§ 46. Fuhr.

Fuhr established a new theory in his Die Metrik des westgermanischen Alliterationsverses. Sein Verhältnis zu Otfrid, den Nibelungen, der Gudrun usw. Marburg 1892. Fuhr proceeds from the thought that in OE. alliterative poetry, as in the German Nibelungenlied, we must assume four beats in verses with feminine ending and three in those with masculine ending. We must not, however, assume the same limitation as exists in the Nibelungenlied, where the verses with four beats belong to the first half-line, whilst the second half-line contains verses with three beats, but that verses with three beats or with four beats can be used in any combination.

§ 47. Criticism of Fuhr's Theory.

This theory seems at first inviting. On closer view, however, we find that it can be applied only if we measure the same words at one time with three beats at another with four beats, and for no other reason except for the sake of the theory. According to Fuhr verses of four syllables such as lange hwile, feond mancynnes, mid Scyldingum contain four beats, because the endings are feminine; yet much longer verses ond minra eorla gedriht, atol da geswing, worold-āre forgeaf contain only three beats because the endings are masculine.

Moreover Fuhr has included amongst the masculine verses (those with three beats) all 'disyllabic masculine' verses, i.e. verses which end with two short syllables, although these can belong to quite different types, as Sievers had rightly shown. In types B and E the two short syllables are resolved stresses of a long syllable and make, therefore, only one member of the verse; here, therefore, there is actually 'disyllabic masculine' ending, e.g. pār was madma fela, hel-pegnes hete. In the types C and D1 according to Sievers a short syllable takes the place of a long one; the verse ending ux stands for original x and represents two members of the verse: in gear-dagum, leof landfruma. Fuhr has here, therefore, thrown together verses quite differently constructed.

Fuhr's theory, therefore, is not in keeping with the verses we have, but his book contains many valuable observations. Fuhr stands almost on the side of the four-beat theory, since he scans A, C and D1 verses, which most frequently occur, with four beats, (the 'shortened' types excepted).

§ 48. Ten Brink.

After ten Brink's death his unfinished account of OE. literature was published in Paul's Grundriss II, 2, 510-550. Here something was said on the rhythmical structure of OE. alliterative verse (p. 515 ff.). Ten Brink's use of hebungen often

agrees with Fuhr's, although ten Brink started from a different point of view.

Ten Brink regards an Indo-Germanic original verse of the form ××××××× as the foundation of the alliterative verse. He distinguishes two stronger and two weaker hebungen, which can be used in any order. According to their position ten Brink arrives at five types, which only partially agree with those of Sievers.

a) ××××××××: gewất him þà to wárođe, or without anacrusis and theses: lángè hwilè

B) ××××××××: òfer lágu-strátè, oft Scýld Scéfing 7) ××××××××: ædelìngà gedríht

ε) ××××××××: a) third hebung the weakest: átol ýđà geswing,

b) fourth hebung the weakest: onbánd béadurùnè, lánd-búèndùm.

In addition to these complete verses, which contain four hebungen, there are, however, according to ten Brink also incomplete verses, in which a weaker hebung is suppressed. These contain only three hebungen. Thus the last weak hebung has disappeared in:

a2) ×××××××: a) with independent senkungsyllable at end: mágo-drìht mícel, hél-pègnes héte,

b) with hebung-syllable at end: wlite-bedrhtne wáng, múrnènde mód, wórd-hòrd onléac, ã-lédòn þá. P2) ×××××××: a) with senkung-syllable at end:

pat was god cýning, hwat we Gár-Déna, þùrh rúmne séfan,

b) with hebung-syllable: him on béarme læg, hìm pa Scýld gewắt.

ɛ2) ×××××××: a) with senkung-syllable at end: suutol sáng scopes, léof lánd-frùma,

b) with hebung-syllable: fæder éllor hwèarf, flóta fámig-hèals.

The second weak hebung has fallen out, or the first and second hebung are contained in the same syllable in:

a3) ×××××: land gesáwòn, wil-gesiđàs.

The first weak hebung has fallen out in: p3) ××××××: gehróden góldè.

§ 49. Criticism of ten Brink's Theory.

Ten Brink's theory shows a great advance on Fuhr's in that he reads with four beats some verses, which Fuhr read with three beats, viz. B, D and E verses with masculine endings; but very much the same objections may be raised against ten Brink as against Fuhr. Ten Brink can give no reason why verses of four syllables such as lange hwile, mid Scyldingum should be 'complete' and, therefore, have four beats, and yet verses with five syllables such as fæder ellor hwearf, flota stille bād etc. should be 'incomplete' and have, therefore, only three beats. Further he has put together in his type a2, A and E verses, in type ẞ2, B and C verses, whilst his types 7 and 8 contain verse forms only

rarely found. Ten Brink's theory, therefore, gives us no clear view of the true structure of the allit

erative verse.

NOTE. Frank Heath, a pupil of ten Brink, has explained ten Brink's theory of alliterative verse in a lecture held before the Philological Society in London on 2. Jan. 1893. At the same time he treated of the lengthened lines; cp. Transactions of the Philological Society, 1891-93, pp. 375-395.

§ 50. A new Attempt with the strict Four-Beat Theory.

What is common to the theories of Möller, Hirt, Fuhr and ten Brink is that these scholars do not accept the two-beat theory, but assume a mixture of four, three and two beats. Sievers, too, does not adopt the strict two-beat theory, but for one part of his verses (A2a, A2b, Aok, D1, D2, E) assumes one extra nebenhebung, and for another part (A2ab) even two nebenhebungen. Thus in his system, too, there is a variety of verses with two, three and four beats. But if, as is clear from what has been said, neither the strict two-beat theory nor a theory, which assumes a mixture of two, three and four beats, gives a satisfactory view of the rhythmical structure of alliterative verse, free from all objections, it is necessary once more to see if the old four-beat theory of Lachmann, in spite of the objections raised against it, cannot be established for OE. verse. The chief difficulty, as we see from the discussions of Fuhr and ten Brink, is