, tō brímes fároðe Beow. 28, swā féla fýrena Beow. 164; sometimes with resolution of the second arsis only, tō sǣ́es fároðe An. 236 and 1660, for frḗan égesan An. 457, but not in Beowulf. The first thesis may have two, three, or four syllables, þā wið góde wúnnon Beow. 113, ofer lágustrǣ́te; with two resolutions, ic þæs wíne Déniga Beow. 350, hū sē mága frémede An. 639, þæt him his wínemā́gas Beow. 65, ne hīe hūru wínedríhten Beow. 863. (iii) C3 is type C with short second arsis, ×–́|⏑́×, and is pretty common, in gḗardágum Beow. 1, of fḗorwégum Beow. 37; the first thesis may have from two to five syllables, þæt wæs gṓd cýning Beow. 11, þæt hīe in bḗorséle Beow. 482, sē þe hine dḗað nímeð Beow. 441, ne meaht þū þæs sī́ðfǽtes An. 211, þonne hē on þæt sínc stárað Beow. 1486. Resolution seems to be avoided, though it occurs here and there, of hlíðes nósan Beow. 1892, on þǣm méðelstéde Beow. 1083. Thesis with secondary accent is not found. The number of hemistichs of type C in Beowulf is, according to Sievers, 564

§ 28. The type D always ends with a disyllabic thesis, of which the first is generally the second syllable of a compound and has the secondary accent. There are four sub-types. (i) D1 is the normal form, –́|–́×̀×, as hélm ǽlwìhta An. 118, fḗond máncỳnnes Beow. 164, wī́gwéorðùnga Beow. 176, wéard Scýldìnga Beow. 95, lándbúèndum Beow. 95, hríng gýldènne, Beow. 2810, hóf mṓdìgra Beow. 312, frḗan ū̀sèrne Beow. 3003. The chief variations arise from resolution of the first arsis, cýning ǽlmìhtig El. 145, fǽder álwàlda Beow. 316, mérelī́ðènde Beow. 255, flótan ḗowèrne Beow. 294, cýning ǣ́nìgne Beow. 1851, or of the second arsis, hḗan hýgegḕomor An. 1089, mǣ́g Hígelà̄ces Beow. 738 and 759; resolution of first and second arsis, hláden hérewǣ̀dum Beow. 1898, néfan Hérerī̀ces Beow. 2207. Hemistichs like wiht unhǣlo Beow. 120, which have compounds with un, may be read wíht únhǽlo according to type D2, or wíht unhǣ́lo according to type A, –́×|–́× (Sievers, Paul-Braune’s Beiträge, x. 251, and Kluge in Paul’s Grundriss, i², p. 1051). (ii) D2 is the same form, but with the thesis short and with secondary accent, –́|–́⏑̀× béorht blǽdgìfa An. 84, lḗof lándfrùma Beow. 31, strḗam ū́t þònan Beow. 2546, rǣ́d éahtèdon Beow. 172; with resolution of the first arsis, mǽgen sámnòde El. 55, mága Héalfdènes Beow. 189; with resolution of the second arsis, hórd ópenìan Beow. 3057, the only example. (iii) D 3 is the normal type, but with short second arsis (rare), –́|⏑́×̀×, éorðcýnìnga El. 1174; with resolution of the first arsis, rádorcýnìnges El. 624. (iv) D 4 has the form –́|–́××̀, and is closely allied to the type E (–́×̀×|–́), as it has the secondary accent on the last syllable of the thesis (Sievers, Paul-Braune’s Beiträge, x. 256), brḗost ínnanwèard An. 649, hólm ū̀p ætbæ̀r Beow. 519, fýrst fórð gewā́t ib. 210; varied by resolution of the first arsis, géaro gū̀ðe fràm An. 234, flóta fā́mighèals Beow. 218, súnu dḗað fornàm Beow. 2120; by resolution of the second arsis, wlánc Wédera lṑod 341, and of both first and second arsis, wlítig wéoruda hḕap An. 872; and resolution of the last thesis with secondary accent, wṓp úp āhàfen Beow. 128, wúnað wíntra fèla Ph. 580. Certain hemistichs which may belong to this sub-type admit of an alternative accentuation, and may belong to the following type; for example, scop hwlum sang Beow. 496 may be read –́|–́××̀, or as E –́×̀×|–́, so werod eall ārās Beow. 652.

§ 29. The type E has two sub-types, distinguished by the position of the syllable bearing the secondary accent; this syllable is generally the second syllable of a compound or the heavy middle syllable of a trisyllabic word with a long root-syllable.

E1 has the form –́×̀×|–́, the syllable with secondary accent standing first in the thesis, mṓdsòrge wǽg El. 61, wéorðmỳndum þā́h Beow. 8, Sū̀ðdèna fólc Beow. 463, ḗhtènde wǽs Beow. 159, hǣ́ðènra hýht Beow. 179, ǣ́nìgne þónc Cri. 1498, wórdhòrd onléac Beow. 259, úplàng āstṓd Beow. 760, scóp hwī̀lum sáng Beow. 496 (cf. above, § 29); varied by resolution of the first arsis, héofonrī̀ces weárd El. 445, Scédelàndum ín Beow. 19, wlítebèorhtne wáng Beow. 93, lífigènde cwṓm Beow. 1974, ǽðelìnges wḗox El. 12, médofùl ætbǽr Beow. 625, dúguð èall ārā́s Beow. 1791; resolution of the second arsis is rare, tī́rḕadge hǽleð An. 2 (the MS. reading ēadige must be corrected to ēadge, see Sievers, Beiträge, x. 459 on these middle vowels after long root-syllable), hélþègnes héte Beow. 142; resolution of both is rare, sélewèard āséted Beow. 668, wínedrỳhten frǽgen An. 921; resolution of the accented thesis, glḗdègesa grím Beow. 2651.

E2 has the last syllable of the thesis with secondary accent, and is very rare, –́××̀|–́, mórðorbèd strḗd Beow. 2437; with resolution of last arsis, gḗomorgìdd wrécen An. 1550, bǣ́ron ū̀t hrǽðe An. 1223.

II. Hemistichs of five members.

§ 30. Hemistichs of five members (extended) occur much more rarely than the normal types of four members. The extended types are denoted by the letters A*, B*, C*, &c.

(ii) A*2 –́××̀|–́× may possibly occur in mā́ððumfæ̀t mǣ́re Beow. 2405, wúldorlḕan wéorca Cri. 1080; with resolution of the thesis with secondary accent, mórðorbèalo mága Beow. 1079. Possibly, however, the syllables um in māððum and or in wuldor and morðor are to be written m and r, so that the scansion of the hemistich would be A2 –́–̀|–́× and –́⏑̀͜×|–́× .[77]

Type B* ×̀×–́|×–́ does not seem to occur in O.E. poetry, though it does in Old Norse.

Type C* in the forms ×̀×–́|–́×,×̀×⏑́×|–́×,××–́|⏑́× are also not found in O.E.

Type D* on the other hand does occur, but almost exclusively in the first hemistich. It has three sub-types: (i) D*1 –́×|–́××, sī́de sǣ́næ̀ssas Beow. 223, áldres órwḕna Beow. 1002; with resolution of the first arsis, ǽðeling ā́nhȳ̀dig Beow. 2668; more commonly with resolution of the second arsis, mǣ́ton mérestræ̀ta Beow. 514; with resolution of both, lócene léoðosỳrcan Beow. 1506. (ii) D*2 –́×|–́⏑̀×, mǣ́re méarcstàpa Beow. 103, éaldor Éastdèna Beow. 392; with resolution of the first arsis, ǽðele órdfrùma Beow. 263; with resolution of the second arsis, mṓdges mérefàran Beow. 502, Bḗowulf máðelòde Beow. 505, &c. (iii) D*3 –́×|⏑́×̀× is not found. (iv) D*4 –́×|–́××̀ is rare, grḗtte Gḗata lḕod Beow. 625, þrȳ́ðlīc þégna hḕap Beow. 400; with resolution of first arsis, éaforan éllorsī̀ð Beow. 2452; with resolution of the second, ȳ́ðde éotena cỳn Beow. 421; with resolution of the secondarily accented syllable, wī́n of wúndorfàtum Beow. 1163; this type is varied by anacrusis, ongínneð gḗomormṑd Beow. 2045, and by anacrusis together with disyllabic thesis in the second foot, oferswám þā síoleða bigòng Beow. 2368.

Type E* does not occur in O.E. poetry.[78]

§ 31. To assign the different hemistichs of a poem to these various types we have to follow as a regulating principle the natural word accent and syntactical accent of each sentence. In some cases the similarity or relation with one another of the types renders it a matter of difficulty to determine exactly to what particular type a hemistich may belong. Systematic investigations as to the principles which govern the combinations of the five types in pairs to form the long line have not yet been made. From such observations as have been made it would appear that by preference hemistichs of different rhythmical structure (ascending and descending) were combined with a view to avoid a monotonous likeness between the two halves of the verse.[79].

§ 32. The combination of two hemistichs so as to form a long line is effected by means of alliteration, one at least of the two fully accented syllables being the bearer of an alliterative sound. In no case is an unaccented syllable or even a syllable with a secondary accent allowed to take part in the alliteration. This fact, that secondarily accented syllables are debarred from alliterating, is another proof that it is better to look on them as belonging to the thesis rather than to the arsis of the verse.

§ 33. Quality of the Alliteration. It is an all but invariable rule that the correspondence of sounds must be exact and not merely approximate. A g must alliterate to a g, not to a c, a d to a d, not to a t, and so on. There is, however, one remarkable exception, namely, that no distinction is made between the guttural c (as in cūðe) and the palatal c (as in cēosan), nor between the guttural g (as in god) and the palatal g (as in gierede), not even when the latter represents Germanic j (as in geong, gēar). With exceptions hereafter to be noted, a consonant followed by a vowel may alliterate with itself followed by another consonant: thus cūðe alliterates not only with words like cyning, but with words like cræft, cwellan; and hūs alliterates not only with heofon but with hlēapan, hnǣgan, &c. The fact that different vowels, as ī, ū, and æ in īsig ond ūtfūs æðelinges fær Beow. 33, alliterate together is only an apparent exception to the strictness of the rule, as it is really the glottal catch or spiritus lenis[80] before all vowels which alliterates here. Wherever a vowel seems to alliterate with an h we are justified in assuming a corruption of the text, as in óretmecgas æfter hǽleðum frægn Beow. 332, where Grein improves both sense and metre by substituting æðelum for hæleðum; other examples are Beow. 499, 1542, 2095, 2930. In some cases where foreign names beginning with h occur we occasionally find instances of this inexact alliteration, as Hólofernus únlyfigendes Jud. 180 and 7, 21, 46, contrasted with Hólofernus hógedon āninga 250; in later works as in Ælfric’s Metrical Homilies we find alliteration of h with a vowel not only in foreign names but with native words, as

and h before consonants (viz. r, l, w) is disregarded as

It is important to observe that the combinations st, sc, sp are not allowed to alliterate with each other or with words beginning with s not followed by a consonant, but st can alliterate only with st, sc only with sc, sp only with sp; thus spere and scyld, stillan and springan, sǣ and styrman do not count as alliterations. The invariable practice is seen in the following lines:—

In later times this rule was not so strictly observed. The metrical Psalms alliterate sc with s and sw with s, as

but sp and st do not alliterate with each other or with s. In Ælfric all these combinations of consonants alliterate indifferently with each other or with s + another consonant or with simple s, as in

Sometimes in Ælfric the alliterating letter does not stand at the beginning of the word,

and the alliteration may even fall on an unaccented particle as in

For a full account of Ælfric’s alliteration the reader may be referred to an interesting essay by Dr. Arthur Brandeis, Die Alliteration in Aelfric’s metrischen Homilien, 1897 (Programm der Staatsrealschule im VII. Bezirk in Wien).

§ 34. Position of the alliterative words. Out of the four accented syllables of the line two at least, and commonly three, must begin with an alliterative sound, and this alliteration still further increases the stress which these syllables have in virtue of their syntactical and rhythmical accent.

The position of these alliterative sounds in the line may vary in the same way as their number. The general laws which govern the position of the alliteration are the following:—(i) One alliterating sound must, and two may occur in the first hemistich; (ii) In the second hemistich the alliterating sound (called the head-stave[82]) must fall on the first of the two accented syllables of that hemistich, and the second accented syllable in the second hemistich does not take part in the alliteration at all; (iii) When there are three alliterating sounds in the whole line two of them must be in the first hemistich and only one in the second. Examples of lines with three alliterating sounds:

Lines with only two alliterative sounds, the first of which may coincide with either of the accented syllables of the first hemistich (the second of course coinciding with the first accented syllable of the second hemistich) are very common:

If the first hemistich contains only one alliterative sound this alliteration generally falls on the more emphatic of the two accented syllables of the hemistich which is usually the first, as

In the type A the single alliteration of the first hemistich not unfrequently falls on the second accented syllable, such cases being distinguished, as A3

In types C and D the single alliteration of the first section must always fall on the first accented syllable which in these types is more emphatic than the second. In types B and E alliteration on the second arsis would bring the alliteration too near to the end of the hemistich, and is therefore rare.

Double alliteration in the first hemistich occurs in all of the five types, and chiefly when the two accented syllables have equally strong accents. It is, therefore, least common in C ×–́|–́× where the first arsis predominates over the second, and is most frequent in the strengthened hemistichs, in D, E, A2, and in the five-membered D* types, where it is the rule.[83]

A third form of alliteration, though much less important and frequent than these two, occurs when the second accented syllable of the second hemistich shares in alliteration, in addition to the first accented syllable. There are then two different pairs of alliterative sounds distributed alternately between the two hemistichs. The commonest form of this double alliteration of the whole line is represented by the formula a b | a b, as

less commonly by the formula a b | b a:

verses corresponding to the formula a a | b b are not found in early poetry. No doubt certain instances of this double alliteration may be accidental, but others seem intentional.

The foregoing rules as to alliteration are strictly observed in the early and classic poetry, but in later times certain licences crept in. Three of these may be noticed. (i) The second accented syllable of the second hemistich is allowed to carry the alliteration instead of the first accented syllable,

(ii) Both accented syllables of the second hemistich alliterate with one accented syllable of the first hemistich,[84]

(iii) The four accented syllables of the line all alliterate together,

In the majority of cases the same alliterative letter is not employed in two successive lines, but we find cases like

and earlier in Andreas 70, 197, 372, 796, 815, 1087, &c., or in Beowulf 403, 489, 644, 799, 865, 898, &c.

And even three lines in succession, as

This usage, which in Middle English became very popular, is noticeably frequent in the poem of Judith, probably with a view to emphasis. Many examples of such pairs of verses are to be found collected by Dr. A. Brandeis from Ælfric.

The unaccented words may begin with the same letter as the accented words which bear the alliteration proper,[85] as

or one of the unaccented words may begin with the same letter as an accented word which does not alliterate, as

this of course has nothing to do with alliteration, though in later times it was often mistaken for it.

Verses without any alliteration at all, as

occur only in late OE. poetry like Ælfric’s Homilies, and when rhyme was beginning to creep in.

§ 35. Alliteration in relation to the parts of speech and to the order the order of words. Both alliteration and the whole structure of the alliterative line depend in the first place on the natural or etymological accent of the single words, and next on the syntactical accent which these words bear in their relation to one another in the sentence. Just as only the accented syllable of a single word can take part in the alliteration, so only can those words take part in it which are marked out in the sentence as important and therefore strongly accented.

The relative degree of stress is influenced at times by the rhetorical accent, but generally speaking we find a certain gradation of accent among the accented words depending on their intrinsic and not on their rhetorical importance in building up the sentence.

Two general principles may be laid down: (1) If the syntactical value of the two accented syllables of the hemistich is not equal, then the word which has the stronger accent of the two is chosen to alliterate. In the second hemistich it is always the first accented word (the ‘head stave’), in the first hemistich it is generally the first accented word, though the second accented word may alliterate as well. (2) If the two accented syllables of the section are equal in syntactical value, then the first alliterates, and when double alliteration is allowed the second may also alliterate.

The various grammatical classes of words are treated in regard to the alliteration in the following way:—

Nouns, including adjectives and the infinitives and participles of verbs, have the strongest accent of all words in the sentence. A noun therefore takes precedence over the other parts of speech among which it occurs and has the alliteration, as

If two nouns occur in the same hemistich it is always the first which alliterates,

The only exceptions are when a special rhetorical emphasis is given to the second word.

When a noun and two adjectives or two nouns and an adjective occur in the same hemistich, one of these is always subordinated to the other, and the two together are treated as a combination. In such cases, where there is double alliteration in the hemistich, the position of the alliterating words may be either a a x, or a x a, the subordinate element (x) standing either in the last or the second place in the hemistich,

In the case of single alliteration, it is always the first of the nouns or adjectives which alliterates.

The verb (excluding the infinitive and participles) is usually less strongly accented than the noun. It may therefore precede or follow the noun or adjective without alliteration, either in the arsis or thesis, as

On the other hand, when a hemistich consists only of one noun and one verb, the verb may alliterate, as

When a substantive and an adjective are closely combined, a verb in the same hemistich may alliterate, as

In formulas consisting of noun + verb the noun predominates over the verb and takes the alliteration, as

But if the verb is emphatic it may alliterate though there is a noun in the same hemistich; this occurs chiefly in the second hemistich, as

but a few instances are found in the first hemistich, as

When one of two verbs in the hemistich is subordinate to the other the verb in the subordinate clause alliterates, having a stronger accent than the verb in the main clause,

If the two verbs are co-ordinate the first alliterates,

in the first hemistich both verbs commonly alliterate,

The adverb. Adverbs of degree like micle, swīðe, ful, &c., are commonly found in the thesis, and even if they stand in the arsis they usually do not alliterate, as

When adverbs of this kind have a special rhetorical emphasis they may of course alliterate, as

Adverbs which modify the meaning of the word which they precede alliterate, as

Adverbial prepositions preceding the verb also alliterate,

but not when they follow the verb,

Adverbs derived from nouns are more strongly accented than the verb which they modify and therefore alliterate,

Pronouns (and pronominal adjectives like monig, eall, fela) are usually enclitic, and precede or follow the noun without alliterating, as

With a special rhetorical accent they may alliterate even if they precede the noun,

The pronoun selfand the pronouns compounded with the prefix ǣ (ǣghwā, ǣghwylc, &c.) are usually accented, and alliterate if they form the first arsis of the hemistich, as

Prepositions, conjunctions, and particles are not as a rule accented, but prepositions if followed by an enclitic pronoun take the accent and alliterate, as

Whether words of these classes, standing in the first arsis of the first hemistich along with another alliterating word, were intended also to alliterate is somewhat uncertain, but it is probable that they were so, as in

These laws of accentuation are strictly observed only in the older poetry; by the end of the tenth century, in Byrhtnoth, the Metres of Boethius and the Psalms, they are frequently neglected.

§ 36. Arrangement and relationship of verse and sentence. The following rules hold good in general for the distribution of the sentence or parts of the sentence between the hemistichs of the verse. Two distinct pauses occur in every alliterative line, one (commonly called the caesura) between the first and second hemistichs, the other at the end of the line, and these pauses are determined by the syntactical construction; that is to say, they coincide with the end of a clause or lesser member of the sentence. The hemistich must contain such parts of the sentence as belong closely together; and such coherent parts, as, for example, a pronoun and noun to which it refers or adverb with adjective, must not be separated from one another by the caesura, unless the pronoun or adverb is placed in the second arsis of the hemistich, as

In Beowulf this separation of closely connected words is permitted only if the word standing in the arsis alliterates at the same time. Longer parts of a sentence may be separated both by the caesura and the pause at the end of the line. The syntactical connexion between the parts of a sentence thus broken up makes the unity of the parts clear, and when the division occurs in the caesura between the two halves of the verse, the alliteration common to both hemistichs serves further to emphasize this unity.

The single alliterative lines are connected with one another by the prevailing usage of ending the sentence not at the end of the completed line, but at the end of the first hemistich or in the middle of the line, and of beginning a new sentence with the second hemistich. The great variety of expression, and the predilection for paraphrase by means of synonyms which is so characteristic of OE. poetry, contribute to make such breaks in the line easy. Whatever may be the explanation, it is certainly the fact that in the OE. poetry the metrical and syntactical members do sometimes coincide, but at other times overlap in a way which does not admit of being reduced to rule.[86]

The Lengthened Verse

§ 37. Besides the normal four-beat line (with two beats to each hemistich) there is in OE. and Old Saxon another variety, the lengthened line (Schwellvers) with three beats in each hemistich.[87] These verses occur in almost all OE. poems, either isolated or more commonly in groups, and occasionally we find lines with one hemistich of two beats and the second hemistich of three, like.

and Jud.96, Crist1461, &c., or with a lengthened hemistich of three beats and a normal hemistich of two beats, like

and Sat.605, Gnom. Ex.200, &c.

In the Psalms and in Cynewulf’s Juliana they are wanting entirely, in Cynewulf’s Elene out of 1321 verses there are only fourteen lengthened whole lines, and three lengthened hemistichs. Examples of groups of these lengthened verses will be found in Gen. 44–46, 1015–1019, 2167–2169, 2854–2858; Exodus 569–573, Dan. 59–106, 203–205, 226–228, 238–246, 262–271, 435–438, 441, 448, 452–458; Judith 2–12, 16–21, 30–34, 54–61, 63–68, 88–99, 272–274, 289–291, 338–349; Satan 202, 232, 237, 605, Crist 621, 889, 922, 1050, 1382–1386, &c., and in many of the smaller poems.[88]

Lengthened verses of a looser type occur in Salomon and Saturn, and Genesis B; they have unusually long theses of four or five unaccented syllables after the first accented syllable, as

or have equally long anacruses before the first accented syllable, as