Yet these it is evident might also be read with three accents, as in the following couplet also:
"The se bigan to flowe,
And Horn child to rowe."
According as one reads the Anglo-Saxon verse with two or four accents to the half-line, he will regard the typical half-line of King Horn as made up of two or four accents. If the fundamental number was two, the additional accented syllables were doubtless introduced under the influence of the French eight-syllable couplet. It is not difficult to see how the more regular (Latin or French) and the less regular (native) measures might have been confused, and soon have coalesced in popular use. Ten Brink, reading the King Horn lines with four accents, speaks of them as "formed entirely on the Teutonic principle, with two accents upon the sonorous close of the verse, so that it appears to be an organic continuation of the chief form in Layamon and in Ælfred's Proverbs. This circumstance makes the poem exceptional among the early English romances." He also speaks of "an unmistakably strophic construction in the text as we have it." (English Literature, Kennedy translation, vol. i. p. 227.)
Anon out of þe north est þe noys bigynes:
When boþe breþes con blowe upon blo watteres,
Roȝ rakkes per ros with rudnyng anunder,
Þe see souȝed ful sore, gret selly to here,
Þe wyndes on þe wonne water so wrastel togeder,
Þat þe wawes ful wode waltered so hiȝe
And efte busched to þe abyme, þat breed fysches,
Durst nowhere for roȝ arest at þe bothem.
When þe breth and þe brok and þe bote metten,
Hit watz a ioyles gyn, þat Ionas watz inne;
For hit reled on roun upon þe roȝe yþes.
(Patience, ll. 137-147. ab. 1375.)
Til þe knyȝt com hym-self, kachande his blonk,
Syȝ hym byde at þe bay, his burneȝ bysyde,
He lyȝtes luflych adoun, leveȝ his corsour,
Braydeȝ out a bryȝt bront, & bigly forth strydeȝ,
Foundeȝ fast þurȝ the forþ, þer þe felle byde,
Þe wylde watȝ war of þe wyȝe with weppen in honde.
Hef hyȝly þe here, so hetterly he fnast,
Þat fele ferde for þe frekeȝ, lest felle hym þe worre
Þe swyn setteȝ hym out on þe segge even,
Þat þe burne & þe bor were boþe upon hepeȝ,
In þe wyȝt-est of þe water, þe worre had þat oþer;
For þe mon merkkeȝ hym wel, as þay mette fyrst,
Set sadly þe scharp in þe slot even,
Hit hym up to þe hult, þat þe hert schyndered,
& he ȝarrande hym ȝelde, & ȝedoun þe water, ful tyt;
A hundreth houndeȝ hym hent,
Þat bremely con hym bite,
Burneȝ him broȝt to bent,
& doggeȝ to dethe endite.
(Sir Gawayne and the Green Knight, strophe xviii. ab. 1375.)
These two specimens, both doubtless the work of the same author (to whom are also attributed the Pearl and Cleanness), represent the patriotic revival of the alliterative long line by contemporaries of Chaucer in the latter half of the fourteenth century. In Sir Gawayne the rimeless long line is gathered into strophes, each of which concludes with four riming lines. (See above, p. [109].)
For analysis of this revived alliterative verse, see a valuable article by Dr. Luick, Die Englische Stabreimzeile im 14n, 15n, und 16n Jahrhundert, in Anglia, vol. xi. pp. 392 and 553. Luick analyzes not only the poems of the "Pearl poet," but also the Troy Book, the Alexander Fragments, William of Palerne, Joseph of Arimathea, Morte Arthure, and minor poems. He finds the Troy Book the most regular alliterative poem of Middle English, but in all of the group a general tendency to preserve not only the early laws of alliteration but also the "five types" of Sievers's Anglo-Saxon metrics. Dr. Luick attributes the final decline of the old measure in large degree to the falling away of the final syllables in -e, etc.; without these numerous feminine endings the primitive "long line" was impossible. This he regards as regrettable, since the old alliterative verse, "growing up on native soil with the language itself," represented the natural accent-relations of the Germanic languages, especially the recognition of two principal degrees of accent; whereas the modern rimed verse requires the reduction of this to a uniform "tick-tack" of alternating stress and non-stress.